Basic Training Course
1
Course Outline • • • • • • •
Chapter 1. Simulation basics Chapter 2. Flexsim basics Chapter 3. Flexsim terminology Chapter 4. Flexsim modeling steps Chapter 5. Flexsim object library Chapter 6. Commands and constructs Chapter 7. Concepts and models 2
Chapter 1 Simulation Basics
3
What is Simulation? Simulation is the imitation of a system, based on knowledge or assumptions about the behavior of the parts of that system, with the purpose of obtaining insight in the behavior of the whole system.
4
Problems solved with simulation •
• •
Service problems – the need to process customers and their requests at the highest level of satisfaction for the lowest possible cost. Manufacturing problems – the need to make the right product at the right time for the lowest possible cost. Logistic problems – the need to get the right product to the right place at the right time for the lowest possible cost. 5
Simulation uses • • • • • • • • • • • • • • • • •
improve equipment utilization reduce waiting time and queue sizes allocate resources efficiently eliminate stock-out problems minimize negative effects of breakdowns minimize negative effects of rejects and waste study alternative investment ideas determine part throughput times study cost reduction plans establish optimum batch sizes and part sequencing resolve material handling issues study effect of setup times and tool changeovers optimize prioritization and dispatching logic for goods and services train operators in overall system behavior and job related performance demonstrate new tool design and capabilities manage day-to-day operational decision making plc controls, real-time monitoring 6
Discrete Event Simulation • Used to model systems which change state at discrete points in time as a result of specific events • State variables: – machine status (idle, busy, down, etc.) – queue content – operator availability, etc.
• System events: – – – –
order arrival product movement machine process start machine breakdown, etc. 7
Chapter 2 Flexsim Basics
8
Flexsim Basics • A Flexsim model is basically a system of queues, processes, and transportation. • Processing is simply a forced delay. • Transportation is the movement of an entity from one resource to another. • Queues develop naturally in a model when upstream arrival rates exceed the downstream processing rates. 9
Flexsim Basics • Modeling objects typically include: – fixed resources (i.e. queue, machines) – shared resources (i.e. operators) – mobile resources (i.e. forklifts, elevators).
• Entities flow through the model from one fixed resource to another. We call these entities “flowitems” in Flexsim. • Flowitems hold information (i.e. labels and itemtypes) that can be set and queried in the model to define process times and make routing decisions. 10
Flexsim Basics • Sometimes an operator must be available before the processing time for a flowitem can start, and sometimes the processing time starts immediately upon entry into the fixed resource. • Sometimes the flowitems are carried to the next fixed resource via a mobile resource, and sometimes they are simply moved instantaneously.
11
Flexsim Basics • Most modeling information is defined on the fixed resources in a model. Such as: – how long to process the flowitem – where to send the flowitem next – call a shared resource (operator) for processing the flowitem – call a mobile resource to transport the flowitem to the next fixed resource 12
Chapter 3 Flexsim Terminology
13
Flexsim Terminology • model – a collection of object instances representing an actual system -with just enough detail to answer the questions you are concerned with. • object – a model building block with inherited behavior, and user-definable attributes, variables, and visual properties. • library – a list of class objects to be used in building a model. Instances are created in the model when an object is dragged from the library icon grid into one of the view windows (orthographic, perspective)
14
Flexsim Terminology • flowitem – the entities that flow through a model such as products, customers, paperwork, phone calls, etc. are called flowitems. • FixedResource class – objects which receive and send flowitems – Processor – Queue – Conveyor
• TaskExecuter class – shared resources and mobile resources that perform assigned tasks – Operator – Transport 15
Flexsim Terminology • task – a single instruction or action to be performed by a TaskExecuter object. • tasksequence – a group of tasks to be performed in a set order. A tasksequence is sent to either a Dispatcher or TaskExecuter object. A Dispatcher will queue a tasksequence as necessary and send it to a TaskExecuter as they become available. TaskExecuters will execute the tasks of the tasksequence. 16
Flexsim Terminology • itemtype – all flowitems have an attribute called itemtype. It’s a numeric identifier used to distinguish one “type” of flowitem from another in processing and routing decisions in a model. • labels – information may also be stored on flowitems (and other modeling objects) in the form of labels. You may define as many labels on a flowitem or object as you would like. Labels have a name and data. The data may be numeric, string, list, or table 17
