Technology. Jim Hendrix. Oxford, Mississippi USA

Transcription

1 Survey of Deturbulator t Technology Jim Hendrix Oxford Aero Equipment, LLC Oxford, Mississippi USA

2 About the Author Jim Hendrix, retired research scientist (physical acoustics), electrical engineer, software engineer (digital signal processing), soaring pilot. Since 2003, Jim has logged sixty flight tests measuring effects of Sinha deturbulators on drag and aircraft performance. Resume:

3 About the Inventor Sumon K. Sinha holds the following degrees: Ph.D. in Mechanical Engineering, University of Miami, Coral Gables, Florida, May M.S. in Mechanical Engineering, University of Miami, Coral Gables, Florida, December B.Tech (Honors) in Mechanical Engineering, Indian Institute of Technology, Kharagpur, India, May Sumon has held faculty positions at Florida International University, the University of Nebraska Lincoln, and the University of Mississippi. He now works full time for his company, SinhaTech ( com) Resume: pdf

4 Data are from: Jim Hendrix About the Data Drag measurements from probes and sensors constructed by Jim. Aircraft sink rate measurements from flight data recorder. Dick Johnson (independent tester) Aircraft sink rate measurements from altimeter and chronograph readings. The Hendrix/Johnson data are the best available and are conclusive.

5 Drag Measurement Method Differential pressure sensor measures difference between aircraftpitot and drag probe pressures. Pitot sees stagnation pressure at nose of aircraft. Drag probe sees average stagnation ti pressures of a short array of Pitot tubes in boundary layer at trailing edge of wing. The pressure difference characterizes the reduced velocity of the boundary layer departing the wing on one side only.

6 Drag Measurement Instrumentation

7 Pressure Side Drag Data 2003 First Encouraging Results Lower surface, 53 span Green Yellow Red Clean wing One strip Two strips Dashed lines are percent change (see right scale).

8 More Pressure Side Drag 2004 Lower surface 158 span Green Red Clean wing One strip Dashed lines are percent change (see right scale).

9 Suction Side Drag Data

10 More Suction Side Drag Data

11 About Hendrix Sink Rate Measurements GPS altitudes logged at 4 second intervals. Linear curve fit to 10 samples (typically) gives a slope that indicates the sink rate. Sample decimation as necessary makes the data fully span the speed run. Illustration at right is a special case where performance at a single airspeed was changing and a sliding average trend was sought.

12 Early Performance Measurements 2005 Inboard 60% Suction Side Only Application (average of two data sets) Sink Rates Lift/Drag Progress asp

13 First Full Span Performance 2005 Suction Side Only Application Two Flights Averaged Much later data shows that the 50 kt point should touch the baseline; thus, airmovementraised raised the red curve slightly.

14 Important Issue Regarding Deturbulator Performance Measurements Vertical air movement scatters sink rate data more or less randomly. Normally, this is minimized by 1) carefully selecting weather conditions and flight altitudes, and 2) averaging sink rates from several flights. Deturbulator performance is not (yet) stable, so averaging gacross flights also minimizes real sink rate changes. Deturbulator sink rate changes are typically so large that scatter from air movement is usually small by comparison. Thus, examining individual flight tests is appropriate and necessary to see the true extent of deturbulator potential.

15 Better Full Span Performance 2005 Two Flights Plotted Separately Sink Rates Air movement lowered orange curve slightly. Should touch baseline at 50 kts. Lift/Drag Dots show percent improvement against right scale.

16 A Problem with Condensation Performance declined and even became negative during the summer months. Below is a plot of best performance improvement an any airspeed from different flights over a two year period. The scatter represents engineering and climatic i changes. The overall pattern suggests a humidity issue in the summer months.

17 New Deturbulator Gives 20% Improvement 2006 Two Flights Averaged Sink Rates Lift/Drag Notice that new deturbulator touches baseline at 65 kts instead of 50 kts as before. Also, notice extreme improvement at 40 kts. This repeats in Johnson tests the following December. Sparse data points hide structure that will show up in Johnson tests. Decided to take glider to Dick Johnson for independent testing! Progress asp

18 Pilot since MS in Aero Engineering from Stanford University in years in the aircraft and missile industry. Inventor of birdie head laser guidance device on smart bombs. regarded as one of the country s premier aeronautical engineers (Texas Instruments). Since 1960, published more than100 glider flight test test reports. Known globally for his glider performance reports. Who is Dick Johnson?

