That's a nice bendix assembly you've made and I know you have a good bit of time in it. Only other alternative I can see is to machine the end of compressor nut to a conical profile and machine a matching rubber cone insert for your bendix. A conical shape vs. spherical shape will self-center and eliminate the positive feedback de-centering. ...... /ramble off
Your "conical" idea is a good one, it's pretty much what I'm going to do, John and I spent some time last Friday afternoon discussing it, whether to do and internal or external cone will be the question. We pretty much decided to go the "internal" route but it will mean I'll have to make up another bendix assembly to suit, not too much of a problem, just time consuming.
I made the last bendix by hand and it took quite a while to get it to work correctly/smoothly. I sat in the "Beast" and John believes there will be sufficient room for the starter to be mounted as I had hoped. I just need some time to do the silly thing now.
John, I'll be "popping" the comp wheel off in the not too distant future and bringing to my work to machine it, just have to do some measurements etc..to decide on the cone size so as not to weaken the wheel hub but have enough to get sufficient traction for the "O" rings....watch this space...
The little "round house" on top of the FT with all the holes in it serves two purposes , one to allow unrestricted flow of air to the evaporators and secondly as a space for the 2 fuel manifolds ( kero and LPG preheat) to be mounted within , those holes have at least twice the flow area of the evaps
The actual evaporator inlets "meter" the airflow to ~10 % of the overall FT hole area (inducer area) the remaining 20% of primary air comes in through the more conventional cap/wall holes.
My evaps are only 304 stainless with 1.2mm WT , nothing fancy , the air and fuel flow should keep their temperatures low enough to prevent overheating problems, the 1.2mm thick wall gives a bit of insurance , the evaps in my micro engines are 800 series alloy as they are only 0.5mm-0.63mm WT .
It should be possible to meter the evap airflow by using suitably sized holes in the "round house" , just treat those holes as if they were normal flametube wall holes feeding 10-12% of your air to an "oversized" evaporator crossectional area, there will probably be a bit more pressure drop involved than normal , so definitely use bellmouthed metering holes
I use ~6 times inducer area as the total surface area of my evaporators for heat transfer/evaporation
We don't need to be be too pedantic about the areas , when Cameron got my first turbine kart he installed a much smaller turbo ( ~40% smaller ) onto the original flametube/combustor built for the TV84 turbo , and the combustor behaved quite OK with this "oversized" combination of evaps and wall holes ,........ a 1.1 lbs/sec mass flow into a 1.8 lbs/sec combustor .
Fuel is "discharged" into the flametube , the mass of fuel ignites , but because it isn't finely atomised its slow to burn , resulting in a flame exiting the turbine wheel and igniting the afterburner fuel spray .