Flexsim Terminology • ports – objects have output, input and center ports. Output-to-input port connections define possible flowitem routes. Center port connections are for referencing purposes, and are typically used to define which taskexecuter will be called for processing or transportation. Port connections are made and broken by holding a letter down on the keyboard while click-and-dragging from one object to another. 18
Flexsim Terminology • triggers – when key events occur on an object, a trigger is fired. The user may specify a variety of things they want to have happen when a trigger fires. There are triggers for when a flowitem enters or exits an object. There are triggers for when a process time finishes or a repair time is complete. Each type of object has a unique set of triggers. • message – information sent from one object to another. When the message is received, a trigger is fired. A message is sent with a command. Messages may be sent immediately (sendmessage), or after a specified period of time (senddelayedmessage). 19
Flexsim Terminology • variables – Objects have system variables such as process time, maximum capacity, speed, etc. that can be defined uniquely for each instance. The Parameters window for an object is typically used to edit variables. • attributes – Objects have physical attributes such as size, location, rotation, 3D shape, color, etc. that may be defined uniquely for each instance. The Properties window for an object is typically used to edit attributes. • toolbox – a collection of tools that are global in nature and therefore are not associated with any one object in the model. 20
Chapter 4 Flexsim Modeling Steps
21
Basic Steps of Modeling in Flexsim 1 - Create a layout using library objects 2 - Make port connections 3 - Edit look and behavior of objects 4 - Compile and run the simulation 5 - View the results
22
Model 1 • Purpose – Introduction to the 5 basic steps of model building using a single queue, single server system (post office)
• Description – people arrive at a post office every 60 seconds on average. The statistical probability distribution which best simulates the inter-arrival pattern is an exponential distribution with a location value of 0 and a scale value of 60. – the service time at the one and only service window in the post office is lognormal(31,3.1,0.5) seconds. – if the queue for the service window exceeds 20 people, new arrivals are routed out of the model as “unhappy customers”.
• Questions – – – –
What’s the maximum and average time people wait in line? How many unhappy customers? What’s the utilization of the service window? How many customers were serviced?
23
Model 1 Layout
24
Step 1. Create a layout • Drag objects from the Library (icon grid) and drop them into the Orthographic Model View window using your mouse. • Reposition, resize, or rotate objects as desired using mouse.
25
To scale or not to scale? • Flowitems are passed from one object to the next in zero time regardless of where objects are located. • Convey time depends on the conveyor’s length and speed. • Transport time between objects depends on the distance between the objects and speed, accel, decel. • Grid size can be modified in the Settings menu. • Length units are “user-definable” – consistency is the key! • The VisualTool object can be used to display a CAD drawing as the floor plan for the model. • After scaling the imported image/shape, the modeling objects can be placed on the layout and sized to match. 26
Mouse operations • View Window: – Translate: left click-drag – Tilt and Rotate: right click-drag – Zoom: left&right click-drag (or scroll wheel)
• Object: – – – – –
Translate in X and Y: left click-drag on object Translate in Z: right&left click-drag on object X,Y,Z rotation: right click-drag on object Resize X and Y: left click-drag on yellow dot Resize Z: right&left click-drag on yellow dot (or use scroll wheel when object is highlighted) 27
Highlighting and Selecting Objects • Only one object can be highlighted (yellow box) in a given view window. • Multiple objects can be selected (red box) across views using shift-clickdrag and/or ctrl-click-drag. • Group editing can be performed on selected objects using the “Edit Selected Objects” menu on the Model View windows.
28
Step 2. Make port connections • While pressing the “A” key, click-and-drag port connections between the objects. • These output to input port connections are needed between the objects in your model that send and receive flowitems.
29
The purposes of port connections • Output to input port connections are typically used for passing flowitems. • Center port to center port connections are typically used for calling operators and transporters. • Center port connections are a handy way to reference other objects when writing code (i.e. outobject(current,1) references the object connected to output port 1) • Output to input port connections provide automatic “handshaking” that is used in passing flowitems – Flowitems are either pushed or pulled when port connections are open at both ends. – Opening ports may trigger events on the connecting objects (useful in developing object behavior). 30
Keys used for port connections Outputs to Inputs
Center to Center
Disconnect
Q
W
Connect
A
S 31
Port connections • Objects can have an unlimited number of ports. • A port connection is needed for each possible routing choice in the model. • The actual port that a flowitem passes through depends on the selected rule for the object that is either doing the pushing or the pulling.
