Barracuda User Manual

Barracuda Virtual Reactor User Manual Release 17.0.5

CPFD Software, LLC

August 31, 2016

CONTENTS

1

2

Introduc Intro ductio tion n 1.1 Getti Getting ng star started ted with Barr Barracuda acuda . . . . . . . . . 1.2 Navi Navigati gating ng the Barr Barracuda acuda GUI GUI . . . . . . . . . 1.3 Mod Modeli eling ng wit with h Bar Barrac racuda uda . . . . . . . . . . . 1.3.1 1.3 .1 Creati Cre ating ng a pro projec jectt file . . . . . . . . . 1.3.2 Opening Openi ng an exis existing ting proj project ect file . . . . 1.3. 1. 3.3 3 Sett Se ttin ing g up a mo mode dell . . . . . . . . . . 1.3.4 1.3 .4 Runnin Run ning g a Bar Barrac racuda uda mod model el . . . . . 1.3.5 1.3 .5 Analyz Ana lyzing ing Mod Model el Res Result ultss . . . . . . . 1.4 Refer Reference ence of Barr Barracuda acuda VR GUI Comp Component onentss 1.4.1 1.4 .1 Navig Na vigati ation on tre treee . . . . . . . . . . . . 1.4. 1. 4.2 2 Menu Me nu ba barr . . . . . . . . . . . . . . . 1.4.3 1.4 .3 Shortc Sho rtcut ut bu butto ttons ns . . . . . . . . . . . . 1.5 Com Common monly ly Use Used d GUI Pa Patte tterns rns . . . . . . . . . 1.5. 1. 5.1 1 Comm Co mmen entt Fi Fiel eld d . . . . . . . . . . . . 1.5.2 1.5 .2 Add,, Edi Add Edit, t, Cop Copy y, Del Delete ete . . . . . . . . 1.5.3 1.5 .3 Tab abula ularr Inp Input ut Usi Using ng SFF Fil Files es . . . . 1.5.4 1.5 .4 Applie App lied d Mat Materi erials als . . . . . . . . . . . 1.5.5 1.5 .5 Flux Flu x Pla Plane ne Opt Option ionss . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

1 2 2 3 3 4 4 4 5 5 5 6 9 10 10 10 11 13 15

Setu Se tup p Gr Grid id 2.1 The Set Setup up Gri Grid d Wi Windo ndow w . . . . . . . . . 2.1.1 2.1 .1 Grid Gri d uit uitil ility ity bu butto ttons ns . . . . . . . 2.1. 2. 1.2 2 Grid Gr id vi view ew pa pane ness . . . . . . . . . 2.2 2. 2 Ge Geom omet etry ry Tab . . . . . . . . . . . . . . 2.3 2. 3 Ba Bafffle fless Tab . . . . . . . . . . . . . . . . 2.3.1 2.3 .1 Baffles Baf fles Ma Main in Tabl ablee . . . . . . . 2.3. 2. 3.2 2 Add Ad d Ba Bafffle . . . . . . . . . . . . 2.3. 2. 3.3 3 Remo Re move ve Ba Baffl fflee . . . . . . . . . 2.3. 2. 3.4 4 Bafffle Ma Ba Make kerr . . . . . . . . . . 2.3. 2. 3.5 5 Imp Im por ortt . . . . . . . . . . . . . . 2.3. 2. 3.6 6 Exp Ex por ortt . . . . . . . . . . . . . . 2.3.7 2.3 .7 Vie iew w Baf Baffle fle But Button ton . . . . . . . 2.4 Gri Grid d Con Contro trols ls Tab . . . . . . . . . . . . 2.4. 2. 4.1 1 Opti Op tion onss . . . . . . . . . . . . . 2.4. 2. 4.2 2 Grid Gr id li line ness ta tab b . . . . . . . . . . 2.5 Cha Changi nging ng the Gri Grid d Vie iew w . . . . . . . . . 2.6 Add Adding ing and Mod Modify ifying ing Gri Grid d Line Liness . . . . 2.6.1 Changing Chang ing grid line liness grap graphica hically lly . 2.6.2 2.6 .2 Using Usi ng the gri grid d li lines nes edi editor tor . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

17 19 19 19 21 22 23 24 24 24 26 26 26 26 26 30 31 31 32 33

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . .

i

2.7 2.7 2.8 2.9 2.10 2.11

Checki Chec king ng th thee Gr Grid id . . Genera Gen eratin ting g the Gri Grid d . Vie iewin wing g Gri Grid d Out Output put . Reference Refer ence Grid . . . . ScaleSTL Scale STL util utility ity . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

34 34 36 36 38

3

Global Set Global Setti tings ngs 41 3.1 The Glob Global al Sett Settings ings Wi Windo ndow w . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 3.2 Hea Heatt Tr Trans ansfer fer Mo Model delss . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43

4

Basee Mat Bas Materi erials als 47 4.1 The Bas Basee Mat Materi erials als Wi Windo ndow w . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 4.2 Spe Specif cifyin ying g Bas Basee Mat Materi erials als . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

5

Part Pa rtic icle less 5.1 Pa Parti rticle cle Int Intera eracti ctions ons . . . . . . . . . . . . . . 5.2 5. 2 Dr Drag ag Mo Mode dels ls . . . . . . . . . . . . . . . . . . 5.2.1 5.2 .1 Creati Cre ating ng a dra drag g mod model el . . . . . . . . 5.2.2 5.2 .2 Checki Che cking ng a dra drag g mo model del . . . . . . . . 5.2.3 5.2 .3 Drag Dra g mod model el ref refere erence nce . . . . . . . . . 5.3 5. 3 Vol olat atil iles es . . . . . . . . . . . . . . . . . . . . 5.3.1 Specifyin Speci fying g vol volatil atilee mate material rialss . . . . . 5.3.2 Defining Defini ng vola volatile tiless comp composit osition ion . . . . 5.4 Pa Parti rticle cle Spe Specie ciess . . . . . . . . . . . . . . . . . 5.4.1 Specifyin Speci fying g part particle icle speci species es . . . . . . 5.4.2 Defining Defini ng parti particle cle comp composit osition ion . . . . 5.4.3 Specifyin Speci fying g a parti particle cle size dist distrib ributio ution n 5.4.4 5.4 .4 Drag Dra g Mod Model el Mul Multi tipli pliers ers . . . . . . . . 5.4.5 5.4 .5 Agglom Agg lomera eratio tion n mod model el . . . . . . . . .

6

7

ii

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Initia Ini tiall Con Condit dition ionss 6.1 Fl 6.1 Flui uid d IC ICss . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1.1 6.1 .1 Specif Spe cifyin ying g a flui fluid d IC . . . . . . . . . . . . . . . . 6.1.2 Defining Defini ng init initial ial fluid comp composit osition ion . . . . . . . . . . 6.2 6. 2 Pa Part rtic icle le IC ICss . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2.1 6.2 .1 Specif Spe cifyin ying g a par partic ticle le IC . . . . . . . . . . . . . . . 6.2.2 Adjusting Adjus ting the dist distrib ribution ution of comp computat utational ional part particle icless

. . . . .

. . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Boundary Bounda ry Con Condit dition ionss 7.1 7. 1 Pr Pres essu sure re BC BCss . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.1.1 7.1 .1 Specif Spe cifyin ying g a pre pressu ssure re BC . . . . . . . . . . . . . . . . . . . . . . . . 7.1.2 7.1 .2 Definin Defi ning g a tra transi nsient ent pre pressu ssure re BC . . . . . . . . . . . . . . . . . . . . 7.2 7. 2 Fl Flo ow BC BCss . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.2. 7. 2.1 1 Spec Sp ecif ifyi ying ng a flo flow w BC . . . . . . . . . . . . . . . . . . . . . . . . . . 7.2.2 7.2 .2 Definin Defi ning g a tra transi nsient ent flo flow w BC . . . . . . . . . . . . . . . . . . . . . . 7.3 Part Particle icle Feed Feedss at at Pressu Pressure re and Flo Flow w BCs BCs . . . . . . . . . . . . . . . . . . . . 7.3.1 7.3 .1 Specif Spe cifyin ying g a par partic ticle le fee feed d . . . . . . . . . . . . . . . . . . . . . . . . 7.3.2 Setting Sett ing up a part particle icle feed contr controlle ollerr . . . . . . . . . . . . . . . . . . . 7.3.3 Using an adva advanced nced part particle icle feed contr controlle ollerr . . . . . . . . . . . . . . . 7.4 7. 4 In Inje ject ctio ion n BC BCss . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.4.1 7.4 .1 Specif Spe cifyin ying g an inj inject ection ion BC . . . . . . . . . . . . . . . . . . . . . . . 7.4.2 Creating Creat ing tran transient sient injec injection tion files . . . . . . . . . . . . . . . . . . . . 7.4.3 Import/E Impo rt/Export xport Inje Injector ctor Locat Location ion Data . . . . . . . . . . . . . . . . . 7.5 The Therma rmall Wall BCs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.5.1 7.5 .1 Specif Spe cifyin ying g a the therma rmall wal walll . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . .

53 54 58 60 63 64 69 69 71 71 73 74 76 78 89

. . . . . . . . . . . . . . . . . .

93 93 94 95 96 97 98

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

101 . 101 . 102 . 104 . 104 . 106 . 109 . 109 . 111 . 113 . 115 . 118 . 120 . 126 . 127 . 127 . 129

7.6

7.5.2 Defining a transient thermal wall . . . . . . Passive Scalars . . . . . . . . . . . . . . . . . . . . 7.6.1 Specifying a passive scalar . . . . . . . . . 7.6.2 Creating a transient passive scalar input file 7.6.3 Analyzing passive scalar data . . . . . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . . . . . . . . . . . .

. . . . .

. . . . .

. . . . .

. . . . . . . . . . . . . . .

131 131 133 134 134

8

BC Connections 137 8.1 Secondary Feeds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137 8.2 BC Connectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139 8.2.1 BC Connection Input from Domain (outlet BCs) . . . . . . . . . . . . . . . . . . . . . . . . 140 8.2.2 BC Connector Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141 8.2.3 BC Connection Output to Domain (inlet BCs) . . . . . . . . . . . . . . . . . . . . . . . . . 142

9

Chemistry 145 9.1 The Chemistry Window . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147 9.2 Rate Coefficients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149 9.2.1 Defining chemistry rate coefficients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150 9.2.2 Adding solids dependence to a reaction rate . . . . . . . . . . . . . . . . . . . . . . . . . . 153 9.3 Reactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156 9.3.1 Volume Average Stoichiometric Reaction Dialog . . . . . . . . . . . . . . . . . . . . . . . 157 9.3.2 Volume Average Species Reaction Dialog . . . . . . . . . . . . . . . . . . . . . . . . . . . 159 9.3.3 Discrete Reaction Dialog . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160 9.4 Conversion between volume average and discrete chemistry . . . . . . . . . . . . . . . . . . . . . . 163 9.5 Importing chemistry from another project . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165

10 Numerics 167 10.1 Solver Advanced Settings Dialog . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168 10.2 Turbulence and Advection Options Dialog . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169 11 Time Controls

171

12 Data Output 12.1 Flux Planes . . . . . . . . . . . . . . . . . 12.1.1 Defining a flux plane . . . . . . . 12.1.2 Flux plane data format . . . . . . 12.1.3 Gas species flux data format . . . 12.1.4 Flux plane particle size data format 12.2 GMV Output Options . . . . . . . . . . . 12.3 Average Data . . . . . . . . . . . . . . . . 12.4 2D Plot Data . . . . . . . . . . . . . . . . 12.4.1 2D data filenames . . . . . . . . . 12.4.2 2D data example . . . . . . . . . 12.5 Transient Data . . . . . . . . . . . . . . . 12.6 Wall Erosion . . . . . . . . . . . . . . . . 12.7 Particle Attrition . . . . . . . . . . . . . . 12.8 Raw Data . . . . . . . . . . . . . . . . . . 12.8.1 Particle Output Data . . . . . . . 12.8.2 Cell Output Data . . . . . . . . . 12.9 Solver Output Units . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

175 175 176 178 178 180 181 184 185 187 189 190 192 193 194 194 196 198

13 Run 13.1 Checking the Model Setup . . . . . 13.1.1 Running a single time step 13.1.2 Verifying the model setup . 13.2 Starting a Simulation . . . . . . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

203 203 204 204 204

. . . .

. . . .

. . . .

. . . .

iii

13.3 Restarting a Simulation . . . . . . . . 13.4 Interacting with a Running Simulation . 13.5 Running from the Command Line . . . 13.5.1 Solver flags . . . . . . . . . . 13.5.2 Command line restart . . . . . 13.6 Data file reference . . . . . . . . . . . 13.6.1 History file . . . . . . . . . . 13.6.2 Info file . . . . . . . . . . . . 13.6.3 Flux files . . . . . . . . . . . 13.6.4 Transient data files . . . . . . 13.6.5 Population files . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

14 Post-Run 14.1 Data Visualization Tools . . . . . . . . . . . . 14.1.1 GMV . . . . . . . . . . . . . . . . . 14.1.2 XMGR/QtGrace . . . . . . . . . . . 14.1.3 View log files . . . . . . . . . . . . . 14.1.4 Terminal window . . . . . . . . . . . 14.2 Viewing GMV results . . . . . . . . . . . . . 14.3 Post-run utilities . . . . . . . . . . . . . . . . 14.3.1 Viewing images and animations . . . 14.3.2 Converting images and animations . . 14.3.3 Convert Gmv to Ensight Gold Format 14.3.4 Creating a Support File . . . . . . . . 14.3.5 Colormap editor . . . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . .

207 210 213 213 214 214 215 215 215 215 215

. . . . . . . . . . . .

217 219 219 219 222 222 222 223 223 225 225 226 226

15 Plot Manager 16 Useful Formulas 16.1 Estimating Particle Volume Fraction at a BC . . . . 16.2 Converting from Mass Fraction to Volume Fraction . 16.3 Converting from Mass Fraction to Mole Fraction . . 16.4 Converting from Mole Fraction to Mass Fraction . .

231

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

. . . .

233 233 233 234 234

17 References

235

Bibliography

237

Index

241

iv

CHAPTER

ONE

INTRODUCTION

Barracuda VR is a powerful computational fluid dynamics (CFD) software package for the design, analysis, and optimization of industrial-scale fluidized bed reactors and other particle-fluid systems. The creation of models in Barracuda VR is facilitated by the graphical user interface (GUI) which enables the rapid development of particlefluid dynamics models that fully utilize the capabilities of the software. These capabilities include: Efficient modeling of particle-fluid hydrodynamics Barracuda VR uses the multiphase particle-in-cell (MP-PIC) approach for calculating particle-fluid dynamics in which the massive number of real particles (1e16 or more) that exist in real industrial-scale systems are represented by a smaller number of computational particles for which the mass, momentum, and energy transport equations are solved. Complex industrial-sized geometries While the MP-PIC approach allows the efficient modeling of large numbers of real particles, the built-in grid generator cuts a full three-dimensional geometry from a CAD file that contains all necessary features of chemical processing equipment such as inlet and outlet piping, gas distributors, cyclones, and heating tubes. See Setup Grid for more details on the Barracuda VR grid generator. Calculation of heat transfer within geometry The Barracuda VR solver considers heat transfer at the particle, fluid, and walls while utilizing heat capacity functions of all fluid and solid materials within the model for accurate calculation of temperatures. Furthermore, the endothermicity or exothermicity of any included chemical reactions is also considered. Heat transfer is discussed further in Global Settings. Gas phase chemical mixing The fluid phase can consist of multiple gas components when gas-solid systems are modeled, which allows the mixing and reactions within the gas phase to be studied. The Barracuda VR solver considers all material properties for calculation of mass, momentum, and energy equations: molecular weight, viscosity, temperature-dependent heat capacity expressions, heat of formation, and thermal conductivity. Material properties and the use of Barracuda VR‘s built-in library of gas and solid materials are discussed in Base Materials. Multicomponent particles Particles in Barracuda VR can consist of multiple solid components, which enables the study of particle reactions of industrial interest, such as: • the changing coke content on FCC catalyst particles in a regenerator; • gasification and combustion of coal, biomass, or other organic materials; • oxidation/reduction reaction of solid materials for oxygen capture; • and CO2, SOx, and NOx adsorption for flue gas cleanup. Definition of particle species in Barracuda VR is discussed further in Particles. Chemical reactions in gas phase and solid phase The chemistry module in Barracuda VR provides multiple approaches for including reaction chemistry, phase changes, or adsorption processing in a model. These processes are closely coupled with the hydrodynamics and heat transfer in many reactor systems and are an important component of a complete fluidized bed model. Reaction chemistry in Barracuda VR is discussed in Chemistry.

1

Barracuda Virtual Reactor User Manual, Release 17.0.5

1.1 Getting started with Barracuda For typical Barracuda VR installations, the Barracuda VR GUI is started by double-clicking on the Barracuda VR icon on the desktop or by typing barracuda.17 from a terminal window in Linux systems. When initially opened, the GUI will appear in a state of reduced functionality until a project file is created or opened. Once a project file has been opened, the Barracuda VR GUI will be fully functional: the Navigation tree is populated, the Menu bar commands are enabled, and project comments can be added. The fully functional GUI is shown in Fig. 1.1.

Fig. 1.1: Barracuda GUI with new project file created

Creating or opening a project file A new project file is created by clicking on the New Project button under the File menu which opens the New Project Dialog that prompts the user for a project file name and a project directory. Alternatively, an existing project file can be opened by clicking on the Open Project button under the File menu which allows the project file to be selected through a file open dialog. Additional information on the creation and opening of project files is discussed in Creating a project file.

1.2 Navigating the Barracuda GUI The GUI contains five major functional areas that facilitate all stages of a successful simulation, from model creation to simulation execution and data analysis. These functional areas are: Navigation tree The navigation tree, located on the far left of the GUI, provides the main interface for accessing the different input dialogs for creating a model in Barracuda VR. The navigation tree is organized so that a user can access

2

Chapter 1. Introduction


each section of the tree stepwise from the top of the tree to the bottom, providing input information where necessary. In doing so, the user will specify a complete model in an efficient and methodical manner. For more information, see Navigation tree. Menu bar The menu bar at the top of the GUI provides functionality for file management, grid creation, simulation execution, and data analysis. The features contained within the File, View, Setup, Run, Graphics and Output , Post processing, and Help menus are discussed further in Menu bar . Shortcut buttons The shortcut buttons, which appear below the menu bar in the GUI, provide direct links to commonly used functionality within the menu bar: file open, file save, grid generation, and simulation execution for example. For details, see Shortcut buttons. Main window The main window is the area to the right of the navigation tree which shows the Barracuda VR artwork and the project comment box in Fig. 1.1. During the model setup, the main window area is dynamically changed to Setup Grid is selected provide the functionality of each section shown in the navigation tree. For example, when from the navigation tree, the main window will display the grid controls and viewing panes of the setup grid screen (shown in Fig. 2.2). Directory bar The directory bar, located at the bottom of the GUI, shows the current project file name on the left and the project directory on the right. In Fig. 1.1, the sample project file shown is FBR_model.prj and the path to the directory is C:/Users/lobo/FBR_Model_Directory.

1.3 Modeling with Barracuda Barracuda VR is a powerful engineering tool for simulating, optimizing, and designing particle-fluid systems and a successful Barracuda VR model can be used many times with slight modifications to test different operating conditions, feed materials, or geometry designs. Therefore, it is important when setting up a model to know what output data is necessary, how the data will be analyzed, and what modifications will possibly be made in the future. An overview of data output options is provided in Analyzing Model Results.

1.3.1 Creating a project file A new project file is created by clicking on the New Project button under the File menu which opens the New Project Dialog, shown in Fig. 1.2. The project name should be typed in the Project name field and the project directory can be located by clicking the Browse button which will open a file dialog. It is important that a separate directory is used for each Barracuda VR simulation.

Fig. 1.2: New File Dialog

1.3. Modeling with Barracuda

3


1.3.2 Opening an existing project file An existing project file can be opened by clicking on the Open Project button under the File menu. This will open a file dialog in which the project file can be selected. The most recently opened projects can be easily accessed by hovering the mouse over the Recent Projects button under the File menu. A project file can then be selected by clicking on the file name from the list of the ten most recent project files. Note CPFD Software, LLC, strives to maintain compatibility of legacy Barracuda VR project files. However, the user may be prompted with warnings or notices when opening a file from a previous version of Barracuda VR indicating that changes have been or should be made to the project file which reflect recent improvements to the software. Please read these notices carefully.

1.3.3 Setting up a model The Barracuda VR GUI is designed to help the user create a full model in a quick and methodical manner. The Navigation tree of the GUI is organized top-to-bottom in a recommended order in which information should be provided. The order is: 1.

Setup Grid - In this step, a CAD file containing the geometry to be modeled is imported into Barracuda VR and a grid for CFD calculations is created. ( more info)

2.

Global Settings - Model-wide parameters are established, such as gravity vector and heat transfer correlations. (more info)

3.

Base Materials - Gas, liquid, and solid materials used within the model are imported or created. ( more info)

4. Particles - The composition, size distribution, and other parameters of particles in the model are specified. (more info) 5.

Initial Conditions - The particle locations, fluid composition, temperature, etc. at the start of simulation are specified. (more info)

6.

Boundary Conditions - All inlets, outlets, injectors, and heating elements in the model are specified. ( more info)

7.

BC Connections - The connections between flux planes leaving and entering the system are specified. ( more info)

8. Chemistry - All chemical reactions to be modeled are specified. ( more info) 9. Numerics - The convergence settings used by the solver can be changed. Typically, however, the default settings will be used. ( more info) 10.

Time Controls - The time step, end time, and restart frequencies of the model are entered. ( more info)

11.

Data Output - The simulation data to be recorded by the solver is specified. ( more info)

Often the setup of a model in Barracuda VR is an iterative process and models will benefit from a process of creating the model, simulating it for a short while, tweaking the model to improve it, and simulating it again. However, following the steps above and in the navigation tree will ensure that the model is set up in a logical manner.

1.3.4 Running a Barracuda model Simulation of a Barracuda VR model is initialized from the GUI by clicking the Run or Run Solver Setup button Run window. Additionally, a previously executed simulation can be restarted from an IC file located within the through the Restart Solver Dialog. Run a simulation will open a new terminal window to run the solver.

4



•

Run Solver Setup - A simulation can be run for a single time step which is sufficient for the user to verify that all boundaries, initial conditions, transient data points, and flux planes are correctly located. Frequent execution of a single time step during model creation is recommended, particularly while boundary conditions and initial conditions are being established to verify correct locations. The GUI provides shortcuts to view boundary condition locations in GMV, as discussed in Run .

•

Run - Once a model is fully created and the setup has been verified, the model can be run to the specified Data Output, can be analyzed end time. Once the model is running, the simulation output data, specified in as it is written by the solver.

•

Restart Solver - The solver will periodically write out IC files during the simulation and at the end of the simulation. An IC file contains all of the particle and fluid states at the time that the IC file is written. A simulation that has ended can be reinitialized to the time and state of the IC file through the Restart Solver Dialog. This is discussed further in Restarting a Simulation.

• Interact - While a model is being run to a specified end time, changes to Time Control, CFL and Dump files, among other things, can be made in the Interact Barracuda Dialog . This is discussed further in Interacting with a Running Simulation

1.3.5 Analyzing Model Results Simulation data is output in Barracuda VR as either three-dimensional model files for visualization (using either GMV or Ensight) or as text-based data files. Both play an important role in data analysis of the model: three-dimensional data provides invaluable insight into the operation, flow structures, and behavior of a model, whereas the text-based data is better for quantification of performance and comparison with experimental data. •

Post Run - Three-dimensional data can be output for each Eulerian cell or for each Lagrangian particle and much of the cell data can be time-averaged as a separate output.

•

Plot Manager - Plotting utilities are bundled with Barracuda VR for visualizing text-based data. The most frequently used text data output are flux planes and transient data points. Further details about plotting data from these files can be found in Plot Manager .

1.4 Reference of Barracuda VR GUI Components The following section provides a listing of the contents and functionality of the navigation tree, menu bar, and shortcut buttons in the Barracuda VR GUI.

1.4.1 Navigation tree The navigation tree is located on the far left of the GUI and provides the main interface for displaying the windows through which a Barracuda VR model is created. The tree is ordered top-to-bottom such that a user can access each section of the tree stepwise, providing input information where necessary. In doing so, the user will specify a complete model in an efficient and methodical manner. The contents are as follows: Table 1.1: Navigation Tree Reference

Item

Description

Barracuda VR Setup Grid

GUI start screen with field for typing a model description and comments Imports STL files and creates a model grid

Global Settings

Defines gravity and thermal properties

Base Materials

Manages properties of gas, solid, and liquid materials

Continued on next page 1.4. Reference of Barracuda VR GUI Components

5


Table 1.1 – continued from previous page Description

Item Particles

Defines particle interaction settings

Drag Models

Manages drag models

Volatiles

Defines particle components that release a gaseous mixture

Particle Species

Specifies particle composition and size

Initial Conditions

Defines initial state of simulation

Fluid ICs

Specifies initial fluid composition, pressure, and temperature

Particle ICs

Specifies initial particle locations, loading, and temperature

Boundary Conditions

Defines fluid inlets, outlets, and wall temperatures

Pressure BCs

Creates fluid opening in model with specified pressure

Flow BCs

Creates fluid opening in model with specified fluid flow rate

Injection BCs

Defines injection of particles, fluid, or tracers within geometry

Thermal Wall BCs

Specifies temperatures of solid walls

Passive Scalar BCs

Creates a tracer for tracking fluid movement

BC Connections

Creates connections between flux planes in and out of a system

Secondary Feeds

Specifies components in direct mass feed

BC Connectors

Connects boundary conditions

Chemistry

Specifies chemical reactions in the model

Rate Coefficients

Defines reaction rate coefficients

Reactions

Defines reactions and reaction rates

Numerics

Specifies solver settings and special use models

Time Controls

Specifies run time, timesteps, and restarts

Data Output

Specifies output variables and data formats

Flux Planes

Creates 2D planes for tracking fluid and particle movement

GMV Output Options

Selects particle and fluid variables to write to Gmv files

Average Data

Selects variables for time-averaging during simulation

2D Plot Data

Defines 2D planes of Eulerian data to write out as a grid

Transient Data

Selects cell-level data to be written every time step

Wall Erosion

Tracks particle-wall collisions to estimate erosion areas

Particle Attrition

Tracks particle-wall collisions to estimate particle attrition

Raw Data

Exports detailed fluid and particle data as text for external analysis

Solver Output Units

Specifies units of measure for data output

Run

Verifies model setup and executes simulation

Post-Run

Provides tools for data analysis

Plot Manager

Create plots using data output by solver

1.4.2 Menu bar The menu bar is located at the top of the Barracuda VR GUI and contains the File, Setup, Run, Graphics and Output, Post-processing, and Help submenus below.

File menu

6



Table 1.2: Navigation Tree Reference New Project

Creates a new project file after prompting the user for a file name and location. New projects should always be placed in a new folder to keep the new simulation files isolated from the simulation files of another project. Opens an existing project file. The user will be prompted with the Open File Dialog to browse for the project file to open. Displays the ten most recently opened project files. Click on file to open project.

Open Project Recent Projects Import from Project Save Project Save Project As

A shortcut to the Import from Project Dialog discussed in Importing chemistry from another project . This dialog provides an interface for importing base materials and chemical reactions from an external project file. Saves any current changes to the project file. Saves the current project file with a new filename. This will open a Save File As Dialog through which the filename and location can be identified. In most cases, the project file should be saved in the current project directory to maintain the path to any STL or SFF files that are used. Saves the current project and associated files to a new directory. This will open a Save Case As Dialog through which the associated files can be selected. The location can also be identified at this time. Closes the Barracuda VR GUI. The user will be prompted to save the project file if any unsaved changes have been made.

Save Case As Quit

Setup menu Table 1.3: Navigation Tree Reference Launch scalestl Setup Grid Reference Grid

A shortcut to the ScaleSTL utility for modifying existing STL files. ScaleSTL is discussed in ScaleSTL utility. Setup Grid Window for importing an STL file and creating a grid. Grid creation is Opens the discussed in Setup Grid . Opens the Reference Grid Window for viewing the current grid and determining the coordinates of features within the geometry. The Reference Grid Window is discussed in Reference Grid .

Run menu

1.4. Reference of Barracuda VR GUI Components

7


Table 1.4: Navigation Tree Reference Generate Grid

Runs the grid generator to produce the model grid from the STL file and gridlines. The grid generator is discussed further in Generating the Grid .

Run Calculation Setup

Run Calculation Restart

Interact

Opens a terminal window for the Barracuda VR solver which runs the current model for a single timestep. This will produce an initial Gmv file which can be used to check boundary condition locations and model initial conditions. Initial data contained within log files can be verified, including initial bed mass and boundary condition cross-sectional areas. Opens a terminal window and starts the Barracuda VR simulation of the current model. Opens the Restart Calculation Dialog that enables the user to restart a simulation from a Barracuda VR IC file. The restarting of simulations is discussed in Restarting a Simulation. Opens the Interact Dialog that allows the user to change simulation parameters such as time step, end time, or convergence settings; instruct the solver to reread simulation input files; or write out current simulation data. The Interact Dialog is discussed in Interacting with a Running Simulation .

Graphics and Output menu Table 1.5: Navigation Tree Reference Gmv Output Options Solver Output Units Average Data 2D Data Transient Data Raw Data

GMV Output Options Window for setting the GMV output variables and A shortcut to the file frequency. GMV output options are discussed in GMV Output Options. Solver Output Units Window in which the units of output data are set. A shortcut to the Output units are discussed in Solver Output Units. Average Data Window in which the user selects variables to be A shortcut to the time-averaged as the simulation runs. Time-averaged variables are discussed in Average Data. 2D Plot Data Window in which the user can select two-dimensional planes A shortcut to the of data to be output. The output and use of 2D plot data is discussed in 2D Plot Data. Transient Data Window in which the user can select cell-level data to be A shortcut to the written to a data file after every time step. Transient data is discussed further in Transient Data. A shortcut to the Raw Data Window in which the user can output full data sets to text files for additional analysis or post-processing. Raw data is discussed further in Raw Data.

Post-processing menu

8



Table 1.6: Navigation Tree Reference Gmv

Ensight Open Terminal Launch Plot Utility Launch bldcol Open Log Files View Images/Movies Convert Images/Movies

Opens a submenu with options for opening GMV for viewing data and various GMV shortcut buttons for viewing the geometry (discussed in Viewing Grid Output ), viewing the model setup (Checking the Model Setup ), and viewing simulation results ( Viewing GMV results). Opens a submenu with options for opening Ensight for viewing data and various Ensight shortcut buttons for viewing the geometry. Opens a terminal window with the path set to the current directory. Opens Plot Utility, the 2D plotting program used by Barracuda VR. Opens Bldcol which is a utility for creating GMV color tables. This is discussed in Colormap editor . Opens a text viewer for Barracuda VR log, flux, and transient data files. The user is prompted with a file open dialog from which any text file can be selected. Opens image or movie files for viewing. The user is prompted with a file open dialog from which any image or movie file can be selected. Opens the Image and Movie Converter Dialog discussed in Converting images and animations. This provides an interface for converting images to different formats and creating a movie from a collection of images

Help menu Table 1.7: Navigation Tree Reference Create Support File Unit Converter Units Reference Check License Check HQ Directory Barracuda User Manual Barracuda Release Guide Barracuda FAQ GMV User’s Manual Read License Agreement About Barracuda

Opens the Support Package Dialog for creating a compact file that contains all necessary project files that can easily be emailed. Support files are discussed further in Creating a Support File. Opens the Units Converter Dialog which provides unit conversion calculator. Opens the Units Reference Dialog which provides a table of unit conversions. Provides the user with information on the Barracuda VR license in use including remaining time. Provides the user with information on the CPFD_HQ directory path that is being used A shortcut to the current Barracuda VR user manual that provides information on all aspects of the software A shortcut to the current Barracuda VR release guide that provides information on new features in the software. Past versions’ release guides are also shown. A shortcut to the Barracuda VR FAQ, a reference document with answers to some common questions about the software A shortcut to the GMV User’s manual which provides information on the use of GMV software. Displays the License Agreement for use of the Barracuda VR software Provides information on the current version of Barracuda VR and contact information for CPFD Software, LLC.

1.4.3 Shortcut buttons The shortcut buttons below provide quick access to commonly used features within Barracuda VR. The full description for each button is provided in Menu bar .

1.4. Reference of Barracuda VR GUI Components

9


Table 1.8: Navigation Tree Reference Quit Button

Shortcut to File → Quit

New Project

Shortcut to File → New Project

Open Project

Shortcut to File → Open Project

Save Project

Shortcut to File → Save Project

Setup Grid

Shortcut to Setup → Setup Grid

Generate Grid

Shortcut to Run → Generate Grid

Run Calculation Setup

Shortcut to Run → Run Calculation Setup

Run Calculation

Shortcut to Run → Run Calculation

Gmv

Opens General Mesh Viewer (GMV) for visualizing simulation output

Ensight

Opens Ensight for visualizing simulation output

Interact

Shortcut to Run → Interact

Open Terminal

Shortcut to Post-processing → Open Terminal

Reference Grid

Shortcut to Setup → Reference Grid

1.5 Commonly Used GUI Patterns Consistency is a goal in the design of the Barracuda VR GUI, and several common patterns are used throughout the GUI. This section discusses these patterns, giving detailed explanations about each one. Subsequent sections in the manual refer back to this section, rather than explaining such details repeatedly.

1.5.1 Comment Field

Fig. 1.3: Example comment field In many dialogs throughout the Barracuda VR GUI, a Comment field is available. The Comment is not required, but it is strongly recommended that concise and descriptive comments be used whenever possible. Such comments are very useful when reviewing old project files, or discussing Barracuda VR project setup details with others.

1.5.2 Add, Edit, Copy, Delete Many different lists are maintained by the Barracuda VR GUI. A common design is used for managing the items in these lists, with four buttons:

10



Fig. 1.4: Example list manager with Add, Edit, Copy, and Delete buttons Add Creates a new list item and raises the corresponding dialog window for defining any necessary input fields. Edit Opens the currently selected item in a dialog window for editing. In most lists, an item can also be edited by double-clicking on it in the list. Copy Creates a new list item as a copy of the currently selected item. The item is populated into the list with most properties copied exactly from the source item. However, the Barracuda VR GUI automatically appends a suffix to flux plane names, to maintain unique flux plane names throughout the current project. The Copy functionality allows for convenient creation of new items based on existing items, in cases where only a few details need to be changed. Delete Removes the currently selected item from the list. In some list tables, the currently highlighted row can also be deleted by pressing the Delete button on the keyboard. Warning: Throughout the Barracuda VR GUI, the Delete operation cannot be undone.

1.5.3 Tabular Input Using SFF Files When defining particle size distributions, boundary condition flow rates, and other input parameters, the Barracuda VR GUI often allows specification via a tabular format. By convention, the tabular data is stored in an external file with an extension of .sff (Standard File Format). The file name is specified by the user. A typical example of a location in the Barracuda VR GUI that allows the user to enter data via an SFF file is shown in Fig. 1.5.

Fig. 1.5: Example choice between SFF file and Specify values When a choice is given between using an SFF file or Specify values, it is recommended to always choose the SFF file format. The reason for this is related to the ability to use Interact while a simulation is running. If Specify values is chosen, there is no ability to change the values during a simulation. But if an SFF file is used, the tabular values can be modified, and then the Update Simulation button can be used to signal to the Barracuda VR solver to re-read the table. This is a very powerful feature, and it is best practice to set up simulations to allow for such interaction. In dialogs similar to Fig. 1.5, there are three components related to the SFF file:

1.5. Commonly Used GUI Patterns

11


Filename text box This displays the currently selected SFF file or allows the user to directly type in a file name if the SFF file already exists in the current directory. Edit button If the filename text box is empty, clicking the Edit button will create a new SFF file and launch the SFF Editor (Fig. 1.6). If the filename text box has a valid SFF filename already listed, clicking the Edit button will open the file in the SFF Editor. Open file button This allows you to select an SFF file via a file selection dialog. The selected file will replace any filename already present in the filename text box. The SFF Editor is not opened in this case.

Fig. 1.6: Example SFF editor window Fig. 1.6 shows an example SFF editor window. Depending on the type of SFF file being edited, the columns could be different than those shown in the figure. However, all SFF files follow a certain set of rules: • All rows of data in the file must be complete, i.e. if any cell in a given row has data, then no other cells in that row can be left blank. (The last row in the table can be left blank.) • The first column of data must be monotonically increasing. For PSD tables, the first column contains particle size information. For transient data tables, the first column contains time values. In the first column, the value in each row must be greater than the value in the previous row. • For cumulative PSD tables, the first column must contain a value of 0 in the first row, and a value of 100 in the last row. • For transient data tables, the first column must contain a value of 0 in the first row. • Data is linearly interpolated between successive rows in the SFF file. • The column Particle Feed specifies whether or not particles are fed with the fluid. If Particle behavior at boundary is set to No particle exit or Particle out flow, the solver will ignore this option. If Use BC Connector data, Particle feed(Slip and vol frac), Particle feed(Slip and mass flux) , or Particle feed(Slip and mass flow rate) option is chosen, the solver will turn the feed On or Off based on the specification in the table. On represents a value of 1 and off represents a value of 0. The value is linearly interpolated and particles are fed whenever it is above 0.5.

Buttons The SFF Editor contains several utility buttons to modify and inspect the tabular data in an SFF file: Add Row Adds a new blank data row below the currently selected or active row.

12



Delete Row Deletes the currently selected or active row. Warning: the SFF Editor does not contain undo functionality when a row is deleted. Check Data Checks the SFF table for input errors. Graph Opens a plotting dialog, allowing quick graphing of any SFF columns. Update Simulation Sends an Interact signal to the Barracuda VR solver, causing it to re-read any SFF files and update the currently running simulation with changes. Save Saves the current SFF table to the specified File name. Save As Save the current SFF table to a new user-specified filename. Close Closes the SFF Editor window. If the table needs to be saved, a dialog box will be raised to ask for verification.

Changing Column Headers Some SFF table columns can be modified to change their meanings. For example, in a PSD table, the user can choose between a size basis of Radius or Diameter . Further, the units of measurement can be changed between m, mm, and µm. For columns that support different options, the option selection dialog is raised by double-clicking on the column header.

1.5.4 Applied Materials Since Barracuda VR supports simulating multiple gas species in the fluid domain, and multi-material particles, a commonly required task is the specification of mixtures of materials. Whenever a mixture needs to be specified, there will usually be a Define fluids button (such as Fig. 1.7) or an Applied Materials button.

Fig. 1.7: Example Define fluids button Clicking this button will raise a new window allowing the user to define the desired composition (such as Fig. 1.8). From the Applied Materials Dialog users can add, edit, and delete materials using the buttons in the GUI.

Adding, editing, and deleting applied materials New materials can be added using the Add material button, which will raise the Material Selection Dialog described below. The mass fraction of existing materials can be edited by selecting the material from the list and clicking on the Edit button. A material can be deleted by selecting it from the list and clicking on the Delete button. Once the desired materials have been added to the list, users should verify that their mass fractions add up to 1.0 and then click the OK button, which applies the materials closes the dialog.

The Material Selection Dialog When users add a material to a mixture, a new dialog is raised where a list of available materials is displayed, Fig. 1.9. Users select each material individually and specify its Mass fraction amount before clicking Apply. If a desired material is not shown in the Material Selection Dialog, the user must add the material to the base materials list. To easily add a material, click on the :guilabel:’Material Properties Library’ button. This will raise a Base Materials Manager window in which new materials can be added to the project. The addition of materials to the project material list is discussed in The Base Materials Window.


13


Fig. 1.8: Example Applied materials dialog

Fig. 1.9: Example Material Selection dialog

14



1.5.5 Flux Plane Options

Fig. 1.10: Flux plane options Flux planes are found throughout Barracuda VR projects, and the options available are similar in all cases. Fig. 1.10 shows a typical example of a Flux plane options section. Internal flux planes are discussed in Flux Planes. The typical options available are: Flux plane name Specifies the filename used for the primary flux plane file. The filename specified will also be used as the base name for any secondary flux plane files, such as gas composition, particle Subdivide by radius , and raw particle data. It is recommended to follow these conventions when specifying the Flux plane name : • Start the filename with FLUX_ for internal flux planes, or FLUXBC_ for boundary condition flux planes. This will make the flux plane files show up in the default filters used in other locations in the Barracuda VR GUI. • Do not use spaces in the filename; instead, use underscore characters. This makes writing scripts easier, since spaces in filenames often cause problems. • If no name is specified at a boundary condition flux plane, NO flux plane file is created and no data is output. Gas species flux plane behavior If multiple gaseous species are used, users may wish to have the flux plane track the flux of each species crossing the plane. The desired output (no output, mass flow rate, mass fraction, mass time cumulative, or mass time cumulative plus minus) is selected in the Gas species flow behavior menu. Gas species flux plane data will be contained in a file with the Flux plane name followed by _gasSpc###_###. Subdivide by radius Users may request that the particles crossing the flux plane be subdivided into size groupings by selecting the Subdivide by radius check box. The number of radius subdivisions used in the output file is controlled independently for each boundary condition flux plane by using the Radius divisions spin-box. Subdivide by radius flux plane data will be contained in a file named with the Flux plane name followed by _pSpc_####. Output raw particle data Users may request very detailed output data for particles crossing boundary condition flux planes. Currently, this option is not available for internal flux planes. The raw particle output data is written to a file named with the Flux plane name followed by _raw_particle.

Sign Convention Data for flow into the domain will be written to the flux plane file with positive values while data for flow out of the domain will be written with negative values. Internal Flux Planes follow a different sign convention, where data will be written to the flux plane file with positive values when flow is in the postive axis direction while data will be written with negative values when flow is in the negative axis direction.


15


16


CHAPTER

TWO

SETUP GRID

Barracuda VR uses an Eulerian-Lagrangian approach for simulating particle-fluid dynamics in which the solid particle phase is modeled as discrete Lagrangian points and the fluid phase is modeled on an Eulerian grid of cells. The setup of the grid is often the first task in the creation of a Barracuda VR model because the grid has multiple roles which make it truly the foundation of any simulation: not only does the Barracuda VR grid determine the spatial resolution with which Barracuda VR will calculate all fluid fields such as pressure, velocity, composition, and temperature but the grid also establishes the locations and extents of exterior walls, internal solid features, and openings (inlets and Setup Grid in the outlets) in the geometry. The setup grid functionality in Barracuda VR is accessed through navigation tree. The Barracuda VR grid is generated based on two user inputs: a CAD file and a set of grid line locations. The computer-aided design (CAD) file, commonly referred to as the “STL file” , contains the model geometry to be used and must be created outside of Barracuda VR by 3D design software (most CAD programs have the ability to export geometry in STL format). Based on the extents of and features within the CAD file, the user creates the set of x , y , and z -directional grid lines used by the grid generator to create the Barracuda VR grid. Determination of the optimum location and quantity of grid lines by the user is the major task of creating a grid in Barracuda VR and it is recommended that the user spend an appropriate amount of time on this step. Making changes to a grid after the rest of a model is setup is possible, but can require additional adjustments to boundary conditions and initial conditions locations as well. A good grid has the following characteristics: • Accuracy: Important features of the model are adequately represented on the grid. This includes external solid walls and fluid boundaries as well as internal solid features such as cyclones, tube bundles, and distributors. Ensuring the adequate representation of these features requires the placement of x, y, and z grid lines at appropriate locations within the model domain. • Resolution: The model has enough resolution to accurately calculate the particle-fluid dynamics. The number of grid lines that are used in the model determines the precision with which the fluid flow can be calculated especially in regions with large gradients in pressure, velocity, temperature, or composition. • Uniformity: In Barracuda VR, uniformity of the cell sizes is important for producing a stable and efficient simulation. Changes should be made gradually from regions of small, high-resolution cells to regions of larger, lower-resolution cells. • Number of cells: It is often desirable to keep the number of cells in a grid at a minimum while maintaining the accuracy, resolution, and uniformity of the grid. Increasing the number of cells in the model increases the computing requirements and the calculation time of a simulation as well. Often, the best grid is one that produces the best answer in the shortest amount of time.

Setting up a grid The basic workflow for establishing a good grid consists of importing an STL CAD file into Barracuda VR, setting the dimensional units, placing grid lines, checking for uniformity, generating the grid, and viewing the grid. Typically this is an iterative process by which the grid is modified, generated, and viewed multiple times in an effort to produce a grid with adequate accuracy, resolution, and uniformity, while maintaining an acceptable number of cells.

17


Step 1 - Importing an STL file An STL file is a stereolithography file which contains the model geometry that has been drawn in an external computer-aided design (CAD) program. The Setup Grid Window contains the functionality for adding and removing STL files from the model through the Geometry Tab. An example STL file view in Barracuda VR is shown in Fig. 2.1 (a) and (b). Step 2 - Setting STL units STL files contain only dimensional information about the geometry and no information on the units of length used in the drawing. After an STL file has been imported, the drawing units must be specified (meters, feet, inches, etc). Step 3 - Placing grid lines Once an STL file has been imported into the model, x, y , and z grid lines are placed within the model which determine the extents of the model, capture features important to the model, and provide necessary grid resolution where needed. This can be done either by manually placing grid lines or by importing grid line locations from an existing project file (see Adding and Modifying Grid Lines). In Fig. 2.1 (c) and Fig. 2.1 (d), a sample set of grid lines is overlaid on an STL file. Step 4 - Checking grid line uniformity Once grid lines have been placed, they can be checked for uniformity prior to generation of the grid. This step produces an XY plot of the grid line spacing in each direction and can provide helpful information to verify that the cell sizes have the desired uniformity and that smooth transitions exist within the grid. The Check Grid Button is discussed in Checking the Grid . Step 5 - Generating the grid Once the grid lines have been placed and checked, the grid generator in Barracuda VR is executed. The grid generation process removes cells that fall outside of the geometry, cuts cells to match the STL geometry, and removes or merges cells to produce a three dimensional array of cells that is optimized for the Barracuda VR simulation. The grid generator produces files for viewing the new three-dimensional model in GMV. Step 6 - Viewing the grid Once a grid has been setup and generated, the gridded geometry can be viewed as a threedimensional model in GMV. There are shortcut buttons set up for easily viewing a transparent geometry, the geometry grid lines, the original CAD geometry, and comparisons between the gridded geometry and the CAD geometry. An example grid is shown in Fig. 2.1 (e).

Fig. 2.1: The grid generation process for modeling a cyclone in Barracuda STL file: (a) top view of cyclone STL lines, (b) side view of cyclone STL lines, (c) XY grid lines with STL file (major grid lines in black, minor grid lines in blue), (d) YZ grid lines with STL file, and (e) cyclone grid generated from STL file and grid lines

Grid line types 18

Chapter 2. Setup Grid


The set of grid lines used by the grid generator typically consists of both major and minor grid lines. Both types of grid lines are treated equally by the grid generator but the manner in which they are specified is different. • Major grid lines have an x , y , or z location specified by the user and can be added, moved, or deleted by the user in the GUI. It is recommended that major grid lines be used whenever a grid line is needed at a specific location. • Minor grid lines are automatically placed in the space between major grid lines by the GUI and are dependent on the number of cells and growth factor specified by the user for that space. When a non-zero growth factor is used, the spacing between each minor grid line in the space will continuously increase or decrease at the specified rate which is useful for smoothly growing or shrinking the grid. Since the placement of the minor grid lines is a function of the surrounding major grid lines, it is recommended that minor grid lines be used to provide spatial resolution for the calculation but not precise grid line placement.

The coordinate system During the setup of a model, the user often deals with two different coordinate systems in Barracuda VR: the xyz coordinates and the ijk coordinates. The xyz coordinate system is a simple reference to the spatial location within the model in the specified units. The xyz can be positive or negative and the validity of points is a function of the extents of the STL file and the units specified. When the grid is formed in Barracuda VR, cells are created from the spaces between gridlines and the numbering of these cells in each direction creates the ijk coordinate system. As a result, the ijk coordinate references a specific cell within the model rather than a point in space. The ijk coordinate system is used in Barracuda VR to specify all boundary and initial conditions in a model whereas either coordinate systems can be used in post processing. The Grid View Window and the Reference Grid provide a graphical interface for correlating the xyz and ijk coordinate systems in a model.

2.1 The Setup Grid Window The Setup Grid Window, provides the interface for creating a Barracuda VR grid - from the importing of an STL file to the final viewing of the generated grid - and is shown in Fig. 2.2. Most of the Setup Grid Window space is devoted to grid view panes but also contains the Geometry Tab, Baffles Tab, Grid Controls Tab, and Grid lines tab. These tabs exist on separate panels within the GUI which can be hidden by the user to increase the grid viewing area. Display of these panels can be managed by clicking on the Docks button in the upper right of the setup grid window.

2.1.1 Grid uitility buttons The Grid uitility buttons shown at the top of Fig. 2.2 provide access to the Import Grid dialog box, Undo and Redo functionality, the View STLs utility, the Check Grid utility, the Grid Generator , and View Output shortcut buttons for viewing the generated grid. The grid utilities are shown at the top of the Setup Grid Window in Fig. 2.2.

Importing an existing grid The Import Grid button at the top of the Setup Grid Window allows grid-related information to be imported from another project file. When the Import Grid button is clicked, a file dialog is raised for selecting the project file from which the new grid should be imported. Any combination of the grid lines, primary STL geometry files, and/or baffle STL geometry files can be selected for import. Any changes made with Import Grid can be undone by using the Undo button at the top of the Setup Grid Window.

