Bushing in Honeycomb Panel

[Emuman]

Short of testing has anyone been able to do a sufficient analysis of bearing stress in aluminum honeycomb panels?
I know I need to install a bushing to protect the honeycomb, but how do I show the load that bushing and panel can take is adequate?

 

[WKTaylor]

Emuman...

I think You are actually referring to a potted-in [bonded-in] fastener insert.

Refer to "Airframe Stress Analysis and Sizing" by Michael Niu ["honeycomb" in index].

Also helpful is SAE "Care and Repair of Advanced Composites" for general insert use/installation Guidance.

Regards, Wil Taylor

 

[Emuman]

No actually I want to mount a 0.75" bushing and pot it into the honeycomb.
I've got all sorts of data on potted inserts, I'd like a way to analyse my particular problem without resorting to testing.

 

[SWComposites]

Is the load in the plane of the panel or perpendicular to the panel? What is the bushing configuration? - does it go thru the entire panel with flanges on both sides or is it mounted on one side? Are you concerned with the bearing stress in the aluminum facesheet or in the potted core or both?

 

[Emuman]

The load is in the plane of the panel. I am concerned with the bearing stress on the core and the face sheet. The bushing is a thru type and has flanges on the upper and lower surface similar to a delron insert. The bushing will be potted with Hysol 934. I'm most concerned with the face sheet buckling and the core crushing. I suppose I could just consider the bearing stress on the face sheets, however I'm sure this would give me a bearing stress far below the actual capability of the potted honeycomb.

 

[gwolf]

The core has effectively zero bearing strength so you will have to take the load in the skins either as bearing on the hole or shear through a bonded 'top-hat'. I think you need to look at best practice and suppliers test data in the first instance, an FE model as the last resort.

 

[SuperStress]

I agree with gwolf about the strength of the core, it's insignificant in compression (bearing).

I'm not familiar with Hysol 934 as a potting compound, but let me share this:

In-plane load in your bushing will be transfered to the skins by two mechanisms; bearing into the skins, and shear between the potting compound and the skins.

The ability of the skin to resist the bearing load from the bushing is suspect. It requires the potting compound to sufficiently restrain the skin from buckling/wrinkling/ripping (assuming the skins are thin), and for everything to stay planar as load increases. In practice, they don't do this very well. You would be wise to disregard this component.

The second mechanism, shear between the potting compound and the skins is your primary load path. The difficulty is that your allowables for this are extremely process-sensitive. You have to undercut the core with some sort of fly-cutter, and have good surface prep for your potting compound to adhere well. If you know the shear allowable for your potting compound/skin combination, your shear allowable, based on area, should have a significant knock-down factor applied (like at least 50%) unless some testing can be done to determine your allowables.

My experience with the FAA is that you will need to test a minimum of 12 samples to get approved 'A' and 'B' basis allowables. More is always better.

With respect to gwolf, I personally think that an FEM is inappropriate for this application.

SuperStress

 

[gwolf]

I like doing these sort of analyses but surely there must be some simple fastener type data out there, if not then you should do some FE as SuperStress suggests.

 

[Sparweb]

Emuman,

Is this a one-off project? Are you producing a series of parts for long-terms service?

Testing may not require as much effort as SuperStress indicates, when:

a) You are prepared to overdesign the panels enough that you can load test them ALL, and

b) you are only making a limited number of them.

I've certified a few panels in my time, each one for a unique project, therefore I put them to the test, inspected for damage afterward, and released them into service.



Steven Fahey, CET