32
Viewing port connections • Display of port connections can be turned off by using the “Settings” option in the Model View. • The size of the ports can also be adjusted with the “Settings” option (although not recommended) 33
Full Port Display • Input ports on left • Output ports on right • Center ports on bottom • Ports are numbered top to bottom • “V” click to display input and outputs • “C” click to display center ports 34
Step 3. Edit the objects • Parameters Window – Double click (or right click then choose Parameters) – Used to edit object specific stuff
• Properties Window – Right click then choose Properties – Used to edit and view general info on all objects
• Model Tree View – All objects in the model listed in a hierarchal tree. – Contains the underlying data structure for an object – ALL info is contained in the tree (all other gui’s are simply filters on the tree) 35
Parameters Window • Each object class has its own Parameters Window • Parameters affect the time that flowitems stay in an object and how they are routed from object to object. • Apply will accept changes • OK will accept changes and close the window • Cancel will close the window without accepting changes • Properties button opens the Properties Window
36
Parameter tabs • Parameters are grouped into tab pages by category • Every fixed resource object will have a “Flow”, and a “Triggers” tab (although some objects will have more triggers than others)
37
Understanding the Pick List Combo Box
• The pick list • The code template (friendly editing)
• The code edit (may use Flexscript and C++)
38
Code Template • Brown text may be replaced with any valid Flexscript. • These templates can be customized through the Code Edit window (to be taught later). 39
Properties Window • Every object has the same set of properties tabs • Visual Tab – Look and size
• General Tab – Port connections – Display switches
• Labels – User defined attributes
• Statistics – Throughput, states, queuing, and stay times – Graphs, charts, readouts 40
Step 4. Compile and Run the simulation • Compile the model (create an executable runtime) • Reset, Run • Control run speed with the slide control (has no effect on the simulation results) • Experimenter is used to define run length, number of iterations, and number of scenarios 41
Step 5. View the results • Create a Standard Report or State Report using the “Stats” menu • Other reporting features to be discussed later: – VisualTool • Dynamic text
– Recorder • Time plots, histograms, pie charts, historical and statistical tabular data
– Triggers writing to GlobalTables – Object Properties > Statistics Tab
42
Standard Report • Display a tabular report (csv file in Excel) for all objects in model (or just those objects selected red). • Choose from a list of standard system variables to include in report. • Or type in your own variables to add to the report
43
Standard Report View
44
Flexsim State Report • Percentage of total simulation time an object was in each of its states. • CSV format (default viewer is Excel). • State profile is listed for each object currently selected in the model.
45
Model 2 • Purpose – learn how to duplicate objects – see affects of adding a second service window – introduction to the “Send to port” field
• Description – add another service window to the post office
• Questions – answer the questions asked in model 1 – if the windows differ by the type of service they can provide and 40% of the people need to use window 1 and 60% need to use window 2, how does the queue time increase? 46
Chapter 5 The Flexsim Object Library
47
Flexsim Family Tree
48
Flexsim Standard Objects • Flexsim has 25 objects in the entire library • Your ability to model will be determined by how well you understand these basic 25 objects and their 3 superclasses 49
FixedResource superclass • FixedResource objects are the backbone of a simulation model. They define the flow of the model. Since flow is defined by the connecting of output to input ports, FixedResources are always connected together using the “A” key. • FixedResources can: – – – – –
push or pull flowitems process flowitems (forced delay) queue up flowitems (accumulate up to maxcontent) call for an operator to be utilized call for a transport to carry flowitems to next resource 50
