Transverse Weld Simulation

I have recently been using the Inspire Extrude Metals software to simulate the extrusion of Aluminum as well as a user-defined Plasticine material. The main focus of these simulations has been to observe the output transverse weld length.

The following trends that I have found in my simulation do not match my experimental data:

The simulation predicts that an increase in the extrusion ratio, in an otherwise identical die, results in a decrease of the weld length.  The change in extrusion ratio caused an increase in exit velocity, which when changed to match the lower extrusion ratio example caused a further change in the weld length. The material I was using was strain-rate insensitive so the velocity should have no effect on the transverse weld. 

Changing the material in the simulated model had no effect on the weld length.

I was curious if anyone knows the parameters that have an effect on the weld length within the Inspire Extrude Metals software. Are the results I am seeing expected, or is it possible that I am making mistakes in my model?