National Aeronautics and Space Administration

Astrometric and Photometric Analysis of theSeptember 2008

ATV-1 Re-Entry Event

M.K. Mulrooney (Engineering Science Contract Group(ESCG)/NASA; mark. mulrooney- 1 @nasa.gov )

E.S. Barker (ESCG/NASA; ebarker40@comcast.net )P.D. Maley (United Space Alliance(USA)/NASA; paul.d.maley@nasa.gov )

K.R. Beaulieu (ESCG/NASA; kevin.r.beaulieu@nasa.gov )C.L. Stokely (ESCG/NASA; christopher.stokely@pra-corp.com )

ATV Re-Entry WorkshopCouncil of European Space Societies (CEAS) Congress

Manchester, England28th & 29th October 2009

https://ntrs.nasa.gov/search.jsp?R=20090037547 2018-06-29T13:16:27+00:00Z

NASA utilized Image Intensified VideoCameras for ATV Data Acquisition

Platform:

Gulfstream V Aircraft

12.8 km altitude

Manual (Hand) Tracking

Instrument Pair:

75 and 12 mm lenses (8 and 20 deg FOV)

Gen 3 Micro Channel Plate Intensifier

S20 Photocathode (400-800 nm)

Optical coupling to COTS CCD Camera

Hi-8 (NTSC analog) recorder

Video Encoded Time via GPS

2

ATV Raw Video DataThe high resolution/narrow field (75mm) data was used for this analysis

(NTSC 480x720 pix; 8 deg Horizontal FOV; 0.7 arcminute/pix)

3

National Aeronautics and Space Administration

Data Extraction I

170 seconds of analog video were acquired

The video was digitized to .AVI, then analyzed with a modified commercial SW package:Image System’s Trackeye.

NASA sponsored SW modifications included a circular aperture photometry option andcoordinate transformation from Cartesian image coordinates to celestial (RA DEC)using field reference stars.

Approximately 300 fragments were visible in the full motion video

184 fragments were actually measureable in the video still frames. Only the narrow field(75mm;8 deg) data was analyzed due to excessive saturation in the wide field video.

Astrometric and photometric measurements were obtained for each fragment in eachvideo frame: totaling - 10 6 independent fragment measurements. Fragments weretracked and measured until they faded to near still frame background levels.

84 reference stars (a minimum of 4 in each video frame) were identified for astrometricand photometric calibration of the fragment data.

Astrometric accuracy is - 4 arcminute (6 digitized pixels) Relative photometric accuracy is - 0.5 astronomical magnitude (due to saturation and

the analog nature of the video). Absolute - 1 magnitude with a limit of 10 (V Band)

4

Data Extraction II

The video is divided into two segments – delimited roughly by the ATV tank explosion

The first segment is 76 seconds in length and contains 71 fragments and 43 referencestars

The second segment is 94 seconds in length and contains 113 fragments and 41reference stars

In each video segment we derived time-dependent fragment angular trajectories,velocities, accelerations, and luminosities.

Hans C.S. Nielsen (University of Alaska) kindly provided the coordinate transformationto Inertial Geocentric XYZ from our RA-DEC-Time measurements.

With Nielsen’s transformation we derived time-dependent fragment spatial trajectories,velocities, accelerations.

Ballistic Coefficients (Beta) were derived for each time step based on fragment velocity,acceleration, and altitude dependent atmospheric density.

5

National Aeronautics and Space Administration

ATV Tracked Video Data – Extracted Still Frame In.`,^'Il

(Trackeye Output w/ Markers for Fragments and Stars)

Raw Tracking Data0.75 sec Sequence; Vertical Traverse

(Two reference stars & Three debris fragments)Erratic target motion is removed using fiduciary reference stars

Time (msec)

7

22.5 22 21.520

Coordinate transformation: Cartesian Image coordinates to Celestial RADECReference Stars (vertical streaks) and Fragments (arcs)

RA DEC vs Absolute Time (T plus13:35:34 GMT)(71 fragments – first video segment)

8

7

6

5

t3'NNE 4

3

2

1

-50

-4

Declination (deg)Right Ascension (hours)

8

148 -42

-35

National Aeronautics and Space Administration

Coordinate transformation:Celestial RADEC to Topocentric Lat Long plus Altitude

(Plane View; All data) – Curvature due to Earth Rotation; Separation due to varying Beta values

Longitude (deg) Latitude (deg)

9

Coordinate transformation:Celestial RADEC to Topocentric Lat Long plus Altitude

(Orthogonal View; All data) –Separation due to Explosion and varying Beta

85

80

75

70

65Y_

N

7

Q 60

55

50

45

40

160 158 156 154 152 150 148

L

Longitude (deg)

