The flame thrower was built using about half of the drawings from the construction manual, the rest of the parts (combustor, diffuser, bearing mounting arrangement and tunnel, jetpipe, casing) were designed by the person that built it. He likes to build everything himself and decided to experiment with a few of his own ideas on that one and brought it to show me and others what problems he was having with it.
The engine that i brought was one that i had been working to test a few new ideas. it was not until about 12.30 on friday night before the competition that i decided to stop messing about with it and go home and begin to pack up ready to go to coventry in the morning.
it was brought as it was. the jetpipe happened to be 7mm larger at the nozzle. from the hundred or so runs that it had done that week, and the various tests that had been done, that was the nozzle that just happend to be fitted at the time. Hence the low thrust 15 or 16lb.

the first run was at 145k rpm. thrust arround 14lb egt about 650c. im only guessing at the temp as i didnt write down any figures at the time.

then Rob rowbotham ran his merlin. the results went in the computer and showed that I was slightly ahead.
then i think rob ran again at a different speed and got a slightly worse result than his first run.
i ran mine again at 150k. the thrust was around 15.5 or 16lb, egt 680 or something like that. the SFC was better than my first run.
then rob ran a second engine as it had clearly become a battle between us two by then. he got the lowest results of all three runs that time.

I still dont know how it won "falsely". there is no set speeds or thrust level that has to be acheived. i could have run at idle and still be within the rules.
the idea is to run for 2 mins at any constant speed while the fuel tank is measured and the amount of thrust is divided by the amount of fuel used. this gives the SFC (specific fuel consumption). whoever gets the lowest SFC wins.
Every engine is mounted on the same rig and measured with the same instruments so any calibration errors (of which there was some but kept quiet about) were the same for everyone.