2.1.2 Grid view panes The Setup Grid Window contains three view panes which display the model geometry; show current locations of major and minor grid lines; provide an interface for adding, moving, and deleting major grid lines; and facilitate the

2.1. The Setup Grid Window

19


Fig. 2.2: Setup Grid Window loaded with sample STL file. The setup grid window contains the grid utility buttons (top) and four tabs with information on Geometry, Baffles, Grid Controls, and Grid lines.

20



adjustment of minor grid line spacing in a model. The three-dimensional geometry is shown in the view panes as two-dimensional projections onto the xy , xz , and yz planes; the advantage of this being that all STL lines in a plane can be viewed together which provides important understanding of how grid line placement will affect the model at all depths. At the top of each viewing pane are text boxes which show the xyz and ijk coordinate values for the current mouse location within the box. This allows the user to determine the ijk values of any feature in the model or specific xyz point by moving the mouse to the desired location and noting the coordinate values. The xyz units displayed can be set by changing the selection of the xyz unit drop-down box in the Grid Controls Tab. Each view pane can be sized separately by dragging the pane separator bars to the left or right. And in each view pane the model geometry can be panned and zoomed in order to gain the most convenient perspective possible. The grid view can be panned left, right, up, and down by holding down on the center mouse button and “dragging” the geometry in the desired direction. For mice with scroll wheels, it is usually possible to push down on the scroll wheel itself as if it is a button; in these mice, pushing down the scroll wheel is equivalent to using the center mouse button. The geometry can be zoomed in by holding the right mouse button down and moving the mouse up. The user can zoom out of a geometry by holding the right mouse button down and moving the mouse down. Alternatively, the scroll wheel of a mouse will also zoom in and out.

2.2 Geometry Tab STL files in the model are controlled using the STL file list shown in the Geometry Tab of the Setup Grid Window in Fig. 2.2. This interface allows STL files to be added and removed and STL drawing information to be displayed. Once added, the units of the STL file need to be set.

Adding STL files New STL files can be added to the geometry by clicking Add Geometry and then selecting an STL file from the file dialog box that is displayed. STL files that are selected outside of the project directory will be copied into the project directory by Barracuda VR.

Removing STL files STL files can be removed from the file list by selecting the STL file from the list and clicking Remove Geometry . STL files removed from the file list are only removed from the Barracuda VR model and are not deleted from the project directory.

Viewing STL properties STL file information can be viewed by double-clicking on the file name in the list. This will raise a dialog that displays the STL file encoding, the x, y, and z extents of the drawing, and the number of triangle faces.

Setting STL units The dimensional units of the STL file must be set after the STL file is added to the model because this information is not contained within the STL file itself. The units of the STL file can be set by selecting the units from the STL unit dropdown list on the Grid Controls Tab. Both metric and English units are available: meters (m), centimeters (cm), millimeters (mm), yards (yd), feet (ft), inches (in).

Modifying STL files If basic transformations to an STL file are required, this can be accomplished with the ScaleSTL utility in Barracuda VR prior to adding the STL file to the Barracuda VR model.

2.2. Geometry Tab

21


Fig. 2.3: STL file information box showing file encoding, model extents, and number of facets

2.3 Baffles Tab

Fig. 2.4: The Baffles tab. This tab contains the controls for creating, adding, manipulating, removing, importing, and exporting baffles. Baffles are 2-dimensional (zero thickness) sub-grid structures that can affect particle and fluid flow. Particles bounce off baffles as if they are solid walls. Baffles can induce a pressure drop on fluid flow via a K-factor, in the same way as a porous media. Baffles can be treated as transparent to fluid flow (K=0), or induce a pressure drop to impede fluid flow (K>0). Baffles are intended to model thin walls that would be difficult to capture with the grid. Applications include fluid-particle guide vanes for redirection of flow, and fluidized bed internals. Baffles have the following limitations, which can be important in some models: • Computational cells are not split or cut into non-communicating cells by baffles. • Baffles cannot be used as heat transfer surfaces. • Baffles must be greater than 2% of the cell size in any direction. As shown in Fig. 2.5, baffles are drawn in the 3 grid viewports along with the normal grid lines and STLs. Each baffle can be assigned its own color, and hidden / shown independently. When hiding STL triangles in the grid viewports, it is possible to include or exclude baffles from the hiding settings.

22



Fig. 2.5: Baffles shown in the three grid view panes.

2.3.1 Baffles Main Table The largest component in the Baffles tab is the main table, which lists all baffles in the project, and allows for various transformations of the baffles. By default, a baffle is added to the project as a single instance called the Master Baffle. This instance can be repeated, rotated, scaled, and so forth using the cells available in the main table. The columns in the baffles main table are: • Filename: the name of the 2D STL file for the baffle. • Enabled: an enabled baffle is used in the simulation. A disabled baffle is not used. • Visible: show or hide baffles in the three grid viewing windows. • Color: set the color of the baffle triangles in the three grid viewing windows. • x-Kfact, y-Kfact, and z-Kfact: set the K-factors in the x-, y-, and z-directions. Combined with the blockage factors, K-factors are used to simulate the resistance of the baffle to the fluid flow and result in a pressure drop as fluid flows through the baffle. The bigger the K-factor and the blockage factor are, the larger the pressure drop, according to this formula:

∆P =

1 ρKv 2 2

where ∆P is the pressure drop (Pa), ρ is the fluid density (kg/m 3 ), K is the value of the K-factor, and v is the fluid velocity (m/s). • x-MBL, y-MBL, and z-MBL: set the Master Baffle Location (MBL), which is the location of the master baffle within the system (indicated by the green dot in Fig. 2.6). The units of x-, y-, and z-MBL are the same as those of the geometry STL. All baffle transformations and repetitions are made with respect to the Master Baffle Location.

2.3. Baffles Tab

23


• x-Rep, y-Rep, and z-Rep: set the number of repetitions of the master baffle, in the x-, y-, and z-directions. • x-Intrvl, y-Intrvl, and z-Intrvl: set the spacing interval between repetitions of the master baffle, in the x-, y-, and z-directions. The unit of measure for the spacing interval is the same as that set for the geometry STL. • x-Scale, y-Scale, and z-Scale: set the scaling of the master baffle in the x-, y-, and z-directions. A scale of 1 means to use the original size from the 2D STL file. • x-Rot, y-Rot, and z-Rot: set the rotation of the master baffle with respect to the x-, y-, and z-axes. Rotation is specified as angles with units of degrees. The baffles main table also has the ability to hide/show and rearrange columns: by double-clicking on any of the column headers in the Main Baffles Table, a dialog will be shown that allows columns to be hidden, shown, or rearranged. This can be useful for making the table smaller, in cases where not all columns need to be used. Values specified in any hidden columns are still applied to the baffles in the simulation.

2.3.2 Add Baffle There are three ways to add baffles to a project: make a baffle using Baffle Maker , add a previously-created baffle file, or import a baffle with all settings from another project. Previously-created baffle files can be added to the geometry by clicking Add Baffle and then selecting a baffle STL2d file from the file dialog box that is displayed. Baffle STL2d files that are selected outside of the project directory will be copied into the project directory by Barracuda VR. When a baffle is added, a single instance of its geometry is created in the domain. This instance is called the Master Baffle, and it is the reference for any transformations (such as repeating, scaling, rotating, etc.). For more details on available transformations, see Baffles Main Table. Note that, currently, Barracuda VR only supports the special 2-dimensional STL files created by Baffle Maker . The default extension of the baffle STL file name is stl2d or STL2D to be distinguished from the regular STL files (extension stl or STL) that define the model geometry.

2.3.3 Remove Baffle Baffle STL files can be removed from the file list by clicking in any cell of the baffle’s row in the main table and clicking Remove Baffle. Baffle STL files removed from the file list are only removed from the Barracuda VR model and are not deleted from the project directory.

2.3.4 Baffle Maker The Baffle Maker interface is used to create special STL files that define baffles used in Barracuda VR. A baffle generated by the Baffle Maker is a surface in 3-dimensional space with zero thickness. Because of the zero-thickness of the geometry, the STL file is called a 2D STL file. Baffle Maker is able to create 2D STL files for several different built-in shapes, and each shape has a set of parameters used to control the dimensions of the baffle created. The green dot in Fig. 2.6 indicates the position that will be used as the Master Baffle Location, once the baffle is added to the current project. The Baffle Maker interface includes the following selection and input fields: • Shape: drop-down menu to select between the available built-in shapes of baffles. These include: Arc, Chevron, Ibeam, Ubeam, and Plate. • Baffle Unit: the baffle dimension unit, which is automatically set to be the same unit as the model geometry for consistency. The baffle unit is displayed here as a reminder to the user of the baffle STL unit that will be applied when the baffle is imported into the current project.

24



Fig. 2.6: The Baffle Maker window.

2.3. Baffles Tab

25


• Baffle Shape Parameters: depending on which shape is chosen, certain parameters (angles and lengths) can be specified to control the final dimensions of the baffle. Angles are in units of degrees, and lengths are in the same units as the model geometry used in the current project. The built-in Arc shape also includes an input field for Curve Resolution, which controls how smooth the resulting STL file will be for approximation the circular curvature of the Arc. Lower values give less smoothness, while higher values give more smoothness. • Depth and Direction: all shapes have these parameters. Each baffle shape can be thought of as a 2-dimensional profile (currently a curve) in a plane. The depth, d, determines how far the 2D profile should be extended in the normal direction of the plane to create the final, full-length baffle. The normal direction of the plane is the orientation of the primary axis of the baffle, which is specified by the Align with drop-down menu. Once the parameters for a baffle have been defined, the following options are available for creating the baffle: • Add to current project: check this box to automatically add the created baffle to the main baffle table in the current project. • Load Settings: selects a baffle settings file that contains the parametric information used to create a baffle. • Save Settings: saves the current parametric settings of the baffle to a file. This is useful because the generated baffle STL file does not contain a record of the parametric settings used during its creation. By saving a settings file, you can have a record of the baffle parameters, and easily modify the baffle in the future by loading the settings file. • Generate: generates the baffle STL file based on the current parameters.

2.3.5 Import Imports a comma-separated value (CSV) file of baffle information into the Main Baffle Table. Note that any baffles already in the table will be deleted.

2.3.6 Export Exports the Main Baffle Table to a CSV file, which can then be used in a different project to import the same baffle settings.

2.3.7 View Baffle Button

The magnifying glass button displays 2D STL baffle files in GMV for quick inspection. Note that the baffles are displayed alone, with no computational grid information. To see the grid and baffles together, it is necessary to run Run window shortcut buttons. the simulation for a single time-step, and then use one of the

2.4 Grid Controls Tab The grid controls tab, shown in Fig. 2.7, contains many of the important components used during the interaction with the Setup Grid Window: Options, Grid lines, and View Controls.

2.4.1 Options The Options section contains the following information: the total number of cells in the model; lists for setting the units of length used in the model and display; and access to the Advanced Options and Set uniform grid . The total number of cells in the model is the sum of the real and null cells. The real cells are the cells inside the domain of grid

26



Fig. 2.7: The Grid Controls tab. and CAD, while the null cells are gridded cells outside of the CAD model. No computations are performed in null cells. Units for STL and XYZ can be specified using the dropdown boxes. Advanced Options The Advanced Options dialog box, shown in Fig. 2.8, contains settings for the grid generator and display settings for lines in the grid view panes.

Drawing options The drawing options section contains settings that control the way the user interacts with the setup grid view panes. Mouse tolerance for moving/deleting lines The mouse tolerance controls the required precision when selecting a line in the setup grid view pane for moving or deleting. Lower values of tolerance will require the user to click closer to the mouse to select a line whereas higher values will relax this requirement but may lead to the accidental movement or deletion of surrounding lines. The default value in Barracuda VR is 0.04. Tolerance for culling triangle normals This value controls which triangles will be hidden by the Hide back triangles and Hide front triangles checkboxes on the grid controls menu. Lower tolerance values will cause more triangles to be hidden. The default value in Barracuda VR is 0.2. Pixel widths of grid lines and STL lines The pixel widths for all grid lines and STL lines determine how thick the lines will be displayed in the setup grid view panes. The default settings are for major grid lines to be displayed with a width of 2 and all other lines to be displayed with a width of 1. In some situations it can be helpful to set the Pixel width for STL lines to 2 or 3 to distinguish the STL lines from grid lines. Line colors The display of lines colors for all line types in the view panes can be set by clicking on the line type button. This will raise the color editor dialog for setting a color, shown in Fig. 2.9. Restoring defaults The display settings can be restored by clicking the drawing options Restore defaults button.

Grid generator methods The grid generator method determines how small cells are handled by the grid generator. Typically, small cells are created when a grid cell is only partially within the STL geometry and the grid generator forms a “cut cell” from the full cell to conform to the geometry. If a very small cut cell is formed in the grid it will not be able to contain a statistically relevant number of computational particles in the model and must be either removed or merged with larger nearby cell to maintain model stability. The user can specify the grid generator method to Remove small cells only , in some case it may be appropriate to Merge and remove small cells . Merging and removing small cells is discussed further in Generating the Grid . Restore grid generator defaults Default merge and remove parameters can be restored by clicking the Restore default settings button under the grid generator method. Merge parameters By default, cells that have less than 4% of the volume of the full cell or any cell that is very long (aspect ratio greater than 30:1) will be merged with a nearby cell if possible. The merging process will go through 150 iterations before applying the cell removal parameters. The values used for cell merging will dramatically affect the grid and stability of the model and it is recommended that default values be used in most cases.

2.4. Grid Controls Tab

27


Fig. 2.8: Advanced grid options dialog for setting view pane drawing options and grid generation parameters.

28



Fig. 2.9: Color selection dialog for setting the line colors displayed in the setup grid view panes Remove parameters By default, cells that have less than 4% of the volume of the full cell are removed and any cell that is very long (aspect ratio greater than 15:1) will be removed. The values used for cell removal will dramatically affect the grid and stability of the model and it is recommended that default values be used in most cases.

ADMesh preprocessing ADMesh is a program that detects and repairs problems with STL files, such as holes and reversed normals. A grid generated with a bad STL file will often not conform to the STL file as expected. In these cases, preprocessing the STL file with ADMesh can fix the problems and produce the grid as expected. By default, preprocessing with ADMesh is not done by the grid generator.

Include grid faces in post-processing files Include grid faces in post-processing files: include info about what this is and what options the choices give you

STL normal check When the Color polygons by outward normal checkbox is selected, the grid generator will produce a GMV file with the STL file geometry split into polygon groups based on the magnitude of the vector component selected in the STL normal check drop-down box. These polygon groups, shown in GMV as n1 through n10, can be displayed with individual colors to help the user look for any reversed normal in an STL file. This feature is typically only used when it is suspected that there are errors present in an STL file.

Setting a uniform grid The Set Uniform Grid dialog, shown in Fig. 2.10, is accessed through the grid controls menu. This utility creates a set of grid lines that have approximately uniform spacing, are within the specified x, y , z extents, and will form a grid with a cell count that closely matches the user’s target number. Often, setting a uniform grid is the first step to creating

2.4. Grid Controls Tab

29


a grid in Barracuda VR. Any changes to a grid made with the Set Uniform Grid dialog can be undone with the Undo button at the top of the Setup Grid Window.

Fig. 2.10: Set Uniform Grid dialog establishes a grid of approximately approximately uniform cells within the specified specified extents

Grid extents The grid extents to be used are specified in the textboxes at the top of the dialog. These values can be specified by the user or automatically populated by Barracuda VR to match the extents of the STL geometry when the button next to Reset min and max to the STL file(s) limits is clicked.

Keep major grid lines and growth rates When an existing grid has been created, the Set Uniform Dialog can be used to increase or decrease the total number of cells in the model. model. When the Keep all current major grid lines or Keep all current growth values checkboxes are selected, Barracuda VR will leave the existing grid lines in place and adjust the number of cells in between to achieve the target number number of cells. The Grid lines section contains contains the checkboxes for for modifying, deleting, deleting, and moving grid lines. The View iew Contro Controls ls sectio section n contai contains ns the butto button n for resett resetting ing panel panel views, views, check check boxes boxes for diagon diagonal al STL lines lines (trian (triangle gles), s), and the subset sliders. sliders. The subset sliders sliders are used to limit limit the display display of STL lines lines to a range of x , y , and z values. The use of the subset sliders allows the user to eliminate unneeded lines from the viewing pane to see other parts of the model more clearly.

2.4.2 Grid lines lines tab tab The grid lines tab provides a text-based input for viewing and changing grid line locations, minor grid line spacing, and minor grid line growth growth rates in each dimension. dimension. Usage of the grid lines lines tab is discussed discussed in Using the grid lines editor .

30



2.5 Chang Changing ing the Grid View View During the gridding process, it is typical for users to focus on only a small area of the model and have this area be maximized in the viewing panes to facilitate the accurate placement placement of grid lines. To do this, the user will move the area of interest to the center of the view pane, zoom in to the maximum extent possible, and remove any unnecessary STL lines that are overlaying in the view plane.

Panning The grid view can be panned left, right, up, and down by holding down on the center mouse button and “dragging” the geometry in the desired direction.

Zooming The geometry can be zoomed in by holding the right mouse mouse button button down and moving moving the mouse up. The user can zoom out of a geometry by holding the right mouse mouse button down and moving the mouse down. Alternatively Alternatively,, the scroll wheel of a mouse will also zoom in and out.

Resetting the view panes The grid view can be reset to the initial viewing state by clicking the Reset views button on the grid controls menu.

Filtering STL lines Large Barracuda VR models will often have many STL lines which can cause the view pane to appear cluttered if all viewed at once. In some cases, it is helpful to hide some of the STL lines through either hiding triangles or using the subset sliders. Hiding triangles Often an STL file contains many diagonals and triangles that provide curvature and resolution to the model. It can be helpful to remove these diagonals and triangles from the view by using the Hide back triangles or Hide front triangles checkboxes checkboxes,, located located on the grid controls controls menu. The criteria criteria for hiding hiding the triangles triangles can be adjusted in the Advanced Options dialog box. Examples of using the hide triangle checkboxes is shown in Fig. 2.11. 2.11. Subset sliders The subset sliders limit the displayed STL lines to those within the minimum and maximum values in the x , y , and z directions directions.. Subset Subset values in each directio direction n can be set by either either using the mouse to drag the sliders sliders to the desired desired amount amount or by typing in the minimum minimum and maximum maximum values directly directly.. When the values values are set, lines will appear in the view panes showing the locations of the subset minimums and maximums. The subset sliders can be reset to the extents of the model by clicking the “Reset subsets” button to the left of the interface. An example of using the subset sliders is shown in Fig. 2.11. 2.11.

Changing display colors The colors and thickness of the STL lines and grid lines can be changed through the Advanced Options dialog.

2.6 Addin Adding g and Modifying Modifying Grid Lines Lines Barracuda VR provides many different approaches for adding, removing, or modifying grid lines. The set of grid lines can be changed graphically by using the mouse within the setup grid view panes, textually by using the keyboard in in the grid line editor, automatically using Set uniform grid , or a grid can be imported from an existing model. The user may utilize multiple methods while setting up the grid but when creating a grid from scratch, it is typical to first use the Set uniform grid utility utility to establish an initial grid with the approximate number of cells to be used in the model. This will create a grid with major grid lines at the extents with minor grid lines populated uniformly in the space between from which the user can then add major grid lines as needed. needed. All grid line modificatio modifications ns can be reverse reversed d if needed needed using the Undo and Redo buttons at the top of the Setup Grid Window.

2.5. Changing the Grid View

31


Fig. 2.11: Unneeded Unneeded STL lines can be hidden in the view panes panes to improve improve clarity: clarity: (a) top view (xy projecti projection) on) of cyclone STL lines, (b) top view of cyclone with back and front triangles hidden, (c) top view of cyclone using subset sliders in z direction to hide cyclone dipleg and cone, and (d) side view of cyclone with blue lines showing locations of subset sliders

2.6.1 Chang Changing ing grid lines graphically graphically Additions of major grid lines and modifications of minor grid lines are made graphically by clicking Modify x, Modify y, or Modify z on the controls controls menu and then moving moving the mouse to a view pane. A line under the mouse will be displayed which shows the location of the major grid line to be added or the block of minor grid lines that will be modified. With the mouse in the correct location, a left mouse click button will • add a new major grid line if there are no existing existing grid lines in the modification direction direction • prompt the user for a number of cells to add if the mouse click is outside outside the existing set of grid lines, or • raise the Modify Modify Grid line Dialog box, shown shown in Fig. in Fig. 2.12, 2.12, if the mouse is within an existing block of cells.

Fig. 2.12: Modify Grid Line dialog box for adding adding major grid lines lines and modifying minor minor grid lines When the Modify Grid Line Dialog is opened, the user is provided the option of adding a major grid line at the mouse location or modifying the set of minor grid lines in which the user clicked. Split cells with grid line The line The major grid line will be placed at the desired location and the cells between existing major grid lines will be divided in between the new spaces to keep the new cells as equal in size as possible. Splitting cells with a grid line will not result in the addition of any additional cells to the system. The user can optionally change the growth rate of the cells which will be applied to the lower side of the new cell blocks.

32



Add grid line and cells The cells The major grid line will be placed at the desired location and the existing cells will be moved to the upper side of the new major grid line. Choosing this option requires the user to specify a new number of cells to be added on the lower side of the new major grid line ( Enter number of mesh cells between grid lines). The user can optionally change the growth rate of the cells which will be applied to the lower side of the new cell blocks. Enter cells The cells The existing block of cells will be modified to have a new number of cells, specified as Enter number of mesh cells between grid lines and enter a new growth rate. The current number of cells within the block is displayed. Note that the number of cells will always be one more than the number of minor grid lines within a cell block.

Moving major grid lines The locations of major grid lines can be moved by clicking on Move from the grid controls menu and then clicking and dragging any major grid line visible in the view pane. The moving of a major grid line will not change the number of cells on either side of the line and the spacing of minor grid lines will be adjusted by Barracuda VR. Grid line moves can be undone using the Undo button at the top of the Setup Grid Window.

Deleting major grid lines Major grid lines can be deleted by clicking Delete from the grid controls menu and then clicking on any major grid line that is visible visible in the view pane. The deletion deletion of a major major grid line will cause the two cell block on either either side of the deleted deleted line to merge into one new cell block. block. The number number of cells cells will not change in this operation. operation. Grid line deletions can be undone using the Undo button at the top of the Setup Grid Window.

Shortcut keys The selections for adding and modifying gridlines can also be selected using keyboard shortcuts from within the setup grid window window. For example, example, to modify the x-direct x-direction ion gridlines, gridlines, the user can simply type x rather than clicking the Modify x checkbox. checkbox. This can be confirmed by verifying verifying that the appropriate appropriate checkbox checkbox is selected. selected. The full list of keyboard shortcuts is: • x - Modify x • y - Modify y • z - Modify z • m - Move • d - Delete

2.6.2 Using the the grid lines editor editor The grid line editor, shown in Fig. in Fig. 2.13 provides 2.13 provides an interface for modifying the grid line set directly with a keyboard. The x, y , and z textboxes contain all the information on grid line locations, cell block size, and minor grid line growth rates. rates. The text data in the grid line editor can be edited edited directly directly which which is useful for providing providing precise precise location locationss for grid lines, copying a set of grid lines from a text file or spreadsheet program, or making fast grid line edits with the keyboard. The grid line editor is typically exposed in the Setup Grid Window by clicking the Grid lines tab, shown in the lower left of Figure Fig. Figure Fig. 2.2. 2.2.

Grid line editor format The grid line editor consists of three textboxes corresponding to the grid line sets in the x , y , and z directions. Each grid line is represented by a row of text which consists of three columns separated by spaces. 1 - The first column is the number of cells to place between the current grid line and the grid line on • Column 1 the row above. This column is omitted from the first row of each text box. 2 - The second column contains the location of the grid line in the units of the STL file. • Column 2 -

2.6. Adding and Modifying Grid Lines

33


Fig. 2.13: Grid Line Editor for editing major grid line locations, cell counts, and growth rates • Column 3 - The third column is the growth rate to use when spacing cells in the space between the current grid line and the grid line on the row above. This column is omitted from the first row of each textbox.

Editing the grid line data Edits to the grid line specification can be made by clicking inside a textbox and using the keyboard to edit the text with separation of columns being maintained by at least a single whitespace. When the mouse is moved outside of the textboxes, the changes to the grid line data will be recognized by Barracuda VR, the grid line locations in the view panes will be updated, and the text within the textboxes will be reformatted to maintain proper alignment of columns. Any changes made in the grid line editor can be undone using the Undo button at the top of the Setup Grid Window.

2.7 Checking the Grid During and after making modifications to the grid, it is important to check the created for the cell count and uniformity. Frequent checking of the grid will help create a grid that is optimized for both speed and stability.

Number of cells in model Barracuda VR tracks the total number of cells in the grid as changes to the grid line set are made by the user. This value, displayed at the top of the grid controls menu, is based on the number of cells created by the major and minor grid lines in the model and is independent of the STL geometry. Once the grid is generated, the total cells will be divided into real cells, which are within the STL geometry, and null cells, which are in the set of grid lines but are not part of the fluid domain. The number of real cells in the model is available within the grid.log file after the grid generator has been run.

Using the check grid utility When creating a grid, it is important for numerical stability to make a grid that is as uniform as possible and transitions between small cells and large cells should be made as gradually as possible. The check grid utility is a useful tool for plotting the cell spacing within a model. When the Check Grid button at the top of the Setup Grid Window is clicked, Barracuda VR will analyze the current set of grid lines and produce a grid check plot, shown in Fig. 2.14. The grid check plot contains the cell spacing in the model, normalized by the minimum cell spacing in any direction. The i, j , and k coordinates are on the x-axis, allowing the user to identify any regions of non-uniformity.

2.8 Generating the Grid When the user clicks the Generate Grid button, located at the top of the Setup Grid Window, Barracuda VR will automatically create a grid from the set of grid lines and the STL files loaded into the model. For very large systems or when Merge and Remove is used, the grid generation process can take up to a few minutes and a progress bar will

34



Fig. 2.14: Check grid output showing normalized spacing of cells in x , y , and z directions

2.8. Generating the Grid

35


be displayed showing the percent completion. The grid generator creates the grid from the grid line set and STL file by the following steps: 1. locating intersection points between the grid lines and the STL lines, 2. connecting intersection points to form a grid, 3. removing cells that are outside the STL geometry (null cells), and 4. merging and/or removing small cells that are formed based on settings in the Advanced Options dialog, 5. and writing structure files for viewing the grid in GMV. Information on the grid generation process is logged in grid.log, located in the project directory, which includes information on the number of real cells, null cells, and the location and number of cells removed due to size or aspect ratio.

2.9 Viewing Grid Output Once the grid has been generated, the user can view the grid using GMV. All display information on the grid, original CAD drawing, and any views of the STL file normals can be viewed by opening the VIEWGRID.gmv file, located in the project directory. For convenience, Barracuda VR offers shortcut buttons for viewing the grid and the original CAD model under the View Output button, located at the top of the Setup Grid Window. This shortcut button provides the user with the following view options: • The View Grid button shows the created grid as opaque cells with grid lines in between. This view of the grid is helpful for understanding the formation of the grid and identifying cells. From this view, the cells can easily be colored by the i, j , or k coordinate which is useful when troubleshooting. Fig. 2.15 (a) shows an example of the display produced by the View Grid button. • The View CAD button shows the original model contained within the STL file. While this view does not depend on the set of grid line, it does depend on the grid generator to write out the file in GMV format. The view of the CAD is by default an opaque but can easily be set to semitransparent for viewing model internals. Fig. 2.15 (b) shows an example of the display produced by the View CAD button. • The View Transparent Model button shows the Barracuda VR grid as semitransparent cell walls without any cell lines. This view is useful for examining how well the grid represents the model geometry as well as the easy viewing of geometry internals. Additionally, since the transparent model view is also used by the shortcut buttons for particle results, discussed in Viewing GMV results, viewing the transparent model can provide the user with a sense of how the model geometry will be represented in the post processing phase. Fig. 2.15 (c) shows an example of the display produced by the View Transparent button. • The Compare Grid to CAD button shows the edges of the grid overlaid on the original CAD model. This view can be useful at locating regions of the grid that require more refinement to capture important features. Fig. 2.15 (d) shows an example of the display produced by the Compare Grid to CAD button.

2.10 Reference Grid The reference grid is a dialog for displaying the STL lines and grid lines that is useful for determining i-j-k coordinates for the specification of boundary conditions, initial conditions, and data output. The dialog, shown in Fig. 2.16, is Setup Grid window without the functionality to adjust the gridlines or generate a grid. As with the similar to the Setup Grid window, the user can determine x-y-z locations and i-j-k coordinates in each of the view panes by placing the mouse cursor in the desired location. The coordinates in the plane will be displayed in the boxes at the top of the plane. The display properties for gridlines and STL lines can be adjusted by clicking on the Advanced options which raises a dialog that is identical to the drawing options on the Advanced Grid Options dialog, discussed in Advanced

36



Fig. 2.15: Shortcut buttons for viewing grid in GMV: (a) View Grid shows grid as opaque cells with edges, (b) View CAD shows original model geometry, (c) View Transparent Model shows grid as semitransparent cells, and (d) Compare Grid to CAD shows the cell edges and original CAD geometry together

2.10. Reference Grid

37


Options. Note: Changes made to the display in the Advanced options settings of the reference grid will also apply to the main setup grid display.

Fig. 2.16: Reference grid window

2.11 ScaleSTL utility The ScaleSTL utility can perform basic transformations on existing STL files: scale, translate, and rotate. The dialog, shown in Fig. 2.17, is accessed through the Setup menu in the Menu bar . Once the desired transformations have been specified, the transformed file can be created by clicking the Run button. Selecting STL files The STL file to be transformed can be selected by clicking the ... button to the right of the Input STL text box. Once the transformations have been completed, a new STL file will be created with the file name listed in the Output STL textbox. The default output file name is output.stl. Scaling and translation Each point in the STL file can be modified by in the x-, y-, and z-directions by adjusting values for Scale, Offset Before, and Offset After . “Offset before” values are added to the x, y, or z value prior to scaling whereas “Offset after” values are added after scaling is performed. Rotation The entire geometry can be rotated about the x, y, or z axes by selecting Rotate about X , Rotate about Y , or Rotate about Z and entering a rotation angle in degrees.

38



Fig. 2.17: Scale STL dialog

2.11. ScaleSTL utility

39


Setting lengths The total length in the x-, y-, or z-direction can be set by selecting the Set X Length, Set Y Length, or Set Z Length values. The length is the difference between the maximum and minimum coordinate values in each direction. Setting origin The coordinates can be all translated in the x-, y-, or z-direction so that the minimum value is located at 0 by selecting the Set X origin to zero , Set Y origin to zero , or Set Z origin to zero . Matching lengths The length in the x-, y-, or z-direction can be set to match the length in another direction by selecting the appropriate Set [XYZ] length to [XYZ] length check box.

40


CHAPTER

THREE

GLOBAL SETTINGS

Global Settings, parameters which apply to the entire model are specified. These include the gravity vector, the In global thermal settings, and the manner in which chemistry is initialized.

3.1 The Global Settings Window The Global Settings Window, shown in Fig. 3.1, is used to specify gravity, thermal, and chemistry initialization parameters for the model. The parameters set in this window apply to the whole model.

Gravity The x, y , and z components of the gravity vector must be entered for every model as these values determine the magnitude of the gravitational force in the model as well as the geometry orientation. The typical practice is to specify the gravity as being -9.8 m 2 /s in the z direction for a z -oriented geometry.

Thermal settings A model can be set up to be either isothermal, in which the temperature is assumed to be constant throughout, or thermal, in which heat transfer and energy balance equations are calculated in addition to the particle-fluid dynamics. Isothermal flow model Selecting an Isothermal flow for the model assumes constant temperature for all fluids and particles in the system and allows a simulation to run faster because heat transfer equations will not be solved. If an isothermal model is selected, users must enter the isothermal flow temperature in the isothermal flow text box. In Barracuda, the default isothermal temperature is 300 K. Thermal flow model If Thermal flow is selected, Barracuda VR will calculate temperature gradients within the model due to initial particle and fluid temperatures, boundary condition temperatures, thermal walls, or chemical reactions. If a thermal model is selected, the user must also specify in appropriate areas: • thermal properties of all base materials • heat transfer coefficients • initial fluid temperatures in the model • initial particle temperatures in the model • boundary condition temperatures The heat transfer coefficients are specified by clicking the Heat transfer coefficients button which will raise the Heat Transfer Coefficients dialog, discussed in Heat Transfer Models . Starting a model isothermally When a thermal model is selected, users have the option of completely specifying a thermal model by starting the simulation isothermally. This is done by selecting “Start with Thermal Off” and entering a starting temperature. The model can be switched to a thermal model upon a later model restart (see Restarting a Simulation).

41


Fig. 3.1: Global Settings Window

42

Chapter 3. Global Settings


Temperature warning limits When Thermal flow is selected, the Barracuda VR solver will issue a warning message during the simulation run if the temperature ever goes below the Minimum or above the Maximum values specified by the user. Output minimum and maximum temperatures... If this box is checked, the solver will create a log file named MinMaxTemp.log to record the minimum and maximum temperatures within the system at each time-step of the simulation.

Chemistry settings The chemical reactions in a model can be active at the start of the model ( On), inactive at the start of the model ( Off ), or the reactions can start out inactive but slowly ramp up over a set time. On The chemistry is initialized at the start of model and is active throughout the entire simulation. Users will typically start with chemistry active from the start of the model when reactions are included. Off, ramp on The ramp setting provides a delay and ramp functionality that can be beneficial for higher reaction rate models. When this model is used, time values must be entered in the text boxes next to the ramp label. The first text box indicates the time at which the ramp will become active and the reaction rates will start to be non-zero. The second text box value indicates the time at which the full reaction rate values will be used in the rate calculation. In between the two times, the reaction rates will be linearly increasing. Off The chemistry is fully specified in the model but reactions are not calculated at the start. This may be used if the user wants to simulate the model start up without chemistry and turn chemistry on once the model has reached a steady state. The chemistry can later be turned “On” when Restarting a Simulation.

3.2 Heat Transfer Models The Heat Transfer Coefficient dialog, shown in Fig. 3.2, provides an interface for specifying a convective fluid-to-wall heat transfer model and a fluid-to-particle heat transfer model. While the generalized form of each model can be modified, the default values match literature correlations. For specification of radiative heat transfer from a wall see Thermal Wall BCs.

Convective fluid-to-wall heat transfer The local fluid-wall heat transfer coefficient, hfw , is a combination of contributions from a lean gas phase heat transfer coefficient, h l , and a dense particle phases coefficient, h d . The fluid-to-wall heat transfer coefficient is weighted by the function f d which is the fraction of contact time by the dense particle phase. The time fraction of dense phase contact, f d is a function of the particle volume fraction at the wall, θ p , and the close pack value fraction, θcp . 10(θp /θcp )

−

f d = 1 − e

hfw = h l + f d hd

(3.1)

Lean phase heat transfer coefficient The general form of the lean phase heat transfer coefficient is

hl =

�

(c0 RenL1 Prn2

k f + c1 ) + c2 L

�

W m2

·K

(3.2)

where c 0 , c 1 , c 2 , n1 , and n2 are adjustable model parameters, k f is the thermal conductivity of the fluid, and L is the cell length. The Reynolds number and Prandtl number are defined as ReL =

ρf U f L µf

Pr =

µf c p,f kf

(3.3)

where ρ f is the fluid density, U f is the fluid velocity, µ f is the fluid viscosity and c p,f is the fluid heat capacity. The default lean phase heat transfer coefficient is based on the correlation of Douglas and Churchill [Yan03] [p. 267].

c0 = 0.46 3.2. Heat Transfer Models

c1 = 3.66

c2 = 0.0

n1 = 0.5

n2 = 0.33 43


Fig. 3.2: Heat Transfer Coefficient Dialog Dense phase heat transfer coefficient The general form of the dense phase heat transfer coefficient is

hd =

�

c0 Re pn

1

kf d p

�

W m2

·K

(3.4)

where c 0 and n 1 are adjustable model parameters, k f is the thermal conductivity of the fluid, and d p is the particle diameter. The particle Reynolds number is defined as Re p =

ρf U f d p µf

where ρf is the fluid density, U f is the fluid velocity, and µf is the fluid viscosity. The default dense phase heat transfer coefficient values is taken from [Yan03] [p. 262] where

c0 = 0.525

n1 = 0.75

Fluid-to-particle heat transfer Heat transfer between the fluid phase and the particle phase is modeled by the fluid-to-particle heat transfer coefficient. The general form of the lean phase heat transfer coefficient is

h p =

�(

c0 RenL1 Pr0.33

k f + c1 + c2 d p

)

�

W m2

·K

(3.5)

where c 0 , c 1 , c 2 , and n 1 are adjustable model parameters, k f is the thermal conductivity of the fluid, and d p is the particle diameter. The Reynolds number and Prandtl number are defined as Re p =

44

ρf |U f − U p | d p µf

Pr =

µf c p,f kf Chapter 3. Global Settings


where ρ f is the fluid density, U f is the fluid velocity, U p is the particle velocity, µ f is the fluid viscosity, and c p,f is the fluid heat capacity. The particle Nusselt number, Nu p in fluidized beds is typically lower than the Nusselt number for a single sphere when the Reynolds numbers is less than 20. Theoretically, a single sphere in a quiescent fluid will have a value of Nu p = 2.0 which represents the limit of conductive heat transfer. In a fluidized bed, however, the bubbling phenomenon will cause the observed magnitude of N u p to be lower than 2.0. Low Reynolds numbers correspond to beds of fine particles (small d p and U g ), wherein bubbles tend to be clouded with entrained particles. This diminishes the efficiency of particle-gas contact below that represented by idealized plug flow, resulting in reduced values of N u p . As particle diameter increases (coarse particle beds), “bubbles” are relatively cloudless and gas-particle contact improves. To capture the fluid-to-particle heat transfer in a fluidized bed, Barracuda VR uses a correlation for fluid-to-particle heat transfer coefficient based on the correlation proposed by McAdams in 1954 [FZ98]:

c0 = 0.37

c1 = 0.1

c2 = 0.0

n1 = 0.6

This correlation agrees with Turton and Levenspiel’s experimental data on particle-fluid heat transfer coefficient in fluidized beds with small particles [KL91].

3.2. Heat Transfer Models

45


46

Chapter 3. Global Settings

CHAPTER

FOUR

BASE MATERIALS

Base Materials section, users define all materials (gas, liquid, or solid) used in any part of the model as well In the as their thermal and physical properties. Barracuda VR includes a property library of many commonly used gas, solid, and liquid materials which can be used in the model. Additionally, specific materials can be created or edited for the project. The Base Materials Window provides functionality for importing, replacing, creating, editing, copying, and deleting materials in a project.

4.1 The Base Materials Window The base materials window, Fig. 4.1, includes the project material list, material library, and settings for flow type.

Fig. 4.1: Base Materials Window

Properties The averaging method used for fluid mixture properties can be selected as Mole average or Mass average. The choice is applied when calculating viscosity, thermal conductivity, and mass diffusivity for fluid mixtures. The default

47


selection is Mole average , which causes fluid mixture properties to be calculated as:

pmix =

�

(yi pi )

i

where p mix is the mixture property value, yi is the mole fraction of species i , and p i is the property value of fluid species i. If Mass average is selected, the fluid mixture properties are calculated as:

pmix =

�

(xi pi )

i

where p mix is the mixture property value, x i is the mass fraction of species i , and p i is the property value of fluid species i. Flow type The flow type can be set to compressible flow or incompressible flow. Incompressible flow is generally used for liquids only. If multiple fluids are to be used in a simulation, the flow type must be set to Compressible flow and the fluids must be gas phase materials. Multiple liquid species are not allowed to be present simultaneously in the fluid domain in Barracuda VR. Compressible Liquid Equation of State Though liquids are commonly thought of as incompressible, in reality they do have some degree of compressibility. Compressibility is defined as:

β = −

1 ∂V V ∂p

where V is volume and p is pressure. Compressibility depends on temperature and pressure, but under normal conditions the change of compressibility is small. Barracuda VR assumes constant compressibility and small changes of liquid density due to compressibility. The equation of state used to describe the liquid density is:

ρ = ρ ref [1 + β ( p − pref )] where:

ρ = density of the liquid at pressure p (kg/m3 ) ρref = reference density, specified as the fluid density in the material property definition of the liquid (kg/m3 ) β = compressibility factor for the liquid (1/Pa) p = fluid pressure (Pa) pref = reference pressure corresponding to ρref (Pa) With β = 0, the liquid is fully incompressible. To use a compressible liquid equation of state, specify a non-zero β value. For a system with liquid and solid particles, a reasonably small β value can increase the stability of the simulation. A typical value of compressibility for liquids is in the range of 1e-11 to 1e-09 (1/Pa). When using liquid compressibility, it is recommended that the reference pressure, pref , be set equal to the initial condition pressure in the system.

Importing materials from the library To import a material from the Material library, click on the desired material in the right-side pane to select it, then click on the Import button, see Fig. 4.1. Materials in the library are organized alphabetically. Users can change the sort order by clicking on the column headers. Materials can easily be located in the database by matching an entered search string to the name, state, or description of a material in the library. The Material library is stored in a text file named cpfd_prop.prp in the cpfdHQ/props directory. The full path of the properties file is shown below the material library list. The references for the sources of the material properties are listed in the header section of this file.

48

Chapter 4. Base Materials


Tip Several materials have multiple entries in the library corresponding to different states of matter or different sources for material property data. Be sure to select the correct version of the material you wish to add to the project.

Replacing materials A material in the Project Material List can be replaced with one from the library. From the base materials window, Fig. 4.1, users can replace a material by selecting it in the project material list, then selecting the desired new material from the material library and clicking on the Replace button. The Barracuda VR GUI will propagate this change throughout the project file, to any initial or boundary conditions referencing the original material.

Adding a new material A new material can be created using the Add button in the base materials window. Users must enter a name for the new material and must also specify its physical and thermal properties. When a new material is created, the material properties dialog, Fig. 4.2, is automatically raised. Details on the material properties dialog are discussed in Specifying Base Materials .

Editing material properties The material properties of any material in the Project Material List can be manually edited by the user. To modify the properties of a material, users must first select the material from the project material list by clicking on it, then use the Edit button to raise the material properties dialog (See Specifying Base Materials for details). Clicking Apply saves the changes and closes the dialog.

Copying materials A material may be duplicated by selecting it in the Project Material List and clicking on the Copy button. This allows users to subsequently edit the material’s name and properties by clicking on the Edit button. Note that the chemical name of the material must be distinct from any materials already in the Project Material List .

Deleting a material A material can be removed from the Project Material List by selecting the material and clicking on the Delete button. Note that a material cannot be deleted if it is being used in any boundary conditions, initial conditions, or chemical reactions. The GUI will display a warning if the user tries to delete a material that is in use.

4.2 Specifying Base Materials The material properties dialog, shown in Fig. 4.2, is used to view, edit, and add material properties to a Barracuda VR model. All material properties are either a single value or are calculated from a fourth-order polynomial expression with user-provided constants. Utilities for evaluating and plotting property data is provided within the dialog to allow the user to verify the accuracy of the property data function being used.

Material properties Each base material requires a unique chemical name and state of matter to be specified. Other properties such as density, heat capacity, viscosity, etc, may be required depending on the material type, reactions,and thermal nature of the simulation. Property expressions are calculated as 4th-order polynomials of the form: x = a 0 + a 1 T + a2 T 2 + a3 T 3 + a4 T 4 , where T has units of K. The coefficients a 0 through a4 can be edited by clicking the ax button. Chemical name The chemical name is used to identify a material used in particles, initial conditions, boundary conditions, and chemical reactions. Therefore, a chemical name is required for all materials and must be distinct from any other materials used in the project. Do not use special characters, such as colons, slashes, or percent signs, in chemical names. Some output files use the chemical name directly in their file names, and special characters can cause problems depending on your file system.

4.2. Specifying Base Materials

49


Fig. 4.2: Material Properties Dialog

50



Description Additional notes about the base material can be entered in the description field. Possible uses would be to document the source of the material property data or the use of the material in the model. This field is optional. State (gas, liquid, or solid) The material state specifies the state of matter for the material and is required for all materials. Density The units of density are kg /m3 for all phases (gas, solid, and liquid). For gas materials, the ideal gas law is used to calculate the material density based on the molecular weight specified by the user. Computational cell temperature and pressure are used, such that the gas density will generally vary spatially throughout a system. Based on these parameters, the gas density, ρGas , is calculated as:

ρGas =

kg p M W RT 1000 g

where:

ρGas is the gas density

�� kg m3

p is absolute pressure (Pa) R is the universal gas constant, 8.31446 T is temperature (K) g MW is molecular weight mol

( )

J mol·K

( )

For solid materials, the density must be specified by the user, and must be greater than zero. Particles in Barracuda VR are assumed to be hard spheres, and no porosity is taken into account. Therefore, the density specified in the Base Materials dialog should be the “envelope density” of the material. If you were to wrap a particle in a zero-thickness envelope, such that the wrapped particle is a sphere with no porosity, the density you need to specify in Barracuda VR is the mass of that sphere divided by its volume. Note that this density does not take into account interstitial spaces between a group of packed particles. Thus it is not to be confused with the “bulk density” of a particle species. If you know the bulk density, ρ Bulk , and close-pack volume fraction, θ CP , of a particle species, you can calculate its base material density, ρ, as:

ρ =

ρBulk θCP

For liquid materials, density must be specified, and be greater than zero. If a liquid will be used as a component of a particle species, it will be treated in a similar manner to solid components with respect to contributing to the overall density of the particle species. Molecular weight The molecular weight must always be specified for a gas or any material participating in a chemical reaction. For a non-reacting solid or liquid, specification of the molecular weight is not required. Units of molecular weight are g /mol. Heat of formation The heat of formation is required for any reacting materials in a thermal calculation. Note that the input units for heat of formation are J /kg, while many tabulated values in reference sources may be in J /mol or kJ/mol. Thermal conductivity The thermal conductivity is required for all materials in a thermal calculation. The temperaturedependence of the thermal conductivity is specified by a fourth-order polynomial. The thermal conductivity units are in W /(m · K). Viscosity The viscosity is required for all gas and liquid materials. The temperature-dependence of the viscosity is specified by a fourth-order polynomial. Units of viscosity are kg /(m · s). Mass diffusivity The mass diffusivity can be specified for gases in cases where the mixing is being studied in the model. In many cases, however, the gas mixing is dominated by convection in a fluidized system. The temperaturedependence of the mass diffusivity is specified by a fourth-order polynomial. Units of mass diffusivity are m 2 /s. Specific heat The specific heat is required for all thermal calculations. The specific heat is specified using two fourthorder polynomials (low and high) split by the Low/high break temperature. Units of specific heat are in J /(kg · K).

4.2. Specifying Base Materials

51


Low/high break temperature for specific heat The Low/high break temperature specifies the temperature at which the specific heat calculation switches from the low temperature polynomial and the high temperature polynomial. The high temperature polynomial is used when the temperature is above the low/high break temperature and the low temperature polynomial is used when the temperature is below the low/high break temperature. It is important that the low temperature and high temperature curves have the same values at the low/high break temperature. It is recommended that simulations not be run near the low/high break temperature, as discontinuities in specific heat can cause solver instabilities. Units of the low/high break temperature are in K.

Validating expressions Property values that are calculated by a polynomial expression can be evaluated at the Validation Temperature and Validation Pressure by clicking the Validate button. The validation test values for gas density, thermal conductivity, viscosity, mass diffusivity, and specific heat will be calculated. Furthermore, the integrated value of heat capacity from the reference temperature, 298.15 K, to the validation temperature is displayed at the bottom of the Material Properties Dialog.

Plotting expressions The values of polynomial expressions can be plotted automatically by clicking the Plot to the right of the polynomial of the expression. This will produce a plot showing the property values over a range of temperatures from Tmin to Tmax, spaced every ∆T degrees. The plotting limits are specified at near the top of the material properties dialog.