10

-35

20 second Data Gapwhile fuel tank cloudwas tracked

E 65

Y_N7

Q 60

45

70

55

50

ATV Reentry - Altitude v Time

85

Fuel Tank Explosion

Explosion dV80

75

400 20 40 60 80 100 120 140 160 180

Absolute Time (sec)

11

50

45

40

35

30rna^

25a^JLQ 20

15

10

5

Arc Length (deg) vs Absolute Time (T plus13:35:34 GMT)(Segment 1; 71 fragments; Arc Length along Great Circle)

Beta - slope; A/m - 1/slope

00 10 20 30 40 50 60 70

Absolute Time (sec)

12

50

45

40

35

30rna^

25a^JLQ 20

15

10

5

Arc Length (deg) vs Absolute Time (T plus13:35:34 GMT)Clustering events (fragment clumps) indicated (Segment 1; 71 frags)

00 10 20 30 40 50 60 70

Absolute Time (sec)

13

50

45

40

35

30rna^vLS 25a^JUQ

20

15

10

5

Arc Length vs Relative Time(71 fragments; Beta - slope; A/m-1/Slope)

00 5 10 15 20 25 30 35 40 45 50

Relative Time (sec)

14

400

EY_

Lm 300a^J

Q

200

ational Aeronautics and Space Administration

Arc Length (km) vs Absolute Time (T plus13:35:46 GMT)Clustering events (fragment clumps) indicated; Differential Motion is readily evident

(Segments 1 &2; All 184 frags)

600

500

15

1.6

1.4

1.2

^ 1a^

rna^

0.8ON

L2Q 0.6

0.4

0.2

Arc Velocity vs Absolute Time (T plus13:35:34 GMT)71 fragments; A/m~1/V

00 10 20 30 40 50 60 70

Absolute Time (sec)

16

Fuel Tank Explosion12

UN2EY

10

^UON

8

14

6

ATV Reentry - Velocity v Absolute Time(T plus13:35:46 GMT)

20

18

16

4

2

00 20 40 60 80 100 120 140 160

Absolute Time (sec)

17

Circular Aperture Photometry via Trackeye

18

Brightness vs Absolute Time (T plus13:35:34 GMT)Subset of 39 (unsaturated) fragments

(Unsaturated) fragments commonly exhibit hookappearance due to initial brightening then subsequent

6.5

fading

20

30 40

50Absolute Time (sec)

6

Z5.5

rncc2

U

0 5C0

Q

4.5

4

3.5 010 60

70

19

ATV Reentry – Astronomical (V) Mag v TimeSegment 2 Only – Same Luminosity Hook as Segment 1

Separation => Enhanced Frictional Heating => Dissipation

10

9

8

7

3

6a

Zrnm

m 5UEO

Q4

2

080

100

120

140

160

180Time (seconds)

20

•

•

Conclusions I

Camera and platform motion were well compensated via our analysis softwareâ Astrometric results were limited by saturation, plate scale, and imposed linear plate solution

based on field reference stars

Analog video (8-bit dynamic range, read noise, AGC) limited the accuracy of thephotometry

â 0.5 astronomical magnitude for the subset of unsaturated fragments; -1 mag absolute

It is readily evident that individual fragments behave differentlyâ Differential velocity a dominant featureâ There are derived trajectory families with a common spatial origin, this is clear evidence of

fragment clustering (multiple fragments emanating from a common parent), unfortunately videosaturation prevented extension of fragment tracks back to a common source

â Fragment linear velocities range from 4 to -20 km/secâ As time progresses fragments trend toward linear velocity. As they lose altitude the exchange

of potential for kinetic E dominates drag lossesâ Almost all unsaturated fragments exhibit hook appearance due to initial brightening then

subsequent fading

I

21

Conclusions II

• The fuel tank explosion at T=13:36:10 significantly affected subsequent fragmentbehavior:

â While the camera was trained on the dissipating fuel cloud, the fragments immediately behind the ATV parentbody were not tracked for ~20 seconds after the explosion, hence the gap in fragment trajectory figures.

â ~25 seconds after the explosion (~5 seconds after the fragments were reacquired with NASA’s camera) a +-5km/s velocity dispersion was evident in the fragment field (versus +-1 km/sec prior)

â ~25 seconds after the explosion (~5 seconds after the fragments were reacquired with NASA’s camera) a +- 10km altitude dispersion was observed in the fragment distribution

â A large time window emerged as fragments descended – those with fastest descent passed through 55 km ~50seconds earlier than the slowest

IUnfortunately photometric accuracy was insufficient to confidently assesscorrelations between luminosity and fragment spatial behavior (velocity,deceleration). Use of high resolution digital video cameras in future should remedythis shortcoming.

Via ATV-1, we have developed a comprehensive pipeline enabling us to conductfuture re-entry event analysis in a more timely and efficient manner.

22