I took the actual heading and created a third set of numbers
Run 1 ? (A) DA 8000?
_________|Vwind | WDir | Vtrue
Vg | Track | Kts | Deg | Kts
161 | 359 | 25.1 | 70.0 | 170.6
194 | 270 | 23.8 | 72.0 | 171.4
173 | 170 | 22.7 | 69.0 | 170.1
148 | 084 | 24.6 | 67.0 | 171.7
_____Avg | 24.1 | 69.5 | 171.0
Run-2 (A) - DA 8000'
161 | 355 | 24.7 | 66.0 | 170.7
195 | 256 | 29.4 | 60.0 | 167.0
175 | 170 | 33.6 | 71.0 | 172.7
140 | 086 | 27.6 | 77.0 | 167.4
_____Avg | 28.8 | 68.3 | 169.4
Run 1 - TAS Spreadsheet M: 169.2 | T: 168.5 | A: 171.0
Run 2 - TAS Spreadsheet M: 167.1 | T: 167.0 | A: 169.4
The altitude for each of these legs varied (pressure & density). Is there a method to "normalize" this data to the same altitude & hence have results that are comparable to each other?
RV builder Doug Gray came up with a
very nice method to calculate TAS using ground speeds and tracks from three legs. I showed his method to the National Test Pilot School, and they liked it so much they adopted it as one of their
standard methods. They did come up with a very clever enhancement that uses data from four legs. The four leg method does four TAS calculations, ignoring data from a different leg each time. If the data is high quality, all four TAS calculations are similar. If the four TAS calculations differ, that tells you that at least one of the legs has suspect data. You don't know which leg is bad (or if maybe they are all bad), so you throw out all four legs.
I put this data in their spreadsheet, using the four leg option. It tells me that the first four legs have a standard deviation of 0.7 kt, which is acceptable. I shoot for 0.5 or lower myself, but anything less than 1 is OK. The second four legs have a standard deviation of 2.7 kt, so I would throw out that data, as it is not good quality.
You asked about normalizing data. The conditions at your two test points are so similar (the average DA is only about 70 ft lower for the second set of four legs) that the effect on TAS would be less than the typical error of the test technique. i.e. there is no point to worrying about normalizing when comparing two test points at such similar conditions.
Typically data would be normalized to standard temperature, a standard power, and standard weight. You need a large data set, implying many flights of data, to feed typical data analysis routines. I've done that sort of thing for my RV-8, because I do flight test for a living, and I enjoy it. But it was a lot of work, first to gather good quality data over many flights, then to analyze it, and finally to present it in a format suitable for my POH.
My general approach to speed vs power flight testing is to first
determine ASI instrument error, then very carefully determine the
airspeed system errors over a wide range of speeds, using a four leg method at each target IAS. Then, when doing cruise performance testing, I record IAS, pressure altitude, OAT, weight and the data to determine engine power - this is repeated at a range of powers at various altitudes. I do not do a multiple leg run at each speed, as it takes a huge amount of time for the speed to fully stabilize after the turn onto each new leg. Instead, I base everything on IAS, after correction for the errors in my airspeed system.
I then
normalize each speed vs power point to sea level, standard temperature and standard weight, plot and draw a line through the sea of test points. That line is declared as the average performance, and I expand it to give predicted performance at other altitudes, weights and temperatures. When I'm on a cross country flight I often compare the TAS against the values predicted in my POH, and I'm usually within a kt once I apply the known corrections for the errors in my airspeed system.