Why did distorted elements not erode even when they met the erosion criteria? |
In AUTODYN, if any nodes of a cell are on a defined polygon, that cell can't be eroded because it has to be there to define the polygon. If the cells are distorted severely, you will... |
######## WORK NOTES COPIED FROM INCIDENT 533085 ########
QUESTION:
Do you have an example demonstrating modeling procedure used to simulate the response of a structure to incident acoustic radiation?
ANSWER:
Please see zip file attached to this solution. This is an unsupported plate subjected to harmonically varying acoustic pressure source. A spherical volume of acoustic medium is subdivided so that you end up with a plane of nodes with uniform spacing (area apportioned to each node is a constant). F,,FLOW is used to define fluid 'flow' at each node in this immersed surface to make it serve as an acoustic radiation source. I believe this allows acoustic energy reflected off the plate to pass back through the acoustic source. I'm suspect that a 'hard' acoustic source defined with D,,PRES would not allow reflected radiation to pass through and so not accurately represent incoming radiation from a distant source.
Please note that FLOW on the F command is NOT what you think. From our knowledge base:
#230323 - Acoustics: use & physical significance of "FLOW" in FLUID29/30
From the acoustics tutorial DN-TO44:50 June 11, 1992:
Ff = -A*rho0*d2u/dt2 (eq 23, page 2-13)
In words, the "flow" at a boundary node equals the surface area associated with the node times the nominal density of the acoustic medium times the second derivative of mechanical displacement (acceleration). Flow is applied with the F command, label FLOW.
Physically, this is NOT flow (like FLOTRAN flow, that is, flow through a stationary Eulerian mesh or "control volumes"). Instead, it represents the motion of the fluid associated with acoustic waves propagating through it. I think.
The plate is modeled w/SOLID45. Infinite acoustic boundary elements (FLUID130) are superimposed on the sperical boundary of the fluid domain to absorb outgoing acoustic radiation. |
######## WORK NOTES COPIED FROM INCIDENT 533085 ########<br><br><br>QUESTION:<br><br>Do you have an example demonstrating modeling procedure used to simulate the... |
Why does the customer receive a FLEXlm error message when trying to release a hung license using the ANSLIC_ADMIN Utility?
The following message was displayed after entering the correct information for the fields in ANSLIC_ADMIN Utility:
Feature/User/Host/Display not found
Child process exited abnormally |
Lmremove is not valid for Workbench. This error occurs only when trying to release a hung Workbench license. You will not encounter this error with ANSYS.
|
Q. The section property method disregards the refT parameter for thermal expansion on a layer
by layer basis. Instead it uses the refT of whatever material that happens to match the MAT
number at the time of element creation (or the specified or default Tref if refT is not defined).
A. This was a code defect (#49755) at 10A1. It has been corrected at 11.0. |
Q. The section property method disregards the refT parameter for thermal expansion on a layer<br>by layer basis. Instead it uses the refT of whatever material that happens to match the... |
Class 3 Error
SYMPTOM:
Sinks of mass for certain multi-component flows give incorrect answers. There are two types of set-up that don't work correctly:
1) If you have source points, mass sinks are always at local composition and temperature. It works correctly for subdomains and boundary sources.
2) If you have combustion scalars defining composition, mass sinks are always at local mixture fraction mean, variance and reaction progress.
PROBLEM:
This is due to a bug in the CFX Solver in setting the multi-component flow |
WORKAROUND:<br>There is no workaround.<br><br>FIXED IN:<br>Release 11.0.
|
Q. When user tries to create a 3D rigid-flex, node-surface contact pair between a beam188/9 tip and a 3D surface area, the wizard creates a 2D node-to-surface instead, meshing only the lines attached to the area designated as the rigid target. This only occurs with beam elements. What is the problem?
