GPS Measurement of Aircraft Glide Ratio

Scott Kurowski, Plus One Flyers member, 12/18/2006

While working on a ditch-risk analysis for flights to Catalina Island (AVX) I wondered how reliable the POH glide ratio that I had been using for years might really be.  Ditching at sea in fixed-gear aircraft is a highly unfavorable forced landing.  As a relative confidence benchmark (and to have yet another reason to go flying) I decided to empirically determine typical engine-idle glide ratios in a few aircraft I regularly flew, and performed without any particularly finely-tuned technique.  Below are my methods and findings using a portable Garmin GPSMap 296.

By flying a square box and measuring the GPS ground speeds at a uniform IAS on at least three headings 90o apart, TAS and wind can be mathematically determined and the wind nullified for glide ratio. In the vector parallelogram (figure 1) below, the fourth leg along 090 is omitted for clarity. The resultant vectors g1-g4 are the measured GPS ground speeds for each compass quadrant heading, w is wind speed, and a is aircraft TAS.

Figure 1 – Vector Parallelogram for GPS Wind and True Airspeed Computations

The formulae below can be derived [references 2,3] from the above vector parallelogram figure.


       - OR - 

 

 

 

 

TAS

Wind Speed

Wind Direction using g3

Wind Direction using g4

The C-172N manual diagram gives a 9.1 glide ratio [1]. To test this in a controlled setting I flew out with a friend to the desert east of San Diego in N6360D and N4975F to take data measurements using my GPSMap 296 in fair weather high over Borrego Valley airfield (L08). I selected L08 for the multiple purposes of local wind uniformity, light aircraft traffic, class E airspace, having an emergency landing field below, and to have a well-defined waypoint fix for the GPS unit to measure ground speed. Both aircraft were configured during all measurements with flaps at 0o and trimmed at or near neutral control pressure in both level flight and carburetor-heated engine-idle glides.

We determined that hand-recorded measurements from the GPS unit’s screen worked well enough in cases, but it was far better to simply hand-record the GPS unit’s timestamps of the various measurement events on a printed form for the measurement plan, and let the GPS unit capture our data for us.

After the flights I downloaded the detailed GPS track data log from the GPS unit [6] and loaded it into a spreadsheet [8] to time-weight average measured box leg data, compute the true and magnetic courses and vertical airspeeds as differentials of each sequential pair of GPS data records at relative times t+1 and t, and finally compute the glide ratios.

As an example of time-weighted averaging, two GPS data records having ALT = 7200 MSL for TP SEC = 20 seconds and ALT = 7250 MSL for TP SEC = 10 seconds, using

results in an average altitude of (7200 x 20 + 7250 x 10) / (20 + 10) = 7217 MSL for the combined 30 second interval.

For two nearby positions away from polar latitudes, true course and magnetic course very closely approximate

TC t+1 = arctan( ( LATITUDE t+1 – LATITUDE t ) / ( LONGITUDE t+1 – LONGITUDE t ) ) – 90o ,
MC t+1 = TC t+1 + VAR ,  using local VAR = –13.13o .

I used the hand-recorded data to identify the GPS data records for each leg of each box and verified the derived Montgomery Field (MYF) arrival and departure magnetic courses were 280o as expected.  The figures for the GPS VSI FPM rate of descent for glide boxes are altitude differentials for GPS data given by

GPS VSI FPM t+1 = ( ALTITUDE t+1 – ALTITUDE t ) / ( time t+1 – time t ) .

The figure for GPS ALT FPM gross rate of descent is similarly computed from the starting and ending GPS-recorded altitudes and times of straight glide.  The GPS unit apparently records the altitude at the end of the data record time snapshot, so the time point interval TP SEC of the starting altitude record is excluded from the time total for the GPS ALT FPM differential.

Measurement #1 – N6360D

N6360D is a 160HP C-172N with 50 gallon fuel tanks and a maximum GTW of 2300 lbs.  Approximate weight at the time of data collection was 2200 lbs.  The C-172N manual gives Vg = 65 KTS IAS for GTW 2300 lbs.  According to Kershner [7], Vg as a function of Vg0 at maximum GTW WM for an aircraft at weight W is given as

 ,

for which Vg = 64 KTS but I used 65 KTS.  The glide ratio determination method I used first measures GPS ground speeds in a level flight box along each compass quadrant holding a uniform IAS, then measures the GPS ground speed and descent rate of a glide to a specific waypoint (figure 2).

L08 AWOS reported 30.46 and 16C/-12C. After temporarily checking the altimeter reading at 29.92 and the OAT at 45F, the cockpit density-corrected TAS reading was about 115 KTS.  For the first step, we stayed at 9500 MSL and 100 KTS IAS as we measured the GPS ground speed along each of the four headings of 000, 270, 180 and 090. Only three such legs are needed but a fourth should reduce any systematic errors.

