Notifications
Clear all

[Solved] Deviation between material model and simulation output


rigseb
Posts: 3
Topic starter
(@rigseb)
New Member
Joined: 11 months ago

Hello Jorgen!

I recently tried fitting several material models for a cable sheath model made from a PVC-P material with MCalibration. I tried to fit a TNM and a BB-model, because I wasn't sure which one fits the model better. I basically only want to map the strains and stresses as precisely as possible while I'm not really interested in the time dependent behaviour of the material.

I have uniaxial tension and compression experimental data - unfortunately only for one strain rate - done according to your article about smart mechanical testing of polymers (cyclic tests option 1).

According to MCalibration I get a resonable good fit with both material models (fitness between 5-8%) but when I export the model to Ansys and simulate a tensile test the results are way off: The stresses I get are around 20 times lower than the experimental data. (see figure below)

Now I started altering the parameters manually until I got something which at least remotely resembled my experimental data, but I think there should be a better way. I feel like I missed something.

What's causing the deviation between experimental data and simulation results?

Should I perform the experiments with a different strain rate again?

 

Cheers
Sebastian

2 Replies
Jorgen
Posts: 3955
Moderator
(@jorgen)
Member
Joined: 3 years ago

Hello Sebastian,

That sounds strange. What you tried should just have worked! There is no need to run tests at multiple rates if you did the "smart test" that you mentioned.

Can you provide more info about your problem? Or, can you send us support request with your files (send to: support at polymerfem.com)?

/Jorgen

Reply
rigseb
Posts: 3
Topic starter
(@rigseb)
New Member
Joined: 11 months ago

Hi Jorgen,

While I was revising the data so I could send it to you I stumbled across the cause of the problem:

Somehow an incorrect temperature was transferred when I exported the model to Ansys .xml file. I thought I'm supposed to enter the temperature in MCalibration in kelvin so I put 293K but I realized that in Ansys it registered this temperature in ° Celsius. I changed the temp from 293 to 20°C and now the data from the simulated tensile test fits my experimental data.

There are still negative stresses occuring during the unloading phases, but I guess it comes from the displacement driven boundary condition and the strain rate at which it unloads.

 

Thank you though for your quick reply! I would've probably kept on overlooking the dimension mix up.

 

Best
Sebastian

Reply
Share: