Preload in shear joints

[shingouz]

Hey everyone,
I was wondering if aircraft manufacturers are using preload tensionning to enhance fatigue life into shear joints, What I mean by preload tightening is instead of introducing a torque to induce a preload ( a lot of scattering, non accurate...), they introduce directy tension by using more accurate toolings ( hydraulic tensioning, strain gauges...)
Thanks

 

[SWComposites]

Doubtful. Sounds too expensive, complicated. Torqued up regular fasteners work pretty well.

 

[rb1957]

i think you mean assemblying the joint with a built-in shear load, to the joint would have a constant shear load and service shear loads would be applied on top of it (similar to prelaod and tension loading on a bolt).

other than the practical difficulties of doing this, one thing to consider is the effect this has on the structure around the joint, you'd be introducing tension and compression stresses.

next you need to consider the reduction in shear capacity of the bolt, for static loading.

then there's bolt bending.

this idea runs counter to everything we do in manufacturing (to avoid manufacturing/assembly stresses).

 

[MintJulep]

Yes. The use of Hi-Lok fasteners or similar is common.

 

[ESPcomposites]

Hi-Loks get a preload via torque. As the OP states, that comes with a lot of scatter (frictional effects, etc.) so you really don't know exactly how much preload you get. For aircraft installations, it is too expensive to have more exacting methods, so we just accept that and correlate the data based upon those results.

In other industries, with critical tension loaded joints, they do use more exacting methods of generating a preload. This can be done via bolt extension and in certain cases a "torque to yield" can work (though you have to be careful with that).

Brian

http://www.espcomposites.com

 

[WKTaylor]

shingouz...

Tightly fitted together parts with interference fit fasteners provide ideal shear joint construction... such as solid rivets, or interference fit Hi-Loks, Bolts, etc

CAUTION... here is where tension preload comes in to play...

The stack must remain tightly clamped together. Thin sheet metal has relatively low eccentricity [sheet centroid to sheet centroid] and the retaining force of solid rivet heads/tails, or the relatively low tension preload of shear head/collar Hi-Loks, or shear-head bolt/shear-nut combos, will keep the parts tightly clamped together... even as loads approach ultimate, especially in the end-fastener [sheet peel, induces fastener tension].

As stack thickness increases for lap shear so does eccentricity... and so does end fastener tension loads. This is where I, and knowledgeable stress guys, will introduce high-clamp-up tension rated fastener systems [still in interference fit]. At some point of load eccentricity, even tension rated bolts can fail heads. NOTE: Tight clamp-up/interference is essential to ensure fasteners see minimal bending loads at ultimate.

NOTE. One way to relieve this eccentricity situation is to introduce a double-shear joint... which may not be practical in all cases.

You haven't lived until You have seen shear head/collar Hi-Loks zipper-off heads, or distort/crush collars, and shear joint pulls apart in tension-shear failures.

Regards, Wil Taylor

Trust - But Verify!

We believe to be true what we prefer to be true.

For those who believe, no proof is required; for those who cannot believe, no proof is possible.

 

[shingouz]

Thank you all for you answers;

wktaylor,

for high eccentricity single lap shear joints, what is the preload tightening method (or control method) used to verify the preload installed? Is it still torque tightening?

 

[rb1957]

i'd've thought you could use whatever preload indicating method you want ... torque wrench, PLI washers, bolt extension, ...

 

[Screwman1]

Virtually always torque control for a tightening strategy. Since these are primarily shear joints you are not looking for a critical minimum preload, you are basing on shear allowables; or am I missing your question.

 

[rb1957]

i was confused by the original post ... at first i thought he meant preloading the bolt (like we'd do for a tension application), then i reread it and thought he meant preloading the joint in shear (ie similar to preloading a tension bolt in tension). it seems the thread has drifted to preloading the bolt similar to a tension application. this will reduce the shear capacity of the bolt (Rs^2+Rt^2 = 1); will has pointed out that this (tensioning a shear bolt) helps bolts with a high stack up.

