Help Needed: Discrepancy in 3D FEA Analysis Results

[lthankachen]

Hi All

Subject: Assistance Needed with FEA Analysis Discrepancy in 3D Model

Hi All,

I am new to the FEA and currently working on a Finite Element Analysis (FEA) project focused on a flat sheet rolling process aimed at identifying the root cause of a bolt-hole crack. In my analysis, I am Using Inventor Nastran to conduct both 2D and 3D simulations on a flat sheet with two bolt holes.

As a preliminary step, I have opted for a linear analysis on a flat sheet without bolt holes in 2D and 3D, treating it as a simply supported beam subjected to a load of 500kN. The goal is to cross-verify the FEA results with hand calculations. The hand calculations align well with the 2D analysis outcomes, including maximum principal stress, etc. However, the 3D analysis yields disparate results.

I have attached the relevant files for your reference, and I would greatly appreciate any insights or assistance you could provide in resolving this discrepancy.

NB: I am using Inventor Nastran Version 2023.0.0.23

[John_Holtz]

Hi @lthankachen 

The beam model and hand calculations allow the "face" at the ends of the beam to rotate. Notice how the face in the solid model is not free to "rotate". In order to be allowed to "rotate", the nodes on the top need to be free to move in the X direction relative to the nodes on the bottom.

In other words, the constraints cannot be applied to the face of a solid if they need to rotate. You need to apply the constraint to an edge so that other nodes on the face can translate. The translation is what causes a "rotation" in real life (and in the analysis).

John