ANSYS Tutorial

ANSYS is a general purpose finite element modeling package for numerically solving a wide variety of mechanical problems. These problems include: static/dynamic structural analysis (both linear and non-linear), heat transfer and fluid problems, as well as acoustic and electro-magnetic problems. In general, a finite element solution may be broken into the following three stages. This is a general guideline that can be used for setting up any finite element analysis.

Preprocessing: defining the problem; the major steps in preprocessing are given below:

Define keypoints/lines/areas/volumes Define element type and material/geometric properties Mesh lines/areas/volumes as required. The amount of detail required will

depend on the dimensionality of the analysis (i.e. 1D, 2D, axi-symmetric, 3D).

Solution: assigning loads, constraints and solving; here we specify the loads (point or pressure), contraints (translational and rotational) and finally solve the resulting set of equations. Postprocessing: further processing and viewing of the results; in this stage one may wish to see:

Lists of nodal displacements Element forces and moments Deflection plots Stress contour diagrams Starting up ANSYS Large File Sizes

ANSYS can create rather large files when running and saving; be sure that your local drive has space for it.

Getting the Program Started

Starting up ANSYS in Windows NT is simple:Start Menu Programs ANSYS 5.7 Run Interactive Now

ANSYS 7.0 Environment (Most of later version is similar to this version)

The ANSYS Environment for ANSYS 7.0 contains 2 windows: the Main Window and an Output Window. Note that this is somewhat different from the previous version of ANSYS which made use of 6 different windows.

Within the Main Window are 5 divisions:

Utility Menu: The Utility Menu contains functions that are available throughout the ANSYS session, such as file controls, selections, graphic controls and parameters.

Input Line: The Input Line shows program prompt messages and allows you to type in commands directly.

Toolbar: The Toolbar contains push buttons that execute commonly used ANSYS commands. More push buttons can be added if desired.

Main Menu: The Main Menu contains the primary ANSYS functions, organized by preprocessor, solution, general postprocessor, design optimizer. It is from this menu that the vast majority of modeling commands are issued. This is

where you will note the greatest change between previous versions of ANSYS and version 7.0. However, while the versions appear different, the menu structure has not changed.

Graphics Window: The Graphic Window is where graphics are shown and graphical picking can be made. It is here where you will graphically view the model in its various stages of construction and the ensuing results from the analysis.

Output Window: The Output Window (DOS type window) shows text output from the program, such as listing of data etc. It is usually positioned behind the main window and can be put to the front if necessary.

ANSYS Interface Graphical Interface vs. Command File Coding

There are two methods to use ANSYS. The first is by means of the graphical user interface or GUI. This method follows the conventions of popular Windows and X-Windows based programs.

The second is by means of command files. The command file approach has a steeper learning curve for many, but it has the advantage that an entire analysis can be described in a small text file, typically in less than 50 lines of commands. This approach enables easy model modifications and minimal file space requirements.

The tutorials in this website are designed to teach both the GUI and the command file approach, however, many of you will find the command file simple and more efficient to use once you have invested a small amount of time into learning the code.

For information and details on the full ANSYS command language, consult: Help > Table of Contents > Commands Manual.

FEM Convergence Testing

Introduction

A fundamental premise of using the finite element procedure is that the body is sub-divided up into small discrete regions known as finite elements. These elements defined by nodes and interpolation functions. Governing equations are written for each element and these elements are assembled into a global matrix. Loads and constraints are applied and the solution is then determined.

The Problem

The question that always arises is: How small do I need to make the elements before I can trust the solution?

What to do about it...

In general there are no real firm answers on this. It will be necessary to conduct convergence tests! By this we mean that you begin with a mesh discretization and then observe and record the solution. Now repeat the problem with a finer mesh (i.e. more elements) and then compare the results with the previous test. If the results are nearly similar, then the first mesh is probably good enough for that particular geometry, loading and constraints. If the results differ by a large amount however, it will be necessary to try a finer mesh yet.

