Creating coupled set 1 works, why when I create set 2 do I get a warning that couple (CP) command is ignored? |
If you use the ALL degrees of freedom option on the CP command and specify a set number of 1 (CP,1,ALL), then you will get several CP sets depending on available DOFs: CP1 for UX, CP2 for UY, CP3... |
User installed Workbench on Windows and Mcafee Virus Scan 8.0i Enterprise.
When the user tries to open a new Simulation or tries to open an existing one, Workbench crashes with an AnsysWBU.exe error
ERROR MESSAGE RECEIVED:
AnsysWBU.exe has encountered a problem and must close. Sorry for the inconvenience. |
The program crashes in the McAfee ScriptScan code.<br> McAfee has included a fix for this in 8.0i Patch 11<br><br><br>Workaround 1:<br><br> Install McAfee... |
Q: A user has a surface model of a cylinder. (i.e., A common pressure vessel.). If you apply a frictionless support to the symmetrical edges, the results do not appear correct.
How can I correctly apply a symmetry boundary condition to a surface model?
A. Applying a frictionless support to the symmetrical edges are not correct for a surface model. This will only work for Solid models..
Shell models have rotational degrees of freedom, which must be accounted for.
Workaround
1. Create a Named Selection scoped to the cut edges of the symmetry plane. Name it "cut_edge".
2. Insert a Commands Object using the following
! Commands inserted into this file will be executed just prior to the Ansys SOLVE command.
! These commands may supercede command settings set by Workbench.
! Active UNIT system in Workbench when this object was created: U.S. Customary (in, lbm, lbf, F, s)
cmsel,s,cut_edge,node
d,all,rotx,0
d,all,roty,0
d,all,uz,0
cmsel,all,all
3. The best explanation of how to apply symmetric boundary conditions is in the Ansys Basic Analysis Guide, Section 2.6.4., Applying Symmetry or Antisymmetry Boundary Conditions
4. Simulation V10.0 will have a Beta feature that will allow you to insert a "Symmetry Condition" into the Model Branch that will automatically handle the application of BC's for symmetry. |
Q: A user has a surface model of a cylinder. (i.e., A common pressure vessel.). If you apply a frictionless support to the symmetrical edges, the results do not appear correct.<br> How... |
How can I remove a given displacement load for a particular load step in a sequenced (multiple load-step) simulation? |
In a Workbench Simulation, it is not possible to enter all blanks (-,-,-) to represent "free in all directions" for a given displacement type load.<br><br> Workaround 1:... |
QUESTION:
Do you have an example of a SAW (surface acoustic wave) device?
ANSWER:
Attached is a model of a hypothetical SAW device. In 3D, you can use SOLID5 or SOLID226/227 to model piezoelectric devices. The attached macro (bode02.mac) was used to create the bode plot you see in the presentation. |
QUESTION:<br>Do you have an example of a SAW (surface acoustic wave) device?<br><br><br>ANSWER:<br>Attached is a model of a hypothetical SAW device. In 3D, you can... |
######## WORK NOTES COPIED FROM INCIDENT 393941 ########
**** Entered By: clsalvin @ 04/11/2005 05:18 PM ****
QUESTION 1:
Using the stacking factor approach proposed in the Ansys LF EMAG Training Manual, Release 8, Pg 2-27,
there is a model for mu_rx, and mur_y. I plotted these as a function
of the stacking factor, and the resulting plot is attached as the file
'StackingFactor.bmp'.
Looking at this file, I see that in the y-direction the mu_r is extremely
small. I am wondering if this is expected? It should mean that the
attraction
force would be extremely small, if an axially-magnetized magnet was
placed UNDERNEATH the stack shown on Page 2-27. But our general
observation suggests otherwise!
