Balcony Shoot #9 - Experimental Catherine Wheel

Date: 2003-09-23


I made this wheel mainly from photos I found on a google images search. I've never seen one in real life, or read texts describing their construction.

The tube was spiral wound from dry craft, over-painted with a little flour paste to hold it together, and dried with a hairdryer. The tube was then rammed with the pinwheel driver composition tested previously.

To bend the tube without rupture I gave it a generous coating of dilute PVA glue, then after a few minutes of softening, rolled the tube partially flat with a bottle against the work bench. The tube was then coiled onto a Coke can and twitched into a flat coil with some tie-wraps.

Once the tube dried to slightly tacky I stuck it to a cardboard circle, painted in PVA to harden it. The tube was given 2 days of sun drying before test. The fuse was my standard meal/NC twisted tissue type.


It got stuck and had low thrust.

I actually posted on rec.pyrotechnics asking about this kind of wheel construction, but got impatient and built this prototype first as an experiment. I eventually got a reply but that was after I had finished. The poster was quite informative, and confirmed my suspicions of the critical design details and also told me that commercial wheel tubes are rolled in a mangle. Convergent engineering I guess :-)

I suspect that the generous quantity of PVA glue used in the construction soaked into the powder core and changed its properties. The composition did not appear to produce the same amount of thrust as it had in the dry tube test. I also believe it hampered correct burning-away of the tube with the composition.

Eventually the tube cross-fired, which was not unexpected. If I try this again I'll probably add cardboard or foil thermal insulation between the layers. I'll also take the time to make a larger axle (and more solid - probably wood) so a larger moment is created by the available thrust. I am not sure if the larger moment of inertia will mask any improvement, inertia should only limit acceleration, as long as the static friction is overcome it should at least turn.


title type size
Burn Video video/x-msvideo 8.632 Mbytes
Pre-Test Picture image/jpeg 24.413 kbytes