Q) I have difficulty importing my Pro/e models into Design Modeler, even though I only need to use this to get my model into CFX-Mesh
A) If you plan to use CFX-Mesh, then try importing to CFX-Mesh directly. The reason we run into problems with importing CAD models in DM is that the import into DM is a more rigorous process than import into CFX Mesh. The reason is that DM assumes that the user will be modifying the geometry in DM and hence tries to re-build the whole model in parasolid format - DM runs on a parasolid kernel. Depending on the original source and the complexity of features, this process may fail. Sometimes suppressing a specific feature in the CAD model will make it work. However when importing into CFX-Mesh directly, it assumes the user only wants to analyze existing CAD and hence it does not try to re-build the model in parasolid. So this works in almost all cases , regardless of the CAD source. |
Q) I have difficulty importing my Pro/e models into Design Modeler, even though I only need to use this to get my model into CFX-Mesh<br><br>A) If you plan to use CFX-Mesh, then try... |
Question:
Our customer wants to do a Loadcase combination between two different analyses having different mesh (due to a slight geometry change), supports and loads. When he tries combining the two loadcases concerned, he gets a warning that the loadcases were solved using different mesh, BCs . . . He wants to know
1. How the combination is done in ANSYS for such type of situations, to better understand the combined results and to interpret the results at the locations where the corresponding nodes in two models are not coincident.
2. The best method to combine such type of loadcases.
Personally, I suppose that the best approach would be using the submodelling technique. I would like to have details on this method for the loadcase combination application, if applicable.
Answer:
Let me begin by reviewing for you how a finite element solution is actually done. In the beginning matrices with all of the degrees of freedom for all of the elements are created and passed to the solver for global assembly. Boundary conditions are then applied. Force type boundary conditions are added to the right hand side of the system of equations. Constraints with zero magnitude displacements reduce the size of the problem by eliminating that degree of freedom row and column from the global matrix. For constraints with non zero displacements, that displacement is multiplied by the appropriate degree of freedom column of the global matrix and the result is added to the right hand side; then that row and column are eliminated from the global matrix. In the end, the problem to be solved consists of only the unknown degrees of freedom.
If the user were then to change the constraints, not forces, the problem to be solved would be different. That is, there would be different unknown degrees of freedom and the length of the solution vector could be different. Or if the mesh were changed, the two problems could have a different element order. The program write |
Question:<br><br>Our customer wants to do a Loadcase combination between two different analyses having different mesh (due to a slight geometry change), supports and loads. When he... |
Q. A model with PLANE55 and SURF151 elements has been made. On the SURF151, (keyopt8=1), a heat flux is applied. When an SFELIST command is issued after the application of the heat flux, a convection load with a film coefficient value equal to the applied heat flux has been created as well as the heat flux itself. This does not seem right. How should the user understand the meaning of this convection load?
A. This is a code defect (report #33543) which has been corrected for the ANSYS 10.0 release. |
Q. A model with PLANE55 and SURF151 elements has been made. On the SURF151, (keyopt8=1), a heat flux is applied. When an SFELIST command is issued after the application of the heat flux, a... |
Are there any examples showing how ANSYS handles temperature-dependent
Young's moduli and coefficients of thermal expansion? |
Yes, please see the ANSYS 9.0 example below. Please note the following:<br><br>Total strain is the sum of the various strain components. In this simplified<br>example, the... |
I tried to curve-fit viscoelastic material data, but the fit I get is very different than the experimental data I supply. I tried curvefitting with several terms, but without luck. I only read-in data for shear. I'm probably doing something wrong when curve-fitting, but I don't know what. Could you please have a look and try to create a good curve-fit with the data I attach?
The data
0.0001 100e6
.05 43e6
0.1 22e6
.15 15e6
0.2 11e6
.25 10e6 |
This an example of that old adage in computer science - garbage in = garbage out. In this case, the guess that you are providing is likely sending the algorithm off in the wrong... |
I only have an 8.1 DesignSpace license. I inadvertently set the Contact Options
to allow edge to face contact, which is not supported with this license. I set up
the contact, saved the model, and exited DesignSpace. Now, when I try to open
the .DSDB database, it comes up in read-only mode, since that feature is not
supported with my license. Is this a known error? |
Yes. Defect 29780 was filed and corrected in the UP20050422 build of ANSYS<br> Workbench. Now, when you try to use unsupported features (that were set under<br> Tools... |
If a user installs using "INSTALL -platform" and installs the Linux version to a Sun Solaris 9 application server, the install appears to go flawlessly. However, when running a modal analysis using the Block Lanczos mode extraction method, the analysis immediately fails with the fatal error:
MKL FATAL ERROR: libmkl_def.so: cannot open shared object file: No such file or directory |
To resolve this issue, try the following procedure:<br><br>1) cd /ansys_9.0/v90/ansys/syslib/linop64<br>2) rm libmath.so<br>3) ln -sf libacml.so... |
When I export Workbench Simulation 9.0 Heat Flux results to Microsoft Excel, the values do not match the Heat Flux plots. What is the reason for this? |
This is an error (#34844) that has been corrected by our Development group for the ANSYS 10.0 release.
|
My ANSYS LS-DYNA 9.0 full restart analyses seems to ignore the EDRD,D2R command.
