I don't understand very well the question, could you please make a drawing sketching the problem?.
You can define any constraint you like (pinned or fixed), in geometry or directly in nodes, but your model should be stable, not rigid body motion should exist or the NX NASTRAN solver will give you FATAL error, OK?.
Hi Blas, thank you for replying .
When I trying to define only simple support constraints (4 edges) for my model (some box lid) I get the message "No results are found" if I tying to open the results . It seems that the solver didn't solve the solution because the solve time too short (even for 101 linear).
In case that I add a fix constraint - everything going OK and I get reasonable results .
But than I have some unwanted stress results in the areas where the fix constraints were defined.
Please revise your nx nastran *.f06 output file in the model directory and search for FATAL word, you will see there the reason of the error.
Or better run a Normal Modes/Eigenvalue analysis (SOL103) and animate the first mode shape, you will see the reason of the error, why your model is wrong constrained (to learn more take a look to my post at the following address: https://iberisa.wordpress.com/2011/02/20/mensaje-d
Or post your full NX AdvSim model here --both *.prt, *.fem & *.sim files-- and we will take a look to it, OK?.