Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

Lightning Bolts

Download as pdf or txt
Download as pdf or txt
You are on page 1of 308

lightning

bolts
first manuevering reentry vehicles

william yengst

Tat e P u b l i s h i n g & E n te r p r i s es
Lightning Bolts
Copyright © 20 by William Yengst. All rights reserved.

No part of this publication may be reproduced, stored in a retrieval system or


transmitted in any way by any means, electronic, mechanical, photocopy, recording
or otherwise without the prior permission of the author except as provided by USA
copyright law.
The opinions expressed by the author are not necessarily those of Tate Publishing, LLC.

Published by Tate Publishing & Enterprises, LLC


127 E. Trade Center Terrace | Mustang, Oklahoma 73064 USA
1.888.361.9473 | www.tatepublishing.com
Tate Publishing is committed to excellence in the publishing industry. The company
reflects the philosophy established by the founders, based on Psalm 68:11,
“The Lord gave the word and great was the company of those who published it.”
Book design copyright © 2010 by Tate Publishing, LLC. All rights reserved.
Cover design by Lance Waldrop
Interior design by Jeff Fisher

Published in the United States of America


ISBN: 978-1-61566-547-1
1.History, Military, Weapons
2. Technology & Engineering, Aeronautics & Astronautics
10.01.12
Dedication

This monograph is dedicated to Betty and our children,


Dick, Barbara, and Tom, who endured the experiences
with me, and encouraged me to “write it down.”
Acknowledgements

The author is highly indebted to the many friends, sci-


entists, engineers and administrative personnel who
worked with me through the years on the maneuver-
ing reentry vehicle programs described in this mono-
graph. I am honored to claim most of them as life-
long friends.
My wife, Betty, deserves special credit for put-
ting up with the many days and weeks I was away on
travel, the long hours spent at the office, and the ten-
sions at home as I wrestled with technical, schedul-
ing, and related administrative problems. Our oldest
son, Richard, was first to review and edit this manu-
script, checking for content, dates and best methods
of presentation. Our son, Tom, a current employee
of the Aerospace Corporation, provided ideas and
access to the company’s library sources. Access to
our mid-1960s documents, some of which I authored
or approved, was a great help. About ten percent of
the references refreshed my fading memory, while
roughly one quarter were reports prepared by industry
associates on program details and event dates. The
remaining references filled gaps in my knowledge of
programs in which I had no direct involvement and
the foreign MRV developments.
I am greatly indebted to Stephen J. Lukasik for
critically reviewing, suggesting refinements, keeping
me on subject, and preparing an independent fore-
word. Michael E. Montie reviewed and advised me
on aspects of the Pershing-II weapon system program
evaluation that were long forgotten. Both Dick and
Tom collaborated in keeping my computer running,
word-processing software up to date, suitable photo-
graphic images and provided computer training for an
aging “slide-rule” engineer.
Table of
Contents
Forward
11

Preface
15

1. Genesis of a New Concept


21

2. An Introduction to ABRES
49

3. MBRV Development and Test


89

4. BGRV Development and Test


111
5. MARCAS Development and Test
131

6. Second Generation Maneuvering


Reentry Vehicles
147

7. Pershing-II MARV Development and Test


183

8. Indian Agni-II and -III MRV Developments


207

9. China’s M-11 & M-11A MRV Developments


223

10. Observations and Emerging MRV Threats


237

References
263

Photograph Credits
297
Foreword

Stephen J. Lukasik

This monograph tells three stories. The first, and most


obvious, story is that of the technology of maneuvering
reentry vehicles. It leads the reader through intricacies
of reentry from space into the atmosphere. Told in
terms of military systems, the science and engineering
are equally relevant to the broader set of applications
where man has explored, and will continue to explore,
the universe beyond our home on earth.
It is an instantiation of what are now known as
“systems-of-systems:” the separate systems of the
earth and its atmosphere, surface-to-surface missiles,
conventional and nuclear warheads, targets, and stra-
tegic interactions between sovereign states. Born of
the need to integrate the many subsystems that con-

11
william yengst

tribute to the functioning of military weapons, the


Department of Defense created the discipline of “sys-
tem engineering” as part of its use of technology to
protect the U.S. With the postwar invention of digital
computers small enough to be embedded in weapon
systems, new opportunities became available to link
smaller systems into systems-of-systems of increasing
power and complexity.
But if “systems engineering” was needed to bring
systems into existence, new processes were needed
to manage the simultaneous development of com-
ponents for systems-of-systems. Eschewing formal
descriptions of the process that challenge the interest
of all but a few, the story told here leads the reader by
example through integration of new technology, pro-
gram planning and management, funding, the larger
political processes and requirements, domestic and
international, that govern outcomes.
The second story is one of the lesser-known aspects
of the Cold War. The leading actor in that superpower
drama was the long-range ballistic missile, deployed
by the thousands, each capable of delivering megaton
nuclear warheads across the globe for each side.
While the public was aware that their govern-
ment’s policy of mutual assured destruction, thus
making them hostages to the presumed rationality of
the other side, it was less aware that that posture was
adopted to avoid major deployments of the offensive
systems described here and the necessity of provid-
ing for their continued operation, elaboration, and
refinement. This was an admission that man had cre-
ated technical systems whose interactions were so
complex that they were beyond our ability to under-

12
lightning bolts

stand in all their implications. The strategic analyst’s


description of this situation was that offensive systems
would introduce such dangerous uncertainties into an
already barely stable strategic nuclear balance between
the superpowers that they must be avoided ab initio.
The third story is a look at the enormous techni-
cal and strategic issues from the perspective of those
who had the task of merging these separate domains.
This is a story of the young military and civilian pro-
fessionals in government, industry, think-tanks, and
universities dedicated to these central issues of second
half of the 20th century.
While dealing with details of radio propagation
through plasma sheaths around bodies moving at
hypersonic speeds in regions never before encoun-
tered and thousands of complex technical questions
and tradeoffs, they had to deal simultaneously with
global strategy as well as the minutiae of office space,
commuting, housing and raising their families.
By looking into this aspect of the cold war, the
reader will more fully understand what the phrase
“Cold War” means and how the world found its way
to survive. The journey was not smooth, and there
were opportunities for disaster at every turn. The
microscopic perspective of this monograph provides
new appreciation for the observation that “the devil
is in the details” and provides encouragement that,
however complex, man’s problems can be managed
through the efforts of trained and dedicated profes-
sionals at all levels of society.

13
Preface

As a Duke University electrical engineering gradu-


ate (1954), I received a Howard Hughes fellowship
to do graduate work at the University of California
at Los Angeles (1956), plus a job at Hughes Aircraft
Company. In those days, the first U.S. long-range,
intercontinental ballistic missile (i.e., Atlas) was being
considered by General Dynamics Convair in San
Diego; however, my entry into that industry took a
detour. The Air Force called Second Lieutenant Wil-
liam C. Yengst to active duty at Holloman Air Force
Base (New Mexico) in June of 1956, barely two weeks
after Betty J. Pitman and I were married. Arriving at
Holloman, I was assigned to work under the direc-
tion of Doctor Ernst A. Steinhoff, one of Germa-
ny’s top World War II scientists, developing inertial
guidance gyros and accelerometers for V-2 missiles.
He migrated to the United States at the end of the
war along with Wernher von Braum under “Opera-

15
william yengst

tion Paperclip,” and he was noted for inventing guid-


ance and control components for the V-2. Based on
descriptions of advanced German ballistic rocket pro-
grams and tutelage by their team of engineers, I was
inspired to pursue a career in guidance and control
systems for long-range missiles.
After discharge from the Air Force (November,
1957), I returned to civilian life at Hughes Aircraft
Company, where I worked on inertial and infrared
guidance units for air-to-air Falcon missiles. But
my interest in long-range missiles prevailed; there-
fore, when Congress approved establishment of the
Aerospace Corporation in June of 1960, I was among
the first to apply. I was hired by Aerospace for the
Guidance Systems Department to perform design
studies in support of Minuteman-I plus emerging
terminal guidance concepts. During August of 1962,
I was afforded the opportunity and responsibility for
managing the experimental development and flight-
testing of the first United States maneuvering reen-
try vehicles (MRVs) for ICBMs. The development,
sponsored by the Department of Defense (DOD)
Research and Engineering (DDR&E) office of the
Pentagon, was preparing technology capable of sup-
porting production of future reentry warheads that
could defeat any known Soviet Union anti-ballistic
missile (ABM) defensive weapon systems.
MRV developments were managed by the Air
Force’s Ballistic Missiles Division (BMD) within
its highly classified Advanced Ballistic Reentry Sys-
tems (ABRES) program. My Aerospace office at
San Bernardino (CA) was assigned the non-descript
title “Adaptive Systems Program Office,” and we were

16
lightning bolts

responsible for technical direction and general sys-


tems engineering of the developments. During the
next six years, our office oversaw definition and test
of three maneuvering reentry vehicle concepts known
as: Maneuvering Ballistic Reentry Vehicle (MBRV ),
Boost Glide Reentry Vehicle (BGRV ), and Maneu-
vering Reaction Control & Ablation Studies (MAR-
CAS). Very little was reported about these pro-
grams. Virtually all engineering, development and
test information generated by ABRES was classified
and closely held to insure against it falling into hands
of Soviet or other foreign nations; thereby, revealing
U.S. capabilities to deliver nuclear weapons, penetrate
enemy Anti-Ballistic Missile (ABM) defenses or
defend against enemy ICBM threats.
Within three years after completing the first
experimental programs, I moved to Science Appli-
cations International Corporation (SAIC) as man-
ager of its Advanced Concepts Division. Our work
focused on two high-visibility programs concerning
applications of maneuvering reentry vehicle weapon
systems. The first, Long Range Research and Devel-
opment (LRR&D) study sponsored by the Advanced
Research Projects Agency (1971-1973) involved
evaluating concepts for achieving highly accurate
delivery of conventional and low-yield nuclear war-
heads. The second, Pershing-II Cost and Operational
Effectiveness (COEA), directed by Congress and
U.S. Army (1979-1980), was funded by the Defense
Nuclear Agency to establish requirements, rationale
and potential effectiveness of the Pershing-II maneu-
vering reentry vehicle and the Air Force counterpart,
Ground Launched Cruise Missile (GLCM).

17
william yengst

A half-century has passed since the first MRV


programs were completed and many scientists and
engineers who worked on them are no longer able
to document the history and progress made with
the technology. Because the programs were clas-
sified ‘Secret Restricted Data (SRD),’ there was no
release of information to open media (i.e., no books,
reports, newspapers or television). Except for occa-
sional meetings between people who worked on the
projects, I have rarely heard the subject mentioned.
Now, well past classification downgrade expiration
dates, it is time to present as much information as I
can remember.
Additional motivations for writing this history are
to: (1) present the rationale and goals of the feasibility
and advanced development programs, (2) show how
the programs influenced international arms control
treaties and agreements, (3) honor dedicated people
who devoted years of hard work to making the vehi-
cles successful, and (4) encourage future scientists
and engineers to pursue similar careers. Although
the background history, engineering principles, and
difficulties we encountered during these projects are
presented, it is likely that those capabilities and tech-
nologies described have been surpassed by modern
reentry vehicles.
Examples of recent foreign developments of the
technology are presented in the final chapters of this
monograph. Few nations have the financial means
and patience to implement exquisitely detailed devel-
opment programs such as MRV to achieve weapon
capabilities to counter enemy threats. China, India,
Pakistan, and Russia are the only countries known

18
lightning bolts

to have initiated such programs. Finally, the Wright


Brothers’ experiments with the first airplane are sum-
marized in the concluding chapter to illustrate and
contrast their approach with the sophisticated and
detailed process we adopted in designing and testing
MRVs. Sources of referenced material throughout
the text are listed at the end, but many drawings and
concepts are presented strictly from memory.
I give great credit to the bright, innovative, ener-
getic people who worked on these projects plus the
companies and organizations that contributed to the
formulation, funding of experiments, and promoting
their success. In fact, each project involved hundreds
of dedicated people for several years and cost tens of
millions of dollars…large sums of money for the time.
I want to thank everyone who took part and hope they
find this document entertaining and accurate.
—Bill Yengst

19
chapter 1

Genesis of a New Concept


I have been told and reminded many times during my
engineering career, that totally new concepts or ideas
are exceedingly hard to find or invent. Maneuvering
reentry vehicles (MRVs) represent an idea invented
by German rocket scientists during the most intense
period of World War II; therefore, it is imperative that
one review history to understand the reasons for their
development and some of the difficulties encountered
in making the technology a reality. Like many tech-
nologies, MRV concepts were driven by demands or
threats of warfare. The war ended before the inven-
tion became a reality, and it was not until 1966—some
twenty-one years later—before the first successful
flight tests of MRVs took place.
For those not familiar with long-range missile
system terminology, a reentry vehicle (RV ) is the

21
william yengst

nose section of a missile, which contains the payload


(usually a warhead), heat shield, and structure that
separates from the booster rocket, then follows a bal-
listic projectile trajectory under the pull of gravity
through outer space, and protects the payload through
hypersonic reentry heating and other environments.
By comparison, an MRV includes (in addition to its
payload, heat shield, and structure) a set of aerody-
namic control mechanisms and an on-board guidance
system with built-in logic. It follows the same series
of events as an RV until returning to the atmosphere
where it can make use of astonishing kinetic energy to
steer away from a ballistic trajectory; thereby extend-
ing flight range, changing direction to evade enemy
defenses, and greatly improving accuracy. Military
implications of this capability are enormous; there-
fore, this book is devoted to explaining the genesis
and development history of MRVs, the secret, inter-
continental-range Lightning Bolts of the “Cold War.”
When the U.S. started maneuvering reentry tech-
nology programs to capitalize on German concepts
described in this chapter, there were no visions it
might stimulate the need for an Anti-Ballistic Mis-
sile (ABM) Treaty or provide an incentive for imple-
menting Strategic Arms Limitations Talks (SALT-
II), which eventually reduced nuclear weapon threats
in Europe between Soviet-backed (Warsaw Pact)
and U.S.-backed (North Atlantic Treaty Organiza-
tion -NATO) forces. Chapters 2 through 7 of this
monograph present the background and rationale for
developing the technology, while explaining difficul-
ties encountered during early developments. Chapters
8 through 10 describe how the technology matured

22
lightning bolts

from 1985 through 2009 and is now being used in


weapon systems developed in India, China, Pakistan,
and Russia.
One might wonder why a young engineer or sci-
entist would choose to work long days, endure many
tiring trips (some in terrible weather conditions), be
separated from loved ones during holidays and spe-
cial occasions, and devote years of his or her career
to develop advanced weapon systems. Answers to this
question obviously vary from one person to another.
My experience was that between multiple technical
and political difficulties, frustrating delays, and sudden
catastrophic failures, there were long periods of enjoy-
ment working with intelligent and dedicated friends,
solving challenging technical problems and experienc-
ing moments of soaring inspiration and overwhelm-
ing joy when tests proved successful. Beyond that lies
the enduring satisfaction that comes from contribut-
ing directly to the nation’s security. In reviewing the
history, I enjoyed a sense of continuity in contributing
to the technology development.

German Inspiration
During October, 1933, several unfunded but promising
rocket experiments were conducted by Wernher von
Braun and Reinhold Tiling of the German Rocket
Society. Therefore, in 1934, the German Wehrmacht
(“German Army”) established a research project at
its artillery testing ground near West Kummersdorf
to develop the technology. The Wehrmacht Weapons
Office, dedicated to advancing artillery technology,

23
william yengst

assigned thirty-five-year-old engineer, Captain Wal-


ter Robert Dornberger, to direct the project. Applying
customary German efficiency, Dornberger and his staff
outlined a series aggregate of ten development goals.
The A-1 rocket, based on von Braun’s doctorial thesis
and concepts, was a small device (fifty-five inches long
with a takeoff weight of 330 pounds). It used an engine
designed by Arthur Rudolph that consumed alcohol
and liquid oxygen. The proof-of-concept model rocket
blew up on its launch pad. The A-2 rocket, built in
1934, was slightly larger (sixty-three inches long with
an empty weight of 158 pounds and takeoff weight of
236 pounds). It used the same fuel and engine, but
with careful design using wood and aluminum frames
instead of iron, its weight was reduced. A Gyroscope
was added on its vertical axis between the alcohol and
oxygen tanks to stabilize the device in flight. Two A-2
rockets were built and launched on December 19 and
20, 1934, at Borkum Island in the North Sea to impress
the army brass.1, 2
The idea of using the rocket for weapon delivery
existed by the mid-1930s. A Peenemunde island Design
Office was set up in 1936 off Germany’s Baltic coast
to develop and test the technology. Rockets promised
longer projectile ranges than could be achieved by
artillery. The A-3 rocket was twenty-two feet long, had
a diameter of twenty-six inches and launch weight of
1,650 pounds. It was designed to test components for
a planned, larger A-4 missile; hence, it carried a para-
chute system so that instruments could be recovered.
Wehrmacht General von Fritch witnessed static fir-
ings of an A-3 engine at Peenemunde in August of
1936, after which several rockets were launched, reach-

24
lightning bolts

ing a maximum altitude (i.e., apogee or apex of its


parabolic trajectory) of eleven miles and downrange
distance of 7.5 miles. Three A-3 rockets were launched
in the autumn of 1937, and each carried a three-axis
gyro control system being developed for the next gen-
eration, A-4 missile. It was found that the rockets
were unstable (several toppled over) during the low
speed, first moments after lift-off. This problem was
solved with ingenuity by adding graphite vanes inside
the rocket’s high-velocity exhaust to provide stability
control.1
After Germany occupied Poland, the British naval
attaché in Oslo sent a message (4 November, 1939)
to London’s intelligence office warning of the secret
weapon developments at Peenemunde. However, it
was not until extensive Luftwaffe aircraft losses dur-
ing the Battle of Britain ( July-October 1940), that
Adolph Hitler took more than a passing interest in the
new weapons by visiting the research facility. The first
A-4 was launched in March 1942, but it crashed and
burned, as did the second rocket. The third A-4 per-
formed perfectly on 3 October 1942. During a review
(early November 1943), Hitler established two strate-
gic targeting priorities for the new weapons: (1) Lon-
don and selected British industrial cities and (2) New
York plus eastern United States cities. Consequently,
the first operational A-4 missile (known as Vergltung-
swaffe Zwel or “Vengeance Weapon 2 . . . V-2”) attacks
were launched against London on 2 September, 1944,
then Paris, and after 4 January 1945, against Antwerp
from launch sites near The Hague, Netherlands. The
long-range goal of attacking U.S. targets was beyond
reach of V-2, and by directive, development of a new

25
william yengst

missile was deferred in October 1944. The U.S. and


Allied invasion of France on D-day threatened Ger-
many and placed urgent demands for troops, critical
materials and funding; consequently, development of
longer-range was delayed but not forgotten.1
The A-4 rocket had a range of 109 miles, apogee
altitude of 93 miles and reentry velocity of 930 miles
per hour (about 1,370 feet per second). Its length was
46 feet, diameter of 65 inches, fin span of 140 inches
and launch weight of 28,380 pounds, including its
high explosive warhead of 2,200 pounds (1 metric
ton).2 During powered flight, its engine consumed
30 gallons of alcohol and liquid oxygen per second.3
An important feature of the missile was that its nose
cone, containing the warhead, and the entire airframe,
including engines and empty fuel tanks, were carried
all the way from launch to impact. The heavy warhead
in the nose and fins attached to the aft end acted like
an arrow (i.e., with a heavy tip and light tail feathers)
to provide stability during reentry.
Dr. Ernst A. Steinhoff, Director of Flight
Mechanics, Ballistics, Guidance, Control, and Instru-
mentation at Peenemunde, was manager of the A-4
project.4 He is shown in Figure 1 and was responsible
for designing the guidance instruments and three-axis
gyro stabilization system for A-4. During 1944, he
planned and directed the A-4b Submarine Launched
Ballistic Missile (SLBM).

26
lightning bolts

Figure 1. Dr. Ernst A. Steinhoff ,pioneer


rocket guidance scientist

On the morning of 7 July, 1943, Major General


Dornberger, Von Braun, and Ernst Steinhoff ignored
Luftwaffe air base reports “not to fly because of foggy
weather.” They boarded a Heinkel-111 bomber and flew
to East Prussia, where they met with Adolph Hitler at
the Wolf ’s Lair, his command center. The meeting was
to answer questions about the V-2 project and arrange
a rocket demonstration at Peenemunde. During their
discussion, Hitler was shown color films of an entire
operation, taking the rocket from its hanger, moving
to a launch pad, placing it in the vertical position, and
the rocket rising slowly on vivid, surging flames. The

27
william yengst

missile flew a perfect ballistic trajectory and landed


120 miles away. Hitler was delighted with every word
of Von Braun’s narration. He approved mass produc-
tion of the weapon at a rate of 2,000 per month, based
on estimates made by the Minister of Industry, Albert
Speer.2 The A-4 design (or V-2) was put into full pro-
duction in 1943.1
The A-5 in the rocket development series was
designed in 1938 and was somewhat smaller than A-4
(i.e., length of 19.1 feet, diameter of 30.7 inches, and a
launch weight of 1,980 pounds). It was first tested in
October 1939 but only to verify control system com-
ponents planned for A-4. It reached an altitude of
7.4 miles on several flights. Between 1941 and 1945, 70
A-5 rockets were launched for test purposes.5 The A-6
weapon was suggested in 1940 by Von Braun as an
improved A-4b, using nitric acid and kerosene fuel.
A-7 was a winged missile, built for the Kriegsmarine
(“German Navy”) and similar in structure to A-5 but
with larger tail fins to provide lift and greater range.
It was 19.4 feet long, had a diameter of 15 inches, and
a launch weight of 2,200 pounds. The A-8, designed
in 1944, had a takeoff weight of 22,370 pounds and
diameter of 30.7 inches but was never built or tested.
To implement Hitler’s long-range strategic objec-
tive, Von Braun established ‘Projekt Amerika’ to pursue
modified A-4 missile designs. Kurt Patt of the Design
Office proposed adding a winged-upper stage to con-
vert the rocket’s speed into aerodynamic lift; thereby,
extending its range a few hundred miles. The concept,
known as Flossengeschoss (“finned projectile”), was
included by Walter Dornberger in a 31 July 1940 mem-
orandum to Hitler regarding the “America Rocket.”1

28
lightning bolts

It was the first design to be called ‘boost-glide’ and


resulted in two competing designs (A-9 and A-10),
both initiated in 1943. A-9 and A-10 were the first
practical transatlantic ballistic missiles. Their range
was estimated at 5,000 kilometers (3,100 miles), pay-
load was 1,000 kilograms (2,200 pounds), diameter of
4.13 m (13.5 feet), length of 41 m (134 feet), and launch
weight of 85,300 kilograms (188,000 pounds). Shortly
after A-4 (or V-2) was put into production, initial
flight test plans for A-9 and A-10 were set for 1946.6
Work on A-9 and A-10 was officially suspended
in late 1944, because of the demand to speed up pro-
duction of V-2. However, the A-4b SLBM design
was given top priority. It was to be launched from a
U-boat (submarine) towed platform off the coast of
North America. Despite orders, Von Braun continued
work on A-9 and A-10 by hiding it under the A-4b
SLBM program. A test of A-4b hardware occurred on
4 January, 1945, barely four months before the defeat
of Germany.6
Dornberger’s staff explored all practical methods
for extending the V-2 rocket’s range under “Projekt
Amerika.” Two basic approaches were taken: (1) scale
up the A-4 missile by developing larger engines for
its first stage plus adding ballistic upper stages and
(2) using aerodynamic lift to extend range by glid-
ing. German artillery scientists had observed early in
the war that mountain-top launches of high-fineness
(long and slender) projectiles traveled much faster and
further than would be predicted by classical ballistic
theory. This led to tests of hypersonic lift and drag
of the projectiles, because of their troublesome ten-
dency to fly, which created a major problem in pre-

29
william yengst

dicting artillery impact accuracy. Consequently, the


Alpha Draco concept of a high lift-to-drag, spinning
projectile was identified as a candidate for the A-9
reentry vehicle. The concept was named for the bright
star Draco, which along with the North Star, Polaris,
and several other stars in the Ursa Minor constella-
tion, formed an astrological dragon figure (Draco in
German).21 It subtly conveyed the notion of meteor-
like projectiles streaking across polar skies toward
American cities.
Two separate engineering teams were organized
to work on the two long-range design approaches.
The A-10 design was based on the first approach, scal-
ing up the size of the A-4 by employing a cluster of
six combustion chambers feeding a single expansion
nozzle. A test stand was built at Peenemunde to test
fire this engine with thrust levels up to 200 metric
tons. The A-9 design followed the second approach, a
refined A-4 with swept wings. Both scaled-up missiles
are illustrated on the right side of Figure 2.6 Although
the initial illustration showed a winged vehicle, the
design evolved to a pointed ogival nose (similar to
V-2) attached to a Messerschmitt Me-109 fighter-
like (single seat cockpit) fuselage with narrow lift-
ing strakes down both sides, extending to the rear for
aerodynamic flap and rudder controls.
At intercontinental ranges, time-of-flight and
delivery accuracy became primary considerations.
Consequently, the A-9 missile would be piloted. After
its engine cut off at an altitude of roughly 240 miles
and speed in excess of 11,000 feet per second, the mis-
sile would reenter the atmosphere and begin a long
glide to extend its range. The pilot would be guided

30
lightning bolts

by radio beacons from surfaced German submarines


located near the U.S. Atlantic coast. After reaching
the target area, the pilot would accomplish an optical
sighting of the target and lock the steering, and then,
he would eject in a parachute recovery capsule. Death
or internment capture of the pilot as a prisoner of war
were accepted as probable outcomes.6 These concepts
were never put into hardware or tested, because the
Russian Army captured Peenemunde in April 1945.

Figure 2. Initial designs for the German A-9 and A-10 missiles

Operation Paperclip
In mid-January 1945, the Soviet “Winter Offensive”
reached the forests of Wolgast, which were not far

31
william yengst

from Peenemunde. All German missile testing was


cancelled and by mid-February, the work was moved
to Rethen on the Weser River west of Hannover.2
German rocket research was completely disorganized,
although a new test station was being built inside an
iron ore mine at Bleicherode and accommodations
were prepared in surrounding villages for the scientists
and their families. Nearly 5,000 people were moved to
the new site along with the important equipment and
documents. All available working vehicle transporta-
tion, including trucks, private cars, and barges, were
commandeered to accomplish the move. The com-
mandant was sent a message in early April to defend
the new facilities by every means and hold it at all
costs. However, before the message arrived, American
forces advanced so near the facility that German offi-
cers ordered its evacuation.2
Engineers Dieter Huzel and Bernard Tessman of
Von Braun’s team were traveling from Peenemunde to
Bleicherode in trucks at night on 1 April with fourteen
tons of documents (archives of the research), when
they heard American artillery. The next day, Nazi SS
guards transported five hundred V-2 personnel to
Oberammergau in the Bavarian Alps. The guards were
under orders to destroy everything of use to the enemy
if they were about to be captured. The implication was
that the scientists would also be killed. Von Braun,
who had broken his arm in an automobile accident,
was well aware of the danger to his scientists. General
Dornberger, Wernher Von Braun and Ernst Steinhoff
were put up overnight in the village of Oberjoch. That
night, Dornberger sent Von Barun’s brother, Magnus,
who spoke English, through the German lines to find

32
lightning bolts

any American post. On 2 May, in the town of Reutte,


Magnus arranged to turn 150 German rocket experts
over to the U.S. 44th Infantry Division.2 Three days
later Soviet Marshal Rokossovsky occupied Peene-
munde and captured the remaining personnel, mainly
those working on the scaled-up A-10 design team.
As a result of these fateful events, the United
States obtained a team of 127 scientists (including
Werner Von Braun, Walter Dornberger, and Ernst
Steinhoff ) plus about 150 engineers, who had worked
on the A-9 concepts, while the Soviet Union captured
personnel who were working on large propulsion
engines for the A-10. In the years from 1945 to 1950,
the Soviet Union exploited the German scientists to
design and build the world’s largest rocket engines.
By contrast, the U.S. concentrated on smaller engines
and more sophisticated, lighter weight vehicles using
aerodynamic lift to achieve long-range missiles.
Under Operation Paperclip, 300 trainloads of V-2
missiles and component parts were captured and sent
to the U.S. and Great Britain. The U.S. missile pro-
gram made good use of its V-2s to learn the tech-
nology and test fly improved components. Von Braun
went to work for the U.S. Army’s Redstone Arsenal
in Huntsville (Alabama) where he helped develop the
Redstone, Jupiter, and Pershing-I missiles, plus Saturn
rockets used for the moon-landing programs. Britain
took a small number of V-2s, which were launched
from a site in northern Germany in 1946 under Oper-
ation Backfire.
Walter Dornberger, detested in England for his
role during the V-2 operations, was held captive by
British authorities. During his incarceration, he helped

33
william yengst

with test firings of V-2s during Operation Backfire.


The government released him in 1947, and he moved
to the United States. Dornberger became a consultant
for the U.S. Air Force and an advisor on weapon sys-
tems for the Department of Defense. He worked for
Bell Aircraft Company for a few years but moved to
McDonnell Aircraft Company (St. Louis) in 1950 and
advocated development work on Alpha Draco, leading
to the advanced Weapon System (WS-199/D).7
Dr. Ernst A. Steinhoff and a dozen guidance
experts settled in Alamogordo (New Mexico), where
they helped launch captured V-2s from the White
Sands Missile Range (WSMR). Between March 1946
and September 1952, a total of 64 V-2s were tested,
at a rate of one or two per month.8 The V-2 missile,
shown in Figure 3 with its guidance section open,
was on display at the WSMR Museum in September
2007. Specifically, it used gyros to establish vertical
attitude and azimuth direction, while a simple analog
computer provided steering corrections to follow the
intended trajectory.5 Flight range was determined by
the amount of fuel consumed (i.e., “brenschluss” or end
of burning). Later models of A-4 used radio “guide
beams,” transmitted from the ground to improve navi-
gation accuracy.

34
lightning bolts

Figure 3. V-2 missile at White Sands Missile Range Museum

On 19 April, 1946, the Department of Defense


gave Consolidated Vultee Aircraft (later Convair divi-
sion of General Dynamics) a contract to study designs
for an intercontinental ballistic missile.9 Because of an
early 1950 Central Intelligence Agency assessment of
Soviet long-range missile developments, the first U.S.
ICBM program, “Project Atlas,” was initiated in Sep-
tember 1951 at Convair (San Diego). Initial emphasis
was focused on the primary problem, development of
main booster engines. North American Aviation pro-
vided the design and tests of a 75,000 pound thrust
engine. Convair clustered three of the engines for the
Atlas missile.9
The Atlas (SM-65) missile development was accel-
erated in January 1955. It had an estimated range of
5,500 miles with a W-38 thermonuclear warhead (3.75
megaton yield). The design was unique, consisting of
two booster and one sustainer engine (described as 1.5

35
william yengst

stages), plus five ‘veneer’ rockets, all ignited prior to


lift-off. The rockets all drew liquid fuel (RP-1) from
the same tank plus a liquid oxygen (LOX) tank. The
initial design relied on radio guidance, since the first
inertial guidance system was not fully developed and
weighed nearly one ton. The ballistic reentry vehicle
selected for the operational missile was Mark-2, a
heavy, blunt ‘heat sink’ configuration that relied on
absorbing reentry heating, designed and built by Gen-
eral Electric.
The system was expected to have an accuracy or
Circular Error Probable (CEP) of 4,600 feet. CEP is
defined as the radius of a circle, inside of which half
the shots will fall. The first flight test was conducted at
Cape Canaveral (Atlantic Missile Range) on 11 June,
1957, with a second flight on 25 September.10 Initial
operational Atlas missiles were deployed in 1959 on
soft launch pads in “coffins or canisters” at Shilling
AFB (Kansas). However, information on the system
was not released to the public by the U.S. Air Force
until 15 October, 1958.11 Unfortunately, the missile sys-
tem was considered so vulnerable to attack (i.e., blast,
radiation, and electro-magnetic pulse or EMP) that it
was replaced in 1962 by Atlas-D with a new Mark-4
‘ablation’ conic reentry vehicle that delivered the same
W-38 warhead.

WS-199/A Program and Alpha Draco:


Before the nonprofit Rand Corporation was established
on 2 May 1946, the U.S. government sponsored Project
‘Rand’ under which a study entitled, “Preliminary Design

36
lightning bolts

of an Experimental World-Circling Space Ship,” was


published. The study report made the insightful obser-
vation, “The U.S. military could design and construct
craft which can penetrate the atmosphere and achieve
sufficient velocity to become satellites of the earth.” The
report backed by several years of lobbying from aircraft
industry advocates, including Walter Dornberger, con-
vinced the Air Force and Congress in 1956 to initiate a
research program, to be managed by the new Strategic
Air Command (SAC). Congress supported SAC’s desire
to develop backup technology for its ballistic missile
programs (i.e., Atlas from 1954 to first flight in 1957 and
Titan-I from 1955 to first flight in 1962). Both ICBM
programs experienced test problems and were struggling
to stay ahead of rapid Soviet missile developments.
Consequently, the modest research feasibility pro-
gram known as WS-199A was approved by Congress
and initiated in 1957. Within this program, three teams
of aircraft companies (labeled B, C, and D) would each
build and test hypersonic missiles to gather technical
data, prove feasibility, and generate designs for future
weapon systems.12 If needed, one or more of the designs
would be selected for rapid development as an opera-
tional weapon system.
WS-199/B: Martin Aircraft (Orlando) received a
contract to determine feasibility of an Air-Launched
Ballistic Missile (ALBM) known as ‘Bold Orion.’
WS-199/C: Lockheed (Palo Alto) plus General
Dynamics Convair (San Diego) received a joint contract
to develop an ALBM known as High Virgo for use on
B-58 Hustler bombers.
WS-199/D: McDonnell Aircraft Company (St.
Louis) was given a contract to build and test a two-stage,

37
william yengst

hypersonic rocket known as Alpha Draco for a missile


nose cone using an unpowered, boost-glide reentry vehi-
cle (BGRV).
This monograph is concerned with the third con-
cept (BGRV). Due to serious questions about aerody-
namic performance and structural design survivability
of the boost glide vehicle at hypersonic speeds, Alpha
Draco was only the first step in a ten-year development
plan. McDonnell Aircraft would built three Alpha Draco
projectiles in 1957 and test them by 1959, using Thiokol
TX-20 Sergeant and TX-30 second-stage solid-fuel
booster rockets. The two-stage missile was mounted and
fired from a vehicle launch rail at Cape Canaveral (Flor-
ida), Launch Complex LC-4. Figure 4 shows an Alpha
Draco projectile during launch preparation. Flight tests
involved the first-stage booster burning for 32 seconds
to lift the projectile to about 42,000 feet. After a brief
coast period, the second-stage motor boosted it to over
100,000 feet and to a velocity of Mach 5+ (~5,000 feet
per second).12 The subsequent glide phase of flight ended
with a programmed dive into the ocean about 240 miles
from the launch site.

Figure 4. Alpha Draco (boost-glide) projectile


(WS-199D) on a Sergeant booster

38
lightning bolts

The Alpha Draco projectile was 169 inches long (14.1


feet), 31 inches in diameter, and had a sharp half-cone
angle of 5.36 degrees. There was no great problem with
aerodynamic heating on the one-inch diameter nose
tip, and no heat shield required at velocities of Mach
5; therefore, a Nickel alloy skin was used for the nose
cone. The projectile weight was roughly 300 pounds
with no payload other than diagnostic sensors and a
small, battery-powered telemetry transmitter. Sandia
Corporation provided advanced materials for the tests
and indicated the vehicle had a skirt of movable flaps
that made it possible to trim its attitude with a slight
nose-up angle of attack.13 Two Alpha Draco flights (16
February and 16 March 1959) were completely suc-
cessful, but the final test on 27 April 1959 deviated
from its planned path and was destroyed by “Range
Safety.”13 According to Guenter F. Wendt, a German
engineer who joined McDonnell Aircraft in 1949, the
entire Alpha Draco program contract (1957–1959) was
$5 million.14
In 1961, I learned the major achievement of Alpha
Draco flights was the measurement of the projectile’s
lift (L) and drag (D). The sharp cone achieved a high
L/D ratio of 3.5, at a few degrees angle-or-attack
and hypersonic speeds. This assured it could achieve
very long glide ranges. Furthermore, nose tip stagna-
tion air pressures and body thermodynamic heating
(while not problems for Alpha Draco), would be seri-
ous design difficulties at intercontinental ranges and
reentry glide conditions (i.e., velocities near 18,000
feet per second). Another step of proving feasibility
would be required before a long-range weapon system
could be developed with confidence.

39
william yengst

Insight into Ballistic Missile


Guidance Problems:
As an electrical engineering graduate with a U.S. Air
Force Reserve Officers Training Corps (ROTC) com-
mission, I reported for active duty at Holloman Air
Force Base with my new bride, Betty, in mid-July
1956. At the time, I knew nothing of the above his-
tory or on-going ICBM developments. My immedi-
ate assignment was in the guidance directorate labo-
ratory reporting to Dr. Steinhoff. The assignment had
a profound effect on my future career. I was highly
impressed by the elderly, gray-haired man with a
kindly face, who spoke softly but with an unmistak-
able German accent. He introduced me to the staff
of German scientists in the laboratory, and they pro-
vided me a small wooden desk with a Monroe electric
calculator. I shall never forget the unusual and har-
rowing tales the German engineers told about war-
time conditions, bombing raids, rocket experiments,
and escaping from the advancing Russian Army.
The guidance directorate at Holloman AFB was
contributing to plans and oversight of inertial guid-
ance instrument development to be used in the Atlas,
Minuteman-I, and advanced intercontinental-range
missiles. Active inertial guidance provided rapid
launch response and avoided growing concerns about
radio-guided Atlas ICBMs, which were particularly
vulnerable to electromagnetic radiation pulses (EMP)
and other nuclear weapon effects. The laboratory was
focused on sensitive inertial gyros and accelerometers
needed for the missile guidance system up to separa-

40
lightning bolts

tion of its reentry vehicle (RV ). Most of the staff was


concerned with reducing gyro drifts caused by ‘minute
design imbalances’ that were sensitive to accelerations
experienced during missile lift-off and powered flight.
Errors generated by the instruments due to missile
accelerations accumulate with time and cause an RV
to experience velocity and angular tip-off errors when
it separates from the booster. To illustrate, one must
realize that nearing separation conditions, the RV is
traveling at a speed approaching 18,000 miles per hour
(i.e., 26,400 feet per second). An error in velocity of
only 0.1 foot per second (i.e., 1.2 inches per second)
can propagate during the remaining 20 minutes of
flight to miss the intended target by 120 feet. Angular
errors at separation can also propagate in 20 minutes
of flight to cause the RV to reenter the atmosphere
at an awkward attitude (e.g., sideways or backward)
to wobble or possibly tumble end-over-end. Hence,
extraordinary efforts were made to develop instru-
ments free of error sources under a wide range of tem-
perature, vibration, and acceleration conditions.
Being the new member of the directorate, my ini-
tial task was to learn the theory of instrument opera-
tion and their design characteristics. But, I was also
assigned a project to determine the density of the
atmosphere at altitudes from 50,000 to 200,000 feet.
This project made little sense to me, and I quickly
learned that air density is a very complicated sub-
ject. A few minutes research provided the following
insights:

Air Density

41
william yengst

At sea level, air density is 1.225 kg/m3 (or 0.00764


lb./ft.3) at 15o C (59oF) temperature and 29.92 inches
of mercury pressure. Clearly, the density of the gas
is dependent on its temperature, pressure, humidity,
and particle content. A rule of thumb indicates den-
sity decreases by half about every 18,000 to 20,000
feet of altitude. For example at 40,000 feet altitude
there are roughly one-quarter as many air molecules
per unit volume as at sea level.20 Thus, the reduction
in density is exponential, and even the mixture of oxy-
gen, nitrogen, other gases, and particles changes with
altitude. But, at what altitude does air density vanish
altogether?
The problem was to determine how it varied above
50,000 feet altitude, because it strongly affected the
flight path and range of ballistic missiles. I was sur-
prised to learn how much it affects the measurements
of inertial instruments. I contacted and interviewed
Major David G. Simons, Project Officer for the V-2
monkey flight tests and Chief of Space Biology at
Holloman AFB. Between 1953 and 1959, he directed
over 60 high-altitude balloon flights to measure
atmospheric conditions. About a year after our dis-
cussions, he flew in a balloon to 101,516 feet altitude
(19–20 August 1957).15 He patiently explained to me,
there were a number of phenomena at extreme alti-
tudes that affected air density.
It was not surprising that air density near the
equator is somewhat lower than near polar regions,
since ambient temperatures are higher and cause
atmospheric gases to expand. However, there is a diur-
nal (day-to-night) variation that an RV experiences
as it passes from the dark-to-light side of the earth

42
lightning bolts

(i.e., a bulge in density contours facing the Sun). There


are showers of space dust, cosmic radiation, and winds
at extreme altitudes. Some of these phenomena alter
the intended flight trajectory of a missile and can be
sensed by ultra-sensitive instruments. In short order, I
learned several other factors may also upset the trajec-
tory of a ballistic missile and its accuracy.

Earth Shape
It was common knowledge the Earth is not a per-
fect sphere. In fact, large mountains, deep canyons,
and unbalanced spinning or precession of the earth
can influence alignments and drifts of sensitive gyros.
These factors were sufficiently important that in 1960
the Minuteman-I guidance system computer included
a Fourier series approximation with over thirty terms
for the Earth’s radius and required coefficients for
double precision calculations. Despite these pre-
cautions, the first half-dozen Minuteman-I missiles
launched from Vandenberg AFB (California) to Kwa-
jalein Atoll (Marshall Islands, South Pacific), in 1962,
missed their aimpoints by more than a mile. Analysts
subsequently determined that Kwajalein lagoon’s loca-
tion and orientation on world maps had large survey-
ing errors.
This difficulty is illustrated by the famous Four
Corners marker at the intersection of Arizona, Col-
orado, New Mexico, and Colorado. The marker was
carefully surveyed and put in place by government
surveyors in 1875. However, in April 2009, satellite
and global position system instruments proved it was
1,807.14 feet (0.34 mile) east of the actual state inter-

43
william yengst

sections according to David Doyle of the National


Geodetic Survey.22

Gravity Anomalies
There are segments of the Earth’s crust where large
deposits of iron ore and other minerals distort the
gravity field. The phenomena affects mass and influ-
ence the alignment and calibration of gyros, particu-
larly near a launch site where the guidance system was
intended to operate continuously after installation to
avoid an hour-long warm-up and stabilization times
before launch. Consequently, every Minuteman (and
later MX missile) silo was subjected to detailed sur-
veys to establish their precise locations and surround-
ing gravity fields.

Reentry Environments
As a ballistic RV reenters the sensible atmosphere, it
begins to heat up due to air friction at roughly 200,000
feet of altitude. The nose tip or heat shield experience
temperatures in excess of 3,500 degrees Fahrenheit,
and if the vehicle is not perfectly balanced (or spin-
ning) and aligned with its velocity vector, it can ablate
asymmetrically. This causes the RV to adjust its atti-
tude or angle of attack with respect to the airstream,
and it begins to fly off the intended trajectory. This
phenomena became a serious problem during Minute-
man-III Mark-12 RV development, because nose tips
broke off or developed striations, causing lift and large
impact errors.
Ablation of the heat shield can also cause a sig-
nificant loss of weight and shift the RV’s center-of-

44
lightning bolts

gravity. This can disturb the trajectory, if the vehicle is


not sufficiently stable. Unlike an arrow that relies on
tail feathers at the rear to keep its aerodynamic center
of pressure far behind its heavy nose, a conic RV must
rely on keeping its weight and center of gravity as far
to the nose as possible. This problem is compounded if
the RV must carry a large, heavy warhead that will not
fit close to the nose. In addition, reentry atmospheric
winds, rain (or snowfall), and air pressure or density
variations can deflect the vehicle from its benign, ideal
computed trajectory.
Given these multiple uncertainties, it is not sur-
prising that the accuracy of early ballistic missiles
with inertial guidance systems was somewhat over
one mile circular error probable (CEP). In the early
1960s, those of us studying these parameters often
joked about the desire to fire a missile toward central
Russia with an RV made of ice so that it would melt
and harmlessly turn to rain during reentry. Tests of
this type would permit us to measure the uncertainties
when firing north over the pole.

