Tutorial GT SUITE Introduction Training

__________________________________________ Tutorial General Training __________________________________________

Including the tutorials Tutorial 1: E-Mech_v1 (page 2) Tutorial 2: E-Mech_v2 (page 6) Tutorial 3: E-Mech_v3 (page 11) Tutorial 4: E-Flow_v1 (page 16) Tutorial 5: E-Flow_v2 (page 30) Tutorial 6: E-Thermal (page 42) Tutorial 6: E-Integrated (page 42)

Tutorial General Training This tutorial has been prepared to assist a new user of GT-ISE by giving step-by-step instructions for building simple mechanical and thermal models. Furthermore it gives a short overview about how to use GT-POST for postprocessing.

Tutorial 1: E-Mech_v1 Open the GTmap Document E-Mech_v1_Start and safe this file using the name EMech_v1. Defining objects First we need to modify some existing objects. Find Force in the model tree, double click on it and modify the attributes as shown below.

Click OK to save these changes and exit out of the Force object. Now we have to create a Mass object. Double click on Mass in the model tree and modify the attributes as shown below.

Click OK to save the changes and exit out of the Mass object. 2

Tutorial General Training Next we need to create a Spring object. Double click on Spring in the model tree and modify the attributes as shown below.

At last we have to create a Damper object. Double click on Damper in the model tree and modify the attributes as shown below.

Click OK to save the changes and exit out of the Damper object. Placing and linking parts To place the Force object on the project map click and hold on the Force object and drop it on the map. Repeat this with the left items using the following order. Also create a Ground object with default settings.

When placing templates on the map some icons will be initially placed in different orientations or the icons may differ from the ones which are shown. To change a part icon 3

Tutorial General Training on the map right click the part, select Choose Part Icon and select an image. This option is not available for all templates. To change the orientation of an icon, right click the part, select Rotate/Invert Icon(s) and select an action. It can be seen that when a part is dropped onto the map, it is renamed with a 1 at the end. Whenever a part is used more than once this number is appended for each case. The green symbols on the top-right corner of the parts are warning messages to warn the user that there is at least one link missing in order for the model to run. To link the parts right click on the empty part of the map and select Start New Link or click on Link in the toolbar. Now we start connecting the parts as shown below. In order to displace elements and connections for display purposes, go back on Select in the toolbar.

As a final step we have to create a MonitorSignal object. Double click on MonitorSignal in the model tree and modify the attributes as shown below.

Drag and drop MonitorSignal on the map below the Mass part. Now connect the MonitorSignal from the Mass part to MonitorSignal. When the Link Creation dialog box appears, chose the following ports. 4

Tutorial General Training

Thus a monitor will pop up during runtime and will show the mass displacement over time. Go to the Run menu and click on Run in order to run the simulation. The runtime monitor that we just have connected will now give us the following output. After the simulation has finished you can close the monitor and runtime report.

5


Tutorial 2: E-Mech_v2 If you have skipped Tutorial 1 then you can open the GTmap Document E-Mech_v2_Start instead. Save the file as E_Mech_v2. Defining objects In order to run three different cases we will need to parameterize a value inside Spring. Open the part and edit the attribute Stiffness from 1 to [Stiffness].

Now the Add Parameter dialog will appear. In this window, a description can be added and the parameter can be added to a specific folder in Case Setup. For this tutorial the default settings are fine.

Case setup Now enter Case Setup by clicking on Case Setup in the toolbar or by pressing the F4 key on the keyboard. Then append two cases. Afterwards enter the values of the parameters as shown below. 6


Another way to add the data with an increment of 0.5 N/m spring stiffness would be to type the command =[<1]+0.5 into the second cell. Each cell can reference to the left of it simply by typing in the command "=[
7

Tutorial General Training Now we want to add parameters for Case Label in order to create an easy overview of the results in GT-POST. To do so, add SpringStiff=[Stiffness] into the Case Label of Case 1. This will create a label that corresponds to the stiffness defined in the parameter for each case.

Go to the Run menu and click on Run in order to run the simulation. After closing the simulation report open GT-POST by clicking on View Results

.

