Effect of knife-edge, HiLites

[Zdane]

Hi there:
I hope that some of you guys might be able to direct me to some reference material on quantification of near knife-edge installation of HiLites, all with interference fit.
Structure has passed fatigue test with no cracks reported. Yet, without some quantitative analysis it is next to impossible to handle manufacturing deviations.
TIA
Zdane

 

[bwain]

You probably need to test some of the likely manufacturing deviations. Such tests are a common thing in the structures trade.

 

[Zdane]

Thanks for the reply, bwain.
Testing would be ideal, but it would also take quite long - not the test itself but the red tape before it.
On the references, there is a graph in Niu (design), say page 232 or so, but it is fairly general in nature - and there is always a question how to quantify interference fit effect on life. I mused the idea of running insertion simulation with something like DYNA or MARC and then doing crack initiation analysis, but that too will have to wait for some less busy times.
Regards,
Zdane

 

[jetmaker]

Zdane,

Without your own data to substanciate analysis, I would strongly recommend you take the following approach.

Since the part made it through fatigue test without damage, assume this detail has a zero fatigue margin. Carefully document the detail so that the baseline is well established. Next, if you find in production that the hole was countersunk too deep, recommend using an O/S Hi-Lok to remove the knife edge. This works in that most 1st O/S Hi-Loks do not require a deeper countersink, thus allowing some shank diameter to be obtained. Also, Radius Lead-in bolts are ideal for this application. The goal is to re-establish your baseline configuration again.

Note: IMHO, be careful on using books like Niu for design/analysis data. These books are compliations of many different sources and often do not reference the source, nor provide all the assumptions/criteria that was used to derive the data. In general, Niu is a good book for understanding principals, but not for detailed information.

Regards,

jetmaker

 

[Zdane]

Hi jetmaker;
Thanks for your input, it is appreciated.

Unfortunately, I was not specific enough about my knife-edge holes. Material is .054" thick at low tolerance; head flushness can go to -.005; HiLok shear head for -5 is .041 max deep. That leaves theoretical min shank of .008, not a lot. Further, there is a relief radius on the head-shank transition (in the hole) which is anywhere from .025 to .035, so in the end there is possibly O/S hole in the skin, and hardly any shank.
Now the question is how to handle damaged holes? -5 has no real O/S couterparts, just 3/16 with head .006 deeper (max .047) being the only option. There is no shank support to speak of here at all.
Having said that, fastener has a head-shank fillet rad of .015-.025 (5/32), that will enforce a good contour of the hole upon installation, provided hole is not O/S too much and hole relief rad is on the lower tolerance. And the same reasoning can be applied to 3/16.
Now, my problem here is how to quantify all this stuff? As I stated before, any references to published research would be mostly welcome - as are your opinions and views.

Final note: jetmaker, I cannot agree more with you on Niu
books.

Zdane