52


CHAPTER

FIVE

PARTICLES

In the Particles section, all information about particles to be simulated is specified. This information includes particle size, solid, liquid, and volatile material components, particle packing, and the models for particle drag, collisions, and interactions. The following sections give details on Particle Interactions, management of Drag Models, specification of Volatiles, and Particle Species. Particle species Particle Species are groups of particles that enter the simulation domain with identical composition, density, and particle size distribution. Each particle species has a unique ID number which is used to define particles being initialized within a system (see Particle ICs) or being defined as a particle feed (see Pressure BCs, Flow BCs, and Injection BCs). The composition and density of a particle only applies to the initial state of the particle and may change due to chemical reaction or the release of volatile components during simulation. Component distribution In Barracuda VR, particles can be composed of multiple solid and/or volatile materials, allowing sophisticated chemical reactions or volatile gas releases to be included in simulations of fluid-particle systems. All materials within a particle are assumed to be uniformly distributed and all materials within a particle are assumed to be available for chemical reactions without any mass transfer hindrances. Volatiles Volatiles are groups of trapped gaseous materials or chemically-bound gaseous materials that are contained within a particle species. During simulation, the volatiles are released at a defined rate which affects the particle density and the composition and volume of the surrounding gas. Particle size and shape When a computational particle is initialized in a Barracuda VR simulation, the particle is assigned a particle radius using a Monte-Carlo method. The method uses a random number to determine the radius of the initialized particle based upon the particle size distribution (PSD) defined for the species. As a result, it is important to provide a sufficient quantity of computational particles in the model to accurately represent the defined particle size distribution. All particles are assumed to be spherical and therefore calculations of particle mass and volume are based upon the particle radius. The particle radius is related to particle volume by the equation for the volume of a sphere.

rparticle =

(

1/3 3 4π V particle

)

(5.1)

Although particles are assumed to be spherical in shape, a non-zero sphericity can be defined for a particle species which is used in the calculation of non-spherical drag models (see Drag Models) and surface area dependent chemical reactions (see Rate Coefficients). In Barracuda VR, the sphericity is defined as

Ψ=

Surface area to volume ratio of a sphere . Surface area to volume ratio of the particle

(5.2)

Particle density The interaction between a particle and fluid is strongly dependent on the particle density. There are three definitions of particle density commonly used: skeletal density, particle density, and bulk density. • Skeletal density - The skeletal density of a particle is the mass of a particle divided by the volume of solids within the particle. Gas volume inside a porous particle is not included in the calculation of the skeletal density • Particle density - The particle density is the mass of a particle divided by the volume of the particle. For a porous particle, the volume of the particle includes both the volume of solids within a particle as well as the volume of gas space inside the particle. In Barracuda VR, the particle density is used when defining particles

53


• Bulk density - The bulk density of a particle is defined as the mass of particles per total volume of occupied space. This total volume includes both the volume of particles and the volume of gas in the interstitial particle spaces For multimaterial particles in Barracuda VR, the particle density is a function of the types and quantities of material that constitute the particle. During simulation, the particle composition of the particle may change due to chemical reaction and the release of volatile components. Therefore, the particle density is not a constant during simulation. For each particle, the density is calculated based upon the total mass of the particle, m particle, and the volume of the particle, V particle ,

mparticle =

�

msolid,i +

i

�

mvolatile,i

V particle =

i

�

vsolid,i +

i

�

vvolatile,i

(5.3)

i

where msolid,i is the mass of a solid component, m volatile,i is the mass of a volatile component, v solid,i is the volume of a solid component, and vvolatile,i is the volume of a volatile component. The particle density is calculated as:

ρparticle =

mparticle . V particle

(5.4)

During simulation, the volume of a solid component is recalculated based on the solid material density, but the volume of a volatile component in a particle remains unchanged, even once all of the initial volatile gas has been released.

msolid,i vsolid,i = ρsolid,i

vvolatile,i = v volatile,i

⃒⃒

init

mvolatile,i = ρvolatile,i

⃒⃒

(5.5)

init

During particle initialization, if all materials in the particle have a defined density, then the overall particle density is calculated directly. In the case where a volatiles component with an undefined density is included in the particle, specification of a total particle density is required. Based on this total particle density, Barracuda VR calculates a volatiles volume. As a result of this special treatment of the volume of volatiles in a particle, changes to the particle size or density depend on whether the change in mass is due to a chemical reaction or the release of volatiles. When a particle gains or loses solid mass due to chemistry, the volume and particle radius increases or decreases accordingly. However, when a particle releases volatiles, it decreases in mass but the volume of the volatile component remains in the particle. Therefore, the radius of the particle remains unchanged.

5.1 Particle Interactions In the Particles window, shown in Fig. 5.1, the particle interaction parameters applying to all particle species in the model are specified. Detailed information on each particle interaction field is described below.

Particle-to-particle interaction parameters In the CPFD method, particle interactions are modeled through the use of a computationally efficient particle stress function. The stress function, τ , is a function of the particle volume fraction, θ p , and is given by [Sni01] as

10P s θ p β τ (θ p ) = max[θcp − θ p , ε (1 − θ p )]

(5.6)

where P s is a constant with units of pressure, θ cp is the particle volume fraction at close pack, β is a constant with a recommended value between 2 and 5, and ε is a very small number. Values for the stress model constants, including close pack volume fraction are specified in the Particles window and the Stress Model dialog box, shown in Fig. 5.2. A particle stress model with recommended values for the model constants will have a negligible effect on the particle behavior in the dilute phase. As the particle volume fraction of a region begins to approach close pack, the stress model will begin to affect particles in or traveling towards the region. Ideally, the stress model should prevent particles

54

Chapter 5. Particles


Fig. 5.1: Particles window showing typical values for close pack volume fraction and particle-to-wall interactions

5.1. Particle Interactions

55


from entering a region with a particle volume fraction that is at close-pack. In this event, a particle that moves toward close pack region will be redirected. The nature of the particle redirection is specified by the maximum momentum redirection from collision parameter discussed below. Close pack volume fraction The close pack volume fraction specifies the maximum volume fraction of particles when they are packed randomly. In practice, the close pack volume fraction, θ cp , is often estimated from the bulk density and particle density of the particles to be modeled.

θcp =

Bulk Density Particle Density

Additional information on the bulk density and particle density is provided at the start of this chapter. Typical values for close pack volume fraction range from 0.56 to 0.64. Maximum momentum redirection from collision A particle approaching a region of close-pack will be redirected in a random manner based on the particle stress tensor and particle incidence angle and a maximum redirection value entered in the Maximum momentum redirection from collision box. A value of 40 % is the default. Blended acceleration model for contact force In closely packed, polydisperse granular beds, relative motion between particles of different sizes and densities is inhibited by sustained particle contacts, and this effect is simulated by the Blended Acceleration Model. The model aims to improve the predicted fluidization behavior of polydisperse particles of differing sizes or densities. Without BAM, fluidization tends to exhibit a higher degree of particle segregation than actually occurs. With this model, individual particle accelerations are a blend between the particle acceleration of the original MPPIC method, appropriate for rapid granular flows, and an average particle acceleration that applies to closely packed granular flows. As a result, particles at or near close-pack tend to move together with velocities close to the averaged velocity due to enduring particle-particle contacts. In dilute regions, particles tend to move independently of each other due to less contacts with surrounding particles. The development of this model is discussed in detail by [ORourkeS14]. Stress model advanced options Clicking on the Stress model advanced options button in the Particles window will raise the Stress Model dialog box, show in Fig. 5.2. In the Stress Model dialog box, the parameters of the particle stress model, (5.6), can be modified. Default values for the stress model are Ps (P s ) equals 1 (in Pascals), B ( β ) equals 3, and Eps ( ε) = 1E-8. The Fraction of average velocity can also be adjusted from the Stress Model dialog box. The fraction of average velocity specifies the frame of reference for close-pack cells in the particle stress tensor. When particles are moving inside a cell that is at or near a close-pack volume fraction, the frame of reference will determine whether an approaching particle can enter the cell. Generally, the particle normal stress will prevent an additional particle from entering the cell if it is full. When the densely-packed particles in the cell are moving, however, then this might be possible due to the space being vacated by the moving particles. This behavior is defined by the fraction of average velocity: 0 indicates the frame of reference is the walls (non-moving) whereas 1 indicates that the average particle velocity of the particles in the cell is the frame of reference. In Barracuda VR 16, the default value is 0.8 which is a change from the default value in previous releases of 0.5. This change was made as a result of improvements to the solver algorithm.

Fig. 5.2: Stress Model dialog box with default values

56



Particle-to-wall interaction parameters There are three variables users can control regarding the physics of a particle bouncing off a wall. All three are found on the Particles window of the Barracuda VR GUI, under Particle-to-wall interaction. Each is explained below: Normal-to-wall momentum retention This is the fraction of the normal component of the particle momentum that is retained by the particle after collision with the wall. A value of 1.0 indicates a perfectly elastic collision, where the particle will bounce back with the same velocity it had before collision, when hitting exactly normal to the wall. A value of 0.0 indicates all velocity is lost. A value of 0.3 indicates 30% of the momentum (velocity since the mass doesn’t change) is retained but it should be noted that this is different from the distance the particle will bounce off the wall under gravity. Retaining 30% of the velocity means the particle only retains 9% of the kinetic energy since kinetic energy varies as with the square of velocity. A typical value for the normal-to-wall momentum retention is 0.3 for “soft” particles while 0.9 or higher may be used for “hard” particles. Tangent-to-wall momentum retention This is the fraction of the tangential component of the particle momentum that is retained by the particle after collision with the wall. A typical value for the tangent-to-wall momentum retention is 0.99.

Fig. 5.3: Momentum retention model Fig. 5.3 shows a diagram that is useful for understanding the mathematical implementation of the normal-to-wall and tangent-to-wall momentum retention parameters. The particle velocity, following a wall impact, is described by normal and tangential momentum losses and the angle of impact as:

|un+1 | = [(rT − rN )(1 − cos θ) + rN ] |un | where |un+1 | is the new particle speed, and |un | is the old particle speed. For a normal wall impact, the momentum retention factor is rN , and for a shallow (tangential) wall impact, the momentum retention factor is rT . Typically when a particle hits a wall at a shallow angle ( θ > 85 degrees), the particle “skips” with little momentum loss. The normal component of momentum retention is usually smaller (greater momentum loss) than the tangent momentum retention. Diffuse bounce The diffuse bounce index applies a scatter function to particles after collision with the wall. A value of 0 means that the new velocity vector for the particles after collision with the wall is computed (using the normal and tangential momentum retention coefficients) with no scatter. Thus, if 100 particles hit the wall with the same velocity vector, then they will all have an identical velocity vector to each other after the bounce. However, in the real world, particles are not perfectly round, and walls are not perfectly smooth. Thus, in actuality one would expect an average vector after the bounce to be computed, but each individual particle may have some deviation from the average vector. The diffuse bounce setting induces such a deviation from the average. The higher the number, the more scatter is applied, and the index value specified is a linear multiplier on the amount of scatter. Although the default value is 0 (no scatter) to minimize the computational cost, a higher value is often appropriate for common industrial surfaces. A stochastic diffuse bounce model is used, and the variation from a specular bounce is specified by a pseudo wallroughness coefficient which ranges from 0 (perfectly smooth) to 5 (maximum roughness). This model gives a max-

5.1. Particle Interactions

57


Fig. 5.4: Diffuse bounce model

imum solid angle variation from a specular bounce of 15.5 degrees. The angle variation is maximum for a normalto-wall impact. Particles which hit a wall at a shallow angle have a near specular bounce regardless of the roughness coefficient. The magnitude of the particle velocity is not changed by the diffuse bounce. Only the direction of the particle reflection is affected by the diffuse-bounce model.

5.2 Drag Models The force acting on a particle by the flow of fluid around it is determined by the drag model of the particle. Each particle species that is defined in Barracuda VR has a drag model assigned to it which may be selected from either system models that are predefined in Barracuda VR or user-defined models which are created. The Drag Models window, shown in Fig. 5.5, provides an interface for creating and managing the drag models in a project. Models in the Drag Model window will be available for application to a particle species in the Particle Species dialog box, as discussed in Specifying particle species.

Managing drag models The Drag Model window lists all system drag models and user-defined drag models. Each line in the list indicates the name, description, and the model source (system or user-defined). The system drag models are shown for all projects. Adding A user-defined drag model is created by clicking Add which displays an empty Drag Model Editor dialog box. The Drag Model Editor dialog box ( Creating a drag model ) is used to specify the new drag model. Editing An existing user-defined model is edited by selecting the drag model from the list and clicking Edit which displays the Drag Model Editor dialog box ( Creating a drag model). The Edit button is only available when a userdefined model is selected. Copying A drag model is copied by selecting the drag model from the list and clicking Copy. The two system drag models that cannot be copied are Constant and Stokes. I f a user-defined model is copied, a duplicate user-defined model is created. However, if a system drag model is copied, a new user-defined model will be created containing the system model in the force equation. When creating a new user-defined drag model, it is often convenient to start from a copy of system model since many drag models contain similar terms. Deleting A user-defined drag model is deleted by selecting the model from the list and clicking Delete. This action cannot be undone. The system drag models cannot be deleted. Viewing Information about a system drag model is viewed by selecting the system drag model from the list and clicking View. The View button is only available when a system model is selected.

58



Fig. 5.5: Drag Models window showing both system and user-defined drag models

5.2. Drag Models

59


5.2.1 Creating a drag model A user-defined drag model is specified in the Drag Model Editor dialog box, shown in Fig. 5.6, which is displayed when a drag model is added or edited from the Drag Models window. A drag model Name and an optional Comment are used to identify the drag model. The model drag model is defined in a dimensionless form in the Drag Model Definition box using the avaialable variables, constants, functions, and operators. These components can added using the Model Tools lists or can be typed directly into the definition box.

Fig. 5.6: Drag Model Editor dialog box showing a user-defined drag model

Dimensionless drag function The drag function that is specified, f custom (), is a dimensionless drag function normalized by the Stokes drag law. The use of a dimensionless drag model eliminates the need to consider units and simplifies the drag models since most drag models depend only on the particle Reynolds number and the fluid or particle volume fraction in the dimensionless form. In the Drag model reference , examples of conversion to dimensionless drag are shown. Fdrag = F Stokes f custom ()

60

FStokes =

3πµf d p (uf − us ) Chapter 5. Particles


Variables A particle drag model is typically a function of the particle properties and the properties of the particle’s surroundings. For many drag models, these properties are limited to the particle Reynolds number, Re, the fluid volume fraction, volfracF, and the particle volume fraction, volfracP. More complicated models may also consider the particle sphericity, sphericityP, or the ratio of the particle diameter to the surrounding sauter mean diameter, diamP/diamSauterP. A full list of available variables is displayed in the Variables list along with all defined constants once the button has been clicked. Variables selected from the list will be added at the cursor in the Drag Model Definition box. Alternatively, the variable can be typed directly into the drag model definition. The following is a full list of available variables. Table 5.1: Variables available in drag models densityF

ρf

densityP

ρ p

diamP

d p

diamSauterP

dVelPF

Re

sphericityP

|u p − uf |

Magnitude of relative particle velocity (m/s)

frac|u p − uf |d p ρf µf

Particle Reynolds number (dimensionless)

µf

volfracF

volfracP

Particle diameter (m) Sauter mean diameter of particles in cell (m)

θcp

viscF

Particle density (kg/m 3 )

d32

ψ

thetaCP

Fluid density in cell (kg/m 3 )

Particle sphericity (dimensionless) Particle volume fraction at close pack (dimensionless) Fluid viscosity (Pa · s)

θf

Fluid volume fraction (dimensionless)

θ p

Particle volume fraction (dimensionless)

The Sauter mean diameter of particles in a computational cell is calculated as:

d32 = 6

Constants

∑∑

Total volume of particles in a cell V p =6 Total surface area of particles in a cell A p

The Constants box is a location where constant values can be defined in the model. The use of constants is not required since the values can be entered directly in the drag model expression. However, defining constants has advantages. In addition to often making the drag model function more readable, the defined constant values can be changed for individual particles species to make the drag model more adaptable. This is discussed further in Particle Species. All defined constants are added to the Variables list when the button is clicked. Constant names A constant name can consist of any characters or numbers except for operators or spaces. Addition-

5.2. Drag Models

61


ally, the constant name must start with a character (a-z,A-Z) and the name cannot be identical to a predefined function or variable. Number formats Numbers can be typed as integers, decimal values, or in scientific notation. Input modes By default, constants are entered in a table-based format, as shown in Fig. 5.6. The constants box can be switched to a text box entry format by clicking Switch to Text Entry . This format, shown in Fig. 5.7, is often preferred for copying constants between drag models or project files. When in text-entry mode, the input mode can be changed back to table-based mode by clicking Switch to Table Entry. All constant values are preserved when switching between input modes.

Fig. 5.7: Text-based entry of constants in Drag Model Editor dialog box

Functions The Functions list contains the functions available for use in the drag model definition. Each function has a defined number of arguments (noted as val1, val2, etc.) which can be expressions or single variables. The following functions are available for use in the drag model definition: • ABS(val1) - Returns the absolute value of val1. Usage: ABS(densityP - densityF)

• COS(val1) and SIN(val1) - Returns the cosine or sine of val1 in radians. Usage: COS(sphericityP*2*3.14159) SIN((1-sphericityP)*2*3.14159)

• EXP(val1) and LN(val1) - Returns the exponential or natural logarithm of val1. Usage: EXP(diamP/diamSauterP) LN(Re+1)

• LOG(val1) - Returns the logarithm to base 10 ( log10 ) of val1. Usage: LOG(1+dVelPF)

• MAX(val1,val2) and MIN(val1,val2) - Returns the maximum or minimum of the two arguments ( val1 and val2). Usage:

62



MAX(1e-4, Re*volfracF) MIN(volfracP, thetaCP)

• IF(condExpr,valTrue,valFalse) - Evaluates the conditional expression and if true, valTrue is returned. If the conditional expression is false, valFalse is returned. Usage: IF(1-volfracF < thetaCP, volfracP, thetaCP) IF(Re < 1000, (1+0.15*Re)^0.687, 0.44*Re/24)

Operators Arithmetic operators Standard arithmetic operators are available for use in evaluation expressions including • Addition: + • Subtraction: • Multiplication: * • Division: / • Exponent: ^ Logical operators Standard logical operators are avaiable for use in conditional expressions including • Equal to: == • Less than: < • Greater than: > • Less than or equal to: <= • Greater than or equal to: >= • And: && • Or: || Grouping operators The left and right paranthesis, ( ), are available for specifying order of operations. Other brackets such as square brackets, “[ ]”, or braces, “{ }”, are not recognized. Order of operations Standard order of operations conventions are followed: 1. Grouping: ( ) 2. Exponent: ^ 3. Multiply and Divide: * / 4. Addition and Subtraction: + 5. Comparisons: == < > <= >= 6. Logical And/Or: && ||

5.2.2 Checking a drag model The drag model expression can be checked to ensure that the string can be evaluated (no formatting errors) and that the evaluation of the string produces the expected result. Verifying drag model expression A drag model is verified by clicking Check Model . Barracuda VR will inspect the typed expression and highlight any issues including mismatched parentheses, unrecognized variables or function names, or incorrect number or arguments in a function. The location of of any error will be highlighted in the Drag Model Definition box.

5.2. Drag Models

63


Validating drag model evaluation When the model is run, a drag.log will be created in the project directory. Within this file, each drag model in the simulation will be evaluated at a variety of conditions. These values can be used to validate the entered expression. Sample contents of drag.log are shown. # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # #

Non-dimensional drag coefficients, Cd = D/D(Stokes), at some sample points for drag models The following constant quantities are used. pvf(CP) = 5.300000e-01 Close pack volume fraction rhoF = 1.454508e+00 Fluid density (kg/m^3) visc = 1.348299e-05 Fluid viscosity (kg/s*m) dVel = 1.000000e+00 Magnitude of relative particle-to-fluid velocity (m/s) 1 2 3 4 5 6 7 8 9 10 11 12 13

Re pvf/pvf(CP) fvf Cdp1 Cdp2 Cdp3 Cdp4 Cdp5 Cdp6 Cdp7 Cdp8 Cdp9 Cd1

Reynolds number based on particle diameter Ratio of particle volume fraction to close pack volume fraction Fluid volume fraction Predefined drag model - Wen-Yu Predefined drag model - Ergun Predefined drag model - WenYu-Ergun Predefined drag model - Turton-Levenspiel Predefined drag model - Richardson-Davidson-Harrison Predefined drag model - Haider-Levenspiel Predefined drag model - EMMS1 Predefined drag model - Nonspherical-Ganser Predefined drag model - Nonspherical-Haider-Levenspiel User defined drag model - Beetstra et al. 2007 Model

Species 1 Additional parameters are used for this species diam = 2.000000e-04 Particle average diameter (m) diamSauter = 2.000000e-04 Particle Sauter mean diameter (m) = diam sphericity = 1.000000e+00 Particle sphericity density = 1.570000e+03 Particle density (kg/m^3) Calculated D(Stokes) = 1.063980e+01 (1/s) Re - - 1e-50 1e-50 1e-50 . . . 2000 2000 2000

pvf/pvf(CP) fvf Cdp1 - - - - - - - - - - - - - - 0 1 1.000000e+00 0.01 0.9947 1.014182e+00 0.1 0.947 1.155241e+00 . . . . . . . . . 0.9 0.523 2.042876e+02 0.99 0.4753 2.632126e+02 1 0.47 2.711516e+02

... ... ... ... ... . . . ... ... ...

Cd1 - - - - - 1.000000e+00 1.144941e+00 1.702218e+00 . . . 1.275539e+02 1.369037e+02 1.380281e+02

5.2.3 Drag model reference The drag model in Barracuda determines the force acting on a particle, F p , by a fluid in the model. Users may choose from a variety of different models listed in the Model Name drop-down menu in the Drag model section shown in Fig. 5.8. A description of the various drag models available is shown below. The following drag models are available for use in Barracuda VR: • Constant Drag • Stokes

64



Fig. 5.8: Drag Model Selection dialog box

5.2. Drag Models

65


• Wen-Yu • Ergun • Wen-Yu Ergun • Turton-Levenspiel • Richardson, Davidson, and Harrison • Haider-Levenspiel • EMMS-Yang-2004 • Non-spherical Ganser • Non-spherical Haider-Levenspiel All drag models calculate a force acting on a particle, F p as a function of the fluid and particle properties and flow conditions. For all models below, the force on the particle is a function of the mass of the particle m p , fluid velocity uf , the particle velocity u p , and the drag function D. F p = m p D (uf − u p )

(5.7)

In many of the models the drag function is dependent of the fluid conditions, the drag coefficient C d , and the Reynolds number Re. For purposes of calculating particle drag, the Reynolds number is calculated as

Re =

2ρf r p |uf − u p| µf

(5.8)

where ρ f is the fluid density, r p is the particle radius, and µf is the fluid viscosity. In many models, the drag function D is related to the to the drag coefficient by:

D =

3 ρf |uf − u p | C d 8 ρ p r p

(5.9)

Constant drag The constant drag model calculates the force on the particle using ( 5.7). The drag function D is specified by the Constant value entered in the Drag Model Selection dialog box. Stokes drag The Stokes drag model is based upon an analytical calculation for the drag force acting on a single particle at creeping flow - typically Re < 0.1 [Whi91]. The Stokes drag is written as F p = 6πµf r p (uf − u p ) which is equivalent to (5.7) and (5.9) when

C d =

24 Re

.

(5.10)

Alternatively, the drag function D can be expressed directly as

D =

9 µf . 2 r p 2 ρ p

(5.11)

The drag force can be adjusted by the value of the Multiplier specified in the Drag Model Selection dialog box. The default multiplier value is 1.0. Wen-Yu drag model In the Wen-Yu model [WY66][PPC93], the particle force and drag function are calculated by ( 5.7) and (5.9). The drag coefficient C d is a function of the Reynolds number Re according to the following:

C d =

 

24 24

2.65

−

θ Re f

2.65

−

θ Re f

66

1 + 0.15Re0.687

2.65

−

0.44θf

Re < 0.5

(

)

0.5 ≤ Re ≤ 1000 Re > 1000

(5.12)



The Wen-Yu model in Barracuda VR is based on single particle drag models plus a dependence on the fluid volume fraction θ f to account for the particle packing. The fluid volume fraction multiplier is θ f −2.65 . The Wen-Yu model uses a Stokes drag for a single particle, C d = 24/Re, at low Reynolds numbers, the initial drag coefficient, C d = 0.44, at high Reynolds numbers, the Schiller-Naumann drag, C d = 24/Re 1 + 0.15Re0.687 , in the transition region. The drag force can be adjusted by the value of the Multiplier specified in the Drag Model Selection dialog box. The default multiplier value is 1.0. Ergun drag

(

)

The Ergun Drag model was developed from dense bed data and is therefore only valid for those systems [BvdHK07]. In the Ergun drag model, the particle drag force is calculated by ( 5.7) and the drag function is given by

D = 0.5

�

�

C 1 θ p ρ f |uf − u p | + C 2 θf Re r p ρ p

(5.13)

where C 1 and C 2 are the linear coefficient and non-linear coefficient , respectively. [Erg49] recommended values of C 1 = 150 and C 2 = 1.75 but many different constants have been fit to different data sets. The default Barracuda VR coefficients are C 1 = 180 and C 2 = 2 but can be adjusted depending on the system to be modeled. Additionally, the drag force can be adjusted by the value of the Multiplier specified in the Drag Model Selection dialog box. The default multiplier value is 1.0. Wen-Yu/Ergun blend Since the Wen and Yu correlation [WY66] is appropriate for more dilute systems and the Ergun relationship [Erg49] is appropriate at higher packing fractions, [Gid94] proposed a drag function blending both the Wen-Yu and Ergun functions. In Barracuda VR, the particle force is calculated using ( 5.7) in which the drag function is calculated as

 

D1

D =

θ p < 0.75 θCP

(D2 − D1 ) D2

�

θ p − 0.75 θCP 0.85 θCP − 0.75 θCP

�

+ D1

0.75 θCP ≥ θ p ≥ 0.85 θCP

(5.14)

θ p > 0.85 θCP

where θ p is the particle volume fraction and θ CP is the particle volume fraction at close pack. D1 is the Wen and Yu drag function defined as

D1 =

3 ρf |uf − u p | C d . 8 ρ p r p

(5.15)

The drag function for D 1 is identical to (5.9). The drag coefficient C d is the Wen and Yu drag coefficient defined in (5.12). The Ergun drag function D2 is defined as

D2 = 0.5

�

�

C 1 θ p ρ f |uf − u p | + C 2 θf Re r p ρ p

(5.16)

where C 1 and C 2 are the linear coefficient and non-linear coefficient , respectively. The default Barracuda VR coefficients at C 1 = 180 and C 2 = 2. The Wen-Yu/Ergun drag force can be adjusted by the value of the Multiplier specified in the Drag Model Selection dialog box. The default multiplier value is 1.0. Turton and Levenspiel The Turton and Levenspiel model in Barracuda VR uses the single particle drag function of [TL86] with the dependence on the fluid volume fraction of [WY66]. The Turton and Levenspiel model calculates the drag force on a particle using (5.7) and (5.9). The Turton and Levenspiel drag coefficient is

C d =

24 Re

1 + 0.173Re0.657 θf 2.65 +

(

)

−

0.413 1 + 16300Re

2.65

−

1.09 θf

−

.

(5.17)

The drag force can be adjusted by the value of the Multiplier specified in the Drag Model Selection dialog box. The default multiplier value is 1.0. Richardson, Davidson and Harrison The Richardson, Davidson, and Harrison model in Barracuda VR uses the single particle drag function of [RDH71]. The drag force on a particle is calculated with ( 5.7) and (5.9) where the drag coefficient is calculated as

C d =

5.2. Drag Models

 

24

Re < 0.2

Re

24

0.313

−

+ 3.6Re 0.44 Re

0.2 ≤ Re ≤ 500 Re > 500

(5.18)

67


The drag force can be adjusted by the value of the Multiplier specified in the Drag Model Selection dialog box. The default multiplier value is 1.0. Haider-Levenspiel The Haider-Levenspiel drag model in Barracuda VR is from [HL89].

D =

9 µf f h 2 ρ p r p2

f h = 1 + c0 Ren + 0

(5.19)

c1 Re2 Re + c2

(5.20)

The default coefficient values are: c0 = 0.14017, c 1 = 0.19197, c 2 = 2682.5, and n0 = 0.6529. EMMS-Yang-

2004 The EMMS-Yang-2004 drag model in Barracuda VR is based on [NYWangGe+04] and [LK94]. The EMMS-Yang2004 model constants were generated for the following conditions based on the Li and Kwauk experiment. • Air at atmospheric conditions • 54 micron mono-sized particles • Particle density of 930 kg/m3 • Fluid Superficial Velocity of 1.52 m/s • Solids Flux of 14.3 kg/m2-s Note that EMMS table values for this experiment are also included in the pre-defined multiplier tables for the Fast Fluidized Bed condition.

D =

 �    1 18θf

f e =

θ c0 θf p

+ c1 Re

(c2 + c3 Ren ) ω Re c4 24 ω c5 +

ω =

c9 +

�

0

9 µf f e 2 ρ p r p2

(5.21)

θf ≥ 0.74 and Re < 1000 θf ≥ 0.74 and Re ≥ 1000

(5.22)

θf < 0.74

c6 4(θf +c7 )2 +c8 c10 4(θf +c11 )2 +c12

0.74 ≤ θf < 0.82 0.82 < θf ≤ 0.97

(5.23)

0.97 < θf ≤ 1

c13 + c14 θf

The default model constants are c 0 = 150, c 1 = 1.75, c 2 = 1.0, c 3 = 0.15, c 4 = 0.44, c 5 = − 0.576, c 6 = 0.0214, c7 = 0.7463, c 8 = 0.0044, c 9 = −0.0101, c 10 = 0.0038, c 11 = 0.7789, c 12 = 0.0040, c 13 = −31.8295, c 14 = 32.8295, and n0 = 0.687. Non-spherical Ganser The Non-spherical Ganser model in Barracuda VR uses the single particle non-spherical drag model of Ganser [CAS99] with the dependence on the fluid volume fraction of [WY66]. The Non-spherical Ganser model calculates the drag force on a particle using (5.7) and (5.9). The drag coefficient C d is calculated as

C d = θ f 2.65 K 2 −

�

24 0.4305 0.6567 1 + 0.1118 (ReK 1 K 2 ) + 3305 ReK 1 K 2 1 + Re K K

�

�

1

2

�

(5.24)

where the isometric shape constants K 1 and K 2 are defined as

K 1 =

68

3 1 + 2ψ

K 2 = 101.8148(

−

0.5

−

log ψ)0.5743

(5.25)



and ψ is the particle sphericity. The drag force can be adjusted by the value of the Multiplier specified in the Drag Model Selection dialog box. The default multiplier value is 1.0. Non-spherical Haider-Levenspiel The Non-spherical Haider-Levenspiel model in Barracuda VR uses the single particle non-spherical drag model of Haider and Levenspiel [CAS99] with the dependence on the fluid volume fraction of [WY66]. The Non-spherical Haider-Levenspiel model calculates the drag force on a particle using ( 5.7) and (5.9). The drag coefficient C d is calculated as 2.65

−

C d = θ f

[ � 24

Re

1 + 8.1716 exp(−4.0655ψ)Re

(0.0964+0.5565ψ)

�

73.69 exp(−5.0748ψ)Re + Re + 5.378 exp(6.2122ψ)

]

(5.26)

where ψ is the particle sphericity. The drag force can be adjusted by the value of the Multiplier specified in the Drag Model Selection dialog box. The default multiplier value is 1.0.

5.3 Volatiles In the Volatiles window, Fig. 5.9, users can define the volatile species used to release groups of gases from the particles into the gas phase. For example, if coal particles were part of the system being modeled, a volatile species could be defined to account for the volatile material trapped in the fresh coal particles as well as the gases that would be subsequently released from the particles. The composition of this “Volatiles” species can be specified with respect to the gases which will be released into the gas phase. Additionally, the user can specify the release rate of volatiles in terms of an Arrhenius-type temperature dependence. Volatile species are created in the Volatiles Manager from existing base materials. The Volatiles Manager follows the Add, Edit, Copy, Delete GUI pattern. The composition and properties of a volatile species are entered in the Volatile Material Editor dialog box discussed in Specifying volatile materials.

Fig. 5.9: Volatiles window showing a sample volatile species

Defining volatiles In the Volatiles list, the volatile species currently in the project are displayed as individual lines. Each volatile species in the list has a unique ID number associated with it that is displayed alongside the volatile species name, density, and rate of release.

5.3.1 Specifying volatile materials Volatile species parameters are defined in the Volatile Material Editor dialog box, shown in Fig. 5.10. This dialog box is activated when either a new volatile species is added or an existing volatile species is edited. In the Volatile Material Editor dialog box, users can specify: Name Required for each new volatile species created, and must be distinct from any other materials used in the project.

5.3. Volatiles

69


Fig. 5.10: Volatile Material Editor dialog box showing a sample volatile species definition

70



Specific heat (Cp) The specific heat is the heat capacity of the unreleased volatile material inside the particle. Specify density Here, the user has the option to specify the density of the volatiles in the particles. If the density is empirically known, enter it here. If not, leave it unspecified. A density which isn’t specified, will be considered undefined. Release gases The user must specify which gases are released as the particles devolatilize by clicking on the Release gases button. This raises the Applied Materials dialog box, discussed in Defining volatiles composition, in which the composition of the release gases is specified. Rate of release The release rate of volatiles, k , can be specified in terms of coefficients c 0 , c1 , c 2 , c3 , E , and E 0 . The form of the release rate expression is

k = c 0 T c pc ρcf exp(−E/T + E 0 ) 1

2

3

(5.27)

where T is the temperature in units of K, p is the pressure in Pa, ρ f is the fluid density in kg/m3 , E is the activation energy in K, and E 0 is an additional Arrhenius activation energy term. Note that the temperature used in the calculation of volatile material release rate, is a weighted mix of gas and particle temperatures. The weighting is 80% particle temperature, and 20% gas temperature. The units of the release rate are seconds. The change in volatiles mass mvolatiles due to the release rate is

mvolatiles = −k mvolatiles dt

(5.28)

5.3.2 Defining volatiles composition The Applied materials dialog box, Fig. 5.11, shows which materials make up the volatile species as well as the mass fraction of each material. From this dialog box, users can add, edit, and delete volatile materials using the buttons in the GUI.

Adding, editing, and deleting applied volatile materials New volatile materials can be added using the Add material button, which will raise the Material Selection dialog boc described below. The mass fraction of existing volatile materials can be edited by selecting the material from the list and clicking on the Edit button. A material can be deleted, by selecting it from the list and clicking on the Delete button. Once the desired volatile materials have been added to the list, users should check that their mass fractions add up to 1 and then click on the Apply button, which applies the materials to the volatile species and closes the dialog box.

The material selection dialog box When users add a material to a volatile species, a new dialog box is raised where a list of available materials is displayed, Fig. 5.12. Users select each material individually and specify its Mass fraction amount before clicking Apply. If the desired material isn’t shown in the material selection dialog box, the user may add a material to the list by clicking on the Material Properties Library button and either importing or creating a new material (see The Base Materials Window). As each gas is added, the user must specify the mass fraction amount of that gas relative to composition of the volatiles species and then click on the Apply button. Repeat this process until all the gases that make up the volatile species and their mass fractions have been defined. All the mass fractions must add up to 1.0.

5.4 Particle Species Particle Species window, Fig. 5.13, the composition of single or multi-material particle species is defined. In the From this window, users can Add, Edit, Copy, Delete particle species using the buttons in the GUI.

5.4. Particle Species

71


Fig. 5.11: Applied Materials dialog box showing sample volatile components

Fig. 5.12: Material Selection dialog box showing sample materials available for selection

72



Fig. 5.13: Particle Species window showing a sample particle species

Defining a particle species In the Particle Species list, all applied particle species are displayed as individual lines. Each particle species has a unique ID number that is displayed alongside the particle materials, particle radius, and sphericity.

5.4.1 Specifying particle species A new particle species is defined in the Particle Species dialog box, Fig. 5.14. This dialog box is activated by clicking on the Add button in the Particle Species window.

Fig. 5.14: Particle Species dialog box When defining a new particle species, users may provide the following information:


73


Predefined PSD Here users can select a predefined PSD from a pull down menu. Materials Users may click on the Applied Solids/Volatile Materials button to specify which materials will be used in the particles. See Defining particle composition for details. Species ID This is an internal species ID number that is displayed on the particle species list and is referenced in log files and post processing data. Comment The comment field for a particle species is optional but it is highly recommended that this is used to name a particle species. The contents of a the comment field are used along with the species ID to identify particle species selection for particle ICs ( Specifying a particle IC ), particle injections ( Injection BCs), and particle feeds ( Boundary Conditions). Radius Users can specify a fixed minimum and maximum radius for the particles or use a SFF file to specify a particle size distribution. See Specifying a particle size distribution for more details. Note that particle size is defined by radius and that the units are meters. Sphericity The sphericity of a particle is defined as the ratio of the surface area of a sphere to the surface area of the particle. A typical value for sphericity is 1.0 .

AS π 1/3 (6V P )2/3 ψ = = AP AP 0 < ψ ≤ 1 Where ψ is sphericity, AS is surface area of a sphere, AP is surface area of particle, V P is volume of particle. Emissivity The emissivity of a particle is defined as the ratio of the radiation emitted by its surface to the radiation emitted by a blackbody at the same temperature. A typical value for emissivity is 1.0 . Drag Model Users can select the drag model to be used in the simulation. See Drag Models for details on the different drag models used by Barracuda VR. Drag Model Multipliers can be used to adjust the calculated drag coefficients.

5.4.2 Defining particle composition The Applied Materials dialog box, Fig. 5.15, displays a list of the applied materials currently present in a given particle species. The sum of the mass fractions of all materials listed here is displayed at the top of this dialog box. Users can add, edit, copy, and delete materials using the buttons in the GUI.

Adding, editing, and deleting applied particle materials New particle materials can be added using the Add button, which will raise the Material Selection dialog box described below. The mass fraction of existing particle materials can be edited by selecting the material from the list and clicking on the Edit button. A material can be deleted by selecting it from the list and clicking on the Delete button. Once the desired particle materials have been added to the list, users should check that their mass fractions add up to 1 and then click on the OK button. This will apply the materials to the particle species and close the dialog box.

The Material selection dialog box When users add a material to a particle species, a new dialog box is raised where a list of available solid and volatile materials is displayed. This material selection dialog box is shown in Fig. 5.16. Users select each material individually and specify its Mass fraction amount before clicking OK . A shortcut to the base materials library is provided for cases where a solid material must be added to the base materials library. Models may require feed or initialization of particles that have already undergone a chemical reaction to some extent. In such cases, the composition of particles initialized or fed into the Barracuda VR simulation domain would be different from the “fresh” particle composition used for calculating conversion in reaction. The Age Factor relates the feed mass of particle components to the initial mass of particle components when solids dependence terms are used in chemical reaction rate coefficients:

m0 = Age Factor ∗ m 74



Fig. 5.15: Applied Materials dialog box showing sample materials added to a particle species where m0 is the initial mass of the selected material when a particle is considered new or “fresh” for reaction chemistry calculations, and m is the mass of the selected material when it enters the Barracuda VR simulation domain. The minimum Age Factor that can be specified is 1. This would indicate that the initialized state in Barracuda VR (the feed composition) is equal to the actual fresh state (the initial composition). Age Factor values greater than 1 indicate the particle is being fed into Barracuda at an “older” composition.

Overall particle density In most cases, the particle density will be automatically calculated from the densities of the solid and volatile materials added to the particle. In these cases, the particle density is displayed at the bottom the Applied Materials dialog box. The exception to this is when a volatile material with an undefined density has been applied to a particle. In this case, the particle density will need to be guiText{manually entered} in units of kg /m3 . When the solver automatically calculates the initial particle density ρ p based on the material composition, the following formula is used. X i is the mass fraction of material i , ρ i is the density of material i , and N is the total number of materials in the particle.

ρ p =

1 N

∑

i=1

Volatiles manager, OK, cancel

Xi ρi

At the bottom of the Applied Materials dialog box are the volatiles manager, OK, and cancel buttons. Clicking Volatiles Manager allows users to access the volatiles window and to create or change any volatile species in the project. Clicking OK saves the values specified and closes the dialog box while clicking Cancel closes the dialog box without saving any values or changes.


75


Fig. 5.16: Material Selection dialog box showing available solids in sample particle species

5.4.3 Specifying a particle size distribution Barracuda VR allows users to specify a particle size distribution using a normal distribution or as a full cumulative particle size distribution using an SFF file.

Normal particle size distribution A normal distribution is specified in the Particle Species dialog box. by entering values for the minimum and maximum particle radius in meters. An example particle size distribution is shown in Fig. 5.17. This is a simple way to specify the particle sizes for cases when an accurate size distribution in unknown. Full particle size distribution When more detailed information is available on particle sizes, it is preferable to specify the full size distribution as an SFF file in Barracuda VR. The SFF file consists of a column of distribution information and a column of corresponding particle sizes. Distribution data The distribution can be specified as either a cumulative distribution (default) or a frequency distribution. When a cumulative distribution is specified, each row indicates the weight percentage of particles that are smaller than or equal to the size specified in the row. When a cumulative distribution is specified, the first row must have a cumulative value of 0 and the last row must have a cumulative value of 100 . When a frequency distribution is specified, each row indicates the percentage of the distribution that has the particle size specified in the row, the sum of which must be equal to 100.

Creating and editing an SFF file for PSD To create a new SFF file describing a specific particle size distribution, users must click on the Edit button in the Particle species dialog box, Fig. 5.14, which will raise the PSD editor, shown in Fig. 5.18. The PSD Editor provides a tabular format for entering the PSD data and displays the current distribution and size input types in the column headers. Alternatively, users may select an existing SFF file that contains the particle size distribution information.

76



Fig. 5.17: Normal particle size distribution for particles between 200 and 600 microns diameter


77


This is done by clicking on the file browse button and selecting the desired file. If the chosen file resides outside of the current project directory, Barracuda VR will ask whether the file should be copied to the project directory.

Fig. 5.18: Particle Size Distribution Editor showing sample particle size distribution Changing column definitions The distribution data can be changed to either cumulative or frequency distribution input by double clicking the header of the first column. The size definition can be changed to either radius or diameter with units of m, mm, or µm by double-clicking on the header of the second column. Adding rows An empty row can be added anywhere in the PSD by selecting the row above and clicking Add Row. Deleting rows A row can be deleted from the table by selecting the desired row and clicking Delete Row. Checking data Once data has been entered, the validity of the data can be checked by clicking Check Data. This will notify the user of any errors in the data input. Saving the SFF file The SFF file can be saved by clicking either the Save or Save as buttons. This will raise the file save dialog box. Plotting data The entered data can be plotted using the Barracuda VR plotting dialog box by clicking Graph. This will raise the PSD Plotting dialog box, shown in Fig. 5.19. The plotting dialog box allows the quick selection of variables for plotting and the line display properties for plotting in XMGR. Updating a running simulation If changes are made to a PSD used in a currently running simulation, the user can have the Barracuda VR solver reread the SFF file and incorporate the changes into the simulation by clicking the Update Simulation button. Updates to a PSD will only be used for new particles that enter the model domain after the PSD changes are made. The sizes of existing particles in the domain will not be altered.

5.4.4 Drag Model Multipliers The drag model multiplier is applied to the selected drag model as a multiplier to give the final drag. It can be a constant or an interpolated value from a data table. Each particle species can have its own drag multiplier.

Constant By default, a value of 1 is specified, which results in no modification of the selected drag model. If any other value is specified, the selected drag model’s calculated drag force is multiplied by that value. Predefined

78



Fig. 5.19: PSD plotting dialog box for selecting variables to plot and display information in XMGR A collection of predefined multiplier tables is available for selection, each of which is based on EMMS principles and generated by the Chinese Academy of Science’s EMMS software. The tables contain data for the drag multiplier as a function of two independent variables – fluid volume fraction and particle-fluid slip velocity. During a simulation, Barracuda VR calculates the drag multiplier by interpolation from the local fluid volume fraction and the particle-fluid slip velocity. The tabulated multipliers are based on the EMMS model [LGW+13] and are the ratio of drag predicted by the EMMS approach divided by the drag predicted from the Wen-Yu drag model. The ratio is calculated by the EMMS approach given the following input parameters: • Fluid Density (kg/m3 ) • Fluid Viscosity (Pa · s) • Particle Density (kg /m3 ) • Particle Diameter (microns) • Particle Sphericity • Fluid Superficial Velocity (m/s) • Solid Mass Flux (kg/m2 · s) • Bed Diameter (m) • Bed Height (m) • Minimum Fluidization Voidage The values of Particle Sphericity, Bed Diameter, and Bed Height are generally not as significant as the other parameters. The bed dimensions are used to constrain the cluster diameter in the model. The drag ratio, which is commonly called the Heterogeneity Index, H d , is tabulated as a function of local gas volume fraction and particle slip velocity,

H d = f (θf , U slip ) C d = H d C dW Y where θf is the gas volume fraction, U slip = |U f − U p | is the slip velocity, C d and C dW Y are the actual drag coefficient and the standard Wen-Yu drag correlation, respectively. The drag multiplier H d is generally between 0 and 1, with 1 corresponding to complete homogeneous fluidization. The data table contains H d as a function of θ f and U slip and is currently in the Tecplot data format. Physical and particle parameters are assumed constant in the generation of the table.


79


The slip velocity range is selected to correspond to a Reynolds number range of 0 to 100. The bounds on the gas volume fraction are generally between 0.4 and 1.0. In the simulation, the drag ratio is interpolated from the table data. If conditions (the slip velocity and the gas volume fraction) are outside of the range, the end values are used. The table values are not extrapolated outside of these ranges. Tables of the Heterogeneity Index are included for the following general conditions: • Bubbling Fluidized Bed • Circulating Fluidized Bed • Circulating Fluidized-Bed Boiler • Circulating Fluidized-Bed Riser, Group A • Circulating Fluidized-Bed Riser, Group B • Circulating Fluidized-Bed Riser, Group D • Fast Fluidized Bed • Riser • Turbulent Fluidized Bed The Archimedes Number, Ar, is also calculated for each of these cases. Ar =

gd p3 ρf (ρ p − ρf ) µ2f

where g is the gravitational constant, d p is the particle diameter, ρ p and ρ f are the particle density and fluid density, respectively, and µf is the fluid viscosity. The Archimedes Number represents the ratio between external (gravity and buoyancy) forces and internal viscous forces and is a representation of the physical properties of the fluid-particle system. The Fluid Superficial Velocity and Solid Mass Flux define the operating conditions. For a given simulation, the Archimedes number can be calculated. For a particle size distribution, the Sauter mean diameter can be used as the particle diameter. If none of the Heterogeneity Index Tables correspond to the conditions of interest, the table with the nearest Archimedes Number may be used. The section for each predefined multiplier shows the conditions for which each table was generated, and a plot of the 2-dimensional multiplier surface calculated by the EMMS software. It is recommended to choose a table with conditions similar to their simulation conditions. Since these multiplier tables were generated with respect to the standard Wen-Yu drag model, the tables should be used with the Wen-Yu model. Bubbling Fluidized Bed Table 5.2: Bubbling Fluidized Bed

Parameter

Value

Fluid Density Fluid Viscosity Particle Diameter Particle Density Archimedes Number Particle Sphericity Fluid Superficial Velocity Solid Flux Bed Diameter Bed Height Minimum Fluidization Voidage Reference

1.225 kg/m3 1.7894e-5 Pa-s 65 microns 1780 kg/m3 18.3 1 0.2 m/s 1.e-5 kg/m2-s (Value must be non-zero in EMMS program) 0.267 m 2.46 m 0.4 [HSWL13]

Circulating Fluidized Bed 80



Fig. 5.20: Bubbling Fluidized Bed

Table 5.3: Circulating Fluidized Bed

Parameter

Value


1.225 kg/m3 1.7894e-5 Pa-s 60 microns 1400 kg/m3 11.3 1 3.5 m/s 86.1 kg/m2-s 0.411 m 8.5 m 0.448 [ZLWL08]

Circulating Fluidized Bed Boiler


81


Fig. 5.21: Circulating Fluidized Bed

Table 5.4: Circulating Fluidized Bed Boiler

Parameter

Value


0.315 kg/m3 4.55e-5 Pa-s 129.74 microns 2598.4 kg/m3 8.5 1 7.236 m/s 39.2 kg/m2-s 3.96 m 24.5 m 0.4 [LZW+13]

Circulating Fluidized Bed Riser, Group A

82



Fig. 5.22: Circulating Fluidized Bed Boiler

Table 5.5: Circulating Fluidized Bed Riser, Group A

Parameter

Value


1.225 kg/m3 1.8e-5 Pa-s 60 microns 1000 kg/m3 8.0 1 1.17 m/s 11.7 kg/m2-s 0.05 m 2.79 m 0.4 [LWL11]

Circulating Fluidized Bed Riser, Group B


83


Fig. 5.23: Circulating Fluidized Bed Riser, Group A

Table 5.6: Circulating Fluidized Bed Riser, Group B

Parameter

Value


1.225 kg/m3 1.8e-5 Pa-s 300 microns 2500 kg/m3 2500 1 7.76 m/s 151.6 kg/m2-s 0.411 m 8.5 m 0.4 [LWL11]

Circulating Fluidized Bed Riser, Group D

84



Fig. 5.24: Circulating Fluidized Bed Riser, Group B

Table 5.7: Circulating Fluidized Bed Riser, Group D

Parameter

Value


1.225 kg/m3 1.8e-5 Pa-s 1020 microns 4000 kg/m3 1.57e5 1 20 m/s 400 kg/m2-s 0.411 m 8.5 m 0.4 [LWL11]

Fast Fluidized Bed


85


Fig. 5.25: Circulating Fluidized Bed Riser, Group D

Table 5.8: Fast Fluidized Bed

Parameter

Value


1.1795 kg/m3 1.8872e-5 Pa-s 54 microns 930 kg/m3 4.75 1 1.52 m/s 14.3 kg/m2-s 0.09 m 10.5 m 0.4 [HSWL13]

Riser

86



Fig. 5.26: Fast Fluidized Bed

Table 5.9: Riser

Parameter

Value


3.65 kg/m3 1.3e-5 Pa-s 65 microns 1500 kg/m3 87.0 1 1.67 m/s 34.6 kg/m2-s 1 m (not specified in Ref.) 1 m (not specified in Ref.) 0.4 [LCL+13]

Turbulent Fluidized Bed


87


Fig. 5.27: Riser

Table 5.10: Turbulent Fluidized Bed

Parameter

Value


1.225 kg/m3 1.7894e-5 Pa-s 90 microns 1375 kg/m3 37.5 1 0.6 m/s 1.e-5 kg/m2-s (Value must be non-zero in EMMS program) 0.05 m 0.75 m 0.4 [HSWL13]

From File Users can provide their own table for the multiplier and apply it to any drag model. The multiplier must be a function of exactly two independent variables (this restriction may be removed in the future), and the table file must be in the Tecplot data format, which is the output format of the EMMS software. One of the independent variables must be the gas volume fraction, and the other must be the particle slip velocity or the Reynolds number based on the particle diameter and the slip velocity. If the multiplier is a function of only one independent variable, the other independent variable can be included as a constant in the table. An excerpt from a file in the correct format is shown here: VARIABLES = "eg", "Uslip", "Hete" ZONE I=1000, J=200 0.4001 0.001 0.0118931

88



Fig. 5.28: Turbulent Fluidized Bed

0.4001 0.00105138 0.0118931 0.4001 0.00110539 0.0118931 0.4001 0.00116218 0.0118931 0.4001 0.00122189 0.0118931 (... many more lines ...) 0.9994 17.4931 2.08863 0.9994 18.3918 2.09417 0.9994 19.3367 2.0996 0.9994 20.3302 2.10493 0.9994 21.3747 2.11015

The first column of data is the gas volume fraction. The second column of data is the fluid-particle slip velocity, but it can be changed to mean Reynolds number (“Uslip” replaced by “Re”). The third column is the drag multiplier.