19 About Johnson Sink Rate Measurements Calibrates airspeed system against tknown instruments t using tethered th Pitot/static bomb. Carefully chooses weather conditions for low inversion and low wind shear above inversion i level. l (Winter conditions) Tows aircraft to 13,000 feet, noting air temperature every 1000 feet. Files 35, 40, 45,... Kt airspeeds (indicated). At each airspeed, notes start and end altitudes and elapsed time. Flies low speeds until minimum 400 foot loss. Flies high speeds until minimum 500 foot loss. Corrects pressure altitude losses to geometric altitude losses. Corrects measurements to sea level equivalents. Averages three to six flights to remove scatter from air movement. For these flights a Cambridge 302 flight data recorder was operated to collect data to corroborate Johnson s visual measurements and for further dtild detailed analysis by Jim Hendrix of Johnson s flights. Johnson Report: SSA 2007.ppt

20 Johnson Average of Six Flights Sink-Rates 13% at 48 CAS

21 Johnson Average of Six Flights L/D Ratios 13% at 48 CAS

22 Johnson Average of Three Flights Sink-Rates for Flights 1, 5 & 6 18% at 48 CAS Johnson discarded flights 2, 3 & 4 for too much scatter.

23 Johnson Average of Three Flights L/D Ratios for Flights 1, 5 & 6 18% at 48 CAS Johnson discarded flights 2, 3 & 4 for too much scatter.

24 Remember Vertical air movement scatters sink rate data more or less randomly. Normally, this isminimizedby 1) carefully selecting weather conditions and flight altitudes, and 2) averaging sink rates from several flights, as Johnson did. Deturbulator performance is not (yet) stable, so averaging across flights also minimizes real sink rate changes, as occurred in Johnson data. Deturbulator sink rate changes are typically so large that scatter from air movement is usually small by comparison, as we shall see upon further examination of Johnson log data. Thus, examining individual flight tests is appropriate and necessary to see the true extent of deturbulator potential, as we shall see....

25 Hendrix Analysis of Johnson s Log Data Plotted individual baseline flights to illustrate normal scatter and to contrast visually with deturbulated data. Plotted individual deturbulated flights to illustrate magnitude of deturbulated changes visually by way of contrast with baseline data. Fit a 2 nd order polynomial to each sink rate polar (solid red in the following plots). Connected selected sink rate points to highlight systematic deviations (blue in the following plots). Fit a 4 th order polynomial to each L/D polar (dashed red in the following plots). Progress Johnson Details.asp

26 Johnson Baseline Flights Illustrates normal data scatter!

27 Johnson Deturbulated Flights 1 4 Day 1 Altitude vs. Time Sink Rates L/D Ratios Compare magnitude of changes to baseline slide!

28 Johnson Deturbulated Flights 5 6 Day 2 Altitude vs. Time Sink Rates L/D Ratios Compare magnitude of changes to baseline slide!

29 Symmetries About 50 knot Airspeed Opposing extreme deviations of equal magnitude (82 fpm, 042m/s) 0.42 at 50 kts (48 kts calibrated) in adjacent flights. Symmetric deviations in three data points on each side of same airspeed. (One exception at 40 kts) These indicate real dynamic processes, not random scatter from air movement.

30 Detailed View of Symmetries in Three Flights Symmetries about 2 nd order polynomial were fit to individual polars for flights 2, 3 & 5. Clear indication of real, non random dynamic process. Half of the six deturbulated dflights showed this symmetry. Deviations at 50 kt are four times the worst errors Johnson normally measures.

31 Conclusions from Symmetry and Magnitude Patterns Flights must be analyzed individually to see true potential of the technology. Johnson s 3 rd flight measurement of 90% improvement at 50 kts is real. Engineering R&D is needed to obtain consistent performance and to broaden performance peaks.

