Oops!
My mistake Simon, you are correct. I was using the program "convert" and hit "pounds (troy)" not "pounds (lb)" - this is why i much prefer metric!
So 0.38 lb/hp * 454 g/lb / 0.746 kW/HP = 231 g/kW (using BHP not PS).
Then using BSFC 83.7 g/kW-hour for 100% efficiency (most aviation fuels), makes the engine 36% efficient when running at optimum rpm.
I was suprised at my original figure since the most efficient small 4-stroke diesel is 45%, and 2-stroke is never as efficient (even with low bore/stroke ratio designs). I was designing a 2-stroke turbodiesel for fun, but dropped it when i realised that the slight power advantage of 2-stroke disappears when you look in detail. The scavenging process of even a uniflow design (like the Wilksh aeromotive) is not as good as 4-stroke, and at high turbo boost makes up for power. CAT have even developed hybrid generator turbos to take off excess blowdown power directly. Besides oil always gets into exhaust with 2-stroke.
I don't know your level of interest, but i believe that high boost turbodiesel has the potential to revolutionise the light helicopter market. The concept is that the turbine does all of the compression and expansion work, with a much smaller piston core replacing what would be the high pressure stage. Either the combustion pressure would be very high, or the engine would run very low compression ratios. This greatly reduces engine size and weight for a given power. There are also emerging technologies which overcome the combustion ignition rpm limitations of diesel engines. I have also considered alternative layouts and crank mechs to reduce the higher piston skirt wear which would occur.
Mart