 

[shingouz]

rb1957,
Yeah perhaps it does reduce the shear capacity of the bolt but in my case it isn't very important because the bolt is dimensioned so that the plates fail before. And yes the original post was about tensioning the shear bolt to increase his fatigue resistance (transmit some load by friction between plates)

 

[rb1957]

woooo there buddy !

you're straying onto thin ice when you could on friction for shear load transfer.

ok, you've designed the joint to be critical in bearing, and now you've found a fatigue problem with the joint, yes?

i would do several other things before a grasped at preloading a shear joint. i'd start with increasing the section size, maybe going to a double shear joint, but i suspect these are non-starters (since they're obvious, and probably the joint is designed and now you've got to "show it good"). next i'd try an interference fit bush.

so what really is the problem ?

 

[WKTaylor]

shingouz... I second RB1957's comment... whoaaaa tiger...

Friction in aircraft structures, is a highly unreliable force, and is NEVER, counted into structural equations. We design all joints as if the spaces in between fasteners is "0" friction... and load transfer only occurs thru fasteners.

Clamp-up forces from fasteners ONLY ensures that the intended shear joint geometry remains stable and viable for the life of the joint. IF the joint becomes loose, for any reason... such as fastener component crushing or failure [head/nut/washer yield, debris in the joint, failure to properly pre-load torque, etc, etc], uneven fastener loading [poor hole quality control], shear-bearing yield, etc... then all fatigue calculations go-out-the-window for a host of reasons: uneven/unintended fastener locding, fretting, corrosion, scoring, etc.



Regards, Wil Taylor

Trust - But Verify!

We believe to be true what we prefer to be true.

For those who believe, no proof is required; for those who cannot believe, no proof is possible.

 

[shingouz]

I agree with what you say guys, but don't you agree that by preloading shear joints, some amount of applied load will be transferred by friction and it would be interesting in term of fatigue resistance? I also know that the interference + high clamping can be combined for better fatigue benefits

 

[rb1957]

sure, some load is transferred by friction. the problem is determining how much, and ensuring that this result is consistent. with extensive testing and QA you will be able to answer these questions.

if you Really want to transfer load over the fitting faces (similar to friction), then you could serrate both faces so load would be transferred by the serration teeth.

 

[shingouz]

Now tensionning techniques are mature enough to tighten to a desired preload (+-10% let's say). my original question is why companies aren't using theses techniques to enhance fatigue life of their joints? ( I heard about LCT (ultrasonic measurment and monitoring of preload )technologies used for boeing but I don't know the proportions)

 

[WKTaylor]

shingouz...

Companies have spent many decades, hundreds-of-millions of $$s and millions-of-man-hours in R&D to develop exactly what You are asking about: highly fatigue resistant shear joints. That design/assy data, developed privately by companies, will be proprietary to the individual companies... and will be access-controled. NOTE. There is some data that has been developed and published by the FAA; however, that data will have restricted usefulness without a broader context [such as design manuals].

CAUTION. Due to the technology critical aspects of this data, I suspect that most of corporate developed data [and design documents] would be considered ITAR restricted [= International Traffic in ARms]... at least in the USA... and would ONLY be transferraable with significant corporate/USA-Govt oversight/restrictions on it's distribution/useage.

Regards, Wil Taylor

Trust - But Verify!

We believe to be true what we prefer to be true.

For those who believe, no proof is required; for those who cannot believe, no proof is possible.

 

[rb1957]

your last post confused me.

sure preload tension control is well understood, and applied to tension joints.

but you're been talking about shear loaded joints ?

 

[shingouz]

Yes rb1955, preload in shear loaded joints

 

[Screwman1]

It seems like the OP wants to use tension control (high clamp load) to provide enhanced fatigue results in shear joints.

I agree with the group that this is BAD idea because of the variations of friction and eccentric loading on the sheets. It would scare me to death to think about what would happen if we depended on interface friction in an aircraft.
Ultrasonics is little used on airframe bolts because the low L-D ratios generally give such limited elongation that the results end up lost in the gage error.