Postprocessing In GT-POST we find in the project tree under CaseRLT → Mass → 1kg-1 → Positions → Momentane Position the current position of the mass for each case. Right click and select View to open the diagram. We can also double click on Mass on the map and chose Momentane Position to open the diagram.

8


In the next step we modify the x-axis in order to get some more representrative results. For this click on the X-Axis button on the top right in the CaseRLT view.

The RLT Selector window will appear. Here select Main → Stiffness.

9

Tutorial General Training Click OK to save the change. Now right click on CaseRLT → Mass → 1kg-1 → Positions → Momentane Position again and the window should look as shown below.

10


Tutorial 3: E-Mech_v3 If you have skipped Tutorial 2 then you can open the GTmap Document E-Mech_v3_Start instead. Save the file as E_Mech_v3. Defining objects In order to change the behavior of Force over time we will need to create a reference object. First, go to the Force part on the map and open it to edit its attributes. Type Force into the Force attribute cell.

The green text Force means that the force is now defined using a reference object (which doesn’t contain any data at this point). Reference objects are the key to GT-ISE. They allow GT-ISE templates to have less attributes because some important properties can be imbedded within reference objects. There are two ways to select a reference object for a particular attribute. The first way is to simply type in the name of the reference object. An important item to remember is that the reference object names are case sensitive, so the name must be typed exactly as it is displayed in the object tree. The second way is to right click in the particular attribute cell and select Value Selector (…) from the pop up menu. The Value Selector dialog contains detailed information about the types of data that can be entered into that specific attribute. In this dialog, there is a list of available reference objects that can be selected instead of typing the name directly in the cell. Also note that the Value Selector can be reached by left clicking on the gray box in the right side of the attribute field. This newly defined reference object still needs to be filled with data. Double click on the green Force. 11

Tutorial General Training A window will pop up. For this case choose ProfileTransient.

Fill the ProfileTransient with the data shown below.

Click OK and return to the project map. Now double click on all parts on the map and turn on all plots in the Plots folder.

12


Run setup and case setup For describing the type of simulation select Run Setup in the toolbar. Go to TimeControl and set the Automatic Shut-Off When Steady-State attribute to off in order to run the full simulation duration even if all steady state convergence criteria have been met. Then fill in the remaining attributes as shown below.

Now go to the the ODEControl folder. By default the explicit Runge-Kutta solver is used for modeling mechanics and coupled mechanics + hydraulics. GT-SUITE offers to select from 9 different ODE Integrators (5 implicit ones for stiff problems).

A few additional infos about the Explicit Runge-Kutta solver: 13

Tutorial General Training    

Fifth order accurate explicit solver Adaptive timestep methodology with automatic relaxation If relative error between fifth and sixth order evaluations exceeds user-defined tolerances, solver will sub step Decreases CPU time compared to fixed timestep solver and maintains high accuracy

Now go to Time Step and Solution Control Object. This attribute should be filled in with the reference object named Explicit-def. This default solution control object automatically uses a maximum integration time step of 0.001 seconds. To change this, click on the Value Selector as shown below.

Next double click on the ODEControlExplicit template to create a new object from this template. Name this template ODE-1E-2 and fill it with the values shown in the figure below.

14

Tutorial General Training Then click on Run in the toolbar in order to run the simulation. Afterwards open GTPOST using the View Results button . Postprocessing In GT-POST you can find the same map as in GT-ISE. Double click on the Mass object and select Position. Select all three cases. You will find the results of all three cases in the plot window as shown below.

15


Tutorial 4: E-Flow_v1 Open GT-ISE and chose New in the Home menu. Then the Creation Wizard will pop up. In the Creation Wizard chose Cooing Systems and Thermal Management. On the next window just select General Flow. After doing this we can see the new project tree for this case.

16

Tutorial General Training Defining objects Start with creating a new PipeRound object and call it Pipe1. Define values of this object as shown below.

Next we have to create the Fluid_Initial reference object. Define this reference object as shown below with parameters for pressure, temperature and composition.

17


Now we want to create a second pipe with same values as Pipe1 but another diamater. For this right click on Pipe1 in the project tree and select Copy and Edit Object. Change the diameter to 50 mm as shown below and call it Pipe2.

