Contact Analysis

[KrishanSingh0706]

Hi everyone,

I am having trouble with boundaries in a 2d analysis, I have two bodies, one inside the other experiencing forces. I can make it work with several friction boundary conditions on pairs of edges, however rather than identifying which faces are/will be touching I would like to apply one boundary condition to all the boundaries as follows:

Code: Select all

Boundary Condition 2
 Target Boundaries (8) = 105 106 107 108 109 110 111 112
 Mortar BC = 3
 Friction Contact = Logical True
 Static Friction Coefficient = Real 0.9
 Dynamic Friction Coefficient = Real 0.8
 Contact Depth Offset Initial = Real 1.0e-4
 Plane Projector = Logical True
 Galerkin Projector = Logical True
 Normal-Tangential Displacement = Logical True
 Initial Contact Velocity = Real 0
End

Boundary Condition 3
 Target Boundaries (4) = 113 114 115 116
 Normal-Tangential Displacement = Logical True
End

When I try this it runs for a few iterations and then fails, is this the correct way to do this?
The MultipleBoundaryConditions.sif file contains the setup that works, and the SingleBoundaryConditions.sif file contains the setup I would like to work.

Thanks very much in advance for any guidance you can give.

Regards,
Krishan

[kevinarden]

1. The normal projector appears to work better for this case since the boundary is all four sides, planer worked OK on multiple BC conditions because each 1 individually is planer, but as a group they are not

2. Elastic analyses are sensitive to element formulation, linear triangles almost never work unless they are very finely meshed. If you want to use triangles I recommend a finer mesh and increases the order to 2.

3. Quad elements are much better, and quad elements of order 2 even better.

4. Therefore I re-meshed, When I re-meshed I combined the boundaries in the mesher, although this is not a required step.

5. I also added damping since it is transient, transient includes accelerations and velocities, and colliding bodies chatter if they are not damped. To reduce the collision I changed the densities to 1, assuming that you are not interested in the collision part of the simulation.

result.png (34.34 KiB) Viewed 301 times

project files are here

https://github.com/mrkearden/contact/tree/main