Why is my ANSYS 8.0 electrostatic input with FMAGSUM not giving correct forces? The documentation does recommend EMFT for calculating these results, but FMAGSUM used to work fine in version 8.1 and there's nothing stating that it should no longer be used. |
Use EMFT with the nodes and attached elements selected for the body you want to calculate the force on. The FMAGSUM documentation has been corrected in 12.0.
|
What is the minimum OUTRES setting to obtain reaction forces at CERIG master nodes? OUTRES,RSOL,1 isn't sufficient and RSOL,NLOAD,1 increases the results file by 10x. |
This was corrected by the resolution of Defect 60140. In uncorrected versions the following is the minimum setting to get the reaction... |
What is the minimum OUTRES setting to obtain reaction forces at CERIG master nodes? I successfully solved a harmonic response problem and contour plots of displacement can be produced in about 6 seconds for this particular model. However, contour plots for stress, even those that have been plotted before, take about 90 seconds to generate. |
This problem was recently identified (Defect 60057) and will be corrected in Simulation at 11.0 SP1. There have also been reports of very slow frequency response plots of... |
Why did FEModeler smooth all corners to create "Initial Geometry?" |
It appears you have set the property "Tolerance Angle" too high. This is accessed via the tree button "Skin Detection Tool." This sets the maximum angle between... |
Why is MPC Contact affecting the load vector applied with LVSCALE in ANSYS 11.0? The model has MPC bonded contact and it is not set to ignore gaps/penetration. |
When MPC contact generates constraint equations, they can include constant terms. These terms wind up on the right hand side of the system equation, so they get scaled by LVSCALE. Defect 59983... |
Why can't I expand my modal cyclic symmetry solution in a model with SHELL41? When "/CYCEXPAND,,ON", a stress listing has both the real and imaginary results (without the usual warning about unprocessed data). The listing of displacements is correct, and the plotting of stress seems to be correct as well. This also occurs with SHELL181 KEYOPT(1)=1. It does NOT occur with SHELL63 KEYOPT(1)=1 or SHELL181 with KEYOPT(1)=0. |
The existence of membrane shells turns out to be one of the many cases where PowerGraphics is turned off without notice, which then turns off /CYCEXPAND. Defect 59982 requests that membrane... |
What does AVSMOOTH do in the command "TINTP,,,,,,,,,AVSMOOTH"? Using AVSMOOTH causes ANSYS 11.0 to differ from a competitor's solution. |
The Newmark accelerations can be noisy for many problems. The Newmark assumption uses the average acceleration [(a(t) + a(t+dt))/2] in its calculation. Initial conditions and nonlinearities can... |
How can I obtain harmonic forces with SOLID117? |
Only the JXB forces in conducting regions are available. For these forces, ETABLE items FMAG,{x|y|z} or NMISC,{4|5|6|7} can be used. Feature Request 59853 has been file to ask for more complete... |
Why doesn't the /CYCEXPAND display of modal cyclic symmetry results with the phase angle given by CYCPHASE match the CYCPHASE max values? |
CYCPHASE uses Full Graphics to determine the maximum value. /CYCEXPAND uses PowerGraphics, which is known to give slightly different values of stress, particularly at stress concentrations. This... |
Why is AL command requiring so much memory in ANSYS 11.0? See an example below.
fini
/clear
/PREP7
k,1
numstr,kp,100000
circle,1,10
al,all
It results in an error:
"This Model requires more scratch space...."
"This error is non-recoverable..."
The same script in 10.0 runs without any problem. |
This error, defect 57808, is fixed in 11.0 SP1. A programming error introduced at 11.0 changed the memory allocation for the AL command to be based on the square of the maximum keypoint number,... |
Does the theory behind the ADAMS interface require interface nodes to have 6 DOF? In I-DEAS, they can specify which DOF are active at each interface point, thus reducing the number of constraint modes needed. So for their I-DEAS run, they have 66 constraint modes plus 50 normal modes, and the analysis takes about 8 hours. In contrast, trying to output the .mnf file in ANSYS, they have 120 constraint modes and are only requesting 10 normal modes, but after 24 hours it is only up to the 30th mode.
Part of the problem is the sparse solver is running out-of-core, but I wanted to find out if in ANSYS you can specify which DOF are active at interface points and thus reduce the total number of constraint modes that must be calculated? |
I have seen a customer-modified version of the ADAMS macro which selectively defined the master DOF for each node, so I think it is possible but it does require modifying the macro.
|
How can I specify the load step in which I want to apply a CFX-Pressure load? For example, if I have several load steps in my analysis, how can I apply pressure from CFX to just one load step? Is there any direct way in Simulation 11.0? |
Once you apply the CFX load, it should be there for each of the load steps by default. You could suppress it if you want to for any particular load step. Note that if you wanted a load from a... |
Is it possible to do a Spectrum analysis in 11.0 Simulation with a template? The 10.0 template doesn't work anymore. |
The change in Simulation at 11.0 to do postprocessing directly from the RST file, instead of from an XML retrieval, has made postprocessing snippets more difficult; however, there is a way to make... |
Can ANSYS solve this plate buckling problem accurately? The following explains a simple test of buckling analysis of a plate in ANSYS.
