Limits of Solidworks FEA

[CharlesGrey]

Hello,

I am wondering if anyone has experience with the limits of Solidworks FEA. My current employer has used it for years but has also outsourced larger studies to a consultant that uses ANSYS.

Typically, anything that we have been able to run ourselves has been okay, with only minor discrepancies with ANSYS results. It has only been for when the models get so large that we do not have the computing power to run them that we have really relied on ANSYS.

But recently we have found a different problem. We have a long weldment (symmetric, with half shown below) that gets a number of hydraulic cylinders and other weldment attached. We have been looking at the deflection in the long tube (around 26 feet) and only get around 63% of the deflection using Solidworks that our consultant gets using ANSYS. We see around 1.1" and they get 1.75". We measured one under load in the plant and ANSYS is correct.

After having gone over the ANSYS model, boundary conditions, and loads, in depth with the consultant we cannot find a difference between the two setups.

We do have large displacement activated when we run the study.

Is there simply too much going on for Solidworks to get this right? Or is there something special that we should be doing with large members under deflection? Any other thoughts?

Sorry, can't share much more of the geometry for IP purposes.

[EngineerM]

The principle of these two software is the same but they differ in the type of partitioning into small parts (mesh). If the boundary condition is set the same, the result should be similar. You can study its limits from the white paper in the folder.

[Frederick_Law]

All CAD FEA are certified to within same tolerance. It's inside help somewhere.
The only reason for different result is the setup are not the same.

[rothers]

How does your mesh and element parameters compare to the ANSYS analysis ?

[mp3-250]

it sounds like overstiffness due to volumetric locking or similar artifacts.
what kind of elements are you using?

we had a young guy making fea with inventor. he used all solids with a bad mesh for a very big structure he got like 3mm deflection, while we had calculated like 20mm both by hand and solidworks with shell elements.

run some benchmark with partial models or a simplified case and compare different approaches.