Honeycomb Failure - Continuation

[irq]

Hello everybody,

Unfortunately the following thread has been closed:

http://www.eng-tips.com/viewthread.cfm?qid=256487

and I need to open a new one.

I have a cylindrical sandwich panel (see a picture) and I want to check the core for crushing under applied pressure.

I have FE-Model in Nastran with PSHELL/PCOMP. As far as I know it is not possible to obtain the forces in normal direction to the shell element, since PSHELL has no E3 strength properties ... Does exists any method to get the core compression stress based on NASTRAN results? Many thanks for any description.

Thanks in advance!

 

[ESPcomposites]

Is the problem this simple? If so, couldn't it be calculated by hand?

If you need to run a FEM for some reason, a solid element would also calculate the stress you seek. That may be time consuming to build though if you want to define it lamina by lamina. Alternatively, you could consider a solid isotropic solution since I don't think you would see much a difference for that stress component. You will have to be the judge of that though.

Brian

http://www.espcomposites.com

 

[RPstress]

If you really need to get through-thickness info from FE then you need to model the third dimension somehow. (As ESP says, usually through-thickness calcs like Brazier stresses on curved material are done by hand. Not entirely sure what your setup is like—picture not there!)

You can do it with shells spaced apart and joined together by spring elements or MPCs, or by modelling the core as solid elements (with orthotropic properties if honeycomb) and shells on either side.

You will always wind up with quite a bit more complication.

If you want to process the simple shell FE results to extract through-thickness forces you can find the in-plane forces in the face sheets, and from the curvatures (remember to use the final deflected-shape curvature) you can calculate through-thickness values (similarly to how the rib-crushing Brazier compression is worked out for wings in bending).

If you have a purely through-thickness situation then you'll have to do it by hand or 3D FE.

 

[irq]

I'm trying to attach the picture again to complete the problem description. This is only a representative outline, not the real construction (real panel has more complicated geometry). Through this discussion I want to get some information about general principles.

RPStress: For core crushing I know this formula: Core_compression = Shell_force / Radius of final shape curvature. I am not sure if it is a correct one, since it does not contain the information to initial curvature.

Let assume e.g. very thin 40x40inch sandwich panel ply-honeycomb-ply with a thickness of 0.2 inch. For the calculation of thin panels I would say that the failure will be caused more by general instability as an effect of transverse bending and not due to the core compression. Are you the same opinion?

Thanks for valuable information to this topic.

 

[RPstress]

For your arched bridge it'll be a trade off between the bending moment causing core tension and overall compression in both skins causing core compression. That's the sort of trade that some FE would be handy for! It would probably be worth the effort of modelling the sandwich in 3D.

For the core crush it is the skin forces/inch and the final curvature that dictate the through-thickness forces. The initial curvature isn't relevant. For the direction of loading shown the curvature will be reduced, so using the initial curvature would be conservative.

You can probably do some rough numbers using arch formulas in Roark to find bending moments and compression and therefore skin forces. If the curvature is sufficient then the moment may not build up enough to put the lower skin in tension, in which case the core will definitely be in compression. Your FE model will of quantify this for you even without modelling the core in 3D. You should then be able to use the skin forces and curvatures to estimate the through-thickness Brazier forces in the core.