In addition to Ed's excellent response, I would like to add that the old technique of casting mounts and bearings in place is
alive and well. This was done for years using lead alloys. The lead alloys have been taken over by epoxies. However, as bearings
require service, it is usually the bearing boss that gets epoxied. An excellent example of this usage is the positioning of engine
supports in large ships. These engines are very often built into the ships during construction. The engine mounts are optically
aligned with laser and cast in place with a product called "Chalk Orange". The advantage of using this epoxy is that when it
cures, there is no shrinkage. I actually used this myself when I positioned my oil bearinged propeller shaft/tube in my sailboat
several years ago, as welding would have disturbed the shaft bearing alignment.
Steve

"Ed Huntress" wrote in message ...

"Existential Angst" wrote in message ...
Awl --

First, is "two-part epoxy" redundant? Does "epoxy" nec. imply two parts?

Not necessarily. See 3M Scotch-Weld.


Next, does the strength of a two-part epoxy vary with its setting time? Intuitively, I would think the longer the stronger.

In very general terms, yes.


The classic is "5-minute epoxy", but I believe I have seen 60 sec epoxy, and I have 90 minute epoxy.
I find 15 minutes to allow for a few parts without rushing, and without waiting forever to cure. Altho I have also found that
the epoxy is not really hard in that time.

Of the consumer-grade polyamine-cure epoxies (most of the stuff you're likely to encounter), expect a 48-hour time at room
temperature
for full cure, and about a week to get really hard.


The stuff I have now is Permatex (a 30 min epoxy), but DevCon and Loctite are big players, with big-time industrial uses, such
as epoxying bearings, instead of press-fitting them into their housings. I believe DevCon stated that epoxying bearings is in
fact the preferred method, being stronger than press fit.

The shear strengths are enormous, but require at least .010 on the radius for good bonding. So a slip-fit for epoxy would be
counterproductive.

You need at least 0.005" for a high-strength bond, metal-to-metal. A little more is better.


In some of my proposed assembly methods, I could use set screws, pins, epoxied pins, or just epoxy -- typically nominal 7/8
alum round in 16 ga SS tubing, or some such.

I thought epoxying would be slam dunk, and indeed it saves machining, but goddamm, you gotta pick yer poison, it seems. I
could see arguments for actually using a mechanical method AND epoxy, in some applications.

That's how many aircraft wing skins are assembled. The epoxy provides sheer strength, and rivets resist peel failure. It's
called rivet-bonding.


A feature of epoxy, which could be both good and bad, is its instability to heat. I think boiling water, or at least not that
much hotter, can undo epoxied parts.

Ordinary amine-cure epoxies lose strength in the mid-200s F. Some of the industrial products are good up to 350 F or more.


Iny thoughts?

Using epoxy for assembling metal parts is not a trivial thing, if you want decent performance. It's time to get on the phone
with Devcon, or 3M, etc., engineers.

--
Ed Huntress