OK. That makes sense. I would put any orifice as close to the Starter as possible, so that the pipework is still as high a pressure as possible.Renegadenemo wrote: ↑Mon Jun 25, 2018 2:29 pmThe problem there is that all the delivery plumbing is for the HP setup so it's all inch bore and that's without all the valves and convolutions in the pipework. It's easy to turn the pressure down to 40psi but the plumbing then won't flow sufficient volume to be of any use. To get sufficient volume down the pipe we need to push it with higher pressures but then it expands inside the start turbine case and mostly escapes around and past the turbine without doing any useful work. Indeed, condensation can be seen forming a foot ahead of the turbine exhaust showing that the gas is still expanding all the way out there. it's incredibly wasteful. The answer looks to be the hybrid LP/HP starter with the nozzle plate. That way the gas can only expand directly out of the nozzles and onto the turbine vanes. It must form an orderly queue to get out of the case and must go via a vane on the way. Our current nozzle plate doesn't have an o-ring seal around its outer edge so we have a loss there but initial tests suggested it halved our gas consumption for the same engine speed. The noise it makes, however, is horrendous. an awful ear-piercing scream.
The LP Starters and Start valves I worked with (Trents & 211's), had connections around 3 inch diameter - in order to flow enough air at 40psi to actually work. The start valves were butterfly types, mounted fairly close to the starter - sometimes even right on the starter inlet. In contrast, the HP systems (Olympus) were around 1 inch or 1.5inch diameter. Could even have had common parts with K7's system.
The HP starter sounds like it was a Impulse turbine, with HP air being 'shot'/expanded through fixed nozzles directly into the turbine rotor (so there are no stator vanes) - whereas the LP starter was probably more of a Reaction (*) Turbine, with a set of stator vanes before the rotor (which increase the velocity of the aire, before it goes into the rotor, where it expands).
(* - strictly probably a 50% Reaction/50% Impulse Turbine - which is normal for gas turbines)
You could perhaps add an O-ring to better seal your plate....? The awful scream is probably unavoidable.....