Boeing 737 3

[rb1957]

i don't know that you can blame all RT resin systems .... Fokker have used RT resins for decades, without problems. in my experience, it it the process compliance that is the root of the problems. Boeing now uses a much better (reliable) systeem (PAA) and much tighter process controls.

the best thing that can be said about SW is that 1st no-one died, and 2nd this is probably the most highly stressed (certainly most highly loaded) part of the fuselage. Interestingly the failure happened without max cabin pressure.

i'd add to the previous posts (noting that i'm not particularly a Boeing fan) that we're looking at an airplane designed 50 years ago and it has stood up pretty well to that experience. it was designed before damage tolerance, and in the early days of fatigue; fail safe was in it's infancy.

the Aloha incident was a combination of errors (as all incidents are) ... the basic design was possibly suspect, the maintenance was not complete (not through negligence), the certification was questionable. this produced the aging aircraft re-cert program, highlighting the (possibly obvious) interconnection between OEM, maintainer, and certiication. remember too that this design is early in the development of mandatory structural inspections, and operators had to understand the importance of these.

In-service usage would be closely monitored by Boeing; yes, each operator is different to the certification spectrum, but their operations are assessed (by Boeing or by themselves) to ensure that the correct assessment of fatigue effects (again, fall-out from Aloha).

 

[Engineerpaii]

the fuselage skin section from Southwest flight 812:
watch it,.,.,

http://www.youtube.com/watch?v=nI78uAsVMEw&feature=player_embedded

 

[blakmax]

Msquared48

In a properly formed adhesive bond, the fasteners will carry negligible shear loads and the bond will carry almost everything. This is because unlike the populous theory that adhesive bonds spread the load over the entire bond, in reality the shear stresses peak at the ends of the joint and decay to zero within a relatively short distance along the bond. Most of the load is transferred at the ends of the joint. The closest a fastener can be installed to the edge of the joint is determined by tear-out considerations and is typically 1.5 times the diameter of the fastener. Hence, all of the load will be transferred by the adhesive before the fastener even experiences any load.

The fasteners attach the joint to the frame and carry out of plane loads.

As for the fears expressed in relation to adhesive bonding, it is by far the best method for joining such structures, provided the processing is correctly validated and implemented, which in the case of Aloha 243 it was not. Given the thickness of the skins on the 737 if they were effectively and correctly designed and bonded, the adhesive itself should never fail. It is possible to design bonded joints such that the metal will always break, not the adhesive.

Adhesive bonds also have the advantage of not introducing the 300% increase in stress caused by fasteners.

With regard to the Fokker use of RT adhesives, they never attempted to use a pre-mixed film adhesive which had to be used frozen. That was the error in the design of the early 737 lap joints.

 

[KENAT]

To all you target builders out there (msquared & hokie in particular), for an example of what happens in a pressurized airliner when manufacturing staff persuade engineering not to use adhesive, take a look at the comet story.

Folks coming from other sectors have trouble in aerospace getting their heads around the mass issue. My first boss's background was in naval dockyards not aerospace, when he got given responsibility for some airborne kit he had to constantly remind himself about mass being the enemy.

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[thruthefence]

"manufacturing staff persuade engineering not to use adhesive, take a look at the comet story."

I have read that the marketing people overruled engineering (who wanted "rounded" window openings) in favor of the Comet's squarish windows, because they "didn't want the windows to look like portholes on ship."

Where were we with bonding technology in the early 50's anyway?

 

[KENAT]

From my ever worsening memory...

They had some proprietary bonding technology or something. They built up sample fuselage sections using this bonding around the windows, I think in conjunction with rivets but don't recall for sure, and subjected them to fatigue testing of pressurization cycles.

Apparently the bonded design was a pig to build or something and manufacturing wanted to move to riveted only design. This was approved without repeating the fatigue testing.

To make it worse I believe holes were punched not drilled/reamed and of course many windows and fuselage cut outs had sharp corners.

DeHaviland were probably amongst the world leaders in aerospace adhesives, look how much they used it on their wartime aircraft.

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[rb1957]

a joint can be properly designed using adhesive, adhesive and rivets, or rivets alone.

they did do a fatigue test on the comet, quite ground breaking (no pun intended) for the time. unfortunately they used the static test specimen (since it was still intact) and that (as we know now) invalidates the results.

 

[blakmax]

rb1957 said"a joint can be properly designed using adhesive, adhesive and rivets, or rivets alone".

This is true, but the strength of that joint follows exactly the sequence you stated. One military aircraft with a composite to titanium multi-step joint was evaluated for bonding, bonding and fasteners and fasteners only. The results from memory were approx. 35,200 lb/in for bonding, and 28,500 lb/in for bolting. For bonding and bolting the strength was 34500 lb/in and the bolts carried just 400 lb of load. So by adding bolts, the strength dropped and the bolts carried next to no load.

