There are two different situation that could be going on here, and I am not sure which to answer. But they actually both result in the same answer, so the short answer is don't do it. Epoxy bedding metals is almost never a good answer for a umber of reasons.
First, and assuming we are talking about a through hull fitting, it is never proper to epoxy bond a through hull fitting into the hull. When used in this application the epoxy becomes a structual element of the hull itself, and in most cases is actually stronger than the surrounding hull structure. This means there is no way to remove the through hull in the future except cut an even larger hole in the boat. Typically these future holes will be so large that they can't simply be filled, but instead require a full fiberglass filled patch job, which can be expensive, time consuming, and unnecessary.
Instead the proper way to bond a though hull into a boat is by first drilling an apropriate sized hole in the hull, then epoxy sealing the edges of that hole with neat (unthickened) epoxy. Finally you insert the flange into the hole, after applying siKaflex or some other Semi-structual caulk (4200, ect) to the flange of the through hull. This sealant is what makes the system water tight, not epoxy. Since it can be removed (though with great difficulty) in the future if it becomes necessary. Remember most of the water will be excluded by the force applied y the backing plate/screw fitting, the sealant just needs to prevent the drips.
Basically here the sikaflex is used as a gasket, while the load is taken by the bolt. Same reason you don't need to epoxy seal oil filters, or impeller housings.
A longer and more detailed set of directions can be found at Replacing Thru-Hulls and Seacocks Photo Gallery by Compass Marine at pbase.com
. This also avoids the problems discussed below.
If this is not for a through hull, but just attempting to bond bronze to fiberglass again epoxy is not the proper tool. This has to do with the characteristics of epoxy and pretty much all metals (though some are better than others). The issue is that while epoxy is an incredibly strong material, and will adhere to almost everything, is it also very brittle, so it can't flex or expand much. Of course nether can bronze, so we are trying to bond two very hard inflexible objects... This would be fine if they were dimensionally stable, but they aren't.
As the materials heat up and cool they actually change size. And since there is no flex in either of these materials the bond where they are joined will be put under tremendous preassure, since the force that is being opposed is found at the atomic level. So how bad is it? Well the change in size combined for the materials is .000,041 inch/inch/degree F. Not a lot really, but more than enough with cycle loading to rupture the joint between the two materials.
This is most easily observed when you apply a torch to stuck nuts To break them free. The heat causes the nut to expand faster than the bolt does, allowing you to break it free.
This requires us to instead join bronze with something more flexible like sikaflex, or 4200, even west systems G-flex. Basically something that has a little bit of give to reduce this thermal expansion load.