Tipping analysis for mulitple stack items....

[muddog2]

What is the best approach to calculating the tipping force for a stack of multiple items? We have been asked to design some Layup Mandrels that can be stacked on top of each other for the Autoclave. The mandrels are supported by outriggers with extension legs to hold them off the surface below them. It hasn’t been determined how they will move them yet – they just want us to analyze the outriggers for tipping with a stack of 8 (truthfully, it’s the seismic analysis that worries us the most…..).

I have no issue doing an analysis with a single rigid body, but what factors should I include for multiple items stacked on each other? There will be a spud to provide some lateral control, but no means to bind the stack together other than gravity. Any advice is appreciated.

Thanks.

 

[cloa]

Why not make some physical scale models of the mandrels etc. with similar surface roughness, geometry, likely maximum mispositioning of the pile, weight and see what results you get?

 

[GregLocock]

I don't know of a hand calculation method, although you might want to check out how masonry pillars and gothic cathedrals were designed.

You could use software like ADAMS or Working Model to create test cases and see what works. These can model mechanical assemblies to any degree of complexity, and are specifically coded to deal with friction and contact.

As an example of the latter, here's a recreation of a Mythbusters experiment -

http://www.geocities.com/greglocock/gallery/Myth2.zip

Cheers

Greg Locock

SIGlease see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[GregLocock]

Here's a thread about a similar problem

http://www.eng-tips.com/viewthread.cfm?qid=228209&page=2

Cheers

Greg Locock

SIGlease see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[Twoballcane]

In a general sense, if you had boxes stack one on top of each other and friction was just so that the boxes will not move respectively from each other you would first calculate the center of gravity and then draw a free body diagram of the center of gravity and the point where you think another force may tip it over (horizontal, vertical, or at an angle). Then find the moments at the corner where it will tip over. From there you should be able to calculate the tipping force.

Tobalcane
"If you avoid failure, you also avoid success."

 

[muddog2]

Thank you all for the input. The mandrels are made of aluminum but the extension legs are made of A36. It should be rather rigid but it would still have to be viewed as a jointed vert beam. I agree with GregLocock that the best approach would be a simple beam FEA model. The extension legs are a solid weldment from one side to the other and if I model these weldments with compression only joints, it should cause separation when tipping begins.

The only reason that I don’t just do a free body diagram is that I am concerned about the interaction of the 8 different bodies on top of each other. If the stack only weighed a few kips I would worry so much, but the total weight could get to 80 k and that’s a lot of hurt if it falls down.

Greg

 

[GregLocock]

A linear FEA won't capture the instability, but I suppose you could use it just to find when one of the reaction forces falls towards zero.

What does your seismic event look like? Just a lateral half sine pulse to the base of the stack?

Cheers

Greg Locock

SIGlease see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[GregLocock]

Had a bit of a look at this. Interestingly, for a stack of disks maybe three times as high as the OD, they can withstand a bigger lateral bump at the base than if they were all rigidly connected.

It's the old pulling the tablecloth out from under the wineglass trick.

That is not a general result, but it does indicate that the complexities are there.

Cheers

Greg Locock

SIGlease see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[muddog2]

Greg,

You are correct that the linear analysis will not fully capture the instability from the pulse at the base of the stack, but it will tell us if the reactions on the opposite side begin to approach zero. If does get close to an uplift, then I will re-create the model in a different software and run an MES on it.

Actually, after considering your check on the disks, I think I will still run the MES analysis just to prove out the customers desire to stack that tall.

Thanks again..

Greg D