Experimental results
We implemented five types of experiments.
3-1) without electric ignition: We ignited the pilot by inserting a flame from the outlet of the combustor and
maintained the pilot flame. We then started the blower to test the stability. In the absence of a choke plate (6 in Fig.
2), the flame could not be sustained. 3-2) without electric ignition: We added a choke plate (6 in Fig.2) and
maintained the flamed pilot as described in case 1. We then started the blower and found that the flame could be
sustained. If the direction of injection was upstream and the injector was located near the center line, then the flame
was sustained. 3-3) We kept the pilot at the same location as used for the electric ignition system, and then we
ignited and generated a shock, but the pilot flame could not be sustained. 3-4) We added a front room. As a result,
the pilot flame was ignited and its stability was successfully maintained, with or without the convergent nozzle and
the turbine. In this setup, we observed that the center velocity of the jet was 25m/sec (for a cold jet) and the outlet
temperature was 280C at 25m/sec and 230C at 12m/sec (for a cold jet). The stable pilot flame and stable main fuel
flame were observed. Blue flames were observed in both cases. 3-5) For the turbine experiment, since the friction of
the ball bearing is rather large, the blower had difficulty in rotating the turbine in a cold combustor; however, once
the combustion started, the turbine also started rotating at a sufficiently fast rate to produce power for the
compressor and fan. The outlet temperature was very high in both these experiments, so we removed the gypsum
shield to increase the amount of cooling air around the shroud. We examined the blade for damage, but found no
damage near the shaft; all damage was slight and localized at the middle and tips of the blade. This suggests that the
amount of cooling air was quite sufficient.