Joining stainless steel to aluminium

[RayW]

I am in the process of sorting out a cooling system for my Sphinx engine and have bought a nice piece of 4inch diameter stainless steel tubing to use as the cooling tank.
If I made a base of aluminium and glued it into the end of the cylinder with Araldite or JB Weld, would I run into any problems with reaction between the ali and the stainless as I understand that this can be a problem where the two metals are in direct contact with each other?

[cnr6400]

Hi Ray, I'd suggest not risking it - you could instead find a small sheet of stainless for your tank end and glue that on. Or better yet, make a flanged plug with an o-ring groove on the od from a slice of stainless round bar. If you have a lathe and faceplate it would not be a big job. No glue and no galvanic corrosion risk that way.

[Allen Smithee]

Stainless and Aluminium have very different expansion coefficients, so as soon at the temperature changed your glued joint would shear and tear apart.

AS

[RayW]

Thanks guys for your replies. You have confirmed what I had suspected. I will buy a thick disc of stainless and turn a step on it to locate in the bottom of the cylinder, then glue it in. The only reason I was considering aluminium was because I had some ideal pieces available.
The other option to gluing, of course, would be to silver solder.
I will post updates on my Sphinx thread.

[cnr6400]

Hi Ray, Just food for thought - you could do the joint without glue, with screws or drive rivets, if the joint were made as in the attached sketch. Also avoids difficulty and discoloration risk with silver solder. Glue may be messy and with repeated heat / cool cycles of this cooling tank the glue may crack and leaks may occur. If you use a good quality o-ring (plumbing store vs dollar store) it can expand and contract with the tank and maintain a seal.

[Jasonb]

The usual old bean tin seems a lot less bother

Would there really be that much differential movement on a 3-4" dia open ali/stainless tank from ambient to say 70deg C  to make it tear apart?

[RayW]

Thanks for the suggestion cnr6400 (sorry, don't know your name). The discolouration issue with silver soldering would not be a problem as the tank will be painted anyway, but your idea is definitely another way to consider.
As Jason says, I am not sure if there would be sufficient expansion/ contraction differential between the two metals to cause a problem.The more relevant issue would be the possible reaction between them if they were in direct contact with each other.

[cnr6400]

Jason, a 100 mm dia stainless steel tank could expand 0.05 mm or more with a change from 20 to 70 degrees C. I say "or more" because the coefficient of thermal expansion for stainless grades varies from 10 to 17 x10^^-6 mm/mm and 0.05 mm number was calculated with 10 as the coefficient. The equation I used is expansion amount = coeff of therm expansion x length/dia of object x change in temperature.

If the base disk is on a cold bench and the tank tube heats faster than the disk (since it is thinner and has large area exposed to the coolant) and moves even 0.02mm relative to the disk, after a few cycles the epoxy will likely crack, because it is brittle and will tolerate almost no movement. I wouldn't risk it, that's why I suggested the o-ring. Silicone sealant might also work, and would be better than glue because it is flexible, unlike epoxy.

Mind you your bean tin idea is far less complicated, just as effective for cooling, and you get a nice bean side dish for your lunch or dinner as well!  I can't beat that! 

CNR6400 aka Jeff

[Jasonb]

I've never found Araldite to be brittle, has quite a bit of flex in it

Would it not be an advantage that the tube heats more than the base as that would counteract the greater expansion coefficient of the aluminium base

[cnr6400]

I'm only making statements based on my own experience in my shop and in the day job (35 years) with glues and sealants and o-rings in many types of systems using warm or hot water based fluids with various metal tubes and vessels, or assys of mixed metals / plastic tubes and vessels.

In my experience glue joints have almost always cracked and leaked in this sort of service, o-rings, rubber flat seals, and silicone sealants have not. If there is hot and cold cycling, cracks in the glue joints happen faster. If there are pressure changes, however small, cracks in glue joints happen faster yet.

Your mileage may vary. My best wishes for Ray to make a successful tank that doesn't ever have leak issues!