Hi, I was wondering how would I be able to define a shape memory alloy in prepomax like nitinol using the properties given by default? I also want to map the martensic and austenite phase, or at the least, one of them
Regards
To model such materials accurately, you would need superelasticity material model (taking phases into account). It’s available in Abaqus but not in CalculiX. So you would have to write a umat.f subroutine to have this model in CalculiX or simplify the problem and use the material models available in CalculiX (such as linear elastic - plastic).
I see, so basically I do this in abaqus, use umat.subroutine in abaqus, save it as an inp file, edit physics in prepomax and I’m good to go?
Not really. In Abaqus, it’s a built-in material model. You would need to write the umat.f subroutine from scratch or find it online (sometimes subroutine codes are included in research papers and theses). Then there can be some differences with UMAT in Abaqus. I guess I would start with elastic-plastic model and go for subroutine only if absolutely necessary.
Ok will try, thanks a lot for the guidance
Hi, so I was looking for existing literature on the matter and in this paper,
These parameters are used in CalculiX:
Any idea of how that may be done? I probably will try to do the same, but I couldn’t find all these properties in PrepoMax or CalculiX helpbook
Since CalculiX doesn’t have this material model, it’s possible that they just used a umat.f subroutine. Or performed the calculations in Abaqus but didn’t mention it for some reason. Maybe try to contact them.
I see, I will, and just to clarify, if I make a umat subroutine in abaqus and import it to calculix via prepomax, I wont be able to solve it?
Subroutines in CalculiX not only may have a different syntax but they require recompiling CalculiX from the source code. But you can find some ready-made examples and workarounds. This thread might be helpful: UMAT option in PrePoMax
PrePoMax is not a problem here as you will only need to configure CalculiX properly and add necessary keywords - you can just use the keyword editor for that.
Also, it might be good to ask on the CalculiX forum. Someone there should have more experience with CalculiX subroutines.
This is extremely helpful, so was the thread, thank you so much!
Also, if I don’t want to take into account the intricacies of phases, but only the initial shape and the final shape, would there be any way to do that? for example the initial phase is a straight beam and the final shape at temp T2 is a deformed bar, this would be something akin to ‘training’ the SMA
You could try with regular temperature-dependent elasticity first.
For reference, here’s how the superelasticity model in Abaqus works:
- no loading - austenite with linear elasticity (E_A)
- loading:
a) transformation to martensite beyond a certain stress - linear elasticity with properties calculated following the rule of mixtures (depending on the fractions of both phases)
b) martensite only after a certain stress - linear elasticity with E_M - unloading - transformation back to austenite (transformation strain fully recovered) at a different stress than from austenite to martensite
Shape setting is done using the special *ANNEAL
procedure to set all state variables to zero.
I see thank you for your reply and sorry for replying so late. I will start learning ABAQUS as well