Mechanical Bonds in Polyester -laminated Glass

I'm helping a friend with their one-off glass boat. It's apprx 30', built on a plug with C-Flex and then cloth. He recently talked to an engineer who wanted him to add a few more layers of glass in some bilge areas and I'm helping with that which brings me to my question.

When doing mechanical bonds he has been basically fully prepping the entire surface to be bonded with a 40-grit flap wheel on a grinder, aiming for 90%+ of the area roughed up. I'm much more familiar with epoxy than polyester and I know the poly is less adhesive but still this seems excessive to me. With heavy cloth or roving there's no efficient way to hit the low spots in the weave without removing quite a bit of material so with epoxy I typically hand 'sand' (really just scuff) the area and my understanding is that around 65-70% coverage is plenty for a mechanical bond. Can anyone enlighten me on what is considered sufficient scuffing for mechanical bonds using polyester? In this case the hull is something like 3/4" thick in this area and we're about to add 3 more layers of 1708 so I'm not particularly concerned about removal of 1/32" or less but I'm curious what the science and/or standard practices are on this. Presumably there is a point at which the mechanical bond is strong enough that the glass will tear rather than delaminate and any grinding beyond that point in order to scuff every last square centimeter is unnecessarily removing material and weakening the structure. Anyone know where that line is drawn when laying polyester-laminated glass on a fully cured poly-laminated glass substrate?

 

Is that a material with a layer of mat on one side ? Certainly having that facing the old glass, will improve the bond, I would try that flexible mesh abrasive that can get down into depressions, 3M and others make it, in various stiffness and aggressiveness of cut, looks like this:

 

A true mechanical bond would have to be keyed in. That means that the hole or groove needs to be larger on the inside. Grinding increases the surface area, which increases the total force needed to delaminate the repair. All bonds are chemical. However, when a new laminate is done over a not yet cured resin, there will be more cross-linking, which is stronger. Styrene will dissolve some of the old resin and help the mers of the new resin chemically bond with some of the old ones. The mers combine to create a polymer. They are the like the links of a chain. Polymers can be linear, branched, crosslinked and networked depending on the geometry of the links. Cleaning the surface is more important for good adhesion than sanding. However, sanding can help if the fibers of the previous laminate are not cut.

 

I use 16 grit discs to scuff up bond areas, works really well!
There’s never any doubt if it’s scuffed up enough.

 

The answer can vary depending on the exact resin, the type of glass (CSM, roving, cloth, uni, or biax), ambient conditions and time.

Some resins, especially in a thin laminate, can have excellent bonding characteristics even weeks later. Other resins are limited to a few days.

Flap disc's don't work well, use a normal sanding disc in a similar grit.

A fully cured laminate should have 100% of the surface sanded for a good bond.

In large boat (yacht) building, the surface maybe sanded about 50% if the laminate has been sitting for a while. 100% if left for extended periods of time.

 

I've successfully used sandblasting to clean the laminate surface without cutting the glass fibers. It is much faster than grinding, but makes a lot more dust.

 

If you have the dimpled surface to sand, those flexible mesh discs are very effective, and you can always use fibre discs in conjunction.

 

Sandblasting leaves a very grippy surface for a great mechanical bond, but I find it works best with vacuum lamination.
Grinding leaves a flatter surface, better suited to hand layup associated with smaller repairs.