I tried to get some reaction forces for a model today. Nothing made sense. So back to the cantilever I go....
Check out this screenshot. What I have is 100 pounds loaded on the split-line circle at the end of the 12-inch beam. In this case I've constrained two edges instead of the whole face (I was playing around with it some).
I selected two opposing faces on the "side" of the beam. Look the reaction forces. Really?
1) Reaction forces should be zero everywhere where there isn't a constraint. My VAR is telling me the side faces "touch" the constrained face/edges so there is a reaction force there. In this case, the only face where this is true is the "free face" directly opposite the constrained face.
2) Reaction forces should balance with the load. So obviously even there are forces on these side faces, they are wrong.
3) I'm told by my VAR I can't get moments because volume elements are tetrahedrons and the nodes have only three DOF. Shell elements would show me torque because they are 6 DOF. Really? I can do this one in my head: the torque is 100 ft-pounds. Your $15k software can't do it? The only time you'd have to do FEA before getting the resultant force is if you prescribe displacements instead of loads.
For the torque question I got the same kind of answer when I asked why can't I have displacement plots as rotation relative to some axis. My take on it is, the whole point of a computer is to do math faster than me, so the only excuse is feature value and market needs. Well, let me tell you, the market probably needs right answers.
Am I missing something here?