A. This is a confirmed defect in contact wizard programming (ONYX Defect #52298). Since no more development and error corrections are pursued in ANSYS Classic contact manager, the defect has been dispositioned as not worth fixing and closed. The workaround is to create the node-to-surface contact pair thru APDL commands. |
Q. When user tries to create a 3D rigid-flex, node-surface contact pair between a beam188/9 tip and a 3D surface area, the wizard creates a 2D node-to-surface instead, meshing only the lines... |
Will ANSYS 10.0 SP1 for the EM64T and Opteron run on an AMD Athlon 64? |
We have not done any testing here in office, but ANSYS and Workbench should run on the ATHLON 64.
|
I ran a case with low inlet velocity (200FPM) and no heat sink. TASPCB ran successfully and produced good results. However, when I added a heat sink on an IC, the air and junction temperature become unrealistically large. What could be the problem? |
Most of the time this is a convergence problem. Always check results for consistency and accuracy.<br>(1) If the solution variables, velocity and pressure, are not reasonable, change the... |
Class 3 Error
SYMPTOM:
There can be a problem of conservation of chemical elements when using CFX RIF. Only the solutions near chemical equilibrium (Chi=0) seem to be affected.
PROBLEM:
This is due to a bug in the code received from Advance |
WORKAROUND:<br>To work around, you can let the solver ignore the lowest Chi values in the flamelet library by adding the "Skip Over Scalar Dissipation Rate Samples" parameter... |
My CFX 10 installation is on a Solaris machine. When I try to exit from CFX-Post I can get one of two error messages, either:
CUEProtocol::receive: Read error: no token on socket 13
or:
An error has occurred in cfx5post:
/apps1/ansys/ansys_inc/CFX-10.0/bin/solaris/PostGui_ogl.exe was interrupted by signal PIPE (13)
This is only on exit and the session has worked fine up to that point.
If I then re-run CFX-Post it exits normally. |
<br>This is a known problem with CFX-Post 10 which only happens on Solaris machines. The session itself works normally but just fails to exit.
|
After installing the unofficial (solver only) CFX-10.0 service pack 2 (SP2), the solver fails with an error message in the output file similar to:
An error has occurred in cfx5solve:
The CFX-5 solver exited with return code 128. No results file has been created. |
<br><br>This failure occurs because CFX-10.0 SP2 has been installed without first installing SP1. The problem can be resolved as follows:<br><br>- Download and install... |
Compilation of a UserFortran File gives the following error message:
User@Fortran > cfx5mkext test.F
/usr/bin/g77 -fpic -I/usr/cluster/CFX-10.0/include -o linux//test.o -c test.F
test.F:1: #include expects "fname" or <fname>
test.F:96: #include expects "fname" or <fname>
test.F:97: #include expects "fname" or <fname> An error has occurred in cfx5mkext:
/usr/bin/g77 exited with return code 1. |
the Version of the gcc Compiler is not up to date. Please, upgrade your gcc package to the actual version.
|
Q. What is the correct way to get the heat flux out of a surface in a 2D axisymmetric radiation problem using the radiosity solver?
A
The best way to get information such as that for a radiosity problem is to use the SURF251 and SURF252 (for 3D) elements. They overlay the radiosity surfaces.
These elements have two primary functions:
1) Provide more output for the user through the use of ETABLE.
2) Allow the user to coarsen the mesh used for the radiosity surfaces. This is referred to in the documentation as decimation.
You can create these elements merely by issuing RSURF just before SOLVE. You need not define the element type before hand. |
Q. What is the correct way to get the heat flux out of a surface in a 2D axisymmetric radiation problem using the radiosity solver?<br><br>A<br>The best way to get information... |
How do I make/prepare an ANSYS mesh for import into CFX Pre?
************************************************************** |
<br><br>If you have an ANSYS db file already you can convert it to a cdb file in ANSYS by:<br><br>1. Opening the ANSYS database in ANSYS.<br>2. Issuing the ALLSEL... |
What conditions lead to a message of finer mesh required near center line for elements n? |
I talked with the developer that coded the ¿finer mesh required¿ message and ran some test cases. The check is only done for non-structural axisymmetric element types that do not... |
User tried to run ANSYS on a Solaris server using sygwin + ssh from a PC.
The GUI could not start because it was receiving the following error.