The Consequences

Finer meshes come with a cost however: more calculation time and large memory requirements (both disk and RAM)! It is desired to find the minimum number of elements that give you a converged solution.

Beam Models

For beam models, we actually only need to define a single element per line unless we are applying a distributed load on a given frame member. When point loads are used, specifying more than one element per line will not change the solution, it will only slow the calculations down. For simple models it is of no concern, but for a larger model, it is desired to minimize the number of elements, and thus calculation time and still obtain the desired accuracy.

General Models

In general however, it is necessary to conduct convergence tests on your finite element model to confirm that a fine enough element discretization has been used. In a solid mechanics problem, this would be done by creating several models with different mesh sizes and comparing the resulting deflections and stresses, for example. In general, the stresses will converge more slowly than the displacement, so it is not sufficient to examine the displacement convergence.

ANSYS: Saving and Restoring Jobs

Saving Your Job

It is good practice to save your model at various points during its creation. Very often you will get to a point in the modeling where things have gone well and you like to save it at the point. In that way, if you make some mistakes later on, you will at least be able to come back to this point.

To save your model, select Utility Menu Bar -> File -> Save As Jobname.db. Your model will be saved in a file called jobname.db, where jobname is the name that you specified in the Launcher when you first started ANSYS.

It is a good idea to save your job at different times throughout the building and analysis of the model to backup your work in case of a system crash or other unforeseen problems.

Recalling or Resuming a Previously Saved Job

Frequently you want to start up ANSYS and recall and continue a previous job. There are two methods to do this:

Using the Launcher... In the ANSYS Launcher, select Interactive... and specify the previously defined jobname. Then when you get ANSYS started, select Utility Menu -> File -> Resume Jobname.db . This will restore as much of your database (geometry, loads, solution, etc) that you previously saved. Or, start ANSYS and select Utitily Menu -> File -> Resume from... and select your job from the list that appears.

ANSYS Files

Introduction

A large number of files are created when you run ANSYS. If you started ANSYS without specifying a jobname, the name of all the files created will be FILE.* where the * represents various extensions described below. If you specified a jobname, say Frame, then the created files will all have the file prefix, Frame again with various extensions:

frame.db Database file (binary). This file stores the geometry, boundary conditions and any solutions.

frame.dbb Backup of the database file (binary).

frame.err Error file (text). Listing of all error and warning messages.

frame.out Output of all ANSYS operations (text). This is what normally scrolls in the output window during an ANSYS session.

frame.log Logfile or listing of ANSYS commands (text). Listing of all equivalent ANSYS command line commands used during the current session.

etc... Depending on the operations carried out, other files may have been written. These files may contain results, etc.

What to save?

When you want to clean up your directory, or move things from the /scratch directory, what files do you need to save?

If you will always be using the GUI, then you only require the .db file. This file stores the geometry, boundary conditions and any solutions. Once the ANSYS has started, and the jobname has been specified, you need only activate the resume command to proceed from where you last left off (see Saving and Restoring Jobs). If you plan on using ANSYS command files, then you need only store your command file and/or the log file. This file contains a complete listing of the ANSYS commands used to get you model to its current point. That file may be rerun as is, or edited and rerun as desired (Command File Creation and Execution).

If you plan to use the command mode of operation, starting with an existing log file, rename it first so that it does not get over-written or added to, from another ANSYS run.

Printing and Plotting ANSYS Results to a File

Printing Text Results to a File

ANSYS produces lists and tables of many types of results that are normally displayed on the screen. However, it is often desired to save the results to a file to be later analyzed or included in a report.

Stresses: instead of using 'Plot Results' to plot the stresses, choose 'List Results'. Select 'Elem Table Data', and choose what you want to list from the menu. You can pick multiple items. When the list appears on the screen in its own window, Select 'File'/'Save As...' and give a file name to store the results. Any other solutions can be done in the same way. For example select 'Nodal Solution' from the 'List Results' menu, to get displacements. Preprocessing and Solution data can be listed and saved from the 'List' menu in the 'Utility Menu bar'. Save the resulting list in the same way described above. Plotting of Figures

There are two major routes to get hardcopies from ANSYS. The first is a quick a raster-based screen dump, while the second is a scalable vector plot.

