Applying strain boundary condition

[cvady]

Hi,

In Abaqus, we can apply displacement, velocity or acceleration as boundary conditions and solve the problem. In the similar manner, how can I apply strain as boundary condition and solve it?
I have gone through the Abaqus documentation and could not find it. Any help is greatly appreciated.

Thanks,
Siva

 

[FEA way]

There’s no such option in Abaqus. In fact, I’ve never heard about strain boundary conditions. For what purpose do you want to use it ?

 

[cvady]

I have the strain data (at all nodes) from the experiments. I would like to apply this strain in FEA at all nodes. Thanks for your quick reply.

 

[SWComposites]

Convert the strains into nodal displacements relative to a fixed point.

 

[cvady]

How can I convert nodal strains to nodal displacements? Is there any built-in Abaqus facility for doing this? I have E11,E22,E12 from experiments at all nodes and would like to convert to U1,U2.

 

[Mustaine3]

What do want to get from the simulation? When you have strain, you can directly calculate displacements and stresses. So what is the purpose of the simulation?

 

[cvady]

Sorry, I have edited the question.
I am having strains(E11,E22,E12) from experiments and convert them to displacements(U1,U2).

 

[FEA way]

This sound like the Inverse Finite Element Method where we use surface-measured strains to obtain 3D displacement field of the structure. But it’s not possible in Abaqus. Is that what you want to achieve or your analysis has different goal ?

 

[cvady]

The goal of the analysis is to apply the experimental strains and perform stress analysis in FEA. Are there any tools for performing inverse FEM?

 

[IceBreakerSours]

It sounds like you have a surface strain field (from DIC, perhaps) and you want to estimate a possible (unknown) displacement field that might result in the measured strain field. As mentioned above, this is an inverse problem i.e., you have to run the forward problem multiple times in an iterative sense. Inverse problems seem easy at a first glance but the devil lies in the details i.e., you have to know the geometry, BCs, material model in detail so that the optimizer that is solving the iterative algorithm has a chance at success. Otherwise, the whole process can easily become a waste of resources.

Try to do this on a very simple problem first. In fact, try to do this on a problem that you have an answer for already (say, a tensile test on a coupon of a simple material).

You can use Isight or Python with Abaqus to do this.

*********************************************************
Are you new to this forum? If so, please read these FAQs:

http://www.eng-tips.com/faqs.cfm?fid=376

http://www.eng-tips.com/faqs.cfm?fid=1083

 

[FEA way]

If you somehow obtained strain values for the whole model (not just its surface) and you want to calculate stresses then it's a different type of problem. In such case check how linear FEA calculates stresses from strains (strain vector is multiplied by elasticity matrix).

 

[cvady]

I do appreciate you guys for your continuous support. I got good insight into the problem.