Why is PML being error-trapped in a modal analysis at ANSYS 11.0? |
That error trap was added because modal results with PML were possibly incorrect and the runs sometimes failed. The PML is like a form of high damping, so doing an undamped modal on it is not... |
What is effect of SED in a rocking spectrum? What are units of rocking spectrum? |
At 11.0, we improved the documentation of the Rocking Spectrum in the Theory Manual. Theory Manual Section 17.7.5 now clearly indicates that the rocking spectrum is added to the un-normalized SED... |
Why do thick pipe elements give lower than expected stress for internal pressure? |
The stress is only available at the outer radius where it is lower for internal pressure (but maximum for moment loading).
|
Why is an unbalanced rotordynamics load with SYNCHRO on solid elements wrong? The customer has a 3D solid model on which he is performing a harmonic analysis with the SYNCHRO command. He notes that changing the rotational velocity value on the CMOMEGA command changes his results when for this analysis type, CMOMEGA input is documented to only be used for defining the axis of rotation. |
This is an error reported in defect 62620 and corrected in 12.0. For solid elements, the CMOMEGA input in SYNCHRO analysis is being used as a centrifugal force and not as a specification of the... |
Can EMAG submodeling be done with the DSP/GSP formulations? |
Possibly not due to defect 15582, which reports that DSP/GSP solutions are incorrect for a boundary condition utilizing non-zero MAG values, which would be the case in a... |
Why does the presence of circuit elements in a harmonic analysis cause slow/incorrect solutions? |
Circuit elements can create ill-conditioned matrices. This and similar problems with ill-conditioned models, such as occur in acoustic-structural harmonic analyses are addressed in the resolution... |
Can table loads be used in a mode-superposition transient? Do you have an example? |
Yes, here is an input file showing the mode-sup transient procedure. It just uses 5 time points, but the procedure would be the same for any number of time points. The only difference would be... |
Can a PSD analysis be done on in Simulation with a model supported with body-to-ground joints? Currently the PSD in Simulation seems to work with a fixed support only. The user has a model that is supported with various spherical joints to ground. |
No, in ANSYS 11.0 the PSD method utilizes unit displacements at the fixed supports and that isn't possible when the model is supported on joints. To run a PSD analysis it would be... |
Can the ADAMS macro be modified to use "/ASSIGN,,,RST"? The ADAMS macro uses the MNF command to generate the .mnf file. I think MNF is also a macro, with the POST1 FILE command hard-coded to look for genCMS.rst in the working directory. As a result, it fails to read a results file written to a different drive with the /ASSIGN command.
I was able to workaround the problem by inserting *ABBR,file,stat just before the MNF command in the macro. Could you please have MNF fixed so it's not hard-coded to look for the .rst file in the working directory? |
Feature request 62282 created.
|
Why is a PSD analysis failing on this large model in ANSYS 11.0? |
For large models, Simulation is using the LANPCG modal solver with MSAVE,ON. This solver option, greatly reduces the memory needed but doesn't support a subsequent PSD... |
Is it possible to perform a harmonic piezoelectric analysis in English units? |
Yes. The input below is a modification of VM231 using a consistent system based on feet, pounds, seconds and Coulomb. A unit conversion table is attached to this solution.<br><br>The... |
How can I get a response surface for a second random output variable with RSFIT? I tried reissuing the RSFIT command with a new name and it wasn't permitted. |
The procedure for obtaining multiple response surfaces is to add new output variables to the existing set of response surfaces using the same name as shown below. The RSlab name, freq_RS1, can be... |
How is the stress at a probe location obtained in Simulation 10.0? You can insert a probe in Simulation 10.0 in various ways; either by clicking on an arbitrary location with your mouse, clicking on a vertex or defining it based on the origin of a coordinate system. When you solve your problem afterwards, how will the probes follow the deformation and how are the results evaluated at the probes?
If you define your probe on an arbitrary position, for instance between two nodes in a two dimensional model, will the probe follow the deformation of the 2 nodes on each side and will the results for the probe be an interpolation between the node results on each side of the probe? Or is the probe position fixed in space so that the result will be more and more inaccurate as the deformation is increased?