Is this a known error in deformable-to-rigid switching? Is there a workaround? |
Defect 34412 was filed and corrected in the UP20050520 build of ANSYS LS-DYNA 10.0.<br> Please see the example below that demonstrates the error in releases prior to 10.0<br> ... |
I am assuming that this license (advmesh) is only in use for the short time that Design Simulation takes to mesh the model. After the model is meshed, the license is released, correct? |
The advmesh license is not released after the mesh. However, it is also enabled by ANSYS Structural and higher licenses at 9.0 and therefore does not require the advmesh license. You can test it... |
Are there any ANSYS LS-DYNA examples showing how to define orthotropic material
properties? The GUI only lists the minor form of the Poisson's ratios and my data
is in terms of the major form. |
Yes, please see the ANSYS LS-DYNA 9.0 example below (which is also attached to<br> this solution record). You can still enter the Poisson's ratio data in terms of the<br> ... |
How can I borrow a license file for a laptop? The license keys in the license file on the license server have the BORROW=nnn option (nnn is the maximum number of hours that licenses can be borrowed for) included. |
You should proceed as follows:<br><br>1) Make sure that you have at least version 9.0 of the ANSYS FLEXlm License Manager installed on the laptop - earlier versions do not support the... |
Question:
What is the version and name of the C, C++ and Fortran compiler for windows and Linux at the 10.0 release of ANSYS
Answer:
The compilers for PC machines for release 10.0 are currently
For WIN32
Fortran INTEL 8.1
C++ .NET 2003
For WIN64
Fortran INTEL 8.1
C++ INTEL 8.1
For WINX64
Fortran INTEL 8.1
C++ Microsoft Platform SDK April, 2003
For Linux - Opteron
PGI 5.2-4 (Portland Group)
For Linux - IA64, IA32 & EM64T
INTEL 8.1 for all |
Question:<br><br>What is the version and name of the C, C++ and Fortran compiler for windows and Linux at the 10.0 release of ANSYS<br><br>Answer:<br><br>The... |
QUESTION:
I'm looking at a circuit coupled LF emag model (2D PLANE53/CIRCU124) that was created by someone else and it is very awkward identifying with circuit element is connected to which coil mesh in the finite element model. Is there an easy way to do this? Also, a macro to automate the circuit coupling on CURR and EMF in the finite element mesh would be helpful, in case I have to redefine these coupled DOF sets.
ANSWER:
Attached are some macros that may help you in this regard. Please note that these are not ANSYS supported macros.
1) conn2.mac:
Prompts user to select a stranded coil circuit element. The macro automatically selects the PLANE53 coil elements to which the circuit element is coupled and displays them (EPLOT). This greatly simplifies determining the connections between circuit elements and finite element meshes of the coil elements. Note that coupled degree of freedom sets need not already be defined in order for this macro to work.
2) conn3.mac (PREP7):
Prompts user to select a stranded coil circuit element. The macro automatically selects the PLANE53 coil elements to which the circuit element is coupled and creates the coupled degree of freedom sets on CURR and EMF that are necessary for solution. This macro may only be used inPREP7.
3) crkt.mac (POST26):
Prompts user to select any circuit element in the circuit model. The macro reports the voltage across, current through, and power supplied by the circuit element (the listing appears in the output window for interactive post processing). This macro may only be used in POST26. |
QUESTION:<br><br>I'm looking at a circuit coupled LF emag model (2D PLANE53/CIRCU124) that was created by someone else and it is very awkward identifying with circuit element is... |
User is unable to update geometry for a .prj file by going through the Ansys Workbench Project Page.
User is able to update it by directly accessing the ICEM HEXA module.
The Ansys Workbench Project Page displays the following error:
error writing "sock1728": connection reset by peer
Also, a dialog is presented which contains several ines of code begining with ;
called "aie_wb_brep_attach" with too many arguments ....
Applies to:
Ansys Workbench 9.0
Advanced Meshing Module
ANSYS ICEM CFD/AI*Environment 5.1 |
This is a known error and has been resolved in SP2<br><br>Download and Install ANSYS ICEM CFD/AI*Environment 5.1 Service Pack 2<br> From the following site:<br><br>... |
When trying to run launcher90 or anslic_admin HP IA64, user is getting the following error:
The ANSYS Launcher could not be started.
