Need help with FEM analysis of assembly and some fixtures

[Adaware]

For a school project i need to do a FEM analysis of a frame that is essentially a marine lift for boats on wheels like this one:



I have drawn a frame as a solid with simulation in mind because i belive it makes it easier and i have no fillets and extravagant stuff on it so the frame is simple. Also i have just drawn half of it to make the FEM analysis faster.



On the underside of the lowest framerail there should be wheels, but since i don't want to do simulation with them i have drawn pilars going straight down to the point where the edge of the rim should be.



This is all good but i now need the correct fixture to simulate the contact between the wheel and the ground. Therefore you can se a uniball type of thing on the end of the pillar in a lighter shade of graycolor.
The idea is to make a roller slider on the underside of this and make the uniball flexible. In that way the frame can move in every way it wants exept for downwards (samme direction as down on the picture)

I have set the component contact to non penetration as i belive it should, but how do i go about with the rest to make the simulation correct?

Have tried the different tutorials, but can't find relevant stuff.


Picture of the forces (purple(?) arrows) and the fixtures (green arrows, only roller sliders):

[tosjduenfs]

Can you attach your solidworks files so I can take a look? Is this supposed to be one half of the crane? Why not mirror it and simulate the entire crane?

[Hochiki]

you should mirror the structure and analyse the whole thing as that front section will bend outwards and effect the angle of the uprights and effect your wheels.

As for fixturing your wheels, your really going beyond the limits of the Solidworks analysis software really- its only intended for basic FEA, you need to use a proper FEA package which would normally be third year mechanical engineering degree level- probably more than your school expect you to do! I would personally do exactly as you have said you have already done.

You are already making some compromises by making the whole thing as one part rather than lots of parts bolted together (as it would be in real life but Solidworks won't let you test an assembly) and you've missed out the radii which make a bit difference to the strength of a component as sharp corners cause massive stress concentrations... so I am sure what you have done already would give you the closest result to what you can achieve on Solidworks.

From my experiences with FEA and engineering reports- as long as you state your assumptions and compromises in the report and blame it on the software, the marker can't fault your results or your efforts!

[JerryFlyGuy]

I haven't gone through the description completely [no time] but I will say that your headed in the right direction, possibly too much allowance for freedom [the balls and what not]... I'd just make sure to use wheels that are rated properly and model everything above that level. Btw, DON'T mirror the structure.. you can tell SW that the model is mirrored and it will do the math in assuming it's mirrored w/out having to mesh twice as much model.. your results will be just as good and the time to run will be half.. anytime you can do FEA w/ the model only a segment of the full model your going to save hours of time.

Fwiw

Jfg