Program 437AP: A Sub-orbital Corona-Derived Satellite Inspector

Ted Molczan - 2016 January 6

1. Introduction
2. Origin within Program 437
3. Satellite interceptor vehicle
    3.1 Configuration and size of payload section
    3.2 Film-return re-entry vehicle
    3.3 Resolution of imagery
4. Program 437AP Flight 1
    4.1 Launch circumstances
    4.2 Orbital elements of target
    4.3 GMAT estimate of interceptor trajectory
    4.4 Mission results
5. Program conclusion
6. Acknowledgments

1. Introduction

The Program 437AP satellite inspector was an alternate payload of the mid-1960's Program 437 ASAT (anti-satellite) weapon system. The ASAT was of the direct-ascent type, which employed a Thor IRBM, that carried a nuclear warhead within its blunt-nosed G.E. (General Electric) Mark II re-entry vehicle. It was designed to fly a sub-orbital trajectory to intercept and destroy Soviet satellites that passed within range of its launch site on Johnston Island in the Pacific Ocean. The satellite inspector was conceptually similar, but only took photos of its target. The existence of the ASAT was made public by President Johnson in 1964.

The existence of the satellite inspector appears to have remained secret until 1991, when USAF historian, Dr. Wayne R. Austerman wrote Program 437: The Air Force's First Antisatellite System. It was published by the Air Force Space Command Office of History in May 1991, after FOIA review and declassification. It is a thorough history of the program, with the expected emphasis on politics, policy and personnel, but with much useful technical information. The present report is primarily on the technical aspects of the satellite inspector vehicle and its test flights, building on the information reported by Austerman.

Austerman described the main components of the Program 437AP satellite inspector vehicle, and provided a clue regarding its overall configuration. Its Corona imagery intelligence satellite heritage became known after that program was declassified in 1997. Consideration of Austerman's information in light of that fact, leads to the realization that the Program 437AP satellite inspector vehicle must have been close in size and configuration to a Corona satellite, without the Agena. This suggested something more than a minor modification of the ASAT's G.E. Mark II RV.

Strong supporting evidence is found in an excellent photo of a Program 437AP launch, provided by Jonathan McDowell, which together with Austerman's information and what is known about Corona, leads to the conclusion that the satellite inspector vehicle was a sub-orbital derivative, similar in size and configuration to Corona-M (aka Mural and KH-4), without the Agena. Elements of the ASAT's G.E. Mark II RV may have been incorporated, but only with considerable modification, and accounting for a fraction of the volume of the new vehicle.

Although 437AP was derived from Corona, they are somewhat distant cousins. Lockheed was the prime contractor on Corona, but was not among the contractors of 437AP listed by Austerman, who reported that G.E. received by far the largest share of program spending.

Section 2 describes the origin of Program 437AP. Section 3 describes the satellite interceptor vehicle. Section 4 presents a detailed analysis of Flight 1. Section 5 summarizes the final three flights and the termination of the program. Section 6 acknowledges those who assisted in the making of this report.

2. Origin within Program 437

In February 1962, the U.S. Air Force Systems Command (AFSC) initiated research and development of an ASAT (anti-satellite) system. This resulted in a preliminary development program for a nuclear-armed direct-ascent ASAT, to be launched by the Thor IRBM, from Johnston Island in the Pacific Ocean. It was designated Program 437 in February 1963, and funding was approved for four launches to demonstrate feasibility in early 1964. (Austerman, pp.12-13)

Colonel Quentin Riepe, the System Program Director, described a crash program that relied heavily on existing hardware:

There is no new technology in the 437 program. Actually it is drawn from equipments developed on other research and development and operational programs, and from the junkpile residue left over from them. The Thor, the guidance system, the warhead and its fusing, were all drawn from other developed systems or from systems in research and development test. ... Our biggest problems were associated with . . . the Secretary of Defense directive that the system would be operational in June of 1964. We found ourselves implementing plans for the operational phase before funding had been received on the R&D program. The first man was put into training before the R&D engineering had been started. (Austerman, p.23)

The Program 437 ASAT could only target satellites that passed within range of the Thor IRBM. A standard Thor IRBM carried a nuclear weapon within a G.E. Mark II re-entry vehicle, which it could deliver to a terrestrial target over a nominal range of 1,500 NM (2,800 km). The maximum range within which the Program 437 ASAT could engage an orbital target was about half this (between 600 km and 1312 km in test flights, Austerman, p.95). Its nuclear warhead had a kill radius of 3 NM (5.6 km) (Austerman, p.46).