There is no question that properly designed and fabricated adhesive bonded joints are stronger, lighter and more fatigue resistant than either bolts alone or bonding and bolting. The only reason we do not see more adhesively bonded joints is the "Grandma's cup syndrome"; I glued Grandma's cup with that glue stuff and it broke again.

The fundamental reason we see bond failures on aircraft is that the FARs require demonstration of static strength and fatigue resistance, and most bond failures can not be prevented by static or fatigue testing. The are due to chemical, time dependent dergadation of the bond interface, usually by hydratrion of metal oxides. Fir the oxides to hydrate, the chemical bonds formed at the interface dissociate, leading to interfacial failure of the bond. There are tests to demonstrate resistance to hydration for bonded joints and if these were mandated, we would not see bond failures in service and we would achieve the weight savings and sign ificant fatigue improvement that can be achieved by bonded structures.

I'd be happy to discuss what a waste of time NDI and damage tolerance are for bonded structures, but that may be getting too far from the topic.

Regards

Blakmax

 

[dik]

I don't have a real problem with adhesives... I'd used them for plywood gussetted trusses 40 years ago because I had more confidence in the glue than I did in nails...

Dik

 

[rb1957]

the order of joint designs was not intended to show an order of preference or acceptability. the point made about adhesive being process dependent and possibly degrading over time this IMHO enough of a hit to question them entirely ... yeah, i know there's a baby in the bathwater but ...

i've learnt enough to be very careful of adhesive joints.

 

[thruthefence]

"I'd be happy to discuss what a waste of time NDI and damage tolerance are for bonded structures"

What can be done, in an inspection environment, to insure the continued airworthiness of bonded structures?


Maybe we need a new thread?

 

[rb1957]

do it right the first time ... a correctly produced adhesive joint should outlast the aluminium around it.

the problem is incorrectly produced adhesive which can degrade very quickly.

I guess what we haven't found is a method to detect an incorrectly produced adhesive joint

 

[msquared48]

Kenat et al:

I am well aware about the mass issue as s Hokie. I too am aware of the Comet. However, I am also well aware of the result of any failure in this current instance.

In structural building engineering, one of the basic tenets is to buy time for the occupants to survive an event of an acceptable risk. The problem I have with the glue, is that the failures are sudden and the weaknesses generally not detectable until the failure. To date, we have been lucky.

To me, the current technique, although apparently the best we have, is just not good enough. There must be a better way to get a more reliable result. Complacency with the current methods and materials is not the answer. Finding other options is.

It is true that many materials we have in buildings use glue, but we are trained not to rely on any contribution, except that which has been adeqauately tested, as in plywood, glulam beams, paralams, etc. Occasionally, I also use glue in other circumstances, but never totally rely on it in a connection. Every coffee cup I have ever mended with epoxy has eventually failed. I just don't trust glue.

This is not a rant on aeronauticals or the aircraft industry in general. I just think we can do better, and should as engineers, that's all. No offense intended guys.

Mike McCann
MMC Engineering
Motto: KISS
Motivation: Don't ask

 

[moon161]

"Every coffee cup I have ever mended with epoxy has eventually failed. I just don't trust glue."

How many coffee cups have you fixed with rivets?

 

[rb1957]

for a coffee cup, failure provides two acceptable futures ...
1) retirement, or
2) accepting reduced functionality (loss of a handle).

the 3rd option (repair) is fraught with possible in-service sudden failure resulting in unacceptable consequences.

 

[KENAT]

Geez, guess I better not mention some of the reserve/safety factors that get used on aircraft then or you'll really get your knickers in a twist;-).

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[rb1957]

all i said was ...

 

[ESPcomposites]

Not to muddy this any further, but bonded repairs are an interesting case as well.

I believe the current strategy (per FAA) for bonded repairs is that structure must still be capable to limit load in the event of a failed bond. Ultimately, we just can't trust them yet. As said, they are great when done correctly, but very poor if not done correctly. In the real world, they are not always done correctly and that possibility must be addressed.

Brian

http://www.espcomposites.com

 

[DHambley]

Documented??? People want "documented" information? The information about the skin panels being stretched is from the engineers whom I worked with on the early -200 line. Boeing will NEVER have this information "documented". So, I guess go beleive only what is "documented" if that determines what you believe.

 

[KENAT]

My best structures prop told similar stories about the Harriers he saw being built at British Waste of Space in the 80's. Some guy under the A/C, on his back using his feet and arms to push the panel into place or something.

Hammer to shape, File to fit, Paint to match right?

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​