1.0 Quick Image Save

When you want to quickly save an image of the entire screen or the current 'Graphics window', select: 'Utility menu bar'/'PlotCtrls'/'Hard Copy ...'. In the window that appears, you will normally want to select 'Graphics window', 'Monochrome', 'Reverse Video', 'Landscape' and 'Save to:'.

Then enter the file name of your choice. Press 'OK'

This raster image file may now be printed on a PostScript printer or included in a document.

2.0 Better Quality Plots

The second method of saving a plot is much more flexible, but takes a lot more work to set up as you'll see...

Redirection

Normally all ANSYS plots are directed to the plot window on the screen. To save some plots to a file, to be later printed or included in a document or what have you, you must first 'redirect' the plots to a file by issuing: 'Utility menu bar'/'PlotCtrls'/'Redirect Plots'/'To File...'.

Type in a filename (e.g.: frame.pic) in the 'Selection' Window.

Now issue whatever plot commands you want within ANSYS, remembering that the plots will not be displayed to the screen, but rather they will be written to the selected file. You can put as many plots as you want into the plot file. When you are finished plotting what you want to the file, redirect plots back to the screen using: 'Utility menu bar'/'PlotCtrls'/'Redirect Plots'/'To Screen'.

Display and Conversion

The plot file that has been saved is stored in a proprietary file format that must be converted into a more common graphic file format like PostScript, or HPGL for example. This is performed by running a separate program called display. To do this, you have a couple of options: select display from the ANSYS launcher menu (if you started ANSYS that way) shut down ANSYS or open up a new terminal window and then type display at the Unix prompt.

Either way, a large graphics window will appear. Decrease the size of this window, because it most likely covers the window in which you will enter the display plotting commands. Load your plot file with the following command: file,frame,pic

if your plot file is 'plots.pic'. Note that although the file is 'plots.pic' (with a period), Display wants 'plots,pic'(with a comma). You can display your plots to the graphics window by issuing the command like plot,n

where n is plot number. If you plotted 5 images to this file in ANSYS, then n could be any number from 1 to 5.

Now that the plots have been read in, they may be saved to printer files of various formats:Colour PostScript: To save the images to a colour postscript file, enter the following commands in display: pscr,color,2 /show,pscr plot,n where n is the plot number, as above. You can plot as many images as you want to postscript files in this manner. For subsequent plots, you only require the plot,n command as the other options have now been set. Each image is plotted to a postscript file such as pscrxx.grph, where xx is a number, starting at 00.

Note: when you import a postscript file into a word processor, the postscript image will appear as blank box. The printer information is still present, but it can only be viewed when it's printed out to a postscript printer.

Printing it out: Now that you've got your color postscript file, what are you going to do with it? Take a look here for instructions on colour postscript printing at a couple of sites on campus where you can have your beautiful stress plot plotted to paper, overheads or even posters!

Black & White PostScript: The above mentioned colour postscript files can get very large in size and may not even print out on the postscript printer in the lab because it takes so long to transfer the files to the printer and process them. A way around this is to print them out in a black and white postscript format instead of colour; besides the colour specifications don't do any good for the black and white lab printer anyways. To do this, you set the postscript color option to '3', i.e. and then issue the other commands as before pscr,color,3 /show,pscr plot,n Note: when you import a postscript file into a word processor, the postscript image will appear as blank box. The printer information is still present, but it can only be viewed when it's printed out to a postscript printer.

HPGL: The third commonly used printer format is HPGL, which stands for Hewlett Packard Graphics Language. This is a compact vector format that has the advantage that when you import a file of this type into a word processor, you can actually see the image in the word processor! To use the HPGL format, issue the following commands: /show,hpgl plot,nFinal Steps It is wise to rename these plot files as soon as you leave display, for display will overwrite the files the next time it is run. You may want to rename the postscript files with an '.eps' extension to indicate that they are encapsulated postscript images. In a similar way, the HPGL printer files could be given an

'.hpgl' extension. This renaming is done at the Unix commmand line (the 'mv' command).

