Rivet countersinking by dimpling 1

[CDR]

I am looking for information on dimpling as a countersinking method in aircraft panels. I have found a lot of information on techniques used for home-built aircraft, but I am more interested in dimpling of thicker sheet materials. Specifically I am looking for information on sheet thickness limits for dimpling, specifics on different dimpling processes for thicker skins, and possible comparisons of fatigue life of dimpled vs machine countersunk riveted joints. Any help would be appreciated.

 

[gunther]

Hi CDR

I am not aware of any fatigue allowable for dimpled fastener installations. I would not approve as a DER this method of joints in primary aircraft structure, subject to fatigue loads, at all. I know only of dimpling of very thin sheets where countersunking is impossible and the second sheet is thick with a countersunk to fit the dimple.

Gunther

 

[liaisoneng]

CDR:
My experience is on older aircraft where quite a few of the skin lap joints are dimpled. The skin stack up is usually two 0.040 sheets of 2024-T3 and a 0.040 7075-T6 stringer for MS20426D6 rivets. In this case the dimple goes through all layers, including the stringer. The problem I have had in the past is rework. If you replace any part of the stack up, other than the top sheet, you will have difficulties getting the new parts dimples to nest with the existing one (a 200+ inch lap joint only compounds this problem). Furthermore, in removing and replacing fasteners, we have had problems with the inside (stringer) portion of the dimple cracking when installing D or E rivets. From a durability standpoint, I haven't seen anything that leads me to believe there is a systemic problem with dimples. Perhaps someone with a classical fatigue background could speak to that more. Hope this helps.
Russ

 

[CDR]

Thanks Russ,

your input is useful. I am interested in some of the manufacturing mwethods you used in producing dimples. If I am correct, 0.040 inch 2024-T3 is thin enough to be cold dimpled. I am curious though whether coin or radius dimpling is typically used. Also, are the dimples typically formed in all sheets being joind at the same time, or are they dimpled independently?

Also, in terms of fatigue, I know it is diffucult to directly compare a machine countersunk and dimpled riveted joint due to the differences in their load transfer mechanism. In a machine countersunk joint, the rivet acts more like a shear bolt while in a dimpled joint, the dimple transmits most of the load and the rivet keeps the sheets clamped together; more like a tensile bolt. I am curious to see if there are differences in the way dimpled and machine csk joints are designed/analyzed based on these differences. Dimpling of thick sheet material also tends to produce microcracking that is detrimental to the fatigue life of the joint.

Any additional insight into dimple countersunk riveted joints is appreciated by all.

Calvin

 

[liaisoneng]

Calvin:
I am not familiar with what Boeing did on this 60's era aircraft (hot vs. cold). You're right that 0.040 stackups could probably use cold dimpling. However, with a harder 7075-T6 stringer underneath, they may have had to hot dimple. As far as replacement of the skin: if replacing the top layer, you go up in gage sufficient to avoid knife edging the machine countersink for the fastener; if replacing the middle layer or stringer, you can oversize/countersink the holes in the new part to allow the dimple above to nest into the part - of course you have to reinforce the part in some fashion due to the increased area out. I have also moved the skin lap to completely remove the dimpled area and replicate it at an adjacent stringer. Many times, if you're not replacing the top part only, the dimples push me toward trying to change everything in the joint. Sometimes explaining that to your customer is a challenge.
Russ