32 New Design Installed after Johnson Tests / / Tested by Johnson 12/2006 Tested by Hendrix 12/2007 fiber reinforced membrane deeper substrate channels

33 Johnson s 3 rd Flight Repeated by Hendrix New Deturbulator Design One Year Later Data points 2.5 kts on each of peak show that t magic airspeed is now slightly faster Error bars show best and worst L/D in the data True performance is greater than 55:1 since flying too slowly

34 Altitude Profiles of Both Flights at 50 kts Johnson 12/2006 Hendrix 12/2007 Performance improves in both cases at constant airspeed. A sliding average performance analysis may reveal a dynamic process.

35 Sliding Average L/D at 50 kts IAS Four Second dintervals Both flights show same trend sequence. Characteristic trend is due to interaction between static pressure and ventilation change rates. Hendrix flight runs through trend faster due to faster ventilation through deeper substrate channels. Hendrix flight maximizes at 70:1 due to flying slower than magic airspeed.

36 Second 50 kt L/D Trend Hendrix 12/1/07 Hunting: angle-of-attack and deturbulator out of phase. Rotational inertia carries AOA past magic point Consistent oscillation period. Bottoms out consistently at baseline performance. Peaks consistently on nonhunting trend line. Not possible for pilot to deliberately fly this pattern. Each brown data point is a 28 second average. Confirms that new deturbulator performance matches old.

37 How is extreme performance possible? The objective is to coerce slightly detached flow over all wing surfaces in order to kill skin friction and reduce trailing edge turbulence! Two modifications accomplish this: Passive tape over leading edge Deturbulator tape at the transition reattach point

38 Normal Transition Bubble and Deturbulated Flow Smoke image of normal transition bubble Greg Cole Imagine this 1) flattened down, 2) detaching near leading edge, 3) skipping with grazing angles over deturbulator tape and 4) detaching the trailing flow. What would happen to skin friction?... Trailing turbulence? How important is skin friction in attached laminar flow? Is attached laminar flow the best we can do?

39 Oil Flow Image of Normal Transition Bubble

40 Oil Flow Image of Deturbulated Flow Forward Top Surface Initial detachment occurs about two inches behind leading edge Initial detachment occurs about two inches behind leading edge. Dappled region is not influenced by overhead flow and has no skin friction. Oil collected behind narrow tape shows reverse flow region where oil is smooth.

41 Oil Flow Image of Deturbulated Flow g Before and After Deturbulator

42 Clear Indication of Flow Detaching Behind Deturbulator Thickened oil immediately behind the deturbulator show where the flow is too high to further influence the oil.

43 Close Up of Deturbulated Flow Behind Deturbulator

44 Normal Pressure Side Oil Flow Showing Transition Bubble and Trailing Attached Turbulent Flow Also, Early Deturbulator Strip Deturbulating the flow

45 Bottom Flow with Leading Edge Tape and Top Side Deturbulator Flow is detached significantly only a few inches from leading edge. Detached flow over entire bottom surface. Transition bubble is removed.

46 Parallel Flight # Standard Cirrus 1969 best glide ratio = 36:1 ASW best glide ratio = 45:1 Test airspeed = 80 knots Deturbulated Standard Cirrus almost matched AS-28 for 4 minutes. 46

47 Parallel Flight # Standard Cirrus 1969 best glide ratio = 36:1 LS best glide ratio = 42:1 Test airspeed = 51 knots Deturbulated Standard Cirrus matched LS3 for 8 minutes. 47

48 Parallel Flight # Standard Cirrus 1969 best glide ratio = 36:1 Diana best glide ratio = 45:1 Test airspeed = 51 knots Deturbulated Standard Cirrus matched Diana 1 for 20 minutes. 48

49 New Team Member Jari Hyvärinen is modeling deturbulator operation using his own Linflow software that quickly simulates both large and MEMS scale aeroelastic processes.

50 Conclusions Extreme performance has been demonstrated! Convincing evidence shows that extreme performance is repeatable. More R & D is needed to fully understand deturbulated flow and to make it a practical reality.