Wet Rivet Shear Strength 4

[YoungTurk]

When installing rivets in aircraft structure (sheet aluminum), what, if any, adjustment do you make to the joint shear capability for solid (AD) rivet installation with a sealant (wet install)?

Niu suggests a 20% reduction if no test data is available. MMPDS specifically discusses the rivet being solvent wiped. I'm inclined to think introduction of a thin layer of sealant will act as a lubricant and reduce shear capability for bearing critical joints. This is being met with skepticism from other engineers. Opinions welcome, references appreciated.

 

[SWComposites]

Sealant will not significantly affect the bearing or shear-out strength of the aluminum sheets.

Sealant will likely affect the strength of the rivet (due to increased joint flexibility and therefore more bending on the rivet).

Joint strength will depend on which failure mode is critical.

I have no idea what it will do to joint fatigue performance.

SW

 

[rb1957]

sealant is there to help prevent corrosion. agree with SWC, i don't see a hit against strength (bearing against the hole, shear of the rivet). disagree that it'll cause more bending in the rivet, don't see a longer off-set. i'd expect a reduction in joint stiffness, which would impact slightly the fatigue performance, but probably in a good way, bluntening load peaks.

just my 2c

 

[YoungTurk]

Thanks for the feedback.

I'm curious why MMPDS makes a point of talking about the rivet joint tests being dry pin if there is no affect on strength? Doesn't sealant between the rivet and the sheet affect the plastic behavior and load distribution after yield?

 

[thruthefence]

Wouldn't the sealant act as an adhesive between the two sheets?

 

[YoungTurk]

I don't think you'll end up with much, if any, sealant between the sheet for a wet rivet install, at least not the way our proces works. They basically gob sealant onto the rivet and insert it into the parts, which should be clamped up with clecos or similar.

The issue I'm dealing with is how to handle a situation where an assembly was built and tested with dry rivets and now wet rivets are being used (years later) as a new standard method to avoid corrosion (as rb guessed/knew). Rivets were not originally the point of failure, but how to be sure that the test results are still valid? Retest is out of the budget.

 

[rb1957]

i really don't think there's much of a problem (that the sealant reduces the shear allowable).

whilst you can't retest the assembly, can you test coupons ? looking to detect a difference in rivet shear strength ??

if you can't do this, could you interest a local university in this research project ?

 

[thruthefence]

I was asking because the tank sealants are so tenacious. I once hoisted a Bell 206B clear of the floor, in an attempt to get the 890 that sealed around the transmission deck fittings to break free.

 

[WKTaylor]

Be extremely careful... this is not really a "simple procedure" as it seems. Materials, assembly and installation techniques are everything when sealant is introduced 1n a joint [wet assembly and wet fastener install]. Rivets are particularly susceptible to installation problems with sealant.

Early testing revealed poor static allowables with sealant in riveted joints. Fatigue allowables were worse. this is why early integral tanks DID NOT assemble structure wet with fayng surface sealant, nor install fastener wet. Everything was assembled "dry"... and the gaps were packed, all edges were fillet sealed and fastener heads/tails over coated. Fuselages relied on tight multi-fastener lap-joints to retain pressurization. Internal edges were also fillet sealed and interior/exterior gaps injected. The resultant airframe was heavier than desired and a maintenance nightmare: and "possibility" that fay-sealed assembly with wet-fasteners could save weight, reduce assembly costs, reduce joint corrosion [fay and hole] and reduce sealant maintenance headaches [fluid and air leaks]; and possibly reduce weight by eliminating bulky/globed sealant... was all seen as a "gold objective"... if achievable.

The DoD, OEMs, fastener vendors and sealant vendors worked toward the objective noted above.

DoD testing revealed that structural pull-down [clamp-up] and tight fastener fit before actually installing the fastener, and/or during the installation process, of lock-bolts, blind-bolts, rivets and blind-rivets also allowed for maximum sealant escape from the hole and around the head/tail resulting in tight/hard joints.

Sealant vendors evolved lower viscosity sealants for fay and wet-fastener assembly... and DoD/OEM structural testing evolved the techniques of HIGH pre-clamp-up of "wet-assembled" parts prior to fastening for MAXIMUM sealant squeeze-out and tightest possible joints.

