FE Model Feedback / Criticism

[Anonymous]

I've been putting together a rather complex FE model and using it for response factor analysis. We're rather confident in the model after we've done some qualitative checks on the mode shapes, deflections, etc. but that doesn't mean we can't improve on the model / method of building the model. Therefore, I've included a screenshot below of one of the more complex areas of the model, and I'm interested in general community feedback. I can't include the Robot model for public distribution for obvious reasons, but I could share it with Autodesk Support agents if requested.

Note the comment on the top-flange offset: I don't think this setting actually had any significant impact on the analysis, but it would help if someone had any comments on this modelling decision. To me, it seems like it should have an effect since it's a 75mm offset!

[Pawel.Pulak · Accepted solution]

Topic discussed outside of the forum - so here repeating my remarks to make them useful for others.

Basing on the screen capture I have 2 remarks.
It seems that steps are modeled by bars and these bars are connected (fixed) to stringer plates in single nodes. If the plates of steps are welded to stringers along all length of the connections it could be useful to model these connections (or at least parts of their lengths) as rigid links. It would increase the stiffness, redistribute the stresses but also make the model more realistic.
As concerns the offset of beams below the landing plate it should increase the stiffness. This increase would be clearly visible when comparing "pure plate" model (containing only the landing plate and beams without offsets) with analogous model but with offsets.
In your complete 3D model with stringers, stairs, edge plate along landing plate this increase of stiffness may be difficult to notice when superposed on complex 3D effects.

I hope these remarks will be useful.

---------------------------------------------
If this post answers your question please click the "ACCEPT AS SOLUTION" button (the button visible when logged to the forum). It will help everyone to find answer more quickly!

Regards,

[Anonymous]

Superb, very helpful response from Pawel. I meant to post this response, but have been side tracked since. I have added one additional comment here, for insight as to how Pawel's suggestions helped in my case, and how the suggestions can be generally helpful to others:

For anyone building a model with a bar adjoining a plate where you will have serious forces in a number of different DOF's, you should especially consider the Rigid Link suggestion. In my case, this helped to stiffen the stair case and increase the first mode natural frequency of the system. To achieve this, if you are new to Rigid Links, you have to place at least a couple of nodes along the interface of the bar section and the panel, and link the main bar node (Master Node) to the added nodes (slave node). When you remesh the panel, the added nodes should be built into that mesh.

I hope that makes sense; please feel free to message me otherwise.