Damage Tolerance Analysis for Composite Structures

[edmeister]

Lately I have several FAR23 aircraft (Composite construction) (Post Amdt. 45/48) certification that I have to demonstrate compliance with FAR23.571 & FAR23.573 ..
We are not dealing with complicated situations - just a few holes to allow the mounting of Antennas - never the less the analysis has to be provided. I will not be so bold as to ask for a sample analysis; as i know a number of participants of this group will spout out proprietary info .. etc & distract from the objective of the question.
Lets just start with a link to any references or discussion / experience that individuals may be aware of that could be useful.
Anything would be appreciated ! thx.

 

[rb1957]

"spout" ?

Is there no substructure supporting the antennas (or are they mounted directly onto the skin) ?

Composite analysis is going to be a PITA. You've done the static analysis ? vibration ??

Composite DTA is an even bigger PITA. How can you say (the easiest solution) that these holes are less critical than other similar locations ?

The simplest DTA approach I can think of is testing from a detectable crack, hoop stress cycling, some conservative factoring,

Of course if you have access to the OEM data then this is much easier, but then you'd've said so if you did (and probably not asked the question).

"Hoffen wir mal, dass alles gut geht !"
General Paulus, Nov 1942, outside Stalingrad after the launch of Operation Uranus.

 

[edmeister]

Neglected to add that this was an pressurized aircraft and additional plies or Doubler would be added to compensate for the .75" Dia hole.
.. also since this is non-pressurized; what is a comparable cyclic loading to calculate the crack growth?
PITA ???

 

[rb1957]

Pain In The A$$

In the non-pressurised portion of the fuselage I'd use hoop stress as a simple, clearly conservative, 1/flt load.

I assume Carbon composite ?

How does the OEM inspect the pressurised fuselage ? tap test ? Is there a feature in the pressurised fuselage that you can use as similar to your hole ? (hole size, ply lay-up ?)

"Hoffen wir mal, dass alles gut geht !"
General Paulus, Nov 1942, outside Stalingrad after the launch of Operation Uranus.

 

[SWComposites]

Ok, this is for a pressurized composite fuselage for a bizjet type aircraft? Obviously not transport a/c or you would be referencing Part 25. Or an unpressurized fuselage?

So, DT “analysis”? Presumably this means fatigue and residual strength, no?

First, what exactly are you doing (analysis + data) for basic ultimate strength? You are adding holes? To pressurized or unpressurized area? Is there an SRM that give allowable hole sizes bigger than your additional holes? If so you should be able to claim magic similarity and be done.

Otherwise, yep, it can be a PITA. Give a bit more details on what data you have for the existing composite structure.

References:
FAA AC 20-107B
CMH-17 Volume 3 Rev G Chapter 12 (buy from sae.org)

If you happen to be a member of ASTM F44 committee, Structures subcommittee - I recently posted to working area a revised draft guide for composite DT.

 

[SWComposites]

Oh, and we (the industry) generally does not do crack growth craziness for composites. Cert is typically done by demonstrating “no growth” of damages. See the references. And no growth is based on test data containing holes, notches or impact damages.

 

[edmeister]

Appreciate the replies. RB1957 input about employing Max Alt press load would be a conservative hoop stress value at 1 cycle/ Flight.
This is for a Cirrus SR20 where its required to add a second .75" dia hole 1.5" away from an existing .75" dia. hole. The location - cabin upper crown is defined as a 'stay-out area' - thus requiring more then an SRM application. My brief review indicates that the layers are most likely glass - but this exact location material composition is not defined in the AMM. The only inspection methods defined are general & detailed visual. Replacement of the removed material cross-section would consist of several plies & an Alum backing plate. (all internal)
- I would would be surprised to see any crack growth in this situation - but i still have to show compliance to FAR23.571(c)

 

[SWComposites]

Doesn’t SR20 fuselage have a longitudinal splice at the crown and belly?

Hole spacing - 1.5 inch edge to edge of the holes? Or center to center?
Edge to edge means you have a w/d = (1.5 + 0.75) / 0.75 = 3 which is small.
Center to center means w/d = 1.5 / 0.75 = 2 which is crazy small.
Any doubler you add is likely going to have to cover both holes. If there is a splice joint there, then its going to get real complicated.
Presumably this is on aft part of the fuselage, so you have longitudinal bending + torsional shear loads to deal with.

So to deal with DT, assuming without any test data, seems you will somehow need to show the stress levels and Kt with the new hole and doublers are no worse than the existing stress and Kt for the existing hole.

Seems it might be better to space the new hole much farther away from 5he existing hole.

 

[rb1957]

I thought the idea was to bond some extra plies (to the IML ?).

Are you going to have a DER look over this ? or a field office ?
whichever, I'd be talking to them to see what they expect.

Have you tried talking to Cirrus ?

"Hoffen wir mal, dass alles gut geht !"
General Paulus, Nov 1942, outside Stalingrad after the launch of Operation Uranus.

 

[LiftDivergence]

A good reference I know of is Vassilopoulos & Keller - Fatigue of Fiber Reinforced Composites.

