Why can't I create a bolt pretension section on the model that I brought from Simulation to ANSYS? |
To create a pretension section in ANSYS your mesh must have a structural element type. The mesh that you brought over from Simulation currently has MESH200 (mesh only) elements. Change these... |
**** Entered By: rlange @ 03/23/2005 12:04 PM ****
Q. How do I implement and use my own RGP table in CFX5.7.1?
A. RGP gas tables for some materials are provided in the materials-extra CFX installation folder. They are used by creating a new material. The Option is Table, and the Table Format is TASCflow RGP. The table name is the path to the RGP file, and you browse to find it. Then be sure to enter the Component Name exactly as it appears in the header of the RGP file (below the $$$Data line). Also choose Gas as the thermodynamic state. |
**** Entered By: rlange @ 03/23/2005 12:04 PM ****<br><br>Q. How do I implement and use my own RGP table in CFX5.7.1?<br><br>A. RGP gas tables for some materials are... |
How can I create a line mesh control with points inside the domain (i.e. by prescribing the points coordinates rather than by selecting an existing point)? |
You should proceed as follows:<br>1. Create a 'point spacing'<br>2. Create a line control. After inserting the line control, you will have in the point entry in the... |
Solving in Workbench Simulation produces the following error:
Ansys - Error - An internal solution magnitude limit was exceeded. Please check your Environment for inappropriate load values or insufficient supports
The Solution Information - Solver Output (solve.out file) shows an error similar to the following :
*** ERROR *** CP = 999.999 TIME= 12:34:00
Smallest negative equation solver pivot term encountered at Ux DOF of
node NNNNNN. Check for an insufficiently constained model.
Note: Setting a different working Units in Simulation may allow the model to solve successfully. |
Pivot term errors mean that a model is unconstrained.<br><br> The simplest way to troubleshoot this error is to create a modal analysis (Frequency Finder) and look for a zero body... |
The following error message appears when running ANSYS 9.0 using *MFOURI. It successfully calculated the Fourier coefficients at release 8.1 but no longer does so at 9.0.
/COM,ANSYS RELEASE 9.0A1 UP20050128 14:16:06 03/14/2005
*** ERROR *** CP = 11.737 TIME= 14:19:19
Error in isimeq called by vopmfr. Please send data to your ANSYS
Technical Support Provider. |
This error was previously reported (defect 30847) and apparently resolved as being a case of error in input (see notes below). This particular problem was resolved by modifying the input file to... |
Q. I have set up a Conjugate Heat Transfer Analysis for CFX by importing a single assembly and defining solid and fluid domains. Since I have a continuous mesh between the solid and fluid domains, I expect that a solid-fluid interface will be automatically setup. However, it appears that this was not done, as I see in the results that the fluid temperature is unchanged. What should I check in my setup?
A. Be sure that you have not applied any boundary conditions to surfaces that will act as solid-fluid interfaces. If you name any of these faces, CFX requires that you set up the 1:1 interface manually. It looks at the default surface when automatically creating the interfaces. |
Q. I have set up a Conjugate Heat Transfer Analysis for CFX by importing a single assembly and defining solid and fluid domains. Since I have a continuous mesh between the solid and fluid... |
Question:
Were there any changes to the format of the results databases (in particular the *.RTH and *.RST file) between 8.0 and 9.0?
Answer:
There was a major change to the results files at release 9.0; the entire file now uses 64 bit pointers. The addition of 64 bit pointers to the solution part of the results files at release 5.5 allowed for files larger than 2 Gb without file splitting and was believed, at the time, to be all that was required. It was assumed at the time that the front end of the file that contains the geometry would not need long pointers. That is the change at 9.0. If you are reading results files, it will be necessary to convert the pointers to long integer form. If you are writing results files, setting the release identifier to 8.0 will solve the problem in the short term. In the long term it will likely become necessary to convert the writing program to long integers.
I am attaching 3 files for you. One identifies the differences. The other 2 are the documentation files for the results files for release 8.0 and release 9.0
The second change is the format of the nodes. Previously, the node record was a single record that contained all of the nodes and their coordinates. Beginning at 9.0, the nodes are written individually so the number of node records increased from 1 to the number of nodes. This is also because of the size of the geometry is now unlimited. |
Question:<br><br>Were there any changes to the format of the results databases (in particular the *.RTH and *.RST file) between 8.0 and... |
QUESTION:
Can't you use TIMINT,OFF command in 3D transient EMAG to establish initial conditions (current distribution + field) when the conductor is modeled with keyopt(1)=1 (time integrated electric potential)?