The G.E. Mark II re-entry vehicle protected its nuclear warhead from the heat of re-entry with a heavy, heat-sink type heat shield, made principally of copper. The precise mass of the heat shield is uncertain; typical reported values range between 1,000 lbs and 1,200 lbs. Since the ASAT was intended to detonate its warhead at orbital altitude, it did not require a re-entry heat shield. Replacing it with a nose cone sufficient to protect against ascent-heating, would have enabled a much greater range; however, judging by the actual range of the interceptions, the G.E. Mark II was not significantly modified for the ASAT. This could have been due to the crash-program nature of the project. Also, missile "accuracy was degraded at extreme range due to the ballistic trajectory and increased time of flight" (Austerman, 55), which may have limited the practical advantage of eliminating the copper heat shield.

The ASAT interception was accomplished by aiming the sub-orbital Thor IRBM at the pre-computed co-ordinates of a point in space, through which the orbiting target was predicted to pass at a certain time. A major technical and logistical challenge was to provide sufficiently precise orbital data on the target, 12 hours before ASAT launch, accurate to 3 NM (5.6 km) along-track, and 0.5 NM (0.9 km) cross-track. It was essential that the ASAT arrive on target and on time. The launch window was a mere +/- 1 s. (Austerman, p.23)

Four demonstration launches were conducted during Feb-May 1964. The first three were judged successful. Initial operational capability was declared the day after the failed fourth launch. (Austerman, p.28)

In May 1963, Program 437 was directed to study the practicality of performing satellite inspection. In early 1964, work began on a satellite-inspector variant. It was initially called Program 437X, but by the fall of 1964, it was designated 437AP, for alternate payload. (Austerman, p.53) It was conceptually similar to the ASAT, with the inspector vehicle replacing the warhead. It presented the greater technical challenge, because of the need to design and build a new sub-orbital vehicle. Fortunately, all of the essential payload components had already been designed, built and flight-proven. General Electric received a contract "for a study and analysis of the program," and received by far the greatest share of program spending (Austerman, pp.53-54).

3. Satellite inspector vehicle

3.1 Configuration and size of payload section

Most of the publicly available information on the Program 437AP satellite inspector vehicle was reported by Dr. Wayne Austerman in 1991:

Its basic aim was to demonstrate the "feasibility of obtaining technical intelligence photos of orbiting objects using a non-orbital interception system." The 437AP payload was to be completely interchangeable with the Thor booster, and was to include an aft bulkhead, a spacer section, a modified General Electric Mark II nonrecoverable re-entry vehicle, and an H-30A recovery re-entry vehicle carrying twin cameras in "dual rotary" configuration, an allied control subsystem, and telemetering equipment. (Austerman, p.54)

The Thor booster's expendable payload was configured something like an oil drum with a bullet-shaped superstructure added to one end. The payload vehicle contained an index camera and a primary panoramic camera housed within its protective carapace, with a clamshell door that opened to permit photography of the target vehicle once interception had been made. The exposed film was taken up on cassettes housed in the recovery capsule, which was ejected from the main payload vehicle during re-entry into the atmosphere. The ejected capsule then deployed a parachute at a pre-determined altitude and was recovered in mid-air by a specially configured C-130 aircraft operating from Hickam AFB, Hawaii (Austerman, p.55)

The description closely matches that of the NRO's (National Reconnaissance Office) Corona satellites. The designation of the re-entry vehicle as H-30A, is similar to the NRO's H-30 designation for the G.E. SRV (satellite re-entry vehicle) used on Corona, revealed in
Robert L. Perry's history of Gambit, and in the minutes of a Sep 1962 meeting that informed his writing. The H-30A probably was the G.E. Mark V SRV, which was the current model at the time the 437AP vehicle was designed, used singly on Corona-M and in pairs on Corona-J (Volume IV of Perry's Corona Program History, p.1-10).