5.4.5 Agglomeration model The aerodynamic drag on small particles ( < 50 µm) is not well defined but data suggests that the small particles will often group together and behave as larger particles. In Barracuda VR, an optional agglomeration model which accounts for this clustering behavior can be applied to any drag model. The user can chose either the default agglomeration model curve or specify a custom model in the agglomeration dialog box, shown in Fig. 5.29. In all agglomeration models, a relationship is established between a actual particle size and an effective particle size, which is used in the drag force calculations.

Default agglomeration model The default agglomeration model was developed from a small set of cyclone data where the “cut grade” was measured. The cut grade is the amount of solids as a function of particle size which exits the top of the cyclone. The effective particle radius vs real particle radius used in the default model is plotted in Fig. 5.30. The default model is used if


89


Fig. 5.29: Agglomeration model dialog box there is no effective radius file specified in the dialog box. In this case, it is recommended that the default radius cut point of 1.8e-5 be used.

User-defined agglomeration model A user-defined agglomeration model can be created by specifying a radius cut point and an agglomeration model SFF file which is only active for particle sizes less than the radius cut point . The agglomeration model SFF file contains columns for the particle size and effective particle size. For particle sizes between the values in the row entries, the values are linearly interpolated. Typically, the particle size and effective particle size are given equal values at the cut point, as the last entry in the table. Particle size The particle size can be specified as either particle radius or diameter with units of meters, millimeters, or microns. Effective particle size The effective particle size can be specified as either particle radius or diameter with units of meters, millimeters, or microns.

Creating and Editing an Agglomeration SFF File To create a new SFF file for the agglomeration model, users must click on the Edit button in the agglomeration dialog box which will raise the agglomeration model editor, shown in Fig. 5.31. The agglomeration editor provides a tabular format for entering the agglomeration model data and displays the current size input format in the column headers. Alternatively, users may select an existing SFF file that contains the agglomeration model information. This is done by clicking on the file browse button and selecting the desired file. If the chosen file resides outside of the current project directory, Barracuda VR will ask whether the file should be copied to the project directory. Changing column definitions The agglomeration model input can be changed to either radius or diameter with units of meters, millimeters, or microns by double-clicking on the column header. Adding rows An empty row can be added anywhere in the PSD by selecting the row above and clicking Add Row. Deleting rows A row can be deleted from the table by selecting the desired row and clicking Delete Row. Checking data Once data has been entered, the validity of the data can be checked by clicking Check Data. This will notify the user of any errors in the data input. Saving the SFF file The SFF file can be saved by clicking either the Save or Save as buttons. This will raise the file save dialog box. Plotting data The entered data can be plotted using the Barracuda VR plotting dialog box by clicking Graph. This will raise the Barracuda VR Plotting dialog box. The plotting dialog box allows the quick selection of variables for plotting and the line display properties for plotting in XMGR. Updating a running simulation If changes are made to an agglomeration model used in a currently running simulation, the user can have the Barracuda VR solver reread the SFF file and incorporate the changes into the simulation by clicking the Update Simulation button. Updates to a PSD will only be used for new particles that enter the model domain after the PSD changes are made. The sizes of existing particles in the domain will not be altered.

90



Fig. 5.30: Default agglomeration model

Fig. 5.31: Agglomeration model editor


91


92


CHAPTER

SIX

INITIAL CONDITIONS

The simulation initial conditions - the state and composition of the fluid and particles in the model domain at the Initial Conditions section of Barracuda VR. For Fluid ICs, the initial pressure, simulation start - are specified in the temperature, velocity, and composition must be specified. While the model may have multiple initial conditions (ICs) distributed within the domain, the user must ensure that every cell within the domain has an initial condition specified. For Particle ICs, the volume fraction, temperature, and particles species must be specified. Specification of particle ICs in the system is optional (a simulation can start with the domain void of solids).

6.1 Fluid ICs Fluid ICs window, shown in Fig. 6.1, which displays a list of all existing Fluid initial conditions are managed in the fluid ICs and allows users to follow the Add, Edit, Copy, Delete GUI pattern with ICs in the model. In the Fluid ICs list, the initial fluid conditions are displayed as individual lines. Each fluid initial condition in the list has a unique ID number associated with it that is displayed alongside the fluid IC coordinates, velocity, temperature, and pressure.

Fig. 6.1: Fluid Initial Conditions Window showing a single initial condition (IC) added to the domain Since a fluid initial condition must be specified everywhere within the system, every new Barracuda VR model is initialized with one fluid IC already created which spans the entire domain (shown in Fig. 6.1 as fluid IC 000 ). Be aware that when fluid IC 000 is created it is incomplete and requires fluid composition and pressure to be defined. Additionally, temperature and velocity may be adjusted if needed.

93


6.1.1 Specifying a fluid IC The Fluid Initial Condition dialog, shown in Fig. 6.2, provides the interface for entering all values necessary for a fluid IC. This dialog is raised when a new fluid IC is added or an existing fluid IC is edited. Once the all values for a fluid IC are entered, clicking Apply will assign the values to the model and close the dialog. Clicking Cancel will close the dialog and any values entered will be lost. The dialog also provides a link to the Reference Grid which is useful for determining i, j , k locations within the model domain.

Fig. 6.2: Fluid Initial Condition Dialog showing sample IC values for a thermal model. The temperature field is inactive in an isothermal model. Fluid velocity The initial fluid velocity is defined as x , y , and z components. The initial fluid velocity always has units of meters per second (m /s). Temperature For thermal flow models, users must specify the initial temperature of the fluid. The initial fluid temperature always has units of Kelvin (K). Pressure Users must specify the initial pressure of the fluid. For a compressible model, the pressure is always an absolute value (not gauge) and is specified in units of Pascals (Pa).

94

Chapter 6. Initial Conditions


Fluid composition The composition of the fluid can be specified by clicking on the Define fluids button which raises the Applied Fluid dialog, discussed in Defining initial fluid composition. Location Each fluid IC is applied to a range of cells specified in ijk coordinates. The range goes from i1 to i2 in the i direction, from j1 to j2 in the j direction, and from k1 to k2 in the k direction. See Setup Grid for details on i, j, k coordinates in Barracuda VR. Note that for location indices, min or [ (left square bracket) denotes the first possible value, while max or ] (right square bracket) denotes the last possible value. The Reference Grid can be useful for determining appropriate i, j , k values within a model. Comment A comment is not required but it is recommended for documenting any special notes about the fluid IC.

6.1.2 Defining initial fluid composition Users specify the composition of the fluid species by clicking on the Define fluids button in the Fluid Initial Condition dialog, ref: fluid_ics_dialog. This activates the Applied Fluids dialog, shown in Fig. 6.3, which displays a list of all materials currently applied to a fluid IC. From the Applied Fluids dialog users can add, edit, and delete materials using the buttons in the GUI to set the desired initial fluid composition. The composition of a fluid can be specified as Mass fraction or Mole Fraction values.

Fig. 6.3: Applied Fluids Dialog showing an IC composed of 100% Air

Adding, editing, and deleting applied fluid materials New fluid materials can be added using the Add material button, which will raise the Fluid Selection dialog described below. The fraction of existing fluid materials can be edited by selecting the material from the list and clicking on the Edit button. A material can be deleted by selecting it from the list and clicking on the Delete button. Once the desired fluid materials have been added to the list, users should check that their mass fractions add up to 1.0 and then click on the Apply button, which applies the materials to the fluid IC and closes the dialog.

6.1. Fluid ICs

95


The fluid selection dialog When users add or edit a fluid material, the Fluid Selection dialog, shown Fig. 6.4, is raised which displays a list of available fluid materials in the model. Users can select the material to be included in the fluid IC and specify the fractional amount. Clicking Apply will apply the changes to the model and close the Fluid Selection dialog. Clicking Cancel will close the dialog and any changes will be lost.

Fig. 6.4: Fluid Selection Dialog showing a sample list of available fluids to add to the IC. Note If a desired fluid material isn’t shown in the Fluid Selection dialog, the user must add the fluid to the base materials list. The addition of materials to the base materials list is discussed in The Base Materials Window.

6.2 Particle ICs It is often necessary to start a simulation with particles already in the model domain. In Barracuda, the initial particles in a model are specified through particle initial conditions, commonly referred to as particle ICs. Each particle IC in Barracuda specifies the initial placement of a particle species in a range of cells at a specified temperature. The amount of a particle species in each cell is specified by particle volume fraction. To achieve a mix of particle species, multiple particle ICs can be overlaid within the domain up to the close pack volume fraction for the model. This gives the user a large amount of control over the initial particle specification and enables gradients of particle composition and temperature to be specified in the domain. Particle ICs window, shown in Fig. 6.5, which displays a list of all existing particle Particle ICs are managed in the ICs and allows users to follow the Add, Edit, Copy, Delete GUI pattern with the ICs in the model. In the Particle initial conditions list, the initial particle conditions are displayed as individual lines. Each particle initial condition in the list has a unique ID number associated with it that is displayed alongside the particle IC coordinates, volume fraction, temperature, number of computational particle per cell, and any user description of the IC. When a particle IC is added or edited, it is specified in the Particle Initialization dialog discussed in Specifying a particle IC . User control over the number and distribution of computational particles for particle ICs is also provided in the Particle

96



Initial Conditions window and within the Advanced Particle Distribution Options dialog discussed in Adjusting the distribution of computational particles.

Fig. 6.5: Particle Initial Conditions Window with sample particle IC Computational Particles and Distribution Options The specification of Initial computational particles per cell and Distribution options is discussed in Adjusting the distribution of computational particles.

6.2.1 Specifying a particle IC The values of a particle IC are defined in the Particle Initialization dialog, shown in Fig. 6.6, which is raised when a new particle IC is added or an existing particle IC is edited. In the Particle Initialization dialog, users must specify the location of the particle IC and the particle species and volume fraction with which to initialize. The initial particle temperature must be specified if a thermal model is being set up (see The Global Settings Window) and the computational particles per cell must be specified if manual input of computational particles selected (see Adjusting the distribution of computational particles). Particle species The particle species to be initialized is selected from a drop-down menu which contains a list of all particle species that have been defined in the model (see Particle Species for information on defining particle species). If multiple particle species are being initialized at a location, there will need to be a separate particle IC for each. Note that in the drop-down list, the particles are listed by the species ID number and any comment that was applied to the species. Particle volume fraction Users specify the volume fraction of the particles. The initial particle volume fraction must be less than or equal to the close pack volume fraction ( Particle Interactions). When initializing a mixture of particles of different species in a system, it is often the case that mass fractions of the various species are known. However, Barracuda VR requires that volume fractions be specified for each species. See Converting from Mass Fraction to Volume Fraction for details about converting from the known mass fractions to the volume fraction values required by Barracuda VR. Temperature For thermal flow models, users must specify the initial particle temperature. The initial particle temperature always has units of K.

6.2. Particle ICs

97


Fig. 6.6: Particle Initialization Dialog showing the sample specification of a particle IC Location The location of the particles at the start of the simulation is specified in i, j, k coordinates. See Setup Grid for details on i, j, k coordinates in Barracuda VR. Note that for location indices, min or [ denotes the first possible value, while max or ] denotes the last possible value. Computational particles per cell If Manual input is selected in the Particle ICs window, the user must specify the number of computational particles per cell here. Additional information on specification initial computational particles is provided in Adjusting the distribution of computational particles. Particle momentum Selecting No particle momentum for this distribution will fix all particles initialized in the system in space. The presence of the particles are still coupled with the fluid phase (possibly causing a pressure drop inside the fluid) but the particles will have zero velocity. Comment A comment is not required but it is recommended for documenting any special notes about the particle IC. Apply, cancel, and the reference grid At the bottom of the Particle Initialization dialog are the apply, cancel, and reference grid buttons. Clicking Apply saves the values specified and closes the dialog while clicking Cancel closes the dialog without saving any values or changes. The Reference grid button raises a reference grid dialog and can be helpful in determining appropriate i, j, k coordinates when specifying particle location.

6.2.2 Adjusting the distribution of computational particles Particles that are initially in the system must have a sufficient number of computational particles for a simulation to have an appropriate level of resolution in the particle phase. Under Initial computational particles per cell in the Particle Initial Conditions window ( Fig. 6.5), the user has the option of either having the number of computational particles initialized in each cell automatically calculated or to manually input a number of particles for each particle IC. Automatically calculated When the Automatically calculated option is selected, the appropriate number of computational particles per cell is determined by Barracuda VR based on one of five levels of resolution selected by the user. Resolution levels of high or medium-high will provide more computational particles in the system but result in longer calculation times. Conversely, low resolution levels will result in faster calculation times but the user should

98



verify that enough computational particles are available to accurately model the system. A default setting of medium is recommended. Manual input When the Manual input option is selected, the user will need to specify a number of computational particles per cell for each particle IC and it up to the user to determine that an appropriate level of particle phase resolution is being used. Particle distribution options The Particle Distribution Options dialog, shown in Fig. 6.7, allows users to specify an initial particle momentum override, adjust initial particle seeding, and the calculation of particle resolution for the particles being initialized in the system. It is important to note that all options specified in this dialog apply to all particle ICs. This dialog is activated by clicking on the Options button in the Particle Initial Conditions window.

Fig. 6.7: Particle Distribution Options Dialog Particle momentum Selecting No particle momentum will fix all particles initialized in the system in space. This will override any no particle momentum settings on any individual particle IC. This setting is typically not used and is included in Barracuda VR to maintain backward compatibility. Initial particle seeding Selecting Seed particles with random locations will place particles randomly within each cell. If unselected, particles are evenly distributed within a cell. By default this option is unselected. Particle resolution This gives users control over the definition of number of computational particles per cell. When computational particles are distributed, the number of computational particles that a cell receives will be dependent on the volume of the cell, with larger cells receiving more particles and smaller cells receiving fewer particles. When By average volume size is used for the particle resolution, an average-sized cell will receive the computational particles specified by each particle IC and by max volume size is selected the largest cell will receive specified computational particles for a particle IC. Typically, the default setting, By average volume size , is used.

6.2. Particle ICs

99


100


CHAPTER

SEVEN

BOUNDARY CONDITIONS

Boundary conditions define how the outer edges of the model geometry are treated. In Barracuda VR, there are a Boundary Conditions section, which describe the inflow and variety of boundary conditions, managed in the outflow of material from the domain, temperature of solid walls, and the introduction of Eulerian and Lagrangian tracers for tracking fluid movement during simulation. Any wall within the simulation domain that does not have either a pressure BC or flow BC attached to it is considered a wall. Pressure BC The pressure BC establishes an opening in the domain with a specified pressure through which fluid and particles can move in or out. Although a pressure BC is typically used as a fluid outlet, it can be used for fluid inflow and as a location for a particle feed. ( more info) Flow BC The flow BC establishes an opening in the domain through which a specified mass rate of fluid moves in or out of the domain. A flow BC is typically used as a fluid inlet with an optional particle feed but can also serve as a fluid and particle outlet. In the case when a flow BC acts as an outlet, the fluid is drawn from the simulation domain at the specified rate. (more info) Thermal Wall BC A thermal wall applies a specified temperature to a solid boundary which transfers heat from the wall to the domain at the rates specified by the fluid-to-wall heat transfer model (see Heat Transfer Models) and radiative heat transfer. ( more info) Passive Scalar BC A passive scalar creates an injection of a non-interacting Eulerian scalar variable which enters with the fluid or particles at a flow or pressure BC. Passive scalars are typically used for determining a residence time distribution for fluid flow. ( more info) Injection BC An injection BC defines a point source of fluid, particles, or Lagrangian tracers into the domain. The point of injection can be at any location within the domain which makes injection BCs an important tool for adding nozzles, distributor shrouds, or other inlets that cannot be easily captured by the grid. ( more info)

7.1 Pressure BCs A pressure boundary condition (BC) specifies a fluid opening in the domain through which fluid can flow in or out with a specified pressure. The fluid velocity at a pressure BC is determined according to the underlying fluid dynamics equations in Barracuda VR. The composition and temperature are calculated from the inner cells at the boundary for fluid outflow whereas a specified composition and temperature are used for fluid inflow. Fluid outlets are typically modeled with pressure boundaries in Barracuda VR models. Pressure BCs are managed in the Pressure BCs Window, shown in Fig. 7.1 , which displays a list of all existing pressure BCs and allows users to follow the Add, Edit, Copy, Delete GUI pattern within the model. Within this window, the pressure boundary conditions are displayed as individual lines. Each pressure BC in the list has a unique ID number associated with it that is displayed alongside the pressure BC locations and other pressure BC parameters.

Vent table fileA vent table specifies a pressure boundary at multiple points within a domain and is a useful tool for cases when there are a large number of discrete locations with a shared pressure boundary condition. A new vent table

101


Fig. 7.1: Pressure BCs Window with sample pressure BC can be created by clicking New Table which will open the vent table editor while existing vent tables can be added to the model and edited by clicking View Table.

K-factor adjustment Each pressure boundary can optionally have a irrecoverable pressure drop at the boundary that is proportional to the local fluid velocity by a unique factor K which is useful for modeling flow through filters or other porous media at a boundary. The K -factor adjustment provides a global change to the value of K to account for particles at the boundary. The equation for the K -factor adjustment is k = (1 + c1 θ p )kclean where kclean is the K -factor specified at the boundary, θ p is the particle volume fraction, and c 1 is a “clogging factor”. For more information on the K -factor, see Specifying a pressure BC .

7.1.1 Specifying a pressure BC All pressure BC parameters are defined in the Pressure BCs Dialog, shown in Fig. 7.2. This dialog is activated when either a new pressure BC is added or an existing pressure BC is edited. Once the values have been entered, clicking Apply will assign the pressure BC to the model and close the dialog whereas clicking Cancel will close the dialog and any values will be lost.

Location Direction Pressure BCs must be aligned normal to either the x, y , or z , direction which must be selected. Pressure BCs can often be applied to an area of a complex boundary shaped by overlaying multiple pressure BCs in different directions. i, j, k Pressure BCs are placed at cells specified by the range of i , j , k coordinates entered. A pressure BC will be applied to all cells that are fully outside of the fluid domain, lie adjacent to a wall in the direction specified, and are within the range of i , j , k coordinates. The range of coordinates is from i 1 to i 2 , j 1 to j 2 , and k 1 to k 2 in the x , y , and z directions. Proper coordinate values can be determined using the Reference Grid . Note that the dialog will also accept min or [ for the minimum coordinate value and max or ] for the maximum coordinate value in the model.

Flux plane options Follows the Flux Plane Options GUI pattern.

Comment Follows the Comment Field GUI pattern.

102

Chapter 7. Boundary Conditions


Fig. 7.2: Pressure BCs Dialog showing sample pressure BC

Fluid behavior at boundary Pressure file An SFF file, which is always recommended based on the ability to use Interact while a simulation is running, can be used to specify time-dependent pressure BC parameters. The file will override any specified values that have been entered in the dialog. Additional information on the use of an SFF file at a pressure BC is provided in Defining a transient pressure BC Specify values Area fraction Area fraction is the physical open area fraction for the pressure BC. This is typically used in the presence of screens or orifices and the default value of 1.0 indicates that the boundary is completely open. Pressure Users can input a single value for pressure at the boundary in units of pascals (Pa). Note that this pressure value will be overridden if a pressure BC file is used. Temperature Users can input a single temperature value at the specified boundary in units of kelvin (K). Note that temperature can be specified only for thermal flow and that a temperature value entered here will be overridden if a pressure BC file is used. K-factor The K -factor adds an additional pressure drop between the boundary value (specified at the BC) and the first interior cell next to the boundary which is proportional to K times the dynamic head, ∆P = 0.5ρKv 2 where ρ is the local fluid density and v is the local fluid velocity at the boundary. A non-zero K factor at a boundary will have the effect of spreading the flow of a fluid across a boundary and damping any local velocity fluctuations at the boundary.

Properties Fluid properties if inflow A pressure BC does not enforce inflow or outflow of fluid in a strict sense. Instead, the user usually has an idea of whether fluid is predominantly entering or exiting the system at the pressure BC. It is possible that some recirculation occurs at the pressure BC, such that fluid is entering through some cells and exiting through other cells. The Fluid properties if inflow list determines fluid properties at a boundary if fluid inflow occurs. If Specified BC values is selected, the temperature and composition entered at the boundary will be used at the boundary; this setting should be used for pressure BCs that are predominantly inflow. Alternatively, if Interior cell values is selected, the fluid composition and temperature will be calculated from the cells in the interior of the domain; this setting should be used for pressure BCs that are predominantly outflow. The default setting is to use Interior cell values.

7.1. Pressure BCs

103


Applied fluid species The composition of the fluid at the pressure boundary is specified by clicking on the Define fluid species button which will raise the Applied Fluid Dialog, discussed in Applied Materials.

Particle behavior at boundary Options are discussed in Particle Feeds at Pressure and Flow BCs .

7.1.2 Defining a transient pressure BC When transient pressure data is entered for a pressure BC, it is done through the use of an SFF file in Barracuda VR. An existing SFF file can be added to the model by clicking the browse button in the pressure BC dialog, Fig. 7.2. Clicking Edit will open an added file in the SFF Editor or create a new SFF file if the file name box is empty. The columns within the SFF file, shown in Fig. 7.3, are time, pressure, temperature, area fraction, particle feed, and K-factor. For more information on creating and editing SFF files, see Tabular Input Using SFF Files.

Transient input data Time The time specifies when the parameters in the other columns should be used. The units of time can be either seconds (s) or minutes (min). Pressure The pressure to use at the pressure BC. Pressure can have units of either pascals (Pa) or pounds per square inch (psi). Temperature The temperature of any inflow fluid. To be used, this value requires that Fluid properties if inflow be set to Interior cell values for the pressure BC. Temperature can have units of kelvin (K), Fahrenheit (F), or Celsius (C). Area fraction Area fraction is the physical open area fraction for the pressure BC. This is typically used in the presence of screens or orifices and the default value of 1.0 indicates that the boundary is completely open. Particle feed The particle feed column determines whether any specified particle feed at the boundary is to be enabled. Like all parameters in the transient pressure BC file, the value of the particle feed is determined by linearly interpolating between time entries in the SFF file. The convention used in Barracuda VR is that the particle feed will be enabled when the interpolated value is greater than 0.5 . K-factor The K -factor adds an additional pressure drop between the boundary value (specified at the BC) and the first interior cell next to the boundary which is proportional to K times the dynamic head, ∆P = 0.5ρKv 2 where ρ is the local fluid density and v is the local fluid velocity at the boundary. A non-zero K factor at a boundary will have the effect of spreading the flow of a fluid across a boundary and damping any local velocity fluctuations at the boundary.

7.2 Flow BCs A flow boundary condition (BC) specifies a fluid opening in the domain through which fluid can flow in or out at a specified rate. The fluid pressure at a flow BC is determined according to the underlying fluid dynamics equations in Barracuda VR. For fluid outflow, the composition and temperature are calculated from the inner cells at the boundary and a specified composition and temperature are used for fluid inflow. Fluid inlets are typically modeled with flow BCs in Barracuda VR models. Flow BCs are managed in the Flow BCs Window, shown in Fig. 7.4, which displays a list of all existing flow BCs and allows users to follow the Add, Edit, Copy, Delete GUI pattern within the model. Within this window, the flow boundary conditions are displayed as individual lines. Each flow BC in the list has a unique ID number associated with it that is displayed alongside the flow BC location and other flow BC parameters.

104



Fig. 7.3: SFF Editor showing sample pressure BC file

Fig. 7.4: Flow BCs Window with sample flow BC

7.2. Flow BCs

105


7.2.1 Specifying a flow BC All flow BC parameters are defined in the Flow BC Dialog, shown in Fig. 7.5. This dialog is activated when either a new flow BC is added or an existing flow BC is edited. Once the values have been entered, clicking Apply will assign the flow BC to the model and close the dialog whereas clicking Cancel will close the dialog and any values will be lost.

Fig. 7.5: Flow BC Dialog with a sample inlet flow BC

Location i, j, k Flow BCs are placed at cells specified by the range of i, j , k coordinates entered. A flow BC will be applied to all cells that are inside the fluid domain, lie adjacent to a wall in the direction specified, and are within the range of i , j , k coordinates. The range of coordinates is from i 1 to i2 , j1 to j2 , and k1 to k2 in the x, y , and z directions. Proper coordinate values can be determined using the Reference Grid . Note that the dialog will also accept min or [ for the minimum coordinate value and max or ] for the maximum coordinate value in the model.

Flux plane options Follows the Flux Plane Options GUI pattern.

Comment Follows the Comment Field GUI pattern. Flow direction Once the cells containing the boundary faces are identified, the Flow direction is used to identify which walls should be changed to flow BC faces. This is only used for the selection of faces. During simulation the flow always enters normal to the boundary face. There are multiple choices for direction. x-, y-, z- direction flow Often the quickest way of applying a flow BC to the face is to select either an x-, y-, or zdirection. This will apply the flow BC to any face that has a normal vector that is within the x/y/z variation angle of the selected x , y , or z direction. Normal to surface flow The flow BC will be applied to all surfaces within the specified range of i, j , k coordinates. If Normal Limit is selected, all available faces within the cell range are used for the flow BC. Alternatively, X zero, Y zero, or Z zero can be selected to eliminate any faces in the respective direction. Direction flow vector The flow BC will be applied to all faces with a normal vector that is within the Vector variation angle of the specified vector, see Example a shown in Fig. 7.8. When the Force absolute direction check box is selected, only the inward normal of a face will be used for comparison, see Example b in Fig. 7.8. Note: The direction

106



Fig. 7.6: Flow BC in x, y, and z direction

Fig. 7.7: A normal to surface Flow BC

7.2. Flow BCs

107


flow vector is only used for the selection of faces. During simulation, the direction of flow is always normal to the vector face.

Fig. 7.8: Vector direction in Flow BC

Fluid behavior at boundary Use BC Connector data The BC Connector provides the fluid mass flow rate, temperature, and composition. See BC Connections for more information on setting up a BC Connector. Flow file An SFF file, which is always recommended based on the ability to use Interact while a simulation is running, can be used to specify time-dependent flow BC parameters. The file will override any specified values that have been entered in the dialog. Additional information on the use of an SFF file at a flow BC is provided in Defining a transient flow BC . Specify values Velocity flow The fluid flow rate at a flow BC can be specified as a superficial velocity by selecting Velocity flow from the list in the Flow BC Dialog under Specify values. The specified superficial velocity, in m /s, will be applied equally across the entire face of the flow BC and the mass rate of the fluid entering at the boundary will be calculated based on the velocity, cross-sectional area of the boundary, and specified pressure and temperature. If particles are present at the boundary, the velocity at the boundary surface will be larger than the specified superficial velocity by a factor of (1 − θ p )−1 , where θ p is the particle volume fraction, which accounts for U nput the volume within the cell occupied by particles. U B = 1i− A positive value of the velocity corresponds to θp inflow of fluid whereas a negative velocity corresponds to an outflow of fluid at the boundary. Mass flow The fluid flow rate at a flow BC can be specified as a mass rate by selecting Mass flow from the list in the Flow BC Dialog under Specify values. The specified mass flow rate, in kg/s, will be applied equally across the entire face of the flow BC. The volumetric flow rate and velocity at the boundary will be calculated based on the mass rate, cross-sectional area of the boundary, and specified pressure and temperature. A positive value of the mass rate corresponds to inflow of fluid whereas a negative mass rate corresponds to an outflow of fluid at the boundary. Pressure The specified Pressure is used to define the fluid density at the boundary. However, if the flow type is incompressible (see The Base Materials Window), the pressure is not used in the calculation. Values of pressure need to be entered in units of pascals (Pa). Temperature The temperature is used to calculate the incoming fluid density for compressible flow problems (see The Base Materials Window) and the incoming fluid enthalpy for thermal problems (see The Global Settings

108



Window). In cases of outflow or where the flow is incompressible and isothermal, the temperature is not used. Values of temperature need to be entered in units of kelvin (K).

Fluid composition The composition of the fluid at the flow boundary is specified by clicking on the Define fluids button which will raise the Applied Fluid Dialog, discussed in Applied Materials.

Particle behavior at boundary Options are discussed in Particle Feeds at Pressure and Flow BCs .

7.2.2 Defining a transient flow BC When transient flow data is entered for a flow BC, it is done through the use of an SFF file in Barracuda VR. An existing SFF file can be added to the model by clicking the browse button in the flow BC dialog, Fig. 7.5. Clicking Edit will open an added file in the SFF Editor or create a new SFF file if the file name box is empty. The columns within the SFF file, shown in Fig. 7.9, are time, mass flow or velocity, temperature, pressure, particle feed, and number density manual. For more information on creating and editing SFF files, see Tabular Input Using SFF Files.

Transient input data Time The time specifies when the parameters in the other columns should be used. The units of time can be either seconds (s) or minutes (min). Flow The flow can be specified as either a mass flow rate in kg/s or velocity in units of meters per second (m /s), feet per second (ft /s), or centimeters per second (cm /s). The flow type and units can be changed by double clicking the column header in the SFF file editor. Temperature The temperature of any inflow fluid. Temperature can have units of kelvin (K), Fahrenheit (F), or Celsius (C). Pressure The pressure to use at the pressure BC. Pressure can have units of either pascals (Pa) or pounds per square inch (psi). Note that the pressure specified here is only used if velocity is the chosen basis for the flow BC definition. It is ignored for mass flow rate-based flow BCs. Particle feed The particle feed column determines whether any specified particle feed at the boundary is to be enabled. Like all parameters in the flow BC file, the value of the particle feed is determined by linearly interpolating between time entries in the SFF file. The convention used in Barracuda VR is that the particle feed will be enabled when the interpolated value is greater than 0.5. Number Density Manual The number density manual controls the rate at which computational particles are fed through a boundary. This overrides the value used in the particle feed, discussed in Particle Feed .

7.3 Particle Feeds at Pressure and Flow BCs The behavior of particles at a boundary is an important consideration when any pressure BC or flow BC is specified in Barracuda VR. Within the Pressure BC Dialog (see Specifying a pressure BC ) and the Flow BC Dialog (see Specifying a flow BC ) there are six main options for specifying particle behavior at a boundary: Use BC Connector data , No particle exit , Particle out flow, Particle feed with slip and volume fraction, Particle feed with slip and flux, and Particle feed with slip and flow rate. The dialog for selecting the particle behavior, common to both pressure BCs and flow BCs, is shown in Fig. 7.10. Use BC Connector Data This selection will allow the current flow BC to feed particles based on information from a BC Connector. Note that the BC Connector itself, which defines the source(s) of this particle information, must be defined in that section of the GUI. This is useful for returning particles to a system when you wish to maintain exactly the same particle size, temperature, composition, etc.

7.3. Particle Feeds at Pressure and Flow BCs

109


Fig. 7.9: SFF Editor showing sample flow BC file with a ramp in mass flow rate up to 10 kg/s over the first two seconds of simulation

Fig. 7.10: Interface for selecting particle behavior at a boundary within the pressure BC and flow BC dialogs

110



No particle exit This selection will allow fluid to exit, but prevent particles from leaving the domain through the boundary. This particle behavior is typically applied to fluid inlets that have no particle feed. Particle out flow This selection will allow particles to leave through the boundary. When this option is selected, two additional text fields will appear which can be used to limit the size of particles that can leave through the boundary. By default, all particles are allowed to exit the domain, regardless of their size ( Min = 0 m, Max = UNLIMITED). To specify lower or upper size limits for particles allowed to exit at a domain, change the Min and Max fields as necessary. Be careful to specify radius in units of meters. Particle feed The particle behavior interface provides three options for specifying particle feeds: Slip and volume fraction, Slip and mass flux , and Slip and mass flow rate . The particle feed types only differ in the information needed from the user to set up the particle feed. For each feed option the slip ratio and either the particle volume fraction, particle mass flux, and particle mass flow rate must be specified. Selecting any of the three options will enable buttons for Edit particle feed and Particle feed control. These buttons enable the specification of a particle feed (discussed in Specifying a particle feed ) and the optional use of a particle feed controller (discussed in Setting up a particle feed controller ). Any particle feed requires that the first cell inside the domain has space available for additional particles. If there is not enough space, a particle feed may be temporarily reduced or stopped. Therefore, it is important to ensure that the flow of fluid to a boundary with a particle feed is sufficient to transport the particles through the feed cells and avoid a close-pack state. The feed boundary maintains a budget of added mass to requested feed. If there is an excess in the particle feed, the feed is suspended. If there is a deficit in the feed mass, the boundary condition will attempt to make up the particle mass in later time-steps.

7.3.1 Specifying a particle feed The Particle Feed Settings Dialog will be raised when a particle feed is selected in the particle behavior interface and the Edit particle feed button is clicked. This dialog allows the user to define the mixture of particles species to feed into the system, the particle feed rate, and the number of computational particles to be used. The Particle Feed Settings Dialog is shown in Fig. 7.11. The particle feed rate that is specified may be modified externally by a transient file at the boundary (see Defining a transient pressure BC or Defining a transient flow BC ) by turning the feed on or off in the Particle Feed column or by changing the number of computational particles being fed in the Number Density Manual column. Additionally, particle feed rates may be overridden if a particle feed controller is specified at the boundary. The setting up of a particle feed controller is discussed in Setting up a particle feed controller .

Defining the particle species mixtureThe particle feed used at a boundary can consist of a single particle species or of a mixture of particles species. Particle species can be added or edited which in both cases raises the Species Editor Dialog shown in Fig. 7.12. Inside the Species Editor Dialog a particle species and the fractional amount of the species in the mixture is specified. Any existing particles can be removed from the mixture by selecting the particle species from the list and clicking the Delete button. Any particle species to be added to the particle feed must be previously specified. The specification of particle species is discussed in Particle Species. Solid fraction When a mixture of particle species is specified, the particles are added to the mixture in fractional amounts. The Solid fraction list allows the fractional amounts to be defined as Mass fractions or Volume fractions of the particle species. The mass fractions are not equal to the volume fractions if the densities of the species are different. The fractions of all species should always sum to one. If there is only one species, the fraction is always one regardless of the fraction type.

Defining a particle feed rate To define a particle feed, the particle/fluid slip ratio and the particle feed per average volume are required inputs and the Particle feed volume fraction, Particle feed mass flux , or Particle feed mass flow rate may be required, depending on the particle feed type selected. Particle/fluid slip ratio The particle/fluid slip ratio is the ratio of initial particle velocity to the fluid velocity at the boundary. After a particle has been fed into the domain, this slip ratio is no longer applied to the particle; the drag model will then determine fluid to particle interaction. The slip ratio can be less than,


111


Fig. 7.11: Particle Feed Settings Dialog for selection slip and volume fraction with sample species added

Fig. 7.12: Species Editor Dialog with sample particle species available for selection

112



equal to, or greater than 1.0 and should be selected based on the expected flow situation at the boundary. In cases where the fluid flow is carrying the particles, it may be appropriate to have a slip ratio of less than 1.0. If, on the other hand, the particle feed is due to particle flow/movement (for example, diplegs return), which is pulling fluid along with it, than it may be appropriate to have a slip ratio is greater than 1.0. The default value is set to 0.5. Particle feed per ave volume Barracuda VR uses computational particles to represent multiple real particles with the same attributes, which allows a large number of particles to be modeled. At a boundary, the particle feed per ave volume value controls the numerical resolution of the computational particles. As was the case with Particle ICs, it is important to balance the higher accuracy of more computational particles with the associated computational cost. This is especially true at particle feed locations, which may cause the total number of computational particles to continuously increase. It is equally as important to remember that too few numerical particles will not adequately represent the physics. The value of n p which the user specifies in the Particle feed settings dialog is related to the number of computational particles expected to be in the system at the end of a calculation. The equation defining n p is based on the number of Eulerian cells and the expected particle volume fraction at the end of the calculation:

n p =

N p N cells θ p

where n p is the value to specify for particle feed per ave volume or number density manual, N p is the total number of computational particles at the end of the calculation, and N cells is the expected number of Eulerian cells which will be filled with particles at a volume fraction of θ p . A frequently used approximation for selecting a value for the particle feed per ave volume is 50/θ p , where θ p is the expected particle volume fraction at the boundary . The default value for Particle feed per ave volume is set to 125 which corresponds to a θ p of 0.4. See Estimating Particle Volume Fraction at a BC for additional information. Note: If a transient file is used to specify the fluid behavior at a boundary (see Defining a transient pressure BC or Defining a transient flow BC ) the value of Particle feed per ave volume will be overridden by the value specified in the Number Density Manual column. Particle feed volume fraction If the slip and volume fraction feed was selected, the user will be prompted to specify a particle feed volume fraction. With this selection, the the particle feed mass rate, m˙ p , will be dependent on the fluid volumetric flow rate, V ˙f at the boundary according to the following expression

m˙ p = V ˙f · S · ρ p

θ p 1 − θ p

where S is the slip ratio, ρ p is the particle density, and θ p is the specified particle volume fraction. Particle feed mass flux If the slip and mass flux feed was selected, the user will be prompted to specify a particle feed mass flux in units of kg /m2 /s. With this selection, the particle feed mass rate, m˙ p will be dependent on the cross sectional area of the boundary, Ab , and the specified particle mass flux, F p . m˙ p = F p · Ab Particle feed mass rate If the slip and mass flow rate feed was selected, the user will be prompted to specify a particle mass flow rate directly in units of kilograms per second (kg /s).

7.3.2 Setting up a particle feed controller A particle feed controller allows a particle feed rate to be automatically adjusted based on flow rates through flux planes in the model, total system mass, or even temperatures and pressures at locations within the model. A particle feed controller is set up within the particle feed control dialog which is raised when the Particle Feed Control button is clicked from within either the pressure BC or flow BC dialogs. The Particle Feed Control Dialog is shown in Fig. 7.13. The most frequently used control schemes, Match particle mass flow rate through flux plane and Keep particle mass within specified limits, can be set from the Particle Feed Control Dialog. Control schemes based on temperature, pressure, or other flux plane variables must be set from the Advanced Feed Control Dialog, discussed in Using an advanced particle feed controller . The Advanced Feed Control Dialog is raised by clicking the Advanced Feed Control button within the Particle Feed Control Dialog. Match particle mass flow rate through flux plane


113


Fig. 7.13: Particle Feed Control Dialog

114



The particle feed rate at a boundary can be set to match the particle flow rate at another boundary using the Match particle mass flow rate through flux plane option in the Particle Feed Control Dialog. When this option is selected, the user must specify a flux plane to match, an efficiency, a slip ratio, and a control time. Match flux plane The Match fluxplane list contains all valid flux plane files that have been defined in the project. The particle feed will be set to match the flux plane selected from the list. Efficiency The particle feed rate at the boundary will be equal to the matched flux plane particle mass flow rate multiplied by the efficiency. The particle feed rate will match exactly at the default efficiency of 1.0. Slip ratio The slip ratio is the ratio of particle velocity to the fluid velocity at the boundary. The value specified in the particle feed controller will override the slip ratio value given when the particle feed was created (see Specifying a particle feed ). Over time The particle matching can be set to happen gradually over the period of time specified in the over time field. This can be useful in situations where a large change in particle feed in one time step may cause instabilities in the model. If the over time parameter is 0.0 seconds (default), the feed change takes place in one time step whenever possible. Keep particle mass within specified limits The particle feed rate at the boundary can be controlled as an on/off controller to maintain the total particle mass in the model within specified limits when Keep particle mass within specified limits is used. When this setting is used, all parameters of the particle feed are used, except for the slip ratio which is specified again in the particle controller. Minimum system mass The particle feed at the boundary will turn on when the total particle mass in the model decreases below the Minimum system mass. The minimum system mass is entered in units of kilograms (kg). Maximum system mass The particle feed at the boundary will turn off when the total particle mass in the model increases above the Maximum system mass. The maximum system mass is entered in units of kilograms (kg). Slip ratio The slip ratio is the ratio of particle velocity to the fluid velocity at the boundary. The value specified in the particle feed controller will override the slip ratio value given when the particle feed was created (see Specifying a particle feed ).

7.3.3 Using an advanced particle feed controller The Advanced Feed Control Dialog, shown in Fig. 7.14, provides an interface for controlling particle feeds based on pressure, temperature, system mass, or flux plane parameters. An advanced particle feed controller is set up by identifying a control variable (temperature at a cell, pressure at cell, flux plane data set, etc) and then specifying the behavior of the controller based on the results of two possible tests. Control variable The control variable, identified in the Control based on list, can be either a temperature, pressure, the system mass, or a flux plane variable. Temperature When temperature is selected, the user will be prompted to specify an i , j , k location at which the temperature should be measured. The Reference grid button at the bottom of the dialog is useful for determining the desired i, j , k location in the model. Pressure When pressure is selected, the user will be prompted to specify an i , j , k location at which the pressure should be measured. The Reference grid button at the bottom of the dialog is useful for determining the desired i, j , k location in the model. System mass When system mass is selected, the entire mass of particles in the model will be used as the control variable. Flux plane When a flux plane is selected, the user will be first prompted to select a flux plane from the Flux plane name list which is populated with all valid flux planes in the model. The user will then specify the flux plane control variable to use. The available options are • Fluid mass flow rate - the total mass flow rate of fluid passing through the flux plane


115


Fig. 7.14: Advanced Feed Control Dialog with example control scheme based on temperature

116



• Particle mass flow rate - the total mass flow rate of all particle species passing through the flux plane • Fluid mass flux - the total mass flux of fluid passing through the flux plane • Particle mass flux - the total mass flux of all particle species passing through the flux plane • Scalar integrated mass - the time-integrated mass of all particles species that has passed through the flux plane Disabled When disabled is selected, the particle feed controller will not be active and the particle feed will be dictated by the particle feed settings and any transient file that is in use. If-then control testsThere are two if-then tests that are used in the advanced feed controller to control the particle feed. The controller starts with Test 1 and compares the selected control variable against the test value using the specified operator . If the result of Test 1 is true, then the specified action will be taken. If the Test 1 result is not true, then the controller will evaluate Test 2. If Test 2 is true, then the action specified under Test 2 will be taken. If Test 2 is not true, then no change to the solids feed is made. If-then operators The operator in each test case can be selected from the following: = > < >= <= ABS = ABS > ABS < ABS >= ABS <=

returns true if the control variable is equal to the specified test value returns true if the control variable is greater than the specified test value returns true if the control variable is less than the specified test value returns true if the control variable is greater than or equal to the specified test value returns true if the control variable is less than or equal to the specified test value returns true if the absolute value of the control variable is equal to the specified test value returns true if the absolute value of the control variable is greater than the specified test value returns true if the absolute value of the control variable is less than the specified test value returns true if the absolute value of the control variable is greater than or equal to the specified test value returns true if the absolute value of the control variable is less than or equal to the specified test value

At times, a user will want to have Test 1 always be evaluated as true. A common trick for accomplishing this is to set the operator as “ABS >” and to set the test value to -1 . This statement will always evaluate true, regardless of the control variable or its value. If-then actions When the result of a test is true, the following actions can be taken to control the particle feed: • ON - turns the particle feed on • OFF - turns the particle feed off • MULTIPLY -multiplies the particle feed variable (particle volume fraction, particle mass flow rate, particle mass flux) by the number specified in the Value over the specified time and at the specified slip ratio. The slip ratio and time parameters are explained below. • ASSIGN - assigns the specified value to the particle feed variable (particle volume fraction, particle mass flow rate, particle mass flux) by the number specified in the Value over the specified time and at the specified slip ratio. The efficiency, slip ratio, and time parameters are explained below. • MATCH - will match the particle feed rate to the rate of the selected flux plane with the specified efficiency over the specified time and at the specified slip ratio. The option for match is only available when a flux plane is selected as a control variable. The slip ratio and time parameters are explained below. Efficiency When MATCH is selected for a flux plane control variable, the particle feed rate at the boundary will be equal to the matched flux plane particle mass flow rate multiplied by the efficiency. The particle feed rate will match exactly at the default efficiency of 1.0 . Slip ratio The slip ratio is the ratio of particle velocity to the fluid velocity at the boundary. The value specified in the particle feed controller will override the slip ratio value given when the particle feed was created (see Specifying a particle feed ). Over time The particle feed multiplication, assignment, or matching can be set to happen gradually over the period of time specified in the over time field. This can be useful in situations where a large change in particle feed in one time


117


step may cause instabilities in the model. If the time parameter is 0.0 seconds (default), the feed change takes place in one time step whenever possible.

7.4 Injection BCs Injection BCs are a versatile boundary condition in Barracuda VR that allows for the injection of fluid, particles, or tracers to be specified at a given location, velocity, mass rate, and direction. The injection point can be at any location within the domain which makes injection BCs an important tool for adding nozzles, distributor shrouds, or other inlets that cannot be easily captured by the grid. Each injection BC specifies a mass rate and velocity of fluid or particles that can be distributed to multiple injection points. Furthermore, the injection locations can be imported and exported as a comma separated values (CSV) file which allows for creation and editing of injection locations in a spreadsheet program.

Particle injection The particle injection creates a source of Lagrangian particles within the model at the specified x , y , z location. The required inputs for a particle injection are • an x, y , z location at which particles are injected, • a particle injection mass flow rate • a particle velocity, • a particle species from which particle composition and particle size distribution are determined, • an average injection direction vector • expansion angles θE1 , θE2 , and αE1 which allow variance in particle injection direction about the specified direction vector • a particle injection temperature, • a specification of Lagrangian phase resolution ( Number density) from which the the size of computational particles is determined. The number of real particles in a computational particle is calculated as N cp = V cell /V particle × (Number density )−1 . Liquid Droplet Injection In versions of Barracuda VR prior to 17.0.0, Liquid Injection BCs were used to introduce liquid droplets into a gas-solid model. In Barracuda VR 17.0.0 and later, this functionality is still available, but is moved into the more general-purpose Injection BCs section of the GUI. One major change in work-flow, if you wish to inject liquid droplets, is that you will now need to define such droplets in the Particle Species section of the GUI. Then, in the Injection BCs dialog, create a particle injection BC and choose the droplet species. Liquid droplets are defined in the same manner as any other particle species. Instead of choosing a solid material, choose a liquid material when specifying the material composition of the particle species. Note, use of a 16-series or earlier project with liquid injection BCs in 17-series or later, requires users to redefine liquid injection BCs within this new framework. Please contact support if you would like assistance with this step. The liquid droplet model in Barracuda VR allows for the injection of liquid droplets into a gas-solid model. Once in the model domain, the liquid droplets can coat nearby solid particles, forming a liquid film on the outside. Particle-toparticle liquid film transfer is also modeled. The liquid film cannot, however, participate in chemical reactions. More detailed information on the model is given in [ZORourkeS09].

Tracer injection Tracer particles are Lagrangian entities that can be injected at a location for tracking fluid flow and fluid flow residence time in the model. They are treated as small particles that are similar to normal computational solid particles in the model except that they • have zero mass, momentum, energy

118



• flow freely with the fluid phase • do not have an injection direction or expansion (due to zero mass and momentum) The Lagrangian phase resolution of the tracers is determined through a specified number density.

Fluid injection Fluid injection is calculated as a source term in the mass, momentum, and energy equations of the Eulerian cell in which the injection point resides. The required inputs for a fluid injection are • an x, y , z location from which the injection cell is determined,

˙ inj , • an injection mass flow rate, m • a fluid velocity, u inj , • a fluid composition with mass fraction of xi for each component i, • an injection direction vector, n inj , • a fluid injection temperature, T inj The continuity equation for mass conservation for each chemical species i is modified by an additional source term S inj for each injection location

∂ (θf ρf,i ) + ∇ · (θf ρf,i uf ) = [Existing source terms ] + S inj,i ∂t

(7.1)

where

S inj,i = m ˙ inj xi /V cell ,

(7.2)

θf is the fluid volume fraction, ρf is the fluid density, uf is the fluid velocity vector, and V cell is the volume of the cell. In the momentum equations, the injection adds an additional force term, F inj .

