I am trying to run a simulation on a 4 layer part that has two base layers of I-beams, a frame I made in the frame generator (below) and a solid sheet on top. I am having some issues with the FEA side and was curios to see if the frame analysis tool would yield the same results as the stress analysis simulation. For the frame generator I saw 4 times the amount of deflection despite using (what I believe) are similar loading and constraints, however now I am unsure if I can trust to run an FEA simulation with a frame generated component. The frame is made up of 2 inch steel box beam with a 0.25 inch wall thickness. For the frame analysis package, I have applied a pressure that equates to a 70 kN load across four different sections of the frame. Supports have been placed under all four corners, with 1 being fixed and the other 3 free to move. For the FEA simulation, I applied a similar pressure by creating 2 inch wide plates and applying the same pressure at the same four points. There are also 2 inch square pads under the four corners of the frame with one edge of one pad being fixed and the bottom of the pads having a frictionless constraint. There are separation contacts between both the pressure plates on top of the frame and the frame as well as between the pads and the frame.
I am more curious if it is an operator error (which is very possible) or an issue with the frame generator and FEA. A few things I have noticed with the frame generater is that all the segments of the frame are bonded together (under contacts) which makes sense since they are welded together. However when I put a sheet on top of the frame, there are 3 contact points created between each beam on the frame and the sheet (one on the face of the frame that is in contact with the sheet and the other two on the rounds of the frame and the sheet, even though the rounds are not actually in contact with the sheet). Either way, I am seeing deflections of less than 0.1 mm when a 70 kN load is applied which I know is not right and just want to know if using a frame generated component is not possible with an FEA simulation. Thanks for any input and sorry for the long post.
Frame in FEA. Maximum deflection of 2.584 in
Same frame in frame analysis. 10.27 in max deflection under identical loading
Can you attach the assembly here?
Unfortunately, the files are too big to upload using pack and go and then zipping the files. Is there some way I can e-mail them to you or someone else? I have also detailed some information about the assembly, constraints, and issues below. Any help would be greatly appreciated.
Base I-beams are welded to the side plates. Two of these sub-assemblies create the base layer for the floor. These are held together by mounting plates on either side. I beams then run perpendicular to these I-beams and will be bolted at the intersection points. The frame and top sheet will be welded together and several bolts will connect the frame and sheet assembly to the I-beams below.
There is a 3.9 m x 3.9 m span that the floor has to cover. The floor will be resting on concrete for everything outside of that. The top floor will experience a loading of 10 kN/m2 and we are hoping to see deflections of less than 7.8 mm. There will also be a punch loading of 20 kN/10 cm2 that the floor must withstand.
With the frame analysis, I exerted continuous loads at four of the locations (as shown in picture). I exerted a loading of 0.215 kN/in across the 81.5 inch sections to get an approximate total loading of 70 kN. The back left position is fixed and the other 3 corners have a floating pinned constraint.
For the FEA simulation, I placed two inch wide plates across the top of the frame and 2 in x 2 in blocks under the same 4 corners. All of these components have a separation contact between them and the frame. On the back left corner the back corner edge has a fixed constraint and the bottom of the 4 pads have a frictionless surface constraint. A pressure of 167 kN/m2 was exerted to each plate, equating to an approximate loading of 70 kN (same as other simulation).
When working in the full assembly (listed in 1st paragraph), I was seeing results of minimal deflection (less than 0.1 mm). When I split the components into various subsets, I get mixed results. For the two layers of I-beams, I see to get good deflection numbers. When I then placed the frame and sheet together to complete the assembly, that is where the minimal deflection numbers came in. It does run with a thin surface warning on the top sheet which as I understand can lead to minimal bending, however when I ran the sheet as a separate component I saw decent deflections. When the frame and top sheet are attached, I get contact points between not only the face of the beam and the sheet, but the rounds on the frame and the sheet as well which are not in contact with the sheet (as far as I can tell). The discrepancies between the FEA on the frame and the frame analysis lead to my original question.
I am relatively new to the FEA process, so I would just like to get some clarification on a few things if you don’t mind. I am wondering if my simulation runs into trouble because of
A) Thin surface identified on top sheet during simulation
B) Additional constraints between frame and top sheet
C) Different deflection in frame analysis and FEA/frames do not run that well in FEA
D) Something else/user error
I figured I could possibly split the top sheet into panels and put a bonded contact between them (the top sheet is 3/8 in thick). I could also suppress the additional contacts in the frame (which would take a lot of time) or if there was some way to import the frame as one welded component that might reduce the number of contacts, that would be nice. Any help you can offer would be greatly appreciated. I attached the top sheet of the floor as well if needed to finish the assembly and see what the results are. The top cover is flush to the frame and the edges are flush on the side with the cutout. The overhang on the sides are also equal (I believe making one edge flush should make the other side flush as well).
Roll up the EOPs of the part files before zipping.
Don't need the analysis files (Inventor will recreate them).
See attached image. Need only the ipt files and the iam file, no other files.
Here are the assembly files. The EOP modification did not seem to change the file size, so in additon I did not include the frame members. The frame is made of ANSI 2 x 2 x 1/4 steel tubing. The top frame can be placed twice to cover the sub structure. There should be about a half inch gap between the ends. The second attachment is the top sheet over the frame. This can be added in place of the FEA blocks that are currently in place to simulate the pressure.