Fuselage skin door corners 1

[egalvan]

In a static test campaing, for some critical load cases we have found that the skin at the fuselage door corner reaches compressive stress values 30%-40% higher than the Fty of the skin material (Al 2024)at limit load, so we have started to think in the necessity of introducing additional reiforcement in order to save the static ultimate test campaing.

In my personal opinion, and staticwise, I would say that local plastification at the corner shouldn't be a problem for a ductile material as the Al 2024 and what probably is going to happen is that the plastification area in the skin will increse for ultimate loads but without failure of the structure.

Anyway, I have been looking for information about sizing criteria of the skin corner of a fuselage door but I could't find anything interesting.

Could somebody help me on this?

 

[rb1957]

1st, the rule 25.305a states "The structure must be able to support limit loads without any detrimental permanent deformation." which i imagine is true in your case.

2nd, stresses this high at limit load suggest more trouble at ultimate load. it is unusual to exceed fcy at limit load. to exceed it by a (un)healthy margin and not to reinforce it would be brave indeed, very brave.

my 2c ... as the skin is getting highly stressed, maybe load is being redistributed into other parts of the structure (longerons, stringers). the skin can buckle and recover, but if the stiffeners cripple or collapse, then the whole house of cards falls down !

break the test specimen and you're down for months (i know for experience !) ... take a week to quickly design a tripler for the corner, or additional stiffeners.

personally, i'm surprised you're in such trouble in compression and the tension corner is fine ??

As for design guidelines, it really depends on your company's experience ... some people (Learjet) design square corner door cut-outs, other (BAe-125) have nice round ones ... either one works with enough Al.

 

[egalvan]

Thanks rb1957 for your answer.

Although permanent deformation is expected according to the residual compressive strains in the strain gages after unloading, visual inspections in the skin corner after test did not reveal any damage or significative deformation in the area, and therefore rule 25.305a has not been violated.

The four skin corners of the door are already reinforced by two (one inner and one outer) doublers. the outer doubler is smaller than the inner and it is exactly there where the outer doubler runs out where that high stress in the skin appears. It is obvious that one possible solution is to extend this outer doubler but since this is also a really complex repair because of all the different parts attached in this area we need to think twice before going for this or similar repair solutions.

About the load redistribution to surrounding elements, after 0.6 limit load was reached in the test, we decided to add additional instrumentation in the elements adjacent to the door corner, i.e, the lower sill beam, stringer and longeron and to our surprise we found out that till 1.25 LL (target load for this test) the stress in those elements remained clearly linear while the stress in the skin corner remained linear only till 0.8 LL (=Fcy) showing a clear non-linear trend after this load level. In my opinion this is showing that even with a really high plastification in the skin corner ( and stress of 492 Mpa was reached for 1.25LL assuming linear behaviour) no skin load is being redistributed to the surrounding elements.

Staticwise, the corner with the highest stress levels of the door is the rear lower one. This is because the down bending case for the rear fuselage it is much more severe then the up bending.

If you were doing a design of the reinforcement doublers for the skin corner from the scratch, and from a pure static point of view, would you size the doubler/s in such a way that your maximum stresses at the skin corner were < fcy or fty at limit and < ftu at ultimate?.

 

[rb1957]

i would design for <fcy at limit ... i realise i don't Have to, but it is slightly more conservative and you're already in an unhappy place ... adding doublers is a resonably inefficient way to reinforce a cut-out corner in part cause you can't attack the problem (the stress near the edge of the cut-out), you can only get close to it (the dblr is only effective at the fasteners, 2D for the edge, which is a long from the stress peak). this is an unusual design problem, having to worry about a corner in compression. something to think about, you want to avoid compression plasticity as it creates tension stresses in skin panel as the load is relaxed; similar (but opposite) to exceeding fty.

i'd design a fastened dblr as a repair to get the test going and for immediate production. for future production i'd look to thicken the skin panel, at least near the corner (an integral dblr) ... machined skin, (hot) bonded doubler. a word of caution, light stiffeners on heavy skin are a bad design ... the stiffeners can't support the skin after it starts to buckle; you need to be really confident that the stiffeners can support the buckled skin, particularly where the stiffeners have cut-outs.

good luck, you've found a challenging design problem !

 

[egalvan]

it looks to me that we still have a little bit of work to do to have this issue solved...

Thank you for the tips rb1957.