The reference to "twin cameras in 'dual rotary' configuration," suggests Corona-M (aka Mural and KH-4), which was the first version to employ a pair of cameras. They were the Itek panoramic C''' (triple prime) cameras introduced with the single-camera KH-3. Mural produced stereo imagery using a pair of C''' cameras, one aimed fore, the other aft of the ground track. The first of 26 Corona-M was launched in February 1962, the last in December 1963. The lenses of all Corona cameras rotated, to provide image motion compensation (IMC). The reference to an "index camera," also suggests a similarity to Corona-M, which was the first model to employ it. (For information on the cameras, see The Corona Camera System, by Frank J. Madden.)

Figure 1 is a scale drawing of the Corona-M payload. Its base was 50 inches (1.27 m) in diameter, and it was 8.5 ft. (2.59 m) tall. It was permanently mated to the 60 inch (1.524 m) diameter Agena B or D second stage of its launch vehicle, via a conical payload adaptor, neither of which are shown in the drawing. It depicts all of the components described above. The index camera is labelled as a vertical framing camera.

[Corona-M line drawing]
Fig. 1.  Corona-M. Source: NRO.

If the Program 437AP satellite inspector vehicle was derived from or similar to Corona-M, then it could be expected to have been similar in configuration and size. Until now, the only publicly available description was Austerman's "oil drum" with "bullet-shaped superstructure." I sought the advice of Jonathan McDowell, who provided me with the scan of a Program 437AP launch photo, shown in Figure 2 (below), which clearly matches Austerman's description.

[Program 437AP launch. Courtesy J. McDowell.]
Fig. 2.  Program 437AP launch. Courtesy J. McDowell.

Figure 3 (below) compares the Program 437AP Thor IRBM with a standard one carrying a G.E. Mark II re-entry vehicle, to which the Program 437 ASAT probably was identical. Both depict the entire missile forebody and payload, and are shown at the same scale. The umbilicals of both appear to connect to the missile at about the same elevation. Several additional umbilicals connect to the 437AP payload.

[Program 437AP compared with standard Thor IRBM.]
Fig. 3.  Program 437AP (left) compared with standard Thor IRBM.

Figure 4 (below) superimposes the outline of the payload section of Corona-M over the visible portion of the Program 437AP payload vehicle. In the left hand image, their re-entry vehicles are aligned, at the top of the stack. Clearly, the RVs match, as expected. In the right hand image, they are aligned at the elevation of the interface between the missile forebody and the 437AP payload. Their cylindrical camera sections are about the same height. The diameter of the 437AP was greater, 60 inches (1.524 m) vs. 50 inches (1.27 m). It was about 13 inches (0.33 m) shorter than Corona-M.

[Outline of Corona-M superimposed over Program 437AP vehicle.] [Corona-M overlayed Program 437AP.]
Fig. 4.  Outline of Corona-M superimposed over Program 437AP vehicle.

Table 1 compares the volume of the payload sections of Corona-M and 437AP, below the base of the H-30A RV's thrust cone. Corona-M was about 67.2 ft3. The 437AP was about 78.6 ft3, assuming there was usable volume beneath the half-foot tall black band at the top of the cylindrical section; if not, then its volume would have been about 68.8 ft3. With or without the black-banded section, the 437AP had roughly the same payload volume as Corona-M; therefore, it seems likely that it could have accommodated the major components of its imaging system.

