Understanding a LISA model of a flat plate

[Japandrew]

So, I'm working on a little project to play with and understand LISA. Essentially the project was to see if a 3/8" plate could replace a 1/2" plate as a base plate for a stand. The plate would be anchored into the ground where the holes are. There is a welded channel that extends vertically from the plate. I set it up as best I could (see pictures below), but for some reason the stresses I get from LISA say they are in the hundreds when the forces are in the thousands. Maybe I'm interpreting the information incorrectly, or maybe I set it up wrong. My goal is just to make sure the stresses remain below the yield for the plate. Any help understanding what I'm not getting would be greatly appreciated.

VM of plate

http://files.engineering.com/getfil...d-45f1-9dc6-9ec40e4a900f&file=VM_of_Plate.PNG

Mesh of Plate

http://files.engineering.com/getfil...42df-9b7a-b3ee85982c1b&file=Mesh_of_Plate.PNG

Material Properties of plate

http://files.engineering.com/getfil...8556a4b&file=Material_Properties_of_Plate.PNG

Force on plate

http://files.engineering.com/getfil...648-b442-8ceafb702f15&file=Force_on_Plate.PNG

Distance from force to the hole

http://files.engineering.com/getfil...3b3&file=Dist_from_force_to_hole_of_plate.png

Dimensions of plate

http://files.engineering.com/getfil...-4753-a897-e95b4b2d114e&file=Dim_of_Plate.PNG

LISA analysis

http://files.engineering.com/getfil...0d-e4ff5506ee80&file=.375_plate_analysis.liml

 

[corus]

I can't view the results but from what you say you don't seem to understand that forces and stress are different. As an aside, don't use tet elements as they give poor results. You'd be better using brick elements with several through the plate thickness.

 

[ThomasH]

Hi
Since I don't know LISA it is difficult to have an informed opinion. The difference in "hundreds" and "thousands" can be due to units, does LISA have units built in?
If it is a flat plate with pure bending you should be able to calculate (by hand) the stresses if you have the moment or vise versa. I would start with that calculation.

Good Luck

Thomas

 

[whitwas]

The points corus and ThomasH mentioned are both very common problems that people have with FEA and both apply here. A bit of elaboration:

LISA is unitless so you have to use a coherent unit system. Your material properties have Young's modulus in psi but density in pounds/inch^3. One is pound-force and the other is pound-mass which aren't part of the same coherent system (with inch and second) so you can't use them together. The ugly looking lb.s^2/in^4 is the coherent unit of density to use with psi. Having said that, it doesn't look like you need density here. A more serious problem is that the thickness of the plate is 9.525 inch, rather than 3/8 inch. This will be another unit problem importing the STEP file. You can open your .liml file with Mecway (see Google) which allows you to specify common units without converting by hand and keeps length units from STEP files correct.

An FEA mesh that looks the right shape will often give completely wrong stresses. That's because accuracy increases with mesh refinement. Your mesh is very coarse for the element type (tet4). If you refine it, you'll probably find the stress increasing significantly. But you'll be refining a long way to get to the correct value because 4 node tet elements are pretty useless. Change them to 10 node tets (quadratic elements) which are more accurate with less refinement.