Epoxy aluminum to aluminum?

These are coolant tubs with -16AN fittings. The fittings have a lip that fits nicely inside the tube and it's quite a tight fit. I can't tig weld. Would I be able to epoxy this joint or am I on the prowl for someone who can do this locally?

You can braze AL with the correct rod and a torch.

Sure you can. There are adhesives meant expressly for sticking fittings in tube. I know I've done some A/C line repairs this way.

I suspect that green Loctite would do the trick. Definitely pressure test it to 50psi minimum.

I would not trust any epoxy to hold pressure in this type of application. Not saying it won't work but I wouldn't trust it not to fail/leak. The aluminum braze rod that you use with a propane torch could work and that's what I would do if I was in a hurry. If I have the time I would look for someone who can TIG weld.

Well what do you know, I have a TIG welder and weld aluminum all the time.

I've been using marine tex to install fittings in aluminum rads and intake manifolds for years. Never had an issue. Many of the metal plumbing fittings come from the factory glued together instead of soldered now. Take a look at a brand new bathtub waste and overflow made from copper and brass. The tub shoe is glued to the copper pipe. I once used epoxy to mount the floor to a furmula ford because the welds and rivets kept failing.

Lotus glued the aluminum tub of the Elise together, so I don't see why not.

Porsche used what looks like green loctite or similar to adhere all the external coolant nipples into various coolant castings on the GT3 and other Metzger design engines.

unfortunately i've pulled 2 engines to remove the original failure prone nipples so that we could weld in new fittings. one of them actually failed on our dyno while i was pulling the first engine to preventatively weld them in.
i'd weld them for peace of mind in my project.

Aluminum brazing rod, uses a propane torch.

EDIT, Might not be a bad idea to braze it with a corresponding fitting torqued on both to act as a heat sink and make sure the sealing face doesn't warp.

Find someone with a MIG

plexus has e36m3 that bonds stronger than the aluminum (probably not stronger than 7075) but I've seen the al fail before the plexus

I've used JB weld on aluminum radiators on my cars, usually to deal with small leaks at the tank ends of the core. You need to use the real stuff, not the 5 minute version and do a bit of surface prep. Never tried this specific application with inserts, though.

If a failure isn't catastrophic, I'd try it here. If these things are impossible to access and/or it's a competition vehicle, I'd get the local fab shop to TIG it.

If gluing:

Prep,prep,prep...

Seriously.

Oxidation on the Aluminum is the biggest cause of failure

Good luck.

These are going on my RX7 so failure is not an option. I'll be heading down to a shop tomorrow to have them tigged up.

In reply to WilberM3:

Some PCA regions do not allow water cooled Metzger engines on track without welding the coolant fittings due to the adhesive failing.

DaveEstey wrote: These are going on my RX7 so failure is not an option. I'll be heading down to a shop tomorrow to have them tigged up.

Yep, this is the best thing to do on critical fittings.

Use 3M 5200 availible at Home Depot/Lowes. Once stuck together you would have to destroy tube and fitting to get them apart. Boat builders use this stuff to bond hulls to deck and anything else they never want to come apart.

I've used JB weld to glue an AL blockoff plate to my AL intake manifold, then ran 10 psi of boost for 40k hard miles. Never failed.

I would say brazing is the way to go unless you can find someone to tig them for you.

The epoxy is all of the strength in this joint and there is nothing else holding it so I see two issues: 1) As Tralfaz mentioned oxidation causes a weak bond. Surface prep (removing oxidation and preventing more from occurring) is critical and not the easiest. 2) The strength of epoxy is very temperature sensitive. The room cure stuff will be weak at normal coolant temperature.

I would find a way to get it welded.

I went with a stronger bond than epoxy

Wow! You went with the liquid aluminum glue. I wish I had a gun that would apply it.

A bit of free advice: Brace that big extension on the cooler. The heavy AN line torquing on that long arrangement WILL crack it.

There's a good reason why fittings are generally directly on a cooler.

This radiator came off a Roush Yates Sprint car and was already arranged that way. It survived 2,000 miles of that action before being retired.

I'll be cognizant of any possible forces when I run the lines. In this case it's about 3' until it turns into hard line.

So, in other words, it has 2,000 miles of stress microfractures and work hardening and they lifed it out before it failed.

I'm not suggesting anything massively overkill. Something as simple as supporting the weight of the AN line by ziptieing it to something nearby would probably work (as long as there could never be relative motion between the two)

Just bear in mind, Mazda built your car with an infamously failure-prone "Logicon", that failed mostly because they stuck a heavy wiring harness on its connector with insufficient support, so the solder points for the connector would crack and fail... It worked okay for a little while, but I remember writeups on how to fix broken Logicons when FCs were still newish cars.