∂ (θf ρf uf ) + ∇ · (θf ρf uf uf ) = [fluid stress terms + existing force terms ] + Finj ∂t

(7.3)

where Finj = m ˙ inj uinj ninj /V cell .

(7.4)

The energy equation is also modified to include the temperature-dependent enthalpy calculation for each chemical species that is injected.

∂ (θf ρf hf ) + ∇ · (θf ρf hf uf ) = [existing energy terms ] + q inj ∂t

(7.5)

where

q inj =

�

xi m ˙ inj hf,i (T inj )/V cell ,

(7.6)

i

hf is the total enthalpy in the cell, and hf,i (T inj ) is the enthalpy of species i at the injection temperature.

The Injection BCs Window Injection BCs are managed in the Injection BCs Window, shown in Fig. 7.15, which displays a list of all existing injection BCs and allows users to follow the Add, Edit, Copy, Delete GUI pattern within the model. Within the window, the injection BCs are displayed as individual lines. Each injection BC in the list has a unique ID number associated with it that is displayed alongside the injection BC type and other injection BC parameters.

7.4. Injection BCs

119


Fig. 7.15: Injection BCs Window with a sample injection BC

7.4.1 Specifying an injection BC All injection conditions parameters are defined in the Injection Specifications Dialog, shown in Fig. 7.16. This dialog is activated when either a new injection set is added or an existing injection set is edited. Injection sets can contain injections of only particles, only tracers, only fluid, or combinations of particles and fluid or tracers and fluid.

Fig. 7.16: Injection Specifications Dialog showing the sample specification of both particle and fluid injection There are three main sets of data that are entered in the Injection Specification Dialog: Particle/Tracer Injection , Fluid Injection, and the injection Locations. Once the values have been entered, clicking Apply will assign the injection BC to the model and close the dialog whereas clicking Cancel will close the dialog and any values will be lost.

Injection Name It is recommended that a unique, descriptive name be given to each injection set created in the project.

Comment Follows the Comment Field GUI pattern. Particle/Tracer Injection A particle feed in the injection BC can be enabled by selecting the Particle/Tracer Injection check box in the Injection

120



BC Editor and clicking Particle in the Type list. A tracer feed in the injection BC can be enabled by selecting Tracer in the Type list. Use BC Connector data Type, species, temperature, mass flow rate, and number density manual will be determined by the BC Connector. Velocity, expansion angles, and orientation angle still need to be specified in the Injection BC Editor. See BC Connections for setting up a BC Connector. subparagraph:Use file An SFF file, which is always recommended based on the ability to use Interact while a simulation is running, can be used to specify time-dependent particle or tracer injection BC parameters. The file will override specified values of velocity, mass flow rate, temperature, or number density that have been entered in the dialog. Additional information on the use of an SFF file at an injection BC is provided in Creating transient injection files. Use specified values Velocity The particle injection velocity should be entered in units of meters per second (m /s). Note that the value entered for velocity will be overridden if Use file is selected. A tracer injection does not need velocity to be specified. Mass flow The particle injection mass flow should be entered in units of kilograms per second (kg /s). Note that the value entered for mass flow will be overridden if Use file is selected. A tracer injection does not need mass flow to be specified. Temperature The particle injection temperature should be entered in units of kelvin (K). Note that the value entered for temperature will be overridden if Use file is selected. A tracer injection does not need temperature to be specified. Number density The volume of a computational particle fed at an injection BC is equal to the volume of the cell divided by the specified value of Number Density. With some arrangement, it can readily be derived that the number density is Number density =

Total real particles in feed Number of real particles per computational particle

�� θ˙ p

1

−

where θ˙ p is the volumetric

flow rate of particles per cell volume. A frequently used rule of thumb is to divide the total real particles into 50 computational particles, which is equivalent to setting the number density to 50/θ˙ p which is equal to 125 at a volume fraction of 0.4. Note that the value entered for number density will be overridden if Use file is selected. Tracer particles require a Number density to determine the number of tracers that will be injected at each location.

Injection type Type The drop-down menu allows selection of particles or tracer. Only one option is permitted per injection BC. Species The particle composition and size distribution of particles injected is determined from the particle species selected in the Species list. This list is populated with all species that have been defined in the model. Only one particle species can be selected per injection BC. Tracers can be viewed in GMV in the same manner that regular computational particles are viewed. Tracers are given species numbers of 2000 or higher to differentiate tracers from computational particles. The user has the option of specifying a species number for display in GMV in the Species field. See Particle Species for details on defining particle species. Expansion and orientation angles Individual computational particles and tracer particles are randomly injected in a direction that falls within within an elliptical injection cone that is constructed with expansion angles θ E1 and θE2 and orientation angle αE1 . The distance from the nozzle exit to the cone front surface is one cell length, and the cone is one half cell in length. Two expansion angles, θE1 and θ E2 , are needed to define the shape of the jet front. θ E1 is the angle from the injection direction to the intersection line between the x ′ -z ′ plane and the cone. Similarly θ E2 is the angle from the injection direction to the intersection line between the y ′ -z ′ plane and the cone. The orientation angle, αE 1 , is the rotation angle of the jet front (about the z ′ axis) in the local coordinates relative to the x′ axis. Typical usage is to specify a round cone in which θ E1 and θ E1 are both specified with an angle between 0 and 15, and α E1 has the default value of 0 . The effect of expansion angle is shown in Fig. 7.19 for symmetric cones with

7.4. Injection BCs

121


Fig. 7.17: Injection BC particle feed expansion angles

Fig. 7.18: Injection BC particle feed orientation angle

122



expansion angles θE = 0◦ θE = 15◦ , and θE = 30◦ .

Fig. 7.19: Effect of expansion angle parameter on a particle injection. Expansion angles of 0, 15, and 30 shown in (a), (b), and (c). The location and direction of the injection is marked by the red arrow For applications with elliptical sprays (fanning spray pattern for example) it can be useful to understand how the expansion angles are related to the global coordinate system in Barracuda VR. When the injection cone is first constructed, the cone is first built in the direction n init = (0,0,1) with θE1 angle aligned in the (cos αE1 , sin αE1 , 0) direction and the θE2 angle aligned in the (sin αE1 , cos αE1 , 0) direction. The cone is then rotated from n init to the specified direction vector n inj and translated to the injection location.

Flux plane options Follows the Flux Plane Options GUI pattern. Fluid injection A fluid feed in the injection BC can be enabled by selecting the Fluid Injection check box in the injection specification dialog. Use BC Connector data Fluid composition, velocity or area, mass flow, and temperature will be determined by the BC Connector. See BC Connections for setting up a BC Connector. Use file An SFF file, which is always recommended based on the ability to use Interact while a simulation is running, can be used to specify time-dependent fluid injection BC parameters. The file will override specified values of velocity or area , mass flow rate, and temperature that have been entered in the dialog. Additional information on the use of an SFF file at an injection BC is provided in Creating transient injection files. Use specified values Velocity The fluid injection velocity should be entered in units of meters per second (m /s). The velocity is applied to each individual injection location within the injection BC. Note that the value entered for velocity will be overridden if Use file is selected. Area Area should be entered in units of square meters (m 2 ). The specified area is the total area of all injection locations within the injection BC. Note that the value entered for area will be overridden if Use file is selected. Mass flow The fluid injection mass flow should be entered in units of kilograms per second (kg /s). Note that the value entered for mass flow will be overridden if Use file is selected. Temperature The fluid injection temperature should be entered in units of kelvin (K). Note that the value entered for temperature will be overridden if a fluid injection file is used for the injection BC.

7.4. Injection BCs

123


Fluid composition The composition of the fluid at the fluid injection boundary is specified by clicking on the Define fluids button which will raise the Applied Fluid Dialog, discussed in Applied Materials.

Locations The mass flow rate of particles or fluids can be distributed to multiple injection locations, which are specified in the Locations table, which is at the bottom of the Injection Specifications Dialog. The Reference Grid is useful for determining x, y , z locations within the model domain. The table can be also expanded for easier viewing and editing by clicking the Expand Locations Table button. The expanded form of the table is shown in Fig. 7.20,

Fig. 7.20: Injection locations table showing 8 sample injection locations Each row of the table contains an injection location and the necessary parameters for that location. Additional rows can be added by clicking the Add Row button and existing rows can be deleted by clicking the Delete Row button. Additionally, the injection locations can be imported from and exported to a comma separated values (CSV) file for editing and viewing in an external spreadsheet program by clicking the Import or Export buttons. The import and export of injection locations is discussed in Import/Export Injector Location Data. Name Each injection location must be given a unique identifier. The GUI will automatically name the injection location if no name is provided by the user. On/Off Injection locations can be individually enabled or disabled. Note that a specified injection will contribute to the fluid and particle weighting, even if an injection is disabled. X, Y, Z The x , y, z coordinates for each injection location are specified in units of meters (m). The Reference Grid button, located at the bottom of the dialog, can be useful for identifying the appropriate ( x, y , z ) coordinates for in jection locations. Injection locations can be placed in any computational cell within the simulation domain, regardless of whether the cell has walls or is a fully interior cell. Further, it is acceptable to place multiple injection locations at exactly the same ( x, y , z ) coordinate. The effects of overlapping injection locations will be additive. nx, ny, nz The values of nx , ny , and nz form the x , y , and z components directional vector for each injection location. The directional vector defined by nx, ny, and nz specifies the nominal flow direction of both particles and fluid injected. A situation that commonly occurs when defining injection BCs is that flow direction angles relative to the primary x , y , and z axes of the model are known, and need to be converted to vector components nx , ny , and nz . In this case, it is useful to think of the known angles in terms of spherical coordinates, such that φ is the polar angle (relative to the vertical axis, often the z -axis), and θ is the azimuthal angle (relative to one of the horizontal axes, often the x -axis). Given this reference frame, the values of nx , ny , and nz can be calculated as: nx = cos(θ)sin(φ)

124



ny = sin(θ)sin(φ) nz = cos(φ) Note that the vector formed by the nx, ny, and nz components does not need to be a unit vector. The formulas presented above happen to result in components that would create a vector with length 1, but this is not a requirement when defining the vector components. Particle Mass Weight The distribution of the total particle mass to the injection locations is determined by the Particle Mass Weight column, which will only be visible if a particle injection is specified. The particle mass flow rate at an injection location is:

m ˙ p,l = W p,l

m ˙ p,inj N

∑

W p,j

j=1

where:

m ˙ p,l is particle mass flow rate at injection location l W p,l is the particle mass weight specified for injection location l m ˙ p,inj is the injection BC particle mass flow rate N

∑

j=1

W p,j is the summation of all particle mass weights in the injection BC ( Particle Weight Sum)

The Particle Weight Sum is displayed at the bottom of the locations table. Particle Temp Multiplier Each particle injection location can have a unique temperature. This can be used to model temperature variations across a geometry. The particle injection location temperature is:

T p,l = M p,l · T p,inj where:

T p,l is the particle temperature at injection location l M p,l is the Particle Temp Multiplier specified for injection location l T p,inj is the particle injection temperature specified for the injection BC The Particle Temp Multiplier column will only be available if a particle injection is specified. Fluid Mass Weight The distribution of the total fluid mass to the injection locations is determined by the Fluid Mass Weight column which will only be visible if a fluid injection is specified. The fluid mass flow rate at an injection location is:

m ˙ f,l = W f,l

m ˙ f, inj N

∑

W f,j

j=1

where:

m ˙ f,l is the fluid mass flow rate at injection location l W f,l is the fluid mass weight m ˙ f, inj is the injection BC fluid mass flow rate N

∑

j=1

W f,j is the summation of all fluid mass weights in the injection BC ( Fluid Weight Sum)

The Fluid Weight Sum is displayed at the bottom of the locations table. Fluid Temp Multiplier Each fluid injection location can have a unique temperature. This can be used to model temperature variations across a geometry. The fluid injection location temperature is:

T f, l = M f,l · T f, inj 7.4. Injection BCs

125


where:

T f,l is the fluid temperature at injection location l M f,l is the Fluid Temp Multiplier specified for injection location l T f,inj is the fluid injection temperature specified for the injection BC The Fluid Temp Multiplier column will only be available if a fluid injection is specified.

7.4.2 Creating transient injection files When transient injection data is entered for an injection BC, it is done through the use of an SFF file in Barracuda VR. An existing SFF file can be added by clicking on the browse button in either the Particle/Tracer injection or Fluid Injection section of the Injection BC Editor, shown in Fig. 7.16. Clicking Edit will open an added file in the SFF Editor or create a new SFF file if the file name box is empty, shown in Fig. 7.21. The columns within the SFF files depend on whether the transient injection file is for an injection of particles, tracers, or fluid. For more information on creating and editing SFF files, see Tabular Input Using SFF Files.

Transient input data Time The time specifies when the parameters in the other columns should be used. The units of time can be either seconds (s) or minutes (min) and is used in particle, tracer , and fluid injection files. On/Off The On/Off column enables or disables the injection. 0 is the value attributed to Off while 1 is the value attributed to On. The value of the On/Off column is linearly interpolated and the injection will be enabled if the value is larger than 0.5. The On/Off column is used in particle, tracer , and fluid injection files. Temperature The temperature of the injected particles or fluid can be set in units of kelvin (K), Fahrenheit (F), or Celsius (C) and is used in the particle and fluid injection files. Velocity The velocity of the injected particles or fluid can be set in units of meters per second (m /s), feet per second (ft/s), or centimeters per second (cm /s) and is used in the particle and fluid injection files. Area The area of the injection locations can be set in units of square meters (m 2 ), square centimeters (cm 2 ), square millimeters (mm 2 ), square inches (in 2 ), and square feet (ft 2 ), and is used in the fluid injection files. Mass flow rate The mass flow rate can be specified in units of kilograms per second (kg /s) and is used in the particle and fluid injection files. Number density manual The number density manual controls the rate at which computational particles are injected and is used in the particle and tracer injection files.

Fig. 7.21: SFF Editor showing sample particle injection file with a ramp in mass flow rate up to 5 kg/s over the first ten seconds of simulation.

126



7.4.3 Import/Export Injector Location Data Barracuda VR allows users to import and export injection location data to and from the injection location table to a spreadsheet application. This feature is useful for editing large sets of location and direction data within an injection set. The typical approach is to first add one or more locations to the GUI directly and then to export those locations to a CSV file. This ensures that the correct column formatting is used in the spreadsheet program for later import of the data back into Barracuda VR.

Exporting location dataData from the Locations table can be exported as a *.csv file by clicking on the Export button in the GUI. This will first open a file dialog for selecting a filename and location for the file. After a filename is selected, the user will be presented with the export preview dialog, shown in Fig. 7.22 which will allow the user to select a delimiter from a choice of comma, semicolon, or space. Typically a comma is selected as a delimiter.

Fig. 7.22: CSV export preview dialog showing file formatting with a comma delimiter applied to a sample file of eight injection locations

Importing location data Location data stored in a * .csv file can be imported into the Locations table by clicking on the Import button. The user will be first prompted to select the file from a file dialog which will open the import preview dialog, shown in Fig. 7.23. In the import preview dialog, the user should select the delimiter that matches the delimiter used by the spreadsheet program. When correctly selected, the import preview window should display a total of 12 columns. If the file has header information included, this should be ignored by selecting the Ignore First Line check box.

7.5 Thermal Wall BCs A thermal wall applies a specified temperature to a solid wall within a model. Depending on the temperature difference between the wall and fluid near the wall, energy will either be transferred into or out of the model through the thermal wall. The rate of heat transfer is determined by fluid-to-wall heat transfer equations discussed in Heat Transfer Models. Heat transfer can also occur through the use of the radiation model in Barracuda VR. Thermal walls are managed in the Thermal Wall BCs Window, shown in Fig. 7.24, which displays a list of all existing thermal wall BCs and allows users to follow the Add, Edit, Copy, Delete GUI pattern within the model. Within this window, the thermal wall conditions are displayed as individual lines. Each thermal wall BC in the list has a unique ID number associated with it that is displayed alongside the thermal wall location and other thermally wall parameters.

Radiation model for walls Barracuda includes a radiation model for radiative heat transfer which can be turned On or Off from within the Thermal wall BCs Window. The radiation model considers radiation between a thermal wall and the particle phase only and it

7.5. Thermal Wall BCs

127


Fig. 7.23: CSV import preview dialog showing file formatting with a correctly identified comma delimiter for an example CSV file

Fig. 7.24: Thermal Wall BCs Window showing sample thermal wall added to the model

128



does not consider radiative heat transfer between particles, between walls, or between the wall and fluid. The radiation between a thermal wall cell and nearby particles, q wp , is calculated as

�

4

q wp = Aw F wp εwp σ T w − T p

4

�

where A w is the area of the thermal wall, T w is the temperature of the wall, T p , is the mass-weighted average of particles in a cell, F wp is a calculated view factor, σ is the Stefan-Boltzmann constant, and εwp is the effective emissivity between the wall and the particles in a cell.

εwp =

�

1 1 + −1 ε p εw

�

1

−

In the calculation of the effective emissivity, ε w is the specified emissivity of the wall and ε p is the volume-weighted average of particle emissivities (see Specifying particle species). For any radiation model, the calculation of view factors can be computationally expensive so to maintain the efficiency of the calculation, the Barracuda VR radiation model only looks at cells near the thermal wall and calculates F wp based on the particle volume fraction, particle diameters, and local geometry.

Output options for wall temperatures Barracuda VR allows users to set the output heat transfer as Flow rate in units of joules per second (J /s) or as Flux in units of joules per square meter per second (J /s/m2 ). The thermal wall data will be written to the trans.data** file and to GMV output files if the Output transient data automatically check box is selected.

7.5.1 Specifying a thermal wall Thermal wall BC parameters are defined in the Wall Temperature dialog, shown in Fig. 7.25. This dialog is activated when either a new thermal wall BC is added or an existing thermal wall BC is edited. Once the values have been entered, clicking Apply will assign the thermal wall BC to the model and close the dialog whereas clicking Cancel will close the dialog and any values will be lost.

Enabling a thermal wall The thermal wall can be enabled by selecting the Enabled (On at start of calculation) check box. The default is for the thermal wall to be enabled. If not selected

Variable or Constant wall temperature Temperature file An SFF file, which is always recommended based on the ability to use Interact while a simulation is running, can be used to specify time-dependent thermal wall temperatures. The file will override any specified values that have been entered in the dialog. Additional information on the use of an SFF file at a thermal wall BC is provided in Defining a transient thermal wall. Temperature The Specify option is the thermal wall temperature used for calculation of heat transfer between the thermal wall and the fluid and particles in the model. Any temperature specified will be overridden if a transient temperature file is used in the model. Temperatures are entered in units of kelvin (K).

Properties Surface area The surface area used in heat transfer calculations at the thermal wall BC can be either Automatically calculated or Specify. If the area is automatically calculated, Barracuda VR will use the surface area of the gridded surface which may be slightly different than the original surface area due to the cutting, merging, and removal of cells. If the actual gridded area is known, it can be manually entered by clicking the Specify radio button and entering the surface area manually in units of square meters (m 2 ). Emissivity The emissivity of the thermal wall is used when the radiation model is enabled, as discussed in Thermal Wall BCs. The default value of emissivity is 1.0 .

7.5. Thermal Wall BCs

129


Fig. 7.25: Wall Temperature Dialog showing sample thermal wall BC

130



Location Thermal wall BCs are placed at cells specified by the range of i , j , k coordinates entered. A thermal wall BC will be applied to all exterior cell faces with normal directions that are aligned with the Direction specified and are within the range of i , j , k coordinates. The range of coordinates is from i 1 to i 2 , j 1 to j 2 , and k 1 to k 2 in the x , y , and z directions. Proper coordinate values can be determined using the Reference Grid . Note that the dialog will also accept min or [ for the minimum coordinate value and max or ] for the maximum coordinate value in the model. Heat

transfer direction Once the cells containing the thermal wall faces are identified, the Heat transfer direction is used to identify which walls should be changed to thermal wall BC faces. There are multiple choices for direction: x-, y-, z- direction Often the quickest way of applying a thermal wall BC to the face is to select either an x- , y- , or zdirection. This will apply the thermal wall BC to any face that has a normal vector that is in the selected x , y , or z direction. Normal to surface When Normal to surface is used, the thermal BC will be applied to all surfaces within the specified range of i, j , k coordinates. If Normal Limit is selected, all available faces within the cell range are used for the thermal wall BC. Alternatively, Z zero, Y zero, or Z zero can be selected to eliminate any faces in the respective direction.


7.5.2 Defining a transient thermal wall When transient flow data is entered for a thermal wall BC, it is done through the use of an SFF file in Barracuda VR. An existing SFF can be added to the model by clicking the browse button in the Thermal Wall BC Editor, Fig. 7.25. Clicking Edit will open an added file in the SFF Editor or create a new SFF file if the file name box is empty. The columns within the SFF file, shown in Fig. 7.26, are Time and Temperature. For more information on creating and editing SFF files, see Tabular Input Using SFF Files.

Transient input data Time The time specifies when the parameters in the other columns should be used. The units of time can be either seconds (s) or minutes (min). Temperature The temperature of the thermal wall which can be entered in units of kelvin (K), Fahrenheit (F), or Celsius (C).

7.6 Passive Scalars Passive scalars are an Eulerian field variable that can be injected with any flow BC to provide a means of measuring the residence time distribution (RTD) of the fluid phase in a model. The passive scalars are transported with the fluid phase but since they are passive, they do not participate in the calculation of particle-fluid physics in Barracuda VR. When passive scalars are used in the model, the value of the scalar is output to any flux planes which allows the passive scalars to be used as fluid tracer for analysis of the residence time distribution within a model. Passive scalars are managed in the Passive Scalars Window, shown in Fig. 7.27, which displays a list of all existing passive scalars and allows users to follow the Add, Edit, Copy, Delete GUI pattern within the model. Within the window, the passive scalars are displayed as individual lines. Each passive scalar has a unique ID number associated with it that is displayed alongside the passive scalar location and other passive scalar parameters.

7.6. Passive Scalars

131


Fig. 7.26: SFF Editor showing sample thermal wall BC file with a ramp in temperature from 500 K up to 800 K over the first 60 seconds of simulation

Fig. 7.27: Passive Scalar BCs Window showing a sample passive scalar added to the model

132



7.6.1 Specifying a passive scalar All passive scalar BC parameters are defined in the Passive Scalar BCs Dialog, shown in Fig. 7.28. This dialog is activated when either a new passive scalar BC is added or an existing passive scalar BC is edited. Once the values have been entered, clicking Apply will assign the pressure BC to the model and close the dialog whereas clicking Cancel will close the dialog and any values will be lost.

Fig. 7.28: Passive Scalar BCs Dialog showing an example passive scalar

i, j, k A passive scalar will be injected with any existing flow BC that is within the range of coordinates specified. The passive scalar will not create a new boundary condition. The range of coordinates is from i1 to i 2 , j 1 to j2 , and k1 to k2 in the x , y , and z directions. Proper coordinate values can be determined using the Reference Grid and also note that the dialog will accept min or [ for the minimum coordinate value and max or ] for the maximum coordinate value in the model.

File An SFF file, which is always recommended based on the ability to use Interact while a simulation is running, can be used to specify time-dependent passive scalar function. The file will override the specified Constant feed value that has been entered in the dialog. The typical usage is to use a transient passive scalar file to specify a pulse input of the passive scalar for tracking of the residence time. The magnitude of the pulse is arbitrary, but it is often helpful to have a pulse with a short width and a known integrated value for later analysis. Additional information on the use of an SFF file at a passive scalar BC is provided in Creating a transient passive scalar input file

Constant feed value 7.6. Passive Scalars

133


It is possible to specify a constant value for the passive scalar feed by clicking the Constant feed value radio button and entering the value in the text field. The typical usage, however, is to specify a pulse using a passive scalar SFF file, discussed above. If a constant feed value is used, the magnitude is arbitrary and will only affect the magnitude of passive scalar values passing through the flux planes. Diffusion coefficient Additional diffusion of the passive scalar can be added by changing the value of the passive scalar Diffusion Coefficient . Additional diffusion is not necessary in typical models and therefore the default diffusion coefficient of 0.0 is a recommended value. Trigger value Use of a trigger value will enable cell level tracking indicating whether the value of the passive scalar has ever exceeded the specified value of the trigger with the data set available for viewing in GMV. When using a passive scalar BC to track the Residence Time Distribution (RTD) for the fluid, it is recommended that the Trigger value be left as the default of 0.0 . Fluid slip ratio When using a passive scalar to track the residence time distribution (RTD) for the fluid, the default Fluid slip ratio value of 1.0 is recommended.


7.6.2 Creating a transient passive scalar input file When a transient passive scalar file is created for a passive scalar BC, it is done through the use of an SFF file in Barracuda VR. An existing SFF can be added to the model by clicking the browse button in the Passive Scalar BC Dialog, Fig. 7.28. Clicking Edit will open an added file in the SFF Editor or create a new SFF file if the file name box is empty. The only columns within the SFF file, shown in Fig. 7.29, are Time and Scalar value. For more information on creating and editing SFF files, see Tabular Input Using SFF Files.

Transient input data Time The time specifies when the passive scalar values should be used. The units of time can be either seconds (s) or minutes (min). Scalar value The scalar value column is where the feed of passive scalar values is defined. The feed rate of the passive scalar tracer into the domain is equal to the scalar value multiplied by the fluid flow rate. The value specified has units of 1/m3 .

7.6.3 Analyzing passive scalar data Passive scalar data is output as a separate column in each flux plane file in the model with the heading "Time integration of all scalar(s) crossing flux plane". As the heading indicates, the flux plane tracks the cumulative amount of the passive scalar that has passed through the flux plane. This data can be used in the raw format or be analyzed to determine a residence time distribution. Cumulative passive scalar data The passive scalar data printed to the flux planes is useful in its raw form for determining the amount of passive scalar that may be still outstanding in the model by comparing with input flux plane totals. Residence time distribution The flow rate of a passive scalar tracer passing through a flux plane can be determined by calculating a time-derivative of the passive flux plane data. Typically, this can be done in XMGR or with any common scripting language. Information on the tracer residence time can be determined by examining the shape and location of the outlet peak compared to the inlet peak. An example of the residence time distribution analysis of passive scalar data is shown in Fig. 7.30.

134



Fig. 7.29: SFF Editor showing sample passive scalar pulse that starts at 1 second into the simulation and lasts for 0.1 seconds

Fig. 7.30: Plotting the derivative of passive scalar data at input and output gives information on residence time distribution within the fluid phase

7.6. Passive Scalars

135


136


CHAPTER

EIGHT

BC CONNECTIONS

BC Connections allow outlet boundary conditions (BCs) to pass information to inlet BCs. Most often, the outlet BC will be a Pressure BC, and the inlet BC will be either a Flow BC or Injection BC. The information passed by BC Connections can include fluid properties, such as temperature, composition, and flow rate, and/or particle properties, such as size, temperature, ID, composition, and flow rate. BC Connections allow some information to be manipulated before being passed on to the inlet BC, such as scaling mass flow rates, offsetting temperature values, and resetting particle residence times. BC Connections are limited to one-way communication: outlet BCs pass information to inlet BCs, but inlet BCs do not pass information back to outlet BCs. Many-to-many relationships are possible, with one or more outlet BCs connected to one or more inlet BCs allowing for simulation of multi-stage cyclones and external circulation loops.

How BC Connectors Track Information Particle and/or fluid mass is tracked via a profile, or snapshot, with a duration and time stamp. Each BC Connector input has a profile for tracking current time step fluid and particle mass movements. All profiles at BC Connector inputs are combined and queued for processing each time step. Fluid mass and Particle mass are treated separately; no interaction between the two phases is available. Fluid attributes tracked through the BC Connector are limited to mass, composition and thermal characteristics (pressure and volume are not available). On the other hand, a BC Connector tracks individual computational particles, including properties such as particle id, mass, composition, volume and other fundamental attributes not dependent on control volume properties. Whether a BC Connector tracks fluid, particles, or both, depends on the BC Connector flags set for each BC Connector output. If any output uses fluid from the BC Connector, then the BC Connector tracks fluid. If any output uses particles from the BC Connector, then the BC Connector tracks particles. Secondary Feeds also promote BC Connector tracking characteristics.

8.1 Secondary Feeds A Secondary Feed can be thought of as a Flow BC with no physical location in the domain, used specifically for direct input to a BC Connector. A Secondary Feed connects to a BC Connector in the same way as any other outlet BC from the simulation, and can only be used as an input to the BC Connector. Fig. 8.1 shows the Secondary Feeds Manager , which follows the Add, Edit, Copy, Delete GUI pattern. Fig. 8.2 shows the Secondary Feed Editor , which is used to specify the properties of a Secondary Feed . Flux Plane Every Secondary Feed must have a flux plane defined, and the Flux plane name is used to identify and select it for use in a BC Connector. The flux plane definition follows the Flux Plane Options GUI pattern. Comment This field follows the Comment Field GUI pattern. Applied fluid materials If a Secondary Feed is used to feed fluid into a BC Connector, i.e. the specified fluid flow rate is non-zero, then the fluid composition must be defined. This follows the Applied Materials GUI pattern.

137


Fig. 8.1: Secondary Feeds Manager

Fig. 8.2: Secondary Feed Editor

138

Chapter 8. BC Connections


Applied particle species If a Secondary Feed is used to feed particles into a BC Connector, i.e. the specified particle flow rate is non-zero, then the particle species composition must be defined. This follows the Applied Materials GUI pattern. Behavior at boundary This is where the fluid and/or particle flow properties are specified, and follows the Tabular Input Using SFF Files GUI pattern. The fluid and particle properties are controlled independently.

8.2 BC Connectors The BC Connectors section of the Barracuda VR GUI is where connections between outlet BCs and inlet BCs are defined. Fig. 8.3 shows the BC Connector Manager , which follows the Add, Edit, Copy, Delete GUI pattern.

Fig. 8.3: BC Connector Manager Fig. 8.4 shows the BC Connector Editor , which is used to specify a BC Connector.

Fig. 8.4: BC Connector Editor

8.2. BC Connectors

139


Enabled By default, a newly created BC Connector will have the Enabled box checked, and the BC Connector will be active in the simulation. If the box is un-checked, the BC Connector definition will be stored in the project, but will not be active in the simulation. A disabled BC Connector cannot be re-enabled on restart.

8.2.1 BC Connection Input from Domain (outlet BCs) The inputs to a BC Connector are defined in the BC Connection Input from Domain (outlet BCs) section of the BC Connector Editor . This section follows the Add, Edit, Copy, Delete GUI pattern, and allows one or more domain outlet BCs to be selected as inputs to the BC Connector. When the Add button is clicked, the BC Connector Input Editor window is raised, allowing the user to select an available flux plane name. The selected flux plane name is then added to the list of inputs for the current BC connector. The Add process can be repeated if multiple flux planes should be combined as inputs to the BC Connector. Advanced feature: Flow BC flux planes available as inputs By default, only Pressure BCs will be listed in the BC Connector Input Editor . However, if this box is checked, then outlet Flow BCs will also be made available as BC Connector inputs.

Fig. 8.5: BC Connector Input Editor To be listed and available for usage in a BC Connector, a simulation outlet must be either a Pressure BC with the Flux plane name defined or a Flow BC with the Flux plane name defined. In either case, only when fluid is flowing out of the simulation domain does it enter the BC Connector. In-flow at the boundary condition takes nothing from the BC Connector and records zero mass to the BC Connector; i.e., a BC Connector allows flow in only one direction, inputs to outputs. Particle split factor This allows computational particles entering the BC Connector to be split, creating more computational particles at the BC Connector outlet. If the split factor is 1, then no splitting occurs. If 2, then each computational particle exiting the simulation will be split into 2 computational particles with equal real particle counts, conserving mass. Non-integer values are handled by tracking the remainder for use in the next time step. Range: [1, inf)

140



8.2.2 BC Connector Properties The BC Connector Properties section of Fig. 8.4 contains the required Name field, and allows for manipulation of some fluid and particle properties before they are passed to the BC Connection outputs. Name This is a required field, and is used as a Flux plane name associated with the BC Connector. It is recommended to use the same naming convention as discussed in the Flux Plane Options GUI pattern. Comment This follows the Comment Field GUI pattern. Time delay A non-zero value forces fluid and particle profiles to remain in a BC Connector some number of seconds before being filtered and staged to BC Connector outputs. Note that staging does not guarantee feed to simulation; for example, if particles are staged to be fed, but the cells into which they are fed are already at the particle close-pack volume fraction, no particle feed will be possible. Fluid temperature control By default, a SCALE value of 1 is used for BC Connectors, meaning that no manipulation of the fluid temperature will take place. However, three options are available if it is desired to change the fluid temperature between the BC Connector inlets and outlets: • SCALE: The mass-weighted average BC Connector inlet fluid temperature is multiplied by the specified Fluid temperature factor , and the result is assigned to the BC Connector outlet. • SET: The BC Connector outlet fluid temperature is directly set to the specified Fluid temperature. • OFFSET: The specified Fluid temperature offset is added to the mass-weighted average BC Connector inlet fluid temperature, and the result is assigned to the BC Connector outlet. Fluid Filter Specifies how the fluid mass from BC Connector inlets is distributed to BC Connector outlets. The option available is: • SCALE: Clicking the Fluid Filter button will raise a dialog in which the Scale factor can be specified. The BC Connector inlet fluid mass is multiplied by the Scale factor , and evenly distributed among all specified BC Connector outlets. No mass is sent to the Exit Flux Plane. The valid range for Scale factor is [0, inf). Particle temperature control Since computational particles are Lagrangian, their treatment in BC Connectors is more sophisticated than for the fluid phase. Whereas the fluid properties from BC Connector inlets are averaged using a mass-weighting, the computational particle properties do not require any such averaging. Instead, properties are maintained distinctly for each computational particle, and are manipulated individually. By default, a SCALE value of 1 is used, meaning that no manipulation of the computational particle temperatures will take place. However, three options are available if it is desired to change the computational particle temperatures between the BC Connector inlets and outlets: • SCALE: The inlet particle temperature is multiplied by the specified Particle temperature factor , and the result is assigned to that particle at the BC Connector outlet. • SET: The BC Connector outlet particle temperature (for all particles, in this case) is directly set to the specified Particle temperature. • OFFSET: The specified Particle temperature offset is added to each particle‘s inlet temperature, and the result is assigned to that particle’s BC Connector outlet temperature. Reset particle residence By default, this box is un-checked, and particles are passed from the BC Connector inlets to the BC Connector outlets without changing their Residence Time values. If this box is checked, the Residence Time value for each particle is set to zero when it is fed back into the simulation domain at a BC Connector outlet. Draw connectors for post-processing By default, this box is checked, and extra polygons are included in the Gmv * output files so that BC Connectors can be visualized. If the box is un-checked, the BC Connector will still function as expected, but the polygons for visualization will not be generated.

8.2. BC Connectors

141


8.2.3 BC Connection Output to Domain (inlet BCs) The outputs from a BC Connector are defined in the BC Connection Output to Domain (inlet BCs) section of the BC Connector Editor . This section follows the Add, Edit, Copy, Delete GUI pattern, and allows one or more domain inlet BCs to be selected as outputs from the BC Connector. When the Add button is clicked, the BC Connector Output Editor window is raised, allowing the user to select an available flux plane name. The selected flux plane name is then added to the list of outputs for the current BC connector. The Add process can be repeated if multiple flux planes should be treated as outputs from the BC Connector. To be listed in the BC Connector Output Editor , a simulation inlet must be either a Flow BC with the flux plane name defined or an Injection BC with the flux plane name defined, and the Use BC Connector data option must have been selected for the BC’s Fluid behavior at boundary or Particle behavior at boundary options.

Fig. 8.6: BC Connector Output Editor Feed limits At the time of simulation feed, all fluid mass in a profile will be fed to the simulation unless otherwise restricted by feed limits. Similarly, all particles in a profile will be fed to the simulation unless otherwise restricted by feed limits; additionally, the close pack volume fraction will always be observed while feeding to the simulation. Fluid mass and particles can build up at BC Connector outputs. Any particle not fed to a simulation will be staged at the BC Connector output for use in the next time step. Fluid is staged with total elapsed time maintained, subsequent feeds of fluid use a fraction of available fluid depending on current time step and available profile time. • Particle volume fraction limit When feeding both particles and fluid to a simulation from a BC Connector output, the maximum allowed particle volume fraction is the minimum of this volume fraction and the close

142



pack particle volume fraction. For example, if a value of 0.1 is specified here, then particles will not feed past 10% of the cell volume at this BC Connector-simulation boundary. Range: [0,1] • Fluid volume fraction limit When feeding both particles and fluid to a simulation from a BC Connector output, this is the maximum allowed fluid volume fraction. This parameter is useful for balancing fluid and particle flow rates; at cyclone diplegs, for example, it is frequently desired to feed the particles at a certain volume fraction, say 0.4, but it is not known beforehand what the corresponding fluid flow rate should be. By setting the Fluid volume fraction limit to a value of 0.6 , the BC Connector will automatically limit the fluid flow rate to a value that results in a particle volume fraction of 0.4. Range: [0,1] • Particle mass flow limit (kg/s) When feeding particles at a BC Connector outlet, the maximum allowed mass flow rate can be limited. The default value is set to 1e+20, which is effectively no maximum limit applied. Range: [0,inf) • Fluid mass flow limit (kg/s) When feeding fluid at a BC Connector outlet, the maximum allowed mass flow rate can be limited. The default value is set to 1e+20, which is effectively no maximum limit applied. Range: [0,inf) • Particle mass threshold off (kg) When feeding particles at a BC Connector outlet, this parameter can be used to specify a mass threshold below which the particle feed should be turned off. The mass referred to in this case is the particle mass within the BC connector at a given point in time. This parameter can be used in conjunction with Particle mass threshold on (kg) to mimic the operation of cyclone dipleg flapper valves. Range: [0,inf) • Partilcle mass threshold on (kg) When feeding particles at a BC Connector outlet, this parameter can be used to specify a mass threshold above which the particle feed should be turned on. The mass referred to in this case is the particle mass within the BC Connector at a given point in time. To mimic the behavior of a cyclone dipleg flapper valve, this parameter can be used in conjunction with the Particle mass threshold off (kg) value. Range: [0,inf) Example: Suppose a cyclone dipleg flapper valve stays closed until 10 kg of particle mass has built up in the dipleg. At this mass, the flapper valve opens and particles begin to drain out. When the particle mass in the dipleg falls below 8 kg, the flapper valve closes again. This situation can be approximated by setting the parameters to: Particle mass threshold off (kg) = 8 Particle mass threshold on (kg) = 10

• Slip The ratio of particle velocity to fluid velocity at the BC Connector outlet BCs. Range:[0,inf) Particle split factor This allows computational particles exiting the BC Connector to be split, creating more computational particles at the BC Connector outlet. If the split factor is 1, then no splitting occurs. If 2, then each computational particle exiting the simulation will be split into 2 computational particles with equal real particle counts, conserving mass. Non-integer values are handled by tracking the remainder for use in the next time step. Range: [1, inf)

8.2. BC Connectors

143


144


CHAPTER

NINE

CHEMISTRY

Chemical reactions can affect all aspects of the behavior of a fluidized bed and it is important to consider the effects of chemical reactions alongside the particle-fluid dynamics and heat transfer being modeled in Barracuda VR. Depending on the system being considered, the chemical reactions may be closely coupled with the particle-fluid dynamics and heat transfer within the bed. For example, a reaction that produces or consumes gases from solids will create a gas volume change that will affect that fluidization within the bed. Conversely, the reaction rate and reactant availability will be a strong function of the gas mixing produced by the fluidization pattern. In thermal models, the reaction rates are strongly coupled to the temperatures in the bed through both temperature-dependent reaction rate expressions and the exothermicity or endothermicity of a reaction. In many cases, a fluidized bed model is much more realistic when the chemical reactions within the system are considered. The range of different reaction types and forms of reaction rate expressions in literature is very broad due to the fact that chemical reaction expressions and mechanisms are sourced from both theoretical derivations and experimental correlations. Furthermore, the chemistry module in Barracuda VR can be effectively used to model many non-reacting processes including phase changes and physical adsorption and desorption of gases. To accommodate these varying requirements, there are multiple methods for inputting and calculating chemical reactions within Barracuda VR. To understand the suitability of these different methods, it is necessary to first identify the characteristics of reactions commonly found in fluidized bed models. Typical chemistry in fluidized bed reactors can be divided into two categories: homogeneous reactions, which occur in the fluid phase alone; and heterogeneous reactions, in which a solid participates as either a reactant, product, or catalyst. The heterogeneous reactions can be divided into additional categories of deposition reactions, consumption reaction, catalytic reaction, and solid reactions. Homogeneous reaction (fluid phase) This reaction occurs in the fluid phase and involves fluid phase reactants and products only. The reaction rate may be dependent on the fluid phase reactant compositions, temperature, or other fluid properties. Deposition reaction This is a heterogeneous reaction which occurs at the surface of a particle and produces at least one solid product from fluid phase reactants. The reaction rate may be dependent on the particle surface area, particle temperature, and fluid phase reactant concentrations. Polysilicon deposition is an example of this type of reaction. Consumption reaction This is a heterogeneous reaction which occurs at the surface a particle and produces fluid phase products from at least one solid reactant. The reaction rate may be dependent on the mass of solid reactant, particle temperature, and the concentration of any fluid phase reactants. Coke combustion in FCC regeneration is an example of this type of reaction. Catalytic reaction This is a heterogeneous reaction with fluid phase reactants and products that is dependent on the presence of a solid component (catalyst) for the reaction to proceed. While the reaction rate is likely dependent on the mass or surface area of catalyst that is present, the catalyst will not be consumed or produced as a result of the reaction. The reaction rate may also be dependent on the fluid phase reactant concentrations or temperatures. Solid reaction (“solids producing solids”) A heterogeneous reaction which occurs at a particle surface or within the particle volume, and involves both solid reactants and solid products. The reaction rate is frequently dependent on the

145


mass of solid reactants, concentrations of fluid phase reactants, and particle or fluid temperatures. The adsorption of a gas onto a solid sorbent is an example of this type of process.

Approaches for calculating chemistry Barracuda VR provides two approaches to calculating chemical reactions within a model. The first approach is volume average chemistry in which the reaction is calculated at the cell level (Eulerian). Because of this, volume average chemistry is best used for homogeneous and catalytic reactions which occur predominately in the fluid phase. While it is possible to use volume average chemistry for other heterogeneous reactions as well, doing so is not preferable since all solids-dependence will be based on cell level averages of particle properties rather than individual particle properties. For most heterogeneous reactions, discrete particle chemistry is the preferred approach. Discrete particle chemistry is calculated at each computational particle (Lagrangian) within the model using individual computational particle temperature, mass, and other properties. While this approach may have a slightly higher computational cost, this cost is frequently outweighed by the benefit of the increased resolution in the model. Discrete particle chemistry cannot be used to model a homogeneous reaction.

Reaction input forms A fully-described reaction specifies the reactants, products, and rate at which the reaction proceeds. In Barracuda VR, there are two approaches for specifying this reaction information. The first is the stoichiometric approach in which the reaction is written as a stoichiometric equation, and the reaction rate is specified for the reaction as a whole. For example, a steam gasification reaction of solid carbon would be written in stoichiometric form as r

C(s) + H2 O −−→ CO + H2

r = 219

3

m T exp kg·K·s

�

−22645 K T

�

ρC [H2 O]

where T is the temperature, ρ C is the local bulk density of carbon, and [ H2 O] is the concentration of water vapor in the gas phase. Alternatively, in the species or differential form, the reaction rate of each reactant or product is explicitly stated as either a reaction rate expression or as a function of of the reaction rate of a different component. For example, the steam gasification reaction could be written in species form as

�

−22645 K T d [H2 O] d [C(s)] = dt dt d [CO] d [C(s)] =− dt dt d [H2 ] d [C(s)] . =− dt dt

d [C(s)] = −219 dt

m3 T exp kg·K·s

�

ρC [H2 O]

While both equation forms describe the same reaction, the stoichiometric form is simpler to enter, and likely more familiar, whereas the species form provides more flexibility. In the Reactions window, discussed in Reactions, volume average reactions can be specified with either approach. Discrete particle reactions must be entered in the species form.

Reaction rate coefficients In Barracuda VR, Reaction rate coefficients are created to describe reaction rate dependencies on temperature, pressure, fluid density, fluid volume fraction, and particle characteristics such as size and composition. These coefficients are then used as building blocks in the full reaction rate expressions. For example, the steam gasification reaction

146

Chapter 9. Chemistry


would be defined in Barracuda VR through the use of a reaction rate coefficient as Rate coefficient:

k0 = 219

Stoichiometric equation:

m3 T exp kg·K·s

�

−22645 K T r

�

ρC

C (s) + H2 O −−→ CO + H2

r = k 0 [H2 O].

Reaction rate:

Note that the gas concentration is not included in the rate coefficient, but is instead included in the reaction rate expression directly. See Rate Coefficients for more details.

9.1 The Chemistry Window The Chemistry window, shown in Fig. 9.1, displays settings that apply to all chemical reactions.

Importing chemistry from a project file Clicking on the Import Materials/Chemistry from another project button will open the Import From Project dialog, discussed in Importing chemistry from another project , through which a user can import existing chemical reactions and base materials from another project file.

Volume-Average Chemistry Reaction Type When using chemical reactions, Barracuda requires that the user only enter all volume-average reactions in either a Stoichiometric or Species form. In most cases, the user will choose the default Stoichiometric setting. The selected option does not affect discrete particle chemistry, which must always be entered in species form. Gas Transport

Limiter For gas phase chemistry, an extreme rate is not a big concern. Gases homogeneously occupy a volume and their reactions do not increase gas mass or energy in the volume. For gas-solid chemistry, particles are at discrete locations which requires gas to move to a particle to react, and the solid-gas chemistry produces or consumes gases and increases or decreases the gas energy. The model is only applicable to first order gas-solid reactions. If the reaction rates for gas-solid chemistry are very high, using this model can help the solution of the reaction equations. The solution for the transport limiter has the appearance of a shrinking-core model. However, the mass transport coefficient considers gas in a cell flowing to a particle. A heterogeneous reaction rate may be limited by the mass transfer of gas-phase reactants to the particle surface. The gas transport limiter considers this potential reduction in a heterogeneous reaction rate by calculating a effective reaction rate coefficient. The reaction rate equation for a discrete particle is:

dC = −k [C] dt ′

kg s

where the gas-transport-limited reaction rate coefficient, k ′ , is calculated as: ′

k =

�

1 1 + k hm A p

�

1

−

(9.1)

where k is the specified reaction rate, hm is the mass transfer coefficient, and A p is the particle surface area. The mass transfer coefficient is calculated as:

hm =

9.1. The Chemistry Window

Sh Dm

l

147


Fig. 9.1: Chemistry Window

148



1/3 l = min(2d min(2d p, 0.1 V cell ), and the Sherwood number is calculated as: where D m is the diffusion coefficient, l = cell

0.6 Re1/2 Sc1/3 Sh = 2 + 0. where Reynolds number, Re, is based on particle size and the turbulent Schmidt number approximation is Sc = 0.9.

Distribute Sensible Heat from Reaction to Particle Phase Discrete particle reactions will consume or produce sensible heat that is distributed to either the particle or the surrounding fluid. Automatic and manual settings are available to specify this distribution. Automatic Automatic (default) Distribution (default) Distribution is calculated according to the model of [MSSH15]. Reactions occur on the particle and sensible heat is exchanged with the fluid by • convection convection of reactant gases to the particle particle at the fluid temperature; and • convection convection of product gases back to the fluid at the particle temperature. temperature. Manual A Manual A fixed percentage of the sensible energy is distributed to the particle phase according the amount specified in the input box. The remaining amount is distributed to the fluid. For example, • 0% = Sensible energy is distributed distributed to the fluid. Particle temperature temperature will not change due to reaction but may change due to heat transfer or other phenomena. • 50% = Sensible heat of reaction is distributed evenly to the particle and fluid. Actual change in temperature will likely be different due to differences in thermal mass between the particle and fluid. • 100% = Sensible Sensible energy is distributed distributed completel completely y to the particle particle.. Fluid Fluid temperat temperature ure will not change change due to reaction but may change due to heat transfer or other phenomena.

Chemistry ODE Settings These settings control the operation of the CVODE solver used solver used for volume-average chemistry. A useful discussion of the tolerance settings can be found on the CVODE FAQ page. page. Tolerance : the maximum • Relative Tolerance: maximum allowed allowed relative relative tolerance tolerance for the iterati iterative ve solution solution of the ODE set. The Sundials webpage recommends using a value less than 1e-03, but greater than the machine precision (1e-15). Tolerance: the maximum allowed absolute tolerance for the iterative solution of the ODE set, specif• Absolute Tolerance: ically useful when dealing with numbers very close to zero. Steps: the maximum number • Max Number Steps: number of iterations iterations the ODE solver will take for a given given computationa computationall cell. After this many steps are taken, if the solution is still not converged, Barracuda will output the “ODE solver failed” message, abandon the chemistry solution for the failed cell, and move on to the next computational cell.