I have done some tests and my conclusion is that if you define your probe to a vertex the result will be evaluated on the vertex since WB always puts a node to a vertex. So when the node is displaced, the probe will give a correct result for the displaced node location. |
As you observed, the stress at a vertex will be the stress at a node and will not be fixed in space. The stress at a coordinate system will use an interpolated value based on the original... |
Why is ADAMS interface failing with message about scalar DOF? |
One of your element types is a BEAM189 with KEYOP(1)=1 activating the WARP degree of freedom. The program seems to treat the WARP DOF as scalar.
|
Why does this transient thermal analysis undershoot the temperature boundary conditions? The model is given a 28 degC initial condition and has convection to negative 140 degC and convection to 28 degC. When it is solved, there are regions that reach temperatures of over 1000 degC. |
It is quite common for thermal models to overshoot or undershoot the boundary conditions. In a transient analysis, the most common reason is that mid-side nodes have been used and the maximum... |
Following a PSD/random vibration analysis, how can I get the contribution made to the 1-sigma value by just one mode? |
It would not include the contribution the mode makes through coupling with other modes, but to see the contribution from mode 7 itself the commands would be:<br><br>!!! Following the... |
Can LSWRITE/LSSOLVE be used in a static cyclic symmetry analysis in ANSYS 11.0? When I issue LSSOLVE,1,2 I get rigid body motion for load step two as if the boundary conditions have been lost. If I run the two load cases separately, the results are ok. |
Yes, the displacement constraints that are created in the first load step for the duplicate sector are not being reapplied in the 2nd load step. Defect 62765 has been resolved by documenting in... |
Why is pressure load stiffness in ANSYS 11.0 not being used for this SHELL181 model? SHELL181 is supposed to use pressure load. |
You may have applied the prestress as an edge load. In Table 2.9 of the Theory Manual Section 2.15, there is the note below referring to SHELL181:<br><br>"x[3] = Same as x, but... |
Why am I getting a warning message that a COMBI214 is not in the plane indicated by its KEYOPT(2)? |
Since the COMBI214 (bearing element) orientation is defined by its KEYOPT(2) and not the ij direction as is common for other line elements, this warning message is issued when the ij direction is... |
Do you have an example using multiple CMS superelements? |
Here is an input file that shows the CMS procedure on a very small... |
Why can't a Simulation-11.0 model with a force at a vertex use DANSYS? |
Simulation creates a follower element (FOLLW201) to apply the force at a vertex. This element isn't supported by DANSYS. Feature Request 62409 asks that this limitation be removed in... |
After a harmonic analysis, why doesn't the POST26 value for the amplitude of S1 match the POST1 value computed with the HRCPLX command? |
The HRCPLX command is a macro which does an SRSS load case combination to obtain the amplitude of a complex solution. By default, the combination is done on the component stresses and derived... |
Why am I getting the error below when I try to abort an analysis using the jobname.abt file in ANSYS 11.0?
***** process error reported by subroutine xdslfa
see xdslfa abstract (ier = -598)
*** ERROR *** SUPPRESSED MESSAGE CP = 1296.422 TIME= 11:31:47
Sparse solver was aborted by jobname.abt or STOP button pressed. |
In ANSYS 11.0, we put abort logic into the Sparse solver based on the same NONLINEAR keyword that is used by jobname.abt. This is defect 61295 and is corrected in 12.0. A workaround is to use... |
Is there a problem size limit of 2 Gb in the ADAMS Interface on Win64? It seems so, based on the test below. A model with the following statistics gives odd messages. Note that 428,555 nodes x 3dof x (20+6x15 modes) x 16 bytes is slightly greater than 2 Gb.
Nodes = 428,555
Elements = 285,026
Interface points = 15
Number of Model to extract = 20
Message obtained:
*** WARNING *** CP = 2942.188 TIME= 18:32:06
Couldn't seek to mode 105 node 385050position in temporary file.
*** WARNING *** CP= 2942.250 TIME= 18:32:07
Mnf_set_mode_shape_vec error [mnf_writeCMS.3130] |
This is defect 63637. Until recently, Windows (or more specifically, Visual Studio that we use to build ANSYS) did not natively support 64-bit file I/O from C *including* Win64, so the MNF... |
Is there a way to re-use the reduced matrices so solve 2 or more loading histories in a reduced transient analysis in ANSYS 11.0? |
No, but a good workaround would be to use a CMS Superelement to do the matrix reduction and that could be re-used in a full transient. An advantage of this approach is that a CMS Superelement... |