To measure the glide, I punched in L08 as my Direct-To waypoint and selected a steady heading – in this test case, 052o according to the panel gyro compass – that resulted in our flight path crossing L08 in the GPS unit display.  At panel altimeter 8500 MSL and IAS Vg = 65 KTS, we glided until the GPS read a distance of 0 NM to L08.  I noted a panel VSI rate of just over –650 FPM during the glide.  We then climbed back to 8500 MSL and repeated the glide from a different direction, gyro heading 325o.

Figure 2 – N6360D Straight Glide Measurement Flight Path

The time-weight averaged level flight box GPS data and derived values are compiled in Table 1.  Time-weighted averages are given in bold font for ALT, ALT FPM and VSI FPM using the summed total TP SEC time as the weight in underscored italic font, at the bases of their columns.

MH

ALT

GS KTS

GS
Label

VSI FPM

TP SEC

000

9458

120.1

g1

-2

46

090

9492

101.9

g4

38

130

180

9444

117.0

g2

0

40

270

9432

130.7

g3

-49

25

 

9471

 

 

15

241

Using Value

TAS
a

WS
w

WD

c0

c1

c2

d

g4

118

16

096

368

14057

-3673

13563

g3

118

13

097

368

14057

3026

13722

AVG

118

14

096

 

 

 

 

Table 1 – N6360D Level Flight Box Data

Using the g4 value to check for consistency should produce very similar results as does using g3, and here the computed values closely agree with each other and with the cockpit TAS estimate.  The wind was determined to be 096o at 14 KTS.

The glide magnetic course computed using the GPS data records differential method for the glide start and end points was MC = 45.3o and the ground speed using time-weighted averaging was GS = 56.0 KTS.  Using the cosine law [3] for the wind vector figure 1, the glide TAS is given by

 ,

 

evaluating to TAS = 65.9 KTS.  To firmly determine actual magnetic heading, the cosine law also requires

 

 ,

 

giving glide MH = 055o, very close to cockpit gyro MH = 052o.  Both glides are summarized in Table 2.

 

GLIDE FROM
8500 MSL

WIND
9500 MSL

GPS DIFFERENTIAL DATA

GLIDE

MH

TAS
KTS

GS KTS

WD

WS KTS
w

ALT
LOSS

VSI FPM

ALT FPM

AVG
FPM

TP SEC

1

055

66

56.0

096

14.3

-1752

-672

-678

-676

174

2

321

71

81.7

096

14.3

-1560

-579

-581

-580

177

Table 2 – N6360D Straight Glide Path Data

The final values for the glide ratio of N6360D at IAS Vg = 65 KTS are determined by dividing TAS by GPS FPM (averaged descent rates), and summarized in Conclusions Table 4.  I wondered how glide ratio would change in denser air at lower altitudes, or with less weight, or relative to the barometric altimeter. To further examine some of these questions, I needed more data from a second measurement.

Measurement #2 – N4975F

To better assess some of the conditions and questions from the N6360D measurement, we took data in another C-172N I regularly fly, one week later.  N4975F is a 180HP STC-modified C-172N with 40 gallon fuel tanks and a maximum GTW of 2550 lbs. Approximate weight at the time of data measurement was 2150 lbs.  The modified POH [5] gives Vg = 62 KTS for that weight, consistent with Vg formulation [7] of 68 KTS at maximum GTW. L08 AWOS reported 30.03 and 20C/-7C.

For this data I used a different method that directly measured GPS ground speeds along each compass quadrant heading in a squared spiral ‘glide box’.  To determine glide ratio variation with altitude and cross-check for errors, I also measured level flight GPS ground speeds, and repeated these measurements in three alternating layers of level-flight boxes and glide boxes at different altitudes from nearly 13000 MSL down to about 3500 MSL, as shown in figure 3.  I noted during the first glide box the panel VSI indicated about –640 FPM.

Figure 3 – N4975F Multi-Layered Glide Box Measurement Flight Path

For each leg of a glide box it’s important to stabilize the aircraft at the same IAS and new heading after each turn before recording the starting altitude and time. During the box leg, measure the GPS ground speed. At the end of the box leg, record the altitude and time again. Longer times in a leg result in more reliable data provided IAS and heading (and therefore TAS) are crisply maintained.

The time-weight averaged level flight box and glide box GPS measurement data and derived calculated values are compiled in Table 3. The winds determined for the level flight boxes provide a cross-check of the glide box winds, and were determined overall to be 217o +/– 20o and 20+/– 3 KTS at all altitudes measured, including glide boxes. Final values for the glide ratio of N4975F at various altitudes for IAS Vg = 62 KTS are determined by dividing TAS by FPM descent rate, and summarized in Conclusions Table 5.

MH

ALT

GS KTS

VSI FPM

TP SEC

000

12663

134.8

22

52

090

12704

125.7

-13

50

180

12746

95.2

0

36

270

12786

103.4

105

9

 

12705

 

10

147

Using Value

TAS
a

WS
w

WD

c0

c1

c2

d

g4

115

22

206

4552

13610

2180

12640