Dirichlet and Periodic Boundary Conditions

[pmxppl]

Hi

I am running an ElasticSolve analysis where I am applying a centrifugal load as a body force on a rotor.

I have periodic boundary conditions at the two sides of a section of the rotor and a 0 displacement boundary condition at the inner of the rotor.

It looks like I have a conflict between the Dirichlet boundary condition and the periodic BCs at the very corners where they meet. I have attached a contour plot of the displacement magnitude and the zoom in on the corner.

It looks like the displacement is non zero on the corner i.e. the 0 displacement condition is not conserved there. This leads to also high stresses in those corners.

My Boundary Conditions and solver settings are e.g.

!###############################################################################
!############################# Boundary Conditions #############################
!###############################################################################

Boundary Condition 1
Name = RotorInner
Displacement 1 = 0
Displacement 2 = 0
Displacement 3 = 0
End

Boundary Condition 3
Name = RotorPeriodicityR
Anti Radial Projector = Logical True
Galerkin Projector = Logical True
Mortar BC = Integer 4
Mortar BC Static = Logical True
Normal-Tangential Displacement = Logical True
End

Boundary Condition 4
Name = RotorPeriodicityL
End

!###############################################################################
!################################### Solvers ###################################
!###############################################################################

Equation 1
Active Solvers(2) = 1 2
Plane Stress = Logical True
End

Solver 1
Exec Solver = Always
Equation = "NonlinElast"
Variable = -dofs 2 Displacement
Procedure = "ElasticSolve" "ElasticSolver"
Nonlinear System Convergence Tolerance = 1E-10
Nonlinear System Max Iterations = 50
Linear System Solver = Direct
Calculate Stresses = Logical True
Apply Mortar BCs = Logical True
Mortar BCs Additive = Logical True
Nonlinear System Convergence Without Constraints = Logical True
End

Thanks and Kind Regards,

Paul

[kevinarden]

From the solver manual

8. Dirichlet boundary conditions and limiters
By default the Dirichlet conditions are set in the order that they just happen to appear in the
boundary element list. This is fine when there are no conflicting boundary conditions. However,
when such is the case the user may want to control the order of the conditions – the last Dirichlet
condition always prevails. This flag enforces the BCs to be set in the order of boundary condition
numbering in the sif file.
This means whatever the last boundary created is at the intersections prevails
if you want the othe BC to prevail then

In the Simulation section use

Set Dirichlet BCs by BC Numbering = Logical True

and make the displacement condition after the periodic one.

[pmxppl]

Hi Kevin,

Thank you very much for your reply.

I tried your suggestion but unfortunately it did not fix my issue. The result was the same as without the change.

I did however find that removing the "Galerkin Projector = Logical True" from the periodic BC gave me a good result.

BR

Paul