Minuteman Missile Difficulties


By the late 1950s, there was great concern for rapidly
expanding Soviet nuclear missile forces. The USSR
launched its first intercontinental ballistic missile
(SS-6) on 27 August 1957. An improved SS-7 missile
was soon to follow with a launch in mid-January 1961.
The Soviet ability to build large rockets and launch
them from Southern Russia—beyond range of U.S.
weapon systems—was militarily unacceptable.
In the spring of 1957, the Air Force reported to
Congress that a solid-fuel ICBM with inertial guid-

45
william yengst

ance was possible. Plans for the Minuteman-I missile


called for Thiokol Chemical Co., Hercules Powder
Co., and Aerojet General Corporation to produce
booster rockets. The missile’s inertial guidance system
would be produced by Autonetics Division of North
American Aviation (later Rockwell International Cor-
poration). Boeing Airplane Company would assemble
and test the missile system. It would have a range of
6,000 miles with a blunt, subsonic impact Mark-5
(cone-cylinder-flare reentry vehicle) carrying a W-59
(1.2 megaton yield) thermonuclear warhead.16, 17
Minuteman-I (LGM-30A/B) was approved for a
crash development beginning in September 1958 and
was scheduled to become operational in 1962 at far
northern U.S. bases (Malmstrom and Minot AFB,
North Dakota). Surveying of sites and initial silo
construction were initiated late in 1957; however, the
initial silo plans were revised such that they would
be dug 13 feet deeper. This was done to permit inclu-
sion of a long, slender, boost-glide reentry vehicle
(i.e., Alpha Draco) on top of the Minuteman missile if
needed. Extra depth of the missile silos was later filled
by larger Minuteman-II and -III missiles.
With respect to missile accuracy, improved instru-
ments and better understanding of the earth’s environ-
ment, made it realistic to predict Minuteman-I would
have a CEP of 1,650 feet.18 Minuteman-II with further
improvements, including a new-cone-cylinder-flare-
reentry vehicle (Mark-11 with a W-56 nuclear war-
head) was hoped to provide a CEP of 1,200 feet.19 The
design goal for Minuteman-III with a conic Mark-
12 reentry vehicle would have a CEP of 600 feet by
the mid-1960s, and its successor, MX, had a goal of a

46
lightning bolts

few hundred feet CEP. Unfortunately, reentry flight


errors alone for most advanced ballistic RV designs
accounted for roughly 150 feet of the total CEP. Hav-
ing lived for several years with the fundamental guid-
ance and environmental error sources described above,
I was drawn to a better solution, a maneuvering reen-
try vehicle (MRV ) that could be guided all the way to
its target.
After my discharge from the Air Force at Hollo-
man AFB (November 1957), I returned to an original
job at Hughes Aircraft Company working on the guid-
ance systems for small air-to-air Falcon air-defense
missiles. However, when Congress moved to set up
The Aerospace Corporation (a new “Not-for-Profit”
company) in 1960 to assist the Air Force with devel-
opment of advanced long-range missiles and satellite
systems, I decided to apply for the more interesting
and challenging opportunity.

47
chapter 2

An Introduction to ABRES
On 25 June 1960, Lieutenant General Bernard A.
Schriever of the Air Force’s Air Research & Devel-
opment Command (ARDC) announced formation
of “The Aerospace Corporation” during a press con-
ference in Inglewood (California). Aerospace would
serve the Air Force in scientific and technical plan-
ning and management of missile and space programs.
Dr. Ivan A. Getting, formerly of Raytheon Corpora-
tion, was named the company’s first president. The
first few months ( June through August 1960) saw
the hallways of the small six-story office building at
the intersection of El Segundo and Aviation Boule-
vards in Los Angeles crowded with people at all hours
of the day. Almost every square foot of space in the
building was occupied by new Aerospace personnel,
former Space Technology Laboratories (STL) people

49
william yengst

(transferring to new ownership as Thompson-Ramo-


Wooldridge Corporation), plus USAF officers orga-
nizing and directing programs. It was a common joke
that no one knew for sure who they worked for and
one’s office was often reassigned from day-to-day as
organizations expanded or changed roles.
Dr. Charles Sarture hired me to work in the Guid-
ance Systems Department of Aerospace. My office
mates (Ken Steffan, Jim Alder, and Bernard Katz,
formerly of STL) were already loaded with guidance
projects responding to USAF questions. Ken was
preparing orbital mechanics equations for satellite
rendezvous and station-keeping while Jim and Ber-
nie were preparing a Minuteman-I accuracy analyses
model. The first half-dozen Minuteman-I missiles
had badly missed their Kwajalein lagoon aimpoints,
and the USAF/SAC wanted an urgent explanation of
the cause and what could be done. Issues uncovered
by their analyses concluded: (1) Kwajalein’s location
on world maps was wrong, (2) existing missile guid-
ance error models ignored some important instru-
ment parameters, (3) a few instruments were drifting
more than laboratory tests predicted, and (4) reentry
vehicle errors were larger than anticipated. It was rec-
ommended that in addition to refining the models,
resurveying the island location, and tightening instru-
ment design specifications, the idea of adding termi-
nal guidance updates should be considered.
My assignment was to evaluate two advanced,
map-matching, terminal guidance concepts for cor-
recting ballistic missile accuracy. Goodyear Aerospace
Corporation (Akron, Ohio) invented “Pinpoint,” a
system that scanned an area of terrain with radar and

50
lightning bolts

correlated the image with a stored reference map of


the target. An offspring of this concept, known as
RADar Area Guidance (RADAG), was developed in
the early 1980s for the Pershing II maneuvering reentry
vehicle. Ling-Temco-Vought (LTV ) of Dallas (Texas)
invented a competitive system, TERrain COntour
Matching (TERCOM). It used a radar altimeter to
measure terrain elevation variations approaching the
target and compared height variations with pre-stored
profile maps. Sixteen years later ( January 1977), the
Navy selected TERCOM for guidance of the Toma-
hawk cruise missile.
As a first step in my evaluation, I traveled to Akron
to learn about Pinpoint and Dallas to be briefed on
TERCOM. I was favorably impressed by the theo-
retical analyses and simulated presentations made by
both companies concerning their concepts. Both con-
cepts promised guidance position location fixes with
accuracies of a few hundred feet (i.e., a factor of 15 to
20 times better than the predicted CEP of Minute-
man-I). Unfortunately, their radar systems relied on
C-band (500 to 1,000 megahertz) or X-band (7 to 12.5
gigahertz) frequencies to scan the earth’s surface and
measure terrain reflectivity or elevation contours from
50,000 to 100,000 feet altitude. This required two-way
radio transmission from the reentry vehicle through
ionized plasma layers that form on the vehicle’s sur-
face at hypersonic speeds.
I could find no Aerospace (or TRW ) experts who
could predict the thickness and attenuation properties
of an RV’s plasma layer during reentry. Estimates ran
from a few millimeters to more than an inch thick-
ness and attenuation levels that varied from almost

51
william yengst

none to full “blackout,” depending on vehicle-heat-


shield-chemical-ablation “ionization products,” and
vehicle attitude changes during flight. Some experts
predicted radar signals would blackout or be badly
distorted. Others claimed there would be little or no
attenuation. I saw no alternative but to recommend
experimental tests be made on the plasma environ-
ment before attempting to develop a system.

San Bernardino Opportunity


Aerospace Corporation general offices were to remain
in El Segundo; however, there was too little space at
the small facility for the expanding company, TRW,
and Air Force offices. Hence, in March 1962, the
Ballistic Systems Division (BSD) of the Air Force
moved to Norton AFB in San Bernardino (Califor-
nia). Because BSD was Aerospace’s primary customer,
Dr. Getting announced on 30 July 1962 that a branch
office of Aerospace would be built near the entrance to
the air base. Dr. Ernst H. Krause was named General
Manager of the San Bernardino Operation, and Dr.
Walter Brewer Jr. was his Vice President and Assis-
tant General Manager. They moved to San Bernardino
in June 1962 amid predictions that the new facilities
would be open for operation by 1963.1 In the mean-
time, Walt Brewer set out to hire people and to form
an organization.
Within days after my terminal guidance system
evaluation report was complete, I received a tele-
phone call from Walt asking me to have lunch with
him at the Airport Inn on Aviation Boulevard (El

52
lightning bolts

Segundo), near the southern runway of Los Ange-


les International Airport. He gave me no clues what
the subject would be but when we met, he asked if I
would consider moving to San Bernardino to man-
age a new maneuvering reentry vehicle program office.
I explained that I knew nothing about aerodynam-
ics, little about thermodynamics, and had no experi-
ence as a program manager. Despite these concerns,
he explained that I had experience with key elements
of the program and did not have to know everything
about the subject. He assured me that Wesley M.
Mann, Jr. and Richard G. Sheffer would join the staff,
and both men were excellent aerodynamics engineers.
We talked for an hour, after which he insisted that I
come to Norton AFB the following day to meet Dr.
Krause and others working on the programs.
My trip to San Bernardino was not overly inspir-
ing. The weather was stiflingly hot (over 950 F with
high humidity), the Aerospace complex at South
Tippecanoe Avenue and Mill Street was a dusty con-
struction site, and the temporary Norton AFB office
buildings were worse than those at El Segundo with
peeling pea green paint, no air conditioning, and less
space per person. However, the place was a ‘bee-hive’
of activity. Dr. Krause was a pleasant, quiet man, who
gave me the impression of a stern German university
professor. I was introduced to Dr. Seth W. Booth,
General Manager of the Reentry Systems Division,
and Dr. Kenneth Pearce, Director of the Penetration
Aids Group (my intended boss).1
Seth Booth was a grandfatherly engineer deeply
involved in managing the Atlantic Research Corpora-
tion development of the Athena missile system. Athena

53
william yengst

was a special, four-stage, workhorse test missile that


would be launched from a remote Green River (Utah)
site with test vehicles into White Sands Missile Range
(WSMR) at Alamogordo (New Mexico), 480 miles
south (see Chapter 5). It attempted to simulate the
reentry conditions of intercontinental ballistic missile
systems.
Ken Pearce was a smiling, slender, bald-headed,
delightful physicist, who immediately pulled me into
a discussion with Dr. Robert J. Whalen, an impos-
ing, tall, heavy-set man with joyful expressions. They
were seriously discussing thermodynamics problems
concerning the Minuteman RV nose tip design, which
was to be flight tested within several days. From the
urgency of their discussion and hurried exit to see
USAF, Colonel Kenneth W. Schultz, the Minuteman
Mark 12 RV Program Director, I acquired a sense of
the vibrancy and enthusiasm of the people and orga-
nization. About two years later, Bob Whalen left
Aerospace to become General Manager of Martin-
Marietta Corporation (Orlando) and lead develop-
ment of the Pershing-II missile system.
Within hours, I learned the Pentagon’s Director
of Defense Research and Engineering (DDR&E)
was planning to investigate and develop all manner
of technologies needed to implement reentry vehicles
for effectively delivering nuclear warheads. Between
50 and 100 small projects or experiments were being
managed by the Air Force as a national program to
support the Strategic Air Command (SAC), Navy
and Army. For administrative purposes, the programs
were combined under the heading ABRES (Advanced
Ballistic REentry Systems). The Air Force selected

54
lightning bolts

Aerospace Corporation to perform General Systems


Engineering and Technical Direction (GSE/TD) of
the private contractors that were involved in develop-
ing the various ABRES technologies.
Aerospace organized its Reentry Systems Divi-
sion to perform its tasks by setting up program offices
devoted to the following technologies:1

• Penetration Aids: Metallic-chaff dispens-


ers to be used in space to hide RVs from
enemy radars. Small, low weight, inflat-
able decoys, which acted like actual war-
head RVs to confuse enemy defensive
radars and optical tracking sensors.
• Reentry Vehicle Signatures: Measure-
ments of RV radar, infrared, and optical
signatures that might be detected and
attacked by enemy defensive systems.
Special efforts were made to reduce all
observable signatures of RVs by selection
of materials, contouring vehicle shapes,
and applying stealth concepts.
• Materials Technologies: Tests on low-
weight, heat-shield materials, survivable
nose tip designs, light weight and high
strength structures, and chemical treat-
ments to reduce reentry ionization and
plasma effects.

55
william yengst

• Warhead Arming and Fusing: More than


a dozen reentry vehicles would be tested
to evaluate altimeters and impact fuses,
plus warhead safety/arming concepts (i.e.,
the Fusing-Arming-Test-Evaluation or
FATE program).

A new maneuvering reentry vehicle (MRV ) office,


known only as “Adaptive Systems,” would be added
to develop and test technologies needed by high-per-
formance delivery vehicles (ballistic and boost-glide).
That was to become my responsibility.
Betty and I talked over Walt’s offer, drove to San
Bernardino over the weekend, and saw a house we
liked in nearby Redlands. Therefore, I accepted the
offer to move with an understanding that much of my
time during the first few months would be spent trav-
eling to support Air Force planning meetings, to learn
what contractors were preparing, and to begin the
job of “GSE/TD.” The job implied we would review
prime contractor designs, perform studies, recom-
mend changes to improve performance, and chances
of success. However, we were always to play a hidden
‘Backseat Role’ to our Air Force Contracting Officer
Technical Representative (COTR) and the prime
contractors. In October 1962, about 150 new employ-
ees met at the temporary AFB office quarters with Dr.
Ivan Getting, Dr. Ernst Krause, and Major General
W. Austin Davis of the Ballistic Systems Division to
officially open the new operation.
On my second trip to San Bernardino (late August
1962), I met Wes Mann, who became a close associate

56
lightning bolts

and life-long friend. Ken Pearce briefly introduced us,


before explaining we were to drive to San Diego by
lunchtime that day to meet Pentagon officials from
DDR&E at the Naval Training Center. Ken was
unable to accompany us, despite the high-level people
we would meet, because of a serious Minuteman prob-
lem that demanded immediate attention. We were
introduced to USAF Captain Mack Mauldin who was
assigned as our COTR, and he would accompany us
to handle any contractual requirements. With virtu-
ally no warning, three of us climbed into my car and
got acquainted with one another during the two-hour
drive to San Diego. Fortunately, Mack Mauldin was
an easy-going, soft-spoken officer who told us what he
knew about the program and that he would handle all
contracts and technical directives we felt were neces-
sary to satisfy program goals.
In San Diego we met Dr. Richard Montgomery,
Deputy to the Director of Defense Research and
Engineering (DDR&E) for strategic systems at the
Pentagon and his assistant, U.S. Navy Captain Robert
H. Wertheim. I shall never forget that meeting, and
in years to come, I enjoyed working closely with both
men on several projects. They were always available
when we had serious problems, wanted to know each
step of our progress, and were instrumental in con-
ducting reviews and obtaining budgetary support.
Bob Wertheim, with experience on nuclear bomb
assembly at Sandia Base (N.M.), moved to the Navy
Strategic Systems Project Office in 1965, where he
became technical director of the Poseidon and Tri-
dent Submarine Launched Ballistic Missile (SLBM)
programs. He was promoted to Vice Admiral during

57
william yengst

the 1980s. Upon retirement from the Navy, he joined


Lockheed Corporation for seven years and Science
Applications International Corporation (SAIC) in
2000, where we worked together again, evaluating
advanced research projects.
The meeting in San Diego took place after lunch in
an empty Naval Training Center cafeteria with five of
us sitting around a white, cloth-covered dining table.
Dick Montgomery and Bob Wertheim described
their ideas concerning maneuvering technologies,
showed us sketches, official papers, discussed contrac-
tors and revealed their plan for a four-year develop-
ment program. In brief, General Electric Company
(GE, Philadelphia) proposed building a ballistic
reentry vehicle that could maneuver during reentry
to evade enemy defenses and achieve high accuracy
with terminal guidance. It would be known as MBRV
(Maneuvering Ballistic Reentry Vehicle). SAC Head-
quarters (Omaha, Nebraska) liked its physical fea-
tures, which were similar to ballistic RVs (i.e., having
shorter length than glide vehicles and possibly solving
Minuteman-I accuracy problems). But, others at SAC
liked BGRV, because of its increased range, higher
payload potential and also solving the accuracy prob-
lem. Both concepts had severe technological develop-
ment problems; therefore, our program was intended
to demonstrate feasibility by four full-scale flight tests
of each concept.
The flight test vehicles would not include provi-
sion for a nuclear warhead but would be highly instru-
mented to measure performance and carry a radar
experiment known as TERSE (TERrain Sensing
Experiment). I was pleased to realize my recommen-

58
lightning bolts

dation for radar experiments had been taken seriously.


The program would be completed in four years to pre-
pare the technology base for an operational weapon
system, depending on emerging Soviet threats.
It is useful to digress briefly to explain differ-
ences in evasion and gliding maneuvers. MBRV was
intended to be an evasion vehicle that during reen-
try could ‘out duel’ enemy interceptor missiles by
performing “jinking maneuvers” (i.e., abrupt turns
or changes in direction within the atmosphere) to
“fake” interceptors out of position and thereby, pre-
vent engagement. Abrupt turns could require MBRV
to experience very high accelerations [e.g., 100 times
the pull of gravity (g)]. The laws of physical dynamics
state the force (F) required to accelerate (a) an object
is proportional to its mass (m) or weight (w), multi-
plied by a constant (k). The constant k can be nor-
malized dimensionally by the acceleration of gravity
(1/g). Hence, Isaac Newton’s well-known relationship
can be written as: F=kma or F=wa/g, where near the
earth, g has a value of 980.665 centimeters per sec-
ond squared (or 32.17 feet/second2). To put this in
perspective, Astronauts experience 3-to-5 times the
force of gravity as their booster rockets accelerate
into space. Fighter pilots, dressed in suitable pres-
sure suits, endure up to 9 times the pull of gravity
(accelerations of 300 feet/second2) for brief periods
during intense combat maneuvers. Air-to-air missiles
designed to shoot-down enemy fighter aircraft are
usually built to survive forces caused by 25-g maneu-
vers. Consequently, to evade enemy anti -ballistic mis-
sile interceptors (capable of 50-g level maneuvers), it
was necessary that MBRV survive 100-g maneuvers,

59
william yengst

far in excess of strength required for ballistic RVs to


survive reentry decelerations. By contrast, BGRV was
intended to be an efficient glide vehicle to fly long
ranges (i.e., avoiding enemy defenses by flying around,
past, or under their coverage) and requiring relatively
low accelerations (e.g., 10–20 g) but resulting in large
radius turns. Throughout this book, I have often used
the terminology “high-g,” when referring to evasion
classes of vehicles and boost-glide, when referring to
glide classes of vehicles.
We were privileged to learn that DDR&E formu-
lated ABRES (approved by Congress on 19 August
1964) and identified its initial budget with about $45
million for MBRV and BGRV through 1968. In addi-
tion, roughly $5 million was earmarked for TERSE
radar experiments. A separate budget was provided to
purchase Atlas-F missile booster rockets to support
the maneuvering vehicle flight tests. The Atlas boost-
ers were retired from operational service, overhauled,
rebuilt, and modified by General Dynamics Convair
(GDC, San Diego), at a cost of $2 to $5 million each
to suit requirements of MBRV and BGRV, respec-
tively. This turned out to be a significant undertaking.
Altogether, $120-$125 million would be spent during
the next four years on maneuvering technology, rep-
resenting nearly half the total ABRES budget. Bud-
get dominance for maneuvering caused stress in other
technology programs developing alternative penetra-
tion aids technologies.

60
lightning bolts

Emerging Soviet Threats


At the time of our program initiation, it was known
the Soviet Union had started development in 1960
of the Galosh Anti-Ballistic Missile (ABM) system.2
This enormous exo-atmospheric intercept missile was
designed to engage and destroy in-coming Atlas or
Minuteman RVs prior to reentry by exploding a nuclear
warhead in their path at about 300,000 feet altitude.
The first test of a Galosh missile was at Sary Shagan
(Kazakhstan) using a new V-1000 rocket booster. It
was displayed in Moscow for the public during the
national May-Day parade in 1964 as shown in Figure
5. Galosh (ABM-1/A-35) would be deployed around
Moscow among several long-range VHF intercept
guidance radar sites. The radars, control system, and
radio communications transmission station became
operational in mid-1966 when the weapon system was
declared fully operational.2
ABRES “Penetration Aids and Maneuvering
Reentry Vehicle Programs,” became important in
providing credibility for the U.S. Mutual Assured
Destruction (MAD) strategic doctrine. It was instru-
mental—in addition to the high cost of defensive
weapon developments—in convincing Soviet lead-
ers and diplomats that a treaty to limit deployment
of anti-ballistic missile systems was attractive. Soviet
defense interceptors could not stop a U.S. ICBM
attack, which used maneuvering reentry vehicles and
penetration aids. Recognizing their vulnerability, the
Soviets signed the Anti-Ballistic Missile (ABM)
Treaty on 26 May 1972. Under the treaty, the U.S. and
Union of Soviet Socialist Republics (USSR) would

61
william yengst

restrict deployment of ABMs to two (later only one)


region of 150 kilometers (93 miles) radius and with no
more than 100 missiles at the launch sites.2 The Sovi-
ets chose to defend their capital, Moscow; whereas,
the U.S. elected to protect its strategic Minuteman
ICBM bases located in North Dakota.
Although details of our program were classified
(Secret Restricted Data), it gave us a sense of pride
knowing that BGRV would glide at hypersonic speeds
and altitudes below the reach of Galosh missile inter-
cepts, including their nuclear weapon effects. Likewise,
MBRV was capable of performing maneuvers up to
100 times the pull of gravity (100 g), and could reenter
the atmosphere away from defense sites, maneuver to
destroy either a target or defensive facilities, such as
command and control center or radars. Analytic stud-
ies showed that an incoming, intercontinental-range
maneuvering vehicle had more energy and could out-
duel any known terminal defense interceptors at the
time.

Figure 5. Galosh ABM-1/A-35 missile


in Moscow (1 May 1964)

62
lightning bolts

The ABM treaty did not end the “Cold War” arms
competition between the U.S. and USSR. The Soviet
Union realized the weakness in its Galosh defense
system and began developing a more capable two-
tier interception system by the mid-1980s. The Galosh
defense system would continue to serve as the upper
defense tier but by the 1980s, a new lower tier of high-
performance endo-atmospheric SA-10 interceptor
missiles would back up the silo-launched Galosh inter-
ceptors. The upgraded system also replaced above-
ground battle-management radars with hardened-
phase-array guidance radars.3 These Soviet defenses
spurred our programs beginning in 1968 to study and
test advanced MRV concepts, including jet-reaction
control (see Chapter 5), low-level-run-in (LLRI)
and low-angle reentry (LAR) ballistic RVs for flying
under radar horizons.
In the fall of 1962, Soviet threats, including the
Cuban Missile crisis, caused the U.S. to place high
priority on bringing Minuteman-I to fully opera-
tional status at the earliest possible date. The USAF/
BMD Minuteman program office, also located at
Norton AFB, caused TRW (the system integration
contractor) to open a sizable office about two blocks
west of Aerospace Corporation on Mill Street. Aero-
space was responsible for directing the development
of the reentry systems for Minuteman-I (Mark IV
with W-56 warhead), Minuteman-II (Mark-11 with
W-56 warhead or W-59 warhead) and Minuteman-III
(Mark-12/12A with the W-62/W-78 warheads).6 Con-
sequently, Aerospace personnel working on the reen-
try vehicles were moved to Norton AFB with their
Air Force counterparts, although large fractions of

63
william yengst

their time were spent at TRW. Ken Pearce and Bob


Whelan, along with other senior scientists working on
penetration aids for Minuteman-II and–III, were not
available for days on end. This left our office with lim-
ited help for organizing work assignments and sup-
porting multiple meetings with contractors, who were
implementing the maneuvering programs.
Our Adaptive Systems Program office was fortu-
nate to acquire four key people by the end of 1962,
about the time we moved into the newly finished
buildings. The first, Richard G. Sheffer (Dick), was
a bright, energetic engineer who took an immediate
liking to the BGRV concept because it challenged
his aerodynamic expertise. Wes Mann was comfort-
able with aerodynamic problems concerning MBRV;
therefore, they began studying the two vehicle designs
with the help of the prime contractors. Ruby Yzagu-
irre, a local secretarial graduate, became our fourth
member. She was a wizard at holding the office on
track, typing our reports, making travel arrange-
ments, handling documents and files security, sched-
uling conference rooms, and answering phones. Ruby
worked with me for seven years, until she retired from
Aerospace to get married. At that time, it was com-
mon practice for Aerospace to employ one secretary
for every three scientists and engineers. A company
newspaper explained that productivity of an office was
proportional to the number of secretaries divided by
the number of professional staff members.
Because surplus Atlas missiles were used to launch
most of the ABRES program test vehicles, Aerospace
maintained a staff of booster experts in its Launch
Vehicles Department, headed by Lee E. Miller. They

64
lightning bolts

managed our many interface meetings with General


Dynamics Convair. Fortunately, William Downey,
a former General Dynamics engineer, joined our
office in December 1963. Bill had previously worked
on Atlas-F missile development and became our pri-
mary booster-interface manager. However, within a
few months, Roy T. Pace, Dale Soulek, and Charles
M. Okerstrom joined us to provide expertise and per-
sonnel to cover the myriad of booster interface issues
between General Dynamics, General Electric, and
McDonnell Aircraft. Early in 1964, Donald Ethering-
ton, an aerodynamics engineer, also joined our staff
and split his efforts between helping Wes and Dick,
plus working on booster interface problems.

Security
From the beginning, security was a constant, overrid-
ing, and nagging discipline. Every scrap of paper, brief-
ing chart, and memorandum dealing with the program
had to be marked (S for “Secret,” C for “Confiden-
tial,” or U for “unclassified”), covered with striped-
color-alert pages. All documents were recorded and
secured in personal safes, whenever they were not in
use or when we left the office. Blackboards on office
walls had to be erased, and wastebaskets emptied
and sent to “burnable waste” at the end of each day.
Telephone conversations were restricted to nontech-
nical subjects. During travel, classified material was
double-wrapped in packages, sealed, and accompa-
nied by courier-authorization papers. Briefings were
made using large (27 inches by 36 inches) white paper

65
william yengst

charts, hand-printed, rolled-up, and stored or carried


in heavy cylindrical tubes and padlocked. We called
the awkward tubes “elephant guns.” Fortunately, by
the late 1960s, view graph slides were invented and
relieved the logistics of carrying charts to briefings
and on travel.
Personnel clearances and “need-to-now” verifica-
tion were required for everyone attending meetings,
a process that usually started days or weeks before
the events. Once, I attended a Los Alamos National
Laboratory (N.M.) conference for the purpose of
meeting personnel who were working on nuclear war-
head designs for our vehicles. Before traveling to the
meeting, I spent several weeks being interviewed and
cleared for Critical Nuclear Warhead Design Infor-
mation (CNWDI). Typically, I arrived several min-
utes early for the meeting at the Los Alamos confer-
ence auditorium and watched with interest as a dozen
security guards armed with radio sensors scanned
the stage, podium, view-graph projection, and audio
equipment for possible “bugs or recording devices.”
After several hundred scientists and engineers from
three national weapon laboratories, selected industry
representatives, and military officers filled the audito-
rium, we were told to take no notes.
I enjoyed the briefings on history of nuclear war-
head developments and background for the W-78
warhead; however, when new design work was to be
presented, guards came to my seat and other selected
attendees and directed us to follow them outside. We
were told CNWDI clearances were not sufficient
“need-to-know” for us to learn about specific design
details. Later, we were allowed to rejoin the meeting

66
lightning bolts

to learn about external warhead considerations. One


Los Alamos scientist explained, “You don’t have to
understand the process of digestion in order to eat!”
We were only shown enough to install the warhead
in the vehicle, protect it, and insure it was properly
delivered.

Guidance System Choices


It fell upon me to manage the guidance systems for
both MBRV and BGRV. Although I had taken grad-
uate-level courses in guidance and control theory, the
subject was still in its infancy. Consequently, I spent
almost every spare moment reading Guidance, an
engineering text on the “Principles of Guided Mis-
sile Design,” by Arthur S. Locke (published in 1956).
It was immediately apparent that the Minuteman-I
guidance system could not perform our missions. The
large NS-17 inertial system, built by Autonetics with
its D-37C computer, was a cylindrical unit 30 inches
in diameter, 24 inches long, and weighed 350 pounds.
More important, it was only qualified for missile
booster maneuvers (i.e., up to 25 times the pull of
gravity or g). We required a smaller and significantly
more rugged system.
I convinced Mack Mauldin to travel to Boston
with me in early November 1962 to visit the Mas-
sachusetts Institute of Technology (MIT), Charles
Stark Draper Laboratory, the acknowledged leader in
making small inertial instruments. The Draper Labs
had invented a new guidance system design known
as FLIMBAL (FLoated Inertial Measurement BALl).

67
william yengst

The design placed all the sensitive accelerometers,


gyros, and their associated electronics inside a self-
contained, floating sphere, hydrostatically centered
in a close fitting outer shell. The “womblike” shell for
the inertial sensors provided full three-axis freedom
of motion during maneuvers up to 100-g. Sadly, we
learned the first Draper Labs gyros, and accelerom-
eters for platform tests were several years and millions
of dollars from becoming a reality.4
We visited Kearfott Corporation (Little Falls,
NJ) on the way home. It was developing the stellar,
inertial-guidance system for the USN Polaris mis-
sile. That design had limited degrees of freedom (two
axis plus part of a third) was qualified up to 25-g
acceleration and weighed hundreds of pounds. We
needed full three-axis (pitch, yaw, and roll) freedom
and 100-g to satisfy reentry maneuver goals. At the
brink of giving up, a young Kearfott engineer—Dr.
Jacques Gansler—suggested we consider a small four-
gimbaled (full, three-axis freedom) platform known
as MINS (Miniature Inertial Navigation System),
which was being designed for aircraft navigation. He
explained that it could be floated in oil to achieve the
desired 100-g capability. The final platform design,
including its instruments and oil filling, would occupy
about one cubic foot and would weigh roughly 50
pounds. Although MINS would not provide the high
accuracy of operational weapon systems, it could be
built and tested within a year, and it would serve the
needs of both MBRV and BGRV.
A contract for a dozen of the small MINS plat-
forms was awarded to Kearfott, six-for-each program
to provide four-flight units plus 2 units for functional

68
lightning bolts

integration and qualification testing. But, the guidance


system needed a digital computer to perform mis-
sile navigation, flight-timing events, and maneuver-
steering calculations. Almost by magic, Robert Palmer
applied for a job in our office near the end of 1962.
Bob was a bright Australian with considerable com-
puter expertise. He became our next hire and was an
easy-going, jovial fellow, who provided us with many
laughs and with an introduction to the world of digi-
tal computers and software. He strongly advocated a
state-of-the-art rugged magnetic memory technology
being developed by Univac that would permit repro-
gramming as software was prepared. It would have
an outside chance of surviving maneuvering environ-
ments up to 100 g, provided it was also floated in oil
to take stress off the components.
In the early winter days of January 1963, Bob, Mack
and I flew to Milwaukee (Wisconsin), to visit the
Sperry-Rand Corporation (Univac Division). Univac
Plant 1 on Sheppard Road and 7th Street was across
the Minnesota River from Fort Snelling, a dreary
environment on the day of our visit. The company
was developing a small Guidance Computer (Model
1824C) for aircraft navigation. The box like prototype
weighed 32 pounds, had an aluminum case (9 x 9 x
15 inches) contained 512–24 bit Destructible Readout
Memory (DRO or RAM) and 3,584–48 bit word Non-
Destructible Memory (NDRO). It could handle up
to 10,750 instructions per second.5 It was not much
by modern standards and at the time, it was not yet
functional, nor qualified for high-g environments. We
convinced Univac to float the computer in oil, and
Mack Mauldin put the company under contract to

69
william yengst

build 12 units, 6 for each program. From that day on,


Bob was our undisputed manager for computer devel-
opment. He attended virtually every vehicle meeting
at both Philadelphia (GE) and Saint Louis (MDAC)
to integrate the computer and software.
The MINS/Univac guidance system had roughly
half the accuracy of the Minuteman-I system but was
not designed for continuous running operation. At a
fraction of the cost of military quality instruments, it
provided several unique features. The system required
only 2.5 cubic feet of space, weighed 30 percent as much
as Minuteman-I guidance and would ultimately per-
form an 80-g maneuver on MBRV and guide BGRV
on 4,500 mile flights. The units consumed relatively
little power and were cooled during flight by mount-
ing them on thick aluminum plates in the vehicles.
During the hectic months from the end of 1964 to
mid-1965, the Minuteman-III Mark-12 Reentry Vehi-
cle program suffered severe developmental difficul-
ties.6 The Mark 12 RV, was being designed and built by
General Electric, the same contractor as MBRV. GE
intended to use a new material (carbon-carbon com-
posite) in a three-dimensional wrap for the Mark-12
nose tip and heat shield. The Air Force’s Arnold Engi-
neering Development Center (Tullahoma, Tennessee)
tested the material in a plasma, arc jet and observed
“surface cracks near the stagnation region of the tip”
and “pitting with surface cracks.”7 Although the cracks
may have occurred during test specimen cool down,
the Air Force was extremely concerned for Mark-12’s
design integrity and ordered us to use the same design
and materials on MBRV to gain early experience on
full-scale flights. This was not a major problem for our

70
lightning bolts

program, since there were few competitive materials


with equal strength and heat insulating capability, but
it locked in a key element of the MBRV design.
Because of the serious nose-tip problem, the Pen-
tagon DDR&E Strategic Systems Office also directed
ABRES to attempt recovery of two of the four MBRV
flight vehicles before impact. They wanted the nose
tips and heat shields back for study. GE promptly
designed a parachute compartment to fit in the base
of MBRV-1 and MBRV-2. The compartment was 20
inches in diameter by 24 inches deep and would carry
three, large parachutes, which could be opened at
12,000 feet altitude to soft land the vehicle in water
with flotation bags, similar to the manned Mercury
capsule.8 The navy would recover the vehicle after
impact in Kwajalein lagoon. The compartment would
also carry an instrumentation telemetry transmitter,
and in the last two vehicles (MBRV-3 and MBRV-4),
it would carry the TERSE radar experiment through
increasingly severe maneuvers. The Air Force was
pleased with this concept, because the compartment
could be modified to carry a warhead if the design was
adopted for operational development at early notice.

Atlas-F Boosters
Early liquid-propelled Atlas missiles were deployed
in semi-buried coffins at dispersed sites and relied on
radio guidance from L-shaped arrangements of trans-
mission towers. The system was known to be vulner-
able to enemy attacks, and the guidance was sensi-
tive to electromagnetic radiation. Between November

71
william yengst

1962 and 1966, an improved generation of Atlas mis-


siles was developed with modified fueling systems and
deployed in underground silos. These Atlas-F missiles
included the first inertial guidance systems, a large
unit weighing over 1,000 pounds but relatively inac-
curate. However, they served as active nuclear weapons
until Minuteman-I became operational in early 1963.9
Because Atlas-F boosters would be used for
both flight test programs, representatives of GE and
McDonnell Aircraft Company (MAC) were invited to
each of the booster interface meetings with Convair.
Most meetings were attended by 25–30 people with
an additional mailing of reports to each organiza-
tion. Because BGRV required the most severe launch
trajectory, a decision was made to simulate the entire
powered flight and missile steering equations between
January 1965 and March 1966 to permit evaluation
of various flight dynamics and environmental con-
ditions.10 This required generation of steering equa-
tions, writing software, and verifying them through a
wide range of parameter values. A similar simulation
already existed at Convair for ballistic missiles, and it
was used for MBRV powered flight.
Bill Downey’s first decision was that the Atlas-F
guidance system would perform all-booster steering
and navigation functions through powered flight and
separation of the experimental vehicles. The MRV
guidance (MINS/Univac) unit would operate in par-
allel from lift-off but would not in any way be used to
steer the booster rockets to avoid computational dif-
ficulties, avoid modification costs, and satisfy Vanden-
berg AFB “Range Safety” concerns about our untested
guidance system. Range Safety also insisted that both

72
lightning bolts

reentry vehicles be designed to carry 6 pounds of liq-


uid explosive (with a separate radio-controlled deto-
nator) to be installed just before launch. That addition
would permit the missile to be destroyed, if it veered
off course. They had never tested a vehicle that could
independently fly thousands of miles after separation.
Booster interface meetings through the early
months of 1966 covered a wide variety of subjects
including the following topics.11

• MRV air conditioning requirements on


the launch pad during booster preparation.
• Launch Missile Service Tower support
strength limits (<4,000 pounds).
• Tower-electrical-cabling shielding and
protection for communications at the pad.
• Atlas booster guidance platform gimbals
limits during powered flight trajectories.
• Arming and safety controls for Range
Safety explosives and destruct signals.
• Interface adapter physical design charac-
teristics and separation mechanics.
• Range tracking and telemetry require-
ments (i.e., antenna patterns, power, and
frequencies) with respect to ground or
ship station locations and equipment.

Atlas-F (SM-65) had a length of 75 feet, base


diameter of 10 feet, fueled weight of 260,000 pounds,
maximum speed of 15,500 miles per hour, typical apo-

73
william yengst

gee of 500 miles, and range of 5,500 nautical miles.9


However, as the first U.S. ICBM, it was an unusual,
extremely light structural design, which featured bal-
loon-fuel tanks made of very thin, stainless steel with
almost no rigid support. The rocket would collapse
under its own weight if not pressurized (5-psia) by
nitrogen in the tanks, even when they had no fuel. The
missile also had an odd staging arrangement in which
two of its engines were discarded, while the third con-
tinued to burn. This made it possible to discard the
balloon tanks in midflight without a weight penalty
for lifting them into orbit. It avoided concerns over
whether the third rocket motor could be ignited in
space.12 These features caused two problems.
(1) The Atlas-F payload (Mark-3 or Mark-4 RVs
with a W-49 warhead, yield of 1.4 megatons) had a
weight of 3,600 pounds.9 This imposed an upper limit
on the weight of our experimental vehicles, which had
to be built from “Off-the-Shelf ” commercially avail-
able components. The interface diameter of the upper-
stage of Atlas was 62 inches. While this was no prob-
lem for the slender BGRV, it placed a limit on the
diameter of MBRV. (2) Our biggest Atlas concern was
that it simply would not hold up to the powered-flight
maneuver required at high-altitude (e.g., 150,000 feet)
to turn into a near horizontal attitude to inject BGRV
on its hypersonic glide trajectory. Although there was
little aerodynamic pressure at those altitudes, heating
would weaken the thin balloon structure before the
missile reached terminal velocity. Bill Downey and
Roy Pace worked with General Dynamics at the San
Diego plant for weeks, studying possible fly-out tra-
jectories for their stress on the missile structure. They

74
lightning bolts

decided the missile’s upper stage skin should be made


thicker, stronger, and covered with phenolic, glass, and
cork insulation to protect it for temperatures up to
700 o Fahrenheit.13

Definition of the Ballistic Coefficient


The ballistic coefficient (BC, often denoted by the
Greek letter beta) is a measure of a projectile’s or
reentry body’s ability to overcome air resistance to its
fight. It is useful to digress briefly to explain the char-
acter of this parameter, since it will be used often in
subsequent chapters. Specifically, BC is defined by the
mathematical expression:

BC=W/CD A

where: W is the weight of the body.


CD is the effective drag coefficient of the body
A is the cross-section area of the body
presented to the air flow (e.g., base area
for an RV )

Obviously, high density and weight materials


travel through the atmosphere much easier and faster
than light objects. Thus, bullets made of lead, copper,
or tungsten travel faster along flatter trajectories than
wooden, rubber, or plastic projectiles.
Very long, pointed, and slender projectiles have
low drag coefficients (e.g., CD values of 0.05 to 0.1)
and travel through the atmosphere with much less air
friction and at higher velocities than blunt bodies or

75
william yengst

spheres (CD~0.5). The smoothness of the body’s sur-


face can dramatically alter the effective drag coeffi-
cient (e.g., dimples on golf balls, fuzz on tennis balls,
and fins or flaps on reentry vehicles increase their
effective drag and slow them down). At the extreme,
the drag coefficient for a flat plate is (CD =1.0), since
it blocks airflow completely.14 Finally, cross-section
area (A) and friction drag presented to the airflow
(as expressed by CD) is a measure of resistance to
movement through the atmosphere, and both have
an inverse relationship to speed. Consequently, if the
drag parameters have small values, the ballistic coef-
ficient increases, and the body moves faster.
With respect to reentry vehicles, low ballistic coef-
ficient designs are characterized by low weight, blunt
bodies having high drag, and large base areas. They
tend to slow down at high altitudes, lose their kinetic
energy quickly, and fall to earth along downward bend-
ing trajectories drawn by gravity and with decreasing
velocity. Because they lose their energy and velocity
quickly at high altitude, they do not experience the
high degree of thermal heating and stress that high
ballistic coefficient bodies endure traveling at hyper-
sonic speeds deep in the atmosphere. By contrast,
high ballistic coefficient designs (characterized by
sharp bodies with low drag and small base areas) tend
to slice through high altitude air, retain their veloc-
ity and kinetic energy to follow nearly straight-line
trajectories from reentry to impact. Therefore, there
is a trade-off between how much energy is retained
for purposes of maneuvering deep in the atmosphere
versus how much stress and heating can be tolerated
by the vehicle structure, heatshield, and nose tip.

76
lightning bolts

To illustrate the above phenomena, consider


examples of reentry vehicles approaching the atmo-
sphere at a velocity of 20,000 feet per second along
a relatively steep (e.g., 60 degrees) trajectory with
respect to vertical. A parametric computer analysis of
this condition for vehicles having ballistic coefficients
of (W/CDA=600 to 2,460 pounds per square foot) is
presented in Figure 6.15 None of the vehicles slow
down very much above 100,000 feet altitude; how-
ever, the design with W/CDA of 600 pounds/ft.2 slows
down to nearly 5,500 feet per second by impact (alti-
tude of zero feet), while the high ballistic coefficient
design (W/CDA=2,460 pounds/ft.2) impacts at 13,000
feet per second. This is a stressful region of design
from a materials survivability standpoint.
Because stress and heating of the RV are propor-
tional to the square of its velocity, slow vehicles expe-
rience several times less thermal and structural stress
during reentry than high ballistic coefficient vehicles.
By contrast, a high ballistic coefficient vehicle retains
several times as much kinetic energy (proportional to
velocity squared), which can be used for maneuver-
ing. For example, the ballistic Mark-2, Mark-3, and
Mark-4 RVs had blunt noses, high drag, and large
base areas (i.e., ballistic coefficients of only several
hundred pounds per square foot) and impacted at sub-
sonic speeds. They could survive with only heat sink or
low-density, ablation materials for thermal protection.

77
william yengst

Figure 6: Variation of velocity versus altitude


for different ballistic coeff icients

Initial MBRV Program Meeting


The GE contract for MBRV was established first, and
within a few weeks (August 1963), the program man-
ager, Raymond L. Welsch, with three staff members
(Frank Spampinato, Jack Cameron, and Fritz Hal-
lowell), came to San Bernardino for an initial meet-
ing concerning the program. They presented early
sketches of the design, which would be steered during
reentry by four aerodynamic flaps mounted around
the vehicle’s base. No photographs of MBRV could
be found; however, two drawings of the vehicle are
shown in Figure 7. The upper design was used for
a B-52H recovery system drop test as described in
Chapter 3, and the lower configuration was used for
the Atlas-F missile flights. A small photograph of

78
lightning bolts

MBRV-1 (i.e., with the flaps covered by the booster


adapter) is shown on its way to the launch pad at the
bottom right of the figure.
The design included two large pitch flaps (18
inches long by 14 inches wide) covered with heat
shield material and mounted with buried hinges at
the top and bottom of the conic base. Two similar size
yaw flaps were mounted on either side of the base, also
with buried hinges. The yaw flaps were split length-
wise, each side being driven by independent actuators
to provide roll control. The vehicle would be allowed
to roll slowly to even out heating and ablation weight
loss during reentry. The flaps including their heat pro-
tection were 1.75 inches thick. Flat elliptical areas were
cut in the vehicle’s heat shield forward of each flap to
permit effective air-flow for steering control.16
We all expressed concerns for aerodynamic heat-
ing and air flow across heat shield joints and around
hinge joints. Finally, four small gas jet nozzles were
positioned on the base between the flaps but outside
the parachute canister perimeter. These jets would
provide exo-atmospheric alignment of the vehicle
prior to reentry. Internally, forward and to the sides
of the parachute canister were three high-pressure gas
tanks to power the hydraulic actuators (4,000 psi),
exo-atmospheric gas jets (3,750 psi), and flotation bal-
loon system (3,000 psi). A high-pressure auxiliary gas
generator was also located between the tanks.

