Pressure on Flat Plate not Matching Hand Calcs

[HouTxMechPE]

I have been trying to model a hydrostatic pressure inside a shipping container for some time now and have been having on again, off again luck. Recently it's been off, even the mesh has been finicky. To back up and check my progress I decided to simply load up a flat plate of equivalent thickness based on matching Moment Area of Inertia values and the deflection in the center of the panel is very low compared to hand calcs and what we are seeing in the field.

The equation in Roarks that I am using is for a simply supported rectangular plate with a triangular pressure distribution, water at a height of 2000mm etc. I am only get a few mm's of deflection in SW, but by hand it's closer to 40mm. Trying to figure out what I'm doing wrong. Even applying the load uniformly across the plate does not increase the deflection much. I have selected the edge of the perimeter of the plate, not the face for my supports.

 

[ShaggyPE]

Make sure your hand calcs are intended for non-linear use (i.e. large deflections). Once that is verified, then you know your hand calc is at least usable. If your handcalcs are comparable to what you are seeing in the field... then you should be good. If it is, then you will need a non-linear analysis in SolidWorks. If you are using a non-linear analysis and not getting appropriate deflections, confirm that your material is correct and that your boundary conditions (restraints) are valid. Your calculation is simply supported, so your restraint should allow rotation (not be moment bearing).

-Dustin
Professional Engineer
Pretty good with SolidWorks

 

[HouTxMechPE]

Thank you for the information. Is there a way to select linear vs non-linear analysis when running a Simulation? I know that if the deflection is over a certain amount a warning notice will pop up asking if I want to enable "large displacement analysis" but I am not aware of any way to manually engage it.

As for the restraints, I can not select an edge to add a roller/slider support so I had tried mating a narrow strip of the same material on the face opposite the pressure to provide a fixture. This seems to work on panels around 2500mm in length but a container sidewall is double that. Simulation fails without using a fixed edge support.

 

[kellnerp]

Can you put an image of your test case up here in png format with the boundary condition display turned on.

There is a setting in the run properties for a linear analysis that takes into account geometric stiffening. Out of plane deflection can radically change the response of a membrane to pressure loading.

I have this check box:

TOP
CSWP, BSSE

http://www.engtran.com

http://www.niswug.org

http://www.linkedin.com/in/engineeringtransport

Phenom IIx6 1100T = 8GB = FX1400 = XP64SP2 = SW2009SP3
"Node news is good news."