Atlas 109D re-entry seen and debris found in Brazil and South Africa

From: Ted Molczan <>
Date: Sun, 16 Mar 2014 20:12:59 -0400
I had long been aware that the core stage of Atlas 109D (1962-003B / 241), that launched John Glenn on 1962 Feb 20, had
been observed to re-enter over South Africa and that fragments were found there, near Aliwal North. That fact was
officially acknowledged by the U.S. on Aug 31, 1962, as documented in Project Blue Book Case 7811:

A photo of one of the fragments is available here:

Just recently, I learned that re-entry sightings and debris falls that correlate with Atlas 109D also occurred over
Minas Gerais, Brazil, less than 20 min. before the South African events. The U.S. promptly acknowledged that the
Brazilian events were due to an Atlas missile, but apparently lacked sufficient confidence to correlate it with Atlas
109D, for reasons I will discuss.

I believe that both cases are of interest, not only because they may be the earliest recoveries of debris fallen from
orbit, but because they involved two widely separated debris fields. The western and easternmost debris locations were
about 67 deg of great-circle arc apart, or more than 7400 km. I know of no similar case.

How might that have happenned? What role might the physical characteristics of the Atlas 109D stage played? Could the
re-entry have begun with the stage intact, or might it have first broken up in orbit, with the Brazilian and South
African events representing two different decays? (The core stage of Atlas SLV-3 (1966-046B / 2188) fragmented in

The remainder of this message documents the archival information I have found about the Brazilian events, with a bit of

Information on Brazilian Events

I learned of the Brazilian events via Project Blue Book Case 7820, regarding things reportedly seen and heard on
February 27, 1962 at 23:00 UTC, followed by debris finds, officially identified as fragments of an Atlas missile. The
nine page case file is available here:

Pages 2-5 are a detailed report of the sightings and debris recovered, telexed by the U.S. Air Force Attaché and
scientific attaché of the U.S. Embassy in Brazil on 1962 Apr 2. Witnesses described visual and acoustic phenomena
consistent with a large meteoric fireball or re-entry from Earth orbit. I have produced the following transcript that is
more convenient to read:

Page 6 of the Blue Book case file is a message from FTD (Foreign Technology Division) at WPAFB (Wright-Patterson AFB),
sent on Apr 3. It is not very legible, but careful reading reveals that it reports no correlation with known foreign
(Soviet) successful or failed orbital or ICBM launches, and lists several recent examples of each.

Pages 7-8 are a telex from NORAD to FTD and others on Apr 6, which references the message from Brazil of Apr 2, and
reports that the decay sighting could not be equated to any known decay of a US space object on or about 27 February,
but allowed that it could possibly be related to a US deep space probe or a vertical firing for which they had no
trajectory data.

Page 9 is a memo for the record, stating that the FTD representative at Patrick AFB AFMTC (Air Force Missile Test
Center) phoned on Apr 18 and reported that the "object was identified as belonging to ATLAS" and that "AFMTC is taking
necessary action for recovery and explanation to Brazilian attaché."

Noting that the Brazilian and South African sightings reportedly occurred within a few minutes of the same time of day,
but exactly 7 days apart, I suspected that Brazil might have got the date wrong. Analysis of the single existing TLE of
Atlas 109D (1962-003B / 241) confirms that it passed almost directly over the locations of the Brazilian debris finds on
Feb 20, less than 20 min. before the South African sightings. The correlation was strong in time and direction of
travel. In reconstructing the events after the debris was found, weeks after the re-entry, the investigators in Brazil
seem to have gotten the day of the week right, but the week wrong.

This was far too obvious for me to have been the first to have figured it out, which I quickly confirmed by visiting The
Aerospace Corporation's table of recovered re-entry debris:

Item 3 is the case in question:

Event: "In March and June 1962, 11 pieces of stainless steel skin (average mass 2.7 kg) and one sustainer rocket engine
spherical pressure bottle (0.56 m diameter, mass 21.7 kg) were found in Brazil and South Africa."

Remarks: "Identified as pieces from Atlas booster for Mercury MA-6 mission, launched Feb. 20, 1962."

The case seems to have fallen into near-obscurity, compared with the South African events.

I did find one highly relevant report available on the web, "Feasibility Study of the Control of the Mercury/Atlas
Booster Re-Entry from Orbital Missions", by D. Pace and T. Shiokari of The Aerospace Corporation:

The opening paragraph of the introduction explains the motivation and purpose:

"Fragments from the Mercury/Atlas orbital missions have been recovered in
foreign countries. These fragments have been thoroughly analyzed by the
manufacturer of the booster and most of the fragments were identified as
parts from the Mercury/Atlas final booster stage. Apprehension involving
the possibility of problems in international relations arising from the impact
of these fragments on foreign soil, prompted the Air Force to request a
feasibility study to determine if the impacting fragments could be controlled
without jeopardizing the present Mercury Program."

Those missions were MA-6 (John Glenn) and MA-8 (Wally Schirra). The study concluded that it was not feasible to control
the de-orbit. It was published on May 15, 1963. (Gordon Cooper's Mercury-Atlas 9 mission lifted-off the same day, and a
fragment of its Atlas fell on Argentina.)

The authors noted that a different study had found that, "the South American parts have not been positively identified
as fragments from MA-6 booster". I suspect that the other study's criteria for positive identification may have depended
on the presence of markings unique to Atlas 109D, which may not have been found. Apparently the authors of the above
study were satisfied with the strong circumstantial case based on the temporal and spatial correlation with the known
trajectory, with allowance for the apparent 7 day error in the date reported by Brazil. They further commented:

"Assuming the fragments were from MA-6, the distance between the two
reported areas of recovery would amount to approximately 4400 N.M. The
parts recovered were fragments from the lox and fuel tank. However, the
South African recovery had a higher percentage from the lox tank and the
South American recovery had a higher percentage from the fuel tank."

Detailed descriptions of the dozen fragments from both continents, including location coordinates, are found in Table II
MA-6 (109D) General Fragment Data, on pg.18.

Figure 2. MA-6 Fragment Location on Missile, on pg.38, may provide clues regarding the dynamics of the re-entry and

The first three references, on pg.15, are to studies of the Atlas 109D debris by General Dynamics that are not on the
web, but perhaps could be obtained via FOIA request.

Ted Molczan

Seesat-l mailing list
Received on Sun Mar 16 2014 - 19:14:04 UTC

This archive was generated by hypermail 2.2.0 : Mon Mar 17 2014 - 00:14:04 UTC