hexagon logo

SPC on rotational DOF in axisymmetric analysis

I'm having troubble constraining the rotational DOF (of an RBE2 reference node) in an axisymmetric analysis. First I noticed that Mentat just ignored this request, just applying the translational constraint on the node; then, after manually editing the .dat file to what I believe should be the correct form of the FIXED DISP card, I still couldn't make Marc actually constrain it (from the .out file I see that only the translational DOF is constrained). Following is the 3rd to 10th block of the FIXED DISP card (excl. blocks 6 & 7) that I tried:
===================================================================
        1        0        0        0        1        0fix_BB-supp_axial-rot
 0.000000000000000+0 0.000000000000000+0
        0        0
        1        6
        2
fix_BB-supp_axial-rot_nodes
===================================================================
Then I tried constraining all the rotational DOFs, in case there would be some convention regarding the (single) rot DOF in an axisymmetric analysis, viz.
===================================================================
        1        0        0        0        1        0fix_BB-supp_axial-rot
 0.000000000000000+0 0.000000000000000+0 0.000000000000000+0 0.000000000000000+0
        0        0
        1        4        5        6
        2
fix_BB-supp_axial-rot_nodes
======================================================================
Still, the .out file tells me I'm just constraining DOF 1...​
 
I have successfully constrained the rotational DOF in Marc in 3d analyses, so I know that is possible; is it not possible to do the same in axisymmetric analyses?
Parents
  • Well, if you want to constrain a surface (an edge in the 2d axisymmetric modelling space)​ from rotating, you need to constrain that DOF...
    It's true you don't have to constrain the y/r direction (since displacement in that direction requires work, ie. induces strain energy), but that does not mean those DOFs will remain zero.
    When there is a possibility for displacements in two directions (i.e. in a plane), there's possibility for rotation, which one could want to constrain.
    I don't follow your implication from not having to fix Ydof => no need to constrain rot Z dof. Am I missing something..?
Reply
  • Well, if you want to constrain a surface (an edge in the 2d axisymmetric modelling space)​ from rotating, you need to constrain that DOF...
    It's true you don't have to constrain the y/r direction (since displacement in that direction requires work, ie. induces strain energy), but that does not mean those DOFs will remain zero.
    When there is a possibility for displacements in two directions (i.e. in a plane), there's possibility for rotation, which one could want to constrain.
    I don't follow your implication from not having to fix Ydof => no need to constrain rot Z dof. Am I missing something..?
Children
No Data