elle53
Structural
- Apr 16, 2015
- 4
Dear all,
I am trying to incorporate a parabolic residual stress field through the thickness of a simple plate model (C3D20R) but uniform in the other two directions. To do that I am using a Fortran script and SIGINI subroutine. My model consists of the following steps:
Step 0: the initial one where the residual stresses are induced
Step 1: a static step without the application of any loads for establishing equilibrium in the case of unbalanced stresses (as suggested by the Abaqus User’s manual) and;
Step 2: a static bending step
If I constrain the whole model then the residual stresses created are correct in both Step 0 and 1. However, this is not desirable as I cannot proceed to step 2 of my analysis with the imposed boundary conditions.
If I use the actual boundary conditions (i.e. supported corners only) then the residual stress field is imported correctly in Step 0 but is almost eliminated in subsequent steps as it is transformed into strain.
Can somebody help with the following questions?
1) Is there any way of incorporating residual stresses in the plate without having to constrain the whole model?
2) The stress field I use in the Fortran script is described by a parabolic function whose definite integral over the thickness of the plate is 0. Is this not considered a balanced stress field? If so why is there a need for the equilibrium step?
3) Is there any way of importing strain in Abaqus that will cause the desired stress distribution instead of importing stress so that I will not have to constrain the whole model?
I would really appreciate your help!
Best,
Korina
I am trying to incorporate a parabolic residual stress field through the thickness of a simple plate model (C3D20R) but uniform in the other two directions. To do that I am using a Fortran script and SIGINI subroutine. My model consists of the following steps:
Step 0: the initial one where the residual stresses are induced
Step 1: a static step without the application of any loads for establishing equilibrium in the case of unbalanced stresses (as suggested by the Abaqus User’s manual) and;
Step 2: a static bending step
If I constrain the whole model then the residual stresses created are correct in both Step 0 and 1. However, this is not desirable as I cannot proceed to step 2 of my analysis with the imposed boundary conditions.
If I use the actual boundary conditions (i.e. supported corners only) then the residual stress field is imported correctly in Step 0 but is almost eliminated in subsequent steps as it is transformed into strain.
Can somebody help with the following questions?
1) Is there any way of incorporating residual stresses in the plate without having to constrain the whole model?
2) The stress field I use in the Fortran script is described by a parabolic function whose definite integral over the thickness of the plate is 0. Is this not considered a balanced stress field? If so why is there a need for the equilibrium step?
3) Is there any way of importing strain in Abaqus that will cause the desired stress distribution instead of importing stress so that I will not have to constrain the whole model?
I would really appreciate your help!
Best,
Korina