Here are some more notes from flight 3 - the first ones were done in a rush as I was getting ready to head to the airport for a road trip. Now I that finally have a few hours of quiet time this evening, I looked at the data I recorded on flight 3. I found some glitches with the scripts I wrote to record and analyze the data, so it has taken some time to decode it. The process should get faster as I fix the bugs in the scripts.

The aircraft performance appears to be fairly good, especially considering that the wheel pants and landing gear leg fairings are not installed. The rate of climb at 80 KIAS 1000 ft MSL at 30 deg C (density altitude of about 2950 ft) at 1620 lb weight was about 1730 ft/mn. The air was a bit bumpy right after take-off, so this is only an approximate value. This rate of climb is perhaps 120 ft/mn lower than that predicted by a hypothetical performance model I created. It will be interesting to see how the actual performance changes as the engine gets broken in, the excess oil gets sucked out of the air filter, and I put the gear leg fairings and wheel pants on. Maybe my hypothetical performance model is not too far off the mark.

At 4500 ft, 75% power gave about 165 kt TAS (calculated from the IAS value, so this assumes zero error in the airspeed system - that testing is yet to come). Builders generally report an increase of about 15 kt when the wheel pants and gear leg fairings are installed. And, the speed for a given percent power will increase with altitude. So, it looks promising that the aircraft should at least come close to Van's performance figure of 182 kt TAS at 75% power at 8000 ft. Also note that the mixture was full rich for these two conditions - the engine would make more power if the mixture was leaned a bit, and the speed would be a knot or two faster. I think the engine is quite rich at full rich, possibly because I put too much oil on the K&N air filter - I found quite a bit of red filter oil had been sucked into the air induction snorkel after the first flight.

While doing the one hour with power varying between 65% and 75% at 4500 ft, I decided to do one airspeed calibration test point. I did a four leg box pattern at 145 KIAS on the analog ASI, using the event marker to indicate the times of interest in the recorded data (if I am recording data, the trigger on the Infinity stick grip will send an event marker to the recorded data). I analyzed the data later, and really wasn't happy with the quality of two of the legs, as the speed varied more than I would want. But, I averaged the data on each leg, then put the averages into the NTPS spreadsheet - it calculated a TAS of 151.5 kt, with a standard deviation of 0.09 kt, which is extremely low. Maybe too low to be true, considering the apparent poor quality of some of the input data. If I take the TAS of 151.5 kt, and assume a reasonable ram temperature recovery factor on my OAT (the actual recovery factor will be determined on a future test flight), I calculate a CAS of 139.5 kt, which means the airspeed system appears to read about 5 kt too fast at 145 KIAS. This is only one data point, based on somewhat questionable data, so it will be interesting to see how the full set of airspeed calibration tests turn out. Update 6 Oct 08 - I discovered that the static line to the alternate static valve had been knocked off, which was allowing cabin air into the static system. I don't know when this occurred, so I consider the airspeed test from this flight to be invalid.

I did two quick tests to see the rise in EGT at 65% power when leaning from full rich - I wanted to ensure that the engine was not running too lean at full rich. I found that the EGT would increase as below (EGTs are in deg C):

 FF      EGT1    EGT2    EGT3     EGT412.9      633     648     631      655  # full rich  -        -       -       -        - 9.3      752     766     748      777 9.2      753     767     749      777 9.1      761     775     755      782 9.0      765     780     757      783 8.9      765     780     757      783 # cylinder 3 peaks 8.8      767     782     757      783 8.7      770     787     752      783 # cylinder 4 peaks 8.6      770     785     757      783 8.5      770     786     757      783 8.4      770     787     755      782 # cylinder 1 peaks 8.3      770     787     754      781 8.2      770     787     752      780 # cylinder 2 peaks 8.1      770     787     752      779 8.0      770     787     749      777 7.9      769     786     746      775 7.8      767     785     745      773

Note that each cylinder had a rise of between 126 to 139 deg C (227 to 250 deg F) which means that full rich mixture was very rich. The cylinders hit peak EGT in a range of 0.7 USG/h fuel flow, which is a bit more than desired for optimum lean of peak EGT operations. Down the road I may work with Airflow Performance to get some injector with different orifice sizes to try to tighten up that spread a bit. Also, when I review the data recorded from the engine monitor, it looks like I need to pull the mixture back a lot more slowly, pausing at each fuel flow to let the EGT values stabilize. I'll repeat this test at various power settings on a future flight.