I have done this a little, but with improvised equipment; all involved using a dry heat source.
After tacking just using the hot bit from an old Weller soldering iron, first attempts to actually hot-air weld were made using a paint stripper hot air gun and then a more controlled source from my heat-shrink sleeving gun to heat two adjacent edges sufficiently to get them to weld together - very hard to get enough, but not too much, focussed heat. Improvising more constricted nozzles and using filler rod (initially thin 3mm square cut from the edge of a sheet, then proper rod) gave better results but the speed of movement along the weld line had to be quite high to avoid over-heating. Sectioning it showed defects and the rod sitting in the apex or alternatively the weld pool had been blown away.
The breakthrough for me was switching to a lower-volume air source using my hot air soldering/rework station which has a highly controllable hot air flow (temp & flow rate) and a range of nozzles. At first I thought the air flow rate was too low, but the slower, more measured work rates allow a far better weld. I did find though that having enough hot air in the vicinity to keep the two workpieces sufficiently hot meant that the filler rod was too soft to get it pushed into the apex, so eventually I modified a soldering nozzle for welding by adding a one-inch filler rod sized metal tube alongside the hot air nozzle to (a) guide the rod down to the weld point, (b) shield the rod from most of the heat until it emerged to the weld pool, and (c) with a flared 'shoe' at the exit point to keep pressure on the rod into the corner. Using this rig I have been able to weld polyprop for various small covers, cases, low-stress fittings etc. (including running multiple beads) but nothing that I've ever done has been for use with chemicals or at pressure etc...
As with heating any plastics - needs good ventilation and usual spatter protection.
Dave