Table 1:  Comparison of payload volume
Payload Module Section Corona-M Pgm 437AP
Cylinder diameter, ft 4.17 5.00
height, ft 3.50 3.80
volume, ft3 47.72 74.61
Conical fairing below RV thrust cone bottom dia., ft 4.17 5.00
top dia., ft 3.34 3.00
height, ft 1.75 0.31
volume, ft3 19.44 3.98
Vehicle total volume, ft3 67.16 78.59

The payload section of Corona-M was integrated with the Agena stage of its launch vehicle, which provided it with electrical power, attitude control, and communications (telemetry and command and control). The 437AP had no Agena, because it was launched on a sub-orbital Thor IRBM. It seems unlikely that those systems would have fit within its payload section. If not, then a clue to their likely location may be found in Austerman's report that the 437AP included an "aft bulkhead" and a "modified General Electric Mark II nonrecoverable re-entry vehicle." Figure 5 (below) compares the 437AP with a drawing of a standard Thor IRBM. The G.E. Mark II re-entry vehicle consisted of a blunt nose cone, below which a conical payload compartment extended into the missile forebody, occupying a portion of the space above the airborne guidance system.

[Program 437AP compared with standard Thor IRBM.]
Fig. 5.  Program 437AP (left) compared with standard Thor IRBM.

The payload compartment of the G.E. Mark II RV appears to have been the only space other than the 437AP's payload section, that could have housed the electrical power, attitude control, and communications systems. Figure 6 (below) superimposes its outline onto the 437AP launch photo. Could this have been what Austerman meant by the 437AP's "aft bulkhead?"

[Program 437AP with outline of payload compartment of G.E. Mk II RV.]
Fig. 6.  Program 437AP with outline of payload compartment of G.E. Mk II RV.

The empty volume of the payload compartment of the G.E. Mark II RV was about 18 ft3. It's dimensions appear to have been limited by a conical bulkhead of similar size, at the top of a Program 437 Thor IRBM forebody, shown in Figure 7 (below). Employing a modified G.E. Mark II RV payload compartment on the 437AP vehicle, seems consistent with the "complete interchangeability" of payloads reported by Austerman. Whether it was actually present on the 437AP, and its purpose, remains to be determined.

[Thor IRBM forebody without payload.] [Thor IRBM forebody without payload.]
Fig. 7.  Program 437 Thor IRBM forebody without payload, on display at Vandenberg AFB. Source J. Page. Originals here and here.

Another part of the G.E. Mark II RV that may have been been used on the 437AP, was its nosecone. In Figure 3, the conical fairing between the 437AP's cylindrical section and its H-30A RV, bears a strong superficial resemblance to this nosecone. The rounded tip of the G.E. Mark II RV nosecone subtended an included angle of 105 deg, resulting in a slope of 37.5 deg relative the top of the missile forebody. Measurements of the 437AP launch photo yield about the same angle. If this was part of a G.E. Mark II RV nosecone, then about 40 percent of the surface area of its heat shield must have been removed to provide an opening for the H-30A RV. It makes more intuitive sense for a new fairing to have been designed and built from scratch, which may well have been done. There is no way to be certain.

3.2 Film-return re-entry vehicle

The H-30A RV probably required minimal modification for Program 437AP. Since 437AP launches were sub-orbital, its retro-rocket would have served no useful purpose; therefore, it almost certainly was removed, perhaps replaced by an inert replica.

The H-30A would experience its maximum rate of deceleration at a much lower altitude than its orbital counterpart, resulting in a far greater maximum rate. Flight 1 is estimated to have peaked at about 29 g (Section 4.4), compared with 7 g to 8 g for an orbital re-entry. Modification may not have been required, given that the April 1965 system performance/design requirements of Corona-J, specified (on page 88) that the cassette assembly with film of its SRV, "shall be capable of withstanding re-entry accelerations up to 35 g's longitudinally, and 5 g's laterally."

The total heating during a sub-orbital re-entry is far less than that of an orbital one, due to its much lower velocity, but the maximum rate of heating is greater. That might have required the H-30A's ablative heat shield to be modified; however, if it was designed for maximum re-entry acceleration of 35 g, then it probably was adequate, since peak heating and maximum acceleration occur at about the same time.