Plate parameters:
Size: b = 112.8 in, a = 720 in.
Thickness: h = 1 in
Young's Modulus: E = 29000 ksi
Poisson's Ratio: nu = 0.3
Yield strength: fy = 50 ksi
Boundary condition: simply supported (moment free)
Our equation for the buckling capacity with
a/b = 720/112.8 = 6.4, k=3.29, gives Fcr = 8.24 ksi. |
The input below file is for the buckling solution for all edges simply supported. Using your equation the buckling stress is 8240psi or 8.24ksi.<br>The ANSYS solution using the file is... |
**** Entered By: dlooman @ 12/28/2007 03:33 PM ****
Q1: Simulation 11.0 is writing the Stefan Boltzman constant at 3.08e-11. What are the units for this number?
Q2: Why is the default value in ANSYS in BTU/hr.in^2.R^4 and not BTU/s.in^2.R^4?
Q3: How can the default value be converted to the value of 3.08E-11?
Unfortunately, the BIN unit system is not consistent between mechanical and thermal units.
A1: The units are in-lb/(sec-in^2-R^4)
A2: Internally we do use secs, but the Btuh is the most commonly used unit for heat rate.
A3: 1.199e-11 (Btu/(hr-in^2-R^4)*(9378.873in-lb/Btu)*(1hr/3600sec) = 3.1e-11 |
**** Entered By: dlooman @ 12/28/2007 03:33 PM ****<br><br>Q1: Simulation 11.0 is writing the Stefan Boltzman constant at 3.08e-11. What are the units for this number?<br>Q2:... |
Is the USUM value from a PSD analysis correct? I know there are certain derived quantities that should not be computed from the 1-sigma components (for example, s1). |
USUM is calculated from the 1-sigma components of displacement. It should be reasonably accurate because the calculation of USUM doesn't dependent on the sign of the components. It is... |
Why does my small modal analysis run slower in ANSYS 11.0? Here are some solution times for a SOLID185 model:
54K 248/262
63K 310/324
73K 395/420 (esize,.00550)
85K 503/2439! (esize,.00525)
89K 485/1222 (esize,.00500) 10.0 times for this model are 546/602) |
This was due to a change in the way I/O is done, which was not efficient for this model size. The problem is corrected in 11.0 SP1. (ref: defect 59253) Development provided a workaround: Set... |
Can Large-Def/Prestressed cyclic modal analysis have contact? Only one harmonic index is being solved. |
This is not supported in ANSYS 11.0 or 12.0. Defect 59036 was filed to have this documented and error-trapped in ANSYS 12.0.<br><br>It appears that one harmonic index can be solved... |
Do you have an example of piezoelectric [c] matrix input for a 2d case? I am having really hard time converting [c] matrix from 3D to 2D case. |
Here is an ANSYS 11.0 input file from the test set that may clarify the input.<br><br>/title, Thermal strain in a 2-D piezoelectric analysis<br>/com, -- Target: piezoelectric... |
Why is my rotordynamic model not giving a linear response in a harmonic analysis? The only load is an acceleration load. I expect the solution to vary proportionately to this acceleration value, but it doesn't. |
The reason for the non-proportional response is that in addition to the acceleration load you have a centrifugal load produced by CMOMEGA and that is constant.<br><br>It is probably... |
Why do solutions with radiation superelements requires so much memory? The customer has a radiation model consisting of a database and several *.sub files of up to 100 MB each (total 784 MB). There are 67,000 nodes. The model seems to need 3.5 GB of memory which is much more than expected. Is this an expected amount of memory for a radiation model? |
For historical reasons, 3 copies of the radiation matrices are stored in memory, which accounts for the unusually large memory requirement. The amount of memory required in ANSYS 12.0 is expected... |
Why is the in-plane reaction moment for SHELL63 almost zero? SHELL181 gives expected moment of 500. |
The in-plane rotational stiffness of a shell finite element is fictitious. Different shell elements resolve this in different ways. SHELL63 uses a weak spring to ground to avoid having zero... |
What is ANSYS policy on upward and backward compatibility of databases? I read a little in the manual in Chapter 13.2: Archiving Models, but I didn't find the information I was looking for. This is not just an ANSYS question but also an ANSYS Workbench question. |
ANSYS databases can be counted on to be resumed. In most cases, an ANSYS db file from as far back as 5.0 can be resumed. ANSYS inputs and CDB files are also typically very reliable. If there is... |
Why does this model produce many small pivot warnings? |
It would be more accurate to use the term "pivot ratio" than "pivot" because the values being reported are ratios of the smallest pivot to the largest pivot. The reason... |
Why am I getting a message about EKILL/EALIVE when I use LSWRITE in ANSYS 11.0? I am not using the birth-death feature. The complete message is below
*** ERROR ***
LSWRITE/LSSOLVE cannot be used with the birth-death option
(EKILL/EALIVE). Use the multiple SOLVE method instead. |
Defect 59548 was recently filed reporting that MPCHG generates this message at 11.0, when it did not at 10.0. This defect has been resolved by making the error message more inclusive in ANSYS... |