 

[40818]

One single word springs to mind here.....COMET.......

Post comet paranoia resulted in reams of testing and research into the localised edge stresses around circular and rectangular cut-outs. Basically, you can carry local stresses to a level of just over 2 x Ftu (or beter put just half the stresses). Carry out neuber stress corrections up to strain failure levels to get more realistic stress levels.

Hatfield stress reports came to the rescue of aerospace companies over 40 years ago or so with this one.

 

[WKTaylor]

egalvan and guys...

"...skin material (Al 2024)..." is NOT a material definition of any use for this forum.

Per MMPDS-03 2024-T3 or -T42 has a substantially lower Fcy than -T361, -T62, -T72, -T81 or -T861.

And bare material [AMS-QQ-A-250/4] has slightly higher allowables relative to clad material [AMS-QQ-A-250/5].

SO... egalvan... what REALLY is your material-temper for the skin??????????

NOTE: A simple combination of temper change and next higher thickness [OK, chem-milled for weight if needed] might get You "over-the-hump".

NOTE: I have dealt with many post-grad researchers. This is a very annoying omission in most of their scholarly presentations. They will say "AA2024 plate" which is about as ambiguous as it comes. They often over-look temper, spec, manufacturer... and quantifying ACTUAL material properties, etc... before manipulating it [IE: FS-welding, corrosion testing, fatigue testing, etc...].

In one case a researcher had a block of 6Al-4V Ti material which he had hot-worked [crushed] to the thickness of a plate.... and then claimed it was "plate" for his tests. He said that it was then annealed... but had no heat treat records to prove the claim. Why did he do this extraordinary work??? Because the block of Ti was given to him "free" and he decided to "make-do" with university resources for processing it to "plate" thickness. His research paper and presentations were very scholarly... but his data (in my humble opinion) was bogus for any real-world application: the material met NO existing production standard for Ti plate... it was 1-off home-brewed piece of material. Grrrrrrrrrr.

Sorry... carried away... I'll stop...

Regards, Wil Taylor

 

[egalvan]

wktaylor, FYI, skin material is Al2024 T42 Clad sheet. Sorry for the annoying amateur omission...

 

[WKTaylor]

egalvan...

I assume You have the skin stretch-formed while in the "AQ" temper... then allow it to age-harden to -T42.

Refer to MMPDS-03 for 2024-T42 and -T62 clad sheet allowables. The Fcy for -T62 is ~28--33% higher than the -T42 Temper [depending on L or LT orientation].

This implies, that skin in -T62 temper [and any doublers and stiffeners] will have a substantial boost in static compression strength [and Fty, Fsu, Fbru, Fbry, etc]. If fatigue durability [fatigue cracking] is not a major issue, then this may be a solution. If durability is an issue, then increased thickness with chem-milling might work... etc.

NOTES.
a. If You have mild compound forming [less than 5% permament strain], You may want to consider stretch forming in -T3 and then aging to -T81 for even higher mechanical properties.
b. Also... You might consider clad 7075-T62 [thin sheet, form in "W" temper and harden to -T62].

CAUTION.
Actually having mechanical tests done on sample sheet materials is advisable: this will leyt Yoiu have a "real-world" look at the final product. If You have high quality materials and processes, You may be able to claim slightly higher mechanical allowables [for specific sheets and controled fabrication processes].


Regards, Wil Taylor

 

[rb1957]

i think the OP has a static problem but there is still an underlying fatigue requirement (being a fuselage).

2024T6 and T8 and 7075T6 would be unusual choices for a fuselage skin; admittedly will does say "if fatigue durability is not an issue".

2024T3 or T4 are used in just about every pressurised fuselage for a reason

 

[WKTaylor]

Sorry guys about my soap-box tirade, earlier... I usually have more gentle-tact in my approach...

I had one of those days where each problem I was handed had "missing basic facts and/or had poor definition".

In one case I had to finally (self-define) baseline my assumptions up-front... and proceeded with a abstract analysis. I guess I was "close-enough" with my answers to satisfy the field engineer...

In 3-other situations, I had to ask several basic/up-front questions, such as part/assembly number?, T.O. Figure?, who is asking?, DLA supply status?, etc...]... just to get started.

Ahhh... mentoring never ceases...


Regards, Wil Taylor

 

[rb1957]

vent, it's good for the soul,

and virtually venting is better