ANSWER:
Here's an input file that shows initial conditions solved for in the first load step with TIMINT,ON (end time 1.000 s). The current is abruptly turned off in load step 2 (end time 1.0005 s). POST26 variable #2 is the current density in an arbitrary element; this decays rather gradually with time. Variable #3 is the TOTAL conductor current which is completely shut off after the second substep of load step 2. This is qualitatively consistent with expected behavior. The current will continue to flow into the conductor in some parts of the cross section, and out of the conductor in other parts of the cross section. The net current in load step 2 is zero, however it is nonzero at specific locations.
It is not clear why the time integrated electric potential will not give a solution with TIMINT,OFF. Perhaps this has to do with the fact that VOLT is NOT physical voltage but is instead the time integral of voltage. Another curious thing I noticed is that when loads are stepped (KBC,1) the first substep applies only half of the full value. I remember seeing this behavior perhaps 8 years ago and I never understood why ANSYS does this.
I am still awaiting reply from development on why this is so. The only thing I can suggest is to leave time integration effects on for the first load step used to establish the initial current and field when modeling conductors with the time integrated electric potental. |
QUESTION:<br>Can't you use TIMINT,OFF command in 3D transient EMAG to establish initial conditions (current distribution + field) when the conductor is modeled with keyopt(1)=1 (time... |
Do you have any examples demonstrating usage of the DMPEXT command (to extract damping coefficients for use with the FLUID136 squeeze film element)? None of the examples in the online documentation contain an example of DMPEXT (although examples of ABEXTRACT, which call DMPEXT, may be found). |
The DMPEXT QA set and a word document are attached to this solution.
|
**** Entered By: akpaul @ 03/28/2005 04:44 PM ****
Q. If I have two boundary conditions at regions next to each other (i.e., sharing an edge) such as an Inlet and Wall, how does the CFX Solver and CFX-Post handle this?
1) For CFX-5, how are both boundary conditions applied to the control volume when they share an edge?
2) For CFX-Post, how is this treated when post-processing "Hybrid" results?
A.
- The flow solver always uses the exact value specified by the user to calculate the flux through the control volume face on a wall or the inlet, when they share an edge. (Similarly for other combination's of BC's).This is because there will be separate control volume faces attached to the wall and inlet for the given node/vertex on the shared edge.
- Hence the control volume value is a function of both the wall value and the inlet value. i.e. both values are applied by the flow solver for the same control volume.
- The "hybrid" value displayed at a node by CFX Post has to be one or the other and hence is not exactly the boundary condition applied by the flow solver (it's one or the other).
Hierarchy for plotting in CFX - POST in such situations for most variables is this (in decreasing order):
Inlet, Fluid Solid Interface, Wall, Opening, Outlet, Interface
Hierarchy for pressure is this:
Outlet, Fluid Solid Interface, Wall, Inlet, Opening, Interface
Hierarchy for wall only values:
Fluid Solid Interface, Wall, all others have same priority |
**** Entered By: akpaul @ 03/28/2005 04:44 PM ****<br><br><br>Q. If I have two boundary conditions at regions next to each other (i.e., sharing an edge) such as an Inlet and... |
QUESTION:
User wants to be able to manually control the OSLM and override the defaults set by SOLCONTROL. When SOLCON is on and TINTP is used to adjust the oscillation limit, ANSYS echoes (output file or window) that the new OSCL setting is accepted. However, it is not being used during solution (the default gets used instead).
ANSWER:
The workaround is to disable SOLCON, do NOT use OPNCONTROL, but turn on AUTO TIMESTEPPING:
C**************************************************************
C*** LIST DYNAMIC LOAD OPTIONS BEFORE TINTP COMMAND
C**************************************************************
solcontrol,off,off
DYNO
STAT
C**************************************************************
C*** LIST DYNAMIC LOAD OPTIONS AFTER TINTP COMMAND
C**************************************************************
tintp,,,,,0.75,0.1 !gamma,alpha,delta,theta,oslm,tol transient int par
DYNO
STAT
C**************************************************************
C*** SOLVE
C**************************************************************
autots,on
time,0.0001
outres,all,all
nsub,500,500,2
alls
save
solv
fini
C**************************************************************
C*** PRODUCE XY PLOT OD OSCILLATION LIMIT
C**************************************************************
/post26
solu,3,dtime
solu,4,reseig
prod,2,3,4,,osc_lim
/gro,ltyp,1
/axl,x,Time
/axl,y,Oscillation Limit
plva,2 |
QUESTION:<br><br>User wants to be able to manually control the OSLM and override the defaults set by SOLCONTROL. When SOLCON is on and TINTP is used to adjust the oscillation limit,... |
Q. How does ANSYS handle cast iron material plasticity beyond the user defined stress-strain data points?
A. Cast iron material behaves same as MISO, elastic-perfectly plastic beyond the last user defined data point in both tension and compression.
