Shear Web Buckling - with holes

[Lescombes]

Hello learned aeronautical community,

I have been asked recently to allow the creation of a wiring loom penetration in the shear web of a fuselage keel beam in a FAR23 aircraft for which we are the OEM. The proposed penetration is circular, 2" in diameter and located on the beam's neutral axis. The keel beam is around 7" high and the distance between frames is 5.25" in this area.

Problem is that two existing penetrations, of 3/8" and 7/8" diameter, both unflanged and unreinforced, are also in the same shear web. They are not located in any sensible position; i.e. they are off axis randomly placed in the upper right hand side quandrant (if I may call it that) of the web.

I've searched high and low for a paper, method or otherwise evaluating the effect upon a shear web's initial buckling stress of 'random' hole position combinations, which has (unsurprisingly) been unsuccessful.

May I inquire as to whether anyone is aware of (a) some sort of paper or (b) whether I can try and use a linear superposition of other known solutions or (c) I'm headed down some sort of testing/numerical analysis path?

Any input would be greatly appreciated.

Thank you,
Greg

 

[Dougt115]

FEA would be my path.

 

[mmggppmgp]

What I've used in the past was ESDU 75034. The caveat is that it is for square plates with central circular holes. However perhaps you can include conservatism to compensate?

Otherwise I would agree with Dougt115 your only other option is FEA

 

[rb1957]

personally, I won't have faith that FE could answer this problem, I'm not sure you can define the problem sufficiently for FEA. for example, do you know the shear in the web already ?

personally, I'd suggest adding stiffeners to the web, maybe quartering the panel, clearly increasing the shear buckling allowable, and add your hole into one of the panels. adding two per side would be possibly overkill but you could nicely bracket your new hole with stiffeners on 4 sides.



another day in paradise, or is paradise one day closer ?

 

[Dougt115]

FEA may not give you "the" answer but you can use it to play with the model to determine relative stress based on where the holes are and the sizes. Simply use the current design and model it as best you can to create a baseline. Then add the hole where you want it and see how the stresses have changed. Move and resize or even reshape the holes for evaluation. You should get a good idea of the feasibility of your options.

 

[SWComposites]

If all you are concerned about is buckling, then FEA is the way to go. But it needs to be calibrated. Dig up the classical solution for shear web buckling and use that for the case with just the 2" hole. Next construct an FEA model and adjust the loading and boundary conditions until it matches the classical hand solution for buckling stress. Then add the extra small holes to the FEM and get the buckling stress for that configuration.

 

[Andries]

Lescombes,

This is an issue that is still being researched, especially using FE analysis. I agree with SWComposites' approach.

Andries

 

[Lescombes]

Thank you to all for your input.

I did feel that FE would be a common suggestion, and being inherently conservative I went along the lines of rb's suggestion and added some additional structure to the web. One motivating factor was that - in answer to rb's question - the shear in the web is not well defined as the original FE model produced in house was a very coarse loads model only.

I also appreciate SWComposites's post regarding calibrating FE models: all too often in my (albeit short) career to date people have blindly trusted FE numbers without having had the model validated or calibrated.

I do thank you all again and often wonder how nice it would be to have another lifetime to be able to devote to learning all one could from read eng-tips posts...

Cheers,
Greg