Dispatcher superclass • The Dispatcher receives tasksequences and either queues them up, OR dispatches the tasksequences immediately to downstream Dispatchers and/or TaskExecuters. • A tasksequence is a list of tasks that must be executed in a defined order one after the other. • Common tasks are actions such as travel, load, unload, utilize, sendmessage, delay and break. • One tasksequence may preempt another tasksequence part way through the current one’s list of tasks. The preemption can either temporarily interrupt the current tasksequence or permanently destroy the current tasksequence and all sequences in queue. • Tasksequences are created and dispatched automatically when the “Use Transport” or “Use Operator” boxes are checked on a FixedResource type object. • Tasksequences can also be created and dispatched using commands in triggers or elsewhere in your model. 51
TaskExecuter superclass • TaskExecuters are subclasses of the Dispatcher and therefore know how to receive, queue up, and dispatch tasksequences. • In addition, TaskExecuters have the ability to execute the tasksequences that they receive. • The Operator, Transporter, Elevator, Robot, Crane, and ASRSvehicle are all subclasses of the TaskExecuter and therefore have similar functionality. 52
Model 3 • Purpose Simple model which uses several objects in the library
• Description: A part enters a Separator every 20 seconds where it is split into 3 separate parts. The first part moves via an “s”-shaped conveyor to a Combiner where a total of 8 parts are packed on a pallet (supplied by another Source). The loaded pallet is then conveyed to a Sink. The second part is conveyed to a Queue where it is then carried by an Operator to a MultiProcessor object where 3 separate processing steps are performed. The first step can complete on it’s own in 3 seconds. The second step requires an operator (same operator that delivered the part) and takes 4 seconds, and the third process can complete on its own in 5 seconds. A Transporter then picks it up and takes it to a Rack where it is stored. The third part travels via FlowNodes to a Queue where 10 are accumulated before being released, one at a time, to a Processor with a 20 second cycle time.They are then placed in a Queue where a Robot transfers one part at a time over to a Sink. 53
Model 3 Layout
54
Model 3 Follow-up Work Once you have finished the model, do the following: • Create a Network path for the transport to follow when travelling between the MultiProcessor and the Rack. • Use a 3D status pie chart to show the state profile for the operator. • Use 3D text to display the content of the Rack. • View the model in the perspective view and experiment with different background colors. • Learn how to apply new 3D shapes to the objects in your model. • Move the objects associated with each of the three flows into a separate VisualTool to create an hierarchal model. 55
Connecting Dispatchers and TaskExecuters
56
NetworkNode • A NetworkNode is used to define the path a TaskExecutor will follow when executing its travel tasks. • A path is made up of a group of NetworkNodes connected together by using the “A” key. • At least one NetworkNode needs to be connected to each load and unload station (FixedResource) using the “A” key. • A Task Executor is assigned to a path by connecting it to any NetworkNode in the path using the “A” key.
57
Connecting NetworkNodes • Click-and-drag a connection while pressing the “A” key
• When connected you will see a green path between the nodes denoting travel is allowed in both directions. 58
Connecting Paths to the Model • Click-and-drag a connection while pressing the “A” key between the NetworkNode and the object where the load takes place. • A blue line will show the connection. The line does not connect to any port. 59
Connecting a TaskExecutor to a Path • Click-and-drag a connection between the TaskExecutor and a NetworkNode while pressing the “A” key. • A red line will indicate that the connection has been made. The line does not connect to any port.
60
Single Direction Paths • A path can be restricted to travel in a single direction by pressing the “Q” key and clicking-and-dragging in the direction you want to disallow travel. • A red line (right side rule) will show the restricted direction.
61
Non-Passing Paths • A non-passing path can be defined by using the “A” key and click-anddragging between two nodes that are already connected. • Non-passing path segments turn a yellow color.
No traveling
Travel but no passing
62
Path Spline Points • Every path has a spline point in the center of the path by default. • Spline points allow paths to be curved. • Add spline points by pressing “X” and clicking on an existing spline point. • Delete spline points by selecting them and pressing “Del” 63
Hide Splines, Paths and NetworkNodes • The display options for a path system can be toggled by holding the “X” key down and then clicking on any NetworkNode in the path system.
