Plastic redistribution around a hole in a panel

[StevenKatzeff]

Hi, I have a question:
When there is a stress concentration around a hole in an aircraft panel, I use the neuber equation to predict the plastic stresses at the hole, and compare that stress to Ftu of the material. However, the neuber relation can factor down very large stress concentration factors to very small values of stress. The plastic region is assumed infinite (my assumption). Is there a way, without using non-linear FE models, to predict the cutoff plastic stress and to ensure that the plastic region does not extend to the boundary of the panel.

 

[rb1957]

personally i've never used neuber on aircraft panels; instead of fatgiue analysis of holes, most of the work i've seen uses damage tolerance preciples.

The most i've ever seen would be net-section ultimate, but 99.99% of the time fatigue requirements dictate a fairly low static limit stress.

 

[CoryPad]

It sounds like you could use the plastic zone size, which is part of fracture mechanics. You can consult a textbook on this subject to see if it will work for you.

 

[rb1957]

IMHO, not really ... two very different concepts/applications.

ultimate tension allowable = Ftu*(net area)

 

[CoryPad]

I didn't read the OP very closely (infinite plastic region, yikes), please ignore my post.

 

[StevenKatzeff]

Thanks for the responses,

I am not worried about fatigue of the panel, only a static check against ultimate loading. I feel like I need to know the extent of the plastic region, and if it extemds to the panel boundary, then that should be the failure criterion, but how do I determine that plastic region?

 

[victorzv]

Steven,

You are definitely confusing something.

What is the goal of your analysis? Why do you worry about the plastic region reaching the End Of Part?

The Neuber's rule should not be used for the entire cross-section. It is valid for a relatively small vicinity of the notch. It is good for fatigue analysis.

For static strength, a simple net section analysis should work. For the first iteration:
Yielding allowable = Fty x Anet
Tensile strength - see rb1957's post.
Note: in aerospace, Ftu should exceed 1.5 Fty, otherwise your ultimate criteria is 1.5 Fty.
Thus, no account for a Kt.
These work for 99.99% of all needs.

For more accurate evaluation, the stress concentration can be taken into account (but this is not a common practice). I beleive, though not 100% sure, Bruhn discusses this phenomenon and Michael Niu reproduces Bruhn's charts.

 

[rb1957]

you've got a finite panel, right ?

start with the net area ULT stress

if you really want to go to town, get the stress distribution around a hole (NACA TN2073), factor for your applied gross section stress, project a line on top of the stress profile at Ftu. the stress above the line cannot happen, and needs to be replaced with area below the line. From trial and error you'll end up with a stress profile that is flattened at Ftu and has the same area (=energy) as the elastic distribution.

or you can base your MS on net section stress

 

[StevenKatzeff]

Sorry for the misunderstanding, here is the problem:

I have a finite panel 508 X 189 mm of Ali. We use a programme (Cannot divulge this info, sorry) to analyse the panel for stresses, buckling etc. However, the programme does not factor in any possible holes that are present in the panel, some of which can have a diameters of 80 mm. I take the FE results of the freebody force fluxes on the panel and obtain the Von Mises stress at the hole edge using a kt value. I then factor this elastic stress down using neuber to a final stress which I compare to Ftu. However as mentioned above, the neuber equation can factor down very large stresses to smaller stresses, and I am sure that this is because it assumes an infinite panel width. I want to check whether the plastic region does not therefore extend past the panel boundary, thus causing the entire panel to fail. I hope this explains things better and eagerly await yoyr responses to my problem.

 

[rb1957]

thx, alots of info ... i was thinking you where talking about small holes.

i take it your model was a coarse grid, without modelling the holes. since you have a fintie panel i think net section stress is still a reasonable measure. other than that i'd use a detail FEA, including material NL if it looked like it needed it.

different thought ... how will the load redistribute in the structure if this panel yields (in a big way) ? In the FE, was the panel modelled as an equivalent thickness, the holes are very large (50% of the width of the panel) ??

 

[RPstress]

Although it's common practise to use Ftu with Anet, an old Hawker Siddeley data sheet (1-41, "Effect of Holes on Nett Ultimate Strength") show that the net stress at failure can be as low as 0.8 x Ftu for L64 (2014-T3). Most alloys come in comfortably above 95% in the same data sheet.

WRT the OP, the neuber factor is not usually used for static stressing. thread2-178465 has more.

 

[40818]

Your problem is a common occurance on Airbus wings. You need to be careful as the extent of plasticity can be quite large and beyond the scope of using Neuber. There are a variety of methods and tools (not FE) which have been developed to accuractly estimate the local effect (at the hole) and also to evaluate the effect on the whole panel.

 

[StevenKatzeff]

Thank you for your posts.

rb1957, I am not sure how the programme deals with the panel yielding, I assume it redistributes the load to the support structure, especially when it buckles. We do use effective thickness of the panel based on the stiffness of the panel with a hole.

RPstress, I took a brief look at the other thread on Net-Area Tension and will go through it more carefully when I have time (which means during my lunch break).

40818, you mention that there are non-FE methods to estimate the effect on the panel, can you elaborate please.

 

[rb1957]

"I assume it redistributes the load" ... that would imply a non-linear FE. is the hole un-stiffened ? is the panel unstiffened ?? it's a pretty big panel (18" x 7") with a 3" hole and obviously high tension stresses ... initially i pictured a sheet, but thinking about it, i expect it's more complicated than that.

 

[40818]

The original methods used by Hawker Siddley (now in various guises at Airbus) use SOR25 to approximate the stress distribution around the edge of the hole cutout. Notably developed post comet disaster..
SOR51 is used to analyse the flat or curved panels (no hole) for a variety of loadings and incorporates post buckled effects, and SOR40 takes into account the effect of the hole.

As a side note, you can support stresses to about 2X FTu at the hole edge without failure.

If the level of plasiticty is beyond the panel boundaries(i.e beyond the panel support stiffeners), then it would become a lot more complicated, however, that level of plastic behaviour would probably be beyond the design requirements.