79
william yengst

Figure 7. MBRV baseline design layout (Summer 1964)

The MBRV airframe was 133.75 inches long from


the nose tip to the base (without the flaps), and its
heat shield had a single plug-together joint, keyed,
and machined (tight fit to prevent erosion) at 84
inches from the nose (63 percent of length) to permit
easy construction and maintenance of the guidance
and control subsystems. The heat shield was bonded
to an aluminum shell. A second, strong, inner-alu-
minum frame at 104 inches from the nose (78 per-
cent of length) separated aft control components

80
lightning bolts

from the guidance units (MINS platform and Univac


computer). TERSE antenna, transmitter, and power
supply components were placed in this section for
MBRV-3 and MBRV-4.16 The nose section was filled
with tungsten ballast (about 800 pounds, extending
to roughly 50 inches from the tip) plus electric power
batteries placed forward to keep the center of grav-
ity forward of the aerodynamic center of pressure for
stability purposes.
The initial design weight exceeded 3,300 pounds,
pressing the ability of an Atlas-F booster to deliver
it. The base diameter of 52 inches permitted a small
adapter to cover the tucked-back flaps to match the
Atlas-F booster, which was 60 inches in diameter.
However, Aerospace thermodynamics personnel at
the meeting expressed concerns over detailed designs
of joints, hinges, nose tip, and surprisingly, anticipated
reentry vibration environments. Although dozens of
ballistic RVs had been tested by that time, none had
measured internal vibration during reentry. One RV
had radio-transmitted the signal from an acoustic
sensor during reentry and when played, it sounded
like a freight train.
Robert (Bob) Hammond, GE Mark-12 RV Pro-
gram Manager, attended the meeting and advocated
using a sharp, 9.0-degree half-angle cone for the design,
because the Mark-12 ballistic RV used that angle and
hypersonic aerodynamic lift and drag test data was
available. He noted it would give MBRV similar radar
and infrared observable signatures as the operational
Minuteman Mark 12 ballistic RV. GE aerodynamist,
Robert B. Hobbs, had made a study and wind tunnel
tests of the stability and control effectiveness of the

81
william yengst

cone with trailing flaps at velocities up to Mach 10.17


Unfortunately, decisions to include parachute recov-
ery, commercial guidance unit, TERSE altimeter, and
stronger structure for 100 g maneuvers allowed no
room for weight growth during development.
Wes Mann observed the vehicle’s ballistic coef-
ficient (W/CDA) approached 1,500 pounds per square
foot (i.e., A=14.75 ft.2 with the flaps tucked back,
W>3,300 pounds, and CD ~0.15 based on Mark 12).
The aerodynamic drag coefficient, CD, represented
the combined effects of nose tip stagnation, surface
friction, and base-drag at near zero angle of attack
(assuming the flaps were tucked back). The Mark-12
RV had a W/CDA of 2,050 pounds/ft.2 (i.e., W=811
pounds, CD=0.15, and A=2.64 square feet), and its
nose tip and heat shield were stressed to the point of
failure by high reentry velocity and thermal stress.18
Higher W/CDA would mean faster velocities deep
in the atmosphere and could invite a nose tip failure.
Therefore, we decided to alter the design to provide a
margin of safety.
With only limited analysis, we directed GE
to increase the cone half-angle to 11 degrees. This
increased the drag to (CD~0.20) while keeping the
base diameter fixed. Then, we imposed a weight reduc-
tion effort (aimed at W of no more than 3,100 pounds)
to lower the ballistic coefficient to W/CDA=1,050
pounds per square foot. A few wind tunnel tests were
added to confirm the lift and drag coefficients; how-
ever, aerodynamics experts at both GE and Aerospace
expressed concern that we had sacrificed some high-g
maneuverability.

82
lightning bolts

While these changes improved the design, quali-


fication of components to survive 100-g acceleration
environments meant that we would centrifuge test
all subsystems. This caused no significant redesign
problems; however, the staff studying reentry environ-
ments came up with an analytic estimate that vibra-
tion would be “white noise” (random) having “root-
mean-square” (RMS) amplitude of 0.1 g2/second2.
This was equivalent to attaching a jackhammer to the
vehicle and shaking it and its subsystems for 30 sec-
onds (expected maneuver flight times). Vibration was
to become our greatest single engineering challenge
during the program.

Initial BGRV Program Meeting


The USAF/BSD requested a revised and updated pro-
posal from McDonnell Aircraft Company (MAC) to
incorporate MINS and Univac computer subsystems,
plus Atlas booster support. The revised proposal was
provided in August 1963.19 Following negotiations
between the Air Force and McDonnell, a contract was
signed and preparations for another kick-off meet-
ing were made in San Bernardino a couple of weeks
after the GE meeting. Carl J. Blattner, BGRV Proj-
ect Manager, and Fred Gais, his assistant, plus a few
members of the McDonnell Aircraft Company staff
arrived from St. Louis in September 1963 to provide
us an initial program review.
BGRV was a sophisticated, sleek approach to
maneuvering reentry compared with MBRV’s “brute
force” engineering solution. Its external configuration

83
william yengst

is shown in Figure 8. There was no ablation of its heat-


shield or erosion of the nose tip. The vehicle had a
metallic skin, cooled from within by a water-gel blan-
ket, and a transpiration cooled nose tip. The transpira-
tion nose tip was an advancement, employing a one-
inch diameter porous and sintered Nickel cap through
which a mixture of water and ethylene glycol was
forced by high-pressure gas. The liquid tank carried
roughly 30 pounds of fluid, which diffused through
the cap to flow back over the first 15–20 inches of the
skin surface. It was forced out by gas stored in a sepa-
rate tank to offset stagnation point airflow pressure.
When flying at a small angle-or-attack, the windward
side of the vehicle heats up faster than the leeward
side; therefore, the vehicle would spin throughout
flight to even the heating. Spinning plus transpiration
were relied upon to keep the temperature even.

Figure 8. Boost Glide Reentry Vehicle test


flown from Vandenberg AFB

The BGRV airframe, close to 168 inches (14.0-feet)


long by 28 inches in diameter, had a forward biconic
shape and weight estimate of about 1,850 pounds. Pre-
cise dimensions were not found; therefore, estimates
are based on scaling from the photograph of Figure 8
and crosschecking against an MAC scale desk model.20
After being boosted onto its glide trajectory at an alti-

84
lightning bolts

tude of 120,000 feet and speed in excess of Mach 18,


four small and base-mounted gas jets would spin the
vehicle and establish its initial few-degree angle of
attack to offset gravity. BGRV would roll continuously
(few revolutions per minute) to maintain even heat-
ing of the nose and airframe throughout its 45 minute
flight. Vehicle control was accomplished with 10 inter-
leaved and hinged flare panels (each panel about 6.25
inches wide at the hinge, 2.25 inches wide spacer, and
18 inches long). I recall the hydraulic actuators powered
the panels like segments of an umbrella. As the vehicle
rolled, the flare undulated in and out to maintain its
angle of attack or change attitude when a programmed
trajectory turn was required. Dick Sheffer and support-
ing Aerospace personnel were amazed with the design
and forced to reevaluate preconceived ideas concerning
reentry vehicles.
Heat protection was our primary concern, since
extrapolation of Alpha Draco flight test data suggested
the airframe could reach a “hot soak” condition during
flight near 2,700 °F.21 John A. Gard, McDonnell struc-
tural engineer, anticipated these concerns and proposed
a project to sort out metals, super alloys, and refractory
materials that would be suitable for the skin. Among
candidate materials were nickel (specific density com-
pared to water of 8.9 and melting point of 2,646 degree
Fahrenheit), columbium (often referred to as niobium
with the chemical symbol, Nb [specific density of 8.4,
melting point of 3,542 o F]), and refractory/ceramic
coated metals.22 One year and many tests later, colum-
bium was selected for the outer skin. The slightly bluish
rare metal resembles platinum and occurs in ores with
Tantalum. Although costly, it gave the vehicle finish a

85
william yengst

rich-looking luster. This choice provided a margin of


thermal safety, which was necessary to maintain struc-
tural strength. To further protect internal subsystems,
the vehicle included—inside its relatively thin metallic
skin—a 1-inch thick “water blanket.” The blanket was
composed of a foam material soaked with water and
ethylene glycol to control fluid viscosity, evaporation,
and flow properties. The liquid circulated by the vehicle
spin plus an electric pump.
Finally, a conic vehicle is difficult to package because
its center of gravity (cg) approaches the base, some-
what behind its aerodynamic center-of-pressure (cp).
Unlike a ballistic arrow, which has a heavy mass near
the tip to keep the cg forward and feathers at the rear to
keep the cp far back, a cone filled with uniform density
material tends to be unstable and can flip around or
tumble. BGRV was a biconic shape (i.e., forward cone
half-angle of 9.50 to 36 inches of the length, midsection
frustum with 2.50 half angle), fitted to a 30-inch, long-
cylindrical base containing the control system actua-
tors and flare structure. This design made it possible to
package heavier elements forward to maintain the cg a
few percent forward of the cp (i.e., small positive stabil-
ity margin). A small positive stability margin (distance
between the cg and cp) is desired for good maneu-
verability, since it does not require excessive force to
change the angle of attack. A large stability margin
(e.g., >10 percent of the body length) makes it difficult
to change the vehicle’s direction, which follows its cg
and resists attitude-changing commands due to strong
aerodynamic pressure or forces at the rear (acting like a
ballistic projectile).

86
lightning bolts

Fortunately, anticipated BGRV heating and pres-


sure environments were considerably less severe than
for MBRV. Maximum turns required at most 25-g and
glide trajectories were considerably smoother (with
order of magnitude, less vibration) than ballistic reen-
try. Therefore, qualification tests on MINS inertial and
Univac computer subsystems for MBRV would pro-
vide sufficient confidence in them for BGRV flights.
The internal arrangement of BGRV subsystems inside
the water blanket from nose tip to base was: (1) tran-
spiration water and pressure tank, (2) battery power,
(3) computer, (4) MINS/Univac, (5) instrumentation,
telemetry, and TERSE bay, (6) auxiliary power unit, (7)
flare actuators, and (8) gas jets on the base. In brief,
heavy components were forward, guidance elements
near the center of gravity and auxiliary power-to-pres-
surize, hydraulic-flare actuators in the rear. The instru-
mentation bay was about 36 inches long, a collection
point for thermal sensor wiring from many locations
along the body and flare, guidance data, plus the telem-
etry transmitter to report nearly three-dozen measure-
ments throughout the flight. This bay also carried the
TERSE radar altimeter and was envisioned as location
for a warhead in an operational design.
Having completed the first design reviews for both
MBRV and BGRV, the prime contractors went home
to begin making detailed designs. Within Aerospace,
our office and staff turned to answering a number of
technical questions concerning the flight environments,
desired flight trajectories, Vandenberg AFB Range
Safety questions, launch-pad preparation and count-
down procedures plus reporting program progress to
the USAF/SAC and DoD/DDR&E. Most subsequent

87
william yengst

design meetings were held at the contractor facilities,


since it was easier to move us to where the hardware
was being built and tested at their plants.
As an air force pilot, Mack Mauldin, was required
to put in a number of hours ‘flying-time’ each month
to maintain his qualification and receive flight pay;
therefore, he would often fly a two-engine C-47 cargo
transport from Norton AFB to our meeting destina-
tions. Wes and I rode along with him and other air
force officers on several trips to permit us to plan in
flight and save travel costs. We would sit in the back
cargo bay, on parachute jump seats, and under frigid
conditions, with two or three refueling stops during an
all-day (or all-night) trip across the country. We paced
the length of the cargo bay to keep warm and feared the
engines would flood out when flying through thunder-
storms. Mack was promoted to major during 1965 and
was assigned an assistant, Second Lieutenant James
Baird, to help with BGRV program coordination and
documentation. Later, Jim was assigned responsibility
as COTR for the MARCAS flight-test program.
To avoid jumping back-and-forth between vehicle
concepts, which were developed in parallel, the next
three chapters address MBRV, BGRV, and MARCAS
programs separately.

88
chapter 3

MBRV Development and Test


My first view of the six-story building at 3198 Chest-
nut and Walnut Street (Philadelphia) reminded me
of a decaying warehouse. It was constructed in 1929
with buff-colored-limestone brick, upper stories that
were darkened by soot, and rain in January 1964. We
arrived early for an 8:00 a.m. meeting at the GE
Reentry Systems plant, which was later honored as a
historic “Aerospace” site by the American Institute of
Aeronautics and Astronautics.1 Six of us (Wes Mann,
Bob Palmer, Bill Downey, Dale Soulek, Captain Mack
Mauldin, and myself ) had flown to Philadelphia from
San Bernardino the night before the meeting.
Jostled by roughly 250 employees trudging in from
the railroad-track end of the building parking lot, we
were directed to a counter for badges to verify clear-
ances. Groups of fifteen to twenty workers cued up to

89
william yengst

ride a single cargo elevator to upper floors. The entire


building, once a meat-processing plant, smelled like
a stockyard. Ray Welsch and a few members of his
staff met us and took us to the fourth floor conference
room, where the two-day meetings would be held. The
elevator and narrow concrete corridors echoed with
ringing alarm bells and flashing red lights to warn
walkers of vehicles carrying heavy equipment or mov-
ing materials in the aisles.
Like most MBRV program reviews, over fifty GE
personnel crammed the conference room, intimidat-
ing those of us at the review table. Ray opened the
session with a summary of vehicle design status,
explained changes made since the previous meeting,
and followed with reports concerning dozens of out-
standing “Action Items” from the previous meeting.
My greatest fear was that our change of the cone angle
might have destroyed the chance for high-g maneu-
vers. Fortunately, analysis and wind tunnel tests con-
firmed the drag coefficient was close to (CD=0.2), and
the vehicle would retain sufficient velocity deep into
the atmosphere for 100-g maneuvers. The next day-
and-a-half (working through lunches and late into the
evening), we were presented one briefing after another
by GE personnel on specific studies, progress, and
design options.

Nose Tip and Heat Shield


As directed, the nose tip and heat shield would be the
same as the Mark-12 RV. The sharp tip would have
a half-inch radius of 3D weave carbon-carbon. It

90
lightning bolts

would have to endure stagnation pressures of about


125 atmospheres (one atmosphere is 14.7 pounds per
square inch). The heat shield would be ablative, one-
inch-thick carbon-phenolic (thicker than Mark-12,
because MBRV would involve flights that were more
severe and last nearly twice as long as a ballistic reen-
try). The heat shield alone would weigh nearly 800
pounds, most of which was in the aft one-third of the
vehicle. Although an analysis showed the design only
needed about three-fourth-inch-thick heat shield, we
elected to use a full inch near the nose and hinge loca-
tions to provide a margin of safety.

Aerodynamic Control Effectiveness


Initial reports were presented by A.H. Boudreau, J.F.
Roberts, and R.B. Hobbs on experimental tests of
flap and roll control effectiveness. Following discus-
sions concerning boundary layer heating and air flow
separation caused by the flaps, it was decided a more
extensive study should be performed of the hinge
joints. Results of the that effort were reviewed during
a later meeting in 1965 and helped refine the flap hinge
design.2

Parachute Recovery
One briefing addressed the “Parachute Recovery Sub-
system,” which would be built for only the first two
vehicles and would be tested that summer (1965) at
Point Magu Naval Station in the Santa Barbara chan-
nel off San Nicholas Island. Originally, the plan was

91
william yengst

to conduct a drop test of a simulated (size and weight)


MBRV, using a helicopter. The helicopter would lift the
vehicle to an altitude of about 15,000 feet and drop it,
and the recovery parachute was planned to deploy at
12,000 feet altitude. The simulated vehicle would have
fixed flaps towed out at 30° angles, like dive brakes, as
planned for the end of the missile flights. The drop test
vehicle would carry a small telemetry package with
seven channels of data to record parachute deployment,
beacon lights, radio beacon operation, flotation activa-
tion, and dye markers plus meteorological data.3
MBRV was heavy (3,100 pounds), almost as heavy
as the manned-Mercury reentry capsule (4,260 pounds)
flown in 1961–1963. Parachute technology was under-
stood, but subsystem engineering was new and needed
to be verified. Upon reentry and after its maneuvers,
MBRV would use its flaps as ‘dive-brakes’ to slow down
below Mach 1, and then it would deploy a ballute (bal-
loon-parachute) to pull out three, large parachutes. The
parachutes would lower the entire vehicle to a soft water
landing, and flotation bags would activate to keep it
afloat for up to 72 hrs. A self-destruct feature was added
to prevent the vehicle from falling into Soviet hands.
Unfortunately, after much planning, we learned that no
helicopters at Point Magu Naval Station could handle
the heavy load under dynamic sea conditions. Mack
Mauldin convinced the USAF/SAC to provide a B-52H
bomber for a high-altitude drop test.
A few months later, Mack, Wes, Dale, Ray, and I
met at Point Magu before sunrise one morning to ride
a Navy Landing Support craft out into the Santa Bar-
bara Channel. Riding 4-to-6 feet waves, we waited off
San Nicholas Island to spot the contrails of the B-52H

92
lightning bolts

flying east to west at about 60,000 feet altitude. After


receiving radio communications from the aircraft, we
watched the drop of the simulated vehicle (steel cone
filled with concrete) and followed its trajectory as the
ballute and three orange parachutes deployed. It drifted
down gracefully and impacted in the water with a splash
about half a mile from our craft. Although it appeared
fine as we approached for recovery, one or more flota-
tion bags failed, and at the last minute, it sank out of
sight. The Navy sent divers down, and sonar was used
for several hours; but the simulated vehicle was lost in
water several hundred feet deep. Improvements were
made to flotation components, but no further drop tests
were conducted, because the fixes were simple and test
costs were higher than the USAF anticipated.

Flight Trajectories
A most interesting feature of the early meetings was the
generation of flight trajectories. We agreed the initial
flight should be relatively simple, building to the most
severe maneuver as the last test. Four trajectories were
formulated: (1) mild pull-out or range extension from
a ballistic path, (2) longer pullout range extension fol-
lowed by a dive to target, (3) significant cross-range
maneuver, and (4) maximum 80-g dive short of the bal-
listic aimpoint. Software for these trajectories would be
prepared for the guidance system computer and vehicle
steering. Mack Mauldin revised the contract to establish
a set of bonus financial incentives for GE, if the flight
vehicles correctly passed through three points selected
at different altitudes along each trajectory.

93
william yengst

Depending on flight range (i.e., reentry velocity and


angle conditions), MBRV would have sufficient kinetic
energy and aerodynamic lift to fly roughly 250–300 mi.
downrange and over 150 mi. crossrange in either direc-
tion from its ballistic aimpoint. This led to a heart-
shaped footprint capability dominated by downrange
maneuvers. A sharp pull-down maneuver was envi-
ronmentally most stressful, and its time of flight from
100,000 feet altitude to impact was barely 10 sec. An
illustration of the four flight trajectories and potential
maneuver footprint are shown below.

Figure 9. Representative MBRV trajectories


and maneuver footprints

94
lightning bolts

Jack Cameron, GE’s guidance engineer, and I pre-


sented the guidance system design, after which Bob
Palmer held working sessions with GE and Univac
personnel to establish ground rules for developing
computer software.4 The software would be loaded
into the computer during prelaunch preparation on
the launch pad, using Mylar tapes to avoid perceived
problems of vulnerability of magnetic-tape media.
During flight, the computer would receive vehicle atti-
tude, velocity, and acceleration data from the MINS
platform. Specifically, gyro gimbals angular readouts
and accelerometer readings along each axis would be
processed to determine steering commands to move
the control flaps including roll and out-of-plane turns.
Separate software would handle timing for MBRV/
booster separation, attitude jet operations, TERSE
altimeter activation, and recovery system initiation.

Control Actuators
Actuators that drove the control flaps were an impor-
tant design consideration. Analysis showed that maxi-
mum dynamic pressure during reentry would occur at
about 35,000 feet altitude, where the vehicle was still
moving at a speed in excess of Mach 15 (see Figure
6) through increasingly dense air. In analytic terms,
dynamic pressure (q) is defined as air density (Greek
letter rho, P) times half the vehicle velocity (v) squared
(i.e., q =Pv2/2). It is closely related to kinetic energy
and provides a measure of stress placed on the vehicle’s
structure subject to aerodynamic forces. If we wanted
to make 100-g maneuvers at the highest-dynamic-

95
william yengst

pressure conditions, the pitch flaps would each expe-


rience nearly 60,000 pounds of force, which had to be
offset by actuator thrust. Fortunately, the GE Loco-
motive Division had developed hydraulic actuators
of that capability for use on diesel railroad engines.
One of these locomotive actuators was attached to
each pitch flap and four 30,000 pounds thrust actua-
tors were used to move both pairs of yaw/roll flaps.
In order to power these large, heavy-steel pistons, an
Auxiliary Power Unit (APU) was added to the design.
The APU was a hefty (e.g., roughly 100 pounds) pack-
age about 1.5 cubic feet in volume. It contained a solid
propellant, gas-driven pump to pressurize a 4,000-psi
hydraulic oil supply in a closed cycle of pipes to move
the pistons and push the flaps. MBRV began to look
and feel more like a freight train than an aerodynamic
vehicle.
In the spring of 1964, Otto Frudenberger and
his staff from LTV (located in Dallas) joined our
design review meetings at General Electric. LTV
was developing the TERrain Sensing Experiment
(TERSE) radar altimeter to be carried on MBRV-3
and MBRV-4 plus the last two BGRV flight tests. The
one-kilowatt, C-band radar would be placed in the
empty MBRV Parachute Canister and would feed a
small (few-inches long by 2-inch wide) slot antenna
that was form fit into the mid-vehicle heatshield (see
Figure 7). The radar would be turned on above 150,000
feet altitude and run continuously through reentry to
measure altitude and two-way plasma attenuation
during the planned maneuvers.
By the fall of 1964, GE developed a vehicle
dynamics simulation that allowed analysis of compo-

96
lightning bolts

nent placement, weights, and balances. To our cha-


grin, it showed the overall weight had crept up over
3,300 pounds. But worse, 800 pounds of Tungsten bal-
last was required immediately behind the nose tip to
offset the heavy heat shield, APU, actuators, and flap
weights at the rear. Ballast was needed to keep the
center of gravity forward of the center of pressure for
stability. Thus, a renewed effort was made to reduce
weight by rearranging internal subsystems, reducing
lengths of electrical wiring, hydraulic plumbing, and
shaving thickness of heat protection in areas of less
concern. The Mark 12 RV nose tip design remained
fixed to verify its integrity. All MBRV subcontractors
were encouraged to reduce weight and size of their
subsystems.

The Subtle Time-Constant Revelation


During a fall of 1964 meeting, GE software engineer,
John Trescott, presented the reentry trajectories and
guidance steering equations. He was justifiably proud
that all four-selected trajectories were realistic, and
flight simulation results showed the vehicle would
pass through three test points at the selected altitudes,
along each trajectory as proof the maneuvers were
being accurately performed. However, as John spoke,
Bob Palmer alertly noticed that a nearly straight-
line, flight path from high altitude to impact would
also pass close to the chosen points along each tra-
jectory. In brief, the simulation was very sluggish and
although it satisfied the intended incentive goals, the
vehicle was not guiding like a 100-g system capable

97
william yengst

of correcting flight path errors or out-dueling enemy


interceptors.
The system, “Time-Constant” (i.e., time required
to respond to steering commands), was much to slow
on the order of a few seconds. The Time-Constant of
a complex system (e.g., one that includes a guidance
unit sensing, computer processing, actuators moving
flaps, and aerodynamic attitude changes) to an abrupt
steering command is defined as the time required for
the system to reach 63 percent of its desired position
(i.e., 37 percent below the intended final position).
A complex system typically requires at least three
“time-constants” of flight to steer out errors and eight
or more time-constants of flight for high accuracy.
Virtually all of us agreed that MBRV needed faster
response during the limited reentry flight period to
satisfy its performance goals. I recall Fritz Hallowell, a
large, usually smiling, and highly respected GE engi-
neer (who stuttered slightly when agitated), made an
impassioned speech to revise the design. Fritz, Frank
Spampinato, and Jack Cameron prepared a presen-
tation on the subject for an industry “Penetration-X
Conference,” held at Lockheed Missiles and Space
Company (LMSC) in March 1965.5
The impact of this redesign was significant, involv-
ing several subsystems, but it was necessary to uphold
good control system engineering standards:

Much of the steering software had to be rewrit-


ten to increase amplification (gains), reduce data
processing times and shorten time-responses
of electrical and hydraulic subsystems. Numer-
ous flight simulations were required to reduce

98
lightning bolts

the overall Time Constant to about one sec-


ond. GE applied extensive effort to accomplish
this rework by mid-1965. Subsequent trajectory
simulations showed that MBRV could faithfully
follow the chosen flight paths, including hard
80-g pull down maneuvers, which lasted barely
10 seconds after entering the sensible atmo-
sphere where steering became effective.

Following rework and verification of software, one


could introduce impulse errors in the steering calcula-
tions, and the vehicle would promptly remove them.
However, the development program schedule had
slipped by a few months, because planned subsystem
assembly and ground tests were delayed. Ray Wel-
sch prepared a new master Program Plan in Decem-
ber 1964 (GE Report 64SD5175), but our engineering
problems were not yet over.

The Great Vibration Challenge


At the end of 1964, prototype MINS and Univac
1824C computer subsystems were delivered to Gen-
eral Electric for functional integration testing. GE
used the actual units to replace digital simulations in
its overall analog/digital system simulation. All went
well, until qualification test units arrived in early 1965.
Centrifuge testing of the massive actuators and guid-
ance units to achieve 100-g capabilities was accom-
plished with only minor difficulties. The MINS plat-
form design was modified to stiffen gimbals, improve
bearings and increase electric power to gyros and

99
william yengst

accelerometers. However, the computer proved to be


extremely sensitive and had to be returned to Univac
for rework and remove particles that short-circuited,
integrated, and electronic circuit boards. A few weeks
later, vibration testing began and a whole new set of
problems emerged. The computer Random Access
Memory (RAM) and integrated circuits, supplied to
Univac by Westinghouse (Pittsburgh), broke into par-
ticles under the intense random noise pounding it had
to endure for a half minute.
During the summer of 1965, Larry Bolton of Uni-
vac and Bob Palmer spent nearly every day at West-
inghouse, flying back to Milwaukee and over to Phila-
delphia to rework circuit designs, removing manufac-
turing particles, refilling with oil, verifying functional
performance, and repeating vibration tests.6 One fail-
ure followed another until we decided that each flight
computer should be subjected to the same qualifica-
tion tests. Without the computer, the system would
not fly. Aerospace vibration experts spent hours with
me reviewing the severe qualification specification to
see if it could be lowered. By that time, the four flight
units were already in production:

As the first scheduled flight date approached and


none of the flight or test verification comput-
ers had passed the vibration tests, Bob Palmer,
in desperation, suggested an alternative. Each
flight computer would be vibrated, functionally
mapped to see which parts of memory remained
undamaged, then the software would be rewrit-
ten to avoid using the damaged sections of the
machine. We proceeded with this risky approach

100
lightning bolts

and when a computer failed, we switched to


another flight unit, including its unique soft-
ware. In this manner, we were forced to make
multiple computer changes prior to each flight
test.

I spent Thanksgiving holiday weekend of 1965 in


the basement of 3198 Chestnut Street (Philadelphia),
observing vibration testing of the MBRV-1 airframe.
Ray Welsch and I watched as the 800 pound Tung-
sten-ballast casting was installed in the nose before
dummy guidance, and other subsystems were bolted
inside the airframe to make the test realistic. A special
fitting attached to the Parachute Canister was con-
nected to an enormous several-feet-diameter, elec-
trical shaker that was bolted to the concrete floor.
The vehicle with vibration sensors attached at many
locations was suspended from the rafters in a sling,
and the shaker was activated. The 3,100-pound vehi-
cle shuddered so fast it appeared blurry. Fortunately,
the holiday weekend had been selected for the test to
avoid frightening employees, since the shaker caused
deafening noise and shook the entire building. To our
relief, the airframe passed with no observable breaks
or cracks.

Flight Test Program


Vandenberg AFB launch pads, 576A1 and 576A2
(latitude 34.7769 degrees north and 120.6017 degrees
west), were built in the early 1960s to launch Atlas
missiles carrying target RVs for Nike Zeus Anti-

101
william yengst

Ballistic Missile tests being conducted at Kwajalein


Island. The first launch was on 26 October 1962, and
the pads were upgraded for use by ABRES during
1965.7 Pad 576A2, shown on the right side of Figure
10, was used for all the MBRV tests plus BGRV-1. Pad
576A1 (shown in the left of the picture) was used for
BGRV-2 through BGRV-4. The launch control center
was located between the pads.
Wes, Mack, and I drove to Vandenberg AFB
in late July 1966 to witness final preflight testing of
MBRV-1, before its scheduled launch near the end of
the month. Ray and his GE team were proud to dis-
play the beautifully painted vehicle (white refrigerator
enamel with a blue GE logo), which had functionally
passed all preflight tests. Two days later, Dr. Earnst
Krause, Vice President of Aerospace Corporation,
arrived at Vandenberg to see the test arrangement. By
that time, MBRV-1 was at the launch pad, mated with
Atlas (149F), standing beside a multistory gantry. We
all climbed the 50 foot gantry to look inside the vehi-
cle and to answer questions about the intended flight.
To our great surprise, Dr. Krause’s face turned ashen,
then red with anger. “Why had we allowed the vehicle
to be painted?” The paint would burn off during reen-
try and chemically destroy the vehicle’s optical and
radar signatures. This would cause increased plasma
and was totally counter to ABRES program goals of
suppressing all signatures.

102
lightning bolts

Figure 10. Vandenberg AFB Launch Pads


576A1 (left) and 576A2 (right)

We had no option but to take the vehicle down


and back to its hangar. GE managed to acquire crushed
walnut shells and preceded to “nut-shell blast” all the
paint off, down to the bare heatshield. The launch date
was moved to 8 August 1966, and once again, we went
through all the preflight preparation and mating with
the Atlas. That evening, Ray, Wes, Mack, and I were
permitted to stand in the back of the blockhouse near
the launch pad to witness the countdown and launch.
We observed Range Safety personnel pouring 6 pounds
of liquid explosive into a small container in the rear of
the vehicle and arming it with an independent radio
transmitter. The countdown was normal, but unfor-
tunately, the flight was held up for a couple of hours,
while a Union Pacific Railroad train, carrying passen-
gers, passed under the flight path along the coast. Dur-
ing the waiting period, we learned from downrange

103
william yengst

radar stations that three Soviet trawlers (intelligence-


gathering, radar ships) had moved close to Kwajalein
lagoon to observe our test.
At 9:47 p.m. PST (Pacific Standard Time or
17:47 Greenwich Mean Time-GMT), the countdown
resumed and MBRV-1 was launched as shown in Fig-
ure 11.8 The USAF, space agencies, astronomical, and
many international time reports are made in terms
of Greenwich Mean Time (i.e., local time at the zero
degree longitude meridian through the Royal Observa-
tory at Greenwich, England). Vandenberg AFB is eight
time zones or hours behind GMT, and therefore, PST
time will also be provided in discussing tests to give
the reader a proper understanding of the time of day in
addition to the 2400-hour day as used by the military.

Figure 11. MBRV-1 during liftoff on


its 8 August 1966 flight test

104
lightning bolts

We were all astonished a few minutes after liftoff


when the Range Safety officer, who was carefully mon-
itoring a two-axis plotter of the trajectory, ordered,
“Destruct!” As I recall, the booster successfully com-
pleted its dual, main engine burn and was finishing
its sustainer engine burn when the decision was made
to terminate the flight. Within moments, the vehicle
exploded and threw debris far out to sea. To our hor-
ror, the first $10 million vehicle was lost. Vandenberg
AFB records were not precise but showed the Atlas-F
booster was off course at an altitude of 310 miles and
might have landed among the Hawaiian Islands.9
Nearly two months and many man-years of effort
were expended studying the failure, since MBRV-1
was guiding and operating as intended at the time of
destruction. Our office reported the failure to USAF/
SAC and DoD/DDR&E. September ended the gov-
ernment fiscal year, and we were required to submit
detailed ABRES budget plans for the next year—not
a good time for a test failure. The program was also
scheduled for review by the Defense Science Board
(DSB), which was chaired by Dr. Edward Teller and
included board members, Dr. Charles Stark Draper,
Albert L. Latter, and brother Richard Latter, plus
several well-known, industry scientists. They gave
us advice concerning the tests, urged us to continue,
but insisted on seeing operational-weapon-system
designs.
Preparation for the second flight test started
before the end of 1966. This time we were not so
closely monitored, and preflight activities went more
efficiently. The launch date was 16 March 1967 using
Atlas (151F).9 On that day, the launch was delayed

105
william yengst

twice by Range Safety but lifted off at 9:56 p.m. PST


(or 17:56 GMT) from the same pad (576A2).7 Van-
denberg records indicate the booster reached an apo-
gee of 800 miles (consistent with a ballistic trajectory
to Kwajalein) and deployed MBRV-2 on its intended
trajectory. The reentry “pullout” maneuver was accom-
plished (i.e., achieving roughly 20 g near 35,000 foot
altitude). However, the recovery system failed, and the
vehicle was lost, including our last chance to obtain
a nose tip and heatshield after reentry. The flight was
declared successful by the Air Force, although little
information was released concerning the maneuver.
Aerospace Corporation referred to the test as, “The
first vehicle capable of intentional deviation from a
ballistic trajectory during reentry was successfully
flown from Vandenberg AFB to Kwajalein.”10
Wes Mann, Dale Soulek, and I drove from San
Bernardino to Lompoc, California, on 27 July 1967,
the night before the scheduled flight of MBRV-3. We
stayed at a motel near the airbase and were pleased to
find several GE and four LTV personnel at the same
inn. They had been there for a week, replacing the
parachute subsystem with the TERSE radar altim-
eter and performing final tests while waiting for one
scheduled flight time-assignment after another. They
had lounged around the motel, watched television,
consumed a small keg of beer, and played endless card
games. None were feeling much pain, but a few had
grown beards and looked tired of the delays.
The flight would be made using Atlas (150F),
and we learned that considerable attention had been
expended by General Dynamics to troubleshoot the
booster. Forty eight minutes after midnight on 29 July

106
lightning bolts

1967 (0:48 a.m., PST or 8:48 a.m., GMT), the Atlas


lifted MBRV-3 perfectly from pad 576A2 and sepa-
rated it at an altitude of 800 miles.4, 7 Radio commu-
nications from the radar-tracking center on Kwaja-
lein Island (5,180 miles downrange) reported picking
up the vehicle reentry about 18 minutes after lift-off.
They tracked the flight path and recorded the vehi-
cle’s telemetry signals. To our pleasure, the TERSE
altimeter worked before reentry but was attenuated by
plasma for only a few seconds at very high altitude (>
50,000 feet). Two-way radio transmission through the
plasma was an achievement.
The infrared imagery of the trajectory, that we
received about a week later, showed the vehicle left a
white streak of hot, ionized air and ablation particles
from over 150,000 foot altitude down to roughly 50,000
feet, where it veered out of plane towards the east for
several miles before pitching downward to impact in a
northern sector of the lagoon. The entire maneuver could
be seen like a white ink stroke on black paper with several
stars in the background. A navy tracking ship reported
picking up the blast of the Range Safety explosive on
its sonar equipment seconds after the expected impact
time. We had a complete success with a maneuver that
exceeded 25 g. Soviet trawlers within about 50 miles of
the lagoon undoubtedly observed the test.
With considerable relief and renewed enthusiasm,
we began preparation for the last vehicle, MBRV-4. Its
flight would be in mid-October on Atlas (118F), and it
would also carry a TERSE-radar altimeter. However,
being our final shot, we decided to perform the severe
80-g pull-down maneuver to stress the nose tip, heat-
shield, and radar plasma environment under extreme

107
william yengst

conditions. It was after dark on 14 October 1967, when


the final flight lifted off. Again, the booster performed
perfectly and reached an apogee of 800 miles.8, 9 To our
joy, the vehicle was tracked by Kwajalein radar as it dove
to nearly a vertical attitude below 50,000 foot altitude.
The one-kilowatt, TERSE altimeter blacked out briefly
but resumed operation at lower altitudes (<25,000 feet).
There was a radio report that the nose tip may have
partially shredded at roughly 35,000 foot altitude (debris
was observed); however, I was not concerned, know-
ing the vehicle could easily fly to impact on its Tung-
sten ballast. Each report of the flight success echoed
through the Vandenberg blockhouse, as we all con-
gratulated one another. Ray Welsch (GE) and Otto
Frudenberger (LTV) lost their voices yelling approval,
and Mack Mauldin was all grins and smiles. Once again,
we received an infrared image of the reentry trajectory
about a week later. It clearly showed the white streak as
MBRV-4 reentered the atmosphere along a ballistic path
then turned abruptly downward. Telemetry confirmed
the guidance accelerometers peaked close to 80 g when
it passed through maximum dynamic pressure around
35,000 foot altitude.
Following our final project review, Ray and the GE
managers honored each of us at a dinner, during which
we were presented small silver MBRV ‘tie-tacks.’ A few
weeks later, we were told they must be returned because
the flap design was considered ‘Classified’ by USAF/
SAC. The MBRV Program was formally completed in
February 1968. Ray and several members of his team
wrote classified reports on the program including analy-
sis of the flight test results.11

108
lightning bolts

Follow-On Developments
Late in 1967, I was promoted to Assistant Group
Director of the Reentry Systems Division of Aero-
space, and Wes Mann took over responsibility for the
Adaptive Systems Program Office. About that time,
Major Mack Mauldin retired from the air force, and
with his wife Anne and their five children, moved to
Arkansas to manage the family farm. Dr. Richard A.
Hartunian was named Director of the Reentry Sys-
tems Group, and with his help, we formulated a new
Advanced Maneuvering Reentry Vehicle (AMaRV )
program.12 AMaRV would use a new, small, nuclear-
hard, and gimbaled-inertial platform for guidance
with the MIT Draper Laboratory designed instru-
ments that were an order of magnitude more accu-
rate than those used in MINS. AMaRV would be the
prototype of an operational maneuvering vehicle for
Minuteman III and MX.

109
chapter 4

BGRV Development and Test


Our group (Dick Sheffer, Wes Mann, Bill Downey,
Bob Palmer, Mack Mauldin, and I) arrived at Lambert
International Airport (Northwest St. Louis) in early
February 1964 for our first program review of BGRV.
I was impressed with the size of Lambert field and
the adjacent McDonnell Aircraft complex of facilities,
including a series of hangars and an office building
sitting on a hill overlooking the 2,000 acre site. This
was the airfield where Colonel Charles Lindbergh
took off in the Spirit of Saint Louis on 12 May 1927 for
his flight to New York before beginning his epic solo
Atlantic Ocean crossing to Paris. The site became the
first municipally owned airport in the United States,
and McDonnell Aircraft purchased the plant and
hangars in the early 1950s.

111
william yengst

Carl J. Blattner, Fred Gais, and other staff mem-


bers met us at the visitor’s entrance to the huge office
building. After checking through security, they took
us on a walking tour through the aircraft production
factory on the way to the conference room. We saw
dozens of Phantom-II F-4C fighter planes being con-
structed by hundreds of workers. Unfinished, green-
sealer-painted fuselages were lined up as far as the
eye could see, while electric-transport carts delivered
large prefabricated modules from side isles. The multi-
story hangar echoed with the stutter of rivet guns and
arc-welding equipment, as men yelled directions and
waved red-warning flashlights to alert us as we passed.
I watched while heavy steel landing gear, designed for
navy aircraft carrier landings, was installed in one air-
frame. Within months, these fighters would be on car-
riers off the coast of Vietnam. Over 3,000 military air-
craft were built there by the end of the Vietnam War.1
Another section of the hangar provided displays of
the NASA Mercury space-flight-recovery capsule and
nearby, a mock-up of the planned Gemini spacecraft.

Refined Design
The BGRV project review was conducted in much the
same manner as MBRV. We were seated in a large
conference room and around a mahogany table, while
Carl introduced the staff and presented the agenda.
The first item on the agenda was a brief but impres-
sive talk by David S. Lewis Jr., president of McDon-
nell Aircraft. He covered the history of hypersonic-
glide vehicles through the Alpha Draco project and

112
lightning bolts

motioned for the aging Walter Dornberger to step


forward, to say a few words, and to wish the program
success. Dornberger was sixty-nine years old at the
time, thin, bald, had a small, gray moustache, sharp
smiling eyes and spoke briefly with a distinct German
accent. I was sorry there wasn’t time to discuss the his-
tory and future goals of BGRV development with him
and David Lewis before they quietly left the meeting.
Since we had seen the vehicle design in San Ber-
nardino, it had matured in definition of internal sub-
systems, although the external configuration looked
the same. A computer simulation showed the overall
weight had grown to almost 2,100 pounds, and some
elements had been moved to maintain balance and a
small, aerodynamic-stability margin. A central bay,
which contained the telemetry subsystem and TERSE
experiment during tests, had gained a hundred pounds
of equipment and electrical wiring. In an operational
design, it would carry the warhead. Joe Deans and Ray
Haynes presented the methods and plans for measur-
ing vehicle performance.2

Thermal Protection System


Heat protection and the nose tip design consumed the
majority of time during the first meeting. Selection of
Columbium for the skin, water blanket, and transpira-
tion nose tip designs were discussed for several hours.
It was decided that a thermal analysis simulation
should be prepared to predict hot spots and compare
with flight-test, thermal-sensor data. Thermal sensors
would be mounted almost every foot along the length

113
william yengst

of the vehicle’s skin and on flare panel segments. J.


P. Rhudy and J. D. Magan Jr. presented the transpi-
ration-cooled, nose-tip design and provided results
of heat-transfer tests to confirm its operation.3 They
showed us porous-nickel samples that had been fabri-
cated and demonstrated the pressure needed to blow
water through the material. J. C. Uselton provided an
analysis of pressure distributions over the nose tip and
vehicle body. These results would be verified by tests
later in 1964.4

Aerodynamic Performance
Dick Sheffer was particularly concerned with the aero-
dynamic lift and drag characteristics of the airframe.
A number of subscale-hypersonic-wind-tunnel tests
were scheduled to map the lift (L), drag (D), coeffi-
cient-of-drag (CD), and L/D as a function of velocity
and angle of attack. Results would be reconciled with
Alpha Draco test measurements.
Paul Czysz, who worked for McDonnell Aircraft
(1963–1966), described in his interview concerning the
advanced Aurora aircraft, having performed tests on
a secret, glide vehicle design in the Wright Patterson
AFB hypersonic impulse tunnel up to speeds of Mach
12. He described the advanced design as having a
potential range of 4,000 nautical miles (4,600 statute
miles).5 This was BGRV.
Subsequent tests showed L/D was about 2.5
at trans-sonic speeds, rising to a peak value of 3.7
between Mach-5 and Mach-7, but falling slightly
to around 3.5 as speeds approached Mach 12. These

114
lightning bolts

results promised relatively efficient glide capabilities;


however, they were sensitive to angle of attack. Drag
increased rapidly when the flare control vanes were
extended. The vehicle was most efficient when flying
at only a few degrees angle of attack to offset gravity.
Higher velocity (Mach 12–18) data would have to wait
until the full-scale, Atlas missile flights.
I recall computing the ballistic coefficient for
BGRV as (W/CDA~50,000 pounds per square foot).
This was meaningless since the vehicle did not fly a
ballistic trajectory, but with a base area (A=4.12 square
feet) and weight (W=2,100 pounds), it implied the
overall drag coefficient was about (CD=0.01). This is
realistic for a slender projectile with no base flare.
However, I was unprepared to see and hear movies
of tests in which shock waves were created at hyper-
sonic speeds. At those velocities, considerable energy
is expended as compressive shockwave cones form at
the nose and tail of a projectile. They represent a jump
in air pressure, draining kinetic energy from the pro-
jectile and increasing its effective drag.6 BGRV would
be seen as a streaking meteor across the sky and would
be heard as multiple sonic booms.
Fortunately, reentry and glide phases of flight
would involve much less severe environments than a
ballistic vehicle; therefore, qualification of guidance,
flare actuators, auxiliary power supply, and other com-
ponents would only have to be tested to 25 g and half
the vibration levels of MBRV. However, the Univac
1824C computer remained a problem in 1966, when it
failed to pass vibration tests. By that time, Bob Palmer
and the McDonnell staff knew how to work around the
difficulties based on experience gained with MBRV.

115
william yengst

Trajectory Simulations
The approach described in Chapter 3 for MBRV was
also used for Atlas-F booster and independent BGRV
guidance. Bob Palmer insisted on software simula-
tions to evaluate steering equations and handle special
events like booster separation, telemetry activation,
and TERSE altimeter initiation. The flight simulation
included vehicle aerodynamics data, and dozens of
analyses were prepared of energy-management, glide-
phase steering to establish the best flight regimes.
Parameters, such as booster separation attitudes and
angles, injection velocities into the glide, preferred
glide-altitudes, and terminal dive maneuvers were
studied.
A typical BGRV flight trajectory and footprint
is illustrated in Figure 12. In profile, the Atlas missile
would fly to an altitude of about 130,000 feet, turn to
horizontal flight to gain speed, and then would sepa-
rate BGRV on a glide-path at over Mach 15 toward
110,000 foot altitude. The vehicle would spin up, estab-
lish a lifting angle of attack to offset gravity, and glide
with little change in attitude, managing its energy to
reach the desired destination. In the target area, pro-
grammed steering equations would take over and turn
the vehicle into a terminal dive and maneuver to the
intended aimpoint.

116
lightning bolts

Figure 12. BGRV powered flight trajectory


and flight-range footprint

Because of its enormous kinetic energy and high


L/D, BGRV could glide downrange nearly 5,000 miles
or turn out of plane cross range about 2,000 miles. At
altitudes near 100,000 feet, there was so little atmo-
sphere; turns took long distances to accomplish and
resulted in higher loss of energy than forward fight.
The entire flight to maximum range was almost 45
minutes, about twice as long as ballistic missile tran-
sit. When these flight capabilities were presented at
Vandenberg AFB a few months later, Range Safety
personnel expressed concerns in several respects.21
Ground-based radars and navy-radar-tracking ships
would not be able to follow the vehicle for long peri-
ods, because of its relatively low glide-altitude (i.e.,
only above the radar horizon for a few minutes at

117
william yengst

each location), flying at high velocity (stressing-radar-


tracking rates) and with almost no radar-cross-section
signature (causing radar acquisition difficulties).