The most likely reason is that your DISPLAY environment
variable is either not set (i.e. issue setenv DISPLAY
dev:0.0 where dev is your device name) or is set to a
device that is not accepting the connection.
DISPLAY is set correctly. What could be causing this error? |
On HP IA64 systems running 11.23 September 2004 patches or newer, any ANSYS Tcl-based application will fail to open a GUI due to bad system xnet libraries. This includes the launcher, ans_admin,... |
A user would like to get advise on how to specify a non-reflective
boundary condition in a structural wave-propagation problem. According
to the Release 5.3 Elements manual, this can be accomplished with a
surface effect element (i.e., SURF19) using MP,DAMP together with the
input of SURT and EFS real constants. When the user tries to do this,
he gets reflections. What's wrong? |
The use of the surface effect elements to absorb energy works as<br>documented in the manual. However, there are several things that<br>need to be understood in order to use this... |
Q. Is there a way to make bonded or no separation contact open when the underlying element stress reaches a user defined limit?
User is looking for something similar to the way TAUMAX works with friction, but in the normal direction, not tangential.
A. There is an undocumented feature which will allow you to define maximum allowable tensile and shear stresses to trigger a bonded or no-separation contact pair to separate. When this stress is reached, ANSYS automatically switches contact behavior from bonded to standard and you get the breakaway you are looking for. At 9.0, you have to define real constant number 27 as the max allowable tensile stress and/or real constant 28 as the max allowable shear. Attached is a simple test model. As with all undocumented features, this has not been fully tested, might not always work correctly and is subject to change without notice. Use at your own risk!
Alternatively, at 10.0 this can be simulated with cohesive zone modeling technology using a new family of interface elements (INTER202 thru INTER205). |
Q. Is there a way to make bonded or no separation contact open when the underlying element stress reaches a user defined limit?<br>User is looking for something similar to the way TAUMAX... |
**** Entered By: akpaul @ 06/09/2005 05:13 PM ****
Q. How can I use the Myrinet Interconnect for distributed parallel runs of CFX?
A. The Myrinet interconnect is supported by CFX at release 10.0 or higher. CFX relies on HP MPI libraries (included with CFX 10 installation) to communicate across Myrinet interconnects. In the Solver Manager, when setting up a distributed parallel run, set the start method to "HP MPI Distributed Parallel for x86 64" - HP MPI method should automatically use Myrinet interconnect if available.
On a system with multiple interconnects, HP-MPI will automatically select the fastest. This choice is based on a default ordering, which is: Quadrics (Elan), Myrinet (GM), InfiniBand (IT-API), or GigE(TCP/IP).
Supported platforms for HP-MPI are Intel Xeon 32-64 bit (Redhat/Suse Linux), Itanium 2 (Redhat/Suse Linux, HP-UX), Opteron (Redhat/Suse Linux), PA-RISC (HP-UX). A full list can be found on http://h21007.www2.hp.com/dspp/tech/tech_TechDocumentDetailPage_IDX/1,1701,1238,00.html
HP-MPI is more robust than the public domain MPICH (From Argonne National Labs) and has been found to be faster even on standard Ethernet. |
**** Entered By: akpaul @ 06/09/2005 05:13 PM ****<br>Q. How can I use the Myrinet Interconnect for distributed parallel runs of CFX?<br><br><br>A. The Myrinet interconnect... |
**** Entered By: akpaul @ 06/09/2005 05:25 PM ****
Q. I am using the "HP MPI Distributed Parallel for x86 64" start method for my distributed parallel run. The system has multiple interconnects. How can I be sure that CFX is using the Myrinet interconnect and not standard Ethernet?
A. There is a command line option which you can manually add to the start methods file in CFX_ROOT/etc/start-methods.ccl file, to get output from HP-MPI on which interconnect is being used.
Make a local copy of this file under ~/.cfx/10.0/start-methods.ccl {on windows %home%/.cfx/10.0/start-methods.ccl)
[If the '10.0' folder does not exist, create one]
Modify this line:
START METHOD: HP MPI Distributed Parallel
Start Command = mpirun -f %{req:appfile}
END
to:
START METHOD: HP MPI Distributed Parallel
Start Command = mpirun -prot -f %{req:appfile}
END
HP MPI will print the communication mode it selects to the std output (command window). If it runs Myrinet it will print "GM" in the output for each node in the parallel run.. A typical output will look like:
Host 0 -- ip 129.40.203.65 -- ranks 0 - 1
Host 1 -- ip 129.40.203.66 -- ranks 2 - 3
host | 0 1
======|===========
0 : SHM GM1 : GM SHM
If you want to force HP MPI to use Myrinet then add the following:
START METHOD: HP MPI Distributed Parallel
...Start Command = mpirun -gm -f %{req:appfile}
END
You might just cut and paste your own start method:
START METHOD: HP MPI Myrinet Distributed
Start Command = mpirun -gm -f %{req:appfile}
END |
**** Entered By: akpaul @ 06/09/2005 05:25 PM ****<br>Q. I am using the "HP MPI Distributed Parallel for x86 64" start method for my distributed parallel run. The system has... |
