I think the procedure that Ron has explained is a good idea.
If you're looking for the stress distribution near a particularly zone you can do a free body analysis, ignoring the rest of the body.
You can use the bearing load instead of the "simple force command".
The welds are ok?