You can do a sequentially coupled analysis. Firstly, analyze the transient tempeature by using a moving heat source (write a subroutine). Then perform static analysis by reading the tempeature as a thermal load. Direct specification of nodal tempeatue as initial condtion is not recomended, as it can produce an inaccurate residual stress field. Good luck.
After having applied the thermal and elastic loading in one direction, you can apply the back superposition of the elastic deformation in negative direction relative to the initial orientation. So you can obtain the thermal residual stress.
- Uncoupled thermal to extract the temperature followed by a structural to calculate the stress based on the material temperature dependant properties. Here, the stress is calculated based on the final thermal results.
- Coupled thermal-structural modelling using element 226 (the 22X series). This element has a displacement (Ux,Uy,Uz) and temperature degree of freedom. And APDL will be needed to set the element. You will also need a temperature dependant material properties (as with uncoupled). The stress will be calculated at each step of the welding process.
You can also add a cooling period within the thermal setting to account for the variation of the stress during this period.
I am trying to simulate a welding joint and find out the residual stress due to the welding for my research work. Does any one have a good document or tutorial on it to do these in ANSYS. Is there a way to simulate welding and welding joint?
What type of elementsis best suited to perform such coupled thermal-structural analysis? I am having the same problem...I have succesfully simulated transient thermal (moving heat flux over the region) and coupled it with structural. The problem is - that at the end of the process - no residual stresses emerge (as if materials is elastic enough to compensate thermal loads) ... some help will be much appreciated...kind regards. M