This detail will be included in 10.0 Theory Manual, Section 1.4.17. |
Q. How does ANSYS handle cast iron material plasticity beyond the user defined stress-strain data points?<br><br>A. Cast iron material behaves same as MISO, elastic-perfectly plastic... |
**** Entered By: akpaul @ 03/29/2005 01:48 PM ****
Q. I want to define 100 or more nozzles at specific X,Y,Z locations. The X,Y,Z data is in a text/data file - how can I do this in a efficient way without having to define each nozzle manaully in CFX-PRE
A. An efficient way of doing this is by using perl scripted CCL (CFX Command Language). The following example is based on Lagragian multiphase models, since it is easier to implement. You should be able to follow the same method for an Eulerian approach using point sources. The example requires a CFX Multiphase license. But users should be able to follow the same approach for any similar application. Required files are attached.
Example :
Specifying point 'Particle Injection Regions' (Nozzles) by reading in the X,Y,Z location of each point from a data file.
Instructions:
1) Create a new simulation in CFX-PRE. Setup domains, boundary conditions and properties as required. In this case include a 'Particle' phase.
2)On the domains form, create one 'Particle Injection Region'. This is a dummy definition, so we can get the required CCL for modification later.
3)Set appropriate solver controls and initial guess
4)Write out a solver 'definition' file - in this case NozzleTest.def.
5) From the CFX-5 Launcher (make sure the working directory is set correctly), go to Tools>Command Line.
6) At the DOS prompt, enter the following command:
cfx5cmds -read -def NozzleTest.def -text NozzleTest.ccl
This will extract a text file 'NozzleTest.ccl' in your working folder.
7) Open the text file in Wordpad and look for the following section:
PARTICLE INJECTION REGION: Particle Injection Regions 1
FLUID: Water at 25 C
INJECTION CONDITIONS:
INJECTION METHOD:
Cone Angle = 25.0 [deg]
Injection Centre = 0.6 [m], 0.5 [m], 0.25 [m]
Injection Velocity Magnitude = 300 [m s^-1]
Number of Positions = 50
Option = Cone
INJECTIO |
**** Entered By: akpaul @ 03/29/2005 01:48 PM ****<br>Q. I want to define 100 or more nozzles at specific X,Y,Z locations. The X,Y,Z data is in a text/data file - how can I do this in a... |
How to define a very large number of boundary conditions (example 100 nozzles) without have to manually create each in CFX-PRE? |
You could create one in CFX-PRE and then edit the CCL file (CFX Command Language) to add a perl loop to generate the actual number of boundary conditions, which are generated at the start of the... |
Q. While working in /POST1 nonlinear diagnostics dialogue box or while trying to save a PGR file, I sometimes get the following fatal error message:
*** FATAL *** CP = 3.703 TIME= 15:50:18
The low level memory manager failed to return a valid memory handle for
a free operation. Pointer value: 135559104. It is likely that the
code is calling for a free with a pointer that was not allocated via
the memory manager.
What is the cause?
A. This is a defect (#29776) that has been corrected for 10.0. |
Q. While working in /POST1 nonlinear diagnostics dialogue box or while trying to save a PGR file, I sometimes get the following fatal error message:<br><br>*** FATAL *** ... |
Q. There seem to be some inconsistencies in what DM considers the starting point of the edge, causing users to have to guess and then make corrections.
How do you determine the frame of reference for point feature dimensions?
A. Starting point of edge depends on how geometry was generated. There is no easy way to determine the frame of reference for point feature dimensions other then trial and error. Development is considering more user friendly point feature preview tools in a furture release. |
Q. There seem to be some inconsistencies in what DM considers the starting point of the edge, causing users to have to guess and then make corrections.<br>How do you determine the frame of... |
Does ANSYS support Physical Address Extension(PAE) of Windows?
http://www.microsoft.com/whdc/system/platform/server/pae/default.mspx |
No
|
Q. Has the tolerance for defining multilinear stress-strain data changed since ANSYS 7.1? It seems to be a tighter tolerance. What is the tolerance at ANSYS 9.0, and can it be changed by the user?
