Material calibration

[liq001]

Hi guys.

Does anyone have good advice or methods for material calibration. Like I have an unknown material, I had a compression test and the output is a force displacement curve, and I need to came up a an stress-strain for this unknown material. Does anyone have good experience for this task?

Best
Lee

 

[SWComposites]

well,
convert force to stress by dividing by the area,
convert displacement to strain by dividing by the gage length
should be documented in the test method standard that was used.

 

[FEA way]

Abaqus/CAE has some basic material calibration tools. Much more advanced app is available on the 3DExperience platform (cloud software with Abaqus and other tools offered by Dassault Systemes). Isight can be used together with Abaqus to perform iterative FE model-based calibrations.

But it’s often sufficient to just use Excel to perform various conversions (e.g. from engineering to true measures) on your test data. You can find some interesting examples in the Simulia Community (official Abaqus forum, look for posts by Tod Dalrymple).

 

[liq001]

well,
convert force to stress by dividing by the area,
convert displacement to strain by dividing by the gage length
should be documented in the test method standard that was used.

The problem is the cross section area is irregular shape which can not be extracted.

 

[SWComposites]

That's a rather poor test specimen design to use to get a material stress-strain curve. you need to repeat the test with a specimen with a gage section with constant cross section over which you can measure strain.

Or, you will need to guess at the material stress-strain curve, input it to a model, run the model and compare to your specimen results, and iterate on the stress-strain curve until you get a sufficient correlation.

 

[liq001]

Abaqus/CAE has some basic material calibration tools. Much more advanced app is available on the 3DExperience platform (cloud software with Abaqus and other tools offered by Dassault Systemes). Isight can be used together with Abaqus to perform iterative FE model-based calibrations.

But it’s often sufficient to just use Excel to perform various conversions (e.g. from engineering to true measures) on your test data. You can find some interesting examples in the Simulia Community (official Abaqus forum, look for posts by Tod Dalrymple).

Thanks so much. Any specific link to the post? The difficulty is I can not know the cross section area as the testing sample is an irregular shape. Manually guess the stress-strain which not idea. Need some guidance here. If I had a start date for stress strain. How can I iterate this curve to fit the displace-force curve? Any method to iterate the original stress-strain curve?

 

[FEA way]

Any specific link to the post?

There are many interesting posts about different material calibration problems from that user of the aforementioned community. Look for his profile and you will find a whole collection of useful resources (including Excel and Abaqus files).

How can I iterate this curve to fit the displace-force curve? Any method to iterate the original stress-strain curve?

You could use the software that I’ve mentioned above - Isight might be available for you as part of the Abaqus UFEA package. Or try with a Python script to automate this task in Abaqus.

Regardless of the approach, you will need a stress-strain curve as input in Abaqus. Modeling an approximate shape of the test sample will let you calibrate this after some iterations.

 

[liq001]

There are many interesting posts about different material calibration problems from that user of the aforementioned community. Looks for his profile and you will find a whole collection of useful resources (including Excel and Abaqus files).


You could use the software that I’ve mentioned above - Isight might be available for you as part of the Abaqus UFEA package. Or try with a Python script to automate this task in Abaqus.

Regardless of the approach, you will need an stress-strain curve as input in Abaqus. Modeling an approximate shape of the test sample will let you calibrate this after some iterations.

Yes, i understand what is your points and methods. I am wondering the methods to iterate my material strsss-strain curve instead of guessing? Manually adjust? If I am using my script.

 

[FEA way]

Yes, i understand what is your points and methods. I am wondering the methods to iterate my material strsss-strain curve instead of guessing? Manually adjust? If I am using my script.

If you can roughly estimate the dimensions of the sample, that may be your initial guess for the stress-strain input derived from the force-displacement curve that you already have. Apply the displacement pattern from the test and record the force. Change the plasticity input until the force variation matches with the test data. This can be automated with a script or, better, with Isight (using the Data Matching component).