Hi there!
I have a uniaxial compression model made in Abaqus. It is just the compression of an elastic cube. This is being done in Explicit since I will later be complicating the model, adding things that require explicit (ie SPH, explicit material models). I have a bigger network of people working in LS-Dyna so it makes more sense for me to work in Dyna.
The Abaqus model runs well and the force-displacement response is linear, as expected (elastic model, as a reminder). The LS-Dyna model however, is modelled similarly but of course the contact definitions are different or whatnot because they are defined differently, and it does not produce a linear output: the response slowly gets stiffer over the course of the compression.
I have outlined some details from the Abaqus model below. I suspect error comes from how the contact is defined in LS-Dyna. Does anyone know the equivalent in LS-Dyna of Abaqus's Surface-to-Surface contact (kinematic contact method with small sliding, default values) including frictionless "Hard" Contact?
Abaqus:
- C3D8R hex elements for the cube, and S4R elements for the plate. The plate is fixed, the cube's top surface is translating downwards, and the force at the interaction is being extracted. Both parts use elastic material properties, with the shell plate being much stiffer.
- Default values for the Step, except for the time period and the time step scaling factor (set to 0.9)
- Contact: Tangential Behavior (Penalty, Friction Coeff 0), Normal Behavior (default), Surf2Surf contact with Kinematic Contact (small sliding)
I have shared the .inp and .k files.
There are two LS-Dyna .k files: one with no *CONTACT card but using *RIGID_WALL, and a second one that more closely resembles the Abaqus model with a shell plate.
- Badges
- Science topic