Discussion Groups

Simulation Mechanical and Multiphysics

Reply
Employee
zhuangs
Posts: 238
Registered: ‎10-04-2010

Re: Failed analysis

07-17-2012 07:36 AM in reply to: ramses061

I will investigate it today.  Thanks.

 

Shoubing

Please use plain text.
Valued Contributor
ramses061
Posts: 61
Registered: ‎05-04-2011

Re: Failed analysis

07-17-2012 07:41 AM in reply to: zhuangs

I really thank you very much. I only want to take my results from start to the end. I attached here my load curve. I want the calculate the stresses at the spesific times.

Please use plain text.
Employee
zhuangs
Posts: 238
Registered: ‎10-04-2010

Re: Failed analysis

07-17-2012 08:08 AM in reply to: ramses061

There is only one surface force in the model, and there is no constraint on he head.  Is it right?

 

Moreover, you don't want to consider 'Gravity", right?

 

Shoubing

Please use plain text.
Valued Contributor
ramses061
Posts: 61
Registered: ‎05-04-2011

Re: Failed analysis

07-17-2012 08:14 AM in reply to: zhuangs

Yes. Only 1 surface load and no constraints on the head. I want to apply a dynamic force to the head. I attached the load curve. The force begins to hit at the time 0,00207 s. And i started to run the calculation at that time to decrease the simulation duration.

Please use plain text.
Valued Contributor
ramses061
Posts: 61
Registered: ‎05-04-2011

Re: Failed analysis

07-17-2012 08:14 AM in reply to: ramses061

And i don't want to to consider gravity.

Please use plain text.
Employee
zhuangs
Posts: 238
Registered: ‎10-04-2010

Re: Failed analysis

07-18-2012 11:24 AM in reply to: zhuangs

The warning message "** Warning: element   3091 in part   7: determinant =-.26223249E-01<= DETTOL!" is related to the bad element shape.  If the element shape is too bad at the very beginning,somehow some deformation will result in too small deternimant.  Actually, the mesh in part 6 and part 7 is not good.  Especially in the impact area.  Finer mesh for part 6 and part 7 is prefered.  However, even the current mesh size make the model too large for my computer to handle easily.  You might to enlarge the mesh size the other parts which is not so important.

 

Shoubing

Please use plain text.
Valued Contributor
ramses061
Posts: 61
Registered: ‎05-04-2011

Re: Failed analysis

07-18-2012 02:00 PM in reply to: ramses061
Thanks for your interest,
But today i suppressed both parts 6 and 7 and i got the same error in parts 8 and 9. I think there is a convergence problem. The time steps are too small. I tried to set a constant time step but that time the program ended with no result.
Please use plain text.
Employee
zhuangs
Posts: 238
Registered: ‎10-04-2010

Re: Failed analysis

07-19-2012 07:20 AM in reply to: ramses061

Well.  It seems that this kind of warning message happens after a convergent step, and the following step has a half step size.  In this case, the warning message might be related to the trial solution based on the the last convergent step.  And the code automatically cut time step.

 

Shoubing

Please use plain text.
Valued Contributor
ramses061
Posts: 61
Registered: ‎05-04-2011

Re: Failed analysis

07-19-2012 09:26 AM in reply to: zhuangs

So what do you suggest?

Please use plain text.
Employee
zhuangs
Posts: 238
Registered: ‎10-04-2010

Re: Failed analysis

07-23-2012 02:09 PM in reply to: ramses061

The model should be symmetric.  We can use a symmetric model with finer mesh to obtain better elemet shape and more stable running.

Please use plain text.