AhmadGuennam
New member
- Jun 25, 2011
- 1
Hi everyone,
I'm working with Abaqus UEL subroutines to define piezoelectric linear 8 node brick elements with non linear material behavior. The code works fine for a single element, however when I try to assemble more than one element I get convergence problems if I use dielectric permitivity below 1.0E-7 F/m. I tried to run an analysis with only two elements and writing the nodal variables to a file. When do this I see that at the begining of each new increment the UEL is calling twice, as explained in the Abaqus manual, but the nodal quantities are modified by ABAQUS before to start the equilibrium iterations. I.e. at the end of increment T, certain nodal dof value is X, when Abaqus calls UEL at the begining of the T+1 increment, that dof value is X, but the second time is X + d and this modification was made before the global Newton iteration. Could you please tell me if this is a normal behavior? Because in my case, the change are so strong that I supose that this is why I get convergence problems (the material model is embeded in the UEL subroutine). Any advise will be of great value. Thnaks. Ahmad
I'm working with Abaqus UEL subroutines to define piezoelectric linear 8 node brick elements with non linear material behavior. The code works fine for a single element, however when I try to assemble more than one element I get convergence problems if I use dielectric permitivity below 1.0E-7 F/m. I tried to run an analysis with only two elements and writing the nodal variables to a file. When do this I see that at the begining of each new increment the UEL is calling twice, as explained in the Abaqus manual, but the nodal quantities are modified by ABAQUS before to start the equilibrium iterations. I.e. at the end of increment T, certain nodal dof value is X, when Abaqus calls UEL at the begining of the T+1 increment, that dof value is X, but the second time is X + d and this modification was made before the global Newton iteration. Could you please tell me if this is a normal behavior? Because in my case, the change are so strong that I supose that this is why I get convergence problems (the material model is embeded in the UEL subroutine). Any advise will be of great value. Thnaks. Ahmad