I am modeling inner grooved tube spinning process in ABAQUS/Explicit, There has been attached a schematic illustration of my model. To reduce the simulation time rollers are chosen to rotate around the tube rather than rotating tube itself. Roller reference points are located on tube axis to able the roller rotate around tube axis.In post processing part I want to extract reaction forces at roller reference point, but when I extract reaction forces, I found out they are somehow meaningless and there is a lot of serrations. Is there any thing wrong? maybe about the coordinate system? I think because of rotation of rollers, it seems that I have to use a specific strategy to extract reaction forces, but which?
Any suggestion would really be appreciated.
Not totally sure of what you are trying to do with your model. The tube is meant to spin and the rollers are pushed inwards in order to groove the tube?
Whether you rotate the tube or the rollers should make no difference. Presumably you need to apply a radial force inwards to groove the tube and I guess you will be using some form of plasticity model for your material...
Hi Dear Angus,
The tube is fixed and the rollers spin around the tube and reduce the wall thickness of tube. this is called tube spinning or flow forming process. In reality, the process is in such a way that tube spins and rollers have idle motion, and contacting with the rotating tube make the rollers to rotate too( likewise the attached image here). But in simulation, to reduce the simulation time the tube is fixed and roller are assigned to rotate, because roller are rigid body and their rotating do not affect simulation time but rotation of the tube could increase the simulation time strikingly.
The material property are assigned according to the uniaxial tensile test up to 50 percent true strain. there is not any kind of applied force boundary condition during simulation. all the boundary condition are just rotation and displacement ones.
Hope I put it simply.
If I understand correctly, the difficulty is that your rollers are moving and thus if you look at the say the reaction forces on the rigid body in the global coordinate system these are meaningless without an appropriately moving coordinate system on each roller. I would recommend looking into connector elements and attaching local coordinate systems to these.
Another approach would be to simply define a cylindrical (or possibly spherical if both altitude and azimuth are important) coordinate system with the center axis of the tube as the origin. Applying the cartesian to cylindrical transformation for the reaction forces reported at the centroid of each rigid body in the global system should give you the correct radial and circumferential reaction forces.
Thanks Scott,
I will try second approach you suggested.
But how would it be if I take abs(absolute values) of Reaction forces in visualization module. because of rotating rigid body the reaction force values fluctuate between negative and positive values and I think if I take the abs of Reaction forces they will be meaningful.
Please let me know your view point about this.
- Badges
- Science topic