I have not really though about that. Perhaps it is the explicit solver, perhaps there is an instability in the incorporation in the boundary conditions and the out-of-plane displacement. Do you have any other ideas?
I ran two simulations, one with umat, one with vumat for the same material model, with same boundary conditions. The results from the umat one showed a smooth line and the vumat one followed the same curve, however, not smooth. I am not sure what is the reason.
I have been able to reduce the oscillations of the VUMAT output by avoiding the *variable mass scaling, and by changing the *density value only. I would guess it might provide uniform kinetic energy for all elements. It is for quasi-static simulations only.
[quote=Wei]I ran two simulations, one with umat, one with vumat for the same material model, with same boundary conditions. The results from the umat one showed a smooth line and the vumat one followed the same curve, however, not smooth. I am not sure what is the reason.[/quote]
Dear all,
I've been trying to reproduce the results in this thread using both Dr. Jorgens and Mr. Hohenbergers subroutines for Neo-Hookean material models. Unfortunately I'm encountering results similar to this thread where the vumat results differ from the results obtained using Abaqus own model.
The models are the same for one 3D element in uniaxial tension as well as simple shear but for plane strain or the same loading with multiple elements I get different results. I have a feeling my input files could have the wrong setup so would anyone be so kind so as to share their test files?
All the best
F