64
Chapter 6 Basic Modeling Functions and Logic Statements
65
Available Help – – – – –
Hints Commands Sample Models (Help menu) Existing picklist options Any C++ book (if you’re so inclined)
66
General rules • language is case sensitive (A does not equal a) • no specific format is required (free use of spaces, tabs and line returns is encouraged for “readable” code) • numbers are double precision unless otherwise specified. • text strings are entered between quotes. “mytext” • parenthesis follow a function call and commas separate the arguments of the function. moveobject(object1,object2); • a function or command will always end with a semi-colon • parenthesis can also be used freely to make associations in your math and logic statements. • curly braces (ie { }) are used to define a block of statements. • to comment out the rest of a line use // • don’t use spaces or special characters in name definitions (_ is ok). • names may include numbers, but may not begin with a number. 67
Math Operators • • • • • • • • •
x+y x-y x*y x/y sqrt(x) pow(x,y) round(x) frac(x) fmod(x,y)
x plus y x minus y x times y x divided by y square root of x x to the power of y (xy) x rounded to the nearest integer returns the decimal portion of x returns the remainder of x/y
68
Comparing • • • •
x>y x= y x <= y
• x == y • x != y • comparetext(string1,string2)
x greater than y x less than y x greater than or equal to y x less than or equal to y
x equal to y x not equal to y string1 matches string2
69
Relating • • • • •
&& || ! min(x,y) max(x,y)
logical AND logical OR logical NOT returns minimum of x and y returns maximum of x and y
70
Updating • • • • • • •
x=y x += y x -= y x *= y x /= y x ++ x --
set x to y set x to x plus y set x to x minus y set x to x times y set x to x divided by y add 1 to x subtract 1 from x
71
Variable Types • • • •
int double string fsnode*
integer double precision text string pointer to a Flexsim tree node
72
Declaring and Setting Variables • • • •
int index = 1; double weight = 175.8; string category = “groceries”; fsnode* forklift = centerobject(current,1);
73
What is a Node? • Basic data structure of Flexsim is a hierarchal tree – main (model and project related objects and data) – view (gui related objects and picklists)
• The node is the basic building block of a tree • Nodes hold all the information behind the scenes for objects, gui’s and data.
74
Node Structure • Nodes have a name • Nodes can have a data item – number – string – object
• If nodes have object data, use > to view a separate node list containing the object info (data members and member functions) • If nodes contain sub nodes, use + to expand 75
Node Symbols Standard Folder Object Object data Function (C++) Function (FlexScript)
76
Sample Model Tree
77
Basic Object Referencing the current object • current the involved flowitem • item rank(current,3) • rank(node,ranknum) inobject(current,1) • inobject(object,portnum) outobject(current,1) • outobject(object,portnum) • centerobject(object,portnum) centerobject(current,1) • model() model()
78
Basic Object Attributes • • • • • •
getname( object ) setname( object, “name” ) getitemtype( object ) setitemtype( object, num) setcolor( object, red, green, blue ) colorred( object ) – blue,green,random...
79
Object Spatial Attributes • xloc( object ) yloc( object ) • setloc( object, xnum, ynum, znum )
zloc( object )
• xsize( object ) ysize( object ) • setsize( object, xnum, ynum, znum )
zsize( object )
• xrot( object ) yrot( object ) • setrot( object, xdeg, ydeg, zdeg )
zrot( object )
80
Basic Object Statistics • content( object ) • getinput( object ) • getoutput( object ) • int inventory = content(current); • int produced = getoutput(current);
81
• syntax
Object Labels
– getlabelnum( object, “labelname” ) – setlabelnum( object, “labelname” , value )
– getlabelstr( object, “labelname” ) – setlabelstr( object, “labelname” , string )
– label( object, “labelname” )
• examples – – – – –
getlabelnum( item, “serialnumber” ) setlabelnum( item, “serialnumber”, 5 ) getlabelstr( current, “category” ) setlabelstr( current, “category”, “groceries” ) inc( label(item, “stepnum”), 1 ) only works with number data nodes 82
Global Tables • gettablenum( “tablename”, rownum, colnum ) • settablenum( “tablename”, rownum, colnum, value )
• gettablestr( “tablename”, rownum, colnum ) • settablestr( “tablename”, rownum, colnum, string )
83
Logical ‘if’ Statement if (test expression) { code block } else { code block }
if (content(item) == 2) { colorred(item); } else { colorblack(item); }
84
Logical ‘while’ Statement while (test expression) { code block }
while (content(current) > 0) { destroyobject(last(current)); }
85
Logical ‘for’ Statement for (start expression; test expression; count expression) { code block }
for (int index = 1; index <= content(current); index++) { colorblue(rank(current,index)); }
86
Logical ‘switch’ Statement switch ( switchvariable ) { case casenum: { code block break; } default: { code block break; } }
int type = getitemtype(item); switch (type) { case 1: { coloryellow(item); break; } case 5: { colorred(item); break; } default: { colorgreen(item); break; } }
87
Code Exercise 1 • Purpose – Gain experience in writing C++ code in the Code Edit windows. – Introduction to PROSESTART and PARAMSTART statements and how they define the Template windows. – Introduction to the bernoulli command.
