Learning Forum

[7121573]

Hello.

 

I always appreciate the valuable feedback I receive from everyone on this forum.

 

I would like to post a question regarding the Menetrey-Willam model of concrete model.

 

I have set up the MW model as shown in the attached image and am conducting compression tests (φ100 x 200 mm), but each time the softening behavior does not converge after the compressive strength of the concrete reaches its peak. In other words, the analysis ends at the peak.

 

On the other hand, a past contributor to this forum (2021, Oct., Mick Mack) and a paper ("Calibration and Validation of the Menetrey-Willam Constitutive Model for Concrete") have shown that the MW model The softening after the peak is obtained by the MW model.

 

How can I set up the model to reproduce the softening behavior after the peak?

Thank you in advance for your advice.

[John Doyle]

Anytime you begin to lose stiffness in a static structural analysis, there is always potential for convergence trouble.  

The sensitivity depends on the application.  How localized is the material yielding?  How well is the structure constrained?

Perhaps the structure is beginning to buckle and non-convergence is a reflection of a physical instability.

You could test this theory by post processing your force-deflection curve to determine if you have lost all resistance to further loading.

You could also try adding stabilization damping to make convergence more forgiving to material softening.

[7121573]

 

 

Dear Mr. John Doyle

 

Thank you for your advice.

 

The simulation setup simulates a typical compression test, so I don’t think there should be any problem, and I have confirmed that the model using CPT215 will converge.

(Reference for CPT215 model:

&t=909s)

 

We also tried stabilization damping with various patterns and restart points, but again, we did not get a tendency for the stresses to soften.

 

It has already been discussed that the softening behavior of concrete is critical in simulating reinforced concrete.

Any further advice on this fundamental issue would be appreciated.

 

 

[John Doyle]

Additional tips:

1. Compare the two solve.out files CPT215 vs SOLID18x to assertain what the differences are in terms of convergence success/failure.

2. Run semi implicit solver.  See 'SEMIIMPLICIT' APDL command for details.  This technology is exposed in Mechanical GUI under Analysis settings.

3. Run Arc Length Method.  See 'ARCLEN' APDL command for details.  This technology would require a command object under Static Environment, to be executed jsut prior to solve.

 

 

[7121573]

 

Dear Doyle

 

Thank you very much for your advice.

 

I tried the SEMIIMPLICIT command. (I have already turned it on in the GUI.)

I have tried many things concerning the Advanced Analysis Guide, a valuable resource from your company, but still, there is no sign of convergence.

 

Since there is no sign of transition after the peak, I wonder if there is some fundamental problem.

 

Has anyone successfully performed a compression test on MW model concrete?

I believe this is a severe problem affecting each paper’s reliability and structural simulation.

 

[Justine Billaud]

Hello,

I was wondering if your model behave successfully with softening after the peak in compression. I'm trying to do the same but I have also convergence issues after the peak.

Thank you

[7121573]

In addition, those who adopt the MW model often use

TB,CONCR,MATID,,,MSOL

using the APDL (/forum/forums/topic/modelling-concrete-with-reinforcement-using-new-feature/など), but I do not understand the meaning of the six values listed after MSOL. I don't understand the importance of the six values listed after MSOL.

 

I have checked TB, TBDATA, and SOLID185 (e.g., Command reference) but could not find any information on these six values.

 

I would appreciate any help from anyone who knows the meaning of the six MSOL values.