Buren has done up a design for a flight rated engine:

It will be good to not have to spend half an hour changing catalyst, and if we can get the catalyst lasting long enough it will make a good engine for the hovering vehicle.

For now I am focusing on getting the catalyst to last longer. My current theiory about the mechanism by which the catalyst material works is that the peroxide is absorbed into it (because its hydroscopic) where it meets the permanganate which also resides in the media. They react and steam leaves the media taking some permanganate with it.

By definition.... in order to have a "consumable" catalyst then you have to settle for loosing some catalyst. The trick will be to ensure that you can force the media to exercise enough physical conservatism.

How do you do that? Well I have a few Ideas. Firstly I think that the way the catalyst material works is probably more complicated than just...it gets absorbed then expelled. Indeed some peroxide must be absorbed in order for any reaction to take place, as the permanganate on the surface of the material would quickly get washed away (this can be seen in the purple plume at the beginning of tests).  This absorption/expulsion system would seem to be unstable, as it can't absorb peroxide and expel gas at the same time. If that is the case I don't really understand how the catalyst can work at all in this case, but it must reache an equliberium.

If we suppose that the catalyst material has a "regression rate", then the question would be how to slow down this rate. At at the moment the catalyst is likely releasing catalyst too quickly to be entirely made use of by the peroxide, how to slow down the regression rate? Assuming the rate is linear than an obvious answer would be to simply increase its thickness. Previously I have only been using what zeolite I could easily get, which is small flakes about 2mm in size, and is designed to absorb oil spills. I have ordered a kew kind of zeolite which is designed as a "molecular sieve" which does come in balls and pellets of all sizes, although I have only been able to find one supplier in Australia and their stock os limited to pellets 3-5mm in diameter. The other interesting thing about the molecular sieve is that it comes in different types, which are able to absorb different sized molecules, hence the name "sieve". The sizes available range from 3 to 10 angstroms. I have ordered some 3 and 10 angstrom pellets. I have been looking for activated alumina but haven't been able to find a supplier who will sell in anything less than a 25kg bag. I would also like to experiment with adding in other inert components to the permanganate solution which could serve to lower the regression rate. This could be anything (e.g. salt) which slows the release of permanganate.