18


Thern create Pipe3 using Copy and Edit Object at Pipe2. Only change bend geometry for Pipe3 as shown below.

Next we want to create a Flowsplit. You can find this object under FlowSplitGeneral. Define this object as shown below. Again choose Fluid_Initial for Initial State Name. Fluid_Initial can be found in the GT-SUITE library. 19


20


Now we have created all flow components. Place them all on the map in the order as shown below.

For the boundary condition at the beginning of flow components we need to create EndFlowInlet. Define this object as shown below and name it FlowBoundary.

Also create EndEnvironment as shown below and name it PressureBoundary.

21


Next add our new components to the map as shown below.

Finally we want to create non-default orifice connections which are called OrificeConn in the project map. Create two objects of this type, call them Orifice50 and Orifice100 and define them as shown below.

22


Place these orifice connections on the map as shown below and make connections between all components.

23


Run setup and case setup Go to Case Setup and append multiple cases. Now we have a total number of cases of eleven. Next start with Volume_Flow of 50 l/min and increase it by 10 l/min per case. Set temperature of all cases to 90°C and also set pressure to 1.2 bar. Finally set Initial_Pressure equal to Pressure (=[Pressure]).

For the Fluid select egl5050 from the database

You can see all settings below.

24


Now go to Run Setup. First change the settings of TimeControl as shown below.

Then change FlowControl settings. Open Value Selector of Part Name List Objects Identifying Circuits Belonging to Column.

25


Next open Guided Circuit Defintion and choose Coolant.

26


At last go to the run menu and click on Run

in order to run the simulation.

Postprocessing Open GT-POST by clicking on View Results . Then show different map modes by clicking on RLT Contour Map on the toolbar. Show Average Volume Flow Rate (with Connection) and use Case Spinner to show results of different cases.

27


Also show results of Average Pressure. If results are the everywhere the same go to options and increase the number of displayed significant figures.

Finally we want to check the convergence of the implicit solver. Therefore go to Circuits, Coolants and set Implicit Convergence Last (all) Timesteps to one. At last click on yes. Now you can see the results.

28


If you are interested you can show plots for number of iterations and residuals for one case.

29


Tutorial 5: E-Flow_v2 Start from E-Flow_v1. The model will be modified to run a classical explicit flow solver case. Therefore a check valve will be build using an orifice and control and a the model will react to a pressure pulse Defining objects Set a parameter [dx] for the Discretization Length using Table Edit Model in order to make changes to all pipes fast. The parameter for [dx] is set to emphasize the trainee that usually runs with the explicit solver which requires shorter discretization length in order to resolve existing pressure waves.

Also turn on the pressure plot of all pipes. You can do this by double clicking on the pipes on the map or double clicking on PipeRound in the model tree and change the settings in the flow window.

30


Now break the link between Pipe2 and Pipe3.

Find OrificeConn in the model tree, drag it on the map and input data as shown below.

31


Also turn on the plot for Effective Area in the orifice.

Now link Pipe2 to Pipe3 again. Next go to the Control Library, find MathEquation, drag it on the map and call it SimpleValveControl. You can find the Control Library by clicking on Template Library in the toolbar. Input equation =if(pUp-pDown>[pCrack],30,0). When you become prompted to add the parameter set the unit to bar and change the description as shown below.

32


Then fill in Variable Description and Variable Name of MathEquation as shown below.

Now connect Pipe2 with MathEquation and select Valve Upstream Pressure. Do this analog for Pipe3 but use Valve Downstream Pressure. That means that we have to connect static pressure sensor from Pipe2 to the pUp equation variable and connect static pressure sensor from Pipe3 to the pDown equation variable. Afterwards go to Averaging and Filtering in the Control Library and find FirstOrderFilter. Drag this object on the map and set Time Constant to 1e-3.

33


Connect the output of the MathEquation to the FirstOrderFilter and actuate the Orifice Diameter of the Valve Orifice. That means that we have to create a connection from FirstOrderFilter to Valve1 and select Orifice Diameter.