Fastener vendors "tweaked" their fastener-systems to account for additional structural and fastener "settling" due to presence of sealant. For instance: pin-pull lock-bolts were developed providing MUCH higher pre-load prior to collar-hammering; and most Hi-Lok collars have collar-break torques about 10--15% higher than recommended for nut-torque installations (assumes pre-load will soften ~10% prior to sealant cure). Blind-rivets and blind bolts were also improved for "sheet pull-up" forces to minimize looseness after joint relaxes. Etc…

Test with wet installation of rivets revealed a “hydraulic lock” effect with trapped sealant. Trapped sealant has to have an escape path and “squeeze-out time” to insure hard contact of the rivet to the structure. If the rivet is bucked before hard metal contact, then the head may actually be “floating” after the cured sealant will crumbles away due to bearing and cyclic loading. The technique for installing rivets “wet” with sealant” involves pressing the head of the rivet hard against the structure [into a countersink] for a few seconds with the rivet gun hammer-head… then driving the bucked-tail (swelling tail will squeeze sealant sideways, forming a natural fillet around the bucked head). For protruding head rivets this is no brainer. HOWEVER… for flush-head rivets, countersink depth is critical: the head must NOT ride below flush in the countersink when driven (may have to deliberately set flush fasteners in shallow countersinks, then mill-flush). The NAS 1097 shear head rivet has a tiny dome element to help ensure that the head of the rivet will see positive contact with the rivet gun hammer-head. NOTE: Douglas developed 100-deg shear and tension head rives with a slight crown to ensure solid contact with driving tools… and allow a slightly shallower countersink “than normal”. Look at the wings and fuselage rivets of an MD-8x and these rivets are clearly visible… and yes the aerodynamicists are “OK” with the crown protruding slightly above flush… because of the smooth rounded-shape of the dome.

NOTEs.

The MS14218 and MS14219 [120-deg flush-interference, solid] rivets styles MUST NOT be installed wet with sealant due to the precise counter-bore/countersink nature of the head recess: the sealant (and even primer) will be trapped under the head and cannot escape (rivets MUST be installed dry, OK to apply alodine holes/countersinks before installing rivets).

A new generation of precision rivets and precision drilling, reaming, countersinking and squeezing (etc) equipment allows OEMs to install rivets so consistent and tightly, that little or NO sealant is required for a life-time integral tank using fully automated assembly. Eliminating sealant has allowed automated assembly to flourish, because wet-sealant assembly and fastener installation is a notoriously time consuming and “dirty” process. Sealant must be controlled, mixed, applied, etc… and then wet sealant sticks to everything like a brother… and must be constantly cleaned-up… otherwise the cured sealant must be aggressively removed by tools or chemicals.

Wet installation of rivets with primer (epoxy) is consider the same as “dry” rivet installation… assuming primer is still “as-thin-as-water” and has not become more viscous like in the latter stages of curing. This has been used effectively for decades.

Some design manuals force a ~ 0.90X strength and fatigue factor if primary structure is assembled “wet with sealant”. Also some torque-process manuals require increased nut-torque to account for inevitable pre-load relaxation as additional sealant squeezes-out of a joint.

Regards, Wil Taylor

 

[YoungTurk]

Wil, great post. Your explanation of issues with wet install match closely to what I recall from Mr. Niu's class. Unfortunately my notes on the topic had nowhere near the detail you're able to give and your post really helps fill in the gaps of my understanding.

I am going to continue to look for existing test/reference data for wet rivets. I believe our process uses a silicone based sealant and most applications are of protruding rivets. Will take extra caution when need arises to use countersinks.

Rb, with respect to testing, I'd love to have the ability to test some coupons but the reality is that there is no company support for testing of any type. I like the idea of working with a college - I'll have to give it some thought.

 

[rb1957]

noting wil's post starts with "Early testing ...", is it reasonable to say that the effect of modern sealants on typical (BB and BJ) rivets is minimal ?

noting in the reply "our process" ... this implies (pretty directly) that your company as a procedure for wet installing rivets. Perhaps you could investigate how this was derived, and what supporting data there is. i suspect that the answer will be "industry standard" meaning that either your company cribbed from Boeing, Douglas, et al (like most) or they just documented what they've been doing for decades (like most everyone else) and pretty much forgetten why the procedure is the way it is (ie the data's been lost to the mists of time).

 

[WKTaylor]

rb1957... another odd one back-at-you...

Try to find ANY engineering data [analysis, test, etc]truly validating the use of countersink fillers in repairs... for Hi-Loks or rivets [solid & blind]. "They seem to work OK"... seems to be the only answer I have.