Flight Test Plans and Range Tracking


The vehicle design was completed during 1964, and
functional tests were performed on its subsystems.
Meanwhile, a series of meetings were held at Van-
denberg AFB to resolve flight test planning. It was
decided the vehicle would broadcast telemetry data
continuously to act as a beacon for radar trajectory
tracking. More important, the flight path would be
shaped to take maximum advantage of ground-based,
tracking radars (i.e., Oahu, Maui, Johnston, Midway,
and Wake Islands), since the U.S. had only a handful
of radar-tracking ships in the Pacific command that
could cover broad-ocean segments of the flight.
The Hawaiian Islands had two useful radar and
optical tracking centers: Kaena Point (Oahu) with an
Ultra High Frequency (UHF) tracking radar, was built
between 1956–1958 to support Western Test Range
monitoring of ballistic missile tests from Vandenberg
to Kwajalein, and also as a Space Surveillance Center
to support NASA. The Maui Space Surveillance Sys-
tem (MSSS) was constructed on three acres of land, at
an altitude of 10,054 feet, on the western rim of Hale-
akala volcano just outside the National Park. It could
provide UHF Mountain Top electronically scanned
tracking of suborbital satellites plus electro-optical
tracking of near-earth objects. Johnston Island, a tiny
U.S. Air Force managed atoll, 765 miles southwest of

118
lightning bolts

Honolulu, had an FPS-16 low-angle, acquisition radar


plus a Doppler radar-tracking station. This FPS-16 was
used by the Western Test Range for “Range Safety”
purposes, because it could pick up and follow ballistic
missiles 5 minutes after launch from Vandenberg and
track them through booster shutdown. Finally, Mid-
way Island, located 3,285 miles west of Vandenberg,
had another FPS-16 radar used to monitor ballistic
missile tests.
It was decided the BGRV flights would be flown
out of Vandenberg on a glide path south of the large
island of Hawaii, turning northwest near Johnston
Island, heading for Midway Island, and perform-
ing a reverse, “hook-slide” maneuver into the vicin-
ity of Wake Island. This route is illustrated in Figure
13. The flight range would be about 4,600 miles and
take 45 minutes. Radars on the islands would be sup-
plemented by at least three range-tracking ships: 1)
between mainland U.S. and Hawaii, 2) between John-
ston and Midway islands, and 3) near Wake Island to
monitor the terminal maneuver.
Navy radar tracking ships in the Pacific were in
constant demand during the mid-1960s. USS Chicago
(CA-136), built in 1943, homeport of San Diego, was
a typical example. With a displacement of 13,600 tons
and compliment of 1,142 personnel, it was assigned to
the Pacific Missile Range from 1967–1968. However,
during this period, it was ordered to the east coast of
Russia to observe Soviet ICBM test launches from the
Plesetsk Missile Test Range. It was sent to the Gulf of
Tonkin to monitor North Vietnamese radio commu-
nications and warn of “MiG” aircraft operations.7 Fol-
lowing stops in the Philippine Islands for supplies and

119
william yengst

maintenance, it was shuttled to Kwajalein for Minute-


man and ABRES missile tests. On 28 January 1968—
following seizure of the Pueblo (AGER-2)—cruiser,
Chicago, was sent to the Sea of Japan to help coordi-
nate air operations of Task Group 70.6 (i.e., less than
a month before the BGRV-4 flight test). Scheduling
the few capable ships (each subject to higher priority
missions) and providing them with sufficient sailing
time to reach stations along the BGRV flight path was
an always-changing proposition.

Figure 13. BGRV flight test route from


Vandenberg AFB to Wake Island

The Soviet Navy was perhaps in better condition


for monitoring BGRV flight tests in terms of avail-
able tracking ships. It had two, older space-tracking
ships (Marshal Nedelin and Marshal Krylow) plus
Dolinsk (1950s vintage), and five newer vessels built
and deployed between 1965–1967.

• Cosmonaut Vladmir Komarov


• Nevel

120
lightning bolts

• Morzhovets
• Kegostrov
• Borovichi

These ships carried Russian P-30 radars, capable


of tracking small reentry vehicles at ranges up to 500
kilometers (> 300 miles).8 They were formerly listed
as “measurement stations,” and most were located in
Pacific waters to support space and missile launches
in addition to collecting data on U.S. tests. The Soviet
ships appeared to have excellent knowledge of U.S. test
dates and experiment types. From three to five of their
vessels would arrive on station near our test sites several
days in advance of our launches to monitor radio com-
munications and telemetry frequency test runs.

BGRV Flight Tests:


The first BGRV flight test was scheduled for 11 Octo-
ber 1966, eight weeks after the MBRV-1 failure. The
test would use Atlas (115F), which had been modified
by General Dynamics Convair to strengthen its upper
skin and add the cork coating to protect against exit
heating. Liftoff from pad 576A2 was planned for early
morning, but delays in fueling the booster and align-
ing the guidance system prevented the launch until
almost noon (11:59 a.m. PST or 19:59 GMT). The mis-
sile with its long, slender, and gleaming projectile lifted
off beautifully, as shown in Figure 14.23 It performed
the worrisome trajectory turn to inject BGRV-1 on
its glide trajectory at roughly 125,000 feet altitude. At

121
william yengst

that instant, the sky lit up with a bright flash, vehicle


telemetry stopped abruptly, and the booster contrail
vanished. Vandenberg AFB records show the booster
reached an apogee of 31 miles (163,680 feet), but this
was probably misleading. BGRV-1 exploded at sepa-
ration, which was at a slightly lower altitude.9

Figure 14. BGRV-1 seconds after lift-off (11 October 1966)

Having experienced several Atlas-F booster fail-


ures, Vandenberg Range Safety and McDonnell per-
sonnel conducted an in-depth investigation of the
event. BGRV telemetry showed guidance, and ther-
mal sensors were all operating normally up to the time
of explosion. Someone at General Dynamics sug-
gested a static-electrical charge might have built up
on the booster with its cork surface during powered
flight. The condition was simulated in a McDonnell
hanger, and it was shown that as the vehicle separated

122
lightning bolts

from the charged booster, a lightening bolt almost 30


feet long discharged between the two bodies. The bolt
apparently detonated the Range Safety explosive in
the vehicle to blow it up. The fix for this unforeseen
accident was to electrically connect the Atlas inter-
stage structure to the base of BGRV with a spool of
thin copper wire over 50 feet long. The wire would
keep the booster, and its payload at the same electro-
static charge potential, until it broke and the two bod-
ies were sufficiently far apart to prevent arcing.
The BGRV-2 flight test was scheduled for 19 May
1967 on Atlas (119F). Vandenberg flight records indi-
cate this missile lifted off at 2:00 p.m. PST (10:00
GMT) from pad 576A1, adjacent to the 576A2 pad
used for MBRV, as shown in Figure 10. It reportedly
reached an apogee of 868 miles, and this undoubtedly
represented a ballistic-missile trajectory to Kwajalein
and reflects a projection of the booster path prior to
injecting BGRV on its glide path.9 As I recall, this
test failed during or shortly after separation from the
booster, and no successful telemetry was obtained.
Test launches and results were rarely reported to local
media; however, some tests could be observed from
outside the Vandenberg base perimeter. Therefore, if a
spectacular flight occurred, it might receive comment
from base officials; otherwise, “successful tests” were
only reported with Atlas-apogee data. Failures were
either given a cover story or were not mentioned.
BGRV-3 was launched on 10 November 1967,
using Atlas (113F) at 4:20 a.m. (12:20 GMT) from pad
576A1.9 This flight functioned perfectly, and within a
few minutes after launch, we learned that separation
occurred and vehicle telemetry was being received.

123
william yengst

About 20 minutes later, personnel on Johnston Island


reported picking up the vehicle on radar and more
amusing, the sonic booms knocked coconuts off palm
trees that lined the island’s narrow runway. Once
again, Vandenberg AFB records showed the booster
reached an apogee of 868 miles (i.e., one that reflected
a ballistic trajectory). Within a couple of weeks, Dick
Sheffer received telemetry data, which showed the
vehicle performed as designed and body temperatures,
were within a few percent of those predicted by the
preflight, thermal simulations.
Scheduling of other ABRES tests delayed the
flight of BGRV-4 until 26 February 1968. Although I
did not witness this final test, Dick provided an exten-
sive presentation of its results for Aerospace corporate
personnel about one month after the flight. The Van-
denberg report on BGRV-4 indicates that Atlas (116F)
lifted off from pad 576A1 and reached an apogee of
868 miles (same report as in previous tests).9 However,
a further brief explanation was provided: “BGRV was
launched from Vandenberg AFB to the area of Wake
Island. It was launched from an Atlas booster and
served to provide data on hypersonic-maneuvering-
flight characteristics.”10
Dick mentioned in his briefing that five Soviet
radar and optical-tracking ships were stationed along
the flight path to monitor the event. U.S. optical-
tracking sensors obtained infrared images of the vehi-
cle, as it streaked past at an altitude of about 100,000
feet and velocity in excess of Mach 10. Vehicle telem-
etry showed the transpiration nose tip worked well
and the guidance system functioned with perfection
past Midway Island and through the “hook-slide”

124
lightning bolts

maneuver off Wake Island. The TERSE radar altim-


eter worked with no “blackout.” We were impressed
with how closely thermal sensor data on the vehicle
fuselage and flare matched predicted temperatures
(within a few degrees at each location) throughout the
flight. I recall Dick’s charts showed skin temperatures
at mid-body and flare thermal sensors as nearly a flat
line of 1,870 degrees Fahrenheit from reentry until the
terminal maneuver. Dr. Ernst Krause and Aerospace
managers jokingly kidded him about “cooking the
books,” (i.e., playing simulated results over the telem-
etry system).
McDonnell engineers, G. L. Winchenbach, and
W. R. Lawrence, prepared a separate report on “Free-
Flight Range Tests of BGRV,” and R.J. Masek, B.M.
Levine, and D.D. Philpott wrote a report explaining
the methods used in reducing and evaluating BGRV
aerodynamic and thermal data.11, 12 These reports,
written in 1967 and 1968, showed that McDonnell was
studying the data in greater detail. Several volumes
of the classified BGRV Final Report, including the
development analyses, ground tests, and flight tests
were submitted to the USAF/SAMSO in September
1968.13
I greatly admired the results of this program as
a technology milestone and assumed Walter Dorn-
berger was exceedingly pleased. The German, “Inven-
tor of Boost Glide,” retired from McDonnell Aircraft
in 1965 and moved to Camden (New Jersey), where he
lived with his wife until his death on 27 June 1980.14
BGRV-4 was recovered from waters near Wake Island
by the U.S. Navy and put on display at McDonnell
Aircraft [see the photograph of Figure 8].10

125
william yengst

New Soviet Defenses


Success of the SA-2 Surface-to-Air Missile (SAM)
in shooting down high-altitude (i.e., 80,000 feet) U-2
reconnaissance aircraft (i.e., Francis Gary Powers on
May Day 1960, over Sverdlovsk and Major Rudolph
Anderson over Cuba on 16 August 1962), caused the
Soviet Union to initiate development of a series of
high-performance, air-defense missiles. The first of
serious concern to U.S. ballistic-missile developers
was SA-4 Ganef (Soviet Krug). It was considered to
have long intercept radius (31–34 miles fly-out range),
medium-to-high altitude performance but with a
maximum velocity of Mach 4. It carried a 300-pound,
high-explosive warhead on its semiactive, radar-
homing missile. The missile system was aided by two
radars: Long-Track E-band early warning radar, which
could detect reentry bodies several hundred miles
out in space, and Pat-Hand H-band fire-control and
intercept radar with a range of 80 miles.15
The Ganef SAM system was first deployed in the
Soviet Union in 1967 (about the time of the BGRV
flights) and was later moved into Warsaw Pact coun-
tries to protect against NATO missile threats. U.S.
intelligence also detected initial development efforts
on SA-6, SA-10 and SA-12 air defense missile sys-
tems.16 Thus, by the late 1960s, the Department of
Defense was searching for improved methods to
defeat these systems. ABRES became the focus of
several new penetration aids and maneuvering reentry
vehicle programs to penetrate these defenses.

126
lightning bolts

Follow-On Developments
McDonnell Douglas was quick to exploit the suc-
cess of BGRV and invented the Low Level Run-In
(LLRI) concept. An early version, LLRI used BGRV
aerodynamics but had a refractory-coated-metallic
skin (Niobium Alloys17), transpiration cooled nose
tip, and weighed almost 3,000 pounds. What made
it unique was the flight trajectory. After flying to the
intended target area, LLRI would dive sharply from
its 100,000 foot, glide-path, pull-out of the dive at
very low altitude (e.g., few hundred foot terrain clear-
ance) with a velocity of Mach 4 to Mach 6, and fly
the last 30 miles to the target below radar detection
and tracking capabilities. This maneuver was strictly
intended to defeat SA-4 class defenses protecting a
30-mile perimeter.
Unfortunately, Soviet development of the SA-10
Grumble class interceptors with significantly increased
performance eroded the LLRI concept by 1980. SA-10
could reach speeds of Mach 8, perform 100-g maneu-
vers, and fly out to intercept at 50-mile ranges. More
important, its FLAP-Lid radar could track 24 incom-
ing targets and engage four of them at the same time.
The radars and interceptors were reported to be capa-
ble of engagements down to altitudes of 80–100 feet.
Their supporting Big Bird and Clam Shell, continuous-
wave Doppler radars, could detect ballistic RVs 620
miles way.16 To remain effective, LLRI would have to
add terminal propulsion and increase stealth features,
causing the cost per vehicle to increase sharply.
During mid-1987 Congressional hearings, it was
revealed that the air force was reducing the number

127
william yengst

of operational SR-71 Blackbird, supersonic-recon-


naissance aircraft, because of their high-operating
costs and age. Air Force Secretary, Edward Aldridge,
explained a new, low-observable signature, lower
cost, manned, strategic-reconnaissance aircraft was
needed.18 Details of the program, code-named Aurora,
were hidden by deep security, but it was reported the
three-man aircraft was hypersonic and would replace
the SR-71. Two contractor teams, having only slightly
different designs, were contracted to develop pro-
totype aircraft: McDonnell Aircraft Company and
Lockheed ‘Skunkworks’ (Burbank, CA).
In August 1989, an unidentified eyewitness
sketched an aircraft seen cruising at Mach 8 (5,300
mph) over the North Sea. The aircraft was preparing
to refuel from a KC-135 air-tanker and was escorted
by two F-111 fighter-bombers. Its plan-form was a
perfect, 75-degree, swept-back triangle, corresponding
exactly to a hypersonic design prepared by McDonnell
Aircraft in the late 1960s. By that time, Paul Czysz, a
professor at St. Louis University, confirmed the air-
craft resembled a drawing he obtained in late 1989. It
was a hypersonic lifting body about 88 feet long with
a 48-foot span. Czysz explained it had an air breath-
ing propulsion system, which burned liquid methane
with a unique low frequency, very high-amplitude,
pulsing sound. The engine had been tested in 1980
at the USAF’s Edwards AFB and a secret flight-test
center at Groom Lake (Nevada).14, 19 Aurora would fly
at 100,000 to 130,000 feet altitude and use structural
materials plus energy management software devel-
oped and tested on BGRV. An artist conception of the
aircraft is shown in Figure 15.20

128
lightning bolts

Figure 15. Artist illustration of Aurora


hypersonic aircraft (1989)

129
chapter 5

MARCAS Development and Test


In the spring of 1964, Captain Mack Mauldin invited
Wes Mann and I to attend a meeting with him at
Douglas Aircraft Company in Santa Monica (i.e.,
before Douglas merged with McDonnell Aircraft).
During our two-hour drive from San Bernardino to
Santa Monica, he explained that Douglas Aircraft had
proposed a new, small, maneuvering-reentry-vehicle
concept to the Advanced Research Projects Agency
(ARPA). The director of ARPA contacted the Air Force
BMD and agreed to help fund research on the concept,
if it looked promising. Specifically, Douglas proposed
to employ “jet reaction” or “external burning” control to
maneuver a reentry body, instead of using aerodynamic
flaps, flares, or fins, which increased the radar cross sec-
tion of the vehicle and made it easy to distinguish from
small, low radar-observable conic ballistic RVs.

131
william yengst

We were met in the lobby of the huge Douglas


complex by Dr. James ( Jim) S. Murphy and D.B.
Harmon Jr. They took us to Jim’s office overlooking
the hangar buildings and adjacent Clover Airport.
We were introduced to several members of the staff,
who showed us artist illustrations of the two con-
trol concepts and results of laboratory experiments.
They argued that jet reaction controls were signifi-
cantly more effective in terms of packaging, weight,
and response time than aerodynamic flaps or flares.
The “jet reaction” concept involved using a slot-nozzle
near the base of the RV to inject inert gas into the
air-stream, changing the vehicle angle of attack and,
thereby, using the inherent body lift to turn. “Exter-
nal Burning” was more exotic in that a hypergolic fuel
(e.g., gasoline or hydrazine) was squirted into the air-
stream to ignite spontaneously and offset high-air
pressures. These concepts are illustrated in Figure 16.

Jet Reaction External Burning

Figure 16. Reaction Control Concepts

Jim’s staff ( J.W. Barnes and G.P. Johnson) assured


us elements of the control system could be placed
forward in the RV to keep the center-of-gravity near
the nose. Reaction controls would avoid drawbacks of
rear-mounted airfoils, which required high pressures

132
lightning bolts

to offset hypersonic airflow. In addition, jets were


ten times faster in response to steering commands.
Finally, they would work at higher altitudes (above
50,000 feet) where airfoils were not overly effective,
thereby, permitting maneuvers earlier in reentry flight.
Clearly the concepts had appealing features for
more effective maneuver controls, although Wes and
I saw difficulties as well. Rolling the vehicle to align
a single nozzle with steering commands would take
time; otherwise, multiple nozzles would be required
with internal plumbing and valves to direct their
exhaust jets. Thermal materials would be required
around exhaust nozzles to prevent erosion of the
heat shield and alter aerodynamic flow. Effectiveness
of reaction jets was questionable since efficiency was
likely to vary with altitude, surface flow, air density
and, in the case of external burning, available oxygen
in the atmosphere and choice of fuel. These concerns
deserved detailed analysis and eventually, flight tests.
Mack reported our findings to the USAF/SAC
and ARPA. Before the month was over, we learned
that a third maneuvering program would be con-
ducted under the title, “MAneuvering Reentry Con-
trol and Ablation Studies (MARCAS).” Douglas Air-
craft would perform three flight tests of the “jet reac-
tion” control concept, plus selected heatshield materi-
als on Athena missiles. As I recall, ARPA committed
$7 million for the program, and the USAF agreed to
supply Athena missiles (i.e., part of its test booster
development program). Our Aerospace office would
be responsible for GSE/TD of the Douglas contract.

133
william yengst

Douglas started design of a jet-reaction-control reen-


try vehicle in the late summer of 1964.

To minimize program costs and avoid serious


range safety concerns, the MARCAS vehicle would
be a ballistic projectile (i.e., positive stability mar-
gin of several percent of body length). There would
be no inertial measurement unit or on-board com-
puter to permit independent steering and maneuver-
ing. Instead, the “jet reaction” control system would
be switched on during reentry to change the angle
of attack for a few seconds, but when off, the vehicle
would aerodynamically return to its ballistic path. In
this manner, the estimated impact at White Sands
Missile Range (WSMR) could be predicted within an
elliptical area about 1.5 miles long by 0.5 mile wide.1
On-board attitude and accelerometer sensors would
measure the vehicle response to the reaction jets and
telemeter the data to ground receivers. These objec-
tives were documented in the program First Quarterly
Report (December 1964).2

Athena Missile Development


Only a few Athena missile flights had been made at the
time of MARCAS program initiation: the first on 25
February and the third on 8 July 1964. William Dono-
van (“Don”) Schutt of Aerospace (San Bernardino)
was responsible for performing technical direction of
the Atlantic Research Corporation development of
Athena. His office was located only a few steps down
the hall from our Adaptive Systems Office, and Don

134
lightning bolts

was very helpful in providing data and coordinating


the MARCAS flight tests.
Athena was a four-stage, solid-propellant, partially
guided missile, designed to simulate flight dynamics
of intercontinental ballistic missiles. It had the fol-
lowing characteristics: (1) deliver about 50-pound
payloads to apogee at 250,000 feet altitude, (2) sepa-
rate the payload including the third and fourth stage
boosters, (3) provide a guidance and orientation phase
to align for reentry, (4) thrust the payload to speeds
approaching 22,000 feet per second at reentry angles
between 18 and 46 degrees and (5) separate the payload
on a ballistic path from the fourth-stage booster. The
Athena Mod-1 (used for the MARCAS tests) could
only loft the payload to a maximum reentry angle of
30 degrees. Most early flights were made from Green
River (Southern Utah) over four states into White
Sands Missile Range (WSMR, New Mexico).3 A typ-
ical flight trajectory is illustrated in Figure 17.
Functionally, Athena Stage-1 and Stage-2 boosters
with aerodynamic control fins, lifted the upper stages
and payload to a “Guidance Phase” of flight where
a ground-based computer working from radio sig-
nals, determined position, velocity, and reentry con-
dition requirements to hit the desired impact point
at WSMR. Overall, the missile was 50 feet long and
had an upper stage diameter of 31.1 inches, payload
capacity of about 50 pounds, launch weight of 16,000
pounds, and it could reach an altitude of 185 miles.4
The upper stage diameter and weight constrained the
MARCAS vehicle design.

135
william yengst

Figure 17. Typical Athena Flight


trajectory with functional events

The time of flight for Athena, launch to impact,


was only four minutes. The first-stage (two Thiokol
Castor rockets) burned for 38 seconds; the second-
stage (Thiokol Pedro rocket) burned for 11.3 seconds
and was followed by a 60 second coast period for the
Guidance Phase. The third-stage (Aerojet Alcor rocket)
burned for 30 seconds and the fourth-stage (Hercu-
les Alcona rocket) burned another 8 seconds.4 Thus,
Athena required half its flight to achieve the desired
reentry velocity and flight path angle to WSMR. The
final 90 seconds of flight involved reentry to an alti-
tude of roughly 50,000 feet, followed by no more than
30 seconds for experiments before impact.
WSMR range safety was leery about lifting vehi-
cles, which could fly if they lost control. During test-
ing of V-2 missiles at WSMR, one flew across the bor-

136
lightning bolts

der and crashed outside Juarez (Mexico) on 15 May


1947. WSMR personnel were greatly relieved to learn
MARCAS was a stable RV, which would not hold an
angle of attack and fly off its ballistic trajectory.

MARCAS Design Reviews


Because of development difficulties with the MBRV
guidance system, Mack Mauldin and I were com-
mitted nearly full-time to supporting the G.E. pro-
gram. However, Lieutenant Jim Baird excelled in the
role of Contracting Officer Technical Representative
(COTR) for MARCAS, and Wes Mann managed our
program tasks. Wes, Jim, and I attended the initial
MARCAS program review in September 1964 at Jim
Murphy’s new McDonnell-Douglas office in Hun-
tington Beach (California). During the meeting, the
vehicle design was presented with supporting analyses
to show it was compatible with Athena missile capa-
bilities. A major RV technical issue to be addressed
was effects of pressure and enthalpy (energy and heat)
imposed on the ablative materials surrounding the jet
nozzle. The presentation by G. P. Johnson was docu-
mented in the program Final Report Volume III ( July
1965).5
Specifically, the RV length was 60 inches, with
a sharp (1-inch diameter) nose tip and base diam-
eter of 31 inches. The nose-cone half angle was 14.0
degrees. This large half-angle caused a higher than
desired drag (i.e., CD > 0.22). Because of its low pay-
load weight (W~50 pounds) and large base area of
(A=5.24 square feet), the ballistic coefficient was only

137
william yengst

(W/CDA~47 pounds per square foot). Normally, this


would cause rapid velocity decay, and the trajectory
would bend downward due to gravity near the end
of flight. However, the vehicle’s loss of energy and
velocity were offset by burning the third and fourth
stages of Athena, as indicated in Figure 17. In fact, the
fourth-stage booster rocket performed a second retro
burn to separate it from the payload to insure unam-
biguous-ground radar tracking of the RV experiment.
It is instructive to consider the MARCAS ballistic
RV design. MARCAS was not actually a maneuver-
ing vehicle, and unfortunately, precise design features
could not be found. However, an approximation can
be derived as suggested below.

Approximate MARCAS Vehicle Description


Nose Tip: Solid, 3-D weave carbon-
phenolic, 14 degrees half-angle cone,
tip radius of 0.5 inch and length of 8
inches 1.8 lbs.
Nose Ballast: Hollow tungsten, conic
frustum, 0.5 inch long 2.5 lbs.
Heat Shield Shell: Aluminum shell (2 mm
thick) covered by carbon-phenolic (2.5
mm thick) 51.5 inch long conic frustum 33.5 lbs.
Structure and Base Cover: Aluminum hoops,
brackets & cover 5.5 lbs.
High Pressure Gas Bottle: Steel (3,000 psi)
container, 10 inches long by 3 inches
diameter 3.4 lbs.

138
lightning bolts

Reaction Jet & Plumbing: Steel tubing,


exhaust nozzle, valve to nozzle, ablation
materials around exit, and altitude
sensor/timer 3.5 lbs.
Battery Power: Wet cells plus wiring to
control/telemetry 2.5 lbs.
Telemetry: Antenna, transmitter, two
accelerometer sensors 3 lbs.

Total Weight 55.7 lbs.

Although this description is only approximate,


it shows the vehicle, and its equipment were severely
weight restricted. The nose tip, heat shield, internal
structure, and small ballast for stability accounted for
roughly 78 percent of the vehicle weight, simply to sur-
vive reentry. Reaction control jet elements were barely
13 percent of the vehicle weight, with the remaining 9
percent devoted to electric power and telemetry. Even
though these estimates may not be precise, they indicate
that heat protection and structure dominated the design
for an experiment weighing less than seven pounds.
A full, three-axis control and guidance subsystems to
permit reentry maneuvers would have weighed twenty
times as much and could not have been flight tested on
Athena.
By the summer of 1965, Athena flights were being
made at a rate of one or two per day from Green River,
Launch Pads 2 and 3. In all, 141 Athena flight tests were
performed from the facility between July 1964 and 2
August 1973.6 Figure 18 shows a typical nighttime launch
from Green River on 16 July 1965.

139
william yengst

Figure 18. Athena missile launch with RV


similar to MARCAS (16 July 1965)

MARCAS Flight Tests


Athena flight test records show the first reaction con-
trol reentry vehicle was launched on 11 October 1965.
The test was performed for the USAF Space and Mis-
sile Systems Organization (SAMSO), new name for
the former Ballistic Missile Division. Athena mis-
sile (CO-16) flew from Launch Pad 2 at Green River
(Latitude 38.9 degrees north and Longitude 110.1
degrees west), reached an apogee altitude of 120 miles
and separated the upper two stages, plus the RV on its
intended trajectory. The test report described it as suc-

140
lightning bolts

cessful with separation of the RV from the 4th-stage


booster, followed by two burns of the reaction control
system.7 I recall seeing telemetry data from the flight
that showed the RV attitude changed abruptly by
several degrees below 50,000 feet attitude and again
around 35,000 feet altitude, when the jet was turned
on and off. Neither Wes Mann nor Dick Sheffer was
overly enthused by the high-altitude performance of
the reaction jet. It was fast to respond but simply did
not work as efficiently in thin atmosphere as originally
anticipated.
The second MARCAS flight was conducted early
in 1966 (apparently 21 January on Athena missile
C0–25). That test, also successful, was documented by
Jim Murphy, R. E. Bloomquest, and F. W. Spaid in
September 1966.8 The third and final MARCAS flight
was on Athena (CO-49) launched on 5 May 1967. This
test, documented by H.F. Lewis in July 1967, showed
the carbon-composite materials surrounding the reac-
tion jets had no excessive erosion. However, the inert
gas jets did not produce sufficient thrusting power at
extreme altitudes to cause a significant change in the
vehicle angle-of-attack. The USAF/SAC, still focused
on the potential operational advantages of jet reac-
tion control, was not satisfied with the limited flight
data obtained on MARCAS and convinced SAMSO
to conduct another reaction-control experiment with
McDonnell-Douglas Aircraft Company using Athena
No. V-1168, Flight 101, on 2 May 1968.9
On 11 July 1970, Athena missile (V-123D) was
launched from Green River at 8:40 GMT, reached an
apogee of 120 miles but suffered an ionization “plasma”-
induced communications blackout. Its booster failed

141
william yengst

to orient properly for reentry and ground radars lost


track of the RV over the horizon. The RV landed in
a remote region of Mexico 150 miles south of Juarez
near the town of Ceballos. During the month-long
search to locate and recover radioactive instruments
from its payload, the event caused a political crisis
between the U.S. and Mexico.1 Demands for larger
reentry payloads plus range safety concerns ranked
high in opposition to subsequent Athena Green River
to WSMR tests.
The Air Force and Army joined in the late 1970s to
resolve several Athena booster deficiencies by develop-
ing a new, larger scale Athena-H (space) missile, built
by Lockheed Space and Missile Company. The new
three-stage missile had a lift-off weight of 30,000
pounds, a payload capacity of 440 pounds, and an
improved guidance and control system.10 It was first
flight tested on 3 April 1971 and was used during the
early 1980s to support antiballistic missile develop-
ment and the Pershing-II medium range ballistic mis-
sile program (see Chapter 7).
The USAF SAMSO decided to perform all subse-
quent ABRES advanced development tests of MRVs,
using retired Minuteman-I boosters launched from
Vandenberg AFB to Kwajalein Island. This decision
was based on three primary considerations:

1. Advanced development of full-scale, oper-


ational MRVs would require larger and
heavier vehicle designs (e.g., weights over
800 pounds).

142
lightning bolts

2. Potential range safety concerns for high-g,


lifting bodies would be less severe on the
Pacific Missile Test Range.
3. Long-range, high-resolution radars and
optical-tracking stations had been built at
Kwajalein. Specifically, large, very expen-
sive TRADEX and ALTAIR radars for
testing Spartan and Sprint Anti-Ballistic
Missile defense systems were built by MIT
Lincoln Laboratories, Raytheon Corpora-
tion at Kwajalein, and Wake Islands dur-
ing the early 1960s.11 Since those radars
were better suited for detecting and mea-
suring reentry vehicle low observable sig-
natures, maneuvering reentry tests would
be made to Kwajalein Island.

Follow-On Developments
Before the MARCAS program was complete (early
1967), the Air Force Systems Command (AFSC)
expressed the need for an operational-maneuvering-
reentry vehicle to defeat Soviet defenses. Knowing
development would take several years, AFSC directed
SAMSO to initiate an “Operational MRV Concept
Study.” Two contracts were awarded to insure that
more than one vehicle producer would remain at the
end of the project. Further, the designs would be lim-
ited to ballistic delivery (like MBRV and MARCAS),
while the boost-glide concept would be pursued
under the Strategic Boost Glide Vehicle, (SBGV ),
an air-launched missile concept to extend the range
of B-52H bombers (see Chapter 6). SAMSO selected

143
william yengst

General Electric and McDonnell-Douglas Aircraft


Corporation to perform the Operational MRV Study.
Aerospace Corporation performed technical direction
and management of both contractor projects. The fol-
lowing classified reports were prepared.12

• General Electric: “Operational MRV Con-


cept Study: Technical Progress Report-
Phase A (including three Volumes),”
written by R.R. Herrick, GE 67SD742,
May 1967 and “Operational MRV Con-
cept Study: Final Report-Phase B,” by the
same author in August 1967.
• McDonnell-Douglas Aircraft Corporation:
“Operational MRV Concept Study: Final
Report,” written by H.F. Steinmetz, Sep-
tember 1967.

As expected, the G.E. operational designs were


based on its MBRV flap-control experience, whereas
McDonnell-Douglas favored jet reaction controls.
The technical debate concerning which mode of con-
trols was most effective (flaps or reaction jets) was not
resolved by design studies or tests. However, a new,
three-flap configuration emerged from the study that
proved to be competitive with reaction jets in weight
(i.e., it had one less flap, less hydraulic plumbing and
valves, more reliable operation, and it was equally capa-
ble of performing high-g maneuvers). Dick Hartunian,
Aerospace Corporation director of Reentry Systems
Division, wrote in his 2002 paper “Reentry Systems:
The Critical Years:”

144
lightning bolts

“Vehicles that used reaction jets to maneuver


were also considered, but design studies and wind
tunnel tests (plus flight tests) indicated the simpler
flap arrangements could perform all the maneuvers
required.”13
Although emphasis shifted to three-flap designs
for long-range ballistic missiles (see ACE and AMaRV
in Chapter 6), reaction jet controls did not van-
ish. U.S. Army programs for ballistic missile defense
interceptors continued reaction control research for
anti-ballistic missile interceptors. During 1969–1970,
Troy A. Street of Army Redstone Arsenal Laborato-
ries (Huntsville), Missile Research Development and
Engineering, performed a series of wind tunnel “oil
flow” experiments on sonic reaction jets ejecting from
a body of revolution into a free air stream at speeds
from Mach 1.75 to 4.5.14 Further evidence of continu-
ing research occurred in 1982, when David C. Sayles
of Huntsville obtained a patent (Number 4316359) for
adding ammonium-metavanadate (liquid oxidant) to
rocket fuels, such as Red Fuming Nitric Acid (RFNA)
or common turpentine to give them hypergolic prop-
erties. Sayles found that adding ammonium liquid
(about 1 percent by weight) permitted non-hypergolic
fuels to spontaneously combust in an oxygen-depleted
atmosphere.15 This research was aimed at improving
fuels for “external burning” reaction control.
Finally, Patent Number 7002126 (2 February 2006)
for an “external burning” concept, entitled “Projectile
Steering by Plasma Discharge,” is further evidence of
continued research. This patent noted, “a drawback of
airfoils is that they require considerable force, which
increases proportionally with the vehicle or projectile

145
william yengst

velocity at high dynamic pressures.”16 The proposed


concept was to use small rocket devices, which burn
once to provide terminal steering. The application was:
“steering a supersonic projectile or cone having a more
or less pointed end and capable of creating a plasma
discharge near the end of the projectile over a lim-
ited sector of the outer surface.” 16 The patent included
analytic simulation of airflow over sharp, hypersonic
projectiles having drag coefficients of (CD =0.0949 to
0.1157). The conclusion was “jet reaction and exter-
nal burning” controls were effective technologies for
steering hypersonic vehicles.
Although the choice between flaps and jet reaction
controls for maneuvering-reentry vehicles was not
conclusively resolved, subsequent U.S. ICBM-range
ACE and AMaRV vehicle designs employed three-
flap configurations. It will be seen that medium-range
Pershing-II and foreign intermediate range missiles
with maneuvering-reentry vehicles evolved to use four,
aft-mounted, cruciform fins, primarily because they
were simple, easy to construct, and readily adapted to
steering logic.

146
chapter 6

Second Generation Maneuvering


Reentry Vehicles
Dick Montgomery retired from DDR&E as deputy
for strategic systems in 1967 and was replaced by Dr.
Lloyd Wilson. When Lloyd came to Norton AFB
that fall to review the ABRES programs, he applied
a new methodology and criteria for further develop-
ment. Due to a tight budget, he stated by way of a
revised homily, “Something that is not worth doing,
is not worth doing well!” In other words, if a program
did not have a well-founded rationale and probable
application, it would not be funded. This guideline
was endorsed by the new ABRES commanding offi-
cer, Brigadier General John C. Toomay. John had a
strong background in radar technology and a man-
date to reduce costs. He favored suppressing reentry,
observable signatures, developing effective chaff and

147
william yengst

decoy penetration aids and trimming costly maneu-


vering projects. Although we argued on several occa-
sions, John was always reasonable, and we became
good friends during the next twenty years. He retired
from the air force with the rank of major general and
advised us at Science Applications International Cor-
poration (SAIC) on a number of projects into the
1990s.
The U.S. Department of State and Soviet diplo-
mats had made progress preparing the Anti-Ballistic
Missile (ABM) Treaty. Consequently, a shift occurred
in USAF, SAC, and USN strategic planning concern-
ing missile forces. Two primary considerations were
responsible for the change:

1. The ABM Treaty would restrict Soviet


defenses to protect only the Moscow
region. Hence, most Soviet targets would
have little or no defenses. Further, there
was less concern about defeating high-
performance interceptors, because the
U.S. had sufficient numbers of missiles to
saturate local defenses or could beat them
with penetration aids. Because the treaty
had not yet been signed and the Soviet
Union could abrogate it, maneuvering
technology was continued into advanced
development of an operational system in
case it was needed.

148
lightning bolts

2. Suggestions of treaty limits on the size


of deployed forces was an incentive to
develop Multiple Independent Reentry
Vehicles (MIRVs) to enhance delivery
potential by smaller numbers of missiles.
Minuteman III would carry three Mark
12A RVs and Poseidon-C3 SLBMs would
carry 10 to 14 Mark 3 RVs with W-68 war-
heads. These systems began development
during the mid-1960s.1, 2 Reentry vehicles
would focus on small, light weight ballis-
tic designs (i.e., USAF Mark 18 and Mark
21 and USN Mark 3 and Mark 4). Up to
seven Mark 21s would fit on a Minute-
man III, and eight Mark 4s could be car-
ried on Trident-I C-4 SLBMs. These bal-
listic RVs depended on penetration aids,
including chaff puff packages in space,
multiple decoys (some capable of high-
altitude reentry before burning up) and
suppression of observable signatures by
sharp conic nose tips, smooth low-radar-
reflecting surfaces and no ablation chemi-
cals or ionization trails.

The primary argument accepted for continu-


ing development of maneuvering technology was to
provide high-accuracy, terminal delivery for small-
yield, nuclear warheads. Clearly, the design, qualifi-
cation, and flight-testing of a guidance subsystem
was the long-pole in the development process for
an operational-maneuvering-reentry vehicle. Under

149
william yengst

the considerably reduced 1968 budget, the ABRES


program initiated two operationally oriented system
developments.

1. Inertial Guidance for Advanced Ballistic


Missiles: The Draper Laboratory’s AIRS
(Advanced Inertial Reference System)
version of its FLIMBAL concept was
given top priority. The small, high-accu-
racy, high-g inertial instruments for this
system took eight years to evolve and were
intended to give the planned Peacekeeper
(MX) and Trident missiles high delivery
accuracy. The one-foot diameter, AIRS
FLIMBAL platform is shown in Figure
19. From October 1967 through January
1968, I was assigned to oversee this project
at the MIT Draper Labs. My counterpart
at the Labs, Kenneth Fertig, developed the
weapon-certified version of AIRS. Dur-
ing this period, Dick Hartunian (General
Manager of the Reentry Systems Division
at Aerospace and my supervisor) was com-
mitted to definition of the Global Posi-
tion System (GPS). GPS was an alterna-
tive method for achieving high-accuracy,
terminal navigation. Dick moved to El
Segundo in 1970 to manage the satellite
system program.

150
lightning bolts

Figure 19. Draper Labs AIRS Flimbal

2. Advanced Development of Maneuvering-


Reentry Vehicles: This program had a long-
term goal of refining and validating MRV
aerodynamics, control system effective-
ness, and packaging subsystems to fit an
operationally sized vehicle. It would be
accomplished in two steps. An Advanced
Control Experiment (ACE) vehicle would
be flown to test reduced size and weight
of control mechanisms and qualify small,
strap-down, inertial-guidance compo-
nents. The second step, an Advanced
Maneuvering Reentry Vehicle (AMaRV ),
would be designed as an operational pro-

151
william yengst

totype and serve as test bed for a small,


nuclear-hardened inertial guidance and
computer navigation subsystem. The
guidance unit would weigh no more than
30–40 pounds, while retaining its accu-
racy through 100-g reentry maneuvers. It
would employ small Draper Laboratory
gyros and accelerometers in a radiation
hardened gimbaled platform, immersed in
a liquid to relieve reentry forces.3

Advanced Weapon System Concepts


I transferred to the Aerospace Missile Systems Divi-
sion in early 1968 to manage the Advanced Weapon Sys-
tems group, performing studies of future, land-based
strategic ICBM forces. Ruby Yzaguirre and I moved
back to the deplorable Norton AFB facilities to join an
exceptional team of engineers and scientists. Almost
before the move was complete, a project was initiated
by Air Force Systems Command (AFSC) to define
and evaluate a new generation of ICBM weapon sys-
tems.4 Dr. Brian D. Henshall, newly appointed Asso-
ciate General Manager of Aerospace’s San Bernardino
Operations, chaired the project, which became known
as, Alternatives for Strategic Missile Force Improve-
ments Study. Our group would perform the key role
of inventing concepts and evaluating their technical
merits.
Our first activity was to invent a series of five-
generic-weapon-system concepts designed to improve
the survivability of land-based ICBM forces in the

152
lightning bolts

event of a Soviet surprise first strike attack. In defin-


ing the concepts, we applied all classic survivability
methodologies: (1) mobility, (2) hardness, (3) prolif-
eration, (4) defense, and (5) deception or concealment.
The concepts included maneuvering reentry vehicles
in addition to MIRVs with penetration aids as pay-
load options. At Brian’s insistence, each concept was
given a name (Name the Baby).

• Ranger: A small, three-stage, single


AMaRV payload, missile about half the
size of Minuteman-I. The missile was car-
ried on a mobile wheeled or air-cushion
vehicle launcher to be operated on or off
roads at military bases throughout the U.S.
It employed dispersal-on-warning tactics
to achieve survivability.
ºº The concept was defined by William
(“Bill”) S. Vance with help from Earl
B. Anderson, Cecile L. Crews, and
Robert T. Scott. It was based on the
Soviet SS-20 IRBM weapon system
and was later selected by SAC as the
basis for the Midgetman ICBM proto-
type built and tested in the early 1970s
by Boeing Aircraft.
• Vulcan: A large, 120-inch diameter mis-
sile with up to 24 Mark 21 ballistic RVs
or 12 AMaRVs. It would be deployed in
new super-hard (1,000 psi hard rock or
concrete) silos at existing air force bases
and would be continuously active with

153
william yengst

an option to fly-out upon attack warning.


A defensive system overlay could also be
added to increase survivability.
ºº This concept was prepared by Victor
Kebely with help from Wes Eberle,
including missile design provided
by Boeing Aircraft Company. Dur-
ing definition, it evolved to become
the baseline for the Peacekeeper (MX)
weapon system.
• Nemesis: This very small 6,000-mile-
range missile was about one-third the size
of Minuteman-I with a single Mark 21 bal-
listic RV. However, the missile was encap-
sulated in a steel container and emplaced
in a hard (1,000 psia), unitary reinforced-
concrete silo under dormant-operating
conditions on existing Minuteman bases.
The missile could be activated by a coded-
radio signal to an antenna sealed in the
concrete silo and connected to its capsule.
The guidance computer contained a small
set of pre-stored targets for only assured
destruction attacks. The low-cost missile
and dormant operation permitted pro-
liferation of the force in large numbers
(e.g., 2,000–10,000 missiles) to maintain
a similar number of warheads as the large
MIRV missile options.
ºº This concept, invented by Frank
Newman and Bill Vance, appealed to
those who saw dormant operation as

154
lightning bolts

a means for achieving near total con-


cealment, low life-cycle costs, and
excellent hardness. A full-scale Nem-
esis silo was constructed for the Dial
Pack simulated nuclear test event on
23 July 1970 at Alberta (Canada).5 The
silo survived the 500 tons of TNT-
surface burst with no cracks or dam-
age at the rim of the crater. The USAF
liked this concept sufficiently that
SAMSO directed the MIT Draper
Labs to work on a dormant version of
the AIRS guidance system that could
be launched with only a few minutes
warm-up.
• Janus: A Minuteman-sized missile with
a single-maneuvering payload that could
be launched as either an offensive strike
weapon or (like the Soviet Galosh) as an
exo-atmospheric defensive interceptor.
It would be based in Minuteman silos to
defend existing forces and to be launched
on warning of attack. The missile could
be redirected during powered flight from
defense to offense using coded radio
signals.
ºº This concept, prepared by Nick F.
Kfoury and Edward Knight, was not
considered attractive by arms control
advocates. Surveillance systems could
not determine the weapon count
(offensive or defensive) for purposes

155
william yengst

of control monitoring. It also ran


counter to the ABM Treaty and those
trying to negotiate reductions in stra-
tegic offensive forces.
• Brimstone: This concept was essentially the
Vulcan-large missile with MIRV payload
but was based deep underground (e.g., half
mile or more deep in abandoned copper
or coal mines) as a citadel complex. The
complex had multiple, long, blind launch
shafts to the surface, normally filled with
sand, which could be drained into under-
ground chambers during a crisis. Each
shaft could be used to launch one or more
of the missiles. The key concept advantage
was supreme survivability through hid-
den but very hard launch portals, ability to
move missiles to adjacent shaft portals if
one or more were destroyed, and optional
dormant or active missile operation and
self-contained command and control.
ºº This concept proposed by Lockheed
Missiles and Space Company (Palo
Alto) had many interesting features.
However, it was very expensive to con-
struct, required extensive manpower
to operate and was slow (hours) to
respond in event of attack. Dr. Prem
Mathur of our group led an evaluation
including the survey of abandoned
copper mine sites in Michigan State
for the citadel complex.