Similar to metallic fatigue and damage tolerance although there are semi-empirical aspects, most analysis is heavily reliant on test data.

As SWComposites alluded to, often time OEMs will conduct cyclic test of coupons for their ply stackups which include intentionally induced damage to see if there is any "growth"

What happens with composites is not really flaw growth in the same sense as a metallic crystalline solid... instead, "damage" tends to be delaminations which propagate.

You have to be careful, because you might initially think, adding some additional plies to reinforce the cutout statically should take care of the stress. But that does not necessarily substantiate continued airworthiness because now you are changing the ply arrangement and the symmetry of the laminate. So you can't really guarantee that flaws are less likely to occur without conducting cyclic tests for your repair laminate.

For something like a 787 with a composite fuselage, there are standard repairs in the SRM, but I have to assume Boeing has qualified these repair stackups via cyclic testing.

This is a major drawback of composites... they tend to have more complicated repairability and make continued airworthiness harder to substantiate in post-production.

Keep em' Flying
//Fight Corrosion!

 

[SWComposites]

Yep, there were lots of static and fatigue tests of repairs for 787. And for A350 and previous models that used composites for primary structures.

Bonding additional plies on the IML will likely be viewed by the FAA as a bonded joint. So see the bonded joint sections of FAR 23.573 and AC 20-107B. Likely need to show ultimate capability with detectable disbond sizes, and limit capability with the bond failed (up to the location of fasteners). It might be easier to substantiate a fully bolted doubler patch.

 

[Sparweb]

Isn't there an ESDU paper that allows you to determine the Kt of holes in orthotropic plates?

 

[edmeister]

Happy to see that lots of useful information has been provided here.
But can we restrict this to FAR23 aircraft (unpressurized) & keep it moderated.
.. unlikely that someone who just wants to install his GPS antenna will want to dish out for a fuselage mockup & a full series of cyclic testing.
My objective is to obtain sufficient information in order to submit an acceptable report to a FAA field office.
- Composite construction in the past was 'assumed' to be DTA -free .. but now with the advance of technology; it now requires us 'old hacks' to learn new things..

 

[rb1957]

you need to talk to the DER, or whoever is making the finding.

This is going to be either ...
1) "PBC" ... passed by comparison with some "story" about how it's not the end of the world, or

2) a mountain of work, 'cause the DER says "Prove it's good".

You may need to talk to Cirrus, unless you know the ply lay-up in the pressure cabin and in your unpressurised tailcone. If the same then you have a "sensible" story (that your unporessurised fuselage is much lower stressed than the pressure cabin( and this gives you some room to manoeuvre.

You say there's a similar OEM? feedthru nearby. You can show that unless the two holes are Very close together then the Kt is still reasonably 3 (there's an ESDU with Kt for multiple holes). Then your installation is not much different to the OEM's. Check that the OEM didn't add plys.

Is this a small L-band or a large VHF antenna ? If VHF, I'd much rather see the antenna supported by internal structure (a U channel, frame to frame; doesn't need to be riveted to the skin (fayseal would be enough).

"Hoffen wir mal, dass alles gut geht !"
General Paulus, Nov 1942, outside Stalingrad after the launch of Operation Uranus.

 

[edmeister]

RB1957 .. to simplify the story even more ..
Cabin is non-pressurized & antenna is a 17" slim line COM-DAT
- will add couple plies & a .050 alum plate so satisfy any 'non-believers' & provide addition stiffness internally.
Antenna (CI-2580-200) just happens to have an additional BNC connector 1.5" away from the existing OEM hole.
- replacing 1 type of antenna for another ..

 

[rb1957]

well that's a pretty "wimpy" antenna !

So you're replacing an OEM antenna with this, and need to add a 2nd hole. Picking up OEM provisions ? for a similar antenna ? well, that's a good story ! (at least the beginnings of a good story) ... "this may not be the end, or even the beginning of the end, but it is at least the end of the beginning" ?

have you got the Kt effect of multiple holes (should be very small, if there is even 1D between the hole edges).

"Hoffen wir mal, dass alles gut geht !"
General Paulus, Nov 1942, outside Stalingrad after the launch of Operation Uranus.

 

[SWComposites]

ed - for us to help you more, please answer the questions above and give some more details of dimensions and layups. Also, do you have a DER involved? Or any help from Cirrus?

 

[MNLiaison]

ed,

As SW stated, the SR20 fuselage is made from two halves bonded at top and bottom with paste adhesive. Depiction in this article. 99% positive original material is 3M SP381 resin with 7781 glass weave fabric and S2 glass tape. 'B' basis properties for these two were part of AGATE and reports are listed below. You'll find Cirrus listed throughout.

7781 / SP381 - E-glass Fabric, AGATE-WP3.3-033051-098
S2 / SP381 - Uni tape, AGATE-WP3.3-033051-099

Good luck

 

[rb1957]

number of plies ?,
orientation ?,
BVID ?,
loads ?? ...

"Hoffen wir mal, dass alles gut geht !"
General Paulus, Nov 1942, outside Stalingrad after the launch of Operation Uranus.