9.2 Rate Coeffic Coefficients ients Rate coefficients are defined functions in Barracuda VR to specify reaction rate dependence on temperature, pressure, fluid density, fluid volume fraction, and particle characteristics such as size and composition. Once defined, a reaction rate coefficient can be used in one or more reaction rate expressions. Rate coefficients are managed in the Fig. 9.2 9.2,, which displays a list of all Rate Coefficients window, Coefficients window, shown in Fig. existing volume-average and discrete rate coefficients, and follows the Add, Edit, Copy, Delete GUI pattern. pattern. Within Within this window, window, each rate coefficient is displayed on an individual line. Each rate coefficient has a unique name of the form k# which is used to reference reference rate coefficien coefficients ts in a reaction reaction rate expression expression.. The formatting formatting of reaction reaction rate expressions is discussed in detail in the stoichiometric, species, and discrete sections below. Rearranging Rate Coefficients The Barracuda VR GUI appends each newly defined Rate Coefficient to to the list of currently defined coefficients; the ID number is incremented for each Rate Coefficient automatically. automatically. In some cases, it is useful to rearrange the coefficients so that they are in a better logical order. The Up and Down arrows to the right

9.2. Rate Coefficients

1 49


Fig. 9.2: Rate Coefficient Window Window with a volume average average rate coefficient added of the Add , Edit , Copy, and Delete buttons buttons allow for such rearrangement. The currently selected Rate Coefficient is moved up or down in the list.

9.2.1 Definin Defining g chemistry rate coefficient coefficients s Rate coefficient properties are entered in the Rate Coefficient Dialog, shown in Fig Fig.. 9. 9.3 3. This dialog dialog is activated activated when either a new rate coefficient is added or when an existing rate coefficient coefficient is edited. While a common dialog is used to define both volume-average and discrete rate coefficients, some of the solid units will change depending on the selected rate coefficient type. As such, the user must select whether the rate coefficient will be used in Volume-Average or Discrete reaction rates.

Selecting the rate coefficient Type Multiple Types of rate coefficients are available in Barracuda VR to accommodate the wide range of possible reaction dependencies. Arrhenius Chem Rate The Rate The Arrhenius form is the most frequently used rate coefficient form as it provides functionality for temperature, pressure, fluid density, and fluid volume fraction through the c 0 , c 1 , c 2 , c 3 , and c 4 constants. The solids terms are added by clicking the Solids Dependence button which raises the Solids Dependence Dialog, discussed in Adding solids dependence to a reaction rate. The discrete discrete chemistr chemistry y form also includes includes the additional additional Np/V dependence dependence through the c5 constant that is needed for converting some rate expressions to a discrete chemistry form (discussed in Conversion between volume average and discrete chemistry ). Note that activat activation ion energy energy term, exp(E/T exp(E/T + E 0 ), does not contain a universal gas constant, R, in the denominator. Volume average: Discrete:

k = c = c 0 T c P c ρf c θf c exp(−E/T + E/T + E 0 ) { solids terms } 1

2

3

4

k = c = c 0 T c P c ρf c θf c (N p /V ) /V )c exp(−E/T + E/T + E 0 ) { solids terms } 1

2

3

4

5

Polynomial The polynomial form allows the specification of a fourth-order temperature polynomial through the c0 , c1 , c2 , c 3 , and c 4 constants along with solid dependence. The solids terms are added by clicking the Solids Dependence button which raises the Solids Dependence Dialog, discussed in Adding solids dependence to a reaction rate . When discrete chemistry is used with the polynomial form, an additional Np/V dependence through the c5 constant is included for converting some rate expressions to a discrete chemistry form (discussed in Conversion between volume average and discrete chemistry ). Volume average: Discrete:

k = (co + c1 T + c2 T 2 + c3 T 3 + c4 T 4 ) {solids terms }

k = (co + c1 T + c2 T 2 + c3 T 3 + c4 T 4 ) (N p /V ) /V )c {solids terms } 5

Table Based Chemistry The Table Based Chemistry reaction rate coefficient type allows the user to define a rate coefficient based on a tabulated relationship between an independent variable and the value of the coefficient. Linear interpolation is performed between the rows in the table, and the first and last entries of the table are used when the independent variable falls outside the range of the table.

15 0



Fig. 9.3: Rate Coefficient Coefficient Dialog


1 51


Depending on the Coefficient is for reaction type selection, the independent variable will have either a volumetric basis (for Volume-Average) or not (for Discrete). The Unit dropdown dropdown box will show units that reflect this difference between the two types of chemistry. The Import and and Export buttons buttons can be used to read and write the table in comma-separated value (CSV) format. This is useful if the table has been created in a spreadsheet program, in which case it can be exported in CSV format, and directly imported to Barracuda VR. Catalyst Deactivation This is a reaction rate coefficient with a form commonly used to describe the deactivation of fluid catalytic cracking (FCC) catalyst as coke is deposited on the particles [BPN+94]. In literature, the deactivation coefficient is commonly written as:

φ =

BC + 1 BC + exp( exp(AC C ci ci )

In the Barracuda VR GUI, the coefficient is not denoted as φ, but rather as k0 , k1 , etc., similar to other reaction rate coefficients. The default values of A C = 4.29 and B C = 10 10..4 are based on [NJR05]. These values can be changed by the user. The C ci percentagee of one or more solid species, species, relative relative to the total mass of all solid species. species. In ci term is a mass percentag literature, the species is often simply “coke”, but in Barracuda VR the user can specify any species currently in the project Base Materials list. At least one material must be selected. If the coefficient is being used for volume-average chemistry, the C ci ci term is defined as:

C ci ci =

Mass of selected solid species in cell × 100 Mass of all particles in cell

If the coefficient is being used for discrete particle chemistry, the C ci ci term is defined as:

C ci ci =

Mass of selected solid species on particle × 100 Mass of particle

Specifying rate coefficient units Reaction rate coefficients can be formatted in a variety of units, which are specified in the dropdown boxes on the right side of the dialog. When calculating the rate coefficient, Barracuda VR will convert a property to the specified units and use that value in the calculation of the rate coefficient. Gas phase units The units The value of the rate coefficient can be calculated for different units of temperature, pressure, and fluid density. The units available for selection are: • Temperature: kelvin kelvin (K ), ), degrees Fahrenheit ( F ), ), or degrees Celsius ( C ); ); kPa), pounds per square inch ( psi), bars, or atmospheres ( atm); • Pressure Pressure:: pascals ( Pa), kilopascals ( kPa

• Density: Density: kilogram kilogramss per cubic meter meter ( kg/m^3), grams per cubic centimeter ( g/cm^3), pounds per cubic foot (lb/ft^3). Solids units for a volume average rate coefficient The units of solids dependence terms of diameter, mass, and area are also specified in the Rate Coefficient Dialog. For a volume-average volume-average rate coefficient, the units available for selection are: • Diameter Diameter:: meters ( m), centimeters ( cm), microns, feet ( ft ), ), or inches (in); • Mass: Mass: kilogr kilograms ams per cubic cubic meter meter ( kg/m^3), grams per cubic centimeter ( g/cm^3), or pounds per cubic foot (lb/ft^3); • Area: Area: square square meters meters per cubic cubic meter ( m^2/m^3), square centimeters per cubic centimeter ( cm^2/cm^3), square feet per cubic foot ( ft^2/ft^3), or square inches per cubic inch ( in^2/in^3). Solids units for a discrete particle rate coefficient The units of solids dependence terms of diameter, mass, and area are also specified in the Rate Coefficient Dialog. For a discrete rate coefficient, the units available for selection are: • Diameter Diameter:: meters ( m), centimeters ( cm), microns, feet ( ft ), ), or inches (in);

15 2



• Mass: kilograms (kg), or grams (g), or pounds (lb); • Area: square meters (m^2), square centimeters ( cm^2), square feet ( ft^2), or square inches ( in^2).

Adding solids dependence Dependence on solids mass, area, volume fraction, and diameter are added by clicking on the Solids Dependence button, which opens the Solids Dependence Dialog. The addition of solids dependence terms through this dialog is discussed further in Adding solids dependence to a reaction rate . Temperature weighting For discrete rate coefficients and volume average rate coefficients with a solids term, the temperature used in the calculation of a rate coefficient is a blend of the fluid temperature and particle temperature which is determined by the Fluid temperature weighting factor and the Particle temperature weighting factor , located at the bottom of the Rate Coefficient Dialog. The temperature used within a rate coefficient, T , is calculated as T = w fluid T fluid + wparticle T particle where T particle is the cell-average temperature of all particles in a cell for a volume average rate coefficient or the individual particle temperature for a discrete chemistry rate coefficient; T fluid is the fluid temperature of the cell; and the weighting factors w particle and w fluid are the values specified in the dialog. For a volume average rate coefficient with no solids terms, the temperature weighting functionality will be inactive and T = T fluid .

9.2.2 Adding solids dependence to a reaction rate Solids dependence is added to a rate coefficient by clicking on the Solids dependence button in the Rate Coefficient Dialog which raises the Solids Dependence Dialog, shown in Fig. 9.4. The Solids Dependence Dialog is used for both volume average and discrete particle rate coefficients to create solids dependence terms based on the volume fraction, area, diameter , or mass of all solids or any individual solid component in the Solid Species List . The solid term can also be raised to an exponent.

Fig. 9.4: Solids Dependence Dialog with a carbon mass term with exponent of 1 added to the Species List On the right side of the dialog is the Solid Project Species List , which contains all solid materials that have been Base Materials window and an additional component “all”. The “all” component defined in the project through the represents the summation of all solids components in a cell (for a volume-average rate coefficient) or the summation of all components in a particle (for a discrete particle rate coefficient). Adding a solids dependence term A solids-dependence term is added to the rate coefficient by selecting a solid component or “all” from the Solid Project Species List and clicking the Import button to add the term to the Species List . Once a term has been added, the user then selects material coefficient type from the Material coefficient type dropdown box and specifies the Exponent on material.


153


Managing solids dependence terms Solids terms that have been added to the Species List can be copied or deleted by clicking the Copy and Delete buttons. Alternatively, users can replace a species in the Species List by selecting a new species from the solid project species list, and clicking on the Replace button. This will replace the component type while retaining the coefficient type and exponent.

Material coefficient type The material coefficient type can be either the volume fraction ( volfrac), surface area ( area), diameter ( diam), mass, m/m0, 1-m/m0, m0, -LN(1-m/m0), or -LN(m/m0) of the selected solid. The definition of these terms depends on whether it added to a volume average or discrete particle rate coefficient and whether the term applies to an individual component or to all solids in the system. The solids terms are calculated for every cell in volume average chemistry and for every particle in discrete particle chemistry and are therefore dependent on the particle properties within each scope. All coefficient type definitions are calculated in terms of the following variables: • a p , the surface area of particle p; • m p , the mass of particle p; • m p,i , the mass of solid component i on particle p; • V cell , the volume of the cell in which the chemistry is being calculated; • v p , the volume of particle p; and • ρi , the density of solid component i. Volume fraction For both volume-average and discrete particle chemistry, the solid volume fraction term is calculated as the volume of selected solid component in a cell divided by the volume of the cell, as shown in ( 9.2). When “all” is used, the volume fraction is the volume of all particles within the cell divided by the volume of the cell. The volume fraction term is dimensionless and has no units.

θi =

1 V cell

� p

θall =

m p,i for selected component i ρi

1 V cell

�� p

i

(9.2)

m p,i for all ρi

Area For volume-average chemistry, the area term is calculated for individual and “all” terms as shown in (9.3). The individual term is summation of all particle areas, averaged by the volume of the selected component i, divided by the volume of the cell. If “all” is used, the area term is the total area of all particles in a cell divided by the volume of the cell. The units of the volume average area are square meters per cubic meter ( m^2/m^3), square centimeters per cubic centimeter ( cm^2/cm^3), square feet per cubic foot ( ft^2/ft^3), or square inches per cubic inch ( in^2/in^3) which are specified in the Rate Coefficient Dialog, discussed in Defining chemistry rate coefficients .

ai,volavg =

1 V cell

� p

m p,i a p for selected component i ρi v p

aall,volavg =

1 V cell

�

(9.3)

a p for all

p

For discrete particle chemistry, the solids area term is surface area of the particle multiplied by the mass fraction of the selected component i on the particle. If “all” is used, the area term is the surface area of the particle, as shown in (9.4). The units of the discrete particle area are square meters ( m^2), square centimeters ( cm^2), square feet ( ft^2), or square inches (in^2) which are specified in the Rate Coefficient Dialog, discussed in Defining chemistry rate coefficients .

ai,discrete =

m p,i a p for selected component i m p

(9.4)

aall,discrete = a p for all Diameter For volume-average chemistry, the diameter term is calculated as a Sauter mean diameter. For an individual component, this is calculated as the volume of solid component i in the cell divided by the volume-weighted surface

154



area of component i in the cell. If “all” is used, then the diameter term is calculated as the ratio between the volume of all solids in the cell divided by the total solid surface area. Both definitions are shown in ( 9.5). The units of the volume average diameter are meters ( m), centimeters ( cm), microns, feet ( ft ), or inches (in) which are specified in the Rate Coefficient Dialog, discussed in Defining chemistry rate coefficients .

di,volavg = 6

�� p

m p,i ρi

p

dall,volavg = d 32 = 6

1

−

� ∑ ∑

m p,i a p ρi v p

for selected component i

p v p

p a p

(9.5)

for all

For discrete particle chemistry, the diameter term is the diameter of the particle multiplied by the mass fraction of the selected component i on the particle or the diameter of the particle itself if “all” is selected, as shown in (9.6). The units of the discrete particle diameter are meters ( m), centimeters ( cm), microns, feet ( ft ), or inches (in) which are specified in the Rate Coefficient Dialog, discussed in Defining chemistry rate coefficients .

di,discrete =

m p,i d p for selected component i m p

(9.6)

dall,discrete = d p for all Mass For volume-average chemistry, the solids mass term is calculated as the total mass of the selected solid component i in all particles within a cell divided by the volume of the cell, as shown in (9.7). When “all” is used, the mass term is the mass of all components in all particles within a cell divided by the volume of the cell. The units of the volume average mass are kilograms per cubic meter ( kg/m^3), grams per cubic centimeter ( g/cm^3), or pounds per cubic foot (lb/ft^3) which are specified in the Rate Coefficient Dialog, discussed in Defining chemistry rate coefficients .

mi,volavg =

1 V cell

�

m p,i for selected component i

(9.7)

p

mall,volavg =

1 V cell

�� p

m p,i for all

i

For discrete particle chemistry, the solids mass term is the mass of the selected component i in the particle or the mass of the entire particle if “all” is selected. The units of the discrete particle mass are kilograms ( kg), grams (g), or pounds (lb) which are specified in the Rate Coefficient Dialog, discussed in Defining chemistry rate coefficients .

mi,discrete = m p,i for selected component i mall,discrete =

�

(9.8)

m p,i for all

i

m/m0 For discrete particle chemistry, this term can be used to define chemical reactions that depend on the conversion of a certain component. m is the current mass of the particle or the specific material(s) chosen in the Species List. m0 is explained below. 1-m/m0 For discrete particle chemistry, this term is a more convenient form to use if the reaction rate depends inversely on the current mass of the particle or the selected species. m0 For discrete particle chemistry, this term is the initial mass of a particle or the specific material(s) chosen in the Species List with default units of kg. Note that the Age Factor defined for each material in the Particle Species definition will be multiplied by m0. This allows initialization of “old” particles in a simulation. For example, suppose an Age Factor of 2 has been set for Carbon in a particular particle species, and a reaction rate depends on m0 of Carbon. When calculating the reaction rate, m0 will be multiplied by 2, because the Carbon has been set to behave as if it originally had twice as much mass. -LN(1-m/m0) For discrete particle chemistry, this term is the natural log of (1-m/m0). -LN(m/m0) For discrete particle chemistry, this term is the natural log of (m/m0).

Solids term exponent 9.2. Rate Coefficients

155


Users may enter an exponent to be included with the solids term in this textbox. Exponent values can be positive or negative. Note that any solids terms that evaluate as “0” and have a negative exponent will cause the entire rate coefficient to be zero.

9.3 Reactions All chemical reactions in a model are managed in Barracuda VR in the Reactions window, shown in Fig. 9.5. The window displays a list of all existing volume-average and discrete reactions that have been defined in the model and allows users to add, edit, copy, and delete reactions in the model. Reactants and products are displayed either in a stoichiometric form (for volume average reactions) or in a differential form (discrete particle chemistry) along with the reaction rate expression and user comments.

Fig. 9.5: Reactions Window with sample reactions for coal gasification Adding a reaction New reactions can be added by clicking the Add button which will display a dropdown box from which the user selects either a Volume Average Stoichiometric Equation, Volume Average Species Rate Equation , or Discrete Particle Rate Equation. This will raise the appropriate reaction dialog for the reaction type selected. The volume average stoichiometric dialog, volume average species dialog, and discrete particle reaction dialog are discussed in Volume Average Stoichiometric Reaction Dialog , Volume Average Species Reaction Dialog , and Discrete Reaction Dialog, respectively. Note: Only one type of volume average reaction (stoichiometric or species) will be available at a time. This is selected in the Chemistry window, discussed in The Chemistry Window. Editing a reaction Existing reactions can be edited by selecting the reaction from the list and clicking on the Edit button or by double clicking on the reaction line in the list. Doing so will raise the appropriate equation editor dialog for the reaction type. The volume average stoichiometric dialog, volume average species dialog, and discrete particle

156



reaction dialog are discussed in Volume Average Stoichiometric Reaction Dialog , Volume Average Species Reaction Dialog, and Discrete Reaction Dialog, respectively. Copying a reaction An existing reaction may be duplicated by selecting it from the list and clicking on the Copy button. This will create a new reaction line with a unique ID number that is otherwise identical to the original reaction. Deleting a reaction A reaction can be removed from the reactions list by selecting it with the mouse and clicking on the Delete button.

9.3.1 Volume Average Stoichiometric Reaction Dialog The Volume Average Stoichiometric Reaction dialog, shown in Fig. 9.6, provides an interface for defining the stoichiometry, rate expression , and equation units for a volume average reaction in stoichiometric form. This editor is displayed when a volume average stoichiometric reaction is added or edited from the Reactions Window.

Fig. 9.6: Volume Average Stoichiometric Reaction Dialog showing sample reaction for water-gas shift reaction

Defining reaction stoichiometry The stoichiometric equation defines the reactants and products of a reaction. For a volume average reaction, this is entered into the reaction text box in a standard chemistry form in which reactants are on the left side, products are on the right side, and reactants and products are separated by a right arrow. Fig. 9.6 shows a sample reaction in which CO and H2O are reactants while CO2 and H2 are products. Other valid stoichiometric equations are the deposition of

9.3. Reactions

157


solid silicon from silane gas, SiH4 => Si(s) + 2 H2; and the combustion of solid carbon, C(s) + 0.5 O2 => CO. The reaction mass balance must be satisfied for a reaction to be valid. Formatting the stoichiometric equation There are a few conventions that have been adopted for entering a stoichiometric equation in Barracuda VR: • the right arrow separating reactants and products is represented as “ =>”; • any solid material is noted as such by placing a “ (s)” after the chemical name; • stoichiometric coefficients are placed before a chemical name, separated by a space; and • when multiple reactants or products are present in a reaction, they are separated by a “ +”. The formatting and mass balance of a stoichiometric equation can be verified by clicking on the Check button. Adding chemical names Stoichiometric coefficients, plus signs ( +), and the right arrow ( =>) must be typed directly into the reaction rate textbox; a chemical name, however, can be typed directly or be added by clicking the Add Chemical button. When clicked, this button will raise a material selection dialog from which a chemical name can be selected. This will then place the chemical name, with correct formatting, at the cursor location in the text box.

Specifying the reaction rate The reaction rate equation determines the speed at which a reaction occurs and may be dependent on the concentrations of reactants, temperatures, pressures, particles properties, or other factors. Gas concentration terms are used directly in a Barracuda VR rate equation whereas other fluid and solids factors are captured through the use of rate coefficients. The creation and management of rate coefficients is discussed further in Rate Coefficients. Formatting the rate expression Rate coefficients and chemical species can be used to create complex rate expressions within Barracuda VR. The expanded format for a rate expression is shown in ( 9.9), where k represents any valid rate coefficient, [C ] represents any valid chemical species and c is any constant. Furthermore, building blocks of “basic groups”, BG; “summation groups”, SG; and “product groups”, PG, are used to describe the full range of reaction rate expression formats. Reaction rate = PG + PG + ...

(9.9)

PG = SG c ∗ SGc ∗ ... SG = BG + BG + ... BG = c ∗ (c ∗ k + c ∗ k + ...) ∗ [C ]c ∗ [C ]c ∗ ... The Barracuda VR GUI will accept a wide range of reaction rate expressions. During the syntax-checking process, expressions are transformed into a standardized format by the GUI. The examples below show both the original expression typed by the user, and the resulting (equivalent) expression after transformation by the GUI. Table 9.1: Rate Expression Parser

Typed by User

Transformed by GUI

R00 = k0 * [A] R00 = ( k0*[A] ) /(1+k1*[B]) R00 = (k0[A])/(1 + k1[B])/(k2*[B]+k3*[C])^2 R00 = k0*[A]*[B]+k1*[C]*[D] R00 = (k0*[A][B])^2+ ( 1 ) /(k1*[C][D])

(k0)[A] ((k0)[A])*(1 + (k1)[B])^-1 ((k0)[A])*(1 + (k1)[B])^-1*((k2)[B] + (k3)[C])^-2 ((k0)[A][B])+((k1)[C][D]) ((k0)[A][B])^2+(1)*((k1)[C][D])^-1

Note: Bold parenthesis ( ) in the examples above are required by the GUI in order to parse the expressions correctly. If the GUI is showing error messages about a particular reaction rate expression you are trying to enter, try using an extra set of parenthesis around your groups. This often helps the parsing process. Adding gas concentration terms Gas concentration terms can be either a molar concentration, mass concentration, mass fraction, or partial pressure, as defined in the Equation Units within the Reaction Rate Dialog. Regardless of the definition, the format of the gas concentration is the same: a valid gas species name inside of square brackets, [ ].

158



The name of the chemical species can be typed directly into the reaction rate text box or can be added using the Add Chemical button. When clicked, this button will raise a material selection dialog from which a valid chemical name can be selected. This will then place the selected chemical name, with correct formatting, at the cursor location in the text box. Adding a rate coefficient A rate coefficient can be added by typing the rate coefficient name or by clicking the Add Volume-Average Coefficient button. Only volume average rate coefficients may be used in a volume average reaction rate.

Setting equation units The units of the reaction rate and gas concentration terms are selected from the Equation Units dropdown boxes. When the reaction rate expression is evaluated, internal gas concentration values will be converted based on the type and units selected and the calculated value will be used within the expression. After evaluation, changes in reactant and product concentrations within a cell will be calculated based on the units of the reaction rate selected. The units of any rate coefficients within the reaction rate expression are implied from the user-selected units of gas concentrations and reaction rate. Reaction rate units Possible reaction rate units are mol/m3/s, mol/m3/min, mol/cm3/s, mol/cm3/min, and kmol/m3/s. Gas species units The gas species terms can be either in units of molar concentration ( mol/m3, mol/cm3, kmol/m3), mass concentration ( kg/m3, g/cm3), mass fraction, or partial pressure ( Pa, kPa, atm , bar , psi).

9.3.2 Volume Average Species Reaction Dialog The Volume Average Species Reaction Dialog, shown in Fig. 9.7, provides an interface for defining the reaction rate of a single chemical component. This editor is displayed when a volume average species reaction is added or edited from the Reaction Window.

Fig. 9.7: Volume Average Species Reaction Dialog

9.3. Reactions

159


Selecting a species The first step in defining the reaction rate for a volume-average species reaction is to select the species upon which the rate will be based. This is accomplished by clicking on the Select species button and selecting the appropriate species from the list.

Specifying the reaction rate The reaction rate equation determines the speed at which a reaction occurs and may be dependent on the concentrations of reactants, temperatures, pressures, particles properties, or other factors. Gas concentration terms are used directly in a Barracuda VR rate equation whereas other fluid and solids factors are captured through the use of rate coefficients. The creation and management of rate coefficients is discussed further in Rate Coefficients. Formatting the rate expression Rate coefficients and chemical species can be used to create complex rate expressions within Barracuda VR. The expanded format for a rate expression is shown in ( 9.9), where k represents any valid rate coefficient, [C ] represents any valid chemical species and c is any constant. Furthermore, building blocks of “basic groups”, BG; “summation groups”, SG; and “product groups”, PG, are used to describe the full range of reaction rate expression formats. Adding gas concentration terms Gas concentration terms can be either a molar concentration, mass concentration, mass fraction, or partial pressure, as defined in the Equation Units within the Reaction Rate Dialog. Regardless of the definition, the format of the gas concentration is the same: a valid gas species name inside of square brackets, [ ]. The name of the chemical species can be typed directly into the reaction rate text box or can be added using the Add Chemical button. When clicked, this button will raise a material selection dialog from which a valid chemical name can be selected. This will then place the selected chemical name, with correct formatting, at the cursor location in the text box. Adding a rate coefficient A rate coefficient can be added by typing the rate coefficient name or by clicking the Add Volume-Average Coefficient button. Only volume average rate coefficients may be used in a volume average reaction rate.

Setting equation units The units of the reaction rate and gas concentration terms are selected from the Equation Units dropdown boxes. When the reaction rate expression is evaluated, internal gas concentration values will be converted based on the type and units selected and used within the expression. After evaluation, changes in reactant and product concentrations within a cell will be calculated based on the units of the reaction rate selected. The units of any rate coefficients within the reaction rate expression are implied from the user-selected units of gas concentrations and reaction rate. Reaction rate units Possible reaction rate units are mol/m3/s, mol/m3/min, mol/cm3/s, mol/cm3/min, and kmol/m3/s. Gas species units The gas species terms can be either in units of molar concentration ( mol/m3, mol/cm3, kmol/m3), mass concentration ( kg/m3, g/cm3), mass fraction, or partial pressure ( Pa, kPa, atm , bar , psi).

9.3.3 Discrete Reaction Dialog The Discrete Particle Reaction dialog, shown in Fig. 9.8, provides an interface for defining the stoichiometry and reaction rate expression for a discrete particle reaction. This editor is displayed when a discrete particle reaction is added or edited in a model. Discrete particle reactions are specified in a differential form in which the stoichiometry of a reaction through differential rates of each components. In the carbon combustion example shown in Fig. 9.8, the stoichiometry of the equation C + O2 → CO2 is entered into Barracuda VR as

d[C(s)]/dt = negative reaction rate d[O2]/dt = 1 × d[C(s)]/dt

d[CO2]/dt = −1 × d[C(s)]/dt

Specifying the reaction rate 160



Fig. 9.8: Discrete Reaction Dialog showing sample reaction for carbon combustion

9.3. Reactions

161


In discrete particle chemistry, the reaction rate, which determines the speed at which a reaction occurs, is defined for a single solid component. The solid component, which may be either a reactant or product, is then used to define the stoichiometry through participating reactions. The reaction rate equation may be dependent on the concentrations of reactants, temperatures, pressures, particles properties, or other factors. The convention used in Barracuda VR is that gas concentration terms are used directly in the rate expression whereas other fluid and solids dependencies are captured through the use of rate coefficients. The creation and management of rate coefficients is discussed further in Rate Coefficients. Formatting the rate expression Rate coefficients and chemical species can be used to create complex rate expressions within Barracuda VR. The expanded format for a rate expression is shown in ( 9.9), where k represents any valid rate coefficient, [C ] represents any valid chemical species and c is any constant. Furthermore, building blocks of “basic groups”, BG; “summation groups”, SG; and “product groups”, PG, are used to describe the full range of reaction rate expression formats. Adding gas concentration terms Gas concentration terms can be either a molar concentration, mass concentration, mass fraction, or partial pressure, as defined in the Equation Units within the Reaction Rate Dialog. Regardless of the definition, the format of the gas concentration is the same: a valid gas species name inside of square brackets, [ ]. The name of the chemical species can be typed directly into the reaction rate text box or can be added using the Add Chemical button. When clicked, this button will raise a material selection dialog from which a valid chemical name can be selected. This will then place the selected chemical name, with correct formatting, at the cursor location in the text box. Adding a rate coefficient A rate coefficient can be added by typing the rate coefficient name or by clicking the Add Discrete Coefficient button. Only discrete particle rate coefficients may be used in a discrete particle reaction rate.

Defining reactants and products The remaining reactants and products in a discrete particle reaction are defined by adding Participating Reactions which define the production or consumption of a material as a function of the main solid reaction rate. In the example shown in Fig. 9.8, the reaction rate of oxygen, O2, and carbon dioxide, CO2, is specified as a function of the reaction rate of carbon.

Fig. 9.9: Participating reactions editor showing sample participating reaction Adding a participating reaction New participating reactions can be added by clicking the Add button which will raise the Participating Reactions Editor, shown in Fig. 9.9. In this dialog, a participating chemical species is selected by clicking the button the left (showing d[O2(G)]/dt in Fig. 9.9) and entering a coefficient that scales the participating reaction rate to the main solid reaction rate. Editing a reaction Existing participating reactions can be edited by selecting the reaction from the list and clicking on the Edit button or by double clicking on the reaction line in the list. Doing so will raise the Participating Reactions Editor, shown in Fig. 9.9. Copying a reaction An existing participating reaction may be duplicated by selecting it from the list and clicking on the Copy button. This will create a new participating reaction line with identical settings to the original.

162



Deleting a reaction A participating reaction can be removed from the reactions list by selecting the it with the mouse and clicking on the Delete button.

Setting equation units The units of the reaction rate and gas concentration terms are selected from the Equation Units dropdown boxes. When the reaction rate expression is evaluated, internal gas concentration values will be converted based on the type and units selected and used within the expression. After evaluation, changes in reactant and product concentrations within a cell will be calculated based on the units of the reaction rate selected. The units of any rate coefficients within the reaction rate expression are implied from the user-selected units of gas concentrations and reaction rate. Reaction rate units Possible reaction rate units are mol/s, mol/min, kmol/s, kg/s, and g/s. Gas species units The gas species terms can be either in units of molar concentration ( mol/m3, mol/cm3, kmol/m3) mass concentration ( kg/m3, g/cm3) mass fraction, or partial pressure ( Pa, kPa, atm , bar , psi ). Shrinking Core

Model Barracuda VR has a shrinking core model for more accurate modeling of some types of reactions. For example, analysis of partially reacted carbon particles shows an ash region surrounding an unreacted core ( [YK55]; [Lev72]). The Barracuda VR shrinking core model is on a per particle basis and assumes that the solid material in a particle reacts in the presence of a gas species. The shrinking core model can only be used with discrete particle chemical reactions in Barracuda VR. The rate of reaction is controlled by the first-order reaction rate, the transport of the gas through the non-reacting material to the core, and the transport of gas the boundary layer. Each particle has its own history and a “fresh” particle will have a higher reaction rate than an “old” particle. To use the shrinking core model, check the Shrinking Core Model checkbox, and specify a diffusion coefficient for the non-reacting material.

9.4 Conversion between volume average and discrete chemistry Often reactions will have rate expressions that are measured based upon the volume of gas that is present and specify that a prescribed amount of material per volume of gas will be consumed, produced, or otherwise transformed per division of time. These reaction rate expression will often have units such as mol/m 3 /s or kg/m3 /s and will often work seamlessly with volume average chemistry in Barracuda VR. A different convention, however, has been adopted for discrete particle reactions in Barracuda VR. Since a discrete particle reaction is applied to an individual particle, it makes sense to specify the amount of material (mass or mole) that will be consumed, produced, or otherwise transformed per division of time for that particle. In this case, the reaction rate units may be mol/s, kg/s, or similar. Unfortunately, it will be necessary at times for the user to convert a rate coefficient with a volume-basis to a discrete form suitable for use with discrete particle chemistry. To derive equations for this conversion, the volume of a cell is divided into separate control volumes for each particle in the cell. If N p is the number of particles in a cell with volume V cell , then the control volume for each particle is vcp = V cell /N p . Using this control volume, the mass and area density around each discrete particle becomes

ρs = m p /vcp = m p N p /V cell

and

as = a p /vcp = m p N p /V cell

(9.10)

where ρ s is the equivalent volume-average solids density, m p is the mass of the particle, a s is the equivalent volume average area density, and a p is the surface area of the particle. Similarly, for a volume-average particle reaction with a gas phase basis occurring in the cell, r , the reaction rate on the particle itself is calculated as

r p = rθ f vcp = rθ f V cell /N p

or

r = r p θf 1 N p /V cell . −

(9.11)

A volume-average reaction rate can be converted to a discrete particle reaction by substituting the relationships in (9.10) and (9.11). The table shown in Fig. 9.10 shows the conversions needed, where N p is the number of individual particles in the cell and V is the cell volume.

9.4. Conversion between volume average and discrete chemistry

163


Fig. 9.10: Volume Average to Discrete Chemistry Conversion Table Conversion of chemistry from volume average to discrete form is demonstrated in the following two examples.

Conversion example #1 In the first example, a simple reaction rate expression for steam gasification is converted to a discrete particle form. The natural units of the reaction rate are mol/m 3 /s. Reaction rate with volume basis

r =

�

m3 219 kg · K · s

� � T exp

−22645 K T

�

ρC [H2 O]

Discrete chemistry conversion To convert the reaction rate to a discrete form, (9.10) (for mass) and (9.11) are substituted into the reaction rate equation. 1

−

r p θf N p /V cell =

�

219

m3 kg · K · s

� � T exp

−22645 K T

�

mC N p /V cell [H2 O]

Rearranging and canceling out N p /V cell terms yields the final form of the discrete particle reaction rate:

r p =

�

m3 219 kg · K · s

�

θf T exp

�

−22645 K T

�

mC [H2 O]

Conversion example #2 A more complicated reaction rate is used in the second conversion example. Unlike the first example, the N p /V cell will not cancel out and would need to be included in the reaction rate coefficient. The natural units of the reaction rate are mol/m3 /s. Reaction rate with volume basis

r =

164

c0 ρs 2 [A] 1 + c1 as [B] Chapter 9. Chemistry


Discrete chemistry conversion To convert the reaction rate to a discrete form, (9.10) and (9.11) are substituted into the reaction rate equation. 1

−

r p θf

c0 (m p N p /V cell )2 [A] N p /V cell = 1 + c1 a p N p /V cell [B]

Rearranging terms yields the final form of the discrete particle reaction rate:

r p =

c0 θf N p /V cell m p 2 [A] 1 + c1 N p /V cell a p [B]

9.5 Importing chemistry from another project The Import Chemistry Dialog allows the user to import the reactions and base materials from an external project file to the current project file. The Import Chemistry Dialog, shown in Fig. 9.11, is opened by clicking on the Import Chemistry button on the Chemistry window. The Chemistry Window is discussed in detail in The Chemistry Window. Note Since the importing of chemistry cannot be undone, the saving of a backup project file prior to chemistry import is recommended.

Selecting a project file The user can select an external project file by clicking the Browse button and navigating an external project file.

Importing base materials The stoichiometry of chemical reactions uses the names from the Base Materials list and it is therefore often necessary to import the materials list along with the chemical reactions. The options for importing base materials are • None - The base materials list will not be imported from the external project file • Append - All materials will be imported from the selected project file. Any imported materials that duplicate an existing material name will be added to the list with an incrementing counter added to the end of the material name. • Merge (overwrite) - All materials will be imported from the selected file and any duplicated material names in the existing base materials list will be overwritten with the incoming base material properties. • Merge (keep) - Only materials with names that do not exist in the current project file will be imported.

Importing chemistry The chemistry settings, reactions, and rate coefficients from will be imported based on the option selected. • None - No chemistry settings, reactions, or rate coefficients will be imported. • Replace - The chemistry settings, reactions, and rate coefficients in the selected project file will replace those in the existing file. • Append - All existing settings, reactions, and rate coefficients will be preserved and the reactions and rate coefficients in the selected file will be added.

9.5. Importing chemistry from another project

165


Fig. 9.11: Import Chemistry Dialog

166


CHAPTER

TEN

NUMERICS

The Numerics Window, shown in Fig. 10.1, provides an interface for enabling specialty models for viscosity and contact friction as well as access to dialogs for adjusting internal solver methods and convergence parameters. The default Barracuda VR settings in this section are appropriate for typical models and should not be adjusted unless necessary.

Fig. 10.1: Numerics Window

Solver settings The Solver Settings button will open the Solver Advanced Settings Dialog which provides access to solver convergence parameters.

Turbulence and advection options

167


The Turbulence and Advection Options button opens the Turbulence and Advection Options Dialog through which the turbulence model and fluid advection settings can be adjusted.

Contact friction The Contact friction model is designed to account for particle-particle contact forces that are present when a cell is at or near a close-packed condition. These forces, which are not fully captured by the multiphase particle-in-cell approach used by Barracuda VR, are usually negligible compared to other forces acting on a particle and therefore the contact friction model is not often used. When the model is enabled, the static value of this force is equal to the gravity vector multiplied by the Friction coefficient when the particle is at rest. This force decreases exponentially as the relative particle velocity increases.

Viscosity Barracuda VR contains two specialty models that adjust how fluid viscosity is calculated in the model: inviscid fluid flow and no viscous drag at solid walls. While these specialty models are typically not enabled when running simulations of real systems, they can be useful in determining the effects of viscous stresses in a fluid flow model. Inviscid fluid flow When enabled, Barracuda VR will not calculated viscous stresses in the fluid flow equations. No viscous drag at solid walls The Barracuda VR solver will not calculate viscous forces at a solid wall when this setting is enabled.

10.1 Solver Advanced Settings Dialog During each time step, Barracuda VR uses an iterative solution method to calculate fluid and solid states (pressure, temperature, velocities, etc.) that satisfy the underlying differential equations for mass, momentum, and energy transport. Each iteration that the solver runs will progressively increase the degree to which the differential equations are satisfied, thereby decreasing the error at each point. The summation of errors existing within the solution forms the residual which can be used to determine whether the solution has converged to an acceptable level. The Solver Advanced Settings Dialog, shown in Fig. 10.2, allows the user to change the maximum number of iterations and the maximum residual value used for the different systems of equations being solved by Barracuda VR. The Solver Advanced Settings Dialog is accessed by clicking on the Solver Settings button in the Numerics Window.

Fig. 10.2: Solver Advanced Settings Dialog Iterations and residual values The values specified under Iterations are the maximum number of solver iterations that will be performed to achieve a solver convergence that is lower than the specified Residual value.

168

Chapter 10. Numerics


10.2 Turbulence and Advection Options Dialog The methods used by Barracuda VR to calculate fluid advection and subgrid turbulence can be adjusted in the Turbulence and Advection Options dialog, shown in Fig. 10.3. Turbulence model

Fig. 10.3: Turbulence and Advection Options Dialog

The drop down menu allows users to select from three options for turbulence model: None, Algebraic, or Large eddy simulation (LES). LES is the default turbulence model in Barracuda VR. None No turbulence model will be active in the model. This is appropriate for laminar flow conditions. Algebraic A zero equation turbulence model in which the eddy viscosity is calculated from model length scales using the Boussinesq approximation. Large eddy simulation (LES) turbulence In the LES model, the large eddies are calculated from the flow equations and the subgrid turbulence is captured with a model. As with other turbulence models, there is no accepted choice for a subgrid turbulence model for dense particle flow. Barracuda VR uses the [SMA63] subgrid scale (SGS) model which calculates an eddy viscosity based on the notion that the effect of the SGS Reynolds stress is increased transport and dissipation. The form of subgrid scale eddy viscosity is

¯ µt = C s ∆2 S S ij =

1 2

⃒⃒ � � ∂u i ∂ uj + ∂x j ∂x i

1/3

where ∆ is the filter length. The common choice of filter length scale is ∆ = (δx δy δz) . The constant C s may be a function of Reynolds number, geometry, and solids volume fraction. For channel flow, it is on the order of 0.005 and can be larger or smaller with a typical value on the order of 0.01. Advection numerical scheme

10.2. Turbulence and Advection Options Dialog

169


In a typical finite volume calculation, fluid velocities values are tracked on cell faces whereas quantities such as pressure, density, mass concentration, and temperature are calculated within cells or at cell centers. When the advection term in a mass, momentum, or energy transport equation is calculated, it is done at cell faces as a face velocity multiplied by a cell average quantity estimated at the face location. It is the advection scheme that determines how the value of a cell average quantity is estimated at the face. The default advection scheme in Barracuda VR is the Partial Donor Cell method. In all advection schemes, the quantity of interest, Q, is calculated from the values of this quantity in two adjacent cells, Q1 and Q2 . A variable ϕ is used as an indicator of the direction of mass flow

ϕ =

u1 A1 θ1 ρ1 + u2 A2 θ2 ρ2 2

where u is the velocity, θ f is the fluid volume fraction, ρ is the density, and A is the face area. In the case where ϕ is positive, the mass flow is from cell 1 to cell 2. When ϕ is negative, the mass flow is from cell 2 to cell 1. Upwind In the first order upwind scheme, the quantity of of interest, Q1/2 , at the face is between two adjacent cells is assumed to have the value of the cell from which mass is flowing.

{

Q1 Q2

Q1/2 =

ϕ > 0 ϕ < 0

Deferred correction central difference and upwind An upwind with a deferred central difference correction adds a correction term ∆Q to the upwind advection scheme. The correction term is calculated at a previous time step.

{

Q1 Q2

ϕ > 0 ϕ < 0

Q2 − Q1 Q1 − Q2

ϕ > 0 ϕ < 0

Q1/2 = ∆Q +

1 ∆Q = 2

{

Partial donor cell (PDC) The partial donor cell scheme is a weighted average of central difference and upwind convection. A limiter is applied to automatically weight the central difference and upwind quantities [AORourkeB89]. The PDC method defines a donor cell, Qd , and an acceptor cell Qa as

Qd =

{

Q1 Q2

Qa =

{

Q2 Q1

ϕ > 0 ϕ < 0 ϕ > 0 ϕ < 0

The quantity of interest at the face is the calculated as

Q1/2 =

1 1 Qd (1 + Φ) + Qa (1 − Φ) 2 2

where the weighting factor Φ is calculated as

Φ = α + βC C = ∆t

|u1 A1 + u2 A2 | θ1 V 1 + θ2 V 2

and Φ is limited between 0 and 1. The quantities of α and β can be adjusted in Barracuda VR. If α = 0 and β = 0, then the convection is center differencing which is unconditionally unstable without sufficient amount of diffusion. If α = 1 and β = 0, then the convection is upwind which tends to be too diffusive. Values for α and β must be between 0.1 and 1. The default settings in Barracuda VR are α = 0.3 and β = 1.

170

Chapter 10. Numerics

CHAPTER

ELEVEN

TIME CONTROLS

Time Controls window, shown in Fig. 11.1, allows users to specify the simulation end time, time step settings, The and the frequency with which Barracuda VR will generate restart files as the simulation is running.

Fig. 11.1: Time Controls Window

Time step and duration settings The time step and duration settings allow the user to specify up to five sets of time steps and end times that will be used sequentially as the simulation progresses. The simulation starts on the first pair ( ∆t = 0.0001, tend = 0.1 in Fig. 11.1) and progresses with the specified time step until the end time is reached. At this point, the simulation moves on to the next set ( ∆t = 0.005, t end = 0.2), using the new time step until the next end time is reached. This repeats until the simulation reaches the last specified end time ( tend = 100) at which point the simulation is complete. The specified

171


time step in each set is effectively a maximum time step since the actual time step may be limited by the CFL number or temperature changes. These time step limiters are discussed in Advanced time step settings.

Advanced time step settings Clicking on the Advanced time step settings button raises the Advanced Time Step Options dialog, Fig. 11.2 where the user can adjust time step limiters. Max temperature change per time step (K) Barracuda VR uses a piecewise

Fig. 11.2: Advanced Time Step Options Dialog linear approximation of fourth order enthalpy expressions to calculate temperature changes. As the temperature change during a single time step becomes larger, the accuracy of the temperature calculation can decrease. Therefore, it is necessary to eliminate large temperature changes from the simulation. As the solver is running, Barracuda VR will automatically reduce the time step when a temperature change over a specified maximum is measured in a cell. The two choices for this are: • Solver controlled - The internal Barracuda VR maximum temperature change will be used when the solver controlled setting is selected. This value is currently 250 K. • Manually entered - The user can specify a maximum temperature change in the cell. The user may want to set the temperature change to a higher value in cases when the heat capacity of the materials in the system are constant or very linear. u∆t CFL control The Courant-Friedrichs-Lewy (CFL) number is calculated as CFL = ∆x and is a measure of how cell many cells of dimension xcell the fluid with velocity u will pass through in a single time step. The CFL number should be maintained within lower and upper limits by adjusting the time step to maintain the stability, accuracy, and speed of the calculation. The limits can be set by the user but the default values of 0.8 for the Min CFL and 1.5 for the Max CFL are recommended.

Restart File Intervals Barracuda VR will periodically generate IC (Initial conditions) files which contain all of the current simulation data. An IC file can be used later to restart the simulation at the time that the file was generated using the Restart Solver Dialog, discussed in Restarting a Simulation. The

172

Time Controls window allows the user to generate IC files at two different rates.

Chapter 11. Time Controls


• The Restart interval will create a new IC file at each interval of simulation time specified. The IC file will be created in the project directory with the filename IC_### where ### is simulation time at which the file was generated. • The Backtrack interval creates a restart file based on an actual elapsed time interval. The solver will generate an IC file with the filename IC_ at the interval specified which will overwrite any existing IC_ file that already exists. This backtrack IC file is beneficial because it will always maintain a relatively recent restart point for the simulation, irrespective of simulation progress.

173


174

Chapter 11. Time Controls

CHAPTER

TWELVE

DATA OUTPUT

Data Output section is where the user specifies the type and form of the data that should be output from a The Barracuda VR simulation. Flux planes Flux Planes are two-dimensional planes through which the movement of particles and fluid is tracked. GMV output options The Eulerian and Lagrangian data to be written in 3D visualization files is selected in the GMV Output Options window. Average data The selection of data for time-averaging is done through the Average Data window. 2D plot data Users can select data to be regularly written to 2D data output files during the simulation in the 2D Plot Data window. Transient data Users can specify cell data to be written to a file every time step in the Transient Data window. These data points act as virtual probes for measuring temperature, pressure, velocity, gas composition and many other variables. Wall erosion The wall erosion model in Barracuda VR tracks particle-wall collisions to estimate erosion areas. This model is enabled in the Wall Erosion window. Particle attrition The particle attrition model in Barracuda VR tracks particle-wall collisions to estimate particle attrition. This is enabled in the Particle Attrition window. Raw data Users can specify raw model data to be exported in a text-based format in the Raw Data window. Solver Output Units Users can specify the units of measure for data output in the Solver Output Units window.

12.1 Flux Planes A flux plane is a two-dimensional region through which the movement of particles and fluid is tracked. Since flux planes can be located anywhere inside the geometry and at boundary conditions, they are useful for analyzing the movement and accumulation of solids, gas components, and enthalpy within a system. During simulation, fluid and gas flow rates passing through the flux plane at each time step are recorded. Data can be split into fluid materials and particle species as well to obtain more depth of information. Flux Planes Window, shown in Fig. 12.1, which displays a list of all existing flux Flux planes are managed in the planes and allows users to follow the Add, Edit, Copy, Delete GUI pattern for the model. Within the window, the flux planes are displayed as individual lines. Each flux plane in the list has a unique ID number associated with it that is displayed alongside the flux plane name, location, direction, and other parameters. Note Flux planes at boundaries are automatically created when a flux plane name is specified for a boundary and are therefore not managed in this window. See Boundary Conditions for information on boundary flux planes.

Number of radius divisions

175


Fig. 12.1: Flux Planes Window

All flux plane specifications contain a Subdivide by radius check box which instructs the solver to separately divide the solids passing through the flux plane into groups based on particle diameter. The number of groups that the solver Flux Planes window. reports is specified by the Number of radius divisions selector box in the

12.1.1 Defining a flux plane Flux plane parameters are defined in the Flux Plane dialog, shown in Fig. 12.2. This dialog is activated when either a new flux plane is added or an existing flux plane is edited. Once the values have been entered, clicking Apply will assign the flux plane to the model and close the dialog whereas clicking Cancel will close the dialog and any values will be lost. The dialog also provides a link to the Reference Grid which is useful for determining i, j, k and x, y, z locations within the model domain. File name The flow of fluids and solids crossing the plane will be written to the file with the name specified in the Flux plane name field. The flux plane name must be unique from other filenames in the project directory. Tip: For easier file management and post-processing, the adoption of a flux plane naming convention is recommended. A typical convention is to start all flux file names with FLUX_ or flux_ . Plane direction Flux planes inside the domain must be aligned along either the x, y, or z direction, which is selected in the Surface plane direction dropdown box. The Reference grid can be useful for determining the appropriate direction of a flux plane. Gas species data If multiple gaseous species are used, users may wish to have the flux plane report the flux of each species crossing the plane. Gas species flux plane data will be written to one or more files in the directory with the flux plane name followed by _gasSpc###_###. The following options are available for selection from the Gas species flux plane behavior dropdown box: • No Output - gas species flux data is not reported • Mass Flow Rate - the mass flow rate of each gas species through the flux plane is reported • Mass Fraction - the average mass fraction of each gas species passing through the flux plane is reported • Mass Time Cumulative - the total mass of each gas species that has passed through the flux plane is reported • Mass Time Cumulative Plus Minus - the total mass of each gas species that has passed in each direction through the flux plane is tracked separately. Subdivide by radius Users may request that the solids flow crossing the plane be subdivided into size groupings by selecting the Subdivide by radius check-box. The number of radius subdivisions for all flux planes is controlled in the main Flux Planes window.