156
lightning bolts

Over a six-month period, all five concepts were


defined and studied in considerable depth. Surveys
were made of potential deployment sites, develop-
ment plans, and 10-year lifecycle costs were esti-
mated for different force sizes, including technology
development programs schedules for each concept.
Bill Vance evaluated Ranger deployment sites in the
U.S., assuming wheeled transporter-erector-launchers
(TELs), while I studied similar bases in Alaska for
air-cushion vehicle, TEL operations. When the study
was complete, Brian Henshall and I briefed its results
to a Defense Science Board (DSB) Strategic Systems
Committee. The review was attended by representa-
tives of the Nuclear Weapon Laboratories plus a hun-
dred, Air Force-invited guests, who met in the Aero-
space Corporation main auditorium (El Segundo).
Dr. Edward Teller chaired the DSB review and
was supported by a dozen renowned scientists, includ-
ing Dr. Stark Draper, Drs. Albert and Richard Lat-
ter, and Aerospace President, Dr. Ivan Getting. I will
never forget the all-day series of briefings for several
reasons:

Brian Henshall got sufficiently excited dur-


ing his “executive summary” that he picked up
blackboard chalk and wrote answers to a ques-
tion on the face of the view-graph, projection
screen. When I started to brief the Ranger con-
cept, Edward Teller said he was tired of looking
at charts. “Just tell me about it,” he directed in a
gruff voice! Fortunately, we had built a wooden
model of the wheeled launcher designed by Cat-
erpillar Tractor Company, so I put the model

157
william yengst

on the table in front of the panel and used it to


describe the concept. The panel genuinely liked
Ranger; however, a few members began invent-
ing improvements. After several minutes of dis-
cussion, it was interrupted by an elderly guest
sitting in the front row seats. The short, bald-
ing, smiling man stood up and, in a strong voice,
addressed the panel, “Gentlemen, You have
been presented a ‘race horse’ idea—don’t create
a camel!” The unexpected guest was Lieuten-
ant General James H. ( Jimmy) Doolittle, World
War II hero of the Tokyo bombing raid and
Chairman of the Board of the Space Technology
Laboratories.

Within weeks, we briefed the study to General


Bruce Holloway, Commander in Chief of SAC, at his
headquarters in Omaha (Nebraska). He favored the
Vulcan concept, because it provided flexibility to per-
form all three offensive missions (i.e., assured destruc-
tion, damage limiting, and defense suppression).6 By
contrast, Nemesis (the lowest cost concept by far) had
two primary drawbacks: it could only perform assured
destruction attacks and was totally counter to arms
control proliferation constraints. Ranger could per-
form all the missions but was more expensive than
Vulcan and introduced land mobility, a form of opera-
tion new to the air force. Many officers liked mobil-
ity and chose the concept in 1980 for developing the
Midgetman ICBM and Ground Launched Cruise
Missile (see Chapter 7). Although Brimstone could
perform all the missions, it was ruled out, because of
excessive ‘up-front’ development costs and Janus was

158
lightning bolts

set aside due to its inability to satisfy arms control


constraints.
Additional briefings were given at AFSC Head-
quarters and the Pentagon for the Joint Chiefs of
Staff ( JCS). In general, the study was acclaimed,
and actions were taken to initiate several of its tech-
nologies. With a tight budget and billions of dollars
already sunk in existing bases, land, and infrastructure,
the Air Force was not prepared to develop a revolu-
tionary concept. Vulcan, renamed Peacekeeper (MX),
was endorsed by JCS and submitted to Congress for
approval and funding. Two years of detailed engineer-
ing, plus repeated debates within Congress, followed
before full-scale development of MX was initiated on
2 November 1971.7

ACE and AMaRV


Development Programs
Although not directly involved with these develop-
ment programs, I followed them closely, because
AMaRV was an important element of our advanced
weapon system concepts. Surprisingly, the rationale for
an operational Advanced Maneuvering Reentry Vehicle
(AMaRV ) was to insure high accuracy for Minuteman-
III and MX weapon systems. As noted by the Center
for Defense Information, “AMaRV ’s ability to cor-
rect its trajectory during reentry and terminal phases
of flight will give it nearly 100-percent accuracy.”8
However, the design goals retained high-g maneuver
capability to evade enemy defenses. The program was
accomplished in two steps: (1) a project to evaluate

159
william yengst

aerodynamics and control effectiveness of a weapon-


size vehicle, known as Advanced Control Experiment
(ACE) and (2) prototype operational AMaRV.
Within the Ballistic Missile Division ABRES
program office, Lieutenant Colonel Michael G.
Buchen managed AMaRV, assisted by Major James
G. Traeger.9 To insure that more than one contrac-
tor would be able to produce the operational weapon,
two contractor teams were selected for the first phase
of development [General Electric (GE) and merged
McDonnell-Douglas Aircraft Company (MDAC)].
They were provided all data and reports from MBRV
and MARCAS programs to support the design tasks.
Aerospace Corporation performed GSE/TD for the
program. In August 1968, Wes Mann became group
director for Special Purpose Systems, and with his
assistant, Richard (Dick) D. Daniels, was responsi-
ble for managing the maneuvering reentry vehicles.
Dick Sheffer directed “Special Vehicles,” George R.
Costello covered inertial guidance systems, and Dr.
W. Paul Thompson handled electromagnetic systems
(i.e., terminal guidance).10

• ACE Vehicle Experiments: Because ACE


was the first step in generating an opera-
tional weapon system, few references were
made to its characteristics, and progress
reports were highly classified. The pro-
gram began in July 1972 but design and
aerodynamic performance were not docu-
mented until January 1975.11 No pictures
or descriptions of ACE were released to
the media. McDonnell-Douglas Aircraft

160
lightning bolts

Company (MDAC), under program man-


ager D. V. Maghill, was selected during
the design phase to build and test three
full-scale ACE vehicles. The conic design
was wind tunnel tested at Arnold Engi-
neering Development Center (AEDC),
Arnold AFB (Tennessee), at velocities
from Mach 7.5 to 19. In April 1973, L. G.
Siler described static stability and drag
characteristics of the design.12 ACE vehi-
cles had “open-loop” guidance consisting
of an attitude reference system, autopi-
lot, accelerometers, and rate measurement
unit. Astronautics Company built the
autopilot, Systron-Donner Corporation
provided accelerometers, and Hamilton
Standard produced the rate measurement
unit.9 The simplified “open-loop” guidance
involved a preprogrammed trajectory in
the computer subsystem, but there was no
“closed-loop” steering correction, if errors
occurred. The vehicle incorporated sev-
eral MBRV features; however, only three
flaps were used to save weight (i.e., one for
pitch plus two for yaw/roll). The design
was reportedly close to the final configu-
ration of AMaRV described below. ACE
vehicles were flown on Atlas-F boosters
from Vandenberg AFB, Launch Site 576
A1.13 Flight dates were as follows: ACE-1
on Atlas (108 F) launched on 30 Septem-
ber 1973, ACE-2 on Atlas (97 F) launched

161
william yengst

on 23 March 1974, and ACE-3 on Atlas (80


F) launched on 9 August 1974. All three
missiles reached an apogee of 868 miles
and were declared “successful” by the Air
Force; however, some problems did occur.
The first vehicle broke up late in reentry
due to faulty prediction of atmospheric
data. The second flight was a total success,
but the third vehicle was lost due to an
undetermined-booster-interface problem.
Donald V. Magill reported, “The vehicle
did not reach the ground but was oper-
ating normally through reentry to a fairly
low altitude and provided adequate flight
data.”9

• AMaRV Development and Test: Detailed-


vehicle designs plus wind tunnel and scale
model tests for the biconic AMaRV con-
figuration began in early 1975. A biconic
was chosen to permit the W-78 nuclear
warhead (minimum diameter of about
20 inches), to fit forward in the vehicle
for stability purposes without requiring
excessive ballast. By May 1975, McDon-
nell-Douglas built and wind-tunnel tested
52-percent scale models of the design to
determine aerodynamic heating and pres-
sure characteristics.14 The new Theodore
von Karman (VKF) Gas Dynamics Facil-
ity at Arnold Engineering Development
Center (AEDC) provided hypersonic
wind tunnel testing for the lifting con-

162
lightning bolts

figuration under direction of James C.


Uselton.15 GE provided analytic stability
measurements for the biconic design.16
Aerospace Corporation prepared wind-
tunnel test plans and combined the results
of configuration refinements by Novem-
ber 1976.17 These tasks led to an Opera-
tional Vehicle (OPV ) design by the end
of 1976.18 A detailed design disclosure was
held at Lockheed Missile and Space Com-
pany (LMSC) to support the navy Mark
500 (or Mark 5) program in May 1977.19
In March 1981, Arthur Martelluci and S.
Weinberg (Science Applications Interna-
tional Corporation) provided an excellent
audit trail of classified reports concerning
this program.20

AMaRV had a greatly reduced weight (W=1,035


pounds) compared with MBRV and relied on a single,
split-body flap (called windward flap) for pitch control.
Two yaw-flaps mounted on the sides provided yaw and
roll control. The flaps were powered by a high-pres-
sure, hydraulic-oil subsystem. Use of only three flaps
for control reduced weight at the rear and moved the
center of gravity forward. The pitch flap was probably
scaled down to roughly 8 inches long by 10 inches wide.
The two yaw-roll flaps were located on the vehicle sides
and were either unitary components (operated differ-
entially to create roll) or split as in the MBRV design.21
The heatshield was thinned (lower safety margin), and
internal structure used less steel to save weight. As a
result, ballast required in the nose to achieve stabil-

163
william yengst

ity was dramatically reduced, saving several hundred


pounds of weight compared to MBRV.
The nose radius was 2.34 centimeters (0.92 inch),
and the forward conic section had a half angle of 10.4
degrees (compared with 11.0 for MBRV ) to a distance
of 31 inches from the nose tip. The aft cone frustum had
a half-angle slope of 6 degrees. The base diameter was
established identical to the ballistic Mark-12 RV (i.e.,
diameter of 22 inches or reference area of A=2.64 square
feet). The AMaRV length was 82 inches. Three of the
vehicles would fit on existing Minuteman III, post-
boost-vehicle (PBV ) MIRV positions, and ten would
fit on the planned MX PBV.21
The biconic shape made it possible to move iner-
tial guidance elements forward of the warhead. The
Dormant Inertial Navigation System (DINS), devel-
oped by Honeywell Corporation during 1977, provided
full, ‘closed-loop’ navigation with a digital autopilot.
Kearfott Division of the Singer Company designed its
small, inertial platform.22 It employed strapped down
Ring Laser Gyros, which were essentially insensitive to
vehicle accelerations and flight-time drifts during pow-
ered and coast phases of flight. Bell X1 accelerometers
would measure (>100-g) reentry maneuver levels. Litton
Industries provided the Reentry Guidance Computer
(RGC). It had a central processor developed for the B-1
bomber but was repackaged to withstand high-g and
vibrations. Consequently, the significant advancement
of AMaRV was its small inertial, navigation system,
prompted by the dormant Nemesis missile concept. An
improved version of DINS would be nuclear hardened
for the operational system.
AMaRV would carry the W-78 nuclear warhead

164
lightning bolts

(same as the Mark 12A), which was a spherical implo-


sion device weighing roughly 350 pounds. Its maxi-
mum diameter was 20 inches based on scaling from
the Mark 12A RV.23 It required adjacent Arming and
Fusing electronics plus independent battery power,
causing the combined elements to occupy an ellipsoi-
dal space as far forward as the frustum permitted.
The vehicle had a slightly higher drag coefficient
(CD~0.24) but lower weight and smaller base area (A =
2.18 square feet) than MBRV. The ballistic coefficient
was 2,000 pounds per square foot (i.e., intended to be
the same as the ballistic Mark 12 and planned Mark
21 RV ). This was to insure its reentry trajectory and
observable signatures could not be distinguished from
ballistic RVs by enemy defenses. The configuration
had a lift (L) to drag (D) ratio that approached unity
(L/D ~ 1.0) and provided lower peak deceleration
during reentry than MBRV. No photographic images
of AMaRV were published; however, the aerodynamic
configuration was used by MDAC twelve years later
to define a MARS reentry and soft-landing vehicle
for the National Aeronautics and Space Administra-
tion (NASA) as described below.
Four AMaRVs were built for ground and flight
test purposes. Three were flown from Vandenberg
AFB to Kwajalein on surplus Minuteman-I mis-
siles. All the flights were made from ABRES Launch
Facility (LF03) at Latitude 34.8427 degrees north and
Longitude 120.5699 degrees west, and occurred on
20 December, 1979; 8 October, 1980; and 4 October,
1981.24 In place of a warhead, the first AMaRV fight
carried telemetry and a plasma experiment, while
the second and third vehicles carried elements of the

165
william yengst

DINS operational guidance subsystem.9 Apogee for


all three flights was reported as 868 miles; however,
again, there were a few problems.
AMaRV-1 suffered a booster separation failure
that caused the vehicle to reenter the atmosphere in an
unstable condition. The vehicle’s control system func-
tioned until its telemetry signal was lost near impact
in the vicinity of Kwajalein. The vehicle attempted to
make a diving turn but started at too high an altitude,
and when stability was restored, it had already passed
the atoll and reverted to a coasting mode to impact in
the ocean.9 The nighttime infrared image of that flight
suggested it followed a trajectory similar to MBRV-3,
as shown in Figure 20. Between 50 and 75 percent of
the flight maneuver and objectives were accomplished.

Figure 20. AMaRV-1 reentry flight trajectory (upper light


streak) and booster missile stage (below) over Kwajalein

The Air Force transferred management of ABRES

166
lightning bolts

to Thompson-Ramo-Wooldridge (TRW ) at a new El


Segundo facility during 1979. This management tran-
sition may have delayed some AMaRV flight sched-
ules; however, nearly a year between flights permitted
time to evaluate data and make design refinements
before the next test. Dick Hartunian noted the three-
flap design worked very well in all tests.21 Following
the three “prototype flights,” AMaRV was declared
ready for operational service on Minuteman III or MX
missiles in 1982.3

DC-X Reentry Vehicle


Twelve years after the successful AMaRV flight tests,
NASA conducted a design study for a space vehicle
capable of carrying people to the planet Mars to make
a soft landing on its surface and return. Three con-
tractor teams were funded between 1994 and 1996 to
design competitive vehicles: (1) Lockheed-Martin
with a delta-shaped, lifting body known as X-33, (2)
Rockwell Aerospace Corporation with a winged lift-
ing-body, and (3) McDonnell Douglas Aircraft Cor-
poration (MDAC) with a maneuvering biconic vehicle
known as DC-X. The DC-X was actually a scaled-up
(by roughly ten times) version of the AMaRV aerody-
namic design with only minor changes.
The aerodynamic configuration was identical to
AMaRV in nose shape, biconic, half-angle, heat-
shield slope and aerodynamic performance. However,
the three aft-mounted control flaps were replaced by
three aft-body, mounted fins. To support its proposal,
MDAC built a one-third scale model of DC-X to

167
william yengst

test for low speed (soft) aft-landing performance.25


A series of landing and takeoff tests was conducted
between 18 August and 30 September 1993. The aero-
dynamic configuration was unclassified in 1996 and
although the NASA MARS program was discontin-
ued, the technology was used by MDAC in preparing
its design for the reusable space shuttle.

Figure 21. NASA vertical-takeoff and landing vehicle (DC-X)

Navy Mark 5 Maneuvering


Reentry Vehicle (MRV)
The Poseidon Sea Launched Ballistic Missile (SLBM)
C-3 program began in January 1965, and was designed

168
lightning bolts

to deliver the W-68 nuclear warhead in a ballistic


Mark-3 RV. When the program was upgraded in 1984
to provide increased operational range to 2,000 nauti-
cal miles, the Mark 4 ballistic RV (with W-76, 100 Kt
warhead) was developed to survive the increased reen-
try environments and to add a sharp nose to reduce
reentry observable signatures. The Trident-I (C-4)
missile was intended to deliver eight Mark 4 RVs
from its MIRV post-boost-vehicle (PBV ). The mis-
siles employed an inertial measurement unit, located
on the PBV plus a “star-sighting or celestial” fix before
RV separation to update the inertial system concern-
ing submarine launch location. The resulting system
accuracy was about (CEP=1,250 feet).26 However,
problems arose during development.
The original, Mark 4 nose-tip design was based
on a boron-carbide-coated graphite material, which
ablated at an uneven rate in some tests. When this
happened, the RV could assume an angle of attack
and fly off its ballistic path, causing significant impact
errors. Redesign of the nose provided a metalized
center core (or plug) surrounded by carbon-carbon to
cause more even ablation. The degree to which this
would restore accuracy was uncertain; therefore, Vice
Admiral Bob Wertheim, now in command of Trident
development, directed Lockheed Missile and Space
Company (LMSC) to initiate design of a maneu-
vering Mark 5 MRV (initially known as Mark 500
“Evader”) as an alternative means for restoring accu-
racy and providing defense evasion capability.
Plans for the longer range Trident-II D-5 mis-
sile, capable of delivering either the ballistic Mark 4
or Mark 5 MRV, were submitted to Congress on 6

169
william yengst

March 1981. The Mark 5 was approved and its funding


was released by Deputy Secretary of Defense, Frank
Carlucci in December 1982. The intended develop-
ment schedule was to achieve “Initial Operational
Capability (IOC) by January 1990.”27 The D-5 missile
would use the new AIRS inertial measurement unit
plus the PBV “celestial” (or GPS) fix to achieve an
accuracy of (CEP=300–400 feet).26 The test program
would include 19 full-scale launch-pad flights plus 9
Trident submarine launches to verify operation. By 30
September 1985, 15 flights had been completed (11 suc-
cessful, 1 partial success, 2 failures, and 1 destroyed by
range safety).28
Little was written about the Mark 5 MRV design.
It would have the same base diameter as Mark 4 and
could be used on either Poseidon C-3 or Trident C-4
missiles. It would be capable of ranges up to 4,000
nautical miles.27 One description of the missile indi-
cated, “Trident D-5 is more sophisticated and can
carry a heavier payload.”9 The MRV featured a new
guidance system known as Reentry Inertial Measure-
ment Unit (RIMU) built by Honeywell, which was
constrained to fit the size and space of an earlier
design that employed spinning-mass gyros and a gim-
baled platform. RIMU employed Ring Laser Gyros
and accelerometers tested on the AMaRV flights.22
The MRV would carry the Navy W-88 (475 Kt yield)
nuclear warhead. “It was equipped with an ‘accuracy
adjunct,’ using a three-axis flap system, developed by
the newly merged Lockheed-Martin Missiles and
Space Company.” 30
The Mark-5 MRV design used an aluminum
structure, carbon-phenolic heatshield and metalized

170
lightning bolts

core nose tip (similar to Mark 4). The nose tip radius
was less than 2 inches but became more blunt due to
ablation during reentry.28 The MRV overall character-
istics were as follows:31

Weight, W . . . . . . . . . . . . . . . < 800 pounds

Diameter, D . . . . . . . . . . . . . . . . 21.8 inches

Length, L . . . . . . . . . . . . . . . . . 68.9 inches

Nose half-angle . . . . . . . . . . . . . . . . . . . 8.2 o

Ballistic Coefficient (W/CDA) . 2,050 lbs/ft.2

Because ABRES supported both air force and


navy technology needs, as evidenced by exchange dis-
closure meetings, it is not surprising the Mark 5 MRV
was similar in many respects to AMaRV.
Following a long series of negotiations and treaties
(i.e., SALT-I [1969–1972], ABM Treaty [1972], SALT-
II [1972–1979], INF Treaty [1987], Start-I [1991], and
Start-II [1993]), between the U.S. and Russian Fed-
eration concerning reductions in strategic offensive
forces, meetings were held in Moscow in 2002 lead-
ing to the “Moscow Treaty,” also known as Strategic
Offensive Reductions Treaty or SORT.32 The SORT
agreement updated the 1991 START-I agreement,
which called for reducing U.S. and Russian missile
forces to 2,200 deployed, offensive, nuclear weapons.
This cut the number of Mark 4 RVs or Mark 5 MRVs
carried on each Trident missile to 4 or 5 warheads. The
treaty constraints were not fully accepted by either

171
william yengst

country through 2007, but President George Bush


ordered reductions in U.S. warheads to just over half
the 2001 total.
In discussing Mark 5 MRV program develop-
ments with Bob Wertheim during the late 1980s, he
explained that Lockheed evaluated several alternative
design concepts.

• One vehicle concept used a moving mass


(several kilograms of heavy metal) on a
screw-driven or electric-powered radial
track inside the vehicle to provide roll con-
trol. The radial mass imbalance caused the
vehicle to assume a trim angle of attack
to generate roll torque. A second-moving
mass, located along the vehicle center-
line, could control the center of gravity
(cg) to produce pitch maneuvers. Thus,
internal moving masses could accomplish
flight trajectory changes without exter-
nal fins, flaps, or jets and with identical
external signatures as ballistic RVs. This
concept was proposed again in 1999 by
TRW Space and Missile Systems of San
Bernardino working with Textron Systems
Division (Wilmington, Del.) as a low-cost
method for converting ballistic RVs to a
maneuvering capability.33
• Another concept was known as “bent
nose.” This design was in fact, a biconic
with the forward 8- to 12-inch conic tip,
tilted several degrees from the main body

172
lightning bolts

axis. The resulting asymmetrical shape


caused the vehicle to seek an angle of
attack during reentry. Resulting lift was
cancelled by spinning the vehicle with
roll control to follow a tight spiral trajec-
tory. Turns could be made by slowing or
stopping the spinning motion. Finally, a
patent was granted for injecting electro-
philic gas through the bent nose to remove
free electrons in the plasma sheath sur-
rounding the vehicle during reentry. The
patent disclosure claimed the bent-nose
reduced plasma and prevented blackout
disruption.34

I found no evidence that these designs were devel-


oped or flown as full-scale tests.

Strategic Boost Glide Vehicle (SBGV)


The enthusiasm for glide vehicles with long-range
and high-payload capacity did not end with BGRV.
The USAF/SAC initiated a hypersonic-glide pro-
gram in the early 1980s as an alternative to the Min-
uteman ICBM system. It would permit bombers, such
as modified B-52H, to stand outside Soviet borders
while attacking targets beyond the range of land-based
ballistic missiles. The Strategic Boost Glide Vehicle
(SBGV ) would use the upper stage of Minuteman-I to
launch the vehicle from the bomber at an altitude of
about 68,000 feet, to achieve a speed of Mach 18, and
to sustain a glide range of 5,000 miles. The bomber and

173
william yengst

SBGV combination could reach targets at ranges up to


9,300 miles from U.S. airbases.
Martin Marietta, General Dynamics and Lock-
heed Missiles and Space Company prepared compet-
ing designs for SBGV. Although a final configuration
was not published, reports indicate it had a highly-
swept (75 degrees) delta plan form design controlled
by four vertical tail fins.35 A full-scale model of the
General Dynamics design is shown in Figure 22. The
SBGV program intended to use the DINS guidance
subsystem developed for AMaRV.

Figure 22. General Dynamics model of


the Strategic Boost Glide Vehicle

Thermal protection for SBGV was provided by a


carbon-carbon composite skin over a titanium struc-
ture. Approximate vehicle characteristics were as
follows:

174
lightning bolts

Length (L) . . . . . . . . 30.0 feet (w/o booster)

Launch Weight (W) . . . . . . . 25,000 pounds

Wing Span (S) . . . . . . . . . . . . . . . 134 inches

Warhead . . . . . . 2 or 3 thermonuclear bombs

The entire missile including its Minuteman booster


was 46 feet long. Of the three competitive designs,
the Lockheed design was selected; however, there is
no evidence that it was developed or flight tested.
Because of strategic weapon agreements between the
U.S. and Soviet Union plus the end of the ‘Cold War,’
SBGV never progressed beyond the design stage.

Long Range Research and


Development Program
Late in 1970, the air force moved the headquarters for
the Space and Missile Systems Organization back to
El Segundo. Aerospace Corporation closed its San
Bernardino Operations, and several hundred staff
members were faced with the unhappy prospect of
commuting 100 miles to work or moving. Five of us
(Bill Vance, Brian Henshall, Don Schutt, Ed Knight
and I) chose to join a new company, Science Appli-
cations Incorporated (SAI, later renamed Science
Applications International Corporation, SAIC). The
company, formed in San Diego by Dr. John Robert
(Bob) Beyster, had headquarters in San Diego (CA).
The five of us set up a temporary office in Redlands

175
william yengst

(California) during March 1971 to market analytic


studies to the air force but with no contracts; in
August we moved to San Diego. Under Bob Beyster’s
direction, our careers as military operations analysts
was about to expand.
During 1970–1971, Dr. Albert Wohlstetter, Henry
S. Rowen, Andrew W. Marshall, and other Penta-
gon senior security advisors noted the U.S. military
forces did not possess proper nuclear weapon capa-
bilities or sufficient doctrine to handle critical war-
fighting situations. The Defense Nuclear Agency
(DNA) and Defense Advanced Research Projects
Agency (still known as ARPA) jointly formulated
and funded a two-year program (1972–1974) with
multiple contractors to address these issues under a
study entitled, “Long Range Research and Develop-
ment (LRR&D).”36 The program, conceived by Ste-
phen J. Lukasik (ARPA) and Dr. Peter Haas (DNA)
with inputs by Fred Wikner, Joseph Braddock, Don
Hicks, Jack Rosengren, and Dom Paolucci, focused
on methods for precision targeting of enemy facilities
and forces but with minimum collateral damage and
unwanted civilian casualties. The study was organized
in four panels, supported by four primary contractors
and a number of independent consultants.
Albert Wohlstetter chaired a strategic alterna-
tives panel, Don Hick chaired an advanced technol-
ogy panel, Jack Rosengren headed a munitions panel,
and other senior-level personnel from the Office of
the Secretary of Defense (OSD) and military services
filled a fourth multidisciplinary panel as needed.36
SAIC and Braddock-Dunn-&-McDonald (BDM)
were responsible for the majority of panel-directed

176
lightning bolts

studies. During the program, I worked closely with


Steve Lukasik for more than two years and grew to
admire his ability to see through problems to practical
solutions. We became friends and have enjoyed work-
ing together for over 35 years. Although Joe Braddock
and his staff at BDM were SAIC competitors on sev-
eral directed studies, we also became good friends, and
I still use a leather notebook he gave me.
With respect to maneuvering reentry technol-
ogy, it is useful to review a few studies we preformed
under the LRR&D program. Drawing on experience
with MRVs, I prepared a study of alternative meth-
ods for achieving high precision delivery of subkiloton
nuclear or high explosive warheads to destroy hard
enemy targets with minimum collateral damage and
casualties.

Precision Guidance of Long Range Missiles: At the


first LRR&D meeting held in La Jolla (CA) at the
SAIC main conference room in early 1973, I presented
descriptions of several methods for guiding ballistic
and cruise missiles with ranges of several thousand
miles to achieve delivery accuracies measured in tens
of feet. Each method was supported by a statistical
error analysis, assessment of technology risks, and
estimates of times and costs required for development.
That was probably the first time Albert Wohlstetter,
Steve Lukasik, Bob Beyster, Joe Braddock, and other
panel members learned details of the ABRES MRV
development programs.

Replay of the Cuban Missile Crisis: One of several


hypothetical warfare situations our group addressed

177
william yengst

during the studies was a replay of the Cuban Missile


Crisis. To review history:
On 14 October 1962, a U-2 surveillance aircraft
photographed Soviet MRBM missile sites that were
under construction in Cuba. By the morning of 17
October, 16 SS-4 missiles were installed on launch
pads, when General Curtis Le May, Commander in
Chief of SAC, called upon President John F. Ken-
nedy to launch immediate air attacks to destroy the
sites. By 19 October, four sites were fully operational,
and new SS-5 IRBM sites were under construction.
SAC placed U.S. nuclear missile and bomber forces on
DEFCON-2 (high strategic alert). Some B-52 bomb-
ers carrying nuclear bombs were sent aloft on airborne
alert. Several Squadrons of B-47 medium bombers
were dispersed to civilian airfields. The Joint Chiefs of
Staff formulated and seriously debated five response
options.17

1. Do nothing.
2. Use diplomatic pressure to have the mis-
siles removed.
3. Air attack against the sites to destroy the
missiles.
4. A full military invasion of Cuba.
5. A naval blockade (or redefined “Quaran-
tine”) of the island.

USAF SAC analysts assessed the missiles could


be destroyed with confidence by an air attack with
nuclear weapons but 10,000 to 20,000 Cuban civil-

178
lightning bolts

ians would be killed and several times as many might


be injured.37
The problem was, “How to destroy the missiles
without suffering an immediate horrific counterattack
against American cities?” It was impossible to perform
air strikes, using conventional weapons, with high con-
fidence that all the missiles would be destroyed before
any could be launched. The air force and navy did not
possess precision conventional (high explosive) weap-
ons such as laser-guided Maverick, electro-optical TV
Walleye or Paveway ‘smart-bombs’ (e.g., accuracies of
CEP~10–15 feet) until several years later. Commando
raids would take hours to implement, could be spotted
during penetration to targets, required precision tim-
ing at several sites, and would not guarantee success.
Higher confidence attacks could be achieved using
tactical nuclear weapons, such as F-4 or A-7 aircraft
delivering W-28 nuclear gravity bombs (CEP > 300
feet and yields of 70 Kt or 350 Kt), Matador MGM-1
cruise missiles (CEP>100 feet and yields of 11 Kt or
47 Kt) or Pershing-I ballistic missiles with W-50 war-
heads (CEP>1,300 feet with yields of 60 Kt or 400
Kt). Matador and Pershing-I were newly operational
in 1961 and could destroy the targets but only with
poor accuracy and large yields. They would cause sig-
nificant collateral damage and high casualties among
generally friendly civilian populations. Missiles with
high accuracy conventional or subkiloton yield nuclear
warheads were needed to accomplish precision low
collateral damage and minimum casualty attacks.

179
william yengst

Population Variations and Casualty Uncertainties:


Bill Vance performed two studies to assess variations
in civilian populations of major cities that could lead
to large uncertainties in collateral damage, even with
precision delivery attacks. The first study used the city
of San Diego as an example (since data was available)
and showed hourly changes in civilian populations of
downtown areas and around major military facilities,
such as airfields and naval bases. As expected, popula-
tions varied from a few hundred people per square mile
during night hours to over 60,000 people per square
mile during daytime (particularly at rush hours). He
selected several representative targets in the city and
computed potential casualties as a function of attack
weapon explosive yields, assuming perfect delivery
accuracy.
In a second study, Bill evaluated several German
cities and found the average heights of downtown
buildings had increased from three or four stories
before World War II to over ten stories by the early
1970s, due to reconstruction following the war. Growth
in height of modern buildings plus use of light weight
construction materials made the new structures more
vulnerable to nuclear weapon effects and fires. Results
of Hiroshima and Nagasaki nuclear attacks plus
Nevada weapon tests proved that more casualties were
caused in built-up areas by falling debris and fires
than by prompt blast, thermal, radiation, or fall-out
weapon effects. Therefore, civilian populations, con-
centrated in high-rise structures, were likely to suffer
significantly higher collateral casualties from nuclear
attacks than previously assumed based on census data.
Similar conclusions were obtained for battle-

180
lightning bolts

field scenarios. Large yield weapons used to destroy


spread-out (nuclear posture) enemy forces resulted in
much higher casualties and large areas where friendly
forces could not enter due to radiation contamination
to secure facilities or advance lines of engagement.
Precise delivery of low-yield weapons that could cra-
ter targets without causing high casualties or cause
unwanted infrastructure damage became a technol-
ogy challenge.38 The LRR&D studies led to missiles
with maneuvering vehicles having terminal guidance
for precision delivery of conventional or subkiloton
nuclear warheads. The Pershing-II missile was the
first U.S. nuclear weapon system to incorporate these
goals.

181
chapter 7

Pershing-II MARV
Development and Test
The Pershing-I missile was developed by Martin Com-
pany (Orlando, Florida) in 1958–1960 for the U.S. Army
Redstone Arsenal at Huntsville (Alabama). It featured
a two-stage, solid propellant booster launched from a
modified M-474 tracked vehicle designed for off-road
operations throughout Western Europe. Its ballistic
RV carried the large W-50 nuclear warhead (yields of
60, 200 or 400 kilotons of TNT) to a maximum range
of 460 miles. However, it had relatively poor accuracy
(Circular Error Probable of 1,310 feet).1 Battlefield
vulnerability was a great concern to the U.S. Army,
because its operations required four-tracked vehicles
(size of main battle tanks) per missile, a large num-
ber of personnel, and over an hour to park in order to
erect the missile, to lay cables between launch control
vehicles and to countdown for launch.

183
william yengst

While Pershing-I gave the army experience with


mobile operations, the missile was expensive, the large
yields were greatly overpowered for battlefield mis-
sions, and accuracy was too poor to employ low yields
effectively against hard targets. When deployed in
Western Europe, its short range prevented Pershing-I
from reaching targets deep in Warsaw Pact nations or
the Soviet Union. To complicate matters, the Soviet
Union deployed its new Pioneer (SS-20) mobile-inter-
mediate range (2,700 miles) ballistic missiles through-
out Warsaw Pact nations in 1976. Its range, plus the
ability to deliver two MIRV warheads, represented a
threat, which could deliver greater firepower against
all European NATO nations. Consequently, in March
1975, the U.S. Army and Martin Aircraft established a
task force to initiate definition of Pershing-II.
Sensing a missile force gap in Europe, the U.S. Air
Force, with support from General Dynamics Convair
(San Diego), proposed an alternative weapon system,
the Ground Launched Cruise Missile (GLCM). Four
GLCMs (i.e., modified Navy BGM-109 Tomahawk
cruise missiles) would be carried and launched from a
large-wheeled vehicle. They carried W-80 nuclear war-
heads (yields of 5 and 200 kilotons) and could fly to a
range of 2,500 miles. With TERCOM map-matching,
terminal guidance, they could achieve an accuracy of
(CEP~260 feet).2 The cruise missile, shown in Figure
23, would be deployed and operated from six U.S. Euro-
pean air bases in a manner similar to the Ranger system
described in Chapter 6.3

184
lightning bolts

Figure 23. Tomahawk cruise missile over southern California

Pershing-II Cost and Operational


Effectiveness Analysis (COEA)
Realizing there would be competition between Persh-
ing-II and GLCM for the same missions, Congress
directed the Defense Nuclear Agency (DNA) in 1977 to
perform a “Cost and Operational Effectiveness Analy-
sis (COEA), of the two concepts. The implication was
that only one weapon would be developed for opera-
tional deployment. Dr. Peter Haas, Director of DNA,
accepted a rapid response proposal from Science Appli-
cations International Corporation (SAIC) to perform
the study. Jack Goldstein, a retired U.S. Army Colonel
with Pershing-I experience—who joined SAIC late
in 1977—would manage the study and our “Advanced
Concepts Division” would perform the technical
evaluation.
Pershing-II was a two-stage, solid propellant mis-
sile launched from a wheeled transporter-erector-
launcher, and although its design requirements had not
been fully established, an advanced development firing
was conducted by the Army Ballistic Missile Agency

185
william yengst

and Martin Company on 18 November 1977.4 Within


a week after receiving the COEA contract (December
1977), Jack Goldstein and I met at SAIC’s Huntsville
office. DNA’s contracting officer technical representa-
tive (COTR), Major Michael E. Montie who would
sponsor the evaluation, joined us to define the intended
project. Mike was a bright, young, aggressive army offi-
cer, decorated Vietnam War hero who walked with a
slight limp and had an infectious smile and highly ener-
getic attitude. He was particularly efficient in arranging
meetings and handling contractual tasks. Jack, being a
retired army colonel, took a more serious and business-
by-the-book approach to the project, knowing that
Congress wanted fully documented results by Septem-
ber 1978.
After Mike explained conditions and objectives of
the COEA, we drove to Redstone Arsenal where we
met Colonel Larry Hunt, Army Director of Pershing-
II development for the army’s Ballistic Missile Agency.
Not surprising, Jack and Larry were long-time friends,
dating back to their days at West Point. Therefore, much
of our meeting was consumed with recounting previ-
ous relationships and stories. However, Mike explained
that the air force (sponsoring development of GLCM)
insisted that the Defense Intelligence Agency (DIA)
should be responsible for defining Soviet and War-
saw Pact defensive capabilities and providing “official”
target lists for the various communist countries. He
agreed to arrange a mid-January 1978 meeting for us
at DIA Headquarters in Washington (D.C.) to obtain
its inputs. Larry Hunt then gave us an intense lecture
to stress, “the study was to be totally objective and was
to fully answer critical questions posed by Congress!”

186
lightning bolts

Later, I learned that results of our study would also


be used to provide justification for deployment of the
selected system throughout Europe and NATO.
I left the Huntsville meeting with two important
and time-urgent action-items:

1. Determine the radar signatures for both the


Pershing-II reentry vehicle and Tomahawk
cruise missile. This would be a major point
of contention concerning in-flight surviv-
ability of both weapons when penetrating
enemy defenses.
2. Evaluate and recommend a design flight
range for Pershing-II. Army staff offi-
cers wanted a range of roughly 700 miles.
Department of State representatives advo-
cated a shorter range to avoid stressing
negotiations with the Soviet Union over
arms control. Army operational command
personnel preferred a range of about 500
miles to permit a smaller, more mobile mis-
sile capable of being deployed and operated
throughout Europe under all-weather, bat-
tlefield conditions.

Jack, Mike, and I flew directly from the Huntsville


meeting to Orlando (Florida) to be briefed by Mar-
tin Marietta on preliminary Pershing-II development
work. At the Martin facility entrance, we were met by
Vice President and General Manager, Robert (Bob)
Whalen, my former associate at ‘Aerospace Corpora-
tion.’ He had worked on MBRV and fully understood

187
william yengst

maneuvering technology. It was our turn to talk about


previous experiences and technical subjects, while he
showed us a mock-up of the Pershing-II system and
explained its features.

Pershing-II Baseline Design


The design objective for Pershing-II was to give the army
the capability to destroy large battlefield deployments
of troops and military vehicles, using the W-85 nuclear
warhead (yields of 5–80 kilotons) but also to destroy
hard underground command bunkers and other targets
with the W-86, earth-penetrating warhead (EPW with
a yield of 1 Kt). Therefore, the two warheads could be
interchanged at field deployment sites. Both warheads
could be delivered by the same MAneuvering Reentry
Vehicle (MARV ), having active radar terminal guid-
ance to achieve high accuracy. The missile is shown in
Figure 24.

Figure 24. Pershing-II missile showing the MARV payload

188
lightning bolts

The Strategic Arms Limitation Talks (SALT-II)


Treaty with the Soviet Union restricted Pershing-II to
be launched from previous but upgraded Pershing-I
vehicles. Therefore, the prime mover for the missile
was either a U.S. M-983 HEMIT or German M-1003
vehicle. A new inertial guidance system (considerably
smaller than the Pershing-I system) was designed by
Singer-Kearfott for Pershing-II and it permitted the
launcher to park, self-align by ‘gyro-compass’ and pre-
pare to fire within minutes. The Goodyear Aerospace
Company (RADar Area Guidance or RADAG) sys-
tem was selected for terminal navigation. It scanned
the target area with J-band (i.e., 10 GHz to 20 GHz)
frequency radar during reentry, compared the image
with a pre-stored radar map, updated the inertial posi-
tion and permitted the MARV to steer to the target.
The overall missile final design had the following
chartacteristics:1

• Length: 34.8 feet


• Diameter: 40 inches
• Engine: two-stage solid propellant
• Launch Weight: 16,513 pounds
• Accuracy: CEP~150 feet
• Launcher: M-1003 TEL
• Range: 1,150 miles
• Guidance: Singer Kearfott Inertialplus
Goodyear Aerospace active RADAG
terminal

189
william yengst

MARV Design Characteristics:


The MARV was designed in three structural and func-
tional sections as illustrated in Figure 25. Its weight
was close to 1,400 pounds with a length of 144 inches.
Because precise design data was not available, dimen-
sions of the three sections described below were scaled
from missile photographs.

1. Conical nose containing the RS (Radar Sec-


tion): This included the radar unit and
antenna, altimeter, digital correlation unit
(DCU), ballast, and cable adapter to send
steering signals to the G&C/A (Guidance
and Control/Adapter) section.1 The cone
was 31 inches long, with a one-inch, nose-
tip radius (aft-radius of 8.5 inches) and
conic half-angle of 15 degrees. Its ablative
heat shield contained the RADAG “real
aperture’ radar antenna.”5
2. Nearly cylindrical Warhead Section: This
section contained the W-85 warhead, with
space within the section for the inertial
measurement unit (IMU), warhead-bat-
tery power and connecting cables.1 The
W-85 could be replaced by the six-foot
long (W-86 EPW, 6-inch diameter and
400-pounds), mounted on brackets within
the section.6 The section was 78 inches
long and had a small (3.5 degree) slope to
a base radius of 13.5 inches.

190
lightning bolts

3. Frustum-flare G&C/A Section: This sec-


tion contained two separate parts: guid-
ance control actuators (G&C) and booster
adapter (A). At the forward end of this sec-
tion, a quick access joint attached it to the
warhead section. At the aft end, a grooved
V-band connected the MARV to the mis-
sile booster.1 The section was 35 inches
long with a base radius of 30 inches. Tri-
angular fins on the G&C portion were 24
inches long and extended 8 inches wide at
their tips. They were about one inch thick,
coated with carbon-phenolic for heat pro-
tection. The adapter (12 inches long) flared
to the 40-inch diameter of the booster.

The adapter flare extension resulted in an over-


all length of 156 inches, and the reported total weigh
was 1,467 pounds.3 Separation of the MARV from the
boosters was accomplished by a linear-shaped charge
ring at the G&C/A interface connection directly
behind the four aerodynamic fins. A protective col-
lar on the outer surface of the adapter was mounted
over the shaped-charge and provided personnel safety
during G&C/A handling.1, 3 Because the P-II MARV
flew only to medium range (i.e., 1,150 miles), its heat
shield was somewhat thinner and lighter weight than
that of AMaRV. Estimates suggest the heat shield was
1/8 inch-thick aluminum covered by 3/8 inch-thick
carbon-phenolic and weighed approximately 650
pounds. The drag coefficient was about (CD~0.2), base
area of (A=4.91 square feet), and weight of (W~1,400

191
william yengst

pounds), to give a ballistic coefficient of (W/CDA=1,425


pounds per square foot). This was sufficiently high to
assure a Mach 8 velocity, ballistic reentry trajectory,
but without overly stressing the nose tip or heatshield.
The vehicle retained sufficient energy below 50,000
feet altitude to perform a 25-g pullout maneuver and
30-mile range extension (during which RADAG per-
formed its map-matching accuracy update), or it could
perform a high-g evasion maneuver.

Figure 25. Approximate Pershing-II MARV design layout

DIA Meetings
Our mid-January 1978 meeting with the Defense
Intelligence Agency (DIA) was less helpful than
Mike, Jack or I had hoped. Two subjects were on the
agenda: (1) in-flight survivability of Pershing-II versus
Tomahawk (GLCM) and (2) distribution of important
targets in Warsaw Pact nations. The meeting began
with a blunt declaration by a DIA senior analyst, who
claimed, “Warsaw Pact radar systems would never see

192
lightning bolts

Tomahawk cruise missiles.” He declared the radar cross


section of the cruise missile, as seen from the forward
hemisphere, was under 0.1 square meter (roughly one
square foot), similar to the size of a “sparrow.” With
no measured data but only analytic estimates, we con-
tended the cross-section of Tomahawk was closer to
a few square feet nose on, increased near the beam to
over 10 square meters (i.e. about 100 square feet) and
reduced somewhat through the rear hemisphere. The
wings and tail surfaces would have a large radar cross
section, particularly if the missile flexed or vibrated
during flight. By contrast, experience with observable
signatures of reentry vehicles suggested Pershing-II
would have an order of magnitude smaller radar cross-
section throughout its reentry trajectory.
This technical disagreement became heated and
persisted through the meeting and until the end of
the evaluation. DIA personnel insisted Soviet radars
would never detect Tomahawk (i.e., even with mul-
tiple cross-range radars deployed). They claimed visual
and thermal observable signatures would be even less
likely to be seen because the cruise missile flew very
low (<100-foot altitude) and high velocity (~550 mph).7
Upon return to San Diego, I enlisted Terry G. Smo-
lin to help review the British experience in detecting
and defending against 2,300 German V-1 buzz-bomb
cruise missiles fired against London between 13 June
1944 and March 1945. British success in finding and
intercepting the 25-foot long, 400 mph, V-1s was low
in June 1944 but within five weeks of experiencing the
threat, they redeployed antiaircraft artillery (AAA),
added ground observers, developed aircraft intercep-
tor search-and-attack tactics to engage the missiles.

