Kt values for CSK Fasteners

[edmeister]

Require a comparison of Kt values for a straight hole vs Fastener holes for NAS1097 / MS20426 Heads. Also does the Kt value change if the fastener is installed? I believe the interface between the bottom of the CSK and the straight bore is significant. Also Briles (+ Aerolok) types are an improvement. Any published info would be appreciated. Understandably, this is proprietary info - but some may be available.

 

[Sparweb]

Peterson's Stress Concentration Factors, 3rd Edition, section 4.6.1 (page 240).

eg. depth of countersink = 50% of plate thickness, increase in Kt is 27% over a straight hole, hence Kt for CSK is ~4. More if the CSK is 75% of the plate thickness.

Peterson also includes corrections for edge distance.

http://www.sparweb.ca

 

[LiftDivergence]

edmeister,

Not only does the CSK reduce the net area at the hole compared to a protruding head with the same shank diameter, and create a sharp notch, but it also changes the bearing stress profile and distribution based on the csk angle. This can actually get complex when you consider most aircraft holes are not just installed in areas of far-field tension, but due to joint geomety are subject to secondary (eccentric) bending. This also has an effect on the bearing stress dsitbribution. Furthermore, due to variability in installation, using a csk head fastener can also have an effect on the hole filling factor.

You can read about this in:
Schijve, Fatigue of Structures and Materials, Section 18.5
de Rijck, Stress Analysis of Fatigue Cracks in Mechanically Fastened Joints


And yes the stress concentration will change if the fastener is installed. You are dealing with rivets it sounds like, but this is especially true for transition fit or interference fit holes and holes which have been cold worked prior to installation.

There is a difference between the Kt of an open hole and a filled hole.

Useful for this is the concept of the stress severity factor. This is presented in Niu Airframe Structural Design page 238. Like most things with Niu it is basically just a bibliography as he has taken the equations and figures straight from Jarfall.

One thing you will note is that the SF presented there does not account for CSK. In industry this is commonly dealt with using generally OEM data for additional "gamma" factors which can be applied to the SF based on the depth ratio. These range from 1.0 for a straight shank to 2.0 for severe cases of CSK.

If you don't have availability of this, you can try modifying the SF based on Sparweb's suggestion, or you might find this paper useful:

Stress Concentrations for Straight-Shank and Countersunk Holes in Plates Subjected to Tension, Bending,and Pin Loading
Shivakumar and Newman

https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19920016754.pdf

Another interesting paper which discusses the fatigue life of joints including countersinks is:
NACA TM 1395



Keep em' Flying
//Fight Corrosion!

 

[rb1957]

There's a lot of literature out there. From my research you can get whatever Kt you want (!?).

In my experience stick with the golden rule (to avoid knife-edging, CSK depth < 2/3 thickness) keeps you out of trouble. Many airplane tests demonstrate this.

Yes, some fasteners "fuss" over the intersection of the CSK cone and the shank and "claim" an incredible fatigue improvement. Not "denying" their claims, but DT is more important than fatigue in today's world. And sure, that's a sentence that can be discussed and dissected until the cows come home.

And of course, with composite skins all bets based on metal experience/analysis/tests are off !

another day in paradise, or is paradise one day closer ?

 

[crackman]

edmeister

You have received some very good advice and information from the other posts. I would only add a few things to this.

For one, Jarfall generated the SSF method back in the early 1960's and he also worked at Boeing back then. The first public paper he published on SSF is in the ICAF proceeding of 1967 entitled "Optimum Design of Joints: The Stress Severity Factor Concept" and provides a good background on his method. However, the best compilation of his work are his papers while he was at the Aeronautical Research Institute of Sweden in the late 60's and 70's. The main point I would make is that the SSF method needs to be used with joint Sn data. The common mistake is to use MMPDS Kt data from notched specimens, and this just doesnt correlate to his method. Additionally, a very good update to the SSF method to account for shear reacted fastener load transfer was performed by Darby and Cornell of Lockheed and published in an AIAA paper if I recall back in the early 1970's I believe.

The reason I mention the above is that by using the joint Sn data, the effects of various countersink depths can be developed, as well as other effects. This is done both analytically and also by validating thru testing as previously mentioned by LiftDivergence. However, there are lots of Countersink test reports in the public domain one could utilize.

Long story short, if you are an OEM, this method has already been fully developed. If you are doing it yourself, it takes quite some time to fully understand the method and how to couple it with joint Sn data and develop the necessary factors resulting a solid method. One reference I can point you to for the countersink effects is a paper by Shivakumar and Newman entitled "Stress Concentrations for Straight-Shank and Countersunk Holes in Plates Subjected to Tension, Bending, and Pin Loading". And on one final note, the Countersink Kt curve in Niu's book (fig. 7.7.3) was actually developed by a well known (at the time) fatigue engineer named Spaulding who worked for Lockheed. He published several good papers in SAE, ASME, etc. back in the 1950's and his curve I believe was based on available NACA test data.

Good Luck