Next delete the inlet EndFlowInlet FlowBoundary. Create and EndEnvironment named InletPressure.

For the Pressure (Absolute) Attribute create a profileAngle called pressurePulse. 34


Final model:

Run Setup and Case Setup Go to Run Setup, change Time Control Flag to periodic, Maximum Simulation Duration to 30. and Main Driver to Reference Object.

35


Then create a driver reference object called cycleFrequency and set the Rotational Speed to 50 Hz.

36

Tutorial General Training Next go to the Flow Control folder, put PartNameList to def and select the Explicit Default Object from the GT-SUITE library for Timestep and Solution Control Object.

At last go to Case Setup. Delete all Case3 - Case11. Include a Case Label for the cracking pressure of the valve (Cracking Pressure = [pCrack]bar) Change the Fluid parameter to air2. Set dx to 10mm and pCrack to 0.1bar (Case1) and 0.7bar (Case2)

Now you can run the model.

37

Tutorial General Training This is what is going to happen: A pressure pulse of 2 bar is going to be given to the flow system. In dependence of the set valve cracking pressure the valve will open earlier or later. Also, in difference the implicit flow solver the result quantities, for example pressure, now show oscillations on small time scales as the explicit solver is able to resolve them. They are a result of the excitation of the fluid by the inlet pressure ramp and the changing valve area. Postprocessing Click on View Results. In GT-POST click on New Report File in the toolbar. Rename the default group to Valve Opening Comparison Use the Plot Data Combine Macro to create a plot of the valve opening for both cases.

38


Crate a group called DIff (RightClick & Add Group) Use the Math Operation Macro to calculate the pressure difference of Pipe3-Pipe2 for Case1.

Rename the Difference [Y1-Y2] using the [CASELABEL] special parameter

Copy and paste the data set and use right click option Change Data Source to change the data source to Case2

39


Create the new group Pressures Use the Combine Data Macro to create a plot of the Pressure in all pipes for Case1

Copy and paste that plot. Use the right click option Change Data Source to change the data source to Case2 Use the right click option Children Properties to change the name of the plots to the special parameter [CASELABEL].

40


Change the properties of the Pressure group to create a 2x1 Layout and replot the group.

41

Tutorial 6: E-Thermal Open GT-ISE, go to Home, choose New, then go to Cooling Systems and Thermal Management and select all templates. On the picture that is shown below you can see the whole system which we are going to create in this tutorial.

Defining objects First we create a ThermalMass named Iron. Fill out the attributes as shown below. When you have finished place the object on the map.


Then we create a ThermalBlock named Aluminum. Again fill out the attributes as shown below and place the object above the ThermalMass Iron on map with enough space between for a ConvectionConnection. Do not forget to set the emissivity folder which is called Source Heat Rate to ign.

Next we have to create a ConductanceConn named ThermalGrease. Place this object between Iron and Aluminum. A default value for thermal conductance per unit area (contact) would mean ideal conduction with a very high thermal conductivity. In our case Aluminum and the Iron are connected with thermal grease. The thermal grease has a conductivity of 10 W/m-K and a layer thickness of 1 mm is assumed. This gives us a thermal conductance per unit area (contact) of 10000 W/(m^2-K). Connect Iron to ThermalGrease to Aluminum as shown below. Make sure to use Contact2Alu Port #2 of Iron and Port #2 of Aluminum.

43


Next create a non-default ConvectionConnection, name it Convection and place it above Aluminum on the map. Add a RLTDependenceXY for the Convective Heat Transfer Coefficient attribute. Consider that the average temperature of Iron must set to °C and should be pulled as RLT from the part Iron.

44


The Initial X Input should be set using the formula editor, so that unit will always be in °C

45


The XYTable must set to HTC-vs-Temp. Connect Aluminum Port #6 to ConvectionConn to Temperature-BC.

Now pull Temperature part from the Introduction-Training.gto and place it above Aluminum with enough space for a further object between on the map. Fill out the attributes as shown below.

46


At last drag and drop HeatRate from the gto into the model and place it below ThermalGrease. Finish defining objects by connecting HeatSource to Iron.