193
william yengst

By the end of July 1944, they intercepted 60 percent


of the V-1s, and by March 1945, over 90 percent of
the missiles were prevented from reaching targets.8
We concluded that Tomahawk would eventually be
detectable and could be destroyed by advanced defen-
sive systems. However, we noted that Tomahawk
demanded a different type of defense than was needed
to stop Pershing-II MARV.
The second controversial issue with DIA was
agreement on a Warsaw Pact target list, and target
distributions from the borders of European NATO
countries. DIA assumed that only military targets
such as army, navy and air force headquarters, force
deployments, weapon depots, command and control
centers, and logistics facilities should be considered.
We argued that some civilian facilities such as air-
fields, railroad centers, vehicle transportation choke
points (bridges and tunnels), and petroleum refiner-
ies or storage sites should also be included. The result,
after several iterations and discussions, was a compro-
mise list with over 1,500 facilities, the large majority
in East Germany, western Poland and Czechoslovakia
Republic.
Two of our SAIC scientists performed unique
weapon effectiveness analyses, using the agreed target
lists. Peter R. Ward prepared a computer model and
performed a parametric evaluation of Pershing-II and
GLCM weapon yield options and delivery accuracies
against all the targets. It showed that a single weapon
of either type, because of high accuracy (CEPs of 150–
260 feet) could insure a high probability of destroying
more than 90 percent of the targets. However, Persh-
ing-II with its W-85 (yield options of 5–80 kilotons)

194
lightning bolts

and low-yield (1 kiloton) EPW W-86 warhead could


destroy more hard targets with considerably less col-
lateral damage and civilian casualties than could be
achieved by GLCM with its W-80 warhead. Lonnie
Nesseler invented a computer code known as DCAPS
(Dual Criteria Aimpoint Selection), which permitted
an extension of the analysis. This code generated off-
set aimpoints for each weapon, simultaneously maxi-
mizing target damage while minimizing damage to
nearby civilian facilities and population centers. When
the analysis was presented for review, Steve Lukasik
(Director of DARPA) agreed to fund development of
DCAPS for use by NATO staff officers to accomplish
European Command (EUCOM) target planning.
Further analysis indicated that Pershing-II
deployed in West Europe, with a range of 700 miles,
could cover nearly all the important targets, includ-
ing those in the Soviet republics of Estonia, Latvia,
Lithuania, Belarus, and Western Ukraine. Hence, a
range of 700 miles would be adequate for accomplish-
ing intended missions; however, we argued that the
maximum range should be 1,000 nautical miles (1,150
statute miles). This recommendation was presented in
a separate report for the Defense Nuclear Agency,
Congress, and Army Redstone Arsenal early in 1978.
It was based on three considerations:

• When launched from forward deploy-


ments, Pershing-II with a 1,150-mile range
could reach Moscow to hold the Soviet
capital at risk. This would provide a deter-

195
william yengst

rent to Soviet aggression and use of its


SS-20 long-range, mobile missiles.
• If NATO was unable to prevent Soviet
invasion and penetration of central
Europe, Pershing-II with an 1,150-mile
range could still reach all major logistics
and rail centers along the Soviet/Warsaw
Pact borders from West Europe.
• In the event Pershing-II did not achieve
its design weight goals for MARV, there
would be sufficient range that could be
exchanged for higher payload weight.

The recommendation was approved and became


the official design range for Pershing-II.

Albert Speer’s Precedence


As the September 1978 deadline for the Pershing-II
COEA Report approached, I realized there had already
been a ‘historical precedence’ for evaluating the choice
between a ballistic and cruise missile development.
During two weekends, I investigated conditions in
Germany between 1943 through early 1944. Hitler had
not made up his mind about mass production of V-1
or V-2 in May 1943. He set up a special commission
known as Long-Range Bombardments Development
Committee, which would visit Peenemunde, acquaint
itself with the facilities, witness test demonstrations
and only then, decide what should be done about pro-
duction of V-1 or V-2.9

196
lightning bolts

The V-1 ‘buzz-bomb’ achieved initial, air-launched


flight test successes in September 1943 after several
development failures and death of one test pilot. V-1
used a simple, low cost autopilot for height and speed
control but relied on two (fore- and aft-mounted)
pendulums, plus gyrocompass damping for pitch con-
trol.10 The original design was unstable at low speeds
and also in roll control, although both problems were
corrected before production. Meanwhile, the V-2
(A-4) rocket development, a more expensive program,
also suffered several engine explosions and launch
flight-control stability failures.
On 7 July 1943, General Walter Dornberger and
Werner von Braun obtained a personal interview
with Adolph Hitler and Albert Speer, Minister for
Armament and War Production.11 At the time, Hitler
favored and gave top priority to the low cost and sim-
pler V-1; therefore, he ordered slowing the develop-
ment of V-2. The May 1943 production rate of 437 V-2s
was cut to only 86 test missiles in July.9 At the meet-
ing, Albert Speers was directed to determine which
weapon development (V-1 or V-2) should be contin-
ued into full production.
When the Allies bombed Peenemunde (17 August
1943) and mounting military losses occurred along the
Russian front, Hitler ordered work cutback on V-2 to
save funds.12 However, two factors changed this deci-
sion by the end of 1943.

• The British began a campaign of bombing


V-1 launch sites in France, Belgium, and
the Netherlands in addition to develop-

197
william yengst

ing defenses (i.e., AAA, aircraft intercep-


tors, barrage balloons, coastal radars, and
observer networks) to detect and intercept
the V-1 ‘buzz-bombs.’
• Albert Speer’s report on V-1 and V-2
became available in early January 1944.
It highlighted two facts: the two weapon
systems used different pre-launch surviv-
ability concepts and required dramatically
different defenses to intercept delivery.
The British had no anti-ballistic mis-
sile defense. Both weapon systems could
be made highly survivable (V-1 through
multiple, quickly moved launch sites or air
launched from He-111 bombers and V-2
through super-hard concrete citadels or
mobility). General Dornberger strongly
advocated mobile V-2s and built proto-
types of the first vehicle launchers.

Albert Speer and Hitler met again with Werner


von Braun, General Dornberger, Field Marshall Her-
man Goering, and other senior officers at Peenemunde
on 8 January 1944. They reviewed a (1:100 scale) model
of the planned V-2 citadel launch facility to be built by
Organization Todt at Watten (France) on the English
Channel. Hitler liked the hard structure concept but
wanted three facilities built (i.e., Watten, Wizernes, and
Sottevast). Watten had been surveyed and initial con-
struction started under the code name “NorthWestern
Power Plant.” Hitler approved its construction and
energized V-2 production; although changes were still
being made to the missile design.13 The Fuhrer alone

198
lightning bolts

reserved the right to designate targets that would be


attacked by either weapon type.9 He only specified
London, Paris (briefly), and Antwerp.
I prepared an Appendix for the Pershing-II
COEA Final Report to summarize the German his-
tory and used it to emphasize the military and politi-
cal advantages of deploying both Pershing-II and
GLCM. Congress accepted the recommendations and
between 1981–1989, the U.S. Army built and deployed
108 Pershing-II launchers with a force of 276 missiles.1
In parallel, the USAF deployed 464 GLCMs on 116
launchers at bases in England, Belgium, Italy, Neth-
erlands, and Federal Republic of Germany (FRG)
between October 1983 and October 1988.14

EPW Development and Cancellation


Starting in 1978, Department of Energy weapon labo-
ratories, Los Alamos National Laboratory (LANL)
and Lawrence Livermore National Laboratory
(LLNL), designed a prototype W-86 Earth Penetrat-
ing Warhead (EPW ) for use on Pershing-II.15 A series
of earth penetration tests were conducted during
1979–1980 at Tonopah Test Range (Nevada) to mea-
sure the stress and strain on electronic components
needed for the warhead arming and fusing. One test
involved a Pershing-1A missile, launched from Fort
Wingate (New Mexico) with a simulated EPW to
impact at White Sands Missile Range. The test device
contained arming and firing electrical circuits, a deep
burial fusing component, and telemetry package but
no nuclear device. Upon impact, the projectile trav-

199
william yengst

eled a diagonal trajectory through 187 feet of soil and


came to rest at a depth of 108 feet.16 The hardened
components were recovered and confirmed; the EPW
design could sustain high impact environments.
Additional EPW tests were conducted at Green
River, Tonopah, and in Alaska. Specifically, the tests
proved the W-86 design could survive the decelera-
tion of impact (about 10,000-g peak) without destroy-
ing arming, fusing, and firing components. The LLNL
design was favored and was prepared for production
by the end of 1980.15, 17
Our COEA Report included an evaluation of
W-86 EPW utility and concluded it should be devel-
oped. However, DIA claimed there were very few
super-hard targets that would require the weapon (e.g.,
half-dozen command and control centers, national
leadership bunkers, and weapon storage facilities).
Consequently, it argued there was little justification
for developing and fielding an EPW. To counter this
position, Bill Vance and I spent weeks searching maps,
literature, and strategic-target facilities in Warsaw
Pact countries that would require either an EPW or
very large yield surface-burst weapon for destruction.
We identified roughly three-dozen facilities and some
dating back to World War II, which were thought
to be in use by the Soviet military or its allies. DIA
refused to accept our list as valid because the status of
many facilities was not fully verified.
One facility in particular was the East German
Kohnstein Mountain underground Nordhausen com-
plex, constructed (1936–1938) by I. G. Farben with
slave labor and used for production of V-1 and V-2
missiles.9 The enormous, deep underground complex

200
lightning bolts

had two tunnels (A and B), each 2,600 feet long and
with a combined floor space of over 500,000 square
feet. It was captured by Soviet troops in March 1945
and was reportedly sealed after the war. However, an
escaping East German border guard in 1978 reported
the facility was being used to store hundreds of heavy
tanks, artillery, and armored vehicles for several divi-
sions of troops in event of a war with NATO. DIA
could not confirm the claim; hence, Nordhausen, plus
several World War II command and control bunkers
on our list that were reportedly being used by the East
German military, were set aside as not being verified.
At a project review in mid-1978, Department of
State personnel expressed concern that the EPW’s low
yield (1 kiloton of TNT) and underground detonation
with little (or no) radiation and limited fallout effects
might lower the nuclear threshold. They argued mili-
tary leaders might be encouraged to use the weapon
in time of conflict, whereas they would be reluctant
to use the large-yield, W-85 warhead. Finally, the
Department of Energy weapon laboratories admit-
ted that development of EPWs might require new
underground nuclear tests. This was counter to State
Department efforts to negotiate a Comprehensive
Test Ban Treaty (CTBT).
Although the P-II EPW (W-86) option would
cost only about 4 percent more for development, test-
ing, and ten-years of operation than using only the
W-85 warhead, Congress and the army cancelled the
option in September 1980.1, 6 All work on the W-86
EPW was stopped. I was disappointed, because the
technical risk was limited and military utility of the
highly accurate, low yield weapon was exactly the type

201
william yengst

of weapon our 1973 LRR&D study advocated. EPW


was a unique feature of Pershing-II, which could not
be duplicated by a cruise missile. I very quickly learned
that decisions by our civilian leadership are foremost
political and not necessarily the best technical or cost
effective choices when other important priorities and
programs are considered.
The concept of an EPW and its technology
did not die with this decision. In 1992, the Defense
Nuclear Agency funded SAIC to study “Hard Tar-
gets That Could Not Be Destroyed by Conventional
Weapons.”18 This study showed that throughout his-
tory, there had been numerous important military and
leadership targets that were sufficiently hard or deep
underground so that they could not be destroyed by
conventional explosives. Cordelia (Deedee) White
and I performed an empirical analysis of all manner
of delivery weapon systems (U.S. and foreign) used
against hundreds of targets dating from the U.S. Civil
War, World War I, and World War II. We concluded
the U.S. military needed a small force of “state-of-the-
art,” accurate, deep-earth-penetrating weapons armed
with high explosive or low-yield, nuclear weapons to
hold super-hard targets at risk. The intent was to pro-
vide the military with a hard-target kill capability and
require enemy nations to expend enormous funds and
resources to construct and harden facilities against
attacks.
We were invited by DNA to brief the study at
USAF SAC Headquarters (Omaha) and to the Joint
Chiefs of Staff (Pentagon) during March 1993. Sup-
porting technology briefings by the Department of
Energy weapon laboratories helped convince Con-

202
lightning bolts

gress to approve development of a shallow EPW


(B-61 Model 11 tactical bomb) in August 1995. This
bomb was a modification of the B-61-Model 7 weapon
already in the U.S. weapon inventory, thus avoiding
new development and testing. Its design could pen-
etrate only 10 to 20 feet of soil following aircraft deliv-
ery, but with a sizable nuclear yield, it could destroy
underground targets to depths of over 300 feet.19

Selling of Pershing-II and GLCM


When the COEA study was complete (September
1978), a formal review of its results was prepared for
“general staff ” officers of the army and air force at the
Pentagon (Washington). A tight, two-hour agenda was
planned, including an introduction by Colonel Larry
Hunt, statement of objectives by Major Mike Montie,
summary of recommendations by Jack Goldstein, and
technical evaluation, which I presented. Senior repre-
sentatives of DNA, DIA, CIA, and State Department,
in addition to officers from the European commands,
were present. Nearly everyone in the crowded con-
ference room was startled when Mike took the floor
to make a brief announcement, “I intend to run this
review like The Gong Show! He was referring to the
1976 Comedy Central TV program, in which contes-
tants with varying degrees of talent performed for a
panel of celebrities. If their performance was consid-
ered below panel standards, a large gong was rung to
stop the performance. Fortunately, our COEA brief-
ings ran smoothly, with no gongs and few challenging
questions, and we finished within the two-hour limit.

203
william yengst

Shortly after the review, the Defense Nuclear


Agency, U.S. Army, and Air Force sponsored a technical
disclosure tour in the spring of 1979 to present results
of the COEA to European NATO countries. Three of
us (Mike Montie, Jack Goldstein, and me) spent two
weeks traveling to Stuttgart and Bonn (Federal Repub-
lic of Germany), Amsterdam (Netherlands), Mons
NATO Headquarters (Belgium), and London (Eng-
land), where Larry Hunt joined us, to present the analy-
sis and rationale for developing and deploying both Per-
shing-II and GLCM. In each country, large contingents
of senior military officers, and in a few cases, diplomats
or government leaders were briefed. At each review, we
received comments and advice on how best to present
the material to the next scheduled authorities.
During the two-week tour, my wife Betty and I
rented a Volkswagen Bug and drove autobahns from
one city and country to another. We attempted to fol-
low Mike and Jack, who drove a second vehicle and
led us to remote locations for lunch or dinner at places
they recommended based on years of service in Europe.
Although we had good times and accomplished consid-
erable sightseeing between meetings, we got lost several
times. Betty and I spent one rainy night in the loft of
an Amsterdam bar, when our hotel reservations expired
before we found its location.
At NATO Headquarters (Mons, Belgium), I recall
waiting nervously in the lobby of the Supreme Allied
Commander Europe (SACEUR), Four-Star General
Alexander M. Haig Jr., while he finish receiving com-
ments and positions on our study from Allied command-
ers. When invited into his elaborate office for less than
fifteen minutes, he quickly thanked us for the results,

204
lightning bolts

indicated his staff was satisfied and told us, “Good Job!
It will be approved.” What a relief !
Development, testing, production, and approval
by NATO allies for both weapon system deployments
took more than two years. Initial units began arriving
in Europe in October 1983.4 Despite numerous anti-
nuclear weapon protests at U.S. bases, the ability to
hold Moscow and other large Russian cities at risk from
several directions fostered a willingness by the Soviet
Union to sign the Intermediate-Range Nuclear Forces
(INF) Treaty in October, 1988. In 1985, Jack Goldstein
and I preformed a follow-on study for U.S.N. Captain
Dennis West of DNA to evaluate basing of GLCM on
Guam to deter possible North Korean nuclear weapon
developments. Jack, Denny, and I briefed results of that
study at Pacific Command (PACOM) Headquarters in
Honolulu (Hawaii) for Admiral William C. Crowe Jr.
(later appointed Chairman of the Joint Chiefs of Staff ).
We also briefed the results at Republic of Korea Com-
mand Headquarters (Seoul, South Korea). No deploy-
ments were made, although a couple of squadrons of
fighter/bomber aircraft were reassigned by PACOM to
support South Korean war contingencies.
By 1991, all Pershing IIs and Soviet SS-20 missiles in
Europe were removed from NATO and Warsaw Pact
countries. They were ceremoniously destroyed. GLCM
transporters were destroyed, but some Tomahawk cruise
missiles were converted to conventional warhead capa-
bilities (i.e., Tomahawk Land Attack Missiles-TLAM,
BGM-109B and D).2, 14 Without the commitment to
develop and deploy these weapon systems, I doubt the
Soviet Union would have agreed to remove or reduce
its nuclear forces. Pershing-II and GLCM were helpful
in deterring and slowing further “Cold War” expansion.
205
chapter 8

Indian Agni-II and -III MRV


Developments
China’s first nuclear weapon test at Lop Nor on
16 October 1964, was an alarming event for India,
because the two nations had several contested border
areas including Aksai Chin in the Kashmir region.
Indian Prime Minister, Jawaharal Nehru, ordered the
Army to “take back our territories,” after open warfare
erupted along the border of Aksai Chin in 1962. China
improved its nuclear warhead capabilities through a
series of tests in the late 1960s and began develop-
ment of missile delivery systems by the mid-1970s. It
subsequently deployed long-range nuclear missiles
along the borders of Tibet and Mongolia with cover-
age of strategic targets in India.1 India’s first nuclear
test on 18 May 1974 was intended to send a message

207
william yengst

for China to deter further aggression. India’s develop-


ment of short-range battlefield and long-range strate-
gic missiles was aimed at countering Chinese as well
as emerging Pakistani nuclear weapon threats.
Both China and India initiated high priority
domestic missile research and development programs,
but their approaches were decidedly different. India
implemented a relatively independent and refresh-
ingly innovative program, because the U.S. and United
Nations placed sanctions on trading and exchanging
nuclear and military technologies with India after 1970.
Diplomatically upset about the sanctions, it refused to
join the United Nation’s Treaty on Non-Proliferation
of Nuclear Weapons. By contrast, China’s program was
heavily dependent on importing advanced technologies
with cooperative assistance from several nations plus
active espionage to obtain warhead and missile designs
as described in Chapter 9.

Agni (Fire) Missile Test-


Bed Development
In 1979, Agni Medium Range Ballistic Missile
(MRBM) system development was initiated. India’s
Defense Research and Development Organization
(DRDO) at Kanchanbagh Research Complex (Hyder-
abad’s Old City) contracted Bharat Dynamics Lim-
ited to produce the missile and its solid rocket booster.
The development intended to produce a test booster
for reentry vehicle technology under the Integrated
Guided Missile Development Program (IGMDP) ini-
tiated in 1983 (similar to the U.S. ABRES program).2

208
lightning bolts

The objective of IGMDP was to establish technology


for long-range missiles. Professor P. Narain Aggarwal,
director of the program, described the original missile,
as weighing 12 metric tons with a range of 430 miles,
carrying a payload of 100 kilograms (220 pounds) and
capable of providing 10 to 12 seconds of reentry flight
at velocities of 16,400 to 26,000 feet per second.3
However, Indian leaders openly acknowledged from
the beginning that Agni would eventually provide
their Air Force with a nuclear capability.
In 1985, Indira Gandhi accepted advice from
political and military leaders and directed DRDO to
increase the Agni payload to one metric ton (2,200
pounds). This was required to carry the first-genera-
tion, Indian nuclear warhead (i.e., BARC-developed,
boosted-fission device with a yield of 200 kilotons).3,4
At the time, Narain Aggarwal was building a super-
sonic wind tunnel and high-altitude test facility and
was reluctant to make changes because they would
delay the technology program. However, he launched
a redesign, and Agni was ready for testing with a bal-
listic RV on 1 May, 1989. Three tests were conducted
at the Interim Test Range in Chandipur on 22 May,
1989; 29 May, 1992 (partial failure); and 19 Febru-
ary, 1994. The long periods between tests permitted
modifications and ground-test verification; however,
delays were also caused by diplomatic pressure from
the United States and other nations to suspend the
program. An improved, smaller Agni-I missile with a
range of 700 kilometers (430 miles) was flight tested
on 25 January, 2002. It was intended for use in con-
flicts with Pakistan, which had armed with Chinese-
designed, short-range M-9 and M-11 missiles.5

209
william yengst

The test program verified several useful reentry


technologies, including a carbon composite heat-
shield, the first without a metal back-up. Aggarwal
explained, “It saved about one-third of the heat pro-
tection system weight.” A new “Portmanteau” radar
dome (or Radome) was also tested for reentry envi-
ronments (i.e., suggesting that a terminal guidance
unit was a potential design feature). Avionics for
MRV controls were also demonstrated.3 During the
tests, concepts for improving missile transporters and
reducing launch reaction time were evaluated.

Agni-II Missile Development and Test


Development of Agni-II was authorized in March,
1998, for the purpose of producing a two-stage Inter-
mediate Range Ballistic Missile (IRBM) with a range
of 2,500 kilometers (1,550 miles). The purpose of
Agni-II was to provide a “credible deterrence” against
the People’s Republic of China and Pakistan.4 How-
ever, its range permitted coverage of Iran, Afghani-
stan, several Arab Persian Gulf states, Iraq, most of
Saudi Arabia, Burma, Malaysia, Thailand, Cambodia,
Vietnam, and part of Indonesia.6 A few days after the
Indian underground, nuclear tests at Pokhran (i.e.,
three events on 5 May and two on 13 May, 1998), Dr.
A.J.P. Abdul Kalam, Director of DRDO, explained
that a plutonium-based, boosted-fission nuclear war-
head with a range of yields was under development
for the Agni missile.2 The high-yield (200-kiloton)
warhead primary was tested in the Pokhran-II event.

210
lightning bolts

The “all-up” reentry vehicle, including the warhead,


was expected to require a 500-kilogram (1,100 pounds)
missile payload.4
The ‘Kargil War’ (Kashmir) between Pakistan and
India (May-July 1999) caused India’s government to
direct DRDO to quickly “operationalize” Agni-II to
deliver the first-generation, Indian nuclear warhead.7
The missile would incorporate several innovative fea-
tures verified on Agni-I. It would be launched from a
highly mobile off-road vehicle or from a special rail-
road car. Unlike Agni-I, which took all day to pre-
pare for launch, Agni-II would be continuously active,
could be erected, and fired within 15 minutes. It would
achieve high accuracy, using a terminally guided MRV.
The first stage solid booster weighed about 9 tons
but had more energetic fuel than Agni-I. The second
stage solid booster represented a decided departure
from conventional design practice. It weighed 4.6 tons
and incorporated flexible engine nozzles to provide
thrust vector control during the powered flight trajec-
tory. This stage was allowed to burn out completely,
which means the velocity prior to reentry could be
more or less than needed for the intended target
range. Variations in burn time and thrust energy of
solid fuels due to manufacturing differences and envi-
ronmental (temperature) conditions, were sensed and
accepted. Compensation for separation velocity errors
was provided once the second stage was exhausted
by two methods: (1) rocket motor trail-off gases were
vented in the inter-stage structure and (2) multiple
thrusters were fitted to the MRV payload adapter
(known as the integrated High Altitude Motor or
HAM), if additional thrust was required. The ballistic

211
william yengst

Mark-1 RV thrusters were powered by 110–170 pounds


of liquid-fuel burning in small increments in appro-
priate directions to shape the trajectory to the desired
reentry conditions.4 The thrusters also provided lim-
ited maneuverability (trim velocity) during early
reentry and served as a Post-Boost Vehicle to deploy
penetration aids. These thrusters were controlled by
on-board sensors plus software, similar to a hot-gas,
attitude-control system.8
Terminal maneuvers were primarily intended
for accuracy correction and not for defense evasion.
One report describes the MRV as having an atti-
tude control system, aerodynamic fins, and possibly
(unconfirmed) optical or radar terminal phase scene-
matching correlation system. Initial accuracy esti-
mates predicted the CEP would be 130 feet, but test
results were closer to 330–650 feet.4 Heavy emphasis
was placed on accuracy; therefore, speculation sug-
gested the system relied on Global Position System
(GPS) updates. Dr. Abdul Kalam explained, “This
was not true! GPS might make the system vulnerable
to electromagnetic interference and it would not be
used for a nuclear weapon system.” However, GPS
was considered acceptable for Agni-II when a conven-
tional warhead was being delivered. Several conven-
tional warheads, including bulk high explosives (HE),
combined effects bomblets (CEBs), fuel-air-explosive
(FAE), and guided sub-munitions, were developed for
the missile.8
Agni’s project manager, Avinash Chander, reported
the original design was too complex, requiring 600
different electrical wiring channels (i.e., 15 miles of
wires and cables). To insure greater reliability and

212
lightning bolts

reduce costs, a standard commercial aircraft Data Bus


(MIL Standard-1553) was adopted for the navigation
system. DRDO sources claim that with the Data Bus,
the number of connections was cut dramatically for
circuit routing, mounting brackets, and wiring was
cut to one-eighth its initial length. Hardware weight
was reduced, software was simplified, and the system
became more rugged.8

MRV Characteristics
Agni-II has an overall length of 20 meters (65.6 feet),
booster diameter of one meter (39.37 inches), weighs
35,200 pounds, and carries an MRV payload that
weighs 2,200 pounds.8 The missile was first shown
on its mobile transporter-erector-launcher during a
1999 parade. An artist drawing of the Agni-II MRV
is presented in Figure 26. The payload has an overall
length of 4.2 meters (165 inches) of which the MRV
has a length of 155 inches from its nose tip to the base
of its fins. A flare adapter (described as a skirt) is 10
inches long and attaches the MRV to the booster sec-
ond stage. The High Altitude Thruster (HAT) noz-
zles are located inside this flare adapter, which has
an estimated weight of 110–150 pounds. Separation
of the MRV from the booster is probably between
the adapter and control fins (similar to Pershing-II),
although the adapter may be retained through reentry.

213
william yengst

Figure 26. Agni-II MRV layout

The forward MRV conic section, including the


nose tip, is 78 inches long and has a half-cone slope
of 7.15 degrees.7 The nose tip is a spherical hemisphere
with a radius of 5.9 inches, to generate separating
shockwaves during reentry, keeping body air friction
and heating down. It is made of multi-directionally
woven, reinforced carbon-carbon fiber composite.
Analysis showed that at Mach 8 velocity and small
angles of attack, the base flare section of the MRV
is inside the aerodynamic flow, resulting shock waves.
The heatshield is a carbon composite proven on the
Agni-I ballistic Mark-I RV. It is manufactured in five
sections, each of which is made of two-layers of com-
posite carbon-phenolic filament material. The objec-
tive is to keep the internal temperatures below 60
degrees Celsius (140 degrees Fahrenheit), protecting
the warhead and electronic components. Abdul Kalam
reported that complete reentry, hypersonic flow was
simulated in India’s Computational Fluid Dynamics
laboratory to minimize required heat protection and
weight.

214
lightning bolts

The forward half-cone angle is a few degrees less


than most maneuvering vehicles, which helps offset
high frontal-drag of the blunt nose. The diameter of
the conic section aft end is 31.5 inches (80 centime-
ters) and connects to the cylindrical main body. The
cone contains the active radar terminal guidance
subsystem, inertial sensors, data correlation, steer-
ing computer, communications bus, and power sup-
ply batteries. The guidance unit is a strap-down set
of instruments (gyros and accelerometers) rather than
full-freedom inertial measurement unit (IMU).10 This
conclusion is based on successful use of strap-down
instruments for steering the Agni-I ballistic RV. It
required considerably less development and is lower
cost than a full inertial guidance system.
Arun S. Vishwakarma noted in an article on “Stra-
tegic Missiles,” the MRV internal structure was inher-
ited from the Agni-I Mark I ballistic RV; and there-
fore, it was larger and heavier (i.e., 1,000 kg payload)
than needed for a new warhead. The new warhead is
described as being lower density, elongated, and hav-
ing a weight between 200 (minimum) and 450 (maxi-
mum) kilograms, with a most likely weight at about
300 kilograms (660 pounds).7 Figure 26 shows the
warhead located near the rear end of the conic section
to keep the center of gravity as far forward as pos-
sible for stability. The warhead is a spherical implosion
device and elongated to include arming, fusing, and
firing circuitry plus independent electric power.
The aft MRV section is a cylinder 31.5 inches in
diameter and 77 inches long. The four triangular con-
trol fins are about 22 inches long and extend 18 inches
from the cylindrical body at their tips. Control actua-

215
william yengst

tors, their power source, and wiring for steering com-


mands are located under the fins. Because the new
warhead is smaller and lighter than the first genera-
tion design, the MRV has space and capacity in the aft
section to carry up to 200 kilograms (440 pounds) of
liquid fuel in pressurized vessels, supplying the HAM
propulsion system.7 Nozzles for the reaction jets are
spaced between the fins on the flare adapter as illus-
trated at the left end of Figure 26. An assembled MRV
is shown at the Bharat Rakshak plant in Figure 27.

Figure 27. Agni-II Mark 2 MRV in


production at Bharat Rakshak Limited

The reported ballistic coefficient for the MRV is


440 kilograms per square meter (or, W/CDA = 2,150
pounds per square foot).7 This is somewhat high to
survive IRBM reentry heating and deceleration envi-
ronments, while retaining ample energy to perform
atmospheric maneuvers. Assuming the overall payload
weight is 2,200 pounds, if one removes the adapter
(~150 pounds) and HAT subsystem fuel (440 pounds),
the MRV including its warhead would weigh roughly
1,610 pounds. Of this weight, the heatshield would be
about 500 pounds, warhead roughly 660 pounds, leav-
ing 450 pounds for guidance, actuators, electric power,

216
lightning bolts

wiring, and structure. Although this is realistic, an


MRV weight of (W~1,610 pounds) and base area of
(A=5.41 square feet), implies an effective drag coef-
ficient of (CD~0.14), a somewhat low value for a blunt
nose design. One would expect the drag coefficient to
be higher, like that of Pershing II. A drag coefficient of
(CD~0.22) appears more realistic, but it would imply
the MRV weight is 2,560 pounds, considerably higher
than reported. This inconsistency may be explained by
one (or a combination) of the following:

1. Reported weight and/or ballistic coeffi-


cient may be in error.
2. HAT thrusters may be used to maintain
velocity during reentry (e.g., like stages 3
and 4 of the Athena missile described in
Chapter 5). This would reduce effective
drag below wind-tunnel measured values
and is a likely answer.
3. Frontal drag may have been reduced by
using a smaller nose radius. A sharper
nose design was used for Agni-III and -IV
(as described below).

The first flight test of Agni-II with a ballistic RV


was made from a railroad launcher on 11 April 1999.
The missile flew 11 minutes from the Orissa coast in
Balasore district and landed in the Bay of Bengal, a
distance of 2,000 kilometers (1,240 miles). The sec-
ond test, on 17 January 2001, was designed to fly 2,100
kilometers (1,310 miles). These successful flights were

217
william yengst

described by P. N. Aggarwal as, “the final configu-


ration for Agni-II.” A third flight occurred in July
2004.3 Early tests showed Agni-II was roughly three
times more accurate than Agni-I.11
Bharat Dynamics Limited established a pro-
duction capability for Agni-II at Hyderabdad and
planned to build 10–12 missiles per year. In May and
again in July 2001, Indian Defense Minister, Jaswant
Singh, told a Cabinet Committee on Security that the
system was operational. On 14 March, 2002, Defense
Minister. George Fernandes, reported to Parliament
that Agni-II was in production.11 Estimates suggest
that India intends to produce several dozen nuclear
Agni-II missiles for deployment in the northeastern
region (555th Army Missile Group).2 From that area,
the weapon can reach most of China, but not Beijing
or Manchuria, which requires a range greater than
3,000 kilometers (1,860 miles).11
The Indian development of Agni-II MRV was a
well-run and highly successful program. Three reasons
for this conclusion are:

• It learned from a carefully structured


advanced development program (IGMDP)
plus observing U.S. Pershing-II and other
missile programs.
• Use of a simplified design (e.g., highly
reliable, low cost “strap-down” guid-
ance subsystem and aircraft electronics)
to avoid test failures, plus allowing suf-
ficient time between flight tests to make
modifications.

218
lightning bolts

• Acquisition of tested commercial technol-


ogy (obtained legally or illegally) plus use
of verified test results of Agni-I for inno-
vative elements of the design.

India acquired some technology from foreign


sources. For example, in March 2003, German pros-
ecutors arrested two men in Montanhydraulik for
violating export laws. The men illegally exported nine
consignments of hydraulic cylinders to India between
1997–2000, declaring them as bridge-laying equip-
ment. The cylinders were actually used for mobile-
missile-launcher erection mechanisms for Agni-I and
Agni-II missiles.12

Agni-III and -IV


To provide a deterrent against Beijing and north-
ern China, a longer-range variant missile, Agni-III,
was developed. This design was initiated late in 2001
and included several improvements over Agni-II to
achieve ranges of 3,500 to 4,000 kilometers (2,170–
2,480 miles). The first and second stage boosters used
higher specific impulse fuels and flex nozzles to aid
powered-flight steering. Motor cases were made of
wound carbon fiber composite to reduce weight and
increase the fuel-mass fraction.7 Because of longer
length, Agni-III may only be rail mobile. The missile
can deliver a third-generation, lighter-weight, ther-
monuclear nuclear warhead (200-kiloton yield), in a
modified Agni-II MRV, known as MRV Mark-2.

219
william yengst

Four modifications were made during develop-


ment of the MRV Mark-2. First, the vehicle length
was shortened to 130 inches (permitted by a smaller
warhead). Second, the blunt nose was sharpened to
a radius or 3.35 inches but the half-cone slope was
increased to 10 degrees. These offsetting changes sug-
gest the overall drag coefficient would remain nearly
the same for longer range and higher velocity tra-
jectories. Third, internal structure was reduced from
supporting a 1,000 kilogram to a 500 kilogram pay-
load.7 Fourth, the heatshield was redesigned to satisfy
reentry velocities consistent with the longer ranges.7,11
One source reported the Agni-III MRV airframe,
including the heatshield weighs 250 kilograms (550
pounds).13
In summary, Agni-III MRV Mark 2 probably
weighs about 1,660 pounds, has a drag coefficient of
(CD ~ 0.22), base area of (A = 5.41 square feet), and
ballistic coefficient of about 1,400 pounds-per-square
foot. Although the design is primarily focused on
achieving high accuracy, one source indicated pene-
tration of enemy defenses would be enhanced by addi-
tion of penetration aids. Specifically, “Larger internal
volume allows more sophisticated Anti-Ballistic Mis-
sile countermeasures.”4
A first flight test of Agni-III was scheduled for
late 2003 but was delayed until July 2006 due to
development problems. The test failed, but two suc-
cessful flights were made before the end of 2007. Dr.
V. K. Saraswat, India’s Chief Controller of Missiles
and Strategic Systems, reported that Agni-III would
be provided new small, Multiple Independent Reen-
try Vehicles (MIRV ) instead of the MRV Mark-2.

220
lightning bolts

According to DRDO, the design will deliver a few


nuclear ballistic RVs to different targets and will be
supported by decoy warheads to confuse defenses.14
A new Agni-III+ (or Agni-IV) missile with an
increased range to 5,000 kilometers (3,100 miles)
was started in 2008. It planned to incorporate almost
identical technology, as that developed and tested for
Agni-III. However, it probably needs a heavier heat-
shield and stronger internal structure to accommodate
higher reentry environments. The first flight of Agni-
IV was planned for 2009.14

221
chapter 9

China’s M-11 and -11A


MRV Developments
China’s development of nuclear weapons was motivated
by the Korean War, during which the U.S. threatened
use of nuclear weapons against Chinese airfields and
forces. Thus, beginning in 1954–1955, China acquired
nuclear technology from the Soviet Union. Through a
series of tests in the late 1970s, the size and yield of its
nuclear warheads were reduced (e.g., tests in Septem-
ber 1977, March 1978, and October 1980—all produced
yields under 20 Kt). A Chinese Army manual that
identified differences between tactical and strategic
nuclear weapons was published in 1983. China officially
declared in 1994 that its nuclear weapon tests were for
improving safety and reliability of existing warheads;
however, it is clear the real objective was to develop

223
william yengst

smaller warheads for tactical missiles and strategic


ICBM MIRV or MRV applications.1
China viewed its tactical missile forces as a tool
for intimidation in the Asian region. Examples include
deployments along the Sino-Soviet border after a
March 1969 crisis in Manchuria, operational exercises
to influence Taiwan political affairs, and threaten U.S.
Pacific naval operations in support of Taiwan (1995–
1996). In addition to providing war-fighting capabili-
ties, the Peoples Republic of China (PRC) developed
warheads and missile delivery systems for export to
friendly nations, including Pakistan, Libya, Syria,
North Korea, and Iran.

The “Mosiac” Process


In 1976, Chinese leaders judged that their scientists
and engineers were behind those of the U.S. or Rus-
sia; therefore, it began a long-term process of collect-
ing intelligence using a process known as “mosaic.”2
The mosaic approach was to develop advanced missiles
and warheads by semi-empirical methods, using data
obtained from foreign nations through cooperation,
purchase, sales, and espionage as necessary. Information
concerning key technologies was gathered by rigorous
review of unclassified technical and academic pub-
lications, interaction with U.S. and foreign scientists,
defections of scientific and engineering personnel and
covert spying. Nuclear warhead and missile informa-
tion was pieced together, confirmed, and updated con-
tinuously to provide detailed design and performance
characteristics.

224
lightning bolts

Dr. Tsien Hsue-shen, a protégé of Theodore Von


Karmen at California Institute of Technology (1930s)
and a U.S. Air Force colonel who helped evaluate cap-
tured V-2 missiles (late 1940s), was prosecuted by the
McCarthy security committee and deported to Bejing
on 17 September 1955. According to the Federal Bureau
of Investigation, he became father of the Chinese bal-
listic missile program in the 1970s.3
Examples of technologies derived from U.S. sources
were as follows:

• At least four QBM 34A Firebee recon-


naissance drones with their optical-sen-
sor units were launched from Bien Hoa
Airbase (South Vietnam), shot down or
crashed in Southern China and put on
display in Beijing (2 April 1965). These
drones, used by the CIA in lieu of U-2
reconnaissance aircraft, were vulnerable to
antiaircraft artillery and SA-2, air-defense
missiles. They were used to monitor PRC
missile tests and logistics transport routes
into North Vietnam.5
• Commercial cooperation with LORAL
Corporation and Hughes Aircraft Com-
pany in 1995 provided satellite guid-
ance designs and improved reliability
of space-launch vehicles. The resulting
scandal investigation became known as,
“Chasing-the-Dragon.”2
• The U.S. fired roughly 280 Tomahawk
cruise missiles into Iraq (1992) and

225
william yengst

Afghanistan (2003), some of which were


recovered from debris. China purchased
TERCOM guidance units from those
recovered in Afghanistan.2
• Pershing-II RADAG terminal guidance
information was downgraded and pub-
lished after the weapon system was with-
drawn under the INF Treaty (1989).
• PRC acquired classified information on
several U.S. nuclear warheads from the
DOE weapon laboratories (e.g., W-70
Lance [late 1970s], W-62 Minuteman-II,
W-78 Minuteman-II [late 1980s], W-87
Peacekeeper MX, and W-88 Trident D-5).
The later two warhead designs were
obtained in December 1999 from Los
Alamos during the Wen Ho Lee scandal.
China also obtained “neutron bomb” data
in the late 1970s and tested a design (1988),
according to U.S. intelligence reports.2

Other missile technologies were obtained from


the U.S., Russia, Israel, Germany, Great Britain, and
France, to support warhead and missile developments.
Richard D. Fisher Jr., after visiting the Zhuhai Air
Show in 1998, claimed China had cut its develop-
ment times for small, reentry vehicle designs by 2 to
10 years. Clearly, China relied extensively on foreign
technologies for its rapid missile progress.

226
lightning bolts

M-11 Missile Development


Development of the M-family of solid propellant,
short and intermediate range, ballistic missiles is
believed to have started in the late 1970s. The M-11
design team focused on a solid-propellant version of
the Russian R-17 (liquid-propelled Scud 1B), which
operated from a Soviet MAZ 543 Transporter-Erec-
tor-Launcher. The MAZ vehicle was modified by
changing its launcher cradle and roof assembly; per-
mitting it to carry the Chinese missile, which had a
different diameter and length. The Chinese intent was
to compete directly with R-17 sales to export markets.
Specifically, the missile was reported to have a con-
ventional warhead and range of 280 kilometers (175
miles), below the limits of the 1987 Missile Technol-
ogy Control Regime (MTCR) Agreement. The agree-
ment, signed by 32 nations, restricts export of weapon
systems and related technologies capable of delivering
payloads of more than 500 kilograms (1,100 pounds)
to ranges in excess of 300 kilometers (186 miles).6
The M-11 program carried an internal Chinese
designator DF-15 but was known in NATO as CSS-
7. Originally, the design focused on conventional,
high-explosive, fuel-air explosive, sub-munitions, and
chemical warheads. Photographs of M-11 obtained in
1999 showed the modified MAZ-543 with clamshell
doors on top of the transporter to protect the missile
from weather, although subsequent sightings did not
include the doors.
Chinese missile exports to Iran began during
1985, at the height of the Iran-Iraq War, and included
HY-2 (Silkworm) cruise missiles. In 1989, a delivery of

227
william yengst

150–200 short-range, M-7 ballistic missiles was made


to Iran.7 The ‘Kargil War’ of 1999 in Kashmir caused
Pakistan to seek longer range missile technology from
China to offset Indian ballistic missile threats. China
offered to assist with development of Pakistan’s Sha-
heen-I medium-range ballistic missile (MRBM), a
direct extension of the M-11.3 This missile has a range
of 750 kilometers (465 miles) with a conventional
payload of 800 kilograms (1,760 pounds).2 The U.S.
imposed MTCR sanctions on China in August 1993,
when it learned about delivery of 34 complete M-11s
to Pakistan. MTCR terms only required “strong pre-
sumption of denial,” which China promptly provided;
hence, the sanctions were waived in November 1994.
The Chinese continued to assist Pakistan in setting up
an indigenous M-11 production capability at Kahuta
in 1996–1997.7
A contract for M-11 sales to Pakistan was signed
in 1988. The first missile flight test with a conven-
tional warhead was conducted in 1990 by the People’s
Liberation Army (PLA).8 Development time from
inception to first flight for this missile was barely over
three years. The system became operational with the
Pakistani Army in 2003 under the name, Shaheen-I.
A manufacturing facility was built in the outskirts of
Rawalpindi during 1995 for producing M-11/Shaheen
missiles and maintaining operational systems.9 Chi-
nese technicians visited Pakistani missile sites to help
train army units with operational missiles starting in
1994.7
The missile is shown shortly after lift-off in Figure
28. It is 24.6 feet long, has a diameter of 80 centi-
meters (31.5 inches), launch weight of 8,360 pounds,

228
lightning bolts

and can carry 1,760 pounds of payload.5 Its MRV with


delta-shaped base fins is clearly evident.

Figure 28. Chinese M-11 conventional


missile during flight test (1990)

Shaheen-I Missile Design


The National Defense Complex (subsidiary of Paki-
stan’s Atomic Energy Commission [PAEC]) was
formed in 1993 and initiated the Shaheen missile pro-
gram in 1995.9 It is useful to use Shaheen-I, an extended-
range version of M-11, to describe characteristics of
the MRV. Because the missile was designed to deliver

229
william yengst

conventional weapons, as well as nuclear warheads, it


required the highest possible accuracy to be effective.
Dr. Samar Mubarakmand in an interview on
GEO-TV (5 March 2004) described the reentry sys-
tem as, “probably having a radar correlation terminal
seeker to achieve accuracy on the order of (CEP=100
to 150 feet).”10 The guidance system can also employ
Global Position System (GPS) updates to its inertial
navigation further improving accuracy. The vehicle
has four-moving control fins at the rear. Inside the
vehicle is a small, solid-propellant thrust motor, used
to orient the reentry vehicle after booster burn out
and improve stabilization during the terminal phase.
He also noted, “The MRV thrusters can be used to
fly evasive maneuvers to prevent intercept by existing
antiballistic missile defense systems. Shaheen-I may
change its trajectory several times during the terminal
phase, preventing pre-calculated intercepts based on
ABM radar tracking.”
The first flight test for Shaheen-I occurred on 15
April 1999.10 Although detailed design features of the
nuclear version are not available—due to uncertainty
in warhead size and weight characteristics—it is pos-
sible to estimate the configuration with some confi-
dence. Therefore, it is instructive to reverse engineer
the MRV design based on available information and
results of the preceding chapters.
The first step is to scale the vehicle’s dimensions
from available photographs. It is a cone-cylinder con-
figuration 155 inches long and 31.5 inches in diameter
at the base. The nose tip radius is about one inch and
the forward conic section is 110 inches long. It’s conic
half-angle is 8 degrees, suggesting an effective drag

230
lightning bolts

coefficient close to (CD=0.2). Consequently, with an


overall weight of 800 kilograms (W=1,760 pounds)
and base area of (A=5.41 square feet), the ballistic
coefficient must be (W/CDA=1,625 pounds-per-square
foot). This is realistic for a vehicle capable of relatively
high-g terminal maneuvers.
Assume, for evaluation purposes, that the Paki-
stani warhead is spherical with a maximum diameter
of 80 centimeters (31.5 inches). It would have an adja-
cent smaller arming, fusing, and detonating electron-
ics package (A&F) plus independent electric power
batteries. The A&F subsystem might weigh about 50
pounds depending on its degree of sophistication.
These components would extend the warhead’s elon-
gated shape to a length of about 45 inches as illustrated
in Figure 29. Its weight would be at least 660 pounds
based on a density of 54 pounds per cubic foot, typi-
cal of early Pakistani warhead designs. The warhead
would be as far forward as the aerodynamic structure
permits for stability (e.g., maximum diameter at the
break-point between the conic and cylindrical sec-
tions of the MRV and with the smaller end forward).
Fortunately, short-range missiles do not require
heavy heatshields, since their reentry bodies do not
experience severe thermal conditions; although the
vehicle requires a nose cone for aerodynamic and sta-
bility purposes. The aerodynamic shell is probably a
thin layer of aluminum (< 0.125-inch thick and density
of 0.0975 lb/in3) coated with carbon-phenolic (~ 0.25-
inch thick and density of 1.5 g/cubic centimeter or
0.0542 lb/in3). With an approximate volume of 3,730
cubic inches, it would weight 250 pounds. Internal
aluminum structures to mount the guidance, warhead,

231
william yengst

and control system plus brackets, clamps, cables, and


electric wiring might weigh an additional 200 pounds.
Therefore, the overall heat shield and structure prob-
ably weigh roughly 450 pounds.
The missile relies on a combined inertial naviga-
tion system (INS) and Global Position System (GPS)
to provide booster launch reference, powered flight
steering, separation and sequencing of events includ-
ing reentry control and steering.8 The INS/GPS, its
computer and data processing units, would be located
forward of the warhead as shown in Figure 29. By
today’s technology, they might weigh 100 pounds
and require battery power (perhaps 50 pounds). The
system would be capable of guiding the vehicle to an
impact accuracy of roughly 50 meters (CEP ~150 feet).