176

Chapter 12. Data Output


Fig. 12.2: Flux Plane Dialog

12.1. Flux Planes

177


Directional flux It is possible for there to be simultaneous movement of material in both directions across a flux plane. The basic flux plane output, however, reports only the net transfer of material across the flux plane. If the Directional Flux box is checked, the solver will separately report the mass flow in both directions across the flux plane (positive and negative). Plane location The location of the flux plane can be specified in either i, j, k ( Node) or x, y, z ( xyz) coordinates. When defining flux planes, it is recommended that users Enter Location by xyz rather than Node because if the grid is later changed, a flux plane defined in x, y, z coordinates will remain in the same location while one defined in i, j, k coordinates may have a different location after regridding. For more information on the coordinate system in Barracuda VR, see Setup Grid .

12.1.2 Flux plane data format The main flux file will be in the project directory with the specified filename. As the simulation progresses, the flux plane data will be reported as a separate row for each time step. The number of columns in the file is dependent on the output options selected, the number of particle species, and the thermal nature of the model. Therefore, the column numbering will vary with each model. A typical flux plane header is shown below which contains the following types of data: Time and flow data, (1-7) The simulation time of the data and material flow data is reported for every flux plane. Time integrated particle data, (8-25) The time integrated data is reported for all species automatically. The additional data in the “+” and “-” directions is only reported when the Directional flux check box is selected in the Flux Plane Dialog. Energy data, (26-27) The fluid and particle energy will reported only for thermal models. See The Global Settings Window for more information on thermal settings. # Cyclone bottom # 1 Time (s) # 2 Fluid mass flow rate (kg/s) # 3 Fluid volume flow rate (m^3/s) # 4 Fluid m ass f lux ( kg/s*m^2) # 5 Time integrated fluid mass crossing flux plane (kg) # 6 Average mass 5 time steps per time of all species (kg/s) # 7 Average mass 5 time steps per time of all species per area (kg/s*m^2) # 8 Time integrated particle mass of all species (kg) # 9 Time integrated particle mass of all species in + dir (kg) # 10 Time integrated particle mass of all species in - dir (kg) # 11 Time integrated particle mass of species 1 (kg) # 12 Time integrated particle mass of species 1 in + dir (kg) # 13 Time integrated particle mass of species 1 in - dir (kg) ... ... ... # 23 Time integrated particle mass of species 5 (kg) # 24 Time integrated particle mass of species 5 in + dir (kg) # 25 Time integrated particle mass of species 5 in - dir (kg) # 26 Time integrated fluid energy crossing flux plane (J) # 27 Time integrated particle energy crossing flux plane (J)

12.1.3 Gas species flux data format Gas species data at a flux plane will be reported in a file with the flux plane file name and _gasSpc###_### appended when Mass Flow Rate, Mass Fraction, Mass Time Cumulative, or Mass Time Cumulative Plus Minus is

178



selected for the Gas species flux plane behavior . The file is not created if None is selected. Sample flux files headers for each type are shown below.

Mass flow rate Net mass flow data is reported for each gas species along with a column for time and total mass flow. # # # # # # # # # # # # # # # #

Cyclone bottom Gas species 000 to 006 1 Time (s) 2 Total mass flow rate of all gas species (kg/s) 3 Mass flow rate (kg/s) of gas species 1 --- CH4 4 Mass flow rate (kg/s) of gas species 2 --- CO 5 Mass flow rate (kg/s) of gas species 3 --- CO2 6 Mass flow rate (kg/s) of gas species 4 --- H2 7 Mass flow rate (kg/s) of gas species 5 --- H2O 8 Mass flow rate (kg/s) of gas species 6 --- N2 9 Mass flow rate (kg/s) of gas species 7 --- O2 Area = 1.3789120e-03 (m^2) x1=-9.69318e-01(m) x2=-8.73304e-01(m) i1= 13 i2= 34 y1=-5.50742e-02(m) y2= 5.50754e-02(m) j1= 7 j2= 30 z1= 3.00993e+00(m) z2= 3.00993e+00(m) k1=121 k2=121

Mass fraction Average mass fraction data is reported for each gas species along with a column for time. # # # # # # # # # # # # # # #

Cyclone bottom Gas species 000 to 006 1 Time (s) 2 Mass fraction of gas species 1 3 Mass fraction of gas species 2 4 Mass fraction of gas species 3 5 Mass fraction of gas species 4 6 Mass fraction of gas species 5 7 Mass fraction of gas species 6 8 Mass fraction of gas species 7 Area = 1.3789120e-03 (m^2) x1=-9.69318e-01(m) x2=-8.73304e-01(m) y1=-5.50742e-02(m) y2= 5.50754e-02(m) z1= 3.00993e+00(m) z2= 3.00993e+00(m)

---------------

CH4 CO CO2 H2 H2O N2 O2

i1= 13 i2= 34 j1= 7 j2= 30 k1=121 k2=121

Mass time cumulative The net mass of each gas species that has passed through the flux plane is reported along with a column for time. # Cyclone bottom # Gas species 000 to 006 # # 1 Time (s) # 2 Time integrated fluid # 3 Time integrated fluid # 4 Time integrated fluid # 5 Time integrated fluid # 6 Time integrated fluid # 7 Time integrated fluid # 8 Time integrated fluid

12.1. Flux Planes

mass mass mass mass mass mass mass

(kg) (kg) (kg) (kg) (kg) (kg) (kg)

of of of of of of of

gas gas gas gas gas gas gas

species species species species species species species

0 1 2 3 4 5 6

------------------------------------

CH4 CO CO2 H2 H2O N2 O2

179


# # # #

Area = 1.3789120e-03 (m^2) x1=-9.69318e-01(m) x2=-8.73304e-01(m) y1=-5.50742e-02(m) y2= 5.50754e-02(m) z1= 3.00993e+00(m) z2= 3.00993e+00(m)

i1= 13 i2= 34 j1= 7 j2= 30 k1=121 k2=121

Mass time cumulative plus minus The net total mass, total mass in the positive direction, and total mass in the negative direction are reported for each gas species along with a column for time. # Cyclone bottom # Gas species 000 to 006 # # 1 Time (s) # 2 Time integrated fluid mass (kg) of # 3 Time integrated fluid mass (kg) of # 4 Time integrated fluid mass (kg) of # 5 Time integrated fluid mass (kg) of # 6 Time integrated fluid mass (kg) of # 7 Time integrated fluid mass (kg) of ... ... ... # 16 Time integrated fluid mass (kg) of # 17 Time integrated fluid mass (kg) of # 18 Time integrated fluid mass (kg) of # 19 Time integrated fluid mass (kg) of # 20 Time integrated fluid mass (kg) of # 21 Time integrated fluid mass (kg) of # 22 Time integrated fluid mass (kg) of # Area = 1.3789120e-03 (m^2) # x1=-9.69318e-01(m) x2=-8.73304e-01(m) # y1=-5.50742e-02(m) y2= 5.50754e-02(m) # z1= 3.00993e+00(m) z2= 3.00993e+00(m)

gas gas gas gas gas gas

species species species species species species

0 0 0 1 1 1

------ CH4 in + dir in - dir ------ CO in + dir in - dir

gas gas gas gas gas gas gas

species species species species species species species

4 5 5 5 6 6 6

in - dir ------ N2 in + dir in - dir ------ O2 in + dir in - dir

i1= 13 i2= 34 j1= 7 j2= 30 k1=121 k2=121

12.1.4 Flux plane particle size data format When the Subdivide by radius box is checked in a flux plane definition, particle size data is reported for the flux plane for each species in a file with the flux plane file name and _pSpc_ and the species ID number appended to it. Each particle species in the model will have an individual file in which the total mass that has passed through the flux plane for each size range is reported. A sample file header is shown below. # # Species ID 1 # # 1 Time (s) # 2 Mass (kg) # 3 Mass (kg) # 4 Mass (kg) # 5 Mass (kg) # 6 Mass (kg)

180

radius radius radius radius radius

(micron): (micron): (micron): (micron): (micron):

6.580 37.464 68.348 99.232 130.116

to to to to to

37.464 68.348 99.232 130.116 161.000



12.2 GMV Output Options In the GMV Output Options window, shown in Fig. 12.3, the Eulerian and Lagrangian data to be exported to the General Mesh Viewer (GMV) is selected. Only the data selected here can be viewed in GMV during post-processing.

Fig. 12.3: GMV Output Options Window

Output file interval The Plot interval is the amount of simulation time between which GMV files are written. Depending on the size of the simulation domain, a GMV file can take a substantial amount of space in computer storage. Therefore, an appropriate value for the plot interval should be entered. Recommended values are: • The default plot interval is 0.1 seconds . • A plot interval of 0.05 seconds can be used to produce a real-time animation at 20 frames per second. • For high frequency animations, a plot interval of 0.01 to 0.02 seconds would be appropriate. • For long running simulations, a larger plot interval of 0.2 to 1.0 seconds may be selected to conserve disk space. Tip The GMV plot interval can be changed by Interacting with a Running Simulation, or when Restarting a Simulation from an IC file. This can be used to produce a high-frequency animation partway through a long running simulation.

Eulerian output data Eulerian output data is grid-based data that can be displayed in GMV as cell or node values. Cell averaged values of particle data can also be displayed in this form as an Eulerian field. The user can select as many Eulerian output options as needed.

12.2. GMV Output Options

181


Parameter

GMV Variable

Description

Particle volume fraction Fluid velocity Particle velocity

p_volFra

Cell-averaged particle volume fraction. Recommended for all calculations

U,V,W

Fluid velocity components. Recommended for all calculations

P_xVel, P_yVel, P_zVel Pressure DynPres

Cell-averaged particle velocity components. Used for study of particle motion

Pressure Dynamic pressure Fluid density Cell indices

f-dens i, j, k

Fluid pressure. Recommended for all calculations Kinetic energy per unit volume of fluid. Used for specialized fluid flow studies Useful for gas phase calculations with large changes in pressure, temperature, or molecular weights Writes the i, j, k coordinates for each cell to GMV. This is a valuable tool for determining cell locations in three dimensions for gridding or the setting of boundary conditions The total particle mass in a cell divided by the cell volume

Particle bulk density Turbulent viscosity CFL

p-dens

Particle species Fluid temperature Particle temperature Cell volume dp/dx dp/dy dp/dz

Species

The turbulent viscosity resulting from the selected turbulence model. Used for specialized fluid flow studies The Courant-Friedrichs-Lewy (CFL) number is commonly used to control the time step. Can be used to determine areas of high CFL for model optimization Cell-averaged particle species. Can be used to study mixing

f-Temp

Fluid temperature. Recommended for all thermal models

p-Temp

Cell-averaged particle temperature. Recommended for all thermal models

cellVol

The volume of each cell. Can be used for post-processing analysis in GMV and is recommended for all calculations Pressure gradient in x direction Pressure gradient in y direction Pressure gradient in z direction. Recommended for all calculations with gravity vector in z direction Cell-averaged particle flux. Recommended for study of fluidization patterns

Solid mass flux Fluid mass flux Wall heat transfer

ViscTurb CFL

dp/dx dp/dy dp/dz P_xMass, P_yMass, P_zMass F_xMass, F_yMass, F_zMass wallHeat

Cell-averaged fluid mass flux. Recommended for study of fluidization patterns

Display of heat transfer at thermal walls. Recommended for models that include a thermal wall

Lagrangian output data Lagrangian output data can be used to color or select particles in GMV. The user can select as many Lagrangian output options as needed.

182



Parameter

GMV Variable

Description

Particle volume fraction Particle speed Particle radius Constant color Particle material Particle density Particle species Visual

VolFrac

Recommended for all calculations

Speed

Recommended for all calculations

rad

Particle radius in microns. Recommended for all calculations

Particle

Particles all have a zero value which can be useful for visualization. This is typically not used. Outputs the mass fraction of each particle component. Recommended for all multimaterial calculations Useful for calculations containing particles with different densities

Velocity

Residence time Residence time by species Temperature

Density Species Visual velx, vely, velz ResTime ResTime## Temperat

Shows the particle species ID assigned in the Particle Species Window ( Particle Species). Recommended for all multiple species calculations A legacy output option combining volume fraction and location. This is typically not used. Particle velocity vector components. Recommended for all calculations

Displays the amount of time that a particle has been inside the calculation domain. Particles initialized in the domain will have a residence time equal to the current simulation time. Recommended for all calculations with a particle feed Shows a separate residence time for each species with other particle species assigned a value of -1. While this may make particle filtering easier, filtering can also be done in GMV. Therefore, this output option is typically not used. Particle temperature. Recommended for all thermal calculations

Gas Species The gas species option determines the form in which gas composition data is output. Only one value can be selected.

Output Choice

GMV Variable

Description

Mass fraction Mass concentration Mole fraction Mole concentration

(species).mf (species).mc (species).nf (species).nc

Gas species are output as mass fractions Gas species are output as mass concentrations Gas species are output as mole fractions Gas species are output as mole concentrations

Compressing graphics output: When the Compress graphics output checkbox is clicked, GMV files will immediately be compressed into .tar.gz files after writing. This option is only used in cases where disk space is extremely limited as the files must be first be uncompressed before use.

Generate predefined GMV attribute files: Several predefined GMV files will be copied from the cpfdHQ directory to the project directory when the Generate predefined GMV attribute files checkbox is clicked. These are the same attribute files that are used when results are viewed through the Post-Run window.

12.2. GMV Output Options

183


12.3 Average Data Most of the data output in Barracuda VR is instantaneous data that displays the conditions at a specific location and time. While instantaneous data is immensely valuable, additional insights into fluidization patterns and other global trends are more readily obtained through the study of time-averaged data. For this reason, Barracuda VR can track time-averaged values for select GMV output data. The selection of data for averaging is done through the Average Data window, shown in Fig. 12.4.

Fig. 12.4: Average Data Output Window

Start time The Start time for average parameter specifies the simulation time at which averaging should begin. Typically, data averaging should begin once the system has fluidized or otherwise reached some sustainable mode of operation. This may not always be known, so it is often best to make a reasonable guess. Tip The time-averaging of all variables can be reset to zero when Restarting a Simulation.

Minimum volume fraction The Minimum particle volume fraction for averaging particle data parameter sets a volume fraction threshold for particle data above which averaging will occur. This may be used to eliminate small volume fraction data from consideration.

Output options The averaging is handled slightly differently for Eulerian and Lagrangian data. For Eulerian data, the averaging is

184



weighted by the length of the time step whereas Lagrangian data is weighted for each particle in a cell by the volume fraction of the particle, θ p , and the length of the time step. The Lagrangian data is also subject to the specified minimum volume fraction.

∑∑ ∑∑ ∑∑

i φ∆ti

φ¯ (Eulerian) = φ¯ (Lagrangian)

i ∆ti

p φ p θ p ∆ti

i

i

p θ p ∆ti

For a description of the following average data options, see GMV Output Options. Eulerian Particle volume fraction, pressure, fluid velocity, fluid mass flux, fluid temperature, dp/dx, dp/dy, dp/dz, wall heat transfer rate/flux, and gas species Lagrangian Particle velocity, solid mass flux, and particle temperature

12.4 2D Plot Data 2D Plot Data section, shown in Fig. 12.5, users can select data to be regularly written to 2D data output files In the during the simulation. 2D data files contain information at many points on a user-selected plane and are convenient for creating line or surface plots. Whereas transient data points are useful for collecting high-frequency output data at a few points in space, 2D data files are better suited for collecting low-frequency output data across a plane that spans the entire computational domain.

Fig. 12.5: 2D Plot Data Window

12.4. 2D Plot Data

185


2D plotting planes At least one of the available data plane directions must be checked in order for 2D data output files to be generated. The available planes are Plot xy, xz, yx, yz, zx, and zy. Only planes normal to the primary x-, y-, and z-axes are available for 2D output data, and only one plane can be chosen in each of the available directions. The choice of plane direction determines which axial direction is used as the first column in any resulting output files, as illustrated in Fig. 12.6. The arrows indicate the arrangement of data columns in the output files for each choice of plane orientation. For example, if Plot xy is chosen, 2D data will be output with x-position as the first column. The remaining columns in the data file will correspond to evenly-spaced y-locations in the domain.

Fig. 12.6: 2D Plot Data Orientation Index location The locations of the output data planes are specified by i-j-k indices, using the text boxes labeled i-index, j-index, and k-index. A direct index value can be entered to specify a particular cell-layer. Or, the special keywords nx/2, ny/2, and nz/2 can be used to specify that the data plane should be in the middle of the domain. For example, if the simulation has 20 x-cells, and nx/2 is selected, a data output plane will be generated at i=10. Including boundary data An additional row will be printed at the top and bottom of the data file for boundary data when the Include boundary data in 2D plot checkbox is clicked.

Plot interval options By default, the Time interval for creating 2D data files is 0 s, which means that 2D data files will be created each time a GMV file is written. The frequency for GMV file output is set in the GMV Output Options window. If a different output interval is desired, the Time interval can be changed accordingly.

186



It is possible to limit the number of columns of data reported in the 2D data files by adjusting the Max # of columns value. The default value is 29 columns, which usually works well.

2D data to display Many variables are available for outputting to 2D data files. Any combination of the available variables may be chosen, with a separate file written for each variable at each plot time interval. If a selected variable is not consistent with the setup of the simulation, output files will not be generated for that variable. For example, if an isothermal simulation is being run, no output files will be generated if Fluid temperature or Particle temperature are selected. Similarly, Average Data window if any time-average 2D data it is necessary that time averaging be properly set up in the variables are selected, such as Ave particle volume fraction or Ave pressure . If information about gas composition is desired, use the options in the Output Gas Species region. Note that only a single basis can be chosen for reporting the gas composition: Mass fraction, Mass concentration, Mole fraction, or Mole concentration .

12.4.1 2D data filenames 2D data will produce data files with a file name followed by a .dat extension. The data output and corresponding filename are shown in the table below where ** represents a plane direction (xy, xz, yz, etc.); ##... represents a time stamp; and [GAS] represents a gas species name. For vector data, three separate files will be created with x, y, or z located in the [xyz] position.

12.4. 2D Plot Data

187


Variable

File Name

Fluid volume fraction

Variable

File Name

Gas species rate f0_ ** _##...

Particle volume fraction f1_ ** _##...

Pressure

p_ ** _##...

Fluid velocity

[GAS]_ **MassRate_##...

Ave particle volume frac tion

fAv1_ ** _##...

Ave Pressure pAv_ ** _##...

Ave fluid velocity u[xyz]_0** _##...

Particle velocity

u[xyz]Av_0** _##...

Ave particle velocity u[xyz]_1** _##...

Fluid temperature

u[xyz]Av_1** _##...

Ave fluid temperature T0_ ** _##...

Particle temperature

TAv0_ ** _##...

Ave particle temperature T1_ ** _##...

Fluid mass flux

TAv1_ ** _##...

Ave fluid mass flux massFlux_[xyz]_0** _##...

Particle mass flux

massFlux_[xyz]Av_0** _##

Ave particle mass flux massFlux_[xyz]_1** _##...

Passive scalar s[#]_ ** _##...

Fluid density

Stoichiometric rate

massFlux_[xyz]Av_1** _##

equation

R[##]_ ** _##...

Liquid concentration rho0_ ** _##...

liqConcn_ ** _##...

Particle density rho1_ ** _##...

Gas mass fraction

Ave gas mass fraction [GAS]_ **MassFrac_##...

Gas mass concentration

[GAS]_ **MassFrac_Av_##.

Ave gas mass concentration [GAS]_ **MassConc_##...

Gas mole fraction

[GAS]_ **MassConc_Av_##.

Ave gas mole fraction [GAS]_ **MoleFrac_##...

Gas mole concentration

Ave gas mole concentration [GAS]_ **MoleConc_##...

188

[GAS]_ **MolesFrac_Av_##

[GAS]_ **MoleConc_Av_##.



12.4.2 2D data example Suppose you wanted to plot the vertical pressure profile in a vessel. Instead of defining many transient points along the vertical axis of the vessel, you could request 2D Plot Data using Plot zx, and specify a j-index of :kbd‘ny/2‘ to automatically select the middle plane of the domain. Also, under 2D Data to Display, select Pressure. For this example, let us consider the 2D data output file generated at 100 s of simulation time. Based on the choices above, the output file will be named p_zx_00100.000.dat, and will contain data like this (where many rows and columns have been omitted so that the data can be more easily understood): # Time = 1.00000e+02 (s) # y = -1.52578e-03 (m) (j=17) # First column z-distance (m). Following columns are pressure (Pa) at x-locations # 1 z # 2 x=-1.45617e+00m (i= 1) # 3 x=-1.25138e+00m (i= 3) # 4 x=-1.06082e+00m (i= 5) # 5 x=-8.66964e-01m (i= 7) # 6 x=-6.63088e-01m (i= 9) # 7 x=-4.59213e-01m (i=11) # 8 x=-2.55338e-01m (i=13) # 9 x=-5.14624e-02m (i=15) # 10 x= 1.52413e-01m (i=17) # 11 x= 3.56288e-01m (i=19) # 12 x= 5.60164e-01m (i=21) # 13 x= 7.64039e-01m (i=23) # 14 x= 9.64748e-01m (i=25) # 15 x= 1.15596e+00m (i=27) # 16 x= 1.34459e+00m (i=29) 5.109293e-02 2.145529e+05 ... 2.143982e+05 2.143254e+05 2.144238e+05 1.532788e-01 2.135674e+05 ... 2.130621e+05 2.134436e+05 2.133121e+05 2.543134e-01 2.126510e+05 ... 2.121992e+05 2.125659e+05 2.124333e+05 3.544176e-01 2.116283e+05 ... 2.116586e+05 2.116075e+05 2.116506e+05 4.597586e-01 2.108635e+05 ... 2.112145e+05 2.109796e+05 2.109576e+05 5.756338e-01 2.100927e+05 ... 2.100725e+05 2.099118e+05 2.097871e+05 7.030965e-01 2.093017e+05 ... 2.088005e+05 2.088166e+05 2.085325e+05 8.433054e-01 2.081575e+05 ... 2.075921e+05 2.076447e+05 2.075209e+05 ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... 8.152327e+00 2.000321e+05 ... 2.000317e+05 2.000317e+05 2.000317e+05 8.304892e+00 2.000321e+05 ... 2.000312e+05 2.000312e+05 2.000313e+05 8.457457e+00 0.000000e+00 ... 2.000307e+05 2.000307e+05 2.000313e+05 8.610022e+00 0.000000e+00 ... 2.000302e+05 2.000303e+05 0.000000e+00 8.762587e+00 0.000000e+00 ... 2.000299e+05 2.000303e+05 0.000000e+00 8.915152e+00 0.000000e+00 ... 2.000299e+05 0.000000e+00 0.000000e+00 9.067717e+00 0.000000e+00 ... 0.000000e+00 0.000000e+00 0.000000e+00

The header comments tell us several important things, including: 1. The simulation time at which the data was written. 2. The y-location, in both absolute distance (meters) and index notation. 3. The meaning of the first column (in this case, z-distance), and the type of data contained in the remaining columns (in this case, pressure with units of Pascals). 4. The x-locations corresponding to columns 2-16. Back to the example task of plotting the vertical pressure profile in the vessel, we can create a plot of the data in column 1 (z-location, or height in the vessel) versus the pressure at the smallest x-location (column 9 in the above

12.4. 2D Plot Data

189


example output file). Such a plot is shown below in Fig. 12.7.

Fig. 12.7: 2D Plot Data Example

12.5 Transient Data Barracuda VR will write specified cell data for each time step to a transient data file. These data points act as virtual probes for measuring temperature, pressure, velocity, gas composition, and many other variables. Transient data is Transient Data window, shown in Fig. 12.8. managed in the

Specifying a transient data point Each data point is managed on a separate line in the transient data table which provides textboxes and dropdown boxes on each line for transient data point specification. Each transient data point must have a type and location. Transient data types The following data types are available for specification as a transient data type: • Pressure • Particle volume fraction (p Vol Frac) • Fluid velocity components (x-Vel, y-Vel, z-Vel) • Particle velocity components (p x-Vel, p y-Vel, p z-Vel) • Time step length (dt) • Fluid temperature ( Fluid temp) • Particle temperature ( Particle temp) • Fluid density • Fluid mass flow rate components (x Massflow, y Massflow, z Massflow) • Pressure gradient components (dp/dx, dp/dy, dp/dz)

190



Fig. 12.8: Transient Data Output Window • Gas mass fraction (Gas mass frac) • Gas mole fraction (Gas mole frac) • Gas mass concentration ( Gas mass conc) • Gas mole concentration ( Gas mole conc) • Wall erosion index (Wall wear) • Stoichiometric reaction rate (Stoich eq rate) • Wall heat transfer Location The location of a transient data point can be either specified with xyz (in meters) or ijk coordinates and is selected in the dropdown box to the right of the transient data type. Comment A comment field is provided for noting the significance of the transient data point. The comment is also printed in the transient data file for later reference.

Removing transient data points A single transient data point can be removed by selecting the blank line at the top of the Type dropdown box. This point will be removed the next time the Transient Data window is opened. All transient data points can be removed by clicking the Clear All button.

Transient data options Output file name The output file name can be entered in the Output file name textbox or navigated to by clicking on the Browse button. The default trans.data is recommended. Write frequency The write frequency controls how frequently the data is recorded. The default write frequency of 1 means that data from every time step will be written to the output file. A write frequency of 2 means that data from every second time step will be written to the output file, and so forth. Flush frequency The flush frequency has to do with the computer input/output functions. When Barracuda VR writes to a file, the computer operating system places the data in a buffer, and the data is not actually written to the file until the buffer is flushed. The flush frequency controls how frequently these buffers are flushed. A flush frequency of 1 means the buffer will be flushed with every write. A flush frequency of 2 means that the buffer will be flushed with

12.5. Transient Data

191


every second write, and so forth. The default flush frequency of 0 means that the buffer flushing is controlled by the operating system. It is recommended that the default flush frequency of 0 be used to maximize computational efficiency.

12.6 Wall Erosion Wall Erosion model in Barracuda VR collects data on particle-wall collisions for the purpose of studying the The erosion and wear of walls. Wall erosion is a function of particle mass, particle speed, and the angle at which particles strike the wall, the form of which is dependent on the wall materials. To accommodate the range of materials in used in particle-fluid systems, the Barracuda VR wear GUI provides an interface for adjusting the form of the wear model and the angular dependence. The wear model is enabled by selecting the Enable Wall Erosion check box within the Wall Erosion Window, shown in Fig. 12.9.

Fig. 12.9: Wall Erosion Window with default settings

Erosion model When enabled, the erosion model in Barracuda VR tracks the accumulated impacts of particles on each wall patch in the geometry. The magnitude of the impact value, I p , is calculated for each particles as I p = w(θ p ) · m pa · u pb where m p is the particle mass, u p is the particle speed, a and b are constants, and w(θ p ) is a weighting factor that is a function of the impact angle, θ p .

192



Angular dependence The angular dependence is a property of the wall material and therefore should be set based on available information on wall material and erosion properties. According to [Til69], the angular dependence of a ductile material (aluminum), will have a maximum erosion angle at glancing angle of approximately 20 ◦ whereas a brittle material (glass or refractory) will experience maximum erosion at impacts normal to the surface. In the Wall Erosion Window, the weighting function, w(θ p ), can be specified by either moving the sliders in the Graph input or by typing the weighting values directly for each angle in the Text input boxes. The values of the Barracuda VR default model are shown in Table 12.1 along with a selection of other material angular weighting models adapted from the data of [Til69]. It should be noted that the erosion models in Table 12.1 are normalized such that maximum erosion angle has weight of 1.0. Table 12.1: Angular weights for different materials. Material properties adapted from the data of [Til69]

Impact Angle (degrees)

0.0

5.7

11.5

17.5

23.6

30.0

36.9

44.4

53.1

64.2

90.0

sin(Impact Angle) Barracuda VR default Aluminum (ductile) Glass (brittle) Steel (both)

0.0 0.00 0.00 0.00 0.00

0.1 0.71 0.41 0.00 0.24

0.2 1.00 1.00 0.04 0.64

0.3 0.97 0.97 0.07 1.00

0.4 0.90 0.82 0.16 0.84

0.5 0.80 0.74 0.23 0.64

0.6 0.72 0.53 0.36 0.52

0.7 0.61 0.44 0.48 0.48

0.8 0.52 0.33 0.63 0.40

0.9 0.38 0.18 0.80 0.36

1.0 0.30 0.03 1.00 0.36

Wear exponents The values of the exponents on particle mass and particle velocity are set by selecting the desired value from the Mass exponent and Velocity exponent lists. Literature differs on values for the mass and velocity exponents but values between 1.0 to 1.5 for the mass exponent and values between 2.5 and 5.0 for the velocity exponent are likely reasonable [MM79]. Minimum limit Users can also set a threshold below which impact values will not be included in the summation by typing the value in the Minimum Limit box. The concept behind this is that small particle impacts may not be sufficient to cause any erosion and a minimum threshold might be necessary for erosion to begin. In the absence of data indicating a proper threshold value, it is typical to use the default value of 0 . Calculation delay Users can delay the start time for the wear model by typing a time delay (in seconds) in the Start calculating wear at time box. This is typically used in cases where steady-state erosion is being studied.

Wear model output data The wear model creates an erosion index for each wall patch on the model surface. For each particle hitting the wall patch, the impact is calculated from the wear model and added to the index. The total index is normalized by the area of the wall patch and then annualized when output to the post-processing files. Note that if a cell has more than one wall patch, then the value written to the post-processing files is the total for all wall patches divided by the total area of all wall patches. The wall erosion data appears in GMV as the variable Impact which is typically used to compare the erosion of wall faces relative to the erosion at other places within the geometry. A common post-processing technique is to create isosurfaces of the erosion index in GMV to show levels of high, medium, and low erosion on geometry walls or other internals, as shown in Fig. 12.10.

12.7 Particle Attrition The attrition model tracks the potential for particle attrition due to particle-wall collisions during a simulation and is enabled through the Particle Attrition window, shown in Fig. 12.11. The potential for particle attrition is tracked through an “attrition index” which is added to each time a particle hits a wall. The attrition index is a function of particle mass raised to the Mass exponent and the normal component of the impact velocity raised to the Velocity exponent . When the particle attrition model is enabled, the index value can be viewed in GMV as the particle variable named “trauma”. Users should be aware of two important limitations in this model:

12.7. Particle Attrition

193


Fig. 12.10: Erosion levels near cyclone inlet: (a) low levels of erosion (b) medium levels of erosion (c) high levels of erosion 1. Particles do not actually attrit. The model creates an attrition index related to the likelihood of attrition. However, regardless of the value of this function, the size of the particle does not change from impact. 2. Only impacts with walls are considered. Particle to particle collisions are not part of this model. Delaying the attrition model Users can delay the start time for the attrition model by entering a time delay in the Start calculating attrition at time textbox. This would be used to ignore the start-up effects of the model, for example. Minimum index value Users can set a minimum attrition value below which the values are not added to the sum. This is typically set to zero in the absence of theory or experimental data required to accurately set this value.

12.8 Raw Data When Raw Data is used, Barracuda VR will write out simulation data in a text-based format for external analysis. Raw Data window, Data selection and specification of the time interval for file writes is set by the user in the shown in Fig. 12.12.

Time interval The raw data files will be written at the time interval specified in the Time interval for data dump text box. The default interval, 1e+20 seconds, is large enough that no data files should be written.

12.8.1 Particle Output Data Particles

194



Fig. 12.11: Particle Attrition Window

Fig. 12.12: Raw Data Output Window

12.8. Raw Data

195


Selecting this option causes raw particle data to be written at the interval specified for the Time interval for data dump to the following file: Raw.particle._ is the file’s number in a sequential counter of raw data files is the simulation time at which the file was created

In the raw particle data file, a line of data is written for every computational particle currently in the simulation. For each particle, many variables are output in a space-separated column format (similar to most other Barracuda VR output files). See the header section of the raw particle data file for a full listing. As an example, here are some typical output columns: #@ #@ #@ #@ #@ #@ #@ #@ #@ #@ #@

1 2 3 4 5 6 7 8 9 10 11

"x" "y" "z" "Particle u" "Particle v" "Particle w" "Fluid u at particle" "Fluid v at particle" "Fluid w at particle" "Cloud volume" "Cloud mass"

"m" "m" "m" "m/s" "m/s" "m/s" "m/s" "m/s" "m/s" "m^3" "kg"

Raw particle composition is included. Note that the “Particle mass” column gives the mass of an individual real particle, and “Cloud mass” gives the mass of a computational particle. Discrete chemistry rate Selecting this option for raw data output will cause raw particle data to be written at the interval specified for the Time interval for data dump. One file will be written at each interval, named as follows: Raw.particle.chemrate._ is the file’s number in a sequential counter of raw data files is the simulation time at which the file was created

A line of data is written for every computational particle currently in the simulation. For each particle, a column of data is written for each discrete particle chemical reaction defined in the project. The output is in a space-separated column format (similar to most other Barracuda VR output files). See the header section of the file for a listing of the reaction rates available in your particular project. As an example, here are some typical output columns: #@ #@ #@ #@ #@ #@ #@

1 2 3 4 5 6 7

"pid" "x" "y" "z" "Discrete chemistry rate for discrete chemistry reaction "Discrete chemistry rate for discrete chemistry reaction "Discrete chemistry rate for discrete chemistry reaction

"" "m" "m" "m" 0" 1" 2"

"kg/s" "kg/s" "kg/s"

If a particle is participating in any of the discrete chemistry reactions, the reaction rate for that particle, with respect to the primary solid material used when defining the discrete chemistry reaction, will be recorded. If a particle does not participate in a given discrete chemistry reaction, a reaction rate of zero will be recorded.

12.8.2 Cell Output Data General

196



Selecting any of the raw data cell options causes raw cell data to be written at the interval specified by Time interval for data dump to the following file: Raw.cell._ is the file’s number in a sequential counter of raw data files is the simulation time at which the file was created

In the raw cell data file, a line of data is written for each fluid control volume in the simulation. For each cell, output variables depend on selected options and are in a space-separated column format (similar to most other Barracuda VR output files). However, variables i , j , k , locCv, and cvType will always be included and placed as first five columns. As an example, here are some typical output columns: #@ #@ #@ #@ #@ #@ #@

1 2 3 4 5 6 7

"i" "j" "k" "locCv" "cvType" "Pressure" "Volume fraction"

"" "" "" "" "" "Pa" ""

Cell types The cell type codes for cvType are as follows: • 1 - Real cells with non-zero volume • 100 - Pressure boundary cell • 101 - Solid cell, located next to a real cell • 120 - Null cells outside of the domain Face types The face type codes for xFaceType, xFaceType, and zFaceType are as follows: • 3 - Internal face with real cells on either side • 5 - Special flow BC with a real cell on either side • 6 - Face between a real cell and a pressure BC cell • 8 - Flow BC face • 100 - Faces with zero area and real cells on either side • 110 - Face with zero area and a solid on both sides • 120 - Faces outside of the computational domain

Volume-average discrete chemistry rate Selecting this option for raw data output will cause raw cell data to be written at the interval specified for the Time interval for data dump. One file will be written at each interval, named as follows: Raw.cell.chemrate._ is the file’s number in a sequential counter of raw data files is the simulation time at which the file was created

A line of data is written for every computational cell in the simulation. For each cell, a column of data is written for each discrete particle chemical reaction defined in the project. The output is in a space-separated column format (similar to most other Barracuda VR output files). See the header section of the file for a listing of the reaction rates available in your particular project. As an example, here are some typical output columns:

12.8. Raw Data

197


#@ #@ #@ #@ #@ #@ #@ #@ #@

1 2 3 4 5 6 7 8 9

"i" "j" "k" "locCv" "x" "y" "z" "Volume-average rate for discrete chemistry reaction "Volume-average rate for discrete chemistry reaction

0" 1"

"" "" "" "" "m" "m" "m" "kmol/s*m^3" "kmol/s*m^3"

The volume-average reaction rates recorded in this file are calculated from summations of discrete particle reaction rates. For all particles in a computational cell, the discrete chemistry reaction rate is summed for each discrete chemistry reaction, and a volume-average equivalent reaction rate is calculated based on the gas volume in the cell.

Material properties Selecting this option for raw data output will cause raw cell data to be written at the interval specified for the Time interval for data dump. One file will be written at each interval, named as follows: MaterialProp_ is the simulation time at which the file was created

A line of data is written for every computational cell in the simulation. For each cell, detailed fluid material properties are recorded, such as viscosity, thermal conductivity, and specific heat. The output is in a space-separated column format (similar to most other Barracuda VR output files). See the header section of the file for a listing of the reaction rates available in your particular project. As an example, here are some typical output columns: # # # # # # # # # # # # # # # #

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

cell index i cell index j cell index k cell center x cell center y cell center z cell volume fluid temperature pressure fluid density Mass fraction CO2 Mass fraction O2 fluid viscosity fluid mass diffusivity fluid thermal conductivity fluid specific heat

"" "" "" "m" "m" "m" "m^3" "K" "Pa" "kg/m^3" "" "" "kg/m^3" "m^2/s" "J/(s*m*K)" "J/kg"

12.9 Solver Output Units Solver Output Units window, Barracuda VR can output data in a wide range of units that can be selected in the shown in Fig. 12.13. The selected units will be applied globally to log files, flux files, transient files, and GMV files. Note that the selection of units in this window does not affect any units for model specification or input files.

Setting units Units for a data type are set by selecting the type from the Output Units List. Doing so will display checkboxes for each of the available units from which the desired units can be selected. Alternatively, all data types can be set to SI units by clicking the Apply all SI Units button or to imperial units by clicking the Apply all Imperial Units button.

198



Fig. 12.13: Solver Output Units Window

12.9. Solver Output Units

199


Available units The following lists the units available for selection for each data type. Time second (s), millisecond ( ms), minute ( min), hour , day Length meter ( m), millimeter ( mm), centimeter ( cm), foot ( ft ), inch (in) Mass kilogram (kg), gram (g), pound (lb), ton , slug Temperature kelvin ( K ), Celsius (C ), Fahrenheit ( F ), Rankine ( R) Mole mole (mol), kilomole ( kmol) Pressure pascal (Pa), kilopascal ( kPa), pounds per square inch absolute ( psia), bar Pressure Gradient pascal per meter ( Pa/m), pascal per centimeter ( Pa/cm), pounds per square inch per foot ( psi/ft ), pounds per square in per meter ( psi/m) Velocity meters per second ( m/s), feet per second ( ft/s), millimeters per second ( mm/s), centimeters per second ( cm/s), inches per second ( in/s) Mass flux kilograms per second per square meter ( kg/s*m^2), grams per second per square meter ( g/s*m^2), grams per second per square centimeter ( g/s*cm^2), pounds per second per square meter ( lb/s*m^2), pounds per second per square foot ( lb/s*ft^2) Density kilograms per cubic meter ( kg/m^3), pounds per cubic meter ( lb/m^3), pounds per cubic foot ( lb/ft3), grams per cubic centimeter ( g/cm^3) Heat transfer coefficient joules per second per square meter per kelvin ( J/s*m^2*K ), watts per square centimeter per degree Celsius (W/cm^2*C ), erg per second per square centimeter per kelvin ( erg/s*cm^2*K ), BTU per second per square foot per kelvin ( Btu/s*ft^2*K ), BTU per second per square foot per degree Fahrenheit ( Btu/s*ft^2*F ), BTU per second per square foot per degree Rankine ( Btu/s*ft^2*R) Heat transfer rate joule per second ( J/s), watt (W ), BTU per second ( Btu/s), BTU per hour ( Btu/hr ), erg per second (erg/s) Chemical reaction rate kilomole per second per cubic meter ( kmol/s/m^3), mole per second per cubic meter (mol/s/m^3), mole per second per cubic foot ( mol/s/ft^3), kilomole per second per cubic foot ( kmol/s/ft^3) Particle radius micron, millimeter ( mm), centimeter ( cm) Mass flow rate kilograms per second ( kg/s), pounds per second (lb/s), grams per second (g/s), kilograms per minute (kg/min), pounds per minute ( lb/min), kilograms per hour ( kg/hr ), pounds per hour (lb/hr ) Volume flow rate cubic meters per second ( m^3/s), cubic meters per minute ( m^3/min), cubic meters per hour (m^3/hr ), cubic feet per second ( ft^3/s), cubic feet per minute ( ft^3/min), cubic feet per hour ( ft^3/hr ) Mass concentration kilograms per cubic meter ( kg/m^3), pounds per cubic meter ( lb/m^3), pounds per cubic foot (lb/ft^3), kilograms per cubic foot ( kg/ft^3) Mole concentration moles per cubic meter ( mol/m^3), kilomoles per cubic meter ( kmol/m^3), moles per liter ( mol/L) Viscosity kilograms per second per meter ( kg/s*m), poise, centipoise, pounds per second per foot ( lb/s*ft ) Kinematic viscosity square meters per second ( m^2/s), square feet per second ( ft^2/s), stokes, centistokes Thermal conductivity joules per second per meter per kelvin ( J/s*m*K ), joules per second per centimeter per degree Celsius ( J/s*cm*C ), BTU per hour per foot per kelvin ( Btu/hr*ft*K ), BTU per hour per foot per degree Fahrenheit ( Btu/hr*ft*F ) Specific enthalpy joules per kilogram ( J/kg), joules per g ( J/g), BTU per pound ( Btu/lb), BTU per kilogram ( Btu/kg) Molecular weight grams per mole ( g/mol), kilograms per kilomole ( kg/kmol) Area square meters ( m^2), square centimeters ( cm^2), square millimeters ( mm^2), square feet ( ft^2)

200



Volume cubic meters ( m^3), cubic centimeters ( cm^3), cubic millimeters ( mm^3), cubic feet ( ft^3) Energy joules ( J ), erg, British thermal units ( Btu), calories ( cal) Wear kgâ*m^b/s^b/m^2 Attrition kgâ*m^b/s^b CPU time seconds (s), minutes ( min), hours (hr ), day

12.9. Solver Output Units

201


202


CHAPTER

THIRTEEN

RUN

Run window, shown in Fig. 13.1, includes shortcut buttons which allow users to run the solver for one time The step, check the project setup, run a simulation, restart an existing simulation from an IC file, and interact with a simulation that is currently running.

Fig. 13.1: Run Window

13.1 Checking the Model Setup It is always recommended that users check the setup of the Barracuda VR model before starting the simulation. This is important to verify that the locations of boundary conditions, initialized particles, flux planes, and transient data

203


points within the model are correct. To do this, the solver must run a single time step in which it reads the project file and produces the GMV file and log files required for checking the model.

13.1.1 Running a single time step Run window will open a command window in which the BarClicking on the Run Solver Setup button from the racuda VR solver will run the simulation for one time step, write out the first Gmv file ( Gmv.00000), and create all log, flux, and transient data files with the first time step data appended. Note that the solver command window must be left open while the solver is calculating the first time step; if it is closed, the calculation will stop. Once the date and time stamp are displayed at the bottom of the run window, the calculation is finished and the window may be closed.

13.1.2 Verifying the model setup Run window contains GMV shortcut buttons for displaying boundary conditions, the initial particles in the The system colored by volume fraction, and the initial particles colored by species ID. Additional information required for validating the model setup is contained within the history.log and flux files. Boundary conditions Clicking on the View Boundary Conditions button on the Run window will open GMV and display the applied boundary conditions in the model. This GMV view, an example of which is shown in Fig. 13.2 (a), can be used to verify that the boundary has been applied to desired cells or faces in the model. The default view settings will show pressure boundaries as yellow cells, flow BCs as red faces, thermal walls as magenta faces, and point source injections as blue arrows. Additionally, the surface area of the boundary will be written to the info.log file or the boundary flux file, if one has been specified. Initial particles Clicking on the View Initial Particles, Volume Fraction or View Initial Particles, Species buttons will open GMV and display the particles in the system colored by volume fraction or species ID, respectively. Examples of these views are shown in Fig. 13.2 (b) and (c). Additionally, total system mass can be verified by looking in the history.log file. Flux planes Clicking on the View Flux Planes button will open GMV and display the flux plane locations. Flux are by default colored purple, as demonstrated in Fig. 13.2 (a). Transient data points Clicking on the View Transient Data Points button will open GMV and display the transient data points as blue dots, as demonstrated in Fig. 13.2 (a).

13.2 Starting a Simulation Run window. Once set up completely, the simulation can be started by clicking on the Run Solver button from the Upon clicking this button, the user may be asked what should be done with existing data files in the project folder, if any exist. After these prompts, which are discussed below, a command window for the solver will be automatically opened in which the solver will run. This window must remain open during the execution of the solver. Once the end time of the simulation is reached, a date and time stamp will be displayed at the bottom of the run window and the window may be closed.

User prompts at simulation start The GMV file prompt If GMV files exist in the project directory from a previous execution of the solver, the user will received the following prompt from the GUI: GMV graphic files exist, do you want to remove all GMV files? • Yes - GMV files in the project folder will be deleted. This is typically selected for a new simulation. • No - GMV files in the project folder will not be deleted. This is typically selected when a project is being restarted from an IC file. • Cancel - The solver startup will be ended.

204

Chapter 13. Run


Fig. 13.2: Check Setup in GMV (a) View Boundary Conditions (transient point and flux planes available through polygons menu); (b) View Initial Particles, Volume Fraction; and (c) View Initial Particles, Species The 2D data file prompt If 2D data files (.dat extension) exist in the project directory from a previous execution of the solver, the user will received the following prompt from the GUI: Data files with the extension .dat exist. Do you want to remove all of these files?. • Yes - 2D data files in the project folder will be deleted. This is typically selected for a new simulation. • No - 2D files in the project folder will not be deleted. This is typically selected when a project is being restarted from an IC file. • Cancel - The solver startup will be ended. Rate equation mass balance prompt In models with chemical reactions, the user may be prompted with the following warning that the reaction is not balancing. Your chemical rate equations have a summed error of #.###e## kg/s-m3 with minimum & maximum reaction rates being #.###e## & #.###e## kg/s-m3 under certain conditions. The error could be due to molecular weights not giving an exact mass balance. See info.log for more detailed information. If you think this error is acceptable, continue the calculation. Press 'c' to continue, 'q' to quit:

Since this can cause an unphysical increase or decrease in total mass due to the reaction, it is recommended that this type of error be fixed and therefore the user should quit the calculation by typing q and pushing Enter on the keyboard. Only in special cases would the user want to continue with the calculation by typing c and pushing Enter on the keyboard. If no errors are found in the stoichiometry of the reactions, it is possible that the source of error may be with the molecular weights of the reactants and products. To calculate the mass balance, Barracuda VR uses the molecular Base Materials library. The default molecular weights in the library, however, weight of each material in the

13.2. Starting a Simulation

205


are average values that factor in the presence of naturally occurring material isotopes. Unfortunately, these average molecular weights will not always balance exactly for a given chemical reaction. A typical remedy for this in Barracuda VR is to modify the molecular weights of those materials participating in reactions to ensure a proper mass balance, assuming a constant atomic weight. For example, if the following atomic weights are assumed: H = 1; C = 12; O = 16; then molecular weights of CO 2 , CO, H2 O, and O2 would be modified to values of 44, 28, 18, and 32, respectively. This would ensure proper mass balance for any reactions involving only these four components. The transient data file prompt If transient data points are defined or the model contains a thermal wall, there may be an transient data file in the project directory that was created during a previous running of the solver. If this file is detected by the solver, the user will receive the following prompt: One or more transient data files exist. (A)ppend (O)verwrite (N)ew (Q)uit?

• Append - New transient data will be added to the end of the existing file. This is typically used if the solver is restarting a simulation and continuing from the previous IC point. This option can be selected by typing a and pushing Enter on the keyboard. • Overwrite - The existing transient data will deleted and the new data will be written to the file. This is typically used for a new simulation. This option can be selected by typing o and pushing Enter on the keyboard. • New - The existing transient data will remain and the new data will be written to a different file: the existing transient data file with _00 appended to it. This option can be selected by typing n and pushing Enter on the keyboard. • Quit - The solver startup will be ended. This option can be selected by typing q and pushing Enter on the keyboard.

The solver window The solver window, shown in Fig. 13.3, displays the progress and status of a running simulation. After the completion of each time step, the solver will write out line to the window indicating the current solver state, convergence, and particle packing information in a series of columns.

Fig. 13.3: Solver Window Time (t) This column shows the current simulation time. The simulation will run until the specified end time is reached. Time step (dt) This column displays the length of the current time step. As discussed in Time Controls, the time step length is set by the user and may be subject to automatic adjustments based on the CFL number and temperature changes. Volume iterations (Vol itr) This column displays the number of iterations required to reduce the volume error to a value below the residual specified in Solver Advanced Settings Dialog or the maximum iterations if that was reached. Volume error (Vol err) This column displays the volume error at the end of the time step.

