Discussion Groups

Simulation Mechanical and Multiphysics

Active Member
6 Posts
0 Kudos
Registered: ‎09-26-2012
Accepted Solution

Large loads, plastic deformation and convergence

727 Views, 4 Replies
09-26-2012 12:18 PM

I am working on a ROPS model which requires plastic deformation to meet energy specs. The analysis quits after some materials are beyond UTS. (I know that a material is has fully failed in tension at UTS.)


For my confidence in the analysis, I want to know if it is not converging because of my modeling, or because of large displacements due to plasticity.


Will all analyses complete even if materials are beyond UTS? Does Simulation even know what the UTS is?


In a perfect model, will all analyses complete even with near infinite loads?

4 Posts
0 Kudos
Registered: ‎01-24-2012

Re: Large loads, plastic deformation and convergence

09-28-2012 05:18 AM in reply to: lesmfunk

What material model are you using ?

Active Member
6 Posts
0 Kudos
Registered: ‎09-26-2012

Re: Large loads, plastic deformation and convergence

09-28-2012 08:07 AM in reply to: algor_neil

I am using MES with non-linear materials. All parts are shell elements for sheet, plate, and formed steel.


Incremental displacements at last solved step were more than 5 times element size for some elements. Strains are also very high as a result.


Just wondering what happens when you add an absurdly high force. I don't know how Simulation is supposed to act in a material failure and whether it can even detect such an event. Or, is it left up to my interpretation when an analysis quits.


I just want confirmation before I tell my client that their design will fail a physical test.

*Expert Elite*
596 Posts
129 Kudos
Registered: ‎08-30-2012

Re: Large loads, plastic deformation and convergence

10-01-2012 10:27 AM in reply to: lesmfunk

Hi Les,


Let me elaborate on algor_neil's question "what material model are you using?". The material model is a setting under the part's "Element Definition" dialog. In short, the material model describes the part's stress-strain curve. I am guessing that the material model in your simulation is the default (isotropic) which means the stress-strain curve follows the modulus of elasticity (E) all the way to infinity. Since your parts are experiencing plastic deformation, you may want to use one of the "von Mises with ___ hardening" material models to account for the plastic deformation (or whatever your energy specification says is allowable.)


To answer your question "Does Simulation even know what the UTS is?", you can answer that question by viewing the material properties for the part. The analysis only knows about the material properties listed for each part. (Do not look at "Tools > Options > Manage Material Library". That shows the complete properties in the library, not the properties assigned to the part and used in the analysis.)


It is difficult to say what reason causes an MES analysis to quit because it cannot converge; that is, is the problem related to the mesh, to the model, to the input, or something else? But it is theoretically possible for a "perfect" model to converge even with unrealistic loads. It probably does not happen very often because the elements become distored which will cause a problem with convergence, or the time step to resolve a change in acceleration etc would need to be so small that the analysis just gives up ("AutoTM over limit").


John Holtz, PE
Mechanical Engineer
Pittsburgh, PA

Simulation Mechanical user since Dec 1997
Active Member
6 Posts
0 Kudos
Registered: ‎09-26-2012

Re: Large loads, plastic deformation and convergence

10-01-2012 11:14 AM in reply to: AstroJohnPE



Thank-you for your reply.


I am using von Mises with isotropic hardening.


The reason I asked if Simulation recognizes the UTS, is because it doesn't show up in that dialog you talk about even though I had to include it in the Material Library to calculate the Strain Hardening Modulus. Thank-you for confirming that.


I forgot to mention that it does reach AutoTM limit.


It looks like I am going to have to make a judgement call. My force (time step) vs displacement curve goes almost flat in the plastic range before it quits. The curve does not have any anomolies. I believe it isn't converging because, when it goes plastic, the deformation is so large that it can't compute the "acceleration" as you suggest.


Thanks John.


I will mark this post solved.