Figure 29. Hypothetical M-11/M-11A


reentry vehicle design characteristics

Four aerodynamic fins extend from the MARV


body about 5 inches from its base. They are 14 inches

232
lightning bolts

long at the surface and taper to 10 inches wide at a


height of 10 inches above the body. At 1-inch thick,
the fins have similar heat protection as the heat-
shield and weigh about 8 pounds apiece (32 pounds
total). Four high-power actuators (either electrical or
hydraulic) are needed to move the fins. These actua-
tors would be located in the cylindrical section imme-
diately beneath the fins and might weigh 12–15 pounds
each (60 pounds total). The fins may be driven by a
high-pressure gas generator, using an oil supply and
plumbing (or alternatively an electrical power supply)
with an estimated weight of 40 pounds. Thus, the con-
trol subsystem total weight would approach roughly
100 pounds.
The MRV attitude control system is probably a
solid-propellant, high-pressure gas generator with
base-mounted jets to align the vehicle for reentry.
This would include four symmetrically mounted base
jets connected by plumbing and valves from its gas
generator. Located at the aft end of the vehicle, its
weight would be about 30 pounds. An aluminum base
cover, structural braces and separation fixture around
the 99-inch base perimeter (connecting the MRV to
the booster) would weigh an estimated 40 pounds.
A well-engineered design would require only mini-
mal ballast to provide stability; however, this hypotheti-
cal design has a problem. The large, heavy warhead, con-
trol subsystem, and exo-atmospheric attitude control
only fit near the rear. Therefore, the center of gravity
(cg) of the MRV is near the 102-inch mark on its body.
Aerodynamic center of pressure (cp) occurs between
two-thirds and three-quarters of the body length for a
cone (i.e., between 104 and 115 inches from the nose).

233
william yengst

Therefore, tungsten or lead ballast may be added inside


the nose to keep the cg forward of the cp to insure a
small positive stability margin (i.e., a few percent of
the body length). Otherwise, the vehicle could lose
control, tumble, or require excessive fin power to per-
form maneuvers. A rough estimate suggests 250 pounds
of ballast might be needed depending on the warhead
weight and its placement. In conclusion, this hypotheti-
cal design with a total weight of about 1,760 pounds
satisfies the reported payload and appears to be feasible.
One report indicated the M-11 MRV may employ
terminal guidance, and if so, it would probably be
electro-optical scene-matching.11 This system would
include a lens or mirror, electro-optical sensor, magnetic
image storage, and digital correlation computer to pro-
vide updates to the INS. It would be colocated with the
INS unit and use the same power supply. It’s weight
would be small (i.e., perhaps 15 pounds) and could
provide accuracies of (CEP=10–20 feet). Nuclear and
chemical warheads do not require precision accuracy to
be effective. Likewise, conventional submunitions and
Fuel Air Explosives (FAE) warheads attributed to the
missile provide significant area effects and would not
sacrifice much effectiveness with only INS/GPS accu-
racy. By contrast, a high-explosive warhead would lose
appreciable effectiveness, if the high accuracy capability
is not included.

M-11A Design Characteristics


Syria and Iran were interested in acquiring M-11 mis-
siles; however, each nation had slightly different design
requirements. China initiated a design in 1995 for a mis-

234
lightning bolts

sile known as CSS-7 Mod 2 or M-11A, as a Syrian vari-


ant. Like the original design, M-11A had a single-stage,
solid-propellant booster, but with low cost, heavier
steel-motor casing. It could be launched from Russian
“Scud B” SS-21 TELs already in the Syrian military
inventory. The missile was believed to have a range of
280 kilometers (174 miles) with 1,000 kilograms (2,200
pounds) payload. However, the design was capable of
flying 350–530 kilometers (217–329 miles) with a payload
of 800 kilograms (1,760 pounds). The same MRV design
described above was used and equipped with high-
explosive, FAE, submunitions, and chemical-warhead
options. The missile provided sufficient range to reach
all of Israel from deployments in Syria.5 China flight
tested M-11A in 1999. It shipped 20 M-11s to Syria and
built a manufacturing facility at Hama (Syria) to sup-
port and to produce the improved M-11A variant.12 This
violated terms of the MTCR Agreement.
The Iranian variant of M-11, known as Tondar,
had a launch weight of 9,240 pounds with a decreased
payload of 500 kilograms (1,100 pounds) to provide an
operational range of 350–530 kilometers (217–329 miles).
Although this design included the same conventional
warheads, it reportedly can carry a small nuclear war-
head. The missile has a length of 28 feet, diameter of 80
centimeters (31.5 inches), but requires a smaller, lighter
weight MRV than M-11. China shipped 30–50 M-11
missiles to Iran along with mobile launchers. Some were
displayed in the “Holy Defense Week” parade south of
Tehran for President Mamhoud Ahmadinejad on 22
September 2005.5
The Iranian design is believed to have inertial guid-
ance with Global Position System updates or possibly

235
william yengst

an optical scene-matching, terminal-guidance subsys-


tem. The MRV appears to have the same external con-
figuration as described for M-11. However, its design
may be modified in two respects: (1) a smaller diam-
eter and lighter warhead that can be moved forward in
the airframe to keep the center of gravity forward and
(2) reduced ballast to maintain nearly the same stabil-
ity margin. For example, if Iran develops a spherical
warhead 26 inches in diameter and weighing roughly
370 pounds (same density as above), it could be moved
forward in the airframe by 10–12 inches. The airframe
structure could be made of titanium rather than alumi-
num to save 50 pounds, and the ballast could almost be
eliminated. Thus, the MRV weight would be reduced to
1,100 pounds, consistent with the reported payload. The
lower weight would also reduce the vehicle’s ballistic
coefficient to (W/CDA=1,020 pounds per square foot).
This would be sufficient to maneuver for accuracy pur-
poses but would sacrifice some high-g, evasion capabil-
ity. Consequently, it is possible that the M-11A MRV is
a minor variant of the original design and would require
little additional flight testing.
The M-11A entered operational service with the
People’s Liberation Army in 1997 and approximately
200 missiles were deployed by 1998. In September 2004,
Taiwanese Defense Minister Lee Jye observed that
400 M-11 and M-11A missiles were deployed along
the coastal region facing his country.13 The U.S. 2002
National Intelligence Estimate, published by the Central
Intelligence Agency, observed that Beijing, “enabled
emerging missile states to accelerate missile develop-
ment timelines and sold turnkey systems to nations
which previously had non-existent capabilities.”13

236
chapter 10

Observations and Emerging


MRV Threats
The objectives of this monograph were to describe the
difficulties, successes, and roles played by many dedi-
cated people in the development of military maneu-
vering reentry vehicles. During the first twenty-five
years of this technology, most of the information was
classified. Therefore, the history and impacts of the
developments on international arms control agree-
ments, and weapon deployments largely escaped
notice by the general public. To be assured, scientists
and engineers involved in these programs met enor-
mous technological challenges and used innovative
methods in solving very difficult technical problems.
They have been rewarded with moments of exhilarat-
ing joy during successful tests, while contributing to

237
william yengst

the advancement of precise navigation, control sys-


tems, reentry materials, and design concepts applied to
civilian aircraft and space vehicles. Each day, they take
pride in knowing the work helped to deter aggression
and maintain peace.

Summary Assessment
German scientists, under stress in 1944 to develop
long-range weapon delivery systems as extensions of
the V-2 missile, are given credit for inventing maneu-
vering-reentry vehicles (MRVs). Large-scale ballistic
rockets and boost-glide concepts were postulated and
advanced in the United States by expatriate German
scientists. Three glide vehicle feasibility experiments
were conducted under the WS-199D Alpha Draco pro-
gram by McDonnell Aircraft Company (1957–1959).
In 1964, the Boost Glide Reentry Vehicle (BGRV )
concept was accepted as a means for extending the
range and payload of U.S. intercontinental ballistic
missiles. However, competing designs emerged, when
General Electric Reentry Systems Division invented
the Maneuvering Ballistic Reentry Vehicle (MBRV ).
It was followed quickly by the Douglas Aircraft Com-
pany jet reaction control MARCAS concept. These
three concepts posed serious technical feasibility
problems before a weapon system could be developed
with confidence.
In studying Chapters 2 through 9, it is possible to
identify three distinct stages of development, which
can be described briefly as follows.

238
lightning bolts

• Feasibility Demonstrations: The first step


in developing a new system is to design,
fabricate, and demonstrate that theory
and intended performance goals or opera-
tional features can be made to work. This
usually involves extending existing theory,
materials state-of-the-art, and inventing
subsystems, such as navigation, transpira-
tion nose tips, terminal-position guidance,
and aerodynamic controls. All four initial
programs between 1957 and 1964 (Alpha
Draco, MBRV, BGRV, and MARCAS)
were feasibility demonstrations. They
required three or four flight tests, 30 to 49
months to prove feasibility, and sufficient
data gathered to move to the next stage of
development with confidence. Feasibility
is usually demonstrated, using modified
commercially available or newly invented
equipment to minimize costs. Although
one always strives for success, feasibility
tests are more likely to result in failures
than subsequent development, because
performance goals push the limits of exist-
ing theory and equipment capabilities.
• Advanced Development: The second stage
involves design, fabrication, and test of
prototype equipment for the intended
system. It builds on feasibility results but
attempts to tailor and qualify subsystems
for their specific functions (e.g., guidance
units, control actuators, data processing,

239
william yengst

and thermal protection). This is usually


more expensive, requires a few to several
tests to confirm performance goals and may
take longer to accomplish. Results should
insure greater reliability and functional
capability. ACE, AMaRV, and Navy Mark
5 MRV were primarily advanced devel-
opment programs, each involved three
flight tests to validate design prototypes
and performance over periods of three to
five years. At completion of AMaRV and
Mark 5, the designs were declared ready
for operational development.
• Operational Development: The third stage
is focused on preparing the weapon system
for mass production and verifying its full
range of capabilities. Designs and subsys-
tems must satisfy military specifications
(i.e., Mil Specs). Six MRVs (U.S.’s Persh-
ing-II MRV and Mark-5, India’s Agni-II
and–III, China’s M-11 and M-11A (or alter-
natively Pakistan’s Shaheen MRV) reached
this stage by 2004. Although only a few
flights of these systems were described
in the text, that portion of the develop-
ment took four to five years. A full-scale,
operational development program often
requires more than five years, is relatively
expensive and requires 10–20 tests to con-
firm performance plus training of military
units under operational conditions. One
should expect relatively few failures (e.g.,

240
lightning bolts

less than 5 percent) during this stage, since


design and production problems have been
corrected during advanced development.

The above observations are reflected in the data of


Table 1. In some cases, program “start” and “end” dates
are only approximate; however, results are realistic for
a new technology moving from feasibility through
advanced development. It is interesting to illustrate
this process by briefly reviewing the Wright broth-
ers’ nearly parallel experience in demonstrating feasi-
bility and advanced development of the first manned
airplane.

241
Figure 29. Hypothetical M-11/M-11A reentry vehicle design characteristics
(Note: This is my own drawing but may need to be redrawn.)

Note: Chapter 10 figures start here.


Table 1. MRV development times and flight tests
Table 1. MRV development times and flight tests

Start Date Flight Period of Flight Program Total


__U.S. Programs__ _(M/Year)_ Tests _Tests (M/Year)_ End (M/Year) Months
Alpha Draco/WS-199D 1957 3 2/59 to 4/1959 1959 ~ 30
MBRV 8/1963 4 8/66 to 8/1967 2/1968 54
BGRV 9/1963 4 11/66 to 2/1968 9/1968 60
MARCAS 9/1964 3 10/65 to 5/1967 7/1967 34
ACE 7/1972 3 9/73 to 9/1974 12/1974 29
AMaRV 1/1975 3 12/79 to 10/1981 ~ 1982 ~ 60

242
Mark 5 MRV 1984 3 1983 to ?? ~ 1989 ~ 60
william yengst

Pershing-II MRV…….12/1977 3+ 7/79 to 6/1980 ~ 1982 ~ 48__


Foreign Systems
India Agni-II MRV* 3/1998 3 1/01 to 7/2004 9/2004 > 30__
India Agni-III MRV 2002 >3 2003 to 2007 2007 ~ 48__
China M-11 MRV 1988 ? …………….. 1992 36-40__
China M-11A MR 1995 ~3 4/99 to 3/2004 2004 ~ 60__
* Development delayed 3 years by decision to change missile payload.

(Note: This is my table…also see separate Excel Table 1, which may need
alignment).
lightning bolts

First Sustained Aircraft Flight Demonstration:


In May 1899, Wilbur Wright wrote a letter to the
Smithsonian Institution (Washington, D.C.) request-
ing information on aerodynamic theory.1 He had
observed birds in flight to formulate a theory con-
cerning wing-warping as a means of controlling
their flight. In July 1899, he built several five-foot
box kites to prove the concept. In 1900, Orville and
Wilbur Wright moved to Kitty Hawk (North Car-
olina) where they established a bicycle shop and, in
spare time, began developing aerodynamic theory
and experimenting with lifting wings. The ability of
birds to change angle of attack by twisting their bod-
ies and wing tips, caused the brothers to invent the
concept of “banking” or “leaning” to turn. This was
the fundamental breakthrough leading to “three-axis
control.” They gained mechanical skills and gathered
aerodynamic “lift” and “drag” data, using three man-
sized, unpowered gliders. Wilbur flew on the larger
gliders to test the controls. Over 1,000 glider experi-
ments were performed by the end of 1901.2 During
these tests, they identified three feasibility obstacles
to manned flight:1

• Strong, light-weight, wing structures of


sufficient size to lift the airframe, propul-
sion, and human pilot. They solved the
first problem by using two (double-decker
or biplane) hollow, wood-frame wings
constructed of spruce, strengthened by
wires, and covered with canvas.

243
william yengst

• A control mechanism to permit banking


and pitch maneuvers. This was done by
twisting the wings and adding a rudder.
The controls were manually operated by
the pilot, using ropes and pulleys.
• Development of a light weight motor
and efficient propeller system capable of
overcoming drag losses of energy. This
was accomplished by two approaches:
(1) a newly designed, wooden propeller
and (2) Charlie Taylor, their bicycle shop
employee, built a cast-aluminum version
of a primitive engine with no carburetor
or fuel pump (gravity fed gasoline through
a rubber hose).

The year 1901 might be considered the beginning


of the program, although much of the brother’s time
and efforts were consumed with inventing, developing
theory, testing alternative concepts, building a box-
like, 6-foot wind tunnel, preparing models, and study-
ing data. Work was funded entirely from donations
and profits from their printing and bicycle shops.3
Feasibility demonstration began in 1902 with a
decision to build three, full-scale Flyer vehicles. Each
vehicle cost less than $1,000, had a wing span of 40
feet, weighed 625 pounds, had two wooden propellers
connected to the engine by belts, and had a 170 pound
engine rated at 12 horsepower.1 They viewed propel-
lers as vertically mounted wings, generating lift in the
forward direction.4 During that year, through a series
of model and ground tests, they improved efficiency

244
lightning bolts

of propellers from below 66 percent to 75 percent and


arrived at a peak efficiency of 85 percent in March 1903.
They suffered weeks of delays when propellers broke.
Most of 1903 was spent building the first full-size
aircraft and testing its controls. They found the value
of a vertical rudder was not in changing direction (as
with a boat) but rather, for aligning the vehicle during
banking turns and leveling off. Turns and roll control
were accomplished using “wing-warping.”
On 8 December 1903, they tested Flyer I, but it
crashed in the Potomac River after fouling on its
launcher ramp. The pilot, Charles M. Manley, was res-
cued. On 17 December, at Kill Devil Hills, Kitty Hawk
Beach (North Carolina), they made the world’s first
sustained and controlled powered flight.2 Starting 26
May 1904, the Wrights began a series of flights using
Flyer II, which had a new 18-horsepower engine.3 They
developed a weight and derrick method of launch-
ing the vehicle, which was tested for the first time
on 7 September 1904. On 9 November, they achieved
a five-minute flight.2 In June 1905, Flyer III with a
larger vertical rudder was flown for the first time, and
on 4 October, it accomplished a half-hour flight. No
further tests were made after 16 October until 6 May
1908.2 By way of summation, three test vehicles were
built, one failed completely, and two (following some
modifications) were flown several times.
Because no one else succeeded in emulating the
Wrights’ first flights, it can be concluded the feasi-
bility demonstration phase was complete in October
1908. Depending on when one selects as the start date,
feasibility took 34 to 46 months. The U.S. military was
reluctant to fund heavier-than-air development in

245
william yengst

1903 and only invested $50,000 with Samuel Langley


in Dayton (Ohio) for experiments on a winged vehi-
cle known as “Aerodrome A.” During 1906–1908, the
Wright brothers performed many flight demonstra-
tions and built five improved versions of the Flyer III,
with more powerful engines and upright pilot seat-
ing. These aircraft could be considered the advanced
development version of the technology. They included
larger engines, independent rudder, and wing-warping
controls.3 The first, formal funding for this capability
came from the U.S. Army and a French syndicate to
build a reconnaissance aircraft. It was demonstrated at
La Mans (France) on 8 August 1908 and at Fort Myers
(Virginia) in September 1908.2
There is no intent to imply that maneuvering
reentry vehicle (MRV ) technology compares with
the important and dramatic Wright brothers’ achieve-
ments with the first aircraft. However, MRV tech-
nology development shows many common features,
including innovation and difficulties, feasibility dem-
onstration times, and number of test vehicles. Most
sophisticated technologies experience a similar devel-
opment process.

Evolution of MRV Design Characteristics:


The physical characteristics and performance of the
MRV concepts described in Chapters 2 through 9
should be compared to assess evolution of the tech-
nology. Table 2 provides a summary of these features
and permits identification of design changes.
The most obvious feature is reduction in size and
weight (by a factor of two to three times) from the

246
lightning bolts

MBRV and BGRV designs to operational MRVs,


such as AMaRV, Mark 5, Pershing-II, India’s Angi, and
China’s M-11 designs. This advance was accomplished
primarily through reducing subsystem (and warhead)
diameters to move them forward in the vehicle: tailor-
designing, packaging guidance and control units, and
custom shaping parts and components rather than
using off-the-shelf, commercially available equip-
ment. Warheads selected for all U.S. operational MRV
designs were smaller and lighter weight by roughly
260 pounds than Indian or Chinese warheads. As sub-
system sizes and weights reduced, it was possible to
place them forward and remove unnecessary ballast
while maintaining stability.

247
william yengst

Table 2. Performance and Characteristics


Table 2. Performance and characteristics
estimates estimates for MRV concepts
for MRV concepts]
Range Accelerat. Payload CEP MRV Weight
__U.S._Systems___ (miles) __(G’s)__ (pounds) (feet) __(Pounds)__
MBRV 4,400 80-100 ~400 ~5,000 3,100
BGRV 4,600 ~ 25 400 ~5,000 2,100
ACE 5,500 ~100 ~ 350 <1,300 ~ 1,100
AMaRV 5,500 100 ~350 < 1,300 1,035
Mark 5 MRV 4,600 > 80 ~350 300-400 ~ 800
Pershing-II MRV___ 1,100 ~ 80 400 ~ 150 1,467____
Foreign Systems
India Agni-II 1,150 < 80 > 660 ~ 330 1,610
India Agni-III 2,480 < 100 > 660 ~ 150 1,660
China M-11 MRV 175 ~ 25 ~660 ~ 150 1,760
China M-11A_ 217-329 < 50 ~ 370` 30-50 1,100____
Nose MRV Base ½ Cone W/CDA
2
__U.S._Systems___ Radius (in) Length (in) Diameter (in) Angleo _(lbs/ft )_
MBRV 0.5 134 52.0 11.0 ~ 1,050
BGRV 0.5 168 28.0 9.5 50,000
ACE ~1.0 ~ 82 ~ 22.0 10.4 ~2,000
AMaRV 0.92 82 21.8 10.4 1,960
Mark 5 MRV < 2.0 69 21.8 10.4 ~800
Pershing-II MRV 1.0 156 30.0 ~ 15.0 1,425_
Foreign Systems
India Agni-II MRV 5.9 167 31.5 7.15 2,150
India Agni-III MRV 3.35 ~ 130 31.5 10.0 1,400
China M-11 MRV 1.0 155 31.5 8.0 1,625
China M-11A MRV 1.0 155 31.5 8.0 1,020

__U.S. Systems___ Control Type Exo-Atmos. Guidance System Terminal Fix


MBRV 4 Flaps Jets MINS/Univac TERSE
BGRV 10 Flares Jets MINS/Univac TERSE
ACE 3 Flaps Jets Open-Loop IN --
AMaRV 3 Flaps Jets DINS/DCU --
Mark 5 MRV 3 Flaps Jets AIRS/DCU --
Pershing II MRV 4 Fins Jets IMU/DCU RADAG
Foreign Systems
India Agni II MRV 4 Fins HAM Open-Loop IN Yes
India Agni III MRV 4 Fins HAM Open-Loop IN Yes
China M-11 MRV 4 Fins Jets IMU/DCU Yes
China M-11A MRV 4 Fins Jets IMU/DCU Yes
Note: Open-Loop IN is inertial navigation instruments, IMU/DCU is inertial
Measurement Unit/Digital Computer Unit, and HAM is High Altitude Motor

(Note: the above Table 2 has been done in Excell (12 point) but I rather like the
smaller 10 point version above.)

248
lightning bolts

Another observation is that weight of MRV aero-


dynamic shells and heatshields is a strong function
of reentry velocity and, hence, range. At short ranges
(e.g., less than 150 miles) aerodynamic shells are made
of light metals that represent only 15–20 percent of the
gross-vehicle weight. At intercontinental ranges (e.g.,
4,500 miles or more), the shell and ablative shield
materials represent 25–35 percent of the gross weight.
A new approach, such as transpiration and/or internal
cooling, will be needed to make further reductions in
heat-protection systems.
The majority of maneuvering ballistic vehicles are
bi-conic configurations, with a sharp cone forward
and small-slope aft frustum or cylinder. This is done
to permit packaging of heavy but sizable diameter
warhead as far forward as possible. Forward half-cone
angles are between 7 and 11 degrees, with resulting
overall drag coefficients of (CD=0.2 to 0.24). Vehicles
designed primarily to achieve accuracy can employ
minimal maneuver capability (e.g., 5 to 10 g and cover-
age footprints of 5–10 miles) with ballistic coefficients
of (W/CDA=500 to 1,000 pounds per-square-foot).
However, vehicles designed to perform evasion of
enemy defensive missiles require ballistic coefficients
of 1,500 to 2,000 pounds per-square-foot to permit
maneuvers of 80 to 100 g and coverage footprints of a
few hundred miles. Ballistic coefficients greater than
about 2,500 pounds per-square-foot are difficult to
achieve, because reentry velocities deep in the atmo-
sphere cause thermal environments that exceed the
strength of available heat-protection materials.
All operational MRVs have employed aerody-
namic flaps or fins for control. Although designs

249
william yengst

using reaction jets, external burning, moving masses,


and bent-nose shapes have been investigated, flaps or
fins are easier to implement and perform effectively.
While AMaRV and Mark 5 intercontinental-range
MRVs employed three flaps, Pershing-II, Indian Agni-
II and–III, and Chinese M-11/M-11A, MRVs are all
medium-range vehicles with four base-mounted fins.
There is little difference in the overall size and weight
of MRV’s using flaps or fins for control.
Early MRV concepts employed conventional iner-
tial measurement units (IMU) and digital computer
units (DCU) for navigation and steering functions.
Likewise, advanced development vehicle designs
(AMaRV and Navy Mark 5 MRV ) relied on full-free-
dom inertial platforms (i.e. DINS, AIMU and RIMU).
India introduced use of strap-down instruments for
the entire missile system with Agni-II and–III. If a
missile is intended to deliver only ballistic RVs and/
or penetration aids, it must use high-accuracy guid-
ance (e.g., IMU plus DCU) through powered flight
and payload deployments. However, if the missile
is designed to deliver highly accurate conventional
or low-yield nuclear warheads and perform high-g
evasive maneuvers, it is better to employ a moderate
accuracy autopilot for booster guidance and achieve
high accuracy with an MRV that includes terminal
fixes (e.g., map-matching or Global Position System).
This approach, using an MRV can reduce the cost by
using high-acceleration (g), strap-down instruments
to perform its missions. Agni-II and -III required less
development, were significantly lower cost, easier to
manufacture, and may be more reliable than previous
design concepts.

250
lightning bolts

There were no new MRV programs initiated by


the United States after the mid-1980s. The primary
reasons for his condition were:

• Fully operational weapon system designs


(Air Force AMaRV, Navy Mark-5, and
Army Pershing-II) were completed with
practical, efficient, and production designs.
AMaRV and Mark 5 were never put into
production or deployed operationally,
but Pershing-II was deployed, only to be
destroyed after the Intermediate-Range
Nuclear Forces (INF) Treaty.
• The end of the “Cold War” plus interna-
tional arms control agreements removed
the need to develop expensive defense-
evasion vehicles. Higher performance
evasion MRV designs including low-level
run-in (LLRI) were invented but never
developed or tested.

ºº The Anti-Ballistic Missile Treaty


signed between the U.S. and U.S.S.R.
on 26 May 1972 restricted defenses to
only two sites and was later amended
to defend only one site.
ºº Strategic Arms Limitation Talks
(SALT-I) between the U.S. and
U.S.S.R. from 17 November 1969 to
May 1972 limited the number of bal-
listic missile launchers (land-based

251
william yengst

and SLBMs).5 This caused both coun-


tries to develop larger missiles with
ballistic MIRV warheads and penetra-
tion aids to maintain or extend force
capabilities.
ºº Strategic Arms Limitation Talks
(SALT-II) between the U.S. and
U.S.S.R. conducted between 1972 and
18 June 1979, curtailed manufactur-
ing of strategic nuclear warheads as a
continuation of SALT-I.5 This limit
caused both nations to focus on higher
accuracy for existing weapons and
encouraged development of interme-
diate-range weapons (U.S.S.R. SS-20,
U.S. Pershing-II, and GLCM).
ºº The INF Treaty between the U.S. and
U.S.S.R., signed on 8 December 1987,
eliminated nuclear and conventional
ballistic and cruise missiles having
ranges of 300 to 3,400 miles.6

High development costs and political pres-


sures against initiating new weapon systems, dic-
tated against the need for MRVs, since there were no
immediate threats.
It is instructive to note that between 1954 and the
mid-1980s (a period of 30 years), the U.S. designed,
developed, tested, and produced over 30 different, bal-
listic-reentry vehicles for Atlas, Titan, Minuteman-I,
-II and -III, Peacekeeper MX, Polaris, Poseidon-C3 and
C4, and Trident-D5 long-range, strategic missiles.

252
lightning bolts

That represents conservative evolution of ballistic RV


technology, with advances being made in small incre-
ments, an average of one new design per year. By con-
trast, in 21 years from 1964 to 1985, the U.S. initiated
seven maneuvering-reentry-vehicle programs [i.e.,
three feasibility (MBRV, BGRV, and MARCAS), two
advanced development (ACE and AMaRV ), and two
operational (Pershing-II and Trident Mark-5)]. This is
an average of only one of the more sophisticated and
costly MRV designs every three years.
It was not the intent of this document to favor a
specific MRV design concept or to suggest that the
technology has reached its zenith. Uneven treatment
of designs is primarily due to limited availability of
information concerning the various programs. In fact,
there is considerable opportunity to extend and refine
the technology in the future. Just as no one would view
the Wright brothers’ aircraft as the zenith of airplane
technology, there are several advanced MRV concepts
that have not yet been explored. An example of MRV
technology advancement may be the Russian Topol-
M MRV development described below.

National Missile Defense and


Emerging Maneuvering Threats
The spread of ballistic missiles with potential deliv-
ery of nuclear warheads by missiles from China to
Pakistan, Iran, Syria, and North Korea during the
2000–2005 time period prompted the U.S. to initi-
ate an ambitious ‘National Missile Defense (NMD)’
development program. This antimissile program was

253
william yengst

estimated to cost $30.2 billion, and initial limited


deployment within the U.S. began in 2004.7 The sys-
tem utilizes both land-based and space-based weapons
to provide a shield against multiple ballistic missiles
launched anywhere on the earth’s surface at any given
time. Interceptor missiles are alerted by a network of
radar stations (located in the United States, Canada,
and selected European nations) to perform the task
of detecting ballistic missiles coming into American
or European territories. Radar stations are augmented
by satellites with infrared and heat sensors in geosyn-
chronous orbit to detect and track ballistic missiles
throughout their trajectories.
The primary element of NMD is a high-per-
formance, land-based missile interceptor capable
of engaging incoming missiles outside the earth’s
atmosphere with hit-to-kill warheads.8 On 19 Janu-
ary, 2006, U.S. Lieutenant General Henry A. Obering
III, director of the Missile Defense Agency, reported
a successful, static-test firing of the second-stage
rocket for the antimissile Kinetic Energy Interceptor
(KEI). It is intended for operational deployment to
occur by 2010.9 NMD was scheduled for development
in three phases. The first phase included building a
single Cobra Dane early warning radar site at Shemya
(Aleutian Islands), and interceptor missile launch silos
on Kodiak Island and Fort Greely (Alaska) by 2006
to counter small attacks (5–10 missiles) from North
Korea or China.18 The second phase was designed to
counter larger threats with up to 100 interceptors with
advanced countermeasures from a second interceptor
site in North Dakota. The third phase was intended
to integrate hundreds of ship-based interceptors and

254
lightning bolts

European land-based units for “Theater Wide” missile


defense. Plans called for building ten-interceptor mis-
sile silos in eastern Poland (Ustka-Wicko Army base)
and an early warning radar station in Czechoslova-
kia.10 Construction at these sites was initiated in 2009
at a cost of $712 million dollars.11
Russia’s government strongly protested NMD
during late 2008. Although the U.S. assured Russia
the theater system was to counter potential attacks
from Iran, Russian President Dmitri Medvedev,
in a state of the nation address (5 November, 2008)
declared, “Russia would respond by deploying short-
range missiles and jamming equipment to its western
border.” Similar protests were raised by China and
North Korea, noting that a new Anti-Ballistic Missile
Treaty was needed to replace the 1972 ABM treaty,
since neither the U.S. nor Russia supported it any
longer.7 Former Russian President Vladimir V. Putin
took a more aggressive view of NMD by stating that,
“Russia had successfully tested a new, strategic-missile
system, which would allow nuclear warheads to avoid
U.S. defenses.” He signaled that Russia was prepared
to commit billions of dollars to continue the arms race
with the U.S.7 Therefore, it is important to understand
the Russian developments.
Russia’s Topol-M missile program was initiated
in the late 1980s to develop a long-range missile that
could be launched from a silo or mobile vehicle. Its
original design, known as SS-X-25, was prepared by
the Moscow Institute of Thermal Technology (MITT)
and scheduled for operational use in December 1997.
When the Soviet Union broke up, MITT upgraded
the missile size and made it road-mobile or rail-

255
william yengst

launched. The stated intent was to replace older stra-


tegic missiles in the force by 2006. A unique feature
of the design is a three-stage, solid-propellant booster,
which accelerates much faster than conventional
ICBMs to be less vulnerable to attack during powered
flight. U.S. early warning radars that detect a launch,
and fire antiballistic interceptors would have little
time to respond. Booster stages also have auxiliary jets
and a three-dimensional flight control system that
permits avoidance maneuvers after the first few sec-
onds of flight. Finally, the reentry vehicle is described
as a ram-jet-boosted, supersonic-cruise vehicle, which
accelerates to speeds of Mach 4 to Mach 5.
In July 2001, observers witnessed a Topol-M test
launch in which the payload performed nonballis-
tic, high-altitude, very high-speed maneuvers.9 Three
years after the test, during the Russian “Security 2004
Command and Staff Training Exercises,” an SS-19,
Stiletto missile, launched on 18 February 2004 carried
an experimental reentry vehicle (HFL-VK) that was
observed to be “bouncing” between the upper atmo-
sphere and outer space. The vehicle defied skeptics
by surviving high heating and maneuver g. Within a
few days, the Russian Ministry of Defense released an
illustration of the flight trajectory, published by Novye
Izvestiya daily news and presented in the United Sates
by Nikolai Sokov of the Center for Nonproliferation
Studies.16 The trajectory, with its Russian Cyrillic
annotations (i.e., solid line for Topol-M MRV and
dashed line for ballistic missile) is provided in Figure
30.

256
lightning bolts

Figure 30. Russian Ministry of Defense


illustration of Topol-M trajectory

Careful consideration of the figure’s trajectory


reveals that powered flight follows a ballistic path and
ejects the MRV onto a hypersonic-glide path simi-
lar to the BGRV trajectory of Figure 12. However,
the terminal phase of flight shows a distinct dive to
low altitude and ram jet sustained low-level-run-in
(LLRI) maneuver to the target. This remarkable, but
not unexpected, maneuver is designed to penetrate
enemy defenses at altitudes below radar detection
horizons, the concept described under BGRV Follow-
On Developments (1968–1970) in Chapter 4.
On 1 November 2005, Russia conducted a test of
Topol-M from Kapustin Yar to the relatively close
Sary-Shagan test range in Balkhah (Kazakhstan).
Sary-Shagan has numerous tracking radars built to
measure warhead and penetration aids signatures. The
choice of this test site, rather than the normal Plesetsk

257
william yengst

ICBM range, was to prevent U.S. radars at Shemya


from monitoring separation of the payload from the
missile.12 The Topol-M missile, now known as SS-X-
27, has the following characteristics:12, 13

• Length: 74.4 feet


• Range: 10,000 kilometers (6,200 miles)
• Propellant: three-stage solid
• Booster Diameters:
ºº 1st Stage: 74.8 inches
ºº 2nd Stage: 63.4 inches
ºº 3rd Stage: 60 inches
• Payload: Maneuvering or gliding vehicle
2,200–2,600 pounds (<1,200 kg)
• Launch Weight: 47.2 tons
• Max Burnout speed: 23,714 feet/second

The very high-altitude, high-speed flight pro-


vides the MRV an ability to avoid NMD intercep-
tors. On 23 October 2008, Russia’s Strategic Missile
Chief, Colonel General Nikolai Solovtsov, claimed
its large rail-mobile, RS-24 missile, currently in the
strategic-operational inventory, was being upgraded
with three targetable Topol MRVs.14 The missile sep-
arates the vehicles in midcourse, changing direction
twice a minute to fool radars as to where the warheads
are heading. Each vehicle is assigned a target within
60 miles of its separation point.12 Solovtsov reported

258
lightning bolts

improved RS-24s were tested twice in 2007, and third


and fourth tests were planned for 2009.
Although little is known about the Topol-M
MRV, the RS-24 missile permits bounding estimates
of its size and weight. The upper stage of RS-24 is
94.5 inches in diameter. This permits three MRVs to
be carried with base diameters of no more than 40
inches. The MRV probably does not have a cylindrical
or biconic shape, particularly if it uses an air-breath-
ing, “ram-jet” engine; therefore, it probably requires
a larger diameter than a ballistic RV. The payload
capacity of RS-24, including its booster aerodynamic
nose faring, post-boost vehicle/guidance, and three
MRVs, is 4.05 tons (8,100 pounds).15 If one assumes
the booster faring weighs roughly 250 pounds and the
PBV with a missile-guidance system inside weighs
roughly 500 pounds, three MRVs must account for
7,350 pounds (or 2,450 pounds each). This is only 30
percent heavier than the U.S. BGRV experimental
vehicle. The length of a Topol-M MRV is less than
120 inches based on scaling from nose section photo-
graphs of the missile.
A Russian video illustration of the system showed
the vehicles cruising after separation and diving to very
low altitude to dash under defense radars and inter-
ceptors to strike targets. Russia may have borrowed
the U.S. LLRI concept to develop a strategic-weapon
system capable of avoiding or seriously degrading
NMD. It took them about two years to prove feasi-
bility of Topol-M (1998–2000), roughly three years
for advanced development (2003–2005), and four or
more years for engineering development of the RS-24
weapon system. The system could become operational

259
william yengst

in 2009. These development times are realistic; how-


ever, the process was slowed down by break-up of the
Soviet Union and economic-funding constraints.
Other nations may find maneuvering technology
a military advantage in solving missile-delivery prob-
lems. To prevent this, the U.S. should restrict docu-
mentation and access to critical technologies (e.g.,
guidance and control systems, aerodynamic theory,
and advanced materials). It is likely MRV technology
will be extended to Iran, Syria, and/or North Korea
by China. It is not surprising that China, India, and
Pakistan adopted the concepts to minimize develop-
ment times and costs for fielding M-11/11A, Agni, and
Sheheen missile systems. The strongest motivation for
continued development of MRVs by these countries is
to achieve high accuracy for effective delivery of con-
ventional or low-yield nuclear warheads.
Within the U.S., an interesting hypothetical case
for extending MRV technology would be to develop
a small force (e.g., 50 missiles) with long-range (i.e.,
greater than 6,000 miles) capable of carrying con-
ventional earth penetrating warheads (CEPWs) and
delivering them with superb accuracy (e.g., CEP < 30
feet). A single-warhead, highly accurate missile force
could be made subject to unconstrained International
Atomic Energy Agency (IAEA) inspections. Such a
weapon system would be politically and militarily use-
ful in holding very-hard targets at risk. The missiles
would be effective against leadership bunkers, major
command and control centers, underground weapon
development, storage facilities and terrorist caves, or
weapon depots. Although an “elite” special missile
force of this type could be highly useful in deterring

260
lightning bolts

aggressive leaders or terrorists, it would raise concerns


about arms control stability. Alternatively, the force
might provide a “peace-keeping” capability, which
could restore stability under serious crisis conditions.
Finally, MRV technology has provided several
highly useful civilian products. Development of small,
low-weight, accurate inertial guidance, and naviga-
tion systems have been adopted by commercial and
private aircraft, ships, and automotive vehicles. More
efficient insulation materials have permitted produc-
tion of lighter weight and smaller refrigerators. High-
temperature ablative materials are used by NASA for
spacecraft reentry vehicles and planetary probes, such
as the Mars Lander and Space Shuttle. Radio com-
munication through ionized reentry environments
is being used to support modern space operations.
Rugged computer memories and circuits have helped
development of laptops and rough-environment mili-
tary field units. Boost-glide energy management con-
cepts, transpiration cooling, and methods for insulat-
ing temperature-sensitive subsystems from high heat-
ing are being used by military aviation.
Consequently, technologies developed for MRVs
are now part of every-day life. Those who have taken
part in the 50-year history of this technology develop-
ment can be justly proud of their achievements.

261
references

First Maneuvering Reentry


Vehicle References

Chapter 1. Genesis of a New Concept:


1. “Aggregate Series,” Wikipedia Encyclo-
pedia, World Wide Web, 5 pages. http://
en.wikipedia.org/wiki/Aggregate_series.
2. “Hitler’s Last Weapons,” Jozef Gar-
linski, Times books, New York, NY,
1978, pages 3–5, 29, 68–70, 183, 187.
3. “How Weapons Work,” Christopher
Chant, Marshall Cavendish Publications,
New York, NY, 1973, pages 239–243.

263
william yengst

4. “Earnst A. Steinhoff,” New Mexico


Museum of Space History, World Wide
Web, 2 pages. http://www.nmspacemu-
seum.org/halloffame/detail.php?id=47
5. “Nazi V-2/A-4 Rocket Technology,”
Time Travel Research Center, World
Wide Web, 8 pages. http://www.zaman-
dayolculuk.com/V2RROCKET.htm
6. “A-9/A-10,” Encyclopedia Astronau-
tica, World Wide Web, 3 pages http://
www.astronautix.com/Ivs/a9a10.htm
7. “The History of the America’s Space
Program,” Mathew Behrens, World
Wide Web, Google search for Walter
Dornberger at McDonnell-Douglas.
8. “White Sands Missile Range Museum
Archival Holdings,” WSMR His-
tory, World Wide Web, 4 pages.
http://www.wsmr-history.org/Fir-
ingRecords1946 (to 1952).asp.
9. “The Exploration of Space: Atlas,”
Century-of-Flight, WWW, 2 pages.
http://www.century-of-flight.net/Avia-
tion%20history/Space/Atlas.htm
10. “SM-65 Atlas,” Wikipedia Encyclopedia.
org., World Wide Web, 3 pages. http://
en.wikipedia.org/Atlas_(missile)

264
lightning bolts

11. “News Release: Atlas Fact Sheet,”


Air Research and Development
Command, Patrick AFB, 15 Octo-
ber 1958, 7 pages. Google Search,
Atlas-missile-development.
12. “WS-199,” Directory of U.S. Military
Rockets and Missiles: Appendix 4:
Undesignated Vehicles, World Wide
Web, 6 pages. http://www.designation-
systems.net/dusrm/app4/ws-199.html
13. “Sandia Materials Enable Hypersonic
Flight,” Bill Sweetman, Defense Tech-
nology International, Aviation Week
& Space Technology Magazine,
15 May 2008, World Wide Web,1 page.
14. “Interview With Guenter F. Wendt,” by
Catherine Harwood, Titusville, Flor-
ida, 25 February 1999, 48 pages. http://
www.jsc.nasa.gov/history/oral_his-
tories/WendtG/GW_2–25–99.pdf
15. “David G. Simons,” International Space
Hall of Fame, New Mexico Museum
of Space History, World Wide Web, 1
page. http://www.NMSpaceMuseum.
org/halloffame/detail.php?id=93
16. “LGM-30A/B Minuteman I,” Global
Security.org, World Wide Web, 2
pages. http://www.globalsecurity.
org/wmd/system/lgm-30_1.htm.

265
william yengst

17. “File: Mark 5 Warhead.png,” Wiki-


pedia Media.org, 1 page. http:://
commons.wikimedia.org/wiki/
File.Mark_5_warhead.png
18. “Minuteman-I,” Missile Threat.com,
World Wide Web, 4 pages. http://
www.missilethreat.com/missilesoft-
heworld/id.77/missile_detail.asp.
19. “America’s War Machine,” Tom
Gervasi, Grove Press Inc., New
York, NY, 1984, page 85.
20. “Density Altitude,” Westport Vir-
tual Charters, World Wide Web, 2
pages. http://www.westportvirtual.net/
forums/viewtopic.php?f=14&t=96
21. “Constellations of Words: Draco,” Con-
stellations of Words.com, World Wide
Web, 1 page. http://www.constellation-
sofwords.com/Constellations/Draco.html
22. “Four Corners Marker Legit, Officials
Say,” The North County Times, Oceans-
ide, CA, 23 April 2009, page C-8.
23. “ABRES/BMRS: Western Missile
Test Range, Vandenberg AFB,” Silo-
world.com, Atlas 115 F, 11 October
1966, 1 page. http://www.siloworld.
com/Missiles%20%20LAUNCHES/
VAFB/ ABRES/abres.htm

266
lightning bolts

Chapter 2

An Introduction to ABRES:
1. “The Aerospace Corporation: 1960–1970,”
Tenth Anniversary Corporate Pub-
lication, 30 March 1970, pp. 1–42.
2. “A-35 Anti-Ballistic Missile Sys-
tem.” Wonderland.org, World
Wide Web, 4 pages. http://www.
wonderland.org.nz/a-35.htm
3. “Ballistic Missile Future Systems
& Technology Workshop,” Nor-
ton AFB, San Bernardino, CA,
14–16 April 1986, 13 pages with fold-
out ICBM Technology chart.
4. “Draper at 25,” Draper. com, World
Wide Web, page14 of 58. http://
draper.com/draper25/draper25.pdf
5. “24 Bit Computers,” Univac 1824 Guid-
ance Computer, World Wide Web,
1 page. http://www.usfamily.net/
web/labenson/Computers24.htm
6. “Minuteman-I,” Missile Threat.com,
World Wide Web, 4 pages. http://
www.missilethreat.com/missilesoft-
heworld/id.77/missile_detail.asp.

267
william yengst

7. “CAT.INIST,” Arc Jet Testing of


a DynasflR Dome,” World Wide
Web, 1 page. http://cat.inist.fr/?aMo
dele=afficheN&cpsidt=17386980
8. “Operations Requirements No.
F1716 (or No. 1) MBRV Drop Tests
Helicopter Drop-72 Hour Flota-
tion,” July 1965, R.G. Gorman, The
Aerospace Corporation, TOR-469
(S5810–47)-14, 2 July 1965, 29 pages.
9. “Atlas F,” Astronautix.com, World
Wide Web, 3 pages. http://www.
astronautix.com/lvs/atlasf.htm
10. “Minutes of the BGRV Guidance
Steering Adapter Critical Design
Review,” The Aerospace Corpora-
tion, L.E. Miller, SA-65–0000–80465,
27 December 1965, 73 pages.
11. “Minutes of the Fourth BGRV Interface
Meeting Held 26–27 May 1966,” J.B. Bly-
ler, The Aerospace Corporation, TOR-
669 (s6815–84)-32, 7 June 1966, 42 pages.
12. “SM-65 Atlas,” Wikipedia Encyclopedia.
org, World Wide Web, 5 pages. http://
en.wikipedia.org/wiki/Atlas_(missile)
13. “Systems Requirement Review for Ther-
mal Protection for Simulated BGRV
Missile Protuberances at GD/C 21
April 1966,” J.B. Blyler, 26 April 1966,
TOR-669 (s6815–84)-23, 18 pages.

