I'm implementing a new element model as well as a new material model to complement it. When I run a simple test under load control, I get reasonable results up until the material softens, which is expected. When I try displacement control, however, I can't even get one step to converge. I have tried the different algorithms and settings to no avail. If I use a very large test tolerance (e.g. 1.0e-2) I can get a few data points.
Is there anything in particular I should look out for? Is there a function or functions I should check for bugs? Any suggestions would be appreciated. Thanks!
New Model works with Load Control but not Displacement
Moderators: silvia, selimgunay, Moderators
Re: New Model works with Load Control but not Displacement
i have never come across this problem before. are the few points with displacement control on the curve (i.e. same as LoadControl if same load increment)?
Re: New Model works with Load Control but not Displacement
No, the points aren't the same. Any ideas?
Re: New Model works with Load Control but not Displacement
i take it you are using nodal loads. could possibly be the resisting force has the wrong sign.
both integrators are just using the element tangent and resisting force. if both are correct it will work. if the model is small, add a 4 to the end of the convergence test (this will spit out the dR and dX vector at each step). you could also mod your element to spit out the K and R at each update to see if anything looks wrong.
both integrators are just using the element tangent and resisting force. if both are correct it will work. if the model is small, add a 4 to the end of the convergence test (this will spit out the dR and dX vector at each step). you could also mod your element to spit out the K and R at each update to see if anything looks wrong.