I know that my approach looks weird but the engine is weird too.
This particular Jett60 is a hybrid of old and new parts and does not run like "normal" Jett. Specifically, it could not reach the ground RPM that are expected from Jett60 on certain props., no matter what I did with variables like fuel, venturis, NVA, glow plugs etc. This deficiency manifested itself very clearly when the engine used Jett's header/muffler only and there was no pipe.
Without going into too many technical details that are described in my previous posts, I can tell you that this engine has some form of fuel transfer timing problem.
You probably know that timing is a very sensitive mechanism and minute differences in ports geometry and surface roughness in critical areas have large impact on the overall engine performance, power, RPM and torque.
Yesterday's tests showed, for the first time, that this engine run better with pipe. Perhaps I was simply lucky when I set the pipe working length (18" and 1/8"), but the ground RPM a little bit below full 2 cycling were much higher than when using only header/muffler with other variables unchanged.
This means that the pipe is "helping" the engine to improve the timing problem, whatever the true nature of this problem is.
The RPM increase on 3blade carbon composite 11' and 7/8'x4" Bolly prop. is especially encouraging and, in my opinion, allows to start thinking about the stunt plane, lets call it the Advanced Stunt Trainer (AST) that will be build around this engine using this (or equivalent) prop and the pipe.
The entire power plant (Jett60 with header, propeller, propeller nut, pipe, pipe silicone connector and the connector clamp) weights about 500 grams (17.6 oz.).
660 in.square wing, lightly build with bellcrank, bellcrank mount, bellcrank to pushrod wire connection, leadouts and light heat shrinkable covering - 375 grams.
Partially build up (front portion only) fuselage with partially hollowed maple engine mounts, Sullivan fuel tank, carbon composite landing gear, light wheels and covered with light covering - 450 grams. Tail surfaces - 80 grams (very conservative). RTF model estimated weight -1405 grams (49.56 oz.). Even if I underestimated the RTF weight by 3 oz., the plane should fly ok for the training purposes with the wings loading around 11 oz./sq.ft.
I have learned to successfully mix different structural materials (foam, balsa, carbon composite stiffeners, Depron, various coverings etc.) and can build very lightly, maintaining reasonable stiffness and strength of critical elements.