ANSWER 1:
This variation in the in-plane and normal relative permeability versus stacking factor (your plot) is exactly what I would expect. The stacking factor is useful for calculating the overall FIELD, but I can see how FORCES may not be correctly calculated for the normal (y) direction because PHYSICALLY, the flux from the PM will impinge the first iron lamination which has the full value of mu_r. The "smeared properties" modeling approach will in effect present the impinging flux with a much lower mu_iron/mu_air ratio and so calculated forces could be significantly lower than expected (since force increases with difference in permeability at the material interface).
If force is important to you, I would model individual laminations near the surface and transition to a "smeared properties" mesh below the laminations to reduce element count.
QUESTION 2:
I want to model different nonlinear permeability acting in different directions in a 3D model. I see in your documentation that one may only define a single BH curve for any given material, with optional linear permeabilities superceding the BH curve in specified directions. Is there a workaround allowing definition of different BH curves actingin different directions?
ANSWER 2:
I do not h |
######## WORK NOTES COPIED FROM INCIDENT 393941 ########<br>**** Entered By: clsalvin @ 04/11/2005 05:18 PM ****<br><br><br><br><br><br>QUESTION... |
How do I plot a flow field as stream lines on a 2-d plane? |
<br>You can plot streamlines based on a velocity that has no component normal to the plane. For example, if you are plotting on an x-y plane, the velocity in the z direction would be 0.0... |
When I try to write the definition file in CFX-Pre I receive the following error in a pop-up window:
Error on writing definition file:
ERROR
Error writing data to Definition file data set: G/COMMANDS |
<br>The error message implies that there may be a .lck file in your working directory. If there is one then please delete it and try to create your def file again.
|
How do I switch between Conservative and Hybrid values when exporting data or creating a chart in CFX-Post? |
<br>Select the Variables tab in CFX-Post, click on the variable of interest, toggle between Conservative and Hybrid.
|
When I use Tools>Interpolate in the CFX-Solver Manager with "Calculate differences" to compare the solution fields in two results files, how is the difference calculated? |
<br>The difference is calculated as:<br><br>(values in "from" file) - (values in "destination" file)
|
Q. I am running a very large CFX job on a 64 bit UNIX machine and CFX is unable to write the large (well over 2GB) results file. The result of the UNIX "limit" command shows that the file size is unlimited. What could be wrong.
A. There is no limit on a 64 bit Linux/UNIX machine.
If you are writing to a network mounted drive then you should make
sure that the NFS daemon/client software is the 64 bit version.
Also make sure that the file system with which your hard
disks are formatted supports files > 2GB (either networked or not).
Older OSes used 32-bit numbers for storing the length of files, and also
used a negative number as an error indicator: this implies a 2Gb limit (2^31
bytes) on file size, enforced by the OS, and expected by programs.
A few years ago, UNIX vendors produced a new interface for programmers to
access files larger than 2Gb, but of course this can't be used with older
programs, as they only set aside space for a 32-bit file size. So, one way
to get this error is if the application has not been built with support for
these large files. CFX is built with this support.
The other place this change affected was the OS itself: anything which
handles access to files had to be upgraded to support these larger files.
Generally for local filesystems this was easy, but for things such as NFS
which involve multiple computers it's harder to do consistently, and doesn't
always happen by default. So it's quite possible that the OS is simply not
allowing him to write files which are any larger then 2Gb.
Quota problems *should* give a different error, but this does come from the OS too,
and an odd system setup it might not be reporting it properly.
(Info provided by Andy Mortimer and Dan Williams) |
Q. I am running a very large CFX job on a 64 bit UNIX machine and CFX is unable to write the large (well over 2GB) results file. The result of the UNIX "limit" command shows that... |
When changing the solver from TASCflow to CFX-5 there is a problem in convergence. The attached file shows the two residual histories. We have tried to solve the problem by checking all settings etc. The grid is exactly the same in both cases. Why is this happening? |
The problem is related to using stage interfaces perpendicular to the flow direction where recirculating flow exists. This leads to recirculating flow across the interface. The problem is smeared... |
When trying to run Workbench 8.1 or 9.0 on a Windows machine with McAfee Virus Scan's Scriptscan feature enabled, the
user will receive the following error:
AnsysWBU.exe has encountered a problem and must close. Sorry for the inconvenience.