268
lightning bolts

14. “Ballistic Coefficient,” Answers. com.,


World Wide Web, 3 pages. http://www.
answers.com/topic/ballistic-coefficient
15. “Dynamics of Atmospheric Reentry,”
Frank J. Regan and Satya M. Anan-
dakrishnan, Google Books Search,
Chapter 7 (Reentry Vehicle Par-
ticle Mechanics), pages 200–207.
16. “Orbital Requirements Document
Program 627A Maneuvering Ballistic
Reeentry Vehicle,” William C. Yengst,
The Aerospace Corporation, TOR 669
(S6810–47)-14 Revision 1), 18 July 1966,
43 pages, Appendix A (MBRV Profile).
17. “An Experimental Wind Tunnel Evalu-
ation of the Static Stability & Con-
trol Effectiveness of a Re-Entry Nose
Cone With Trailing Flaps at Mach 10,”
R. B. Hobbs, General Electric Report
GE-DIN-63SD531, March 1963.
18. “America’s War Machine,” Tom
Gervasi, Grove Press Inc., New
York, NY, 1984, page 91 and 96.
19. “Background and Supplementary
Data in Support of Boost-Glide Re-
entry Vehicle Study Program Plan,”
C. J. Blattner, McDonnell-Douglas
Report Number A-096, August 1963.

269
william yengst

20. “ Boost Glide Reentry Vehicle,” Friends-


Partners.org, World Wide Web, 1
page. http://www.friends-partners.
org/mwade/craft/boohicle.htm
21. “The X-Vehicles Advancing the Limits
of Technology,” Dr. Richard P. Hal-
lion, Air Force History and Muse-
ums Program, Washington, D.C.,
21 May 2002, 16 pages.
22. “Van Nostrand’s Scientific Ency-
clopedia,” D. Van Nostrand Co.,
New York, NY, 1958, Niobium
and Nickel, pg. 1118 and 1119.

Chapter 3

MBRV Development and Test Program:


1. “GE Re-entry Systems,” Ameri-
can Institute of Aeronautics and
Astronautics, Historic Aerospace
Site, World Wide Web, 11 pages.
2. “An Experimental Investigation of
Flap Induced Transitional Bound-
ary Separation on a Flat Plate at Mach
4.5: Part II,” J. N. Levine and M. L.
Roberts, General Electric Report
GE-ATDM 1:76, February 1965.

270
lightning bolts

3. Operations Requirements No.


F1716 (or No. 1) MBRV Drop Tests
Helicopter Drop-72 Hour Flota-
tion,” July 1965, R.G. Gorman, The
Aerospace Corporation, TOR-469
(S5810–47)-14, 2 July 1965, 29 pages.
4. “24 Bit Computers,” US Family.net,
1824 C Computer, World Wide Web,
1 page. http://www.usfamily.net/
web/labenson/Computers24.htm
5. “Maneuvering Ballistic Reentry Vehicle
Technology,” by F. Hallowell, F. Spampi-
nato, J. Cameron, Pen-X Paper, Num-
ber 46, LMSC-B160, March 1965.
6. “Careers, Page 1: Larry Bolton Compo-
nent Engineer” US Family.net, World
Wide Web, 1 page. http://www.usfam-
ily.net/web/labenson/Careers1.htm
7. “Vandenberg 576A1,” Astronautix.com,
World Wide Web, 2 pages. http://www.
astronautix.com/sites/van576a1.htm
8. “Atlas,” Astronautix.org, World Wide
Web, chronological launches. http://
www.astronautix.com/lvs/atlas.htm
9. “Vandenberg ABRES,” Astronautix.com,
World Wide Web, 3 pages. http://www.
astronautix.com/sites/vanabres.htm

271
william yengst

10. “The Aerospace Corporation: 1960–


1970, Serving America,” Corporate
Publication, Times Mirror Press, Los
Angeles, 30 March 1970, page 37.
11. “Maneuvering Aerothermal Technology
(MAT) Program: Data Bibliography,”
by A. Martelluci and S. Weinberg, Sci-
ence Applications Inc., Wayne, Penn-
sylvania, for USAF?BMO Norton AFB,
Contract Number F04701–80-C-0033,
March 1981, 158 pages. http://stinet.dtic.
mil/cgi-bin/GetTRDoc?AD=ADA1188
76&Location=U2&Doc=GetTRDoc.pdf
12. “Ballistic Missiles and Reentry Sys-
tems: The Critical Years,” Richard
A. Hartunian, The Aerospace Cor-
poration Magazine, February 2003, 6
pages. http://www.aero.org/publica-
tions/crosslink/winter2003/02.html

Chapter 4

BGRV Development and Test Program:


1. “Lambert-St. Louis International Air-
port,” Global Security.org, World Wide
Web, 2 pages. http://www.globalsecu-
rity.org/military/facility/lambert.htm
2. “News from Retired Teammates:
Updated 22 June 2008,” World Wide
Web, harpooner@macsoldteam.com

272
lightning bolts

3. “Heat-Transfer Tests of a Transpiration-


Cooled Boost-Glide Re-entry Vehicle
Nose Cone,” by J.P. Rhudy and J.D.
Magan Jr., AEDC Report No. AEDC-
TDR-64–179, September 1964.
4. “Force Test Results on a Series of
Boost-Glide Re-entry Vehicles at Mach
Numbers from 1.5 to 10,” J.C. Uselton,
AEDC Report No. AEDC-TR-64–70,
April 1964 and “Pressure Distribu-
tion Tests on a Boost-Glide Re-entry
Vehicle at Mach Numbers from 1.5 to
10,” by J.C. Uselton, AEDC Report No.
AEDC-TR—64–204, October 1964.
5. “Aurora & Beyond: Paul Czysz on
Hypersonic Aircraft,” 26 July 2006,
World Wide Web, 13 pages. http://www.
americanantigravity.com/documents/
Paul-Czysz-Hypersonic-interview.pdf
6. “Projectile,” My Site, J. Calvert, World
Wide Web, 2 pages. http://mysite.
du.edu/~jcalvert/phys/ballist.htm
7. “CA-136 Chicago,” Global Secu-
rity.org, World Wide Web, 3 pages.
http://www.globalsecurity.org/mili-
tary/agency/navy/ca-136.htm
8. “Centers: Ground Control Stations
(KIK),” World Wide Web, 4 pages. http://
www.russianspaceweb.com/kik.html

273
william yengst

9. “Vandenberg ABRES,” Astronautix.com,


World Wide Web, 3 pages. http://www.
astronautix.com/sites/vanabres.htm and
“Atlas,” Astronautix.org., World Wide
Web, chronological launches. http://
www.astronautixcom/lvs/atlas.htm
10. “Boost Glide Reentry Vehicle,” Astro-
nautix.com, World Wide Web, 1 page.
http://www.friends-partners.org/
partners/mwade/craft/boohicle.htm
11. “Free-Flight Range Tests of BGRV
Configuration at Hyperveloci-
ties,” by G. L. Winchenbach and W.
R. Lawrence, AEDC Report No.
AEDC–TR-67-233, November 1967.
12. “Methods for Reducing and Evalu-
ating BGRV Aerodynamic Heat-
ing Data,” by R. J. Masek, B. M.
Levine, and D. D. Philpott, AFFDL,
Report No. TR-68–77, May 1968.
13. “BGRV Final Report: Three Vol-
umes,” for USAF/SAMSO, Report No.
SAMSO TR-68–456, September 1968.
14. “Walter Dornberger,” About.com:
Inventors, World Wide Web, 1 page.
http://inventors.about.com/library/
inventors/blrocketDornberger.htm
15. “SA-4 Ganef,” Everything2.com,
World Wide Web, 1 page. http://every-
thing2.co/tital/SA-4%2520Ganef

274
lightning bolts

16. “SA-300,” Wikipedia Encyclope-


dia.org., World Wide Web, 2 pages.
http://en.wikipedia.org/wiki/S-300.
17. “Influence of Boost Glide Re-Entry
Vehicle Environment on Coated and
Bare Refactory Metals,” T. H. Fitz-
patrick, Technical Report AD0806939,
May 1965-February 1966. DTIC Met-
allurgy, Unmanned Spacecraft.
18. “Stealth Technology: The Art of Black
Magic,” J. Jones, Tab Books Inc., Blue
Ridge Summit, PA, 1989, pages 80–82.
19. “Fastest and Highest,” Blade-Tri-
bune Newspaper, The San Diego
Union Tribune, 17 October 1988, page
A-8 (Associated Press photo).
20. “ Hypersonic Aurora: a Secret Dawn-
ing?” Bill Sweetman, Jane’s Defense
Weekly, 12 December 1992, pages 14–16.
21. “Minutes of the Fourth BGRV Interface
Meeting held 26–27 May 1966,” J.B. Bly-
ler, The Aerospace Corporation, TOR-
669 (s6815–84)-32, 7 June 1966, 42 pages.

275
william yengst

Chapter 5

MARCAS Development and Test Program:


1. “Athena Missile Found in Mexico,”
El Paso Times, 6 August 1970, page
B-1. Also, “The Athena That Got
Away,” Jim Eckles, White Sands Mis-
sile Range, World Wide Web, 5 pages.
http://www.wsmr.army.mil/pao/
factsheets/AthenatoMexico.htm
2. “Maneuvering Re-entry Controls and
Ablation Studies,” First Quarterly Prog-
ress Report, D. B. Harmon Jr., USAF/
BSD/TR-64–173, December 1964.
3. “Ballistic-Missile Research With
Athena,” Robert F. Friedman and Jack
Reed, Atlantic Research Corpora-
tion, Journal of Spacecraft, Volume
2, No. 5, September-October, 1965.
4. “Athena,” Atlantic Research Com-
pany, Directory of U.S. Military
Rockets and Missiles, World Wide
Web, 2 pages. http://www.designation-
systems.net/dusrm/app4/athena.html
5. “Maneuvering Re-entry Control and
Ablation Studies: Volume III, Inves-
tigation of Pressure and Enthalpy on
Ablative Materials,” G. P. Johnson,
USAF/BSD-TR-65–312, July 1965.

276
lightning bolts

6. “Digital Collections,” University of


Utah, J. Willard Marriott Library, Den-
ver Post Article, “Athena Launches
Conclude,” 26 August 1973.
7. “Green River: 11 October,” Astronautix.
com, World Wide Web, few lines. http://
www.astronautix.com/sites/greriver.htm
8. “MARCAS-2 Re-entry Vehicle,”
Douglas Report Number DDC-ARB-
N-58454, September 1966 (U) and
“MARCAS II Final Report Volume
II: Jet Interaction Investigation,” by
J.S. Murphy, R. E. Bloomquest, and F.
W. Spaid, BSD-TR-66–246, 1966.
9. “Athena V-1168, Flight 101 (2 May 1968),”
Harold M. Richart, Army Electronics
Command, WSMR Atmospheric Sci-
ences Laboratory. http://www.storm-
ingmedia.us/94/9492/0949238.html.
10. “Athena,” Directory of U.S. Military
Rockets and Missiles,” World Wide
Web, 2 pages. http://www.designation-
systems.net/dusrm/app4/athena.html
11. “Radars for Ballistic Missile Defense
Research,” by Philip A. Ingwersen
and William Z. Lemnois, World
Wide Web, 21 pages. http://www.11.
mit.edu/publications/journal/pdf/
v0112_n02/12_2ballisticmissiledefense.pdf.

277
william yengst

12. “Maneuvering Aerothermal Technol-


ogy (MAT) Program: Data Bibliogra-
phy,” by A. Martelluci and S. Weinberg,
Science Applications Inc.,Wayne, PA,
for USAF/BMO, Norton AFB, con-
tract No. F04701–80-C- 0033, March
1981, 158 pages. http://stinet.dtic.mil/
cgi-bin/ GetTRDoc?AD=ADA118876
&Location=U2&Doc=GetTRDoc.pdf
13. “Reentry Systems: The Critical Years,”
Richard A. Hartunian, Crosslink, Volume
4, No. 1, Winter 2002/2003, pages 7–11.
14. “An Oil Flow Study of a Sonic
Reaction Jet Ejecting from a Body
of Revolution into a Free Stream of
Mach Number Range 1.75 to 4.5,” Troy
A. Street, Technical Report 0557178,
21 April 1970, 355 pages. http://www.
stormingmedia.us/55/5571/0557178.html.
15. “Method of Imparting Hypergolicity to
Non-Hypergolic Rocket Propellants,”
U.S. Patent 4316359, David C. Sayles,
23 February 1982, 2 pages. http://www.
freepatentonline.com/4316359.html
16. “Steering by Plasma Discharge-U.S.
Patent 7002126,” Description, Back-
ground of the Invention, 21 February
2006, 8 pages. http://www.patentstorm.
us/patents/7002126/description.html.

278
lightning bolts

Chapter 6.

Second Generation Maneuvering


Reentry Vehicles:
1. “UGM-73 Poseidon,” Wikipedia
Encyclopedia. org., World Wide
Web, 1 page. http://en.wikipedia.
org/wiki/Poseidon_missile.
2. “Trident-I C-4 FBM/SLBM,” FAS org.,
World Wide Web, 2 pages http://www.
fas.org/nuke/guide/usa/slbm/C-4.htm
3. “Ballistic Missiles and Reentry Sys-
tems: The Critical Years,” Richard
A. Hartunian, Aerospace Corpora-
tion, Crosslink, Winter 2003, 6 pages.
http://www.aero.org/publications/
crosslink/winter2003/02.html
4. “USAF Ballistic Missile Programs:
1969–1970,” Jacob Neufeld, Office of the
Air Force History, June 1971, pages 43–50.
5. “Dial Pack,” Global Scurity. Org, World
Wide Web, 1 page. http://www.globalse-
curity.org/wmd/ops/dial-pack.htm
6. “The Status of U.S. Nuclear Weapon Pro-
grams,” The Defense Monitor, Volume X,
Number 8, 1982, page 3. http://docs.nrdc.
org/nuclear/files/nuc_81010101a_n15.pdf

279
william yengst

7. “The Peacekeeper (MX) ICBM,” World


Wide Web, 2 pages. http://nuclearweapo-
narchive.org/Usa/Weapons/Mx.hml
8. “The Status of U.S. Nuclear Weapon
Programs,” The Center For Defense
Information, World Wide Web, 3 pages.
9. “Advanced Reentry Vehicle Flight
Tested,” Bruce A. Smith, Avia-
tion Week & Space Technology,
11 February 1980, pages 43–45.
10. “The Aerospace Corporation:
1960–1970,” Tenth Anniversary Cor-
porate Book, El Segundo, CA, 30
March 1970, pages 169–170.
11. “ACE Program Final Report,” by D.
V. Magill,” (Program Description,
Volume 2) and (Design Analysis and
Flight Test Results, Volume 3), USAF/
SAMSO TR-75–23, January 1975.
12. “Static Stability and Drag Characteris-
tics of the Final SAMSO/McDonnell-
Douglas Advanced Control Experi-
ment (ACE) Vehicle Configuration
at Mach Numbers 16 and 19,” by L.
G. Siler, AEDC-TR-63, April 1973.
13. “Atlas(SM-65/CGM-16/HGM-16)
ICBM,” skyrocket, World Wide Web,
6 pages. http://www.skyrocket.de/
space/doc_lau/atlas_icbm.htm

280
lightning bolts

14. “Wind Tunnel Measurements of Aero-


heating and Pressure for 52% Model of
AMaRV at Mach 9 to 13,” McDonnell-
Douglas Report MDC-G7608, May 1975.
15. AEDC Fellows: James C. Uselton 2005,”
Arnold Air Force Base, World Wide
Web, 1 page. http://www.arnold.af.mil/
library/factsheets/factsheet.asp?id=13125
16. “Stability for Bi-Conic Configuration,”
by C. J. Harris, General Electric Report,
GE-9154-TDM-76–005, May 1976.
17. “Test Plan for AMaRV Parametric and
Configuration Refinement Wind Tun-
nel Test Series,” Aerospace Corporation
Report EX-76–08760, November 1976.
18. “Advanced Maneuvering Reentry Vehi-
cle (AMaRV ) OPV Baseline Report,”
MDC-G6608, December 1976.
19. “AMaRV Preliminary Design Review,”
LMSC-L512728, June 1977.
20. “Maneuvering Aerothermal Technology
(MAT) Program: Data Bibliography,”
by A. Martelluci and S. Weinberg, Sci-
ence Applications Inc., Wayne, Penn-
sylvania, for USAF?BMO Norton AFB,
Contract Number F04701–80-C-0033,
March 1981, 158 pages. http://stinet.dtic.
mil/cgi-bin/GetTRDoc?AD=ADA1188
76&Location=U2&Doc=GetTRDoc.pdf

281
william yengst

21. “Atmospheric Reentry,” Wikipedia


Encyclopedia. org, World Wide Web,
6 pages. http://en.wikipedia.org/wiki/
Reentry_vehicle. and “Reentry Systems:
The Critical Years,” Richard A. Har-
tunian, Crosslink, Winter 2002/2003,
Volume 4, No. 1, pages 7 to 21. And
“Atmospheric Reentry,” http://wapedia.
mobile/en/Atmospheric_reentry?t=4.3
22. “ICBM Reentry Vehicle Navigation
System Development at Honeywell,” J.
Boutelle, S. P. Kau, and C. J. Marino Jr.,
IEEE Position Location and Navigation
Symposium, 20–23 April 1998, pages 294–
302. http://ieeexpore.ieee.org/Xplore/
login.jsp?url=/ie4/5490/14769/00670094.
pdf?arnumber=670094
23. “America’s War Machine,” Tom
Gervasi, Grove Press Inc., New
York, NY, 1984, page 91 and 103.
24. “Vandenberg LF03,” Astronautix. Com,
World Wide Web, 3 pages. http://www.
asronautix.com/sites/vanglf03.htm
25. “DC-X,” Aeronautix.com, World
Wide Web, 3 pages (http://www.astro-
nautix.com/lvs/dcx.htm) And “Edit-
ing Atmospheric reentry (section),”
Wikipedia Encyclopedia.com, World
Wide Web, 1 page http://en.wikipedia.
org/w/index.php?title=Atmospheric_
reentry&action=edit&section=7

282
lightning bolts

26. “Trident Missile,” Wikipedia Ency-


clopedia.org., World Wide Web,
3 pages. http://en.wikipedia.
org/wiki/Trident_missile.
27. “Trident II D-5 Fleet Ballistic Mis-
sile,” FAS. Org., World Wide
Web, 4 pages. http://www.fas.org/
nuke/guide/usa/slbm/d-5.htm.
28. “Trident II D-5 Fleet Ballistic Mis-
sile,” Lockheed/ Failure Analysis
(FA), World Wide Web, 3 pages.
http://www.failureanalysisco.com/
missile_telemetry_prognostics.
29. “Trident Missile,” Times Daily.
Com, Florence, Alabama, World
Wide Web, 1 page. http://www.
timesdaily.com/apps/pbcs.dil/
section?category=News&Template.
30. “US Nuclear Forces, 2006,”
Google Search, World Wide Web,
8 pages. http://209.85.141.104/
search?q=cache:04i-BtOGPUJ (www.
sipri.org/ Mark+5+Reentry+Vehicle)
31. “The W-88 Warhead,” Nuclear Weapon
Archieve. Org., 1 October 1997, World
Wide Web, 3 pages. http://nuclearweapo-
narchieve.org/Usa/Weapons/W88.html

283
william yengst

32. “U.S./Russia Divided Over Arms Agree-


ment,” Oxford Analytica, 22 November
2007, 2 pages. also, “SORT” Wikipedia
Encyclopedia, org., World Wide Web,
1 page. (http://en.wikipedia.org/wiki/
SORT)http://www.forbes.com/2007/11/22/
us-russia-weapons-cx-11220xford.html
33. “A Novel Approach for Trajectory Shap-
ing Reentry Vehicle Designs,” T.C.
Lin, S.A. McKeel, L.K. Sproul, and
T.P. Shivananda, TRW/Space and
Missile Systems plus B.L.Reeves, J.
Kealos, P. Fote, and N. Thyson/Tex-
tron Systems Division, 1999–2000,
American Institute of Aeronautics and
Astronautics, World Wide Web.
34. “Method and System for Provid-
ing Cruciform Steered, Bent Biconic
and Plasma Suppression for maximum
Accuracy,” U.S. Patent 7267303, World
Wide Web, Abstract. http://www.pat-
entstorm.us/patents/7267303.html
35. “Lockheed HGV,” Directory of U.S.
Military Rockets and Missiles, Appendix
4: Undesignated Vehicles, World Wide
Web, 2003, 2 pages. http://www.designa-
tion-systems.net/dusrm/app4/hgv.html

284
lightning bolts

36. “Nuclear Heuristics: Selected Writ-


ings of Albert and Roberta Wohlstet-
ter,” by Robert Zarate and Henry
Sokoloski, Strategic Studies Insti-
tute, U.S. Government Publication,
January 2009, Pages 52–55 (The Long
Range Research Planning Program).
37. “The Cuban Missile Crisis,” Wiki-
pedia Encyclopedia.org, World Wide
Web, 10 pages. http://en.wikipedia.
org/wiki/Cuban_Missile_
Crisis#Planning_an_ American Report.
38. “ARPA Long Range R & D Program-
Systems Study: Final Report,” W.C.
Yengst and W.S. Vance, Science Appli-
cations Incorporated SAI-75–512-LJ,
(AD C001219), March 1975, 200 pages.

Chapter 7

Pershing II MARV Development Program:


1. “MGM-31 Pershing,” Wikipe-
dia Encyclopedia.org, World Wide
Web, 8 pages.http://en.wikipedia.
org/wiki/Pershing_II#Pershing_II

285
william yengst

2. “Tomahawk,” Astronautix.com, World


Wide Web, 6 pages. http://www.
astronautix.com/lvs/tomahawk.htm
3. “America’s War Machine,” Tom Ger-
vasi, Grove Press Inc., New York,
NY, 1984, pages 106 and 327.
4. “The Pershing Weapon System and Its
Elimination,” Redstone Arsenal, World
Wide Web, 6 pages. http://www.redstone.
army.mil/history/systems/pershing/
5. “Introduction to Scene Matching Mis-
sile Guidance Technologies,” National
Air Intelligence Center Wright-Pat-
terson AFB, Storming Media, World
Wide Web, 1 page. http://www.storm-
ingmedia.us/93/9345/A934513.html
6. “W-86,” Wikipedia Encyclope-
dia. org., World Wide Web, 1 page.
http://en.wikipedia.org/wiki/W86
7. “BGM-109 Tomahawk,” Wikipe-
dia Encyclopedia. org., World Wide
Web, 2 pages. http://en.wikipedia.
org/wiki/BGM-109_Tomahawk
8. “Investigation of German V1 Cruise
Missiles Employed Against England,”
T.G. Smolin and W.C. Yengst, Sci-
ence Applications International Cor-
poration, 28 February 1978, 25 pages.

286
lightning bolts

9. “Hitler’s Last Weapons,” Jozef Garlin-


ski, Times Books, New York, NY, 1978,
pages 68, 104–106, 176–178 and 182–183.
10. “V-1 Flying Bomb,” Wikipedia
Encyclopedia. org., World Wide
Web, 2 pages. http://en.wikipedia.
org/wiki/V-1_flying_bomb
11. “The V-Weapons,” After the Battle
Number 6, Winston G. Ramsey, Strat-
ford, London, 1974, pages 16–29.
12. “Doodlebugs and Rockets (V-1
and V-2),” World Wide Web, 2
pages. http://www.486th.org/pho-
tos/letters/doodlebugs.htm
13. “Hitler,” Astronautix.com, World
Wide Web, 4 pages. http://www.
astronautix.com/astros/hitler.htm.
14. “Ground Launched Cruise Mis-
sile,” Wikipedia Encyclopedia.org.,
World Wide Web, 3 pages. http://
en.wikipedia.org/wiki/GLCM
15. “B61–11 Concerns and Background,”
The Brookings Institution, Washing-
ton, D.C., Los Alamos Study Group
(10 February 1997), World Wide Web.

287
william yengst

16. “NRDC: Countering Proliferation


or Compounding It?” T.B. Cochane,
M.G. McKinzie, and R.S. Norris,
Natural Resources Defense Council
(NRDC), May 2003, Report. http://
www.nrdc.org/nuclear/bush/abb.pdf
17. “Effects of Nuclear Earth Pen-
etrator Weapons and Other Weap-
ons,” National Research Council,
146 page report, Chapter 3 (pages
18–20). http://books.nap.edu/open-
book.php?record_id=11282&page=18.
http://www.brookings.edu/projects/
archieve/nucweapons/lasg.aspx
18. “Hard Targets That Could Not Be
Destroyed By Conventional Weap-
ons,” William C. Yengst and Charles
C. Deel II, Science Applications Inter-
national Corporation, for Defense
Nuclear Agency, March 1993, 277 pages.
19. “Nuclear Bunker Buster,” Wikipe-
dia Encyclopedia.org, World Wide
Web, 4 pages. http://en.wikipedia.
org/wiki/Nuclear_bunker_buster

288
lightning bolts

Chapter 8

Indian Agni-II and–III MRV Developments:


1. “The Kashmir Dispute,” William C.
Yengst and Deedee White, Science
Applications International Corpora-
tion, for Defense Threat Reduction
Agency, April 1999, pages 6–8 and 15–16.
2. “The Indian Drive Toward Weaponiza-
tion: Policy Statements,” Michael Kraig,
Federation of American Scientists Space
Policy Project, World Wide Web, 3
pages. http://www.fas.org/nuke/guide/
india/missile/agni-improvements.htm
3. “A Success Story,” T.S. Subramanian,
Frontline, Volume 22, Issue 20, Septem-
ber 24 to 7 October 2005, 5 pages. http://
www.hinduonnet.com/fline/f12220/
stories/20051007002009200.htm
4. “Agni Missile System,” Wikipedia
Encyclopedia. org., World Wide Web,
7 pages. http://www.en.wikipedia.
org/wiki/Agni_missile_system
5. “Short-Range Agni Missile Test-Fired,”
The Tribune, Balasore, India 25 Janu-
ary 2002, 2 pages. http://www.tribunei-
ndia.com/2002/20020126/main1.htm.

289
william yengst

6. “The Agni Missile is for Nuclear Aggres-


sion,” Air Marshall (Retired) Ayaz
Ahmed Khan, Pakistan, World Wide
Web, 2 pages. http://www.defensejour-
nal.com/2001/feb/agni-missile.htm
7. “Strategic Missiles,” Arun S. Vish-
wakarma, Indian Defense Review,
Volume 22.1, 11 pages. http://www.
indiandefensereview.com/?p=33.
8. “Agni-II IRBM,” Based on Front-
line and India Today, reports, 2
pages. http://www.bharat-rakshak.
com/Missiles/Agni-II.html
9. “Agni-Strategic Ballistic Missile,” Arun
Vishwakarma, Bharat Rakshak.com,
World Wide Web, 8 pages. http://www.
bharat-rakshak.com/Missiles/Agni.html
10. “Indian Nuclear Weapon Pro-
gram: Present Capabilities,” Nuclear
Weapon Archive, World Wide web,
6 pages. http://nuclearweaponar-
chive.org/India/IndiaArsenal.html
11. “Agni,” FAS. org., World Wide Web,
2 pages. http://www.fas.org/nuke/
guide/india/missile/agni.htm
12. “India Profile,” NTI Country Over-
views: India: Missile Imports/
Exports, World Wide Web, March
2003. http://www.nti.org/e_research/
profiles/India/Missile/1756.html

290
lightning bolts

13. “Peter Coats Intelligence Blog:


Update- India’s New K-15 Sagarika
Missile is H-Bomb Capable,” Spying
Bad Things, 26 February 2008, World
Wide Web,11 pages. http://spyingbadth-
ings.blogspot.com/2008/02/india-
successfully- launches-k-15-slbm.html
14. “Agni Missile to Get Multiple War-
heads, Ajai Shukla, Business Stan-
dard, 28 January 2008, World Wide
Web, 5 pages. http://ajaishukla.
blogspot.com/ 2008/01/agni-missile-
to-get-multiple-warheads.html

Chapter 9

China’s M-11 and M-11A MRV Developments:


1. “China’s Nuclear Weapon Development,
Modernization and Testing,” NTI.
Org., World Wide Web, 3 pages. http://
www.nti.org/db/China/wnwmdat.htm
2. “PRC Theft of U.S. Nuclear Warhead
Design Information: page 2,” C-Span:
report on Chinese Espionage, 1 page. And
“China Increases Its Missile Forces While
Opposing U.S. Missile Defense,” Rich-
ard D. Fisher Jr., 7 April 1999, 13 pages.
http://www.christusrex.org/www2/china/
theft/pg2.html and http://www.heritage.
org/research/missiledefense/bg1268.cfm

291
william yengst

3. “The Forgotten ‘Spy’ Case of a Rocket


Scientist,” Perter Grier, FAS. org., World
Wide Web, 3 pages. http://www.fas.org/
irp/news/2000/11/irp-001107-whl.htm
4. “Lightning Bugs and Other Recon-
naissance Drones,” William Wag-
ner, Aero Publishers Inc., 1982,
Chapter 11, pages 75–82.
5. “M-11 Variant,” Missile Threat com.,
World Wide Web, 3 pages http://
www.missilethreat.com/missilesoft-
heorld/id.65/missile_detail.asp
6. “DF-11 (CSS-7/M-11) (China),
Offensive Weapons,” Jane’s Strate-
gic Weapon Systems, World Wide
Web, 2 pages. http://janes.com/
extracts/extract/jsws/jsws0413.html
7. “China’s Missile Exports and Assis-
tance to Iran,” NTI Research.org,
World Wide Web, 6 pages. http://
www.nti.org/db/china/miranpos.htm
8. “DF-11/A (M-11, CSS-7) Short-
Range Ballistic Missile,” Sino-
Defense.com, World Wide Web,
6 pages. http://www.sinodefense.
com/strategic/missile/df11.asp
9. “Hatf-3/Shaheen-I/M-11,” FAS.
org., World Wide Web, 2 pages.
http://www.fas.org/nuke/guide/
pakistan/missile/hatf-3.htm

292
lightning bolts

10. “Shaheen Missile,” Wikipedia Encyclope-


dia.org, World Wide Web, 3 pages. http://
en.wikipedia.org/wiki/Shaheen_missile
11. “Land-Based Ballistic Missiles,”
Aeronautics, World Wide Web, 2
pages. http://www.aeronautics.ru/
archive/wmd/ballistic/css6–01.htm
12. “CSS-7 Mod 2,” Missile Threat.com,
World Wide Web, 2 pages. http://
www.missilethreat.com/missilesoft-
heworld/id.28/missile_detail.asp
13. “Nuclear Arms Race in Asia,” Charles
R. Smith, Newsmax, 10 October 2002,
5 pages. http://archieve.newsmax.com/
archives/articles/2002/10/9/184336.shtml

Chapter 10

Observations and Emerging MRV Threats:


1. “Wright Brothers,” Wikipedia Ency-
clopedia.org, World Wide Web,
18 pages. http://wwwen.wikipe-
dia.org/wiki/Wright_brothers
2. “Know Aviation,” F.M. Mason and M.C.
Windrow, Doubleday & Company Inc.,
garden City, New York, NY, 1973, page 11.

293
william yengst

3. “Human Factors of Powered Flight:


The Wright Brothers’ Contributions,”
Stanley R. Mohler, Aviation, Space, and
Environmental Medicine, Volume 75,
No. 2, February 2004, pages 184–188.
4. “Propeller Propulsion,” NASA
Glen Research Center, World Wide
Web, 2 pages. http://wright.nasa.
gov/airplane/propeller.html
5. “Strategic Arms Limitations Talks,”
Absolute Astronomy.com, World
Wide Web, 2 pages. http://www.
absoluteastronomy.com/topics/
Strategic_Arms_Limitation_Talks
6. “Intermediate-Range Nuclear Weapon
Forces,” Wikipedia Encyclopedia.
org, World Wide Web, 2 pages. http://
en.wikipedia.org/wiki/intermedi-
ate_range_nuclear_weapon_forces
7. “Russia Tests Missile That Could
Evade U.S. Defense,” Los Ange-
les Times, 19 February 2004, 3 pages.
http://articles.latimes.com/2004/
feb/19/world/fg-missiles19
8. “The US Missile Defense Sys-
tem,” Taking it Global-Panorama,
World Wide Web, 1 page. http://
nl.takingitglobal.org/express/pan-
orama/article.html?ContentID=16025.

294
lightning bolts

9. “The Missile That Does Not Care


[re: Topol-M, Missile Defense],”
Johnson’s Russia List, 14 February
2006, 2 pages. http://www.cdi.org/
russia/johnson/2006–42–27.cfm
10. “National Missile Defense,” Wikipe-
dia Encyclopedia.org, World Wide
Web, 5 pages. http://en.wikipedia.
org/wiki/National_missile_defense
11. “Russia Gets U.S. Proposals on Arms,
Missile Defense,” The San Diego Union
Tribune, 8 November 2008, page A11.
12. “Missile Threat: Russia Tests New
Maneuvering Warhead on Topol-M,”
Missile Threat.com, 24 October 2008,
3 pages. http://www.missilethreat.
com/archives/id.1842/detail.asp
13. “RT-2UTTH Topol-M,” Wikipe-
dia Encyclopedia.org, World Wide
Web, 2 pages. http://en.wikipedia.
org/wiki/RT-2UTTH_Topol_M
14. “Russia Modernizes Its Missiles,”
The North County Times, Oceans-
ide, CA, 23 October 2008, page A-8.
15. “RS-24/SS-X-29?” Global Security.
Org., World Wide Web, 2 pages.
http://www.globalsecurity.org/
wmd/World/russia/rs-24.htm.

295
william yengst

16. “Military Exercises in Russia: Naval


Deterrence Failures Compensated
by Strategic Rocket Forces,” CNS
James Martin Center for Nonpro-
liferation Studies, 24 February 2004,
Nikolai Sokov, 2 pages. http://cns.
mil.edu/stories/040224.htm
17. “Russia’s New BMD-Beater,” Defense
Technology. org., World Wide Web,
1 page. http://www.defensetech.
org/archives/cat_missiles.html
18. “CRS Report for Congress: National
Missile Defense and Alaska,” Steven
A. Hildreth, 13 July 2001, 6 pages.

296
photograph credits

The author is greatly indebted to a number of peo-


ple and their organizations for permission to reprint
photographs. Very few photographs were made or are
available for developmental weapons systems dating
back a half century. The following people and organi-
zations deserve special credit and my sincere, Thanks.
This book could not have been prepared without their
dedicated and most helpful research.

Figure 1. The photograph of Dr. Ernst A. Stein-


hoff was published by the New Mex-
ico Museum of Space History. NASA
Images Internet Archive is credited as
the source of the picture and Ms. Bonnie
Real, Project Manager of NASA Images,
reported it is not restricted from reprint.

297
william yengst

Figure 2. This illustration was presented in “Nazi


V-2/A-4 Rocket Technology,” Section
1, Pre-Operational History. The article
by Time Travel Research Center, Cetin
BAL-GSM:+90 05366063183-Turkey/
Denizili, refers to “Encyclopedia Astro-
nautica,” as its source (www.zaman-
dayolculuk.com/cetinbal/v2RRocket.
htm). Encyclopedia Astronautica states
unmarked images used on the site are
believed to be in the “public domain” and
can be reprinted.
Figure 4. This photograph was published in the
Internet article, “WS-199: Directory
of U.S. Military Rockets and Missiles,
Appendix 4,” by Andreas Parsch, at
(www.designation-systems.net/dusrm/
app4/ws-199.html). The article was
copyrighted 2003-2005 and its origi-
nal source was “National Archives and
Records Administration (NARA), Still
Pictures Reference.”
Figure 5. This photo of Galosh missile (A-35) was
taken from Aerospace Power Journal,
Summer 2002, 16:7483, “Maintaining
Friendly Skies,” by Lt. Col. W. Grau and
Dr. Jacob W. Kipp, Figure 7, ABM-350
approved for public release with unlim-
ited distribution. Approval was granted
by Air Force Chief of Staff General John
P. Jumper and Editor Lieutenant Colo-
nel Scott G. Wiershke. Another image

298
lightning bolts

of the ABM-1 (A-35) is provided in a


Soviet article on the missile defense sys-
tem, published in Moscow, 10 May 2000.
Figure 7. The drawings of this figure are from my
files and memory; however, the small inset
photograph taken by USAF/Vandenberg
AFB is credited to Gary G. Baker (Silo-
Man) and reprinted with his permission
from an inset on Atlas 149F, 8 August
1966, launch: “ABRES/BMRS: Western
Missile Test Range, Vandenberg AFB.”
(www.siloworld.com/MISSILES%20
%LAUNCHES/VAFB/ABRES/abres.
htm). This may be the only MBRV image
available in open literature.
Figure 8. This rare, close-up picture produced by
the USAF was first published during
2001 by Mark Wade in his article, “Boost
Glide Re-entry Vehicle,” for Friends-
Partners.org (www.friends-partners.org/
partners/mwade/craft/boohicle.htm). It
is one of several images credited to ©
Mark Wade and Astronautix.com, which
permits reprints at (www.astronautix.
com/articles/conssite.htm).
Figure 10. The image of launch pads 576a1 and
576a2 is a United States Geological Sur-
vey satellite photo available for public
use and found in a University of Wyo-
ming (Vandenberg AFB branch) article
by James Kirk, “Vandenberg AFB.” The
article identifies Launch Complex 65-1

299
william yengst

A, B, and C (also known as ABRES A1,


A2, and A3 in 1965-1971) at (http://asu-
wlink.uwyo.edu/~jimkirk/vandenberg.
html). The image is in ‘public domain’
and close-up pictures can be found at
(www.siloworld.com/) by Gary G. Baker
or “Vandenberg 576 A1” (www.astronau-
tix.com/sites/van576a1.htm).
Figure 11. This photograph of the MBRV-1 launch
was taken by the USAF Vandenberg
AFB and presented in “The Aerospace
Corporation: 1960-1970,” 30 March
1970, “Serving America” tenth anni-
versary book by Crosslink Publications.
Aerospace Corporation, as a government
not-for-profit corporation, and Crosslink
deserve credit for documenting the first
maneuvering reentry vehicle test. Other
pictures of MBRV missiles on the launch
pad can be found at Gary G. Baker’s
Internet site: (www.siloworld.com/MIS-
SILES%20%20LAUNCHES/VAFB/
ABRES/abres.htm). Gary encouraged
use of the pictures.
Figure 14. This dramatic image was also collected
and presented by Gary G. Baker (Silo-
Man) in “ABRES/BMRS: Western
Missile Test Range, Vandenberg AFB”
(www.siloworld.com/MISSILES%20
%LAUNCHES/VAFB/ABRES/abres.
htm). Gary presented all four Boost
Glide Reentry Vehicle (BGRV ) Atlas

300
lightning bolts

F missile tests plus several other early


ABRES tests. Gary was most helpful in
providing high resolution USAF/ Van-
denberg AFB images and encouraging
their presentation.
Figure 15. This illustration of the Aurora hyper-
sonic aircraft was first presented in an
Associated Press, San Diego Union Tri-
bune newspaper on 17 October 1988 and
was published nationwide in the public
domain. It was later described in detail
by Bill Sweetman, “Hypersonic Aurora: a
Secret Drawing?,” Jane’s Defense Weekly,
12 December 1992, page 14.
Figure 18. The picture of an Athena missile launch
was part of a University of Utah Moun-
tain West Digital Library collection
located by Sandra McIntyre. The City
of Green River used the image entitled
“Athena Launch-Complete Success,”
to promote civic advancement (http://
content.lib.utah.edu/cdm4/browse.
php?CISOROOT=/GRMB). Ben
Coomer of Green River provided permis-
sion to copy the image on behalf of the
city.
Figure 19. The photograph “Advanced Inertial Ref-
erence Sphere,” was presented by FAS
(Federation of American Scientists) in an
updated article, 22 October 1997 (http://
nuclearweaponarchive.org/Usa/Weap-
ons/Airs.html). This in-depth article

301
william yengst

stated unmarked images were believed


“not to be under copyright.” The image
was also used by the Charles Stark
Draper Laboratory in presenting AIRS
technology.
Figure 20. This USAF photo was presented in Dick
Hartunian’s paper, “Ballistic Missiles and
Reentry Systems: The Critical Years,”
Crosslink magazine, Winter 2003, (www.
aero.org/publications/crosslink/win-
ter2003/02.html). Crosslink is published
and copyrighted by The Aerospace Cor-
poration, as a government not-for-profit
corporation and deserves credit for doc-
umenting many of the seldom reported
events and advanced technologies.
Figure 21. This dramatic picture of Flight Test 2 of
“DC-X,” was presented in the Internet
(www.astronautix.com/lvs/dcx.htm). It is
a NASA photograph also used by Wiki-
pedia Citizendium in describing reentry
technology, (http://en.citizendium.org/
wiki/Atmospehric_reentry). An Astro-
nautix article on “DC-X” (www.astro-
nautix.com), copyright 1997-2008, indi-
cates Mark Wade is credited with acquir-
ing and making the image available for
reprinting.
Figure 22. This image of General Dynamics Stra-
tegic Boost Glide Vehicle (or Hyper-
sonic Glide Vehicle) was presented by
Andreas Parsch in the “Directory of U.S.

302
lightning bolts

Military Rockets and Missiles: Appendix


4: Undesignated Vehicles.” It was pub-
lished in 2003 with permission of Mark
Wade of Encyclopedia Astronautica
(www.designation-systems.net/dusrm/
app4/hgv.html). Credit for reprinting
the picture is given to © Mark Wade,
Astronautix (www.astronautix.com/) and
the Encyclopedia Astronautica.
Figure 23. This U.S. Navy photograph was presented
by Andres Parsch in a 29 September 2004
article, “Raytheon (General Dynamics)
AGM/BGM/RGM/UGM-109 Toma-
hawk,” by Designation Systems as part
of the Directory of U.S. Military Rockets
and Missiles (www.designation-systems.
net/dusrm/m-109.html). Andres indi-
cated the picture is in public domain with
credit to the U.S. Navy.
Figure 24. This U.S. Army photograph was used
by Redstone Arsenal in presenting the
chronological history of Pershing II
development and test: (www.redstone.
army.mil/history/systems/pershing). The
U.S. Army and “Pershing Photo Gallery”
deserve credit for the dramatic image.
Figure 26. Arun S. Vishwakarma published an
exceptional article entitled “Indian Long
Range Strategic Missiles-Agni III,”
in Indian Defense Review, October-
December 2006. It included this illus-
tration of Agni II Mark 2 RV (Figure 8:

303
william yengst

Reentry vehicle comparison). The illus-


tration is reprinted to describe reentry
vehicle development by Indian Defense
Research and Development Organiza-
tion (DRDO) and with “express written
permission of Arun Vishwakarma and
Bharat-Rakshak (Consortium of Defense
Industries).”
Figure 27. The photograph of “Agni RV-Mk2 on
assembly line,” was from Arun S. Vish-
wakarma’s article “Agni-Strategic Bal-
listic Missile,” at (www.bharat-rakshak.
com/Missiles/Agni.htm). It is provided
with courtesy of DRDO and reprinted
with “express written permission of Arun
Vishwakarma and Bharat Rakshak.” The
author is indebted to Arun for assisting
with obtaining permission to use these
photos.
Figure 28. The article “DongFeng11 (CSS-7)
Short-Range Ballistic Missile,” was pub-
lished by Sino Defense (based in United
Kingdom) as Figure 3 on the Chinese
Internet, 18 February 2009 (www.sino-
defense.com/strategic/missile/df11.asp).
The image was released to ‘public domain’
as shown in Chinese on the World Wide
Web at “DF-11/A (M-11) Short-Range
Ballistic Missile,” (www.armsky.com/
yuanchuangzhuangao/seechina/Chi-
na2ndArtillery200601/3261.html).
Figure 30. This illustration was produced by the

304
lightning bolts

Russian Ministry of Defense for public


release and first presented in Izvestiya
(newspaper) on 17 February 2004 by
Dmitri Litovkin, “Putin Rukovodit Stra-
tegicheskimi Ucheniyami.” Several days
later, Nikolai Sokov used the illustra-
tion in his article, “Military Exercises in
Russia: Naval Deterrence Failures Com-
pensated by Strategic Rocket Forces,”
24 February 2004 (http://cns.miis.edu/
stories/040224.htm) for the James Mar-
tin Center for Nonproliferation Studies
(CNS). Kenley Butler of CNS researched
this background and indicated the trajec-
tory map was in the ‘public domain.’ An
Internet video on the Topol-M missile
trajectory was shown on Moscow’s tele-
vision in mid-February 2004.

All other figures, pictures, illustrations, and


tables used in the book were drawn and presented
based on the author’s memory or analysis of events
including material presented in the text.

305

You might also like