The completed 8-can-pipe. Notice cat inspection and the associated feline brain power.

Prior to putting the components together I laid this out to show how the hot band would slip over the stopper and then the cold can with "PAM" would be tapped into the band, stopping at the wood insert. Notice that I had 4 pairs of can's with narrow bands joining the cans bottom-to-bottom. Now it is time to join these pairs of cans with wide bands, with the cans oriented top-to-top.

I only ran into troubles when I attempted to tap a set of cans into place, but I tapped against a band. They are just not stiff enough to take that abuse and that resulted in a split band. Always start with any can combination as shown above, and then connect groups of cans by tapping them together, with the band between the cans.

Although the "pipe" is meant to just be a ventilation duct, there is benefit to the pipe being able to at least maintain it's shape. As a test, I wrapped the 4 x 4 wood block in a towel and hung it at the mid-point of the cans as they were supported on each end by the chairs. There was very little deflecton so in the hot environment of a solar collector, just supporting it's own weight, which for aluminum is pretty light, I suspect these "pipes" made without glue, fasteners or tape will do quite well.

Update: A COMPLETE SUCCESS !

Before I jumped into the freezer/oven technique I also checked my cat can stopper and discovered if I shaved off about 1/8th inch then that would help the two cans to butt against each other at the center of the band.

I cut a few more bands and opened the ends of a few cans. I tossed cans into the freezer and the bands into the oven (at about 350 degrees F). Being impatient the cans were not inside either oven or freezer for more than a couple minutes. Hey, they are aluminum! They are suppose to conduct heat (or lose it) FAST !

I pulled out the band from the oven, then grabbed the can from the freezer and slid the can into the band with very little trouble. I would guess there is about a 10 second window before the metals acclimated enough toward room temperature to make significant movement difficult without the possibility of twisitng the can. I tossed the can/band combo into the oven, let them heat up and then pulled them out and grabbed another can from the freezer. Rinse and repeat BUT I did not rinse with water.

I attached 5 cans in series, bottom-to-bottom and top-to-top. Very tight fits, and with the 4 bands between 5 cans, there were only 2 "pin hole" size openings where the cans buckled inward. So the final product is almost watertight, definitely structurally much more sound, and the oven/freezer angle buys at least 5 to 10 seconds to get the cans joined very well. Success!

TIME TO IMPROVISE !

Here I'm now focusing on the issue of joining the cans in such a manner as to eliminate the problems of melting glue and mechanical connectors that would no doubt cause corrosion. So lets make the connections out of the same cans that were being used originally.

There is a lot of information in this picture so examine it carefully!  In the center of the picture is the Dremel with the #409 cut-off wheel.The can that is laying on it's side is braced against a piece of wood (darker wood) behind, and another brace at the top of the can toward the Dremel. This way I can just pull the board and can toward me slightly and rotate the can while the cutter does it's job. Notice the band already cut to the right.

If you look at the two cans that have been joined by a band on the left side of the photo, the band is a bit narrower since it's joining two cans that are oriented bottom-to-bottom. If you look real close, you can actually see the impression from the cans beneath, about 1/4-inch inside the cuts. The attachment is much stronger than the glued versions, and will not be affected by the heat.

Referring to the band on the right again, notice the cans, one in the cut band, and one outside the band. I knew that if I were lucky enough to push the can into the band easily, then I needed something to stop the can from going beyond a halfway point. These are "Fancy Feast" cans (cat food) nicely cleaned by the kitties and then cut with the same #409 cut-off wheel. They "can" act as a stop.

Besides having a great connection, there are a few other observations to bring to light. I only put the can in the freezer and left the band at room temperature. Once both cans were in place, I saw that I have probably a 1/4-inch gap between the cans on the inside. They could have traveled another 1/8th inch each to butt-up against each other. So here's the next question. Would air flow over the gap be adversely affected enough to warrant more effort to get the cans against each other, or would the heat that will be generated on the band be extracted more efficiently with the gap?

Heat generated would no doubt transfer from the band and through the cans anyway, so I will try the following. Regardless of the plastic coating on the can, I will heat the band (or the band plus the already attached can) to 300 degrees F in my oven.  The other can (or the first can) will go to the freezer. If the expansion of the heated band, and contraction of the can is more pronounced, I may easily get the extra 1/8th inch (down to the cat food can stopper).

What makes the aluminum can such a good conductor will work against me as the cold can will heat up fast, while the hot band will also contract pretty fast. Lets see what can be accomplished !