A list of all available display commands and their options may be obtained by typing: help When complete, exit display by entering finish

Finite Element Method using Pro/ENGINEER and ANSYS

The transfer of a model from Pro/ENGINEER to ANSYS will be demonstrated here for a simple solid model. Model idealizations such as shells and beams will not be treated. Also, many modeling options for constraints, loads, mesh control, analysis types will not be covered. These are fairly easy to figure out once you know the general procedures presented here.

Step 1. Make the part

Use Pro/E to make the part. Things to note are:be aware of your model units note the orientation of the model (default coordinate system in ANSYS will be the same as in Pro/E)

IMPORTANT: remove all unnecessary and/or cosmetic features like rounds, chamfers, holes, etc., by suppressing them in Pro/E. Too much small geometry will cause the mesh generator to create a very fine mesh with many elements which will greatly increase your solver time. Of course, if the feature is critical to your design, you will want to leave it. You must compromise between accuracy and available CPU resources.

The figure above shows the original model for this demonstration. This is a model of a short cantilevered bracket that bolts to the wall via the thick plate on the left end. Model units are inches. A load is applied at the hole in the right end. Some cosmetic features are located on the top surface and the two sides. Several edges are rounded. For this model, the interest is in the stress distribution around the vertical slot. So, the plate and the loading hole are removed, as are the cosmetic features and rounds resulting in the "de-

featured" geometry shown to the right. The model will be constrained on the left face and a uniform load will be applied to the right face.

Step 2. Create the FEM model

In the pull-down menu at the top of the Pro/E window, select

Applications > Mechanica

An information window opens up to remind you about the units you are using. Press Continue

In the MECHANICA menu at the right, check the box beside FEM Mode and select the command Structure.

A new toolbar appears on the right of the screen that contains icons for creating all the common modeling entities (constraints, loads, idealizations). All these commands are also available using the command windows that will open on the right side of the screen or in dialog windows that will open when appropriate.

Notice that a small green coordinate system WCS has appeared. This is how you will specify the directions of constraints and forces. Other coordinate systems (eg cylindrical) can be created as required and used for the same purpose.

The MEC STRUCT menu appears on the right. Basically, to define the model we proceed down this menu in a top-down manner. Model is already selected for you which opens the STRC MODEL menu. This is where we specify modeling information. We proceed in a top-down manner. The Features command allows you to create additional simulation features like datum points, curves, surface regions, and so on. Idealizations lets you create special modeling entities like shells and beams. The Current CSYS command lets you create or select an alternate coordinate system for specifying directions of constraints and loads.Defining Constraints

For our simple model, all we need are constraints, loads, and a specified material. Select

Constraints > New

We can specify constraints on four entity types (basically points, edges, and surfaces). Constraints are organized into constraint sets. Each constraint set has a unique name (default of the first one is ConstraintSet1) and can contain any number of individual constraints of different types. Each individual constraint also has a unique name (default of the first one is Constraint1). In the final computed model, only one set can be included, but this can contain numerous individual constraints.

Select Surface. We are going to fully constrain the left face of the cantilever. A dialog window opens as shown above. Here you can give a name to the constraint and identify which constraint set it belongs to. Since we elected to create a surface constraint, we now select the surface we want constrained (push the Surface selection button in the window and then click on the desired surface of the model). The constraints to be applied are selected using the buttons at the bottom of the window. In general we specify constraints on translation and rotation for any mesh node that will appear on the selected entity. For each direction X, Y, and Z, we can select one of the four buttons (Free, Fixed, Prescribed, and Function of Coordinates). For our solid model, the rotation constraints are irrelevant (since nodes of solid elements do not have this degree of freedom anyway). For beams and shells, rotational constraints are active if specified.

For our model, leave all the translation constraints as FIXED, and select the OK button. You should now see some orange symbols on the left face of the model, along with some text labels that summarize the constraint settings.