What is the cause of this error and is there a workaround? |
The error occurs due to a defect in McAfee's Scriptscan code.<br>A problem report has been filed with Network Associates but we will add code to detect this<br>situation in... |
Can you install Workbench 9.0 and Workbench ED 9.0 on the same machine? |
No, one must be uninstalled to install the other.<br>Both can not exist on the same machine, as ANSYS 9.0 and ANSYS ED 9.0 can.
|
Question
In UserMat.f, there are a number of parameters that are not clearly identified. In particular, for code parallelization, we need to know the Domain number, integration point number, element element. Would you please explain the meaning of the arguments
kDomIntPt (int,sc,i) "k"th domain integration point
kLayer (int,sc,i) "k"th layer
kSectPt (int,sc,i) "k"th Section point
Answer:
The word DOMAIN as used in USERMAT and the word DOMAIN as used in parallel processing do not refer to the same thing. In USERMAT, domain refers to the range of integration points for a particular element. In parallel processing, domain refers to the CPU number being used for a particular part of the processing. The kDomIntPt argument in USERMAT has nothing to do with parallel processing and can only be used to identify the location within the element where USERMAT calculations are being performed. This SOLUTION address the USERMAT arguments and how the values for the arguments are calculated for the various ANSYS elements that call USERMAT.
The USERMAT subroutine is called once for each integration point in the calling element. The current integration point is identified by the arguments list, including the arguments
elemId ' the element number
kDomIntPt ' the overall identifier for the element integration point
kLayer ' the layer number for elements that have multiple layers
kSectPt ' the section integration point number
USERMAT is only called by some elements new to the program after release 5.0. Elements available in release 4.4 do not call USERMAT as an option but may call other non-linear material user subroutines. This solution does not address any of those other non-linear user subroutines. And not all newer elements call USERMAT. Elements that currently (release 9.0) call USERMAT include the 18x family of elements, SHELL131/132 and SHELL 208/209.
The actual number of integration points varies with element type and depends o |
Question<br><br>In UserMat.f, there are a number of parameters that are not clearly identified. In particular, for code parallelization, we need to know the Domain number, integration... |
I have used CFX4 for axisymmetric calculations using a 2D mesh. In CFX4 I specified a boundary condition at the rotation axis. How can I set up a similar calculation in CFX5 using a 2D mesh from ICEM or CFX4 |
It is not possible to set a rotation axis boundary conditions in CFX5, however it is possible to solve the problem you ask for in one of three different ways.<br><br>1) If you have... |
I am trying to interpret the orientation of the stiffness terms on the .emat file so I can reconstruct the matrix for another application external to ANSYS. Can anyone help me convert the emat data into standard matrix form? |
Please see the information below.<br><br><br><br>To: ANSYS EMAT file Readers<br><br>From: John Thompson, ANSYS Technical Support<br><br>Date: April... |
My PLANE42 thermal strain analysis with temperature-dependent coefficient of
thermal expansion properties gives strange results, which are magnified as the
thermal gradient increases. Is this a known error? |
You could be hitting Defect 32678, which was fixed in the UP20050412 build of<br> ANSYS 10.0. A Class 3 Error report is pending. This error exists because the<br> ... |
How can I use CFX to model the solidification of molten aluminum ... on the perimeter of a pipe?
****************************************************************************************************** |
<br><br>You should be able to do this with an Equilibrium Phase Change Model, with the limiting assumption that the solidification is an equilibrium process.<br><br>1.... |
Q. Two different rigid constraint equations are generated and saved to file.s01 and file.s02 files.
When user attempts to run this solution with LSSOLVE,1,2,1, he gets:
**************
Error:
Constraint equation 1 does not exist. The CECMOD command is ignored.
**************
He can do this without error using sequential solves without leaving solution.
What is the problem?
