First test--and how to test
I finally flew my modified cowl yesterday. Before I get into the result, I want to rant a little about testing.
It is very important to get good data as you test your mods. You can't just go out to the same altitude as last time and look at your ASI, and your temp gauge, and turn the wheel on your E6B, and then claim your bird goes 246 mph.
The errors add up...the instrument error, installation error, pitot static errors, the temp probe heats up with air friction and compression, on and on. Also, in order to compare this run with the last one, you must run each test at the same weight, same density altitude, fuel flow, rpm, and cg.
Here in the rockies, the lowest I can get in my area in the summertime is 9000 ft. d. a. That works out to about 7500 ft msl on a cool summer morning. So, I do all my testing at 9000 ft. da.
In the old days before GPS, I used to fly timed runs back and forth between VORs, and do the math. This is the way Beech and Cessna used to test. Now we have a much better way, the 4 legged GPS run.
First, go to the National Test Pilot School web site...
http://www.ntps.edu/HTML/Downloads/
and download the excel spreadsheet GPS PEC.
It will do the math for either 3 or 4 leg runs, but click on the tab for 4 legs, it is much better. There is an article there by Doug Gray with the background on the math used.
Procedure: Fly a big square in the sky...with each leg roughly 90 degrees apart, and record the ground speed and ground track off your GPS. Fly the pattern as precisely as you can, holding altitude and heading, and wait for at least 3 minutes after each turn for your speed to stabilize. I will often fly 3 or 4 of these patterns, and then go play around for a while.
When you get back home, enter the data for each of the 4 legs into the spreadsheet. It will take 3 of the 4 legs and calculate the wind speed and direction, and your groundspeed. It will take all 4 sets of three legs, and calculate an average for each, and a standard deviation of the groundspeeds.
The standard deviation is a measure of the quality of your data. If it is more than 1.0, either you didn't fly precisely enough, or the wind aloft was varying, or there was verticle movement of the airmass that caused the data to be inconsistant.
Living in the mountains, I need pretty calm winds, or I will get vertical movement of the airflow over the peaks. On a good quiet day, I will get std dev to be less than .5 for every run...sometimes I get lucky and it is zero.
So, now to yesterdays results...
In spite of a winds aloft forecast that said 5 to 10 kts at 12,000 ft, I had more like 20, and variable. I knew from the bouncing around that I was not going to get good data. In fact, the std dev turned out to be 4.7, which is the highest I have seen (I suppose some days would have been worse, but then I didn't even take data).
So, the fact that my resulting speed was a little faster means nothing...being 1.7 kts faster means little when the speed is varying 4.7 kts on average between the runs.
Perhaps meaningful, though, was that my CHTs were about 10 to 15 dF lower than I would expect at that power setting and OAT. Maybe I have improved the airflow after all?
Stay tuned!
John