Defining Loads

In the STRC MODEL menu select: Loads > New > Surface

The FORCE/MOMENT window opens as shown above. Loads are also organized into named load sets. A load set can contain any number of individual loads of different types. A FEM model can contain any number of different load sets. For example, in the analysis of a pressurized tank on a support system with a number of nozzle connections to other pipes, one load set might contain only the internal pressure, another might contain the support forces, another a temperature load, and more might contain the forces applied at each nozzle location. These can be solved at the same time, and the principle of superposition used to combine them in numerous ways.

Create a load called "end_load" in the default load set (LoadSet1)

Click on the Surfaces button, then select the right face of the model and middle click to return to this dialog. Leave the defaults for the load distribution. Enter the force components at the bottom. Note these are relative to the WCS. Then

select OK. The load should be displayed symbolically as shown in the figure below.

Note that constraint and load sets appear in the model tree. You can select and edit these in the usual way using the right mouse button.

Assigning Materials

Our last job to define the model is to specify the part material. In the STRC MODEL menu, select

Materials > Whole Part

In the library dialog window, select a material and move it to the right pane using the triple arrow button in the center of the window. In an assembly, you could now assign this material to individual parts. If you select the Edit button, you will see the properties of the chosen material.

At this point, our model has the necessary information for solution (constraints, loads, material).

Step 3. Define the analysis

Select

Analyses > New

Specify a name for the analysis, like "ansystest". Select the type (Structural or Modal). Enter a short description. Now select the Add buttons beside the Constraints and Loads panes to add ConstraintSet1 and LoadSet1 to the analysis. Now select OK.

Step 4. Creating the mesh

We are going to use defaults for all operations here. The MEC STRUCT window, select

Mesh > Create > Solid > Start

Accept the default for the global minimum. The mesh is created and another dialog window opens (Element Quality Checks).

This indicates some aspects of mesh quality that may be specified and then, by selecting the Check button at the bottom, evaluated for the model. The results are indicated in columns on the right. If the mesh does not pass these quality checks, you may want to go back to specify mesh controls (discussed below). Select Close. Here is an image of the default mesh, shown in wire frame.

Improving the Mesh

In the mesh command, you can select the Controls option. This will allow you to select points, edges, and surfaces where you want to specify mesh geometry such as hard points, maximum mesh size, and so on. Beware that excessively tight mesh controls can result in meshes with many elements.

For example, setting a maximum mesh size along the curved ends of the slot results in the following mesh. Notice the better representation of the curved

edges than in the previous figure. This is at the expense of more than double the number of elements. Note that mesh controls are also added to the model tree.

Step 5. Creating the Output file

All necessary aspects of the model are now created (constraints, loads, materials, mesh). In the MEC STRUCT menu, select

Run

This opens the Run FEM Analysis dialog window shown here. In the Solver pull-down list at the top, select ANSYS. In the Analysis list, select Structural. You pick either Linear or Parabolic elements. The analysis we defined (containing constraints, loads, mesh, and material) is listed. Select the Output to File radio button at the bottom and specify the output file name (default is the analysis name with extension .ans). Select OK and read the message window.

We are now finished with Pro/E. Go to the top pull-down menus and select

Applications > Standard

Save the model file and leave the program.

Copy the .ans file from your Pro/E working directory to the directory you will use for running ANSYS.

Step 6. Importing into ANSYS

Launch ANSYS Interactive and select

File > Read Input From...

Select the .ans file you created previously. This will read in the entire model. You can display the model using (in the pull down menus) Plot > Elements.

Step 7. Running the ANSYS solver

In the ANSYS Main Menu on the left, select

Solution > Solve > Current LS > OK

After a few seconds, you will be informed that the solution is complete.

Step 8. Viewing the results

There are myriad possibilities for viewing FEM results. A common one is the following:

General Postproc > Plot Results > Contour Plot > Nodal Solu

Pick the Von Mises stress values, and select Apply. You should now have a color fringe plot of the Von Mises stress displayed on the model.

Notes by R.W. Toogood