A. It is a code limitation regarding CERIG and LSWRITE/LSREAD. At 10.0, we will document that
CERIGID cannot be deleted by using cedele,all and regenerated, in the 2nd load steps or higher if LSSWRITE
and LSSOLVE procedure is used. The reason is that CERIGID generates CE's directly put into load step files (
jobnames.s01, .s02, ...). The deletion of CE (cedele,all) cannot always maintain the consistency among
load steps. |
Q. Two different rigid constraint equations are generated and saved to file.s01 and file.s02 files.<br>When user attempts to run this solution with LSSOLVE,1,2,1, he... |
Q. How do I accurately measure the volume of the cloud inside an iso-surface defined at a fixed mass fraction? I would ideally like to be able to carry out such calcs using Post, rather than User FORTRAN.
A.
If you create a volume in post (isovolume) using "above value" with your specified criterion, and the inclusive switch on, then you get every element that the isosurface passes through, plus those inside it. That is, when you calculate its volume using the function calculator it is slightly larger than ideally indicated by the isosurface. If you switch the inclusive flag off, you only get those elements who lie entirely inside the isosurface. Hence its volume is slightly smaller than the isosurface might suggest. The problem is, the result is mesh sensitive and only in the limit of very small elements do you get the same (accurate) answer.
Another solution is to do the calculation yourself using a new variable that then better takes into account partial elements rather than only whole ones. Create a new variable in Post called "cloud" (or something) and make it equal to 'step(gas.mf - critval)'. Thus inside the cloud values will be one, and outside they will be zero. Then in the function calculator perform the volumeInt of your new variable cloudwith the entire assembly as the location. This should give a more accurate value for the enclosed volume, and incidentally it will be somewhere between the 2 volume values obtained above from isovolumes.
Q2. How do I now get the volume averaged mole fraction of gas within this iso-surface.
A2.
For the volume average bit, create another variable in Post (called "cloudmf" perhaps?) equal to the mass fraction times cloud, or 'gas.mf*step(gas.mf-critval)' . Then to get the volume weighted average within the cloud, calculate the volume integral of cloudmf, and divide by the volume integral of cloud. Always use the entire assembly as the location for these calculations, and you should be ok. |
Q. How do I accurately measure the volume of the cloud inside an iso-surface defined at a fixed mass fraction? I would ideally like to be able to carry out such calcs using Post, rather than User... |
Why does the License Manager go down when user disconnects from Internet?
The customer's license server is a laptop. The license is tied to the disk serial number.
Whenever the user disconnects from the Internet, the ANSYS license manager goes down. |
The ANSYS exe is connected to the socket through the IP address that was used when user originally started the ANSYS run.<br>This is why they require a static IP address for license... |
A customer has installed ANSYS 8.1 on an SGI platform.
When launching an interactive session, the following message appears:
"Cannot allocate color map gray 40".
The entire output window contains just a white empty window.
How can I correct this error? |
If users get this error, the workaround is to have them add -bg blue, or some available color, to their Additional Options emulation setting in the Launcher.<br><br>Launcher90 >... |
What is the CFX5 NOLIMIT increment mean in a CFX license? |
The 'NOLIMIT' license allows the CFX-5 Solver to run on an unlimited number of nodes.<br>Without this feature, the solver is problem size limited to 125,000... |
Customer has installed ANSYS 8.1 on a Linux machine. This machine is also it's own license server.
When the license manager for 9.0 is installed on the Linux machine, an error is received about utilities for stopping and uninstalling the license manager that cannot be found.
The customer is asked for the correct location but no choice can be made that satisfies the program.
The reason for this appears to be that the license manager for 8.1 was not stopped and uninstalled before installing the ANSYS 9.0 license manager.
This results in an old directory named linuxia32 and a new directory linia32.
If the license manager for version 8.1 is removed before installation of the new license manager, the linuxia32 directory doesn't exist and there is no problem.
What is causing this to happen? |
The 9.0 install can't shut down earlier license managers from 8.1 or prior because we changed the directory naming.<br>The workaround is to shut down the license manager before... |