Dear fellow researchers,
I am attempting to carry out a thermal stress analysis on a 2d quarter cylinder part model I have attached in the image file. The geometry is in XY plane and I want to apply differential heating in the inner and outer radial surfaces to generate thermal stresses. I want the stress responses for a generalized plane strain configuration. However, the results don't match with my closed form solution. I am positive there is some mistake in my FEM model. Particular queries I have during the modeling are the entries I have attached in picture 2 file. Can someone please tell me the meaning of these fields. For my problem I would like the cylinder to expand freely with a constant strain in the Z direction.
To check whether Abaqus does what you expect it to do, check the displacements and strains in 3-direction; they should be identical everywhere.
Do not forget to constrain the defined reference node with the appropriate boundary condition (restrict rotation, DOFs 4,5,6).
The meaning of the three parameters are first the thickness (that should beobvious); but Abaqus also allows the generalized plane strain segment to be wedge-shaped, the other two angles define this wedge.
If you are modelling a cylinder in z-direction, your values should be fine.
Dear Martin,
I am very thankful for your helpful suggestion. Now my stress distributions are matching well with the analytical model, so I believe I am close to the solution. However, I still have a problem that the 3-direction strain is not constant. With regards to your inputs to my model, let me clear what I understood and what I have done so far:
1. The reference node is any node in my 2d section which is to be defined from the assembly module using sets.
2. The boundary condition for the generalized plane strain definition is to arrest the node and the section's rotaions in 1,2 and 3 directions.
I have a query with the second point that my node has only three DOFs U1, U2 and UR3. Am I still doing something wrong?
Excuse me for the ruggedness of my software skill. Most of my experience has been YouTube taught. Any suggestion is highly appreciated. Thank you.
If your problem is defined in the XY plane then the Z axis is axial. In this case I think the generalised plane strain boundary conditions should involve the translation UZ and the rotations RX and RY or in ABAQUS speak, U3, UR1 and UR2. If these freedoms are set to zero then you should obtain the same result as a plane strain model.
- Badges
- Science method