DS, It is even mor horrifying to know that 76% of US manufacturers still use an average shear stress design approach (FAA Workdshop Seattle 2004).
This must be backed up by a plethora of tests at coupon, detail, element, subcomponent and component tests. If they designed using Hart-Smith's equations to calculate the load capacity of the adhesive and then designed the joint to achieve that load capacity then every adhesive joint would NEVER fail. Every coupon..... etc. test would break the metal, so you could reduce the number of tests by an order of magnitude, thus saving millions on certification costs. As well as that materials substitution requires a simple calculation to show equivalent joint strength. In the current method, a complete substantiation program may be necessary.
If the same approach was used for repair design and that was combinded with reliable validation of bonding processes, then such repairs would never fail, thus reducing maintenance costs.
I know of one repair on a 777 TE flap that has been repeated by a repair station crew four times. You'd think that they would relaise that there is a problem.
The real root of the problem is that the FARs JARs etc. do NOT prevent one form of structural failure, that of adhesion failure at the interface. Current advisory circulars also do not address the issue.
