Different Results in Beam Models

[f.stronz]

Hello,
I have created a beam model consisting of various members and want to determine the deflection. In the overall model, it appears that there is a hinge within the beam. When I create a submodel, apply the same loads, and set up similar supports, I get significantly lower deflections. How is this possible? I have not added any element releases.

Thank you.

[John_Holtz · Accepted solution]

Hi @f.stronz 

It is hard to visualize how the horizontal beam can be "split" in the analysis but still attached to the center column. 

What I do see is the real beams are in different planes. Columns "A" and "B" are in one plane, and column "C" and horizontal "D" are in a different plane offset by some distance. In the beam model, it looks like the lines are all in the same plane.  This implies the beam "D" is offset mathematically, and something happens in the analysis when the beam elements are offset.

In the frame generator model, you should make the sketch lines at the true centroid of the beams. The sketch lines "C" and "D" will not connect to "A" and "B".  To connect beam "D" to the columns "A" and "B", you need to put in a dummy beam that has a length of "offset".

John