• Description – See layout slide for model description. – Using the “Send To Port” field of the Conveyor, send 30% of the flowitems to port 1 and 70% to port 2.
88
Code Exercises 1,2,3 Layout
89
Code Exercise 2 • Purpose – Learn how to both set and read the itemtype of a flowitem using C++
• Description – – – –
See layout slide for model description. Set the itemtype of flowitems as they exit the Source. 60% of the flowitems are itemtype 7, the rest are itemtype 3. Using an “if” statement in the “Send To Port” field of the Conveyor, send flowitems with itemtype 7 to port 1, and those with itemtype 3 to port 2.
• Questions – Can you change the color of the Flowitems based on their itemtype using code? 90
Code Exercise 3 • Purpose – Learn how to both set and read the labels on a flowitem using C++
• Description – See layout slide for model description. – Create a numeric label named “routing” on the Textured Colored Box in the FlowItem Bin (see toolbar). – Set the “routing” label of flowitems as they exit the Source with a random number between 1 and 4 (use duniform() command). – Use a switch() statement in the “Send To Port” field of the Conveyor to send flowitems according to the following • Label values 1 and 3 go to port 2 • Label values 2 and 4 go to port 1
• Questions – Can you change the color of the flowitems based on their itemtype using code? 91
Code Exercise 4 • Purpose – Learn how to write your own “for” loop.
• Description – A Combiner packs 8 boxes on a pallet. – Use an infinite supply of pallets from one Source and an infinite supply of boxes from another Source. – The Combiner sends to a Conveyor which then sends to another Conveyor. – Using the OnExit trigger of the first Conveyor, write a for loop which changes the color of all the boxes on the pallet.
• Questions – Can you change the 3D shapes of the boxes on the pallet with the setobjectshapeindex() command?
92
Exercise 4 Layout
93
Chapter 7
Concepts and Models
94
Modeling Concepts • FlowItem Bin • Arrivals (interarrival time, schedule, sequence, initial stock) • Labels and itemtypes • GlobalTables • Distributions (empirical, continuous) 95
Modeling Concepts • • • • • • •
Push versus pull MTBF/MTTR Time Tables Experimenter and Optimizer Manual tasksequence creation 3D shapes and animation Importing CAD layouts 96
Modeling Concepts • • • • • • •
Data import and export Data capturing and reporting Fly through presentations Debugging tips Global object pointers Flow control (close and open ports) Messaging (delayed messages) 97
• Purpose
Model 4
– Learn how to use pull logic in defining routings based on a flowitem’s itemtype.
• Description – Electronic components need to be tested after manufacture. – There are two types of components that arrive in a queue at random with the mix being 40% of one and 60% the other. – The inter-arrival time of components is exponential(0,30,1) – There are two testers for component 1 and three testers for component 2. Components will go to the first available tester. – Service times are all between 120 and 150 seconds uniformly distributed.
• Questions – Are all testers equally utilized? – Can you model this with push or pull? – Try assigning colors to the itemtypes.
98
Model 4 Layout
99
• Purpose
Model 5
– Learn how to assign and use labels on flowitems for routing and processing purposes. Learn how to use an empirical distribution to assign itemtypes.
• Description – Parts arrive in a queue every 30 seconds exponentially distributed. There are four part types distributed as follows: 20%, 30%, 40%, 10%. – The four part types are each processed on one of four dedicated machines. The processing times at each machine is uniform(100,120) seconds for first time parts, and uniform(120,130) seconds for rework parts. – After the parts are machined, they are put in a queue and wait to be tested. The test time is a constant 10 seconds. Parts that pass, leave the model, failed parts are routed back to the first queue to be reworked. 10% of the parts fail. – Parts being reworked are given priority in the queue.