Q. How to correctly compute the solver scale up for parallel runs?
A. When computing solver scale up for parallel runs, the correct 'Clock Time' to use is the "CFD Solver wall clock seconds" printed right after the last iteration in the Output file. Example:
======================================================================
OUTER LOOP ITERATION = 784 ( 200) CPU SECONDS = 3.593E+03 (1.068E+03)
----------------------------------------------------------------------
| Equation | Rate | RMS Res | Max Res | Linear Solution |
+----------------------+------+---------+---------+------------------+
| U-Mom | 0.97 | 1.5E-05 | 3.3E-04 | 6.9E-03 OK|
| V-Mom | 0.98 | 3.1E-06 | 9.4E-05 | 1.6E-02 OK|
| W-Mom | 1.05 | 4.1E-06 | 3.6E-05 | 6.2E-03 OK|
| P-Mass | 0.82 | 2.4E-06 | 3.3E-05 | 8.5 2.9E-02 OK|
+----------------------+------+---------+---------+------------------+
| H-Energy | 0.90 | 4.0E-05 | 8.7E-04 | 5.6 6.1E-03 OK|
+----------------------+------+---------+---------+------------------+
| K-TurbKE | 0.87 | 1.1E-05 | 4.2E-04 | 5.6 1.1E-02 OK|
| E-Diss.K | 0.71 | 1.5E-05 | 4.5E-04 | 7.1 4.1E-03 OK|
+----------------------+------+---------+---------+------------------+
| Water Vapour at 100 | 0.94 | 3.4E-05 | 7.2E-04 | 5.6 1.7E-02 OK|
+----------------------+------+---------+---------+------------------+
CFD Solver finished: Thu Jun 9 13:51:42 2005
CFD Solver wall clock seconds: 1.1380E+03
All other CPU times and Wall clocks times reported will include disk I/O overheads and partitioning overheads and hence will not give a true representation of the the solver scale up as you add partitions. |
Q. How to correctly compute the solver scale up for parallel runs?<br><br>A. When computing solver scale up for parallel runs, the correct 'Clock Time' to use is the... |
I ran a test model and found that EPTO,EQV < EPPL,EQV, which seems counter intuitive, since, from equation 19-170 in the theory manual:
EPTOeqv = EPELeqv + EPPLeqv + EPCReqv
Is there something fundamental about the way these results quantities are related that I am missing? Is it possible for EPPL,EQV > EPTO,EQV? Does the choice of EFFNU in the AVPRIN command affect the reported values? |
The short answer to this question is yes. But the explanation for why this is so is quite involved. Please see the "Appendix" below, and in particular, the findings summarized in the... |
**** Entered By: rlange @ 06/16/2005 10:36 AM ****
Q. I want to interpolate results from one simulation to a new simulation. The problems have the same geometry but a different number of domains. CFX does not seem to want to allow this interpolation. Is there any way to accomplish this interpolation?
A. Yes, it is possible, using profiles. From the first analysis, in CFX-Post, export with BC Profile format, a custom set of result, say u, v, w, p, t, k, e.
The export will be on a domain basis. If there were three domains, for example, the exported file will have three different header sections. You then delete the second and third sets of header lines so that it looks like the file is for a single domain.
Once it is initialized in PRE, (we're in the second case now) you will see the function name in the Library section. Use that in the t=0 Global initialization tab as an expression. i.e. Static.Velocity u(x,y,z) for initializing the U velocity. "Static" is replaced by whatever you see under functions in the library after Profile initialization. |
**** Entered By: rlange @ 06/16/2005 10:36 AM ****<br><br>Q. I want to interpolate results from one simulation to a new simulation. The problems have the same geometry but a... |
I have a created a set of blade, shroud and hub curves from a turbine nozzle vane model in SolidWorks/Rhino and written out an iges file to import into Bladegen through the Wizard. However, the trailing edges give a very poor representation. I have also exported the vanes as an IGES from BladeGen back into Rhino and compared them with the original vanes, but the leading and trailing edges are different, although the overall shape is acceptable.
Can you suggest a more reliable method to get these vanes in to BladeGen? |
<br>Sometimes during the import of IGES files using the BladeGen Import Wizard there can be problems with the definition of the leading (LE) and trailing (TE) edges. Exporting via IGES is... |
Is ICEM CFD Hexa CAA V5 supported on an Itanium II processor with SGI Linux RedHat? |
Catia V5 does not support LINUX, nor do the Catia V5 plug-ins. You must use Windows 2000 or XP for any Catia V5 applications.<br><br>Check out the URL <a target=_blank... |