Countersink or Dimple 1

[Rob130]

Assume damage to pressurized skin, Flush patch required, skin - 2024-T3 .050, .063 doubler. SRM Calls for same size, type, spacing and distance between rows as nearest skin splice.

MS20426AD5 Rivets 1.0 spacing, .8 pitch. Skin Splice is dimpled. .050 existing skin can't be countersunk due to min thickness requirement.

Would the repair have the same strength if NAS1097AD5 rivets were used in csk holes vs. MS20426AD5 in dimpled holes? Dimples will need to be formed using coin dies in rivet gun and bucking bar method.
This is a theoretical problem to be discussed in a structural repair class.

What questions would an engineer have for the technician performing the repair or what questions should the tech ask an engineer?

 

[wes616]

I'm at home so I can't quote the number directly, but Mil-Hdbk-5 give some test data to answer your question. I believe that there is a percentage loss in strength for machine csk vs dimpling (somewhere around 15%).

Wes C.
------------------------------
No trees were killed in the sending of this message, but a large number of electrons were terribly inconvenienced.

 

[rb1957]

personally ... whatever you do, don't dimple.

LZ5 rivets sound big enough for most applications, BB4s. If you feel you must use BB5s, and can't make the doubler thicker, i'd accept the repair, do the DTA and find (almost certainly) that the border-line knife-edge rivets at the edge of the doubler are easy to inspect (and not much of a problem 'cause the CSK is in the doubler not the fuselage skin ... so it is contained).

an important consideration ... is this a repair (with a large cutout) or only a small cut-out ? any clues as to the size of the plane ?

 

[Rob130]

rb1957, cut out is approx 2"x2". The repair is on a C-130. I teach a class in structural repair. We use pieces of c-130 structure to perform the repair. I had 2 technicians perform the repair with 1/2 dimpled and the other half using NAS1097. The tools used are available to most org level techs. The result was the dimpled half looked terrible and the csk half very acceptable.

The problem is with the SRM, same size, type spacing for fasteners is prescribed. There is a statement in the SRM:
"NAS1097 rivets are the same as MS20426 standard flush rivets except for the reduced head size.
The NAS1097 rivets are used to
permit countersinking of thin skins; however, do not enlarge the countersink for an NAS1097 rivet to
accept an MS20426 rivet. When necessary to make a substitution, the same diameter and alloy
MS20426 rivet may be driven into an existing prepared hole for an NAS1097 rivet, and the head may
then be shaved flush."

As a technician I would read this as permission to use 1097 vs 426 when the existing material thickness for a 426 is less than the minimum thickness requirement. SRM min skin thickness for MS20426-5 is .063 and for NAS1097-5 is .040.

Taking into account the difficulty of dimpling the existing skin, it make sense to me to use 1097 and the above statement allows for this.

I am looking for any pro/con statements, basic parameters etc for general discussion with the class. The technicians have direct access to engineers and I encourage the tech avail themselves to their expertise, but I would like to offer them some better explaination other than the short blurb on 1097 in the SRM.

Thanks, Rob

 

[rb1957]

i'm alittle surprised you can dimple the sheet ... an OEM I was working at lost that capability (along with several others, but that is a different lament/rant !)

it's good to teach people how to do things and when to be really careful ... dimpleing and presurised fuselages are two things i wouldn't mix. unpressurised fuselages would be ok to dimple. but why ... it's a complicated (relatively) operation, compared to CSKing.

my personal opinion ... I would never (well, hardly ever) shave BB rivets to create an equivalent 1097 CSK, 'cause it's too prone to operator "error". If I had to, I'd carefully create the 1097 CSK, buck the BB rivet into it, and carefully shave the protruding head. even still you might have some funny interaction happening at the bottom of the CSK (as you drive the oversized head into it).

good luck !

 

[WKTaylor]

Rob130...

"...SRM Calls for same size, type, spacing and distance between rows as nearest skin splice..."

First. There is a problem with the SRM guidance. Match-drilling to a dimpled hole is exceptionally tricky. Thin sheet, capable of being dimpled is relatively flimsy [un-stiff]. Any small misalignment will be magnified when the "new" holes are dimpled. Misalignment by definition degrades strength to an unknown degree. I suspect that these skins were likely match drilled together ("flat-lap")... then separated for dissimilar dimpling; or were co-dimpled together in the same male and female dies (reassembled after deburing... then stack-dimpled together... then fastened).

NOTE: dimpling is a mechanical deformation process that tends to strain the areas between dimples. This tends to distort the sheet metal between holes, causing slight dimpled-hole to dimpled-hole mismatch.

Second.
Standard repair practice dictates that over-size [1/64" or 1/32"] fasteners replace the origional diameter fasteners. This cleans-out and re-aligns/re-sizes the old and new holes together as a "matched hole-pair".

Obviously if the substructure is NOT dimpled, the work-effort to coin or impact dimple all these holes separately then match them back together is "dicey"... even for experienced techs.

Third. An over-lay doubler, matched-drilled to the skin and installed with countersunk NAS1097D5** rivets would require minimum work effort and eliminate the potential for sheet metal distortion that could cause hole misalignment. In theory this is a "piece of cake... and would generate "nice-clean-round holes" in the skin.

Notes:

Dimples have a bad habit of cracking around the head due to fatigue-loading [stress cycles].

If existing dimpled holes are under the doubler, then countersink/dimple-fillers [typically cut-off MS20426 AD rivet heads, drilled-thru] will be needed... and holes will have to be oversized for #6 diameter.

NAS1097 #5 heads are 100-Deg X 0.037"-deep; and #6 heads are 100-deg x 0.046 deep. Countersink depth and location MUST be factored into Your stress calculations for the patch or doubler.

** recommend using D or DD rivet alloys, which are substantially stronger than the AD alloy. This rivet "upgrade" would provide a much higher head-tension rating (pull-thru strength) for the NAS1097 head-style.

Based on this discussion, I think You can see that a countersunk patch of adequate size is likely the "simpler option.

CAUTION regarding installation of repair patches & doublers. The edges of the added part MUST extend beyond the edge of existing substructure (frames, skin-laps, stiffeners, etc). This will allow the skin to be directly inspected (from inside) by visual or NDI methods to determine if cracks are evolving in this high stress-concentration row of fasteners (outer rows of fasteners). Failure to provide for this inspectability MAY lead to undetectable skin cracking that could be disasterous.

Regards, Wil Taylor

 

[Rob130]

Thank you very much folks. What you all have described as the problems with dimpling has confirmed my thoughts on the SRM and best practices for fatigue resistant repairs.

The suggestions and comments are greatly appreciated.

Rob