OK there may have been (1) technical report on testing with solid aluminum rivets that swell to fill the hole and lock the filler in-place [my faded memory seems to recall... but I do not have], but for the others???

Please advise if You know...

Regards, Wil Taylor

 

[thruthefence]

Wil, at the risk of being censured, weren't you looking for one of those old Pratt & Whitney engineering reference 'pocket' books, at one time?

 

[rb1957]

will,

i quite agree, i'm trying to have a local college do this as a bachelor's thesis ... my particular interest, "fatigue quaility of repaired CSKs" we (and anyone repairing the fuselage of a modern plane) generally repair the rivet CSKs with little filler pieces, locktited in-place and slap (ok, position carefully) a dblr on top.

 

[WKTaylor]

Thruthefence... Dang...

Thanks for reminding me... got so busy, tabled getting a copy of the P&W engineering reference 'pocket book" for a couple of weeks too-long... then went brain dead on the subject.

Any idea if the P&W is still available; and where to get it??? I'm (still) building a reference library for liaison engineer training and continuing education... and all knowledge is good.

rb1957, NOTE: a mechanic taught me that gap-filling high temp LocTite... with adhesion promoting primer on the countersink and under-side of the csk-filler head... helped form a stable, micro-gap-free bond. This seemed to help make the csk-filler installation fairly rigid/tough and resistant to aggressive/rough high-friction drilling/reaming ["spin-resistant"]. I estimated that this method reduced moisture intrusion and enhanced sealing for long-term corrosion protection [wet-sealant Assy]. Note: also encouraged the mechanics ensure the csk-fillers were flush to very slightly high [0.000-to-0.003] to ensure light contact with the underside of the mating doubler. They were able to do this when pressing the fillers gently into the gap filling LocTite.

Regards, Wil Taylor

 

[49again]

Wil -

Here's a link to order the P&W 79500 Aeronautical Vestpocket Handbook

http://www.airliners.net/aviation-forums/tech_ops/read.main/93610/

Not sure how current it is.....I mean, who wears a vest these days?

 

[graemew]

G'day Wil
A caution on the use of Loctite primers. I was advised by a Loctite rep. some years ago that these contain a very fine dispersion of copper or iron (depending on the primer type)to catalyse the adhesive, which should be washed off after the Loctite has set. My experience of the use of these materials for aircraft bearing installation, without adequate cleaning rigour, is that corrosion is promoted.
Do you have a Boeing Liaison Eng Hbk?
Regards

 

[WKTaylor]

graemew...

Hmmm hadn't heard of that issue. I am a bit confused: is the issue with the adhesive or the primer... or both?

I strongly recommend using adhesion promoting primer for every installation... over "alodine" surface prep for aluminum. If the promoter, adhesive, primer (on part faying surfaces) and sealants (and sealant adhesion promoters as needed) are properly applied, the potential for corrosion should be eliminated by the principle of "isolation".

I think I can find a copy of the Boeing liaison engineering hdbk?? What specific reference???

Regards, Wil Taylor

 

[graemew]

Gooday Wil
Thanks for the response,
The issue is with the Loctite Primer. It gets sprayed on quite liberally at times and sometimes is not adequately washed off afterwards (it is solvent based and I have seen procedures not calling for a wash afterwards).
You are correct in regard to isolation, but you also know how paint cracks and chips on corners.
The cases of corrosion that I have seen were associated with bearing installation and there was no bead seal applied to the edges of the joint. It was a practice in a very old maunual (old aircraft) that I had updated when I found out the cause of the corrosion.

As for the BLEH, I have a copy and I am happy to share if you are collecting handbooks for training purposes.
Regards
Graeme Willaton

 

[Sparweb]

rb1957: I quite agree, i'm trying to have a local college do this as a bachelor's thesis ... my particular interest, "fatigue quaility of repaired CSKs" we (and anyone repairing the fuselage of a modern plane) generally repair the rivet CSKs with little filler pieces, locktited in-place and slap (ok, position carefully) a dblr on top.

Any time I've seen an external patch repair removed, the countersinks are never filled with those conical fillers. I recently had to listen to a shop worker say that they were there, yet when I insisted on having a rivet drilled out to prove it, the empty cavity was revealed.

Despite what I've said, I'd still be interested in seeing someone test those little conical things.

Steven Fahey, CET