A. The tolerance is 0.001*E. This has not changed since ANSYS 5.7. What has changed is that prior to 7.1, ANSYS just issued a warning if the slope of stress-strain curve was lower then E by 0.001E and let you pass if it was greater then E. After 7.1, the warning was switched to an Error and it was issued for slopes either greater then or less then E by an amount of 0.001E. The reason for the tighter restriction was that there were applications sent to us in which results were considered inaccurate and questionable. The explanation for the discrepancy was traced back to the loose variation between E and the slope derived from stress-strain data. The tolerance is hard coded and justified for quality control reasons. There are no plans to tighten the restriction any further. One of the reasons we added the ability to define stress vs. plastic strain directly was to get around this difficulty. |
Q. Has the tolerance for defining multilinear stress-strain data changed since ANSYS 7.1? It seems to be a tighter tolerance. What is the tolerance at ANSYS 9.0, and can it be changed by the... |
Q. I have done a CFX flow analysis with units of pressure in PSI. I am trying to use the CFX Post method to create the SFE commands to do a one way Fluid Structure Interaction analysis. However, the pressures indicated in the SFE commands are not close to the pressures that I am seeing in the CFD model. What is wrong.
A. In CFX 5.7.1, there are problems in the load transfer if FSI units are not used. The error has been reported as defect 31219. |
Q. I have done a CFX flow analysis with units of pressure in PSI. I am trying to use the CFX Post method to create the SFE commands to do a one way Fluid Structure Interaction analysis. ... |
Are there any examples of plotting internal energy in POST26 and strains
in POST1 using ANSYS LS-DYNA 8.1? I cannot get these to plot for me. |
Please see the model below (which is also attached to this solution<br> database along with the resulting output plots). The model includes<br> both shell and solid... |
######## WORK NOTES COPIED FROM INCIDENT 391529 ########
QUESTION:
I have vendor supplied piezoelectric coefficients relating strain to electric field. ANSYS requires differently defined coefficients (relating stress to the field).
How do I convert the vendor-supplied data to the form required by ANSYS?
ANSWER:
The attached word document describes this in detail. A macro (also attached) has been created to automate the conversion. |
######## WORK NOTES COPIED FROM INCIDENT 391529 ########<br><br><br>QUESTION:<br><br>I have vendor supplied piezoelectric coefficients relating strain to electric... |
Q. When running a structural 2D analysis in Workbench 9.0 with a Professional License, it meshes automatically with PLANE18X Elements for 2d problems and then issues the following error message when trying to solve. Why is this element type not supported by ANSYS Porfessional?
**************
*** ERROR ***
The element option PLANE183 was either not ordered with this ANSYS
installation or the appropriate product was not selected for this
session. Contact your ANSYS support person for more information.
**************
A. This limitation has been removed for Workbench 9.0SP1A. PLANE18X elements as well as PLAN42 and 82 are supported in Workbench 2D with a Professional License. |
Q. When running a structural 2D analysis in Workbench 9.0 with a Professional License, it meshes automatically with PLANE18X Elements for 2d problems and then issues the following error message... |
Q. Using "OUTRES, Item, Freq, Cname" to save only results at entities predefined to component names runs without error at 8.1, but in 9.0 this information is not available.
A. This is a defect (#31127) which has been corrected at 10.0. Below is a demo APDL to reproduce the problem at 9.0.
fini
/clear
/prep7
et,1,185
mp,ex,1,2.1e5
block,0,100,0,10,0,10
esize,5
vmesh,all
nsel,s,loc,x,0
d,all,all,0
nsel,s,loc,x,40,50
esln,s,1
cm,mv_e,elem
cm,mv_n,node
/solu
nsel,s,loc,x,100
d,all,uy,-10
nlgeom,on
nsubs,10,10,10
alls
outres,erase
outres,all,last
outres,esol,all,mv_e
outres,nsol,all,mv_n
solve
nsel,s,loc,x,100
d,all,uy,-20
nsubs,10,10,10
alls
solve
/post1
set,first
plns,s,eqv |
Q. Using "OUTRES, Item, Freq, Cname" to save only results at entities predefined to component names runs without error at 8.1, but in 9.0 this information is not... |
When attaching a Solid Edge model into Workbench Simulation or DesignModeler, I get the following errors:
Plug-in error: Unable to activate document
and
Cannot access specified file. No active document.
The Ansys menu is displayed in SolidEdge. A valid model is loaded in SolidEdge.
Reinstalling SolidEdge and Workbench has not resolved the problem. |
The SolidEdge license level (Design and Drafting or Academic license) may be limiting access to API functionality, which is necessary for the Ansys Workbench plugin.<br><br>Solid Edge... |
Are there any examples of an MSUP transient dynamic analysis? I cannot
get my acceleration load to have any effect in the results. |
Yes, please see the ANSYS 9.0 spring-mass example below (and attached).<br> It is important to note how the loads are applied. In this model, the gravity<br> (acceleration)... |