3.3 Resolution of imagery

There is no public information on the expected or achieved resolution of the Program 437AP satellite inspector. An indication of what may have been possible is found in the memorandum of October 1, 1965, from the Chief of the NRO's Systems Analysis Staff, in apparent response to a query by William A. Tidwell, Chairman of the Committee on Overhead Intelligence, regarding the use of KH-4 to photograph orbiting satellites.

Of the several possible circumstances of encounter evaluated, the estimated best resolution was 2 ft., from a vantage point 10 NM (19 km) above a near-coplanar target. This was without image motion compensation. Resolution from encounters at a 90 deg cross-track would have been 25 ft., apparently limited by the much greater, uncontrolled, image motion. The summary gave an indication of the technical challenges involved:

In summary, it might be estimated that if a launch window of about five minutes could be obtained, pictures of a Soviet on-orbit satellite could be obtained from a KH-4 to a resolution of a few feet. It should be added that every time the KH-4 camera is turned on, an appreciable fraction of its film is expended. An orbiting satellite might be photographed when ground intelligence was collectable also. Finally, the limited on-orbit camera control flexibility of KH-4 might result in several, if not many, attempts being necessary before success. Ultimately the purpose and value of the photograph should be considered. As a trick, it may be great; the substantive intelligence value may be considerably less. It must be obvious that this cursory examination is insufficient. A detailed plan would be needed, both refining these rough calculations and including eccentricity, before the feasibility, chance of success, etc. can be firmly established.
The 5 min. launch window discussed was not nearly as challenging as the +/- 1 s available to Program 437AP. The analyst also noted the possible need for a KH-4 to make multiple attempts before it could obtain imagery, due to limited camera control flexibility. The sub-orbital Program 437AP satellite inspector would have only one, brief imaging opportunity, lasting perhaps 10 s. How it would have acquired its target, and whether and how IMC was provided is unknown. That it was an issue is known, because the maximum crossing-angle at interception was set at 45 degrees, due to relative velocity constraints (Austerman, p.55). Also to be determined is how the second panoramic camera, and the index camera were intended to be used.

The following section provides a detailed account of a Program 437AP mission.

4. Program 437AP Flight 1

Four Program 437AP satellite interceptor vehicles were launched between December 1965 and July 1966. The first flight ended in failure, but it is the most readily studied of the four, because it received by far the greatest attention from Austerman. He provided sufficient detail to accurately reconstruct the circumstances of the launch, the orbital trajectory of the target, and the sub-orbital trajectory of the 437AP vehicle. The results of those analyses are used to provide additional context for the discussion of the mission results.

4.1 Launch circumstances

Austerman provided the following facts about the first flight (p.54, p.57 and p.95), that help to constrain the trajectory of the interceptor.

Table 2:  Program 437AP Flight 1
Lift-off time 1965 Dec 07 19:29 MST = 1965 Dec 08 02:29 UTC
Launch site Johnston Atoll, Pad II
Target Agena with SPADATS Object Number 613 (1963-027A)
Time to interception 8.18 min. (491 s) after launch, after 328 s ballistic flight
Closest approach to target within 0.56 NM (1 km) of planned 3.2 NM (6 km)
Azimuth at interception 153 deg
Range at interception 1528 km slant, 1398 km ground
Altitude at interception 487 km

The above lift-off date and time disagree by exactly one day with those tabulated in Appendix III of Austerman's report. The above information was determined to be correct, as a result of the analysis discussed in Section 4.2.