• Questions – Where is the bottleneck? – Can you change the color and shape of failed parts?
100 100
Model 5 Layout
101 101
Model 6 • Purpose – Learn how to us an Operator to both carry and process flowitems
• Description – Parts arrive at a queue every 20 seconds exponentially distributed and need to be carried by the operator to one of 3 machines. – The operator needs to setup the part which takes 10 seconds. – The part can then run for 20 seconds on its own. – When the part finishes, it will leave the system. – One operator will run all three machines.
• Questions – See if you can get the operator to stay for setup once he has carried the product to the machine. 102 102
Model 6 Layout
103
Model 7 • Purpose – Learn how to use the “Break To” Requirement field on TaskExecuters.
• Description – – – –
A Source continuously sends Flowitems randomly to 3 Queues. Flowitems are assigned an itemtype 1,2,3 uniformly distributed. All three Queues can hold all three itemtypes. A Transporter carries Flowitems to three Queues that each hold 1 itemtype. – These Queues then go to Processors and a Sink. – The Transporter can carry up to 5 items at a time. – The Transporter should only carry items that have the same itemtype at any time.
• Questions – Experiment with the other available “Break To” Requirement pickoptions. How does the Transporter behave? 104 104
Model 7 Layout
105
Model 8 • Purpose – More practice working with Operators, priorities and preempting
• Description – – – –
A single work cell has 5 machines. Each part must be processed at each of the 5 machines There are 3 operators who work in the cell as a team. The operators need to carry each part to the next machine as well as set up the machine for that part. – Setup times are 5 seconds and process times are 15. – New parts arrive in the system every 25 seconds.
• Questions – What’s the throughput rate per hour? – Experiment with priorities and preempting to try and improve the throughput rate. – Assign each operator to a different timetable so their breaks and lunches are staggered. 106 106
Model 8 Layout
107
• Purpose
Model 9
– Small example of using a Global Table to define product routings. – Learn how to use a label to keep track of what step its on.
• Description – – – – –
3 different products (uniform) enter a 5 station work cell. Product 1 needs to be processed at station 1,3, and 2. Product 2 needs to be processed at station 1, 2, and 4. Product 3 need to be processed at station 2,4,3, and 5. Use a global table to make sure the products are sent to the right station in the proper sequence. – Use a single Queue who’s output ports are connected to each of the 5 stations and to a Sink. Each of the five stations have their output port also connected to the input ports of the Queue.
• Questions – Can you change all flowItems in the model to a common color depending on step? 108 108
Model 9 Layout
109
Model 10 • Purpose – Learn how to manipulate the Components List table on a Combiner object to do custom packing on the fly – Learn how to import and use a Source arrival schedule
• Description – A small distributor has 10 product types that he ships to the same 5 customers. – Each customer has a different order profile. – The distributer has a large supply of each of the 10 products so all he has to do is wait for an order to come through the door. – Products are placed on a pallet and then conveyed out. – Use the next two tables to build the model.
• Questions – Can you think of other uses for the Combiner? – If time permits, try importing the tables from Excel. 110 110
Model 10 Layout
111
Model 10 Order Table Customer 1
Customer 2
Customer 3
Customer 4
Customer 5
prod 1
1
2
3
1
4
prod 2
2
0
1
2
1
prod 3
0
1
0
1
2
prod 4
1
2
1
3
2
prod 5
3
3
3
0
1
prod 6
2
4
0
2
0
prod 7
0
0
2
2
3
prod 8
2
3
3
3
0
prod 9
1
2
4
1
2
prod 10
1
1
3
0
0
112 112
Model 10 Daily Arrival Schedule ArrivalTime
ItemName
ItemType
Quantity
Customer 1
0
A
1
10
Customer 2
3600
B
2
10
Customer 3
7200
C
3
10
Customer 4
10800
D
4
10
Customer 5
14400
E
5
10 113 113
Model 11 •
Purpose – Gain experience in building an abstract model – Learn to write some C++ in a trigger field
•
Description – Items arrive at a warehouse each day to be stored. – Orders arrive at a warehouse each day to be pulled. – The average number of items and orders received each day is 10, but both are exponentially distributed. – There are only 5 types of items (and therefore orders) received. The type of received items and orders is uniformly distributed 1 through 5 on a daily basis. – Whenever an order arrives it must search for a matching item and remove it. – Whenever an item arrives it must search for a matching order and remove it. – 1 model time unit = 1 day.
