fatigue life of dimpled sheets 3

[EVdave13]

Regarding the fatigue life of dimpled riveted joints, I know I have seen information somewhere long, long ago (in a state far away), that said it was worse than machine countersinking, but I can't find it or remember what I saw it in. I'm thinking old Douglas SRM's???

Anyway, we dimple every so often here for thin sheets that can't be countersunk, and I was wondering if anybody had any readily available information on the subject.

Does anybody else have info. or, preferably, a good reference on the subject?

Thanks!

 

[A380istoobig]

Dimpling is not a recommended practice. It produces stress concentrations and can cause warping. However, a good source of info is if you can get your hands on a Boeing classic SRM and see 51-30-07.

 

[crackman]

Dimpling may not necessarily be bad so long as installation is performed correctly without overbucking and cracking the sheet. There is quite a bit of data on fatigue of dimpled joints. I dont have the numbers on hand, but just go to the NASA tech report server and search under the old NACA reports. Lots of data on dimpled sheets there. One added fact with dimpled sheets is that if you do get a crack, it is not hidden by the fastener head. As I recall, per the NACA test data, the cracks occurr right at the bend at the edge of the fastener head. So even if the Kt is a bit worse, the cracks are not hidden. Anyways, hope the data helps you.

James Burd

 

[EVdave13]

Thanks, guys. I'll try to check those sources out. In the meantime, keep 'em coming.

Thanks again!

 

[WKTaylor]

EVdave13...

Dimpling is USUALLY done incorrectly... and has gotten a bad "rap". Lots of acft have been built and flown for decades with flush-dimpled construction!

Dimpled holes MUST start with a small match-drilled hole(at 1 if not 2 sizes smaller) in all sheets. Deburr all hole-lips to prevent cracking when dimpled [labor intensive]. Dimple the holes [separately or gang-dimple] for small diameter tension head. Reassemble structure with next larger diameter clecos. Drill-enlarge the thru hole to final diameter [preferably nominal rivet Dia + 0.004--+0.008]; then install a shear-head [NAS1097 style] rivet.

NOTES:

The hole lip in the dimple gets stretched and has a tendency to crack: deburring is critical!

The dimple dies must be polished and have a specified radius for the sheet alloy/thickness to be dimpled. Precision fit/alignment of the male and female dies is critical. Also “best dimple dies” will have a significantly undersized pilot tip [with a tiny pilot hole required]… to allow the [noted] clean-out drilling to size when structure is reassembled.

Drill-enlarging the rivet hole to final diameter. This will “mostly”: (a) eliminate stressed metal; (b) realign hole center [best done with a countersunk drill guide]; and (c) eliminate jagged steps from each dimpled sheet-metal layer that limit driven rivet gripping/clamp-up. A “perfect”, ready-to rivet dimpled hole will “look” like a countersunk hole [straight thru]!!!!

Generally, all thickness of 2024-T6/-T8 and 7075-T6/-T73 (etc)... and 2024-T3, 0.063+ thick sheets... will require "hot" dimpling to be successful… and absolutely require enlarging the hole to clean-out streched metal. Even Cold dimpling 2024-T3 0.050-in(+) thick sheet metal will take a lot of mechanical effort… and still need “care” to avoid splitting the hole-lip or cracking the dimple radius.

Generally dimpled holes should have ONLY driven-rivets or tight-fitting blind rivets. AVOID installing Hi-Loks, flush-head screws/nuts, etc… unless the bottom sheet is thick and countersunk to match top-sheet(s) dimple(s).

Using this approach… as tedious and specialized as it is… will produce highly durable dimpled construction.


Regards, Wil Taylor

 

[cmmuell]

A late posting on this issue - I ran across a study done at the Hughes Tech Center ... DOT/FAA/AR-98/33 - Uniaxial and Biaxial Tests on Riveted Fuselage Lap Joint Specimens

http://aar400.tc.faa.gov/acc/accompdocs/98-33.pdf

It states that tested panels which were dimpled had 1/4 the life of equivalent, un-bonded machine countersunk sheets. I expect methods and hole sizes were carefully controlled in the specimen fabrication, as usual (thus, the initial condition may have a dramaticly negative affect, but no better than 1/4 of a machine countersunk joint).

As well, they noticed remarkable increases in fatigue resistence with bonding. For those requiring a dimpled skin, perhaps a cold bond would give an added margin of durability you're looking for(better yet - hot bonding to ensure a reliable and durable bond line is established).

Best Regards All,

Craig Mueller

 

[WKTaylor]

Cmmuell..

Thanks for link to the report "Uniaxial and Biaxial Tests on Riveted Fuselage Lap Joint Specimens".

As a designer and M&P guy, I was unnerved by the comparison of apples and oranges... and the conclusion that apples would be inferior since they had a thin skin and would be easy to puncture. There were so many discrepancies and inconsistencies in this test program that I shudder to think it was technically acceptable to the FAA.

A few examples.

Dimpled configuration was arbritrarily accepted as a 2-row staggered splice... whereas the NAS 1097 joint was as 3-rwo staggered and in-line layout.

ALODINE [Chem Film] is the surface prep of choice for thin sheet metal. ANODIZE as used is NON-STANDARD for thin sheet metal. By definition, anodize, even very thin anodize is a known surface embrittling agent. also it was unclear as to when the anodize and primer were applied... IE: before fabrication or after fabrication [IE: before or after dimpling and/or countersinking].

How were the dimples made? Hand or machine, cold or hot?? single or "ganged"?? Undersized pilot hole drilled thru to final ID then riveted?? US or nominal diameter pilot hole dimpled then riveted?? extra long rivets over-driven in sloppy holes?? Post dimpling visual inspection [or better yet, penetrant NDI]??

The 0.2-mm [0.008-inch] foil was a mystery... First: Unbonded joints SHOULD have intimate metal-metal-fayiing contact [tight pull-up] such that primer-contacts-primer... or a mirco-thin coating of sealant [0.002 max] for joint sealing and moisture exclusion. Effectively they tested a poorly assembled joint with shanked rivets [aluminum foil in between sheets won't resist/support rivet shanks, alowing the rivet to shank... wheras high strength epoxies tend to minimize the shanking effect and stabilize the sheets.

Which leads me to the last statement I will make here [because I simply have to get-off line and do some work...]: the test did NOT reference ANY established [standard] material or fabrication procedures so that basic "how/what-did-they..." questions could be resolved. Example: hole, countersink and dimple dimensions and quality standards. Specimen fabrication standards. What was the material spec for the 2024-T3 clad aluminum alloy, IE: AMS-QQ-A-250/5, AMS4041 or prEN2090??? ETC....

Regards, Wil Taylor

 

[planedr]

I read the report and thought it was pretty good. Yes there is maybe not the intimate detail of many of the issues at hand. But to dismiss the results I feel is disservice. If they quoted Boeing standards it would be inferred by many that's all it's good for. I kind of thought QQ-A-250/5 was a rather broad specification......

Standard industry practice would lead to this general result seems reasonable. If every rivet had it's tail gauged I would be more prone to dismiss it.

It was nice to see some level of comparison under the same controlled test regime. Good place to start.

There is a report titled "Practical Investigation of Aircraft Pressure Cabin MSD and Corrosion" that is pretty interesting as well, sorry I do not have a link. But the Fokker dimpled sheets didn't fair vary well here either....