Pad II was formally called Launch emplacement 2 (LE2). It and LE1 were used for all Program 437 launches. The following co-ordinates were determined with the aid of the site plan and Google Earth:

Table 3:  Program 437 Launch Pads
Latitude Longitude Altitude
Launch Pad deg N deg E m
Launch emplacement 1 (LE1) 16.72956 -169.53790 -14
Launch emplacement 2 (LE2) 16.73210 -169.53437 -14

4.2 Orbital elements of target

The precise time of the interception was determined using the orbital elements of the Agena 1963-027A / 613, which are available in TLE (2-line elements) form in Jonathan McDowell's public archive. The two TLEs that bracket the launch time, were of epoch 3.8 d earlier and 6.0 d later. They agreed to within 1 s on the time of interception. The closer in epoch of the pair was selected for the trajectory analysis:

1 00613U 63027A   65338.26395386  .00001480  27881-8  64242-4 0   591
2 00613  82.3360 136.6698 0028279 146.4922 213.7506 15.21993418135027
The orbital trajectory of the Agena target was readily calculated from the TLE, using the underlying SGP4 (Simplified General Perturbations) model. It matches the tabulated range, altitude and azimuth of the interception on 1965 Dec 08 at 02:36:52 UTC (at which time it was 487.718 km above 5.395 N, 163.839 W); therefore, that is the probable time of the event, to within 1 s. Table 1 reports that the interception occurred 8.18 minutes (491 s) after launch. Subtracting this yields lift-off at 02:28:41 UTC, which agrees with that of Table 1 to the nearest minute.

The difference between the mission elapsed time of interception (491 s) and the duration of ballistic flight (328 s) was 163 s, just 1 s longer than the nominal time of VECO (vernier engine cut-off) for a 1500 NM (2800 km) Thor IRBM launch. Adding this difference to the time of lift-off yields the start of ballistic flight, at 02:31:24 UTC.

The procedure used to estimate the sub-orbital trajectory of the interceptor is described in the following sub-section.

4.3 GMAT estimate of interceptor trajectory

A sub-orbital interceptor trajectory that closely agrees with the known facts of the mission has been estimated using GMAT R2014a (General Mission Analysis Tool), "developed by a team of NASA, private industry, public, and private contributors."

The analysis was performed using GMAT's Dormand-Prince 78 numerical integrator, with a 90 degree, 90 order gravity field, and the MSISE90 atmosphere model.

The starting point of the propagation was VECO, determined earlier to have occurred at 02:31:24 UTC. The precise range from pad LE2, and altitude at VECO are not available; however, the nominal values for a 1500 NM (2800 km) Thor IRBM launch were approximately 154 km down range, and 126 km altitude. The position vector was estimated on the assumption that the launch azimuth would have been about the same as the 153 deg at intercept, which would have placed the vehicle 126 km above 15.49603 N, 168.88412 W.

No information is available on the ballistic properties of the Program 437AP payload. Since the trajectory of greatest interest occurred well above the dense layers of the atmosphere, nominal values of Cd (co-efficient of drag) and A/m (area to mass ratio) were used.

The initial estimate of the velocity vector was based on the nominal 4.42 km/s VECO velocity of a standard Thor IRBM. Rough Cartesian components were estimated on the assumption that they would parallel the vector from the launch site to the point of VECO. These estimates were refined through a process of trial and error, involving numerous GMAT runs, until a trajectory was obtained that closely matched the time and position of the interception reported earlier. The agreement is to within 2 s of time, and a few kilometres miss-distance. The geocentric velocity at VECO was found to have been 4.88 km/s. That is about 0.46 km/s greater than the standard IRBM. It is speculated that the increase may have been due to a net decrease in payload mass. The theoretical ground range to impact of 2,991 km is not much greater than typical for Thor, and within its known range of performance. The resulting GMAT script is available here.

A subset of the GMAT propagated state vectors at almost exactly 5 s intervals was extracted and entered into an Excel spreadsheet designed to generate ephemerides, including geodetic ground track co-ordinates and altitude above geoid, and topocentric co-ordinates from user-specified locations. The Excel file contains ephemeris sheets for the interceptor and the target. Columns 1-13 contain the computed ephemerides. Changing the observer's co-ordinates near the top of col 15 causes the topocentric portion of the ephemeris to be re-computed. Columns 20-26 contain the GMAT-computed state vectors. Columns 27-29 contain derived velocity and acceleration information. Columns 31-34 are used to format the trajectory for use in kml files. VBA macro, kml_placemark, reads this data to generate a file of place marks, ready to be inserted into a kml file. (The macro includes a number of hard-wired values that need to be edited for each case, e.g. input data sheet, path to output file, and object description.)

