Fiberglassing aluminium.

This is an overly simplistic view isn't it, not to mention not at all practical, about any coating, on any material, let alone aluminum. Coating durability is mostly about care, once the physical attributes are accepted.

Epoxy and epoxy paint aren't the same thing. As I mentioned earlier and Paul has reinforced, a lower modulus formulation (again) like G-Flex or similar. This will let flimsy, less then well support aluminum skins move around without harming the bonded joint.

 

When dissimilar materials are bonded together by adhesive, the CTE of the plate must approximate that of the base material. Aluminum has a CTE of about 12.5 x 10-6.

Epoxy resin is in the high end of CTE being 33 x 10-6 while polyester with a CTE of 14 x 10-6 is closer to aluminum. Solid glass has a CTE of 5 x 10-6 so mixing glass and resin in the right proportion, a laminate can be engineered to have a property closer to that of aluminum. It needs to be isotropic as well as aluminum is isotropic. Using the Rule of mixture method, a resin rich glass/polyester laminate (0.33 Gc/0.17 Fv) with predominant CSM material (low tech) or a glass/epoxy laminate (0.5 Gc/0.32 Fv) with high glass content and a quadriaxial layup (high tech) is will yield a laminate with a CTE of 12.5 x 10-6.

Using the equation posted earlier, a low modulus adhesive (3M 2216) with a CTE of 135 x 10-6, 0.062” thick x 2.0” wide bond will exert a horizontal load of 28 lbs. at 50 degreeF temperature differential. The bondline internal stress F2 would be 28/(w x t)=225 psi, well below the limit of 1700 psi shear of the adhesive. Use the 2216 translucent variety for salt water operation.

If the parent and plate material has a much higher CTE differential, Sika 252 with up to 300% elongation should be investigated. This takes care of the high CTE load. Just don’t use carbon fiber laminate and aluminum. The two are at the extreme end of the thermal stress and the galvanic differential.

 

Thanks
I'm afraid it all puzzle me.
Glass is lesser stretchable then AL, how can that be possible that it becomes more stretchable when surrounded by more flexible material? Also the reaction to temperature of fiberglass and that of AL are dissimilar and when, exposed to the sun, a combination like, a rectangular aluminium beam, fiberglassed at the end, over a catamaran hull, as a mean of bonding the two together, as in the case mentioned at the start of this thread, gives me the feelings that, it will not survive the test of time. I would much appreciate to know from those who have had some go at it and have proven results. Thank you all

 

The rule of mixture formula is CTElaminate= (CTEfiber x Vffiber)+(CTEresin x Vfresin) with VFresin=(1-VFfiber). Note that the proportions is in Volume fraction and not the traditional way of specifying Glass content (Gc) by weight. You have to convert Gc to Volume fraction.

CTE of fiber (or "melted" glass) is held constant as it is common. It is the choice of resin that changes. Polyester a little more "stretchable" than aluminum. epoxy about 2.6x more. Mixing a material with "less" and a material with "more" will yield a "just right" mixture. It is a general formula like mixing rubber and plastic to yield a material with "just right" flex like ABS plastic.

So, mixing the resin and fiber in the right proportion will yield a target value of 12.5 x 10-6 of aluminum.

Of course, the formula is a simplification as other factors (fiber orientation, fiber length and diameter) will skew the result but for preliminary sizing, this works. Math is a more practical way of expressing things rather than fabricating (guessing) which combination of fiber/resin will yield the right CTE. This is called fiddling, not knowing what direction the result will be.