206

Chapter 13. Run


x-, y-, z-direction velocity iterations (u itr, v itr, w itr) This column displays the number of iterations required to reduce the x-, y-, and z-direction velocity errors to values below the residuals specified in Solver Advanced Settings Dialog or the maximum iterations if that was reached. x-, y-, z-direction velocity error (u err, v err, w err) This column displays the volume error value at the end of the time step. Enthalpy iterations (h itr) This column displays the number of iterations required to reduce the enthalpy error to a value below the residual specified in Solver Advanced Settings Dialog or the maximum iterations if that was reached. This column is only output for thermal calculations. Enthalpy error (h err) This column displays the enthalpy error at the end of the time step. This column is only output for thermal calculations. Pressure iterations (p itr) This column displays the number of iterations required to reduce the pressure error to a value below the residual specified in Solver Advanced Settings Dialog or the maximum iterations if that was reached. Pressure error (p err) This column displays the pressure error at the end of the time step. CFL number (CFL) This column displays the maximum CFL number in the model during the current time step. The CFL number is calculated as CFL = u∆t xcell which is a measure of how many cells of dimension x cell the fluid with velocity u will pass through in a single time step. The CFL number may be used to control the time step to maintain stability and accuracy of the model. See Time Controls for more information. QL number (QL) The QL number is a thermal analog to the CFL number that is typically not used. Loops (L) The “loops” number is number of times that the solver calculates the momentum, pressure, and energy equations in that time step. Typically this value is 1. Low, Med, Hi The “Low”, “Med”, “Hi” indicate the number of cells that are experiencing low, medium, or high levels of overpacking. During startup, it is common to see values in the low or medium columns but these should quickly reduce in number. If during a simulation these counts steadily begin to increase, it can be an indicator of potential difficulties and the user should investigate the cause. A common fix is to reduce the time step until the overpacking is eliminated. R This column indicates the number of times the particle stress equation calculations were repeated. Non-zero values can indicate that the solver is struggling which may lead to a failure.

13.3 Restarting a Simulation A running simulation will periodically produce IC files which later can be used to restart the simulation from the Run window, contains the Restart Solver Dialog, shown in Fig. 13.4. This dialog, which is accessed from the interface for selecting the IC file to restart from and the adjustment of solver parameters, model inputs, and data output upon restart.

Steps for restarting a simulation Select a restart file (IC) The IC files can be located in the project directory by clicking the Browse button and selecting the desired IC file from the file selection dialog that appears. The IC file can be either a restart file ( IC_###) or a backtrack IC ( IC_ ). These are discussed further in Time Controls. Change the simulation end time If a simulation is being restarted from a previously specified end time, a new End time will need to be entered for the simulation to continue. If this parameter is not updated, the simulation will run a single time step upon restart and finish, having again reached the previously specified end time. Make adjustments to other parameters, if needed It may be necessary to have the solver reread boundary condition or particle size distribution SFF files, change the wear state, or change other parameters on restart. These options are discussed below. Note: for any parameters that do not need to be changed on restart, simply leave the fields blank. Only items that have values specifically entered in the Restart dialog will be changed.

13.3. Restarting a Simulation

207


Fig. 13.4: Restart Solver Dialog

208

Chapter 13. Run


Begin simulation restart After making the necessary updates to parameters in the Restart Dialog, the simulation can be restarted by clicking the Restart button. This will open a new solver window and the user will be presented with the same series of prompts discussed in Starting a Simulation. In this case, the user will likely want to answer these prompts in a way that preserves the existing data in the project directory. In addition to the regular solver prompts, the user will be prompted about whether the solver should reset the averaging of data on restart. The average data user prompt Average Data is used to calculate time-averages of cell and particle data and writes out the averaged values to GMV files as the simulation runs. Upon restart, it is often desirable to reset the averaging so that the simulation only considers data produced after the restart. For this case, the GUI will prompt the user with the following question: Setting average on restart will force a new start of averaging (all prior averaging is discarded). If average is already set and you do not request averaging here on the restart, then the averaging will continue. At this point it is unknown whether averaging was on. If averaging was on, do you want to reset averaging on the restart?

The prompt provides two options: yes and no • Yes - This will reset the averaging upon restart. Only select this option if you specifically want the averaging to begin at restart. • No - This will not change the averaging. If averaging is currently on, it will not reset it.

Time parameters The Restart Solver dialog provides the interface for adjusting the following time-related parameters. If no value is provided in the fields, the previous simulation values will be used. With the exception of the Print interval and Plot interval fields, all time parameters are discussed further in Time Controls. Time step This is the current time step used in the simulation. End time This is the end time of the simulation. This needs to be set to later value if the previous end time and the IC restart time are the same. Print interval The terminal print interval determines how frequently time step data will be written to the terminal. The default for this is 1E-7 seconds. Plot interval The GMV plot interval is the simulation time interval between GMV file writes, as discussed in GMV Output Options. Restart interval The restart interval is the simulation time interval between the creation of restart IC files. CFL control The values set in the Min CFL and Max CFL text boxes will be used by the solver to automatically adjust the time step.

Model properties Reread BC input files Clicking this check box will instruct the solver to reread all SFF files associated with boundary conditions. This includes any SFF files associated with pressure and flow BCs, thermal walls, passive scalars, liquid injection BCs, and injection BC. Reread BC particle size distribution tables Clicking this check box will instruct the solver to reread all SFF files associated with species particle size distributions. Note that this change only will affect new particles being fed into the system through a boundary. Change thermal state As discussed in The Global Settings Window, a model can specified thermally but be initially started as an isothermal model. In this case, the model can be changed from isothermal to a thermal model on restart by clicking the Turn thermal calculations ON if they are not check box. Change chemistry state A model that has chemical reactions specified can have the chemistry state adjusted upon restart by clicking the Change chemistry state check box. This will enable the following options:

13.3. Restarting a Simulation

209


• Turn on chemistry - Chemical reaction can be turned “on” if chemistry was specified in the model but initial turned “off”, as discussed in The Global Settings Window. • Ramp on from ___ s to ___ s - Chemical reactions can be ramped “on” over the specified time from if chemistry was specified in the model but initial turned “off”, as discussed in The Global Settings Window. • Turn off chemistry - Chemistry can be turned off in the model upon restart. Change wear state If the Wall Erosion model has been specified but is turned off, it can be turned “on” or “off” by clicking the Change wear state check box. The wall erosion model is discussed further in Wall Erosion.

Data output Resetting particle attrition Clicking the Reset attrition to 0 check box will reset the “trauma” value in the Particle Attrition model for each particle. The particle attrition model is discussed further in Particle Attrition. Resetting wall erosion Clicking the Reset wear to 0 check box will reset the “impact” value in the model for each particle. The wall erosion model is discussed further in Wall Erosion.

Wall Erosion

Graphics output variables The GMV output variables can be changed upon restart by clicking the Graphics output variables button on the Restart Solver Dialog. This will display a GMV Output Options Dialog that is a clone of the GMV Output Options Window for selecting GMV output. A description of each variable is provided in GMV Output Options. Average data variables The average data variables can be changed upon restart by clicking the Average data variables Average Data button on the Restart Solver Dialog. This will display an Average Data Dialog that is a clone of the Window for selecting the GMV data to be averaged. A description of GMV data averaging is provided in Average Data. Raw data variables The raw data output selections can be changed upon restart by clicking the Raw data variables Raw Data Window button on the Restart Solver Dialog. This will display a Raw Data Dialog that is a clone of the for selecting the raw data and frequency of output. A description of raw data output is provided in Raw Data.

Solver convergence The solver convergence parameters for Volume, Pressure, and Velocity can be changed on restart by entering new maximum iterations or residuals in the text boxes. The solver convergence parameters are discussed further in Solver Advanced Settings Dialog.

13.4 Interacting with a Running Simulation Run Window. The Interact The user can interact with a running simulation by clicking on the Interact button on the Dialog, shown in Fig. 13.5, provides an interface for adjusting solver time and convergence parameters, rereading SFF files for boundaries and PSDs, and creating output data at the current simulation time. Once adjustments have been made, the user needs to click on the Submit button to pass the changes to the solver. Note: for any parameters that do not need to be changed, simply leave the fields blank. Only items that have values specifically entered in the Interact dialog will be changed.

Time settings The Interact dialog provides the interface for adjusting the following time-related parameters for a running simulation. With the exception of the Print interval and Plot interval fields, all time parameters are discussed further in Time Controls . Time step This is the current time step used in the simulation.

210

Chapter 13. Run


Fig. 13.5: Interact Window

13.4. Interacting with a Running Simulation

211


End time This is the end time of the simulation. The simulation will end if this is set to a value that is less than the current simulation time. Doing so is considered a “safe” approach for immediately ending a simulation because the solver will write out a restart file prior to finishing. Print interval The terminal print interval determines how frequently time step data will be written to the terminal. The default for this is 1E-7 seconds. Plot interval The GMV plot interval is the simulation time interval between GMV file writes, as discussed in GMV Output Options. The new interval will start after the next scheduled GMV file write unless the Start new plot intervals now box is clicked. Restart interval The restart interval is the simulation time interval between the creation of restart IC files. CFL control The values set in the Min CFL and Max CFL text boxes will be used by the solver to automatically adjust the time step. QFL control Control by QFL number is a legacy feature that is not used any more.

Model properties Reread BC input files Clicking this check box will instruct the solver to reread all SFF files associated with boundary conditions. This includes any SFF files associated with pressure and flow BCs, thermal walls, passive scalars, liquid injection BCs, and injection BC. Reread PSD input files Clicking this check box will instruct the solver to reread all SFF files associated with species particle size distributions. Note that this change only will affect new particles being fed into the system through a boundary.

Data output Dump restart (IC_###) Checking this box will instruct the solver to create a restart IC file at the current time. Dump graphics (SnapGmvfile) Checking this box will instruct the solver to write out a GMV file at the current time. This gmv file will have a SnapGmv file name to separate it from the main GMV series. Dump population (POPUL...) Checking this box will instruct the solver to create particle population file at the current time. Population files are discussed in Population files. Dump all particles (p_ all...snap) Checking this box will instruct the solver to create a set of particle snap files at the current time. Particle snap files are discussed in Raw Data. Resetting particle attrition Clicking the Reset attrition to 0 check box will reset the “trauma” value in the Particle Attrition model for each particle. The particle attrition model is discussed further in Particle Attrition. Resetting wall erosion Clicking the Reset wear to 0 check box will reset the “impact” value in the model for each particle. The wall erosion model is discussed further in Wall Erosion.

Wall Erosion

2D plot data 2D plot data files can be written at the current time by clicking the 2D Plot Data button on the Interact 2D Plot Data Window for specifying output Dialog. This will display a 2D Plot Data Dialog that is a clone of data. A description of the 2D plot data dialog variable is provided in 2D Plot Data.

Solver convergence The solver convergence parameters for Volume, Pressure, and Velocity can be changed through the Interact Dialog by entering new maximum iterations or residuals in the text boxes. The solver convergence parameters are discussed further in Solver Advanced Settings Dialog .

212

Chapter 13. Run


13.5 Running from the Command Line The Barracuda VR solver can be run remotely through the command line on Linux systems. Typically, the user will create a project file using the Barracuda VR GUI. This solver can be run from the project directory by typing cpfd.x.17 my_project_file.prj

where my_project_file.prj is an example project file name. The actual project file name must be used.

13.5.1 Solver flags Flags can be added to the solver command to run a single time step or automatically answer the user prompts. The answering of user prompts can be necessary for automated scripting of multiple simulations. Running a single time step The user can add the -setup flag to the solver to instruct it to run a single time step. Additional information on running a single time step is in Running a single time step. cpfd.x.17 my_project_file.prj -setup

Printing version information The user can add the --version flag to the solver command to have it print information on the solver version. This is not used with a project file. cpfd.x.17 --version

Handling the rate equation mass balance prompt The user can automatically answer the rate equation mass balance prompt, discussed in Starting a Simulation, by including the -contChem flag to instruct the solver to continue or the -quitChem flag to quit the calculation. cpfd.x.17 my_project_file.prj -contChem

or cpfd.x.17 my_project_file.prj -quitChem

Handling the transient data file prompt The user can automatically answer the transient data file prompt, discussed in Starting a Simulation, by including the -ow flag to instruct the solver to overwrite existing transient data files if they exist or the -aw flag to append new transient data to the existing files. cpfd.x.17 my_project_file.prj -ow

or cpfd.x.17 my_project_file.prj -aw

Listing all command-line options The solver will list all available command-line options when the --help flag is used: cpfd.x.17 --help Usage: cpfd.x.17 projectfile # [-overWriteFile] [-ow] # [-appendFile] [-aw] # [-setup] [-s] [-contChem] [-cc] # # [-quitChem] [-qc] # [-dataFormatA] [-dfa] # [--version] [-time] # [-minDiskPercent ] #

13.5. Running from the Command Line

Overwrite existing transient data files ... Append to existing transient data files ... Run a single timestep for setup... Continue with simulation even if chemical ... Stop simulation if chemical reactions are ... Output ASCII format for displaying Gmv files Output solver version Output simple timing information to end of ... Pause solver if free harddrive space falls ...

213


[-dieFE] [-help | --help] [-gpu] [-dprompt] [-dlist] [-dauto [method]] [-device ] [-d0] [-d1] [-d2] [-d3] [-binarydiff] [-fallback ] [-dryrun]

# # # # # # # # # # # # # #

Automatically exist solver if floating ... Show this usage brief or more information ... Use NVIDIA GPU to accelerate simulation (Default) Prompt user to select GPU when ... List capable GPUs in -dprompt format without ... Auto-select a GPU by: (Default: waterfall), ... Use CUDA enumerated device for GPU ... Use CUDA enumerated device 0 for GPU acceleration Use CUDA enumerated device 1 for GPU acceleration Use CUDA enumerated device 2 for GPU acceleration Use CUDA enumerated device 3 for GPU acceleration Use slower but deterministic GPU accelerated ... Prompt user for action if CUDA error ... Outputs GPU memory requirements and exits

13.5.2 Command line restart The command line can also be used to restart a simulation from an IC file. This is done with the cpfd.x.17 solver command; a special restart project file; and any necessary Solver flags. It should be mentioned, though, that the reaction mass balance is not checked by the solver on restart. Therefore, the transient data file prompts are the only flags of relevance for a simulation restart from the command line. cpfd.x.17 my_restart_file.prj

Restart project file A restart project file is a text file that contains a reference to the IC file to restart from and possibly a new end time, time step, or any of the other parameters that can be set through the Restart Dialog, discussed in Restarting a Simulation. An example restart project file that instructs the solver to restart from IC_ , change the end time to 200 seconds, and change the timestep to 3 milliseconds is shown below. restart IC_ endTime 200 dt 1e-3 endRestart

While this type of file can be created manually, it is usually easiest to use the Restart Dialog, discussed in Restarting a Simulation. When a calculation is restarted with the Restart Dialog, a hidden file .restart.prj is created in the project directory. This file will contain all of the specified adjustments and can be copied to a new file for later restart from the terminal. cp .restart.prj my_restart_file.prj

13.6 Data file reference The following is a description of files that will be created by the solver. Some files, such as history.log, will be written every time the solver is run. Others, such as the transient data files, will only be created if specified in the Data Output windows of the model. While these files are listed here, the main reference is in the appropriate section Data Output chapter. of the

214

Chapter 13. Run


13.6.1 History file The history.log file is an enhanced record of the command window data output. As such, many of the columns of the history.log file are identical to what is written to the solver command window. Columns 1 through 12 of history.log: # # #

1 t s

2 dt s

3 Vol itr

4 Vol err

5 u itr

6 u err

7 v itr

8 v err

9 w itr

10 w err

11 h itr

12 h err

Columns 13 through 26 of history.log: # # #

13 p itr

14 p err

15

16

CFL

QL

17 pMass kg

18 CPU s

19 L

20

21

Low Med

22 Hi

23 24 25 26 Ave Max Itr R ODE ODE ODE

Columns 1 through 16 and 19 through 23 are discussed for the solver window in Starting a Simulation. The additional columns are the following: Particle mass (pMass) This column displays the total amount of solids in the model during the timestep. CPU run time (s) This column displays the total amount of time that the solver has been running. Ave and Max ODE These columns give the average and max iterations that the chemistry ODE solver has required to converge to a solution in a cell. Itr ODE This column displays the number of times that the ODE solver for chemistry has needed to be reset and run to converge in a cell.

13.6.2 Info file The info.log file is created by the solver and contains valuable information and messages generated during the initialization of the simulation. As the simulation runs, additional information and warnings will be written to the file by solver. Information output to the info.log file includes analysis of the model grid, particle PSDs, solid materials, boundary conditions, chemical reactions, and the size of computational particles in the model.

13.6.3 Flux files Flux files are optionally produced at flow boundaries and at specified flux planes. The output format for these files is discussed in detail in Flux Planes.

13.6.4 Transient data files Transient data files contain the output from transient data points, discussed in Transient Data.

13.6.5 Population files Population logs contain information on the size and distribution of particles in the model. Population files, which have file names starting with POPUL_ , are written for each species into the project directory at the same rate as restart IC files. The header of a population log contains information on the total volume, surface area, particle mass, number of physical particles, and number of compuational particles in the model. Although the information contained within population files can be determined from particle Raw Data, the population files can be a useful reference for verifying that the computational particles are well-balanced in the model.

13.6. Data file reference

215


216

Chapter 13. Run

CHAPTER

FOURTEEN

POST-RUN

A Barracuda VR simulation can produce a wide range of data about a modeled system including three-dimensional GMV files and high frequency frequency transient transient data. To view view and analyze analyze this output output data, data, Barracuda Barracuda VR comes bundled with two additional software packages: General Mesh Viewer (GMV) , a three-dimensional scientific visualization tool for viewing fluid and particle data within the geometry; and XMGR/QtGrace, a two-dimensional plotting tool for Post-Run window, shown in analyzing analyzing transient transient data, 2D plot data, and other other text-ba text-based sed output. The Post-Run window, in Fig. Fig. 14. 14.1 1, contains links to both of these software packages along with a variety of other tools: a simple file viewer for opening log files, a terminal prompt, a tool for viewing images and animations, an image and animation conversion utility, a GMV to Ensight Gold file converter, and a tool for creating support files. On the right side of the window, a number of GMV shortcut buttons are included for quickly viewing output data.

Fig. 14.1: Post-Run Window Window

217


Analyzing model startup Simulation output data can be viewed in both GMV and XMGR/QtGrace as soon as it is written into the project folder by the solver and it is highly recomme recommended nded that the user do so. When a system is just beginning beginning to fluidize fluidize and chemical reactions are beginning, the model behavior will often be less a function of the boundary conditions (which are usually well known) and more a function of the initial conditions (which usually have some assumptions built in). A major component of the simulation run time is the time it takes for the model to move naturally to an established operating condition that is purely a function of the boundary conditions and is independent of the initial state of the model. The rate at which this happens is often a strong function of the initial conditions, and therefore is within the user’s control. There are four areas that the user should examine to determine whether the model is running correctly: solids mass, thermal mass, solid components, and any time step limiters. Solids mass Solids Solids mass refers refers to the total amount of solid particles particles within within the geometry geometry.. A natural natural increase increase or decrease in solids mass will be a long simulation process on systems where the mass feed rate into the system is small compared compared to the total mass. Typically ypically,, it is much more efficient efficient to set a system system mass that is close to the target target up front front by adjusting adjusting the particle particle ICs. When the system value is known exactly exactly, it can easily easily be checked by running running a single time step and viewing the total mass in history.log. Often, however, the system mass is not measured but is instead inferred from some other measure, such as pressure drop or a bed height. To verify these values, the model will need to run until the bed height or pressure drop can be confirmed. Thermal mass Thermal mass Thermal mass refers to total enthalpy within the system. As with solids mass, natural adjustments can take a very long time and it is much more efficient to establish a thermal mass that is close to the target by adjusting the particle and fluid IC temperatures. This can be difficult to set exactly, but the use of transient data points placed at thermocouple locations and analysis of transient changes in the enthalpy can help. This is only a consideration for thermal models, discussed in The Global Settings Window. Solid components For components For complex multicomponent reacting systems, it can be important to start with a good guess for the initial composition of individual solid species in the system. Consider the example of a coal or biomass gasifier. In this system, organic material is fed into the system that subsequently dries and devolatilizes to produce a char material which itself itself will react with water or oxygen in the system. system. During During a steady steady operation, operation, the solids solids in the system are going to be in some state of reaction reaction resulting resulting in a mixture mixture of feed organics organics and char. char. As with the total solids solids mass and thermal mass, natural changes can be a long process, depending on the feed rates and reaction kinetics of the model. Accurate estimates for composition composition can be determined from a high level look at the reaction chemistry and solids in the system and the compositions can be examined for natural changes by examining raw particle data files during model startup. It is often much more efficient to adjust initial compositions through the composition of particle ICs than to let compositions develop naturally. This is mainly a concern for multicomponent models with chemistry. Time step limiters Since the time step is a major factor in the calculation rate, the user should verify that the initial conditions do not create conditions that artificially limit the time step through high CFL numbers or large temperature changes in a cell. For example, a substantial difference between the pressure at a boundary and the fluid IC pressure will create a sharp pressure difference and high flow rates at the start of a simulation that would not exist at an established operating condition. condition. Initial gas compositions are another consideration in this area; the initialization of a combustive mixture (and the chemistry to do so) would likely lead to large temperature changes at the start of simulation. In this case, there will likely be combustion somewhere in the model during an established operation but it may be better to allow this to develop naturally rather than force it at the simulation start. Beyond the four areas discussed above, the simulation startup is also an important time to critically examine the model. The following questions can be important to consider as the model startup is being evaluated. • Is the system fluidizing or otherwise starting correctly? correctly? If not, are there better initial conditions for the particles particles and fluid that could be set? • Are the boundary condition conditionss appropri appropriate? ate? For example, example, if a uniform uniform boundary boundary is used for fluidizing fluidizing air, does this make sense? Should an array of point source injections be used instead? • During startup, are the chemical reaction reaction rates reasonable? Is the initial temperature high enough for the chemical reactions to proceed? Is there too little reactant or too much?

21 8

Chapter 14. Post-Run


Analyzing model results The analysis to be performed after a model has finished is dependent on the type of model and the reasons that a model was setup and run in the first place. place. While While it is ultimately ultimately up to the user to identify identify what the goals of the model are and what is to be learned from the development of the model, there are some common approaches to analysis that can be discussed. Fluidization patterns The patterns The effect of internal structures on fluidization patterns in an FBR is studied. The analysis uses velocity velocity and mass flux data for both particles particles and fluids to examine examine the manner of the flow. flow. Time-a Time-avera veraged ged data is especially useful for smoothing out variations in the flow of a bubbling or slugging system. Thermal distribution in an FBR Temperature FBR Temperature effects are studied in an industrial scale FBR, possibly looking at the interactions between heat transfer and chemical reactions occurring in the bed. Since hot spots need to be avoided to prevent damage to catalyst in the bed, the model is used to study different geometry configurations and flow rates to optimize reactions in the model and reduce thermal gradients. Effect of changing operating conditions in a reactor To study the effects of proposed changes in reactor operating conditions, a model is created for the baseline operation. Flux data at the model outlets and transient temperature data points are compared against experimental plant data to validate the kinetics of the model. Once validated, the model is run again with the proposed changes to study the effects on reactor efficiency, temperatures, and particle entrainment. Cyclone erosion A erosion A cyclone model is setup to study flow of particles within the cyclones and areas of the walls that are susceptible to wear. The model is used to study how potential modifications to the geometry may reduce the erosion. Wall Erosion model, Erosion model, discussed in Wall Erosion. This type of study uses the

14.1 Data Visualizatio Visualization n Tools Tools Post-Run window The Post-Run windo w provides provides links to GMV, GMV, XMGR/QtGra XMGR/QtGrace, ce, a file viewer, viewer, and a terminal terminal window window. These These programs form the foundation of data visualization in Barracuda VR.

14.1.1 14.1 .1 GMV General Mesh Viewer (GMV) is the bundled three-dimensional data visualization tool in Barracuda VR, shown in Fig. 14.2.. GMV is the default program that is opened when the user clicks on the shortcut buttons for viewing the CAD 14.2 and gridded model when setting up the grid ( Setup Grid ), ), boundary conditions and initial particles when checking the Post-Run window. The GMV Users Manual can be accessed through model setup ( setup ( Run), and model results via the Post-Run window. the Help list located on the Menu bar . When GMV is opened, the user will need to set up the display of the cells and particles to view the desired output. Since this can be tedious at times, users will often use the GMV shortcut buttons for viewing output data. GMV shortcut buttons are discussed in Viewing GMV results.

14.1.2 XMGR/Q XMGR/QtGrace tGrace XMGR (on Linux), or QtGrace (a port of Xmgr to Windows), is a two-dimensional plotting utility that can be used for plotting transient data, flux data, or 2D plot data. A sample XMGR/QtGrace window is shown in Fig. 14.3. 14.3. XMGR is documented at the XMGR website: http://plasm http://plasma-gate.weizman a-gate.weizmann.ac.il/Xmgr/ n.ac.il/Xmgr/ . XMGR can be used to view the current data and later save the image as an Encapsulate Encapsulated d PostScri PostScript pt (eps) image file. Often Often a user will convert convert this file to a differe different nt format, such as JPEG or PNG, for distributi distribution on or inclusio inclusion n in written reports. reports. Barracuda Barracuda VR includes includes an image conversion utility, utility, discussed in Converting images and animations.

14.1. Data Visualization Tools

2 19


Fig. 14.2: GMV window showing showing sample sample model

22 0



Fig. 14.3: XMGR/QtGrace sample window

14.1. Data Visualization Tools

2 21


14.1.3 View log files Barracuda VR includes a menu item for opening text-based files such as log files, transient data, 2D plot data files, etc. When the View Log Files button on the Post-Run window is clicked, the user will be prompted with a file open dialog in which a file can be selected for viewing. The system’s default text editor will be used to open the selected file. On Windows, the default text editor is usually Notepad , and on Linux it is usually gedit. However, both systems have options so that the default program can be changed to a different text editor if desired.

14.1.4 Terminal window A terminal window is a useful tool for managing files in the project directory or running automated scripts for postprocessing. Clicking on the Open terminal button in the Post-Run window will open a terminal with the command prompt located in the project directory.

14.2 Viewing GMV results Default Post-Processor The selected post-processor will be started when any visualization shortcut buttons are used in the Barracuda VR GUI. GMV is installed with Barracuda VR and is selected as the default post-processor for new projects. EnSight can be selected as the default post-processor and used if it is installed and a valid EnSight license is available. Note that CPFD Software, LLC, does not sell EnSight. Contact CEI (http://www.ensight.com ) directly for information about purchasing EnSight.

Post-Run Visualization Shortcut Buttons GMV shortcut buttons, located in the Post-Run window, provide a convenient way to quickly view threedimensional Barracuda VR results. The buttons are organized for viewing data for particles, cells, and time-averaged data. Note: The appropriate GMV output data needs to be selected in the GMV Output Options and Average Data windows. Table 14.1: View Particle Data Particle Volume Fraction Particle Solid Region Particle Radius Particle Species Particle Residence Time Particle Temperature

222

Displays particles colored by particle volume fraction. Cells are semitransparent so that the geometry is visible as well. An example of the shortcut display is shown in Fig. 14.2. Displays an isovolume of particle volume fractions above 0.45. Cells are semitransparent so that the geometry is visible as well. Particles are not displayed. Displays particles colored by particle radius in units of microns. Cells are semitransparent so that the geometry is visible as well. Displays particles colored by particle species ID. Cells are semitransparent so that the geometry is visible as well. Displays particles colored by residence time. Cells are semitransparent so that the geometry is visible as well. Displays particles colored by particle temperature. Cells are semitransparent so that the geometry is visible as well.



Table 14.2: View Cell Data Cell Volume Fraction Cell Pressure Cell Pressure Gradient Cell Fluid Temperature

Displays cells colored by particle volume fraction. Particles are not displayed. Displays cells colored by the fluid pressure. Particles are not displayed. Displays cells colored by pressure gradient in z direction (dp/dz). Particles are not displayed. Displays cells colored by fluid temperature. Particles are not displayed. Table 14.3: View Average Cell Data

Cell Average Volume Fraction

Cell Average Pressure

Cell Average Pressure Gradient Cell Average Fluid Temperature

Displays cells colored by the time-averaged particle volume fraction. Note that Particle volume fraction must be selected in the Average Data window. Zeros will be displayed for the averaged data if averaging start time is later than the simulation time of the results being viewed. Particles are not displayed. Displays cells colored by the time-averaged fluid pressure. Note that Pressure must be Average Data window. Zeros will be displayed for the averaged data if selected in the averaging start time is later than the simulation time of the results being viewed. Particles are not displayed. Displays cells colored by the time-averaged pressure gradient in z-direction (dp/dz). Note Average Data window. Zeros will be displayed for the that dp/dz must be selected in the averaged data if averaging start time is later than the simulation time of the results being viewed. Particles are not displayed. Displays cells colored by the time-averaged fluid temperature. Note that Fluid temperature Average Data window. Zeros will be displayed for the averaged must be selected in the data if averaging start time is later than the simulation time of the results being viewed. Particles are not displayed.

14.3 Post-run utilities The Post-Run window contains some utilities for viewing images and animations, converting images and animations, converting GMV files to the Ensight Gold file format, and creating a support file.

14.3.1 Viewing images and animations Images and animations can be viewed by clicking the View images or animations button which will bring up a image viewing dialog, shown in Fig. 14.4. By default, a selected image will be viewed with the ImageMagick program and animations will be played with xanim. This can be changed by clicking on the Change button which will raise a program selection dialog. The user can select from ImageMagick , Gwenview, Gimp, XV , or Kolourpaint for image viewing and xanim, Kaffiene, and Noatun for animation viewing. Viewing images and animations from the command line Although Barracuda VR includes a dialog for viewing images and animations, often users will use the command line for viewing images and animations. To view images using ImageMagick from the command line, the user can open a terminal (discussed in Terminal window) and type the following at the prompt: display my_image.jpg

To view an animation from the command line, the user can type the following at the prompt: xanim my_animation.mpg

14.3. Post-run utilities

223


Fig. 14.4: Select image or animation dialog

224



14.3.2 Converting images and animations Barracuda VR contains a utility for converting images to other formats and creating animations from a set of images. This dialog, shown in Fig. 14.5, is accessed by clicking on the Convert image or animation button in the Post-Run window.

Fig. 14.5: Image and animation converter dialog Converting images Images can be converted to a variety of formats by selecting images from the file selection window, selecting a file format from the Convert image(s) to drop down box, and clicking the Convert button. The available file formats to convert to are jpg, png , gif , xpm , and eps. Creating animations Animations can be created from a set of sequentially-numbered images by selecting the images in the file selection window, selecting Create a movie using images, providing a Movie file name, and clicking the Convert button. The user can also select between using the mencoder and convert utilities, and a variety of codecs listed in the Movie codec drop down box.

14.3.3 Convert Gmv to Ensight Gold Format Note: CEI’s Ensight now supports directly reading GMV files. If you are using Ensight version 10.0.3b or newer, you do not need to use this conversion utility. Barracuda VR includes a utility to convert GMV files to Ensight Gold format. The target application for this conversion utility is when users want to analyze Barracuda VR results in a post-processor that supports Ensight Gold format, but not GMV format. The conversion dialog is shown in Fig. 14.6. This dialog can be opened by clicking on the Convert GMV to Ensight files button in the Post-Run window.


225


Fig. 14.6: GMV to Ensight file converter All files in a series can be converted by selecting the first Gmv file from the series (typically Gmv.00000), selecting the Convert All GMV files start with... radio button, and clicking the Convert button. Alternatively, a specified number of GMV files can be converted, starting with the selected GMV file, by clicking the Convert ## GMV files starting with... radio button. A binary file type is recommended unless the Ensight-format files are to be used in the Paraview program - an open source alternative to GMV and Ensight. Additional information about Ensight is located at http://www.ensight.com. Information about Paraview is located at http://www.paraview.org.

14.3.4 Creating a Support File When contacting CPFD Support for help with a model, a user will often be asked to send a “support file”. A support file is an archive file ( .tar.gz on Linux, .zip on Windows) containing all project files required to run the model that can be created in the project directory by clicking the Create support file button in the Post-Run window and clicking Create package. The support file will include the project file, all SFF files for boundary specifications and particle size distributions, any important scripts, GMV attribute files (which are used to store display settings), XMGR parameter files (which are used to store graphing settings), and some files that are helpful for debugging such as history.log and info.log. The CAD file will be included in the support file as well when the Include CAD check box is selected. While the CAD file is necessary for restarting the model, it can be left out of the support file if there are concerns about sending proprietary design information. When possible, the inclusion of the CAD file is recommended. The support file dialog is shown in Fig. 14.7.

14.3.5 Colormap editor GMV can color particles or cells based on the contents of a color map file ( .cmap). Barracuda VR includes a colormap utility called bldcol for creating a map that transitions smoothly between two colors. The bldcol dialog, shown in Fig. 14.8, can be opened by clicking on Post-processing from the menu bar and selecting Launch bldcol. Selecting colors A top and bottom color are selected by clicking on the color swatch which will raise a color selector dialog.

226



Fig. 14.7: Create Support File Dialog


227


Fig. 14.8: bldcol Dialog

228



Adjusting nonlinearity The color map can be biased towards either the top or bottom color by adjusting the Color nonlinearity value. The bottom color will be more dominant when positive values are used; the top color will be more dominant when negative values are used. Saving a color map The color map will be saved in the project directory with the specified File name when the Build button is clicked.


229


230


CHAPTER

FIFTEEN

PLOT MANAGER

When running a Barracuda VR simulation, it is a good idea to monitor certain output data. Plot Manager is an integrated plotting system in Barracuda VR, where all plotting definitions are stored in the project file. It provides a convenient interface for definition and organization of 2D data plots. An effective way of using Plot Manager is to define plots for any output data that needs to be monitored, periodically use the Graph All button to view all plots, and then make any modifications necessary by Interacting with a Running Simulation. The

Plot Manager Window, shown in Fig. 15.1, displays a list of all plots for the simulation and follows the Add,

Edit, Copy, Delete GUI pattern. Creating

a Plot

Fig. 15.1: Plot Manager Window

Plots are created in the Plot Operations dialog box shown in Fig. 15.2. This dialog is activated when a new plot is added or when an existing plot is edited. Name A unique, identifying name for the plot is required. Title An identifying title is recommended, but not required for creating a plot. Subtitle A subtitle can be recorded but is not required for creating a plot. X-Axis Title A title with dimensions is recommended, but not required for creating a plot. Y-Axis Title A title with dimensions is recommended, but not required for creating a plot. Make Image of Plot This box can be checked when an Encapsulated Postscript (.eps) file of the plot is desired. If other formats are needed, use the Convert image or animation button in the Post-Run menu. An image name must be specified. Legend Specify whether to hide or show legend. File At least one data file must be selected. Use the Browse... button to select the file from which data should be plotted. Use the Add, Remove, and Copy buttons to conveniently manage the files for each plot. Mulitiple columns from the same file or various files can be used in each plot. Use the Check Data button to determine if all required fields are filled in.

231


Fig. 15.2: Plot Operations Preview Use the Preview button to see the comment headers of the data file to identify the X - and Y -columns from which data should be plotted. X The column number corresponding to the X-Axis data. Y The column number corresponding to the Y-Axis data. Color, Style, Width, and Symbol The line formatting options available for the plot. To view the plot, use the Plot button in the Plot Operations dialog or Launch Plot in the Plot Manager window. In the Title, Subtitle, X-Axis Title, Y-Axis Title, and Line Name fields, Xmgr escape sequences can be used for special text formatting or characters. Examples of commonly used formats are:

Formatting

Escape Sequence

Example

Superscript

\S

m/s\S2\N

Subscript

\s

Mass\sall\N

Normal script

\N

V\ss\N Plot

Greek letters

\8

\8abc

Roman letters

\4

\8q\4\sp\N

232

Result m/s2 Mass all Vs Plot

αβχ θp

Chapter 15. Plot Manager

CHAPTER

SIXTEEN

USEFUL FORMULAS

It is often useful to perform some estimates or hand calculations when setting up, running, and post-processing Barracuda VR simulations. This chapter contains some of the most commonly used formulas and equations.

16.1 Estimating Particle Volume Fraction at a BC When feeding particles at a BC, it is sometimes necessary to estimate the particle volume fraction. The most common occasion for needing this information is when specifying the Number density manual parameter for a Flow BC that is bringing both fluid and particles into the system. A rule of thumb for choosing an appropriate Number density manual is: Number density manual =

50 θ p

However, instead of knowing θ p , it is often the case that you will know the flow rates and densities of fluid and particles:

�� kg s

˙ f = fluid flow rate m

kg s

m ˙ p = particle flow rate ρf = fluid density

kg m3

ρ p = particle density

kg m3

Making the assumption that the fluid and particles are being fed at the BC at the same velocity, the particle volume fraction, θ p , can be estimated as:

θ p =

1

� �� m ˙ f m ˙p

ρp ρf

+1

16.2 Converting from Mass Fraction to Volume Fraction For a system of multiple materials, it is often useful to convert from mass fraction to volume fraction. An example where this type of conversion is necessary is when initializing particle species in a Barracuda VR simulation. Often, you will know the desired mass fractions of the initial particle species, but in Barracuda VR it is necessary to specify the initial particle volume fraction for each species, and the sum of all individual species volume fractions must equal some value less than the close-pack volume fraction.

233


Assuming a collection of N distinct species, where the mass fraction, X , and density, ρ, of each species is known, and the final mixture of all species needs to have an overall volume fraction of θTarget , the volume fraction, θ , for a specific material i is given by:

θi = θ Target

X i /ρi N

∑

j=1

X j /ρj

As an example, consider a system of three species, A , B , and C . To calculate the volume fraction of species A , the above formula would be expanded to:

θA = θ Target

X A /ρA X A /ρA + X B /ρB + X C /ρC

16.3 Converting from Mass Fraction to Mole Fraction For a system of multiple materials, such as multiple gas species, it is often useful to convert from mass fraction to mole fraction. Assuming a collection of N distinct materials, where the mass fraction, X , and molecular weight, M W , of each material is known, the mole fraction, Y , for a specific material i is given by:

Y i =

X i /MW i N

∑

X j /MW j

j=1

As an example, consider a system of three materials, A , B , and C . To calculate the mole fraction of material A , the above formula would be expanded to:

Y A =

X A /MW A X A /MW A + X B /MW B + X C /MW C

16.4 Converting from Mole Fraction to Mass Fraction Assuming a collection of N distinct materials, where the mole fraction, Y , and molecular weight, M W , of each material is known, the mass fraction, X , for a specific material i is given by:

X i =

Y i ∗ MW i N

∑

j=1

Y j ∗ MW j

As an example, consider a system of three materials, A , B , and C . To calculate the mass fraction of material A , the above formula would be expanded to:

X A =

234

Y A ∗ M W A Y A ∗ M W A + Y B ∗ MW B + Y C ∗ MW C

Chapter 16. Useful Formulas

CHAPTER

SEVENTEEN

REFERENCES

235


236

Chapter 17. References

BIBLIOGRAPHY

[AORourkeB89] A.A. Amsden, P.J. O’Rourke, and T.D. Butler. A computer program for chemically reactive flows with sprays. Technical Report LA-11560-MS, Los Alamos National Lab, 1989. [BvdHK07] R. Beetstra, M. A. van der Hoef, and J. A. M. Kuipers. Drag force of intermediate reynolds number flow past mono- and bidisperse arrays of spheres. AIChE Journal, 53(2):489–501, 2007. URL: http://dx.doi.org/10.1002/aic.11065, doi:10.1002/aic.11065. [BPN+94] Alexis T. Bell, Isabelle Pitault, David Nevicato, Michel Forissier, and Jean-René Bernard. Chemical reaction engineering: science & technology kinetic model based on a molecular description for catalytic cracking of vacuum gas oil. Chemical Engineering Science, 49(24):4249 – 4262, 1994. URL: http://www.sciencedirect.com/science/article/pii/S0009250905800181 , doi:http://dx.doi.org/10.1016/S00092509(05)80018-1. [CAS99] R.P. Chhabra, L. Agarwal, and N.K. Sinha. Drag on non-spherical particles: an evaluation of available methods. Powder Technology, 101(3):288 – 295, 1999. URL: http://www.sciencedirect.com/science/article/pii/S0032591098001788 , doi:http://dx.doi.org/10.1016/S00325910(98)00178-8. [Erg49] S. Ergun. Fluid flow through packed columns. Chemical Engineering Progress , 48:89, 1949. [FZ98] Liang-Shih Fan and Chao Zhu. Principles of Gas-Solid Flows. Cambridge University Press, New York, 1998. [Gid94] D. Gidaspow. Multiphase Flow and Fluidization. Continuum and Kinetic Theory Description . Academic Press, Boston, 1994. [HL89] A. Haider and O. Levenspiel. Drag coefficient and terminal velocity of spherTechnology, ical and nonspherical particles. Powder 58(1):63 – 70, 1989. URL: http://www.sciencedirect.com/science/article/pii/0032591089800087 , doi:http://dx.doi.org/10.1016/00325910(89)80008-7. [HSWL13] Kun Hong, Zhansheng Shi, Wei Wang, and Jinghai Li. A structure-dependent multi-fluid model (sfm) for heterogeneous gas–solid flow. Chemical Engineering Science, 99():191 – 202, 2013. URL: http://www.sciencedirect.com/science/article/pii/S0009250913003813 , doi:http://dx.doi.org/10.1016/j.ces.2013.05.050. [KL91] Daizo Kunii and Octave Levenspiel. Fluidization Engineering. Butterworth-Heinemann, Newton, MA, second edition, 1991. [Lev72] Octave Levenspiel. Chemical Reaction Engineering. John Wiley & Sons, New York, first edition, 1972. [LGW+13] Jinghai Li, Wei Ge, Wei Wang, Ning Yang, Xinhua Liu, Limin Wang, Xianfeng He, Xiaowei Wang, Junwu Wang, and Mooson Kwauk. From Multiscale Modeling to Meso-Science: A Chemical Engineering Perspective. Springer-Verlag, Berlin Heidelberg, 2013. [LK94] Jinghai Li and Mooson Kwauk. Particle-Fluid Two-Phase Flow, The Engergy-Minimization Multi-Scale Method . Metallurgical Industry Press, Beijing, 1994.

237


[LCL+13] Bona Lu, Congli Cheng, Weimin Lu, Wei Wang, and Youhao Xu. Numerical simulation of reaction process in mip riser based on multi-scale model. CIESC Journal, 64(6):1983, 2013. URL: http://www.hgxb.com.cn/EN/abstract/article_13466.shtml , doi:10.3969/j.issn.0438-1157.2013.06.012. [LWL11] Bona Lu, Wei Wang, and Jinghai Li. Eulerian simulation of gas–solid flows with particles of geldart groups a, b and d using emms-based meso-scale model. Chemical Engineering Science, 66(20):4624 – 4635, 2011. URL: http://www.sciencedirect.com/science/article/pii/S0009250911004039, doi:http://dx.doi.org/10.1016/j.ces.2011.06.026. [LZW+13] Bona Lu, Nan Zhang, Wei Wang, Jinghai Li, John H. Chiu, and Shin G. Kang. 3-d full-loop simulation of an industrial-scale circulating fluidized-bed boiler. AIChE Journal, 59(4):1108–1117, 2013. URL: http://dx.doi.org/10.1002/aic.13917, doi:10.1002/aic.13917. [MM79] D. Mills and J. S. Mason. Evaluating the conveying capacity and service life of pipe bends in pneumatic conveying systems. Journal of Powder and Bulk Solids Technology , 3:13–20, 1979. [MSSH15] Jordan Musser, Madhava Syamlal, Mehrdad Shahnam, and David Huckaby. Constitutive equation for heat transfer caused by mass transfer. Chemical Engineering Science, 123():436 – 443, 2015. URL: http://www.sciencedirect.com/science/article/pii/S0009250914006903 , doi:http://dx.doi.org/10.1016/j.ces.2014.11.036. [NJR05] Subramanya V. Nayak, Saket L. Joshi, and Vivek V. Ranade. Modeling of vaporization and cracking of liquid oil injected in a gas–solid riser. Chemical Engineering Science, 60(22):6049 – 6066, 2005. 7th International Conference on Gas-Liquid and Gas-LiquidSolid Reactor Engineering7th International Conference on Gas-Liquid and Gas-Liquid-Solid Reactor Engineering. URL: http://www.sciencedirect.com/science/article/pii/S0009250905003416 , doi:http://dx.doi.org/10.1016/j.ces.2005.04.046. [NYWangGe+04] * Ning Yang, Wei Wang„ Wei Ge„ Linna Wang„ and Jinghai Li. Simulation of heterogeneous structure in a circulating fluidized-bed riser by combining the two-fluid model with the emms approach. Industrial & Engineering Chemistry Research , 43(18):5548–5561, 2004. URL: http://dx.doi.org/10.1021/ie049773c, arXiv:http://dx.doi.org/10.1021/ie049773c, doi:10.1021/ie049773c. [ORourkeS14] Peter J. O’Rourke and Dale M. Snider. A new blended acceleration model for the particle contact forces induced by an interstitial fluid in dense particle/fluid flows. Powder Technology, 256():39 – 51, 2014. URL: http://www.sciencedirect.com/science/article/pii/S003259101400103X , doi:http://dx.doi.org/10.1016/j.powtec.2014.01.084. [PPC93] M. K. PATEL, K. PERICLEOUS, and M. CROSS. Numerical modelling of cirJournal of Computational Fluid Dynamics, culating fluidized beds. I nternational 1(2):161–176, 1993. URL: http://www.tandfonline.com/doi/abs/10.1080/10618569308904470 , arXiv:http://www.tandfonline.com/doi/pdf/10.1080/10618569308904470, doi:10.1080/10618569308904470. [RDH71] J.F. Richardson, P. Davidson, and D. Harrison. Fluidization. Academic Press, New York, 1971. [SMA63] J. SMAGORINSKY. General circulation experiments with the primitive equations. Monthly Weather Review, 91(3):99–164, 1963. URL: http://dx.doi.org/10.1175/1520-0493(1963)091<0099:GCEWTP>2.3.CO;2, arXiv:http://dx.doi.org/10.1175/1520-0493(1963)091<0099:GCEWTP>2.3.CO;2, doi:10.1175/15200493(1963)091<0099:GCEWTP>2.3.CO;2. [Sni01] D.M. Snider. An incompressible three-dimensional multiphase particle-in-cell of Computational Physics, model for dense particle flows. Journal 170(2):523 – 549, 2001. URL: http://www.sciencedirect.com/science/article/pii/S0021999101967476 , doi:http://dx.doi.org/10.1006/jcph.2001.6747. [Til69] G.P. Tilly. Erosion caused by airborne particles. Wear , http://www.sciencedirect.com/science/article/pii/0043164869900350 , 1648(69)90035-0.

238

14(1):63 – 79, 1969. URL: doi:http://dx.doi.org/10.1016/0043-

Bibliography


[TL86] R. Turton and O. Levenspiel. A short note on the drag correlation for spheres. Powder Technology, 47(1):83 – 86, 1986. URL: http://www.sciencedirect.com/science/article/pii/0032591086800122 , doi:http://dx.doi.org/10.1016/0032-5910(86)80012-2. [WY66] C.Y. Wen and Y.H. Yu. Mechanics of fluidization. Chemical Engineering Progress Symposium, pages 100–111, 1966. [Whi91] F.M. White. Viscous fluid flow. McGraw-Hill, New York, 1991. [YK55] Sakae Yagi and Daizo Kunii. Studies on combustion of carbon particles in flames and fluidized beds. Symposium (International) on Combustion, 5(1):231 – 244, 1955. URL: http://www.sciencedirect.com/science/article/pii/S0082078455800331 , doi:http://dx.doi.org/10.1016/S00820784(55)80033-1. [Yan03] Wen-Ching Yang, editor. Handbook of Fluidization and Fluid-Particle Systems. Marcell Dekker, Inc., New York, 2003. [ZLWL08] Nan Zhang, Bona Lu, Wei Wang, and Jinghai Li. Virtual experimentation through 3d fullloop simulation of a circulating fluidized bed. Particuology, 6(6):529 – 539, 2008. Simulation and Modeling of Particulate Systems. URL: http://www.sciencedirect.com/science/article/pii/S1674200108001454, doi:http://dx.doi.org/10.1016/j.partic.2008.07.013. [ZORourkeS09] Paul (Pinghua) Zhao, Peter J. O’Rourke, and Dale Snider. Three-dimensional simulation of liquid injection, film formation and transport, in fluidized beds. Particuology, 7(5):337 – 346, 2009. URL: http://www.sciencedirect.com/science/article/pii/S1674200109001072 , doi:http://dx.doi.org/10.1016/j.partic.2009.07.002.

Bibliography

239


240

Bibliography

INDEX

A

G

ADmesh, 29 Advanced Options, 27

Global Settings, 40 Gravity, 41 Grid checking, 34 generating, 34 number of cells, 34 quality of, 18 utility buttons, 19 viewing in GMV, 18, 36 Grid generator, 18 Grid line editor, 30, 33 Grid Lines types, 18 Grid lines, 18, 31 adding, 31–33 editing, 33, 34 major, 18, 19 minor, 18, 19 modifying, 31 moving, 33 removing, 33 shortcut keys, 33 Gridding splitting cells, 32

B Baffles, 21 adding, 22 limitations, 22 removing, 24 Base Materials, 45 BC Connections, 134

C CAD, 17 Check grid, 34 Chemical name, 49 Chemistry at model start, 43 ramping at start, 43 Close Pack Volume Fraction, 56 Compare Grid to CAD, 36 Controls setup grid, 26 Coordinates, 19 ijk, 19 xyz, 19 Cut cells, 27

D Density, 54 material, 51 Description field chemical, 49 Drag Model, 74

F Filtering STL lines, 31 Flow Type compressible flow, 47 incompressible flow, 47

H Heat of formation, 51 Heat transfer, 43 fluid-to-particle, 44 fluid-to-wall, 43 Nusselt number, 45 Hide triangles, 31

I Import Grid, 19

L Line colors, 27 Low/high break temperature, 51

241

Barracuda User Manual

Recommend Documents