Users of Google Earth, may wish to download the 3D trajectory kml files of the interceptor and the target, which were used to generate the following graphics.

Figure 8 (below) depicts the trajectory of the target Agena (blue and white line) and the interceptor (red and white line). The trajectory of the interceptor begins with VECO, about 154 km downrange of Johnston Island. Much like the sport of skeet shooting, the interceptor was aimed at a point far down range, where it would meet its orbital target. The interceptor's average velocity was about half that of the target, so it was given a head start of more than 1300 km. As the interceptor neared its apogee of 490 km, the target caught up and overtook it, resulting in a close encounter that lasted several seconds.

After the interception, the orbiting target continued to move downrange, but the sub-orbital satellite inspector could only fall back to Earth, re-entering the atmosphere just 6.5 minutes later.

[Approximate re-entry track of 2008-010B]
Fig. 8.  East-facing profile of interception.

Figure 9 (below) depicts the trajectory of the target Agena (blue and white line) and the interceptor (red and white line) as seen looking north. The objects are moving toward the viewer. The crossing angle at interception is evident in this view. It was 23.9 deg, well within the maximum allowed 45 deg for acceptable relative velocity. After the interception, the satellite inspector re-entered. Any fragments that survived the heat of re-entry fell into the Pacific Ocean. The target remained in orbit.

[Approximate re-entry track of 2008-010B]
Fig. 9.  North-facing view of interception.

Figure 10 (below) depicts an overhead view of the trajectory of the target Agena (blue and white line) and the interceptor (red and white line). The alternating coloured line segments of both trajectories span 5 s of flight. Those of the target are noticeably longer, due to its much greater velocity.

[Approximate re-entry track of 2008-010B]
Fig. 10.  Overhead view of interception.

Figure 11 (below) depicts the final six minutes of the satellite inspector's descent toward impact. The near vertical final drop of about 20 km is due to the loss of all horizontal velocity due to atmospheric drag. This vertical descent occurs with meteors and sub-orbital, as well as orbital vehicles. The heat of re-entry would have subsided, so that the object would no longer have been incandescent. In the case of meteorites (meteors that reach the Earth), this final, non-luminous descent is called dark flight.

[Approximate re-entry track of 2008-010B]
Fig. 11.  East-facing view of final descent of interceptor.

4.4 Mission results

Most of the public information about the mission is due to Austerman (pp.57-58). Table 4 summarizes the key events, augmented by information from the present trajectory analysis.

Table 4:  Program 437AP Flight 1 Mission Events
Time Range Altitude
UTC km km Event
02:28:44 0 0 Lift-off
02:31:24 154 126 VECO - start of ballistic flight
02:36:50 1397 488 Closest approach to Agena target
02:40:00 2098 406 Radio contact with H-30A RV expected by Surface Recovery Unit
02:40:20 2173 384 Passed below horizon of Johnston Island
02:43:00 2801 126 H-30A RV last heard by Surface Recovery Unit
02:43:25 2905 71 Onset of significant re-entry heating
02:43:45 2975 31 Peak deceleration (approx. 29 g); peak heating
02:45:00 2992 14 "Contrails and smoke" spotted by surface and aerial recovery units
02:48:45 2991 0 Theoretical impact at 7.49 S, 157.71 W (without parachute)

Austerman wrote that the mission had been nominal through the interception of the Agena, which occurred shortly before 02:37 UTC. The satellite inspector missed its programmed stand-off distance from the target of 3.2 NM, by 0.56 NM. That would put the closest approach between about 5 and 7 km. Soon after the interception, the film should have been spooled into the H-30A RV (re-entry vehicle), cut, and the film door sealed; however, this failed to occur, and mission control at Johnston Island received no indication that the RV had separated.

