Shaving of Rivet Heads For Countersink Fasteners 2

[alexeu]

Recently, I have been asked of the maximum allowable of material to be removed from a fastener head for a countersink rivet to be considered structural. As an example, A NAS1097AD7 rivet in a 0.040 inch thick 2024-T3 materia, the maximum depth of countersink using Boeing guideline is 66% of material thickness, giving a max depth of 0.026 inch. Therefore, the countersink head might not have sufficient bearing area in the countersink hole and tensile properties. Are there industrial specs or test or maybe guidelines to follow on shaving rivet heads.

Another question I had is can a mechanic shave the head of a MS20426AD7 rivet (tensile head) down to a NAS1097AD7 rivet (shear head)? Any justification? Such as induce stress, heat generated due to shaving process which cause stress corrosion cracking for rivet?

AEU

 

[Sparweb]

The rivet is definitely on the big side. I didn't know you could get a #7 size - there isn't one in the NAS spec. Assuming that the head is nominally 0.053" (interpolating between the #6 and #8), you need to shave off 0.027" of the head. You will lose most of the head where it's most effective in holding the top sheet. Pulling through will be easy, and the only thing saving you is the sheer size of the rivet. The bearing allowable I calculate is about 600 pounds - well below the shear allowable, and I expect testing would allow even less.


Steven Fahey, CET
"Simplicate, and add more lightness" - Bill Stout

 

[koopas]

AEU,

Microshaving of NAS1097 shear head rivets is prohibited by the SRM.

Shaving a MS20426 rivet head is permitted. Refer to Chapter 51 of the SRM for dimension requirements before and after shaving to maintain rivet strength.

Alex

 

[Lindner]

Some SRM's allow you to shave a rivet head as long as the factory marks on the head still visible. (same for removing corrosion from the head of differents still fasteners)

 

[crackman]

Hi All

I wanted to do a bit of research before I answered this string. I dug thru my archives as I remembered some information I found a long while ago. In several of the old Lockheed SRMs I have (Constellation, Electra, etc.) the microshaving of AN426 and AN427 rivets is called out. Here is one of the passages:

"Stop the countersink or dimple the hole, depending upon material and thickness, so that the head will protrude 0.001 to 0.009 inch above the surface. After driving the rivet, grind or mill the head flush with the surface exposed to the airstream."

This was done to ensure proper hole fill. To substantiate this, rivet tension fatigue testing was performed on AN426AD3 and AN426AD4 rivets in U shaped sheet metal channels. This was done way back in 1945. Much of the designs and SRM repairs were extensively evaluated by Boeing, Douglas, Lockheed, Grumman, etc., back in the hay days of aviation unfortunately much of this data is stored in the archives and long forgotten by most everyone. No one seems to have the time to research any more these days and all those gray hairs who knew how the designs were substantiated are long gone now.

The fatigue test findings indicated no impact at low or near endurance levels but more significant impacts as the tension fatigue levels rose (for AD4 at 100 lbs tension life went from 75000 to 25000 cyc while AD3 went from 11000 to 3500 ~ essentially a reduction factor of 3 although this is not constant depending on sheet thickness and rivet tension load, etc). In addition, not only did the sheet crack sooner but the fastener heads began to crack (more so in the AD3 than the AD4 but still enough of each to be significant). Static ultimate allowables were also evaluated and a 5% reduction in tension allowable was observed. Obviously as you shave more off this would increase.

Moral of the story, never design rivets in tension ~ I am sure you all know this already. Alas, this cannot always be done obviously in airframes. In particular, rivets connecting pressure skins to frames regularly see bending due to bulging effects. Soo, shaving rivets should be ok in pure shear designs but great care must be used in pressure boundaries. Make sure you know what stress levels you are operating at, any bending effects, and try shaving off the least amount.

Hope this helps any of you and good luck.

James Burd
FAA DER - Structures/Fatigue and Damage Tolerance

 

[Sparweb]

Thanks to a fellow Eng-Tip expert, I have a copy of Grumman's NAS1097 rivet allowables. There is no mention of life expectancy in these tables, though. Some of the numbers you quote are disturbingly low. When I use Swift and Eastin's stuff, I typically see lives of 10⁵, not 10³. Is any of this data available publicly?

Our office benefitted greatly from a retiring engineer who had worked at Shorts, Canadair, and Boeing. Some of the stuff we inherited from him is priceless.


Steven Fahey, CET
"Simplicate, and add more lightness" - Bill Stout

 

[crackman]

Hi Steven

I will try to summarize more of the test results however the results basically indicate that lives greater than 1e5 are not achieved until you get below 50 lbs of tension in terms of a repeated constant amplitude fatigue load (R ratio was 0.2). Thats essentially why we never try to put rivets in tension but its tough these days for one to put their hands on the data backing this up. The results included both blueprint and the milled/shaved down heads. When I get a chance this weekend, I will try to summarize the Sn curves for you.

One of my passions is history and so I have always made it a point to research and since very few people today appreciate the value of these old reports and data, I try to scarf them up when I can.

James Burd
FAA DER - Structures/Fatigue and Damage Tolerance

 

[tbuelna]

alexeu:

My M&P documentation shows flush rivets on OML should be +.000/-.004.