Run setup and case setup First go to Run Setup and fill out the folders of TimeControl, FlowControl and ThermalControl as shown below.

47


Then go to Output Setup. Set RLT Calculation Interval (Continuous Circuit) to 1 s and turn on TimeRLTs Storage Muliple = 20 and set the attributes as shown below.

48


Turn on the respective plots for each object. Afterwards go to Plot Setup and explain the attributes as shown below.

49

Tutorial General Training Finally go to Case Setup. Set T_Init to Ambient-Temperature. Append six cases. The first one is of transient nature and the last six cases are steady state with varying ambient temperature. Make sure the FLUIDSS is set correctly.

Postprocessing After running the simulation open GT-POST by clicking on View Results. First create a plot based on TimeRLTs including Heat Transfer Coefficient (HTC) of Convection, Temperature (Temperature) of Iron Port #2 Contact2Alu and Heat Rate (Imposed HeatRate) of HeatSource. Next add HTC plot instead of TimeRLTs to the existing plot to compare HTC plot and TimeRLTs. Then add data from the document MeasuredData.xlsx. Create a steady state plot with Iron (Contact2Alu) on the Y axis and Ambient Temperature on the X axis. Finally run TimeRLT Animation for Thermal Library with Temperature.

50

Tutorial 7: E-Integrated The E-Integrated example model demonstrated the multi-physics capabilities of GTSUTIE by modeling a capsel coffee machine. This model combines the physical flow, thermal, mechanical and electrical domain as well as the controls domain. The model that can be seen as a predesign model represents the fluid flow from water tank across pump, tankless heater and capsel to the coffee cup and interfaces with thermal and mechanical domain at the respective sections. In exchange the thermal and mechanical domain interface with the electrical subsystem. Electrical Subsystem

The electrical subsystem consists of a 48V voltage source that is providing power to the electrical heaters as well as a permanent magnet DC motor driving the pump. The electrical side of the tankless heater consists of two identical units with 1.5kW heating power each. This large power output is required in order to be able to quickly (about 15sec) heat up the heater to operating temperature of roughly 85degC from environmental temperature. To maintain the temperature of the heater once it is at operating temperature significantly less power is required. In practice this requires a voltage regulator. For the sake of simplicity the voltage regulator that would probably bbe a more complex switching device is simplified to an ideal resistor dropping the voltage upstream of the heating wire. For the initial heat up of the heater this resistor is short circuited by a switch. Note that the thermal structure of the heating wire has been externalized to mae it accessible to optimization. The electric motor is designed to give the required pump torque at 2A. Thermal Subsystem


The thermal model consists of  ThermalMassPipeRound template that is used as a simple representation of the thermal structure of the tankless heater  ThermalNodeInternal in the HeatingWire parts. Those are mainly used to conduct heat to ThermalNodes “HeatingWireThermal” (and therefore have very low mass and high surface area)  ThermalNodes “HeatingWireThermal” that are used as a thermal representation of the heaters electrical heating wires. Flow and Mechnical Subsystem The mechanical subsystem is representing the electric motor shaft that interfaces the emotor with the pump. The flow system gives a simple representation of the of pipes and flow volume. Water is flowing from a tank component to the pump. Parallel to the pump a leakage component is modeling leakage. Water is then pumped through the tankless heater and PressureLoss template used to model the resistance of the coffee (capsel). Model usage to demonstrate optimizer Besides showing a use-case for a multidomain model in GT-SUITE the model us used to give an optimization example. Therefore in the *-opt.gtm model the Advanced Direct Optimizer is set to minimize the Mass of the heater material with the motivation to lower material usage. This is done my varying the total length of the heater (which will implicitly vary the elec. heating wire length as well as the length of the flow pipe). Additionally three constrains are used to limit the maximum elec. power, the duration of max. power during a coffee making cycle and the minimum heater outlet flow temperature (in order to have hot coffee in the cup).

52

Tutorial General Training Optimizer Settings

53


The prepared *-opt.gtm will run approximately 18min. on one core. Therefore a resut GDX and GU file are supplied to show the optimization process and result.

54


55

Tutorial GT SUITE Introduction Training

Recommend Documents