Extracting nodal forces

[jball1]

I have multiple perpendicular plates throughout my model that are modeled with shells that are tied together (like the web and flange of a T-beam). The tie constraint represents welds. I would like to extract forces and moments at the nodes on the interface at the location of each weld, and calculate the stresses in the weld. Each tie constraint includes multiple welds that I want to treat individually, so I can't just pull the forces and moments from the constraint.

I extracted the forces and moments using NFORC and summed them (there is an option to do this in Field Output/Save As). I also did this using SF. I plotted the summed forces on the connection thru time (this is a transient analysis). I also used the Free Body Cut tool to calculate the forces and moments on the nodes on one side of the tie. All three of these ways I tried are giving me different answers!

I must be thinking of this wrong. But why can't I get the same answers from NFORC and SF? And which do I trust?

 

[jball1]

All set. I can get NFORC and the free body tool to match. Looks like if I use the summing tool (field output/save as), it only counts the average of the two nodes of adjacent connected elements, instead of including them both. Not sure why they would do it this way. But anyways, I can at least get these two to match, so I will use NFORC.