The Surface Recovery Unit, located near the planned impact point nearly 3,000 km SE of Johnston Island, was expected to pick up the telemetry of the H-30A RV at 02:40 UTC, shortly before it passed below the horizon of Johnston Island. Apparently, the signal was received, but was lost at about 02:43 UTC, by which time the RV would have descended to 126 km. About 25 s later, it was down to 71 km, and beginning to experience significant re-entry heating. After another 15 s, it was down to 31 km, where it reached the maximum rate of deceleration, estimated by the trajectory analysis at about 29 g. This was also the point of peak heating, where anything not protected by a heat shield would break up and disintegrate.

At 02:45 UTC, the surface and aerial recovery units spotted "contrails and smoke," which presumably were the dust left by the disintegrated booster stage and the non-recoverable section of the payload, with its cameras. Since this was a daytime re-entry, the trails would have been readily visible, several tens of kilometres overhead. Had the H-30A RV survived, then it would have been down to about 14 km, probably on its main parachute (details of the planned recovery sequence are unknown), but it was never recovered. It initially appeared that the recovery parachute had failed to deploy properly and the water impact destroyed the capsule, but a subsequent data analysis by Space Systems Division found the root cause:

...the malfunction was caused by a momentary short-circuit in the in-flight disconnect cable between the payload and recovery vehicles. After squib firing to provide inflight electrical disconnection, the recovery vehicle cable terminal flew back against its own harness, causing cable damage. The momentary short circuit caused the recovery vehicle programmer to reset, which precluded physical separation of the payload and recovery vehicles. (Austerman, pp.57-58)
Unable to separate from the payload vehicle, the H-30A RV was doomed to burn up with it during re-entry. Depending upon its orientation, its heat shield might have prevented complete disintegration, in which case surviving fragments would have fallen into the Pacific Ocean, in the vicinity of 7.49 S, 157.71 W.

5. Program conclusion

Corrective actions were quickly taken to prevent a repeat of the failure of Flight 1. Flight 2 followed on 1966 Jan 18 UTC, reportedly successfully targeting an orbiting Agena. Film was recovered, but there is no information on image quality. (Austerman, p.58)

Flight 3 was launched on 1966 Mar 12 UTC. All "research and development feasibility demonstration objectives," reportedly were met, but there is no information on image quality. (Austerman, p.58)

Flight 4 was conducted at the request of NASA, in an effort to learn why its OAO-1 (Orbiting Astronomical Observatory) had failed soon after it reached orbit. The scheduled launch occurred on 1966 Jul 02 UTC, but an electrical short circuit sent the satellite inspector off-course, preventing its camera from acquiring the target, resulting in a photo of "the void of space." (Austerman, p.58)

Despite the mixed record of success, there was support for Program 437AP. In March 1966, the Commander in Chief of the Continental Air Defense Command requested a launch in April against a Soviet satellite. This was rejected by the Joint Chiefs of Staff and the USIB (United States Intelligence Board). The USIB recommended against launches from Johnston Island, because President Johnson had publicly identified it as an ASAT base, which somehow would have caused the Soviets to realize that one of its satellites had been targeted by a 437AP launch. The USAF terminated the program on November 30, 1966. (Austerman, pp.59-60)

Much remains to be learned about the design and operation of the Program 437AP satellite inspector vehicle. In Section 3, the attempt to reconcile Austerman's description of it with the launch photo and the known features of Corona, is somewhat speculative, especially regarding the possible incorporation of modified parts of the G.E. Mark II re-entry vehicle. The details of how the imaging was to be accomplished are unknown. How was the target to be acquired by the sensors, i.e. the cameras? How were the three cameras intended to be used? Did the vehicle perform as expected? How good were the images obtained on Flights 2 and 3? This report concentrated on the 437AP vehicle, but on the political and policy front, any role the NRO may have had in this unique program is of great interest.

6. Acknowledgments

I wish to acknowledge the assistance of Vicente-Juan Ballester Olmos, Dwayne Day, Jonathan McDowell and Allen Thomson.

Link to the VSO Home Page