Bonjour, Monsieur Avior;
Je ne parle pas Francais bien, so, je escrit en Anglais...
Anodized surfaces are not the best for adhesive bonding. It is better to have a 'passivated' surface with either CHROMATE or PHOSPHATE anions. I have provided a procedure for treating aluminum previously in this thread, and have also provided a link to a very comprehensive set of procedures for preparing substrates for adhesive bonding. See:
http://www.on-hand.com/Manuals/SurfacePreparation.pdf
So, first, please consider using a chromate cleaning / passivation process. Second, do not use tap water to rinse after cleaning / passivation - DO use distilled / deionized water. If too costly, consider using rainwater (discard first five minutes of flow - this is the roof rinse water).
Lastly, do consider taking advantage of the information posted by Mr. Schwarz (who has shared the results of his efforts with all of us), especially in his conclusion that methacrylate adhesives offer a good compromise in performance qualities. Yes, I know from professional experience that epoxies have and are used for hi-performance bonding in aircraft and race cars, however, the practicality for a home-builder must be considered, and Mr. Schwarz has provided much good information on his testing and using methacrylate adhesives. He has found they provide a substantial bond between aluminum, without extensive or expensive surface preparation.
Also, I would recommend using some rivets (even pop rivets) - maybe one in five or less (from the riveting schedule) - as locators or jigs in the adhesive bonding process, and definitely during the adhesive application / curing part of the operation (not post bonding). My experience (professional) is that lap joints, properly bonded, do not experience 'peel' failure. This is especially true if the joint design is focused upon minimizing surface drag (surface skin is butt-joined & bonded to a member / rib underneath).
Orion, a 'chicken rivet' isn't necessary IF the joint design is done properly. Rivets are stress risers and can accelerate / concentrate fatigue failure unless used in quantity to spread the stresses. As for applying pressure to the 'lap joint' during the adhesive 'set up' time, consider the use of elastic straps, or bags filled with water (spreads the pressure uniformly).
Even a layer of mastic beneath a riveted joint will both keep out water and spread the stresses ('deconcentrate' stress) and increase reliability. Unfortunately, it adds weight to the aircraft.
Well, that's enough for now. Gary, Orion, I thank both of you for your generous & willingness to share your knowledge with us folk who haven't done as much as you. You make this site a pleasure to visit.
D'accord, Avior, J'espere c'est suffice.
All the best & fly safe,
MalcolmW