--------------------------------------------------------------
ERROR MESSAGE RECEIVED:
00204534 VERSION=SUN-USIII RELEASE= 10.0A1 UP20060105
CURRENT JOBNAME=file 13:31:15 SEP 21, 2006 CP= 1.220
/SHOW SET WITH DRIVER NAME= X11 , RASTER MODE, GRAPHIC PLANES = 8
RUN SETUP PROCEDURE FROM FILE= /u10/ansys_inc/v100/ansys/apdl/start100.ans
X Error of failed request: BadAtom (invalid Atom parameter)
Major opcode of failed request: 18 (X_ChangeProperty)
Atom id in failed request: 0xf1
Serial number of failed request: 13
Current serial number in output stream: 16
What is the cause of this error? |
Using "trusted" X11 forwarding (ssh -Y) appears to fix the problem.<br>The ForwardX11Trusted=yes directive may be specified in ~/.ssh/config as an alternative to specifying the... |
How to remove a material from the favorites list in engineering data library? |
To remove a material from the favorites:<br> 1) Import the material into the engineering data project. It will then appear under the Materials folder.<br> 2) Right click on the... |
Q. The solution for a SHELL132 analysis seems to be dependent on its position in 3-D space. If you change the value of the parameter ANGLE in the APDL input below, you get a different solution. This should not be. What is wrong?
A. This is a Class 3 error at 10.0A1. See defect report #44755.
It has been corrected for the 11.0 release.
*****************
finish
/clear
angle=0
/prep7
et,1,132
sectype,1,shell
secdata,0.01,1
mp,kxx,1,50
wprota,,angle
rectng,0,1,0,1
esize,1
amesh,1
!
/solu
nsel,s,loc,x,0
d,all,tbot,10
nsel,all
nsel,s,loc,x,1
esln
sfe,all,4,hflux,,10
allsel
solve
!
/post1
set,last
plnsol,temp
nsel,s,loc,x
fsum
allsel
!************* |
Q. The solution for a SHELL132 analysis seems to be dependent on its position in 3-D space. If you change the value of the parameter ANGLE in the APDL input below, you get a different solution. ... |
Class 3 Error
SYMPTOM:
If the methane air flamelet library CH4_Air_Tf298K_Tox298K_p1bar.fll is used, then there is a 4% error in the species mass balance (C, H, O and N), which leads to an error in the flue gas composition.
PROBLEM:
This is due to a bug in the formulation of the flam |
WORKAROUND:<br>There is no workaround.<br><br>FIXED IN:<br>There is no fix yet.
|
Class 3 Error
SYMPTOM:
In a MFX calculation, if the CFX initial time is less than the MFX initial time, CFX executes a sequence of multiple time steps during the first stagger iteration of a new simulation. These time steps begin at the CFX start time and stop at the end of the first MFX time step, and use only the first set of data/loads sent from ANSYS to CFX. Once CFX has reached the end of the first MFX time step for the first time, it is synchronized with ANSYS and they complete the required number of stagger iterations. Results generated by CFX before synchromisation are incorrect.
PROBLEM:
This is due to the CFX Solver not respecting the MFX initial time for all runs se |
WORKAROUND:<br>There is no workaround.<br><br>FIXED IN:<br>Release 11.0.
|
Why does my CFX job take many hours to partition when other jobs of similar size have partitioned in a few minutes? |
<br>In some cases with many or large General Grid Interfaces (GGIs), the partitioner finds it difficult to predict the amount of memory that will be required to complete the partitioning.... |
SYMPTOM:
When visualising the variables Isothermal Compressibility and Isentropic Compressibilty in CFX Post, the units and values are not what is expected.
PROBLEM:
The standard text book definitions of these quantities are:
Isothermal Compressibility = -(1/rho)*(drho/dp)_T
Isentropic Compressibility = -(1/rho)*(drho/dp)_s
where _T and _s denote constant Temperature and Entropy respectively. However, in CFX, these variables are defined as follows:
Isothermal Compressibility = (drho/dp)_T
Isentropic Compressibility = (drho/dp)_s
Thus both quantities in CFX will differ from the standard definitions by a factor of -(1/rho). |
WORKAROUND:<br>For visualisation purposes, create new scalar User Variables in CFX Post using the above formulae. For use in the CFX Solver, create new scalar Algebraic Additional Variables... |
How to save parameter names and types into other arrays? |
The input file below starts with creating several parameters of various types then the<br>commands that save all the parameter names and types into two arrays. If you only<br>want to... |
Is there a way to get ANSYS support for the new Sun platform with the Opteron processor running Solaris 10? |
AMD Opteron machines running Solaris 10 operating systems are not yet supported with ANSYS products.<br>We are negotiating with them, but nothing has been finalized yet.
|
Will ANSYS run on an IBM POWER5 server (e.g. p5-550) under Linux (either RedHat or Suse)? |
We do not have a Linux port for IBM Power systems. The only Linux ports are for Opteron64, Itanium, Xeon/Pentium machines.
|