Calibrating Gen 2
The heart of the “logic” that drives our second-generation AOA tone generator is detailed in DOT/FAA/TC-18/7 Low Cost Accurate Angle of Attack System. One of the authors, aerodynamicist Dr. Dave Rogers, Professor Emeritus at the United States Naval Academy, has been assisting us with development of the Gen 2 system. One of the key insights in this FAA sponsored experiment was the mathematical relationship between differential pressures sensed by the angle of attack sensor and absolute angle of attack. This relationship is expressed as a simple polynomial function and is used by our software to measure angle of attack based on the coefficient of pressure data provided by the differential sensors. The ONSPEED software damps the AOA signal (which can be quite noisy, i.e., “jumpy”) using mathematical averaging and filtering. Averaging and filtering are handled by the software and occurs real time. The output is the pilot-friendly aural AOA logic that allows you to hear L/Dmax, ONSPEED and progressive stall warning.
Dr. Rogers maintains a series of excellent articles on his web site (nar-associates.com). Of background academic interest are his articles about fundamental AOA (L/Dmax) and how it is engineered into the airplane as well as how to derive ONSPEED and Carson’s Speed from that “built in” alpha. Another describes absolute AOA, which is what our system is measuring using coefficient of pressure. Absolute AOA is the difference between zero lift alpha and actual alpha and is a bit different from “geometric” AOA that we were taught in pilot training. Although most of us are wired to think AOA or airspeed, his article Angle of Attack Controls Airspeed explains how the two are integrally related and essentially one and the same. For anyone that wants to thoroughly understand the theory behind the operation of the ONSPEED aural AOA logic, these three articles are essential reading.
There are also a few other articles on the site worth mentioning. From a flight test perspective, Some Comments on Angle of Attack Systems Calibration was written to help us make sure we got the physics right with the ONSPEED system and forms the basis for how we calibrate. This article is an excellent summary, and if you only click on one link to read—this is the one. There are two other articles worth mentioning. One deals with determining flat plate area (parasite drag) from flight test and other explains the math behind and conduct of GPS “Horseshoe” runs that many experimenters are familiar with.
To simplify things, we have developed an Excel based spreadsheet that allows the pilot to input flight test data and generate the algorithm that the computer uses to measure angle of attack. These parameters include pressure altitude, manifold pressure, RPM and OAT which are used in conjunction with manufacturer’s charts to determine horsepower; IAS, which allows us to derive equivalent airspeed; fuel on board, which is required to determine aircraft weight as well as GPS ground speed and track for each leg. While I prefer to write this information down during test, Lenny has integrated EFIS data recording in the ONSPEED software, so it’s also practical to download data post-flight. So, if your airplane is EFIS-equipped, this is an option. If you fly behind round dials, then you’ll need an old-fashioned run card and pencil in addition to some form of GPS.
To conduct test, the pilot climbs to desired pressure altitude (29.92 set in the reference altimeter) and establishes aircraft configuration and speed (power setting/pitch) for each test point. Aircraft test configurations are gear/flaps UP, gear down/half flaps and gear down/full flaps. Suggested speed points are VMAX (or high-speed cruise), Carson’s Speed, L/Dmax for test weight (maximum range glide speed), estimated ONSPEED (either .8 x L/Dmax or 1.3 x VS), and stall speed plus 3-5 KTS. If gear and flaps are extended, then gear and flap limit speeds should be used for high speed data points. If you fly a Van’s type, keep in mind that for airplanes with a NACA airfoils (RV-3/4/6/7/8) that half-flap Vfe is 10 MPH greater than indicated Vfe for half flaps. Initial testing should be conducted at 6000-8000 feet, and at other pressure altitudes, as desired. Assuming three different aircraft configurations and five to seven speed points, 15-18 runs per pressure altitude are required. A “run” consists of 4 GPS legs.
At the conclusion of test, data is input into the spreadsheet. The critical output is shown in the figure below: the basic relationship between the two pressures measured and how they relate to absolute alpha. Note the linear relationship for the Dynon probe in the lower chart when the pressure difference is normalized for dynamic pressure. It turns out that different types of pressure sensors will have different pressure “signatures” requiring some testing and experimentation to properly calibrate. The bulk of our experience thus far has been with the Alpha Systems and Dynon probes, but future work will include the Garmin probe, AFS-style wing ports and a simple, homemade bent tube added to a conventional 90 degree bent pitot tube.
Our ultimate objective is to automate calibration, making it as simple as possible; but we’re not there yet and that may not prove to be possible. We simply don’t know at this point. We also plan to develop a system that allows folks to upload flight test data to the net and then do the number crunching for them. We’ll provide them with a custom software load for their airplane. In the meantime, we can utilize the GPS horseshoe speed runs and crunch numbers with Excel. We call this calibration technique “hard tuning.” Since we are a spare-time group of volunteers, I won’t guestimate any sort of timeline for automatic calibration or calibration net resources, since everything tends to prove a bit more difficult that first thought. It’s definitely an iterative process!