Firstly thanks to rottenray for the delightful comments. After many years as a young angry man, I was once told by a very wise person that "you catch more flies with honey than you do with vinegar." That's why the responses are reasonably friendly, and I hope they will continue to be.
I noted Volume's comments
We have about half a century of experience with repairing composite gliders.
Volume, I bet that absolutely none of these were bolted repairs, so why do Boeing plan to use bolted repairs for the 787 composite structure?
You are absolutely correct about the higher skill levels for bonded composite repairs. The RAAF in Australia restructured their training in 1992 to actually teach how to perform adhesive bonding (rather than the "follow the cook book" approach) and since that time our major repair facility cut the repeat-repair rate from 43% in 1992 to almost zero, saving five man years per year of effort every year since then. Yet civil training still aims at the lowest common denominator and teach to an appallingly low standard. This is stupid. It does not take much more effort to train people to be proficient composite and bonding technicians. I personally have trained even electronics technicians to perform adhesive bonds and to do the task to the best standards.
Part of the problem with the civil system is that they use the term "composites" to encompass composite materials as well as adhesive bonding, yet they are two separate and distinct technologies. Just because a technician can lay-up composite repair patches does not mean he can prepare a metal surface to bond that repair. The blurring of the boundaries also means that FAA documents such as AC 20-107A treat the certification of adhesive bond the same as they treat the certification of composite structures. Why? The design methodologies, the failure mechanisms and testing methods are dramatically different. Composites don't fail at an interface, but adhesive bonds often do. The only similarity is that the materials are sticky before they are processed.
Consider this: Current certification of adhesive bonded structures IAW AC 20-107A involves thousands of tests at coupon, element, structural detail, sub-component and component level. If as in my previous positing, the joint was designed such that the metal (or composite structure) ALWAYS failed before the bond, then the only thing measured by thousands of tests would be the strength of the metal or composite, not the bond. So why do thousands of tests?
The approach based on designing such that the bond was never the weak element in the joint would save millions of dollars in certification costs for bonded structures. A simple test program to characterise key adhesive properties (together with a change in design methodology) would be all that was required. Further testing to demonstrate bond durability is essential, but current tests required by the FARs do not actually demonstrate bond durability.
So, if we characterise the adhesive, demonstrate that the bonding process actually produces durable bonds, use a design methodology that assures the adhesive is never the critical element, train the technicians how to perform correct bonding processes and that joint or repair will NEVER fail. NEVER. Certification costs could be dramatically reduced and reliability would dramatically increase. Am I out with the faries? No, because using this approach we have had three bond failures since 1992, and in every case, technician error (or to be more precise technician short cuts) have been found as the cause of those failures.
Now we really are off the track about the pressurisation failure, but what the heck? I suggested that the Moderator could pull me into line by moving the discussion to a seperate thread, but I've had no response, so lets keep the discussion going.
Regards
blakmax