Airbag Particle Modeling (Some of the intricacies discussed)
Introduction
This is a discussion of the Airbag Particle Model . However, it is not as detailed in its current form as I would have liked it to be. Let me say in short a few things about it: a) The Mass Inflow Rate is modeled according to the “ Old Style” of Airbag_Hybrid. That means we have to define a separate Mass Flow Curve for the different Gas Components. In order to do that, we have to convert the “Mole Fractions” to “Kg. Fractions” (using the supplied Excel Spreadsheet) and use them as “Ordinate Scale Factors”. b) The Airbag Cards have “certain similarities” with the “ Airbag_Hybrid” model but the fields are still not identical. One has to read the Manual thoroughly to understand all of it. c) The “Fabric Porosity Venting ” works through the Fabric Material Card, exactly as in “Airbag_Hybrid”, with or without “Blockage”. The other factors also work the same way. d) The “Cut Hole Venting” cannot be described through Fabric Material definition as we did in “Airbag_Hybrid”. So far, that has not been implemented in the code. “Cut-Hole Venting” has to be defined in the “ Airbag_Particle” block of cards itself. e) “Blockage” of Cut-Hole Venting is defined in the “ Airbag_Particle” block but it does not work in Version R4.2.1 of the code. Therefore, we have to put that flag as “0” currently. It presumably works in Version R5 of the code but I have not tested it. f) We have noticed that in the parameters A, B and C that define the temperature dependent Cp, “negative values ” can cause numerical problems . Apparently, the equation Cp = A + B*T + C*T**2 should only be “ increasing” in value in this method. Negative values can cause negative slopes in certain cases and the Particle Method cannot handle this condition yet. Therefore, the suggestion is to use “0” in place of any “negative values” of these parameters (note that the value of “A” cannot be negative). This is not a limitation of *Airbag_Hybrid but only one of *Airbag_Particle. We also hope this to be resolved in the future. With this brief introduction, let us study the deck itself.
The Airbag Particle Deck Preparation
Please refer to the Deck “dyna.ParticleWith421.HoleAndFabricVenting.k ” (supplied), for this discussion. The *Airbag_Particle Cards look like the following:
2
*DEFINE_VECTOR $
vid
xt
yt
zt
xh
yh
zh
1
0.000
0.000
0.000
0.000
100.00
0.000
cid 0
The above Vector Card is used to define the “Jet Direction”. This shows that the Jet is in the “Y” direction (because “yt=0.000” and “yh=100.00”) *AIRBAG_PARTICLE $
sid1
stype1
sid2
stype2
block
npdata
fric
irdp
1
1
0
0
0
0
0.000
0
$ $ In the following, “nvent” defines the “Cut Vent Holes”. $ Here “nvent=1” and therefore the “Cut Vent Hole” line is active two lines below. $ For two Vent Hole parts, we would have “nvent=2” and two Cut Vent Hole lines below. $ To eliminate Cut Vent Holes, use “nvent=0” and comment out two lines below. $
np
unit
visflg
tatm
patm
nvent
tend
tsw
100000
0
1 295.00 1.0133E-4
1
0
60.0
$ 100000
0
1 295.00 1.0133E-4
0
0
60.0
$ In this deck, the Method switches to Airbag_Hybrid at 60- ms (because “tsw=60.0”) $ Number of Particles used is 100000 (because np=100000) $
iair
ngas
norif
nid1
nid2
nid3
1
5
1
3473
1393
0
chm
$ $ Part-8 (also Material-8) consists of the Vent Holes of the Airbag. A factor of 0.63 $ makes the Vent Holes equivalent to 2x25-dia (details given in Airbag_Hybrid) $
sid3
styp3
c23
8
0
0.63
lctc23
lcpc23
Comment
enh_v
ppop
out this line if “nvent=0” two lines above.
3
$ The following line exists because “ iair = 1 ” two lines above. $
pair
tair
xmair
aair
bair
cair
1.0132E-4 295.00 0.028970
29.000
0.000
0.000
$ $ Define "ngas" cards below. Note that negative values of “c” are made “0”. To maintain $ consistency, all “c” values have been turned to “0”. $ N2 $
lcidm
lcidt
xmi(mw)
a
b
c
infgi
$
1
7
0.028016
27.522
0.005530
-4.820E-7
0
1
7
0.028016
27.522
0.005530
0.000000
0
$
lcidm
lcidt
xmi(mw)
a
b
c
infgi
$
2
7
0.032000
25.883
0.012750
-3.692E-6
0
2
7
0.032000
25.883
0.012750
0.000000
0
$
lcidm
l cidt
xmi(mw)
a
b
c
infgi
$
3
7
0.044010
25.651
0.042720
-1.372E-5
0
3
7
0.044010
25.651
0.042720
0.000000
0
$
lcidm
l cidt
xmi(mw)
a
b
c
infgi
$
4
7
0.002016
28.799 -0.000360
1.758E-6
0
4
7
0.002016
28.799 -0.000360
0.000000
0
$ $ O2
$ $ CO2
$ $ H2
4
$ H2O $
lcidm
l cidt
xmi(mw)
a
b
c
infgi
$
5
7
0.018016
32.366
0.001620
0.8107E-5
0
5
7
0.018016
32.366
0.001620
0.0000000
0
$ $ The following card exists because "norif=1" in Card 3 above (Number of Orifices). $
nidi
ani
vdi
cai
infoi
3473
100.0
1
30.0
0
$ In the above, “vdi” is the Vector-ID Number specifying the direction of the Jet. $ Please refer to the Manual for the rest of the entries. Also refer to the “Fabric Card’ for the Airbag (Part-1, Mat-1) to understand Fabric Porosity Venting. That is exactly as in *Airbag_Hybrid.
NOTE on Number of Particles
The number of Particles to be used in an Airbag is debatable and comes mainly out of personal experience. For Driver Airbags the number starts at 100,000 while for Passenger Airbags it starts at 200,000. However, these are not fixed numbers. It is my experience that the ideal number of particles to be used in a given simulation lies somewhere between “2500 to 4000 particles per Liter of Airbag ”, with the latter number being the “upper limit”. Since most Driver Airbags are about 60-liters in volume, the range turns out to be 150000 to 240000 particles. Since modern Passenger Airbags are about 125-liters in volume, the range turns out to be 300000 to 500000 particles. The higher the number of particles, the greater is the simulation time. Therefore, a happy medium has to be sought here. I have run a Passenger Side simulation where I got excellent results by using 500000 particles for a 125-liter bag. However, I also realized that using more than that number would give me no added value.
5
We have supplied the following decks for added benefit of the users. The names signify what they stand for, indicating also that they have been run with the R4.2.1 Version: dyna.ParticleWith421.HoleVenting.k dyna.ParticleWith421.FabricVenting.k dyna.ParticleWith421.NoVenting.k
It is recommended that the user runs all of these and checks the “mass outflow” for each case with that of the others. It would also be a good thing if the user makes a few runs increasing the number of particles from the given 100000 to 150000, 200000 and 240000.
The following is a picture of my simulation, with the Particles clearly visible.
We end our discussion of the Particle Method for simulation of Airbags. I hope to further annotate this part in the future, giving clearer explanations covering all the data fields.
6
