Well, as the author of this item, I am somewhat disappointed at the response.
Since I submitted my comments to EASA, I have written two papers on this. The server seems to be down so I can't copy the links but I will as soon as possible.
The regs they are relying on have been in effect for many years and do NOT prevent bond failures unless they are applied in a specific manner, and alternative interpretations abound. In the FAA system, the regulations FAR Part 14 2x.603 and 2x.605 are supported by a number of Advisory Circulars (ACs) AC 20-107B, AC 23-20, AC23-19 and these can also be interpreted in an inadequate manner and rely on a Policy Statement PS ACE 100-2005-10038 before there is even any mention of the appropriate test method to prevent adhesion bond failures. The FAA propose to manage their composites (bonding) issues by establishing a training course and I have been actively involved in developing that course by setting out the road map for people to follow such that they are aware of the requirements.
Adhesive bonding relies on chemical reactions at the interface and for metals, these chemical bonds usually involve surface oxides on the metal. Many metals (aluminum especially) have an affinity for forming a hydrated oxide and to do this the chemical bonds to the unhydrated oxide dissociate leading to interfacial disbonding. The method for preventing these failures is to use a surface preparation process which results in a hydration resistant interface. Short term strength and fatigue tests will not discriminate sufficiently to ensure inadequate processes are rejected. Hence the EASA policy will not prevent bond failures and for principal structural elements such as rotor blades, that policy is flawed.
Another issue I raised was the fact that damage tolerance as it is currently implemented relies on testing where artificial disbonds (teflon etc.) are embedded in the structure and tested to show a sufficient level of structural integrity. That is fine if the surrounding interface is in good condition but in the situation where the interface is partially degraded, the surrounding interface around a disbond may well be below strength already, so tests which rely on pristine adhesive around the defect are meaningless.
Next is the issue of porosity in an adhesive bond. Unless the extent of the porosity is known (and that can be difficult to quantify on a structure) how do you model this by a teflon insert? Again in this case, the assumption with teflon inserts is that the surrounding adhesive maintains full strength and for a porous bond this may not be the case.
I'll post the links to the papers as soon as I can.
Regards
Blakmax
