Analyzing honeycomb panels structures

[brailos]

Hello,

I would like to hear some experiences of some of you guys with the design,analysis and testing of honeycomb structures.
I am not interested in how to analytically analyze a rectangular panel or how to model it in FEM.....you can find that in every book.
What i am interested is what are you following in a analysis. And i specifically said honeycomb structures not individual panels.

For example i have a structure made of several honeycomb panels (a stowage ) and it is fixed in 4 points in the base plate with the floor. The thing is it has to withstand the ultimate load case 9G forward.

For those who don't know the condition is that at Ultimate Loadcase 9G forward the structure has to stand for 3 seconds, it must not fail completed. I am saying this because the yield limit obviously it is not the criterion of failure in this case. My question is how you can determine from analytical approaches and FEM modelling if your structure can withstand this 9G load case. Because some failure is allowed it makes it difficult to analyze it with typical static analysis. The analysis becomes nonlinear due to plastic yielding and the sandwich itself being orthotropic make things more difficult.

Is testing the structure the best and safest approach? If yes how can we improve the design as much as possible so we have better chances of getting a ratio weight/strength without to many reinforcements, inlays and potted areas.

For example , i used in my fem model in post processing the yield limit of the faces when analyzing ply 1 and 3 and when analyzing the core i used Max Shear with a max scale of 0.9 MPa - the Plate shear Strength in L direction.

I would like to hear some opinions about my approach and how it should be done. How it should be done i would like it to hear from people who actually done this in a company ,not in a university on research or etc.

Thank you

 

[SWComposites]

- simplest approach is to size it for no failure at DUL.

- if you want to allow initial failures, you are going to have to run a nonlinear analysis that can track progressive damage, for which you will need non-linear stress-strain response properties for your materials; since these are not likely to be available, see the item above.

- if you are using a FEM for interior structures analysis, the FAA (and probably EASA) are going to want to see test data to validate the FEM.

 

[brailos]

Well to size it for no failure at DUL is not that simple. I mean it is simple from a structural strength point of view, but the weight, costs and complexity of the structure would not make things feasible.

The structure has to withstand the Ultimate load case for 3 seconds, even if it bends , or the core crushes locally or any type of local failure occurs as long as the structure keeps its integrity for 3 seconds. The purpose is not to reuse the structure after that.
So yes, initial failures are allowed and yes i agree with you a non linear analysis having the nonlinear curve for the material it is a solution , but i am afraid it is not a feasible one also.

What i did was to make an FEM model of the structure, used PCOMP card, and i used the yield limit for the faces and the Shear Strength of the core in L direction. In this way we could see how the structure behaves, and made some local reinforcements, reduced the stress values as much as possible in the acceptable limit of costs, weight and complexity. We finally made the static test, and for some reasons the test was stopped at 75% of the ultimate load, and the deformations were a lot smaller then the expected ones, almost unnoticeable.
To be safer we reduced the allowed weight limit that would be stored in the stowage, so we are in the same area with the test results.

In my opinion this is still a little cloudy for me, because i don't see a simple and cheap way to make structure, and this is a big problem for small companies who do not have the money and tools nor the experience in such things.