right
I’m sure in industrial applications there are more precise and exacting methods for flame oxidation, as well as chemical oxidation, that aren’t available to the home hobby person.
For home applications a hand-held propane torch does well enough, but to be successful you need to touch the tip of the inner blue cone of the flame to the surface to be bonded and move it along quickly to avoid overheating the plastic. The goal is not to heat the plastic in any way, and certainly not to melt it, but to temporarily oxidize it. You should adjust the propane torch so as to make the blue cone as distinct as possible, and I find it helps to do this in dim enough light to make the blue cone tip readily visible.
As you know, polyethylene (a polyolefin) is a rather chemically boring linear array of ethylene (C2H4)groups linked in a chain. The interior carbon atoms are all satiated with electrons with only protons sticking out. Oxidation, either chemical or flame, strips away some intact hydrogen atoms leaving at least some of those carbon atoms on the prowl for a stray electron. This makes the polyethylene more chemically active and less chemically inert.
But the effect is temporary. If you don’t go ahead and apply your epoxy soon enough, those hungry carbons will pick up a willing electron elsewhere and go back to sleep.
Good info
The flame processes I'm at least a little bit familiar with (using an oxygen-acetylene torch) are not related to the current topic, and not with working with plastic, but I was sure that a similar general principle regarding oxidation (or by the same token, often needed in metal work, the avoidance of oxidation) would apply. I'd just never paid any attention all those other times you've said what to do! Good info, especially in pointing out how heating of the plastic is to be kept to a mininum, and the time-frame issue. Thanks.
Fixed
So I tried my hand at plastic welding. I smoothed out a few scratches on the surface to get my feet wet. Then I tackled the holes at the bottom, the plastic has worn so thin that as soon as I touched heat to it the hole grew larger. So I got a fair amount of HDPE, heated it up with a heat gun until it was molten, and smeared it in. Then I got a metal tent stake, bent it in half to make a radiused smooth metal surface, heated it with a torch and smoothed the edges. Then for good measure I filled the entire valley with duct sealing putty. Spent an hour on the water with no leaks :)
Welding principles
What you experienced is exactly what I’d been thinking would happen, based only on metal-welding experience. What Pete said about complications when trying to adhere filler to un-melted base material is true in working with metals too, and might have left you with inadequate bonding between the very thin material you wished to avoid melting and your filler material. But, nothing ventured, nothing gained, or as my dad used to say, “we can’t do any worse than ruin it completely”, which always meant it’ll be no worse than it already is. Good for you for giving it a go.
If it eventually starts to leak, you can still build a cover for that indentation using your new-found welding ability, and that ought to be the last time you have to mess with it.
With any luck, THIS time was the last time.