Tutorial OS-T 1600


Hello,
i tried to solve the Optistruct OS-T 1600 Tutorial, but after starting the two solvers (optistruct, acusolve) there is a convergence error of acusolve that stops the job:
Messages for the job:
*** ERROR # 5202 ***
Acusolve stops before FSI iteration achieves convergence.
*** PROGRAM STOPPED: FATAL ERROR(s) ENCOUNTERED.
Does anyone knows the problem? I checked all steps of the tutorial and start it again, but the error still appears.
The next step was to change the exchange parameter of acusolve (min/max_stagger_iterations) and optistruct (min/maxEX) to 30. A Simulation of the Tutorial with only 0.02s worked.
To simulate the tutorial 0.1s or the whole 1s - as it is described in the Tutorial- still doesn´t work with a value of 100 for max_stagger and MAXEX. Another error appeared:
<?xml version="1.0" encoding="UTF-8"?>
Does anyone has got experience with the tutorial OS-T 1600 and solved it successfull?
Thank you very much in advance.
Find more posts tagged with


Hi
Do you mean the following tutorial? Its fem file is in attachment.
In my model, since the supports are not symmetric, there is no symmetry in the whole model. Therefore I have to model whole cylinder. Furthermore, there is no gasket so there are no contact surfaces.


Follow the same process shown in tutorial, but don't add PARAM>>LGDISP and in load case refer NLPARAM card in just NLPARM AND NOT TO NLPARM(LGDISP)
Could you please set up attached fem file according to your suggestions above?
Thank you.


The model is already good to go. I have increased the NINC in NLPARAM


Hi Prakash
Is it possible to have cyclic internal pressure in a cylinder then do implicit analysis on it, please? The internal pressure is a function of time.
Thank you.


Hi
How about having implicit analysis when there is a cyclic internal pressure? Is there any tutorial for it?


If I understood correctly, the pressure loading follows a cyclic frequency (time).
If this is the case you can use Nonlinear transient (implicit) analysis where the load is applied WRT time
Please refer to this post:


Can NL_DTRANS.fem be used for small displacement, as well? I need a dynamic implicit simulation for a cylinder under cyclic internal pressure. According to the physics of my model, there is no large displacement in the cylinder. However it can have nonlinearity in material which is steel.
Thank you.


Could you please look at attached fem file and let me know what the problem is?
Thank you.


HI,
Do you have a Static subcase with a DLOAD?
this is not compatible. Please use dynamic subcase to use DLOAD.


Hi Prakash
Could you please let me know how I can have a static subcase with a DLOAD? Furthermore, how may I use dynamic subcase to use DLOAD?
Thank you.


Hi,
DLOAD is used for dynamic analysis and can't be used for a static subcase.
You can use DLOAD for linear transient, NL transient, FRF analysis, for example.
Tutorials for the same are available on OptiStruct help. Please refer the same on how to set up.


Hi Prakash
Is there any way to have a better displacement contour for attached figure, please? There are some disturbations in contour. It is for a transient analysis. I would like to have a very smooth contour.
Thank you.




Hi Prakash
Even by using Simple Averaging method, still there is no smooth contour. What is the possible source of this problem and how to manage it, please? It happens when a model has high number of elements and specially in transient simulation that needs to do high number of iterations.
By the way, is there any news about the FSI tutorials error, please?
Thank you.


Hi,
Is this happening for all time steps in the transient analysis?
Regarding FSI, our experts are testing the models and I will update you as soon as I get any information from them.


Hi
Yes, it is happening for all time steps in the transient analysis. The model has more than five million elements and the simulation has been run on a cluster with several CPUs; it took a few hours to finish.
Thank you.


Hi Prakash
Is it possible to manage such problems in contour, please? Is it an error in simulation or not? What is the source of such problems?
Thank you.


Hi
Sorry for a late reply.
I think too fine mesh is causing this the problem,
Does the problem involve FSI?


Hi
You're welcome. Already it is just an implicit analysis but I want to do an FSI analysis later on it, as well.
If the source of problem is related to fine mesh, how can we find an optimal number of elements to have a smooth contour and accurate results, please?
Thank you.


Hi
From the last replies, looks like the problem is with too fine mesh.
There is no thumb rule for a better mesh. Maybe a mesh convergence study can help...?!!


Hi
How can a mesh convergence study be done in HW, please?
Thank you.


Right now there is no direct way. One has to change the mesh, analyze, record and study the mesh size Vs results to find out the mesh size.


Hi
Regarding the tutorial with FSI, Since this is a coupled Optistruct – Acusolve run, both need to be running at the same time to exchange information. So after you have started your Optistruct, start running Acusolve also as shown in the section ‘Submitting the AcuSolve job’.


Hi Prakash
Thank you for providing information about running the FSI tutorials. Could you please let me know how to run simultaneously both OptiStruct and AcuSolve?
Thank you.


Hi,
Start AcuSolve after submitting the OptiStruct run. That should help


Hi Prakash
Could you please let me know how to launch HyperWorksSolver Run Manager in HyperWorks 17.2?
I found it in Applications.
Thank you.


Hi,
I don't have access to tutorials right now. But you can search in form for information and examples on small displacement analysis
This is one tutorial I remember for NLSTAT small displacement analysis... OS-T: 1360 NLSTAT Analysis of Gasket Materials in Contact