•
Questions – How big must the warehouse be, so there is always room to store the items received each day? – Plot the number of unsatisfied orders each day for an entire year. – What’s the average number of days an order waits to be satisfied?
114 114
Model 11 Layout
115 115
TaskExecuter Control • requestoperators( taskexecuter, station, involvedobj, nrofops, priority, preempting ) • freeoperators( taskexecuter, involvedobj )
• createemptytasksequence( taskexecuter, priority, preempting) • inserttask(tasksequence, tasktype, involved1, involved2, parameter1, parameter2, parameter3 ) • dispatchtasksequence( tasksequence ) 116 116
Sample code for tasksequences fsnode *forklift = centerobject(current,1); fsnode *origin = centerobject(current,2); fsnode *destination = centerobject(current,3); fsnode *item = first(origin); fsnode *ts = createemptytasksequence(trans,0,0); inserttask(ts, TASKTYPE_TRAVEL, origin, NULL); inserttask(ts, TASKTYPE_FRLOAD, item, pickup); inserttask(ts, TASKTYPE_TRAVEL, destination, NULL); inserttask(ts, TASKTYPE_FRUNLOAD, item, destination); dispatchtasksequence(ts); 117 117
Task Types TASKTYPE_LOAD: flowitem, pickup TASKTYPE_FRLOAD: flowitem, pickup, outputport TASKTYPE_UNLOAD: flowitem, dropoff TASKTYPE_FRUNLOAD: flowitem, dropoff, inputport TASKTYPE_UTILIZE: involved, station, state TASKTYPE_STOPREQUESTFINISH: stoppedobject, NULL TASKTYPE_TRAVEL: destination, NULL TASKTYPE_TRAVELTOLOC: NULL, NULL, xloc,yloc,zloc, endspeed TASKTYPE_TRAVELRELATIVE: NULL, NULL, xloc,yloc,zloc, endspeed TASKTYPE_BREAK: NULL, NULL TASKTYPE_DELAY: NULL, NULL, delaytime, state TASKTYPE_SENDMESSAGE: receiver, NULL, param1,param2,param3 TASKTYPE_MOVEOBJECT: flowitem, container, port TASKTYPE_DESTROYOBJECT: object, NULL Brown parameters are required, Gray parameters are optional ** use profiletasksequence(ts) to print to console full description of ts 118 118
Model 12 • Purpose – Learn how to manually build your own tasksequences
• Description – The model consists of a Source, three Processors in series, and a Sink. – An operator must pick up a part from the source, carry it to the first processor, be utilized for the process time, carry the part to the next processor, etc. until finally dropping it off at the sink. – Use Global Object Pointers instead of center port connections.
• Questions – How many operators does it take to make 1000 pieces a day?
119 119
Model 12 Layout
120 120
Basic Object Control closeinput( object ), openinput( object ) stopinput( object ), resumeinput( object ) closeoutput( object ), openoutput( object ) stopoutput( object ), resumeoutput( object ) sendmessage( toobject, fromobject, parameter1, parameter2, parameter3 ) • senddelayedmessage( toobject, delaytime, fromobject, parameter1, parameter2, parameter3 ) • stopobject( object ), resumeobject( object ) • moveobject( object, containerobject ) 121 121 • • • • •
Model 13 • Purpose – Learn how to control product flow and send delayed messages.
• Description – Assume an infinite supply of A and B components. – The two components must be glued together with epoxy which takes lognormal(0,2.6,0.8) seconds. – The assembled parts must wait in a staging area for at least 15 seconds for the epoxy to cure before they can enter the paint booth. – It takes lognormal(0,2.5,0.5) seconds to paint each part separately. – After each part is painted and exits the paint booth, the booth must be purged for 3 seconds before another part can enter.
• Questions – What is the average and maximum time a part waits in queue? 122 122
Model 13 Layout
123
