Tutorial OS-T 1600

Hi @N_R could you please share the entire .out file with us?

Hi @Prakash Pagadala
Yes, of course. I hope it is the following:

If it is not the right one, please tell me.
Thank you.

Unable to find an attachment - read this blog

Hi @Prakash Pagadala

everything alright. I changed a few parameter in Optistruct and it seems to work.
I missunderstand an instruction in the tutorial.

Thank you nonetheless.

@N_R

Thank you for the update.

I am running the same model on my end and so far no issues were seen. 

@Prakash Pagadala
 

Hello Prakash,

I tried to built up an own FSI Problem and i got the same error as written April, 23.
What does this error generally means? What´s the problem?

*** ERROR # 3423 ***
 MUMPS solver encountered error with code -6.
 please contact support and provide .out and .stat files.
 

 *** INTERNAL PROGRAMMING ERROR ***
   in file 'nltranend.F', at location # 364.
 
 ****  ABORTING RUN DUE TO AN INTERNAL ERROR  ****

Thank you in advanve.

Unable to find an attachment - read this blog

Hi @N_R

Can you run a static analysis and share the results from the same?

Hi @Prakash Pagadala,

everything again alright.

Thank you very much for your effort.

Hi

I want to do tutorial OS-T 1600 FSI but in creating contact surface, when I click on '0 elements > elements' I cannot pick all the faces of the beam. When I select all the faces nothing happens and if I click on add, the message 'No new elements have been selected' appears. Could you please let me know how to select all the faces? Do I need to have 'all shell faces' or 'all solid faces'?

In fact I managed the problem. It needs to have 'all solid faces' to select elements.

Hi,

Can you change to 'add solid faces' and check if that helps?

<?xml version="1.0" encoding="UTF-8"?>

Hi Prakash

Could you please look at attached error file and let me know how to manage the problem? There are several issues need to be fixed.

Thank you.

Unable to find an attachment - read this blog

I think this is the problem...

*** WARNING # 9113
 It appears that the directory used for scratch files:
    /home/ac7754/fsi_slab_1/
 is located on remote filesystem. 
 This could result in very slow solver performance.
   Number of CPU processors        :         32

If possible share the model file to debug...

Hi Prakash

Please find the model file in attachment.

Thanks & regards

Unable to find an attachment - read this blog

Hi Prakash

Do you have a copy of FSI_AS_MANIFOLD.inp and FSI_OS_MANIFOLD.fem, please?

Thank you.

Hi Prakash

Please find the model file in attachment.

Thanks & regards

I am able to reproduce the same but with a different warning message, I will get back to you with more details ASAP

Unable to find an attachment - read this blog

Hi Prakash

Do you have a copy of FSI_AS_MANIFOLD.inp and FSI_OS_MANIFOLD.fem<span></span>, please?

Thank you.

Can you give me the reference to the document about the files?

Hi Prakash

I found the files. Thank you.

Hi Prakash

Are you running 'Slab.fem' on one single PC, please? When I did it there is a crash in simulation.

Thank you.

Hi Prakash

Is it possible to look at attached file and let me know how to manage the problem, please?

It is OS-T: 1610 Thermal Fluid-Structure Interaction Analysis on a Manifold.

Thank you.

 ****  ABORTING RUN DUE TO UNEXPECTED ERROR CONDITION  ****
  
Error output from the solver:
==============================

External Code via CCI: *** ASSERTION in Function <cciNew> File <linux64/src/cci.c> Line <152>
External Code via CCI: *** Did not receive connection request from client after waiting for 2400 sec

hwsolver.mesg output from the solver:
==============================
 ****  ABORTING RUN DUE TO UNEXPECTED ERROR CONDITION  ****
 

optistruct:: Solver run finished after 40 minutes.

Unable to find an attachment - read this blog

Hi @Question

I am checking with experts regarding the same. Looks like the FSI part has some problem. 

Hi Prakash

When I run both tutorials OS-T 1600 and 1610, there is the following error.

acuSolve: *** ASSERTION in Function <ecdSetFatalError> File <ecd.c> Line <4208>
acuSolve: *** unable to match with external surface at 7 quadrature points
acuRun: *** ERROR: error occurred executing acuSolve
 

Could you please let me know how to manage it?

Thank you.

Hi,

I have shared test files with experts. I will share the feedback ASAP

Hi Prakash

I want to run an implicit simulation on a model with around 6 million elements. I set up the simulation as plate model in implicit analysis. However there are the following errors in my run. How can I solve that problem, please?

 *** ERROR # 4002 ***
 Running RADIOSS engine, 1 ERROR and 0 WARNINGs have been detected.
 
 ANALYSIS COMPLETED.
  
 *** SOME ERRORS WERE REPORTED DURING THE RUN,
 *** PLEASE VERIFY RESULTS AS THEY MAY BE INVALID.
 ************************************************************************
  A fatal error has occurred during computations:
  
  *** ERROR # 4002 ***
  Running RADIOSS engine, 1 ERROR and 0 WARNINGs have been detected.
  
  ANALYSIS COMPLETED.
  
  *** SOME ERRORS WERE REPORTED DURING THE RUN,
  *** PLEASE VERIFY RESULTS AS THEY MAY BE INVALID.
 
 *** PROGRAM STOPPED: FATAL ERROR(s) ENCOUNTERED.
 ************************************************************************
Errors reported during program execution

hwsolver.mesg output from the solver:
==============================
 
 *** ERROR # 4002 ***
 Running RADIOSS engine, 1 ERROR and 0 WARNINGs have been detected.
 
 ANALYSIS COMPLETED.
  
 *** SOME ERRORS WERE REPORTED DURING THE RUN,
 *** PLEASE VERIFY RESULTS AS THEY MAY BE INVALID.
 
 *** PROGRAM STOPPED: FATAL ERROR(s) ENCOUNTERED.

Solver FAILED

optistruct:: Solver run finished after 2 hours (10117 seconds).

Thank you.

Hi,

What type of analysis is it?

Can you try with NLSTAT with LGDISP instead?

Hi

It is a NLGEOM implicit static analysis as you did for a plate. I set up my model as the plate one. The plate model is in attachment.

Is NLSTAT with LGDISP different than NLGEOM one, please?

Thank you.

Unable to find an attachment - read this blog

Yes, NLGEOM uses RADIOSS block in the background. 

You can refer to this tutorial for NLSTAT large displacement OS-T: 1520 Finite Sliding of Rack and Pinion Gear Model

My model is a cylinder under internal pressure. The cylinder is fixed in three supports along the length and the material is steel. According to physics of the model, I think there is no large displacement in that model. Do you think it needs to be run with NLSTAT large displacement? How about running with nonlinear material simulation?

Hi,

If the problem doesn't involve any large displacement, you can just try with NLSTAT where materail nonlinearity will be taken care.

Regarding FSI models, I am waiting for experts view. But from .cci file I can see that the solver couldn't make a connection after 2400 secs

Hi

How can I have just NLSTAT without large displacement?

Thank you.

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)

Do I follow OS-T: 1520 Finite Sliding of Rack and Pinion Gear Model or OS-T: 1500 Nonlinear Implicit Analysis of Bending of a Plate?