CAPT. BRYCE D. INMAN, USN (RET.

FROM TYPHON TO AEGIS -

THE ISSUES A N D THEIR RESOLUTION
THE AUTHOR AAW weapon and combat systems and the USS Ticon-
deroga and Arleigh Burke classes. During the concept
retired from the U.S. Navy in 1970 and from RCA in 1986. He design and early development phases of the
has a BChE from Ohio State University as well as BS and MS ASMS/Aegis system there arose a number of key tech-
degrees in engineering electronicsfrom the Naval Postgraduate nical and programmatic issues which had to be resolved
School. An engineering duty officer, he served in various elec-
tronics and RDT&E billets in the Navy and Office of the Secre-
before the program could continue.
tary of Defence including command of the U.S. Navy Under- This paper describes the decision process, identifies
water Sound Laboratory and Naval Underwater Systems the issues, and for each, reviews the options, describes
Center. While the Bureau of Ships Radar Branch head and the resolution and speculates on the consequences of a
Surface Missile System Project search radar manager, he was different decision. The objectives of the paper are: one,
assigned to the Advanced Surface M k i l e System (ASMS) As- to provide a brief historical account of the concept
sessment Group and later as Navy technical director for the formulation and early engineering development phases
ASMWSAM-D Commonality Study. A t RCA he led Aegis of the ASMS/Aegis system; two, to provide insight into
simplification studies and concept designs f o r the the key system engineering issues with a system such as
AN/SPY-IA, B, C & D radars, the CG-47 and DDG-SI Aegis; and, three, to describe system concept review
combat systems and the DG ship. He organized and led the procedures that were valuable for ASMS and should be
Aegis Ship Integration Group through 0 3 - 4 7 contract design.
A t retirement, he was director of submarine combat systems. useful for other programs.
He was awarded the Navy Commendation Medal (2 A wards), For security reasons, some of the system engineering
the Navy Meritorious Service Medal and the 1982 ASNE Gold discussions are qualitative or use typical, but hypo-
Medal. He is a member of ASNE, the Naval Institute and a thetical, values for illustration. The speculation on the
senior member IEEE. consequences of a different decision represents the
author’s judgment (biases) and is included to illustrate
ABSTRACT the possible second and third order consequences of a
key system decision.
The Aegis weapon system, now deployed in the USS
Ticonderoga cruiser class and soon to be deployed in the
Arkigh Burke destroyer class, is generally accepted to
be, by far, the most capable ship AAW weapon system TYPHON AAW WEAPON SYSTEM
in the world. The concept for this system was developed
in the mid-l%0s as replacement for the failed Typhon SYSTEM
ELEMENTS
weapon sysem. During the concept design and early de-
velopment phases there arose a number of key technical Typhon was to have been an all new system with a
and programmatic issues. In retrospect, a different new two-member missile family, a new control system,
decision on many of these issues could have resulted in a and a new agile beam multifunction (search, fire control
much less capable system and, in some cases, to another track, target illumination and missile communication)
system development failure. This paper first provides a radar. The missile family consisted of a long-range
brief description of the Typhon system concept and why ramjet powered TalodTerrier replacement and a
it failed in execution. This is followed by the identifica- smaller, shorter range solid fuel powered Tartar/Terrier
tion of key early technical and programmatic issues. replacement. Missile guidance, however, was the funda-
Each issue is discussed, the options defined, and the mental difference between Typhon and its predecessor
final resolution described. With the benefit of hind- systems. The new guidance required a radically different
sight, the impact on the system development from a dif- control radar and this radar, designated AN/SPG-59,
ferent decision is suggested. Finally, an attempt is made was the focus of the development effort.
to reach some general conclusions that could be of value In contrast to the earlier surface missile systems where
to other similar development efforts. missiles were developed separately from their shipboard
support and combat systems, Typhon was intended as an
INTRODUCTION integrated AAW missile weapon system. The Typhon
system concept was defined by The Johns Hopkins Uni-
I n the late 1950s it was becoming apparent that a versity Applied Physics Laboratory (APLIJHU) and
successor system to the “Three Ts” (Talos, Tartar and later technical direction of the development program was
Terrier) for ship AAW was required. Typhon was to be assigned to APL/JHU [l].
the answer. The Typhon system development subse- The full system with both missiles and a large version
quently failed causing the abandonment of plans for of the radar was planned for installation in a CG, CGN
new AAW capable cruiser and destroyer classes. This or DLGN class ship, and a derivative system with only
led to the initiation of the Advanced Surface Missile the medium-range missile and a smaller radar was envi-
System (ASMS) Program, and ultimately, to the Aegis sioned for a DDG class.
62 Naval Engineers Journal, May 1988
INMAN TYPHON TO AEGIS

MISSILEGUIDANCE
SYSTEM Table 1. Comparison of Required Maximum Illuminator
Power for HAW and C3M/TH Missile Guidance
The missile guidance system was to have been close
control command midcourse with semi-active terminal HAW C’M/TH
homing (C3M/TH). Ship to Target Range at Launch, NM 45 45
At the time Typhon was initiated, the Tartar system Ship to Target Range at Intercept, NM 30 30
and most Terrier missiles used semi-active homing Ship to Target Range at Start of Homing, NM 45 31.5
guidance for the entire missile flight. This is known as Missile t o Target Range at Start of Homing,
NM 45 4.5
home-all-the-way (HAW) guidance. With this system
-2
the fire control radar acquires the designated target using Range Product, NM 202.5 141.8
azimuth and elevation data supplied by the ship’s search Relative Maximum Required Illuminator
radar and a firm track is established. The launcher then Power 204 1
points the missile at the target and, after the seeker in the
nose of the missile acquires the target using illumination Assumptions:
Target average velocity - 1000 ft/sec. (M 0.91)
energy reflected from the target, the missile is launched. Missile average velocity - 2000 ft/sec. (M 1.82)
The illumination function is physically “slaved” to the C,M/TH Homing time - 9 sec.
fire control radar since it uses the same dish antenna.
Since the fire control radar operates in the C-Band of fre-
quencies and the illuminator operates in the X-Band of
frequencies, the sharing of a dish antenna by a pulsed shows the relative maximum illumination power re-
radar and a high power continuous wave (CW) il- quired for a simplified example of a HAW intercept.
luminator was not a difficult isolation problem. (Note: In Range product is defined as range from illuminator to
this paper the old radar frequency band designations S - target times range from target to missile. Illumination
about 3.1 KMc, C - about.5.6 KMc and X - about 10.0 power required varies as the square of range product for
KMc are used rather than the new electronic warfare a constant target radar cross section (reflecting area).
band designations F, G and J, respectively). The example depicted in Table 1 is simplified by assum-
HAW guidance causes two system performance limi- ing that (1) the target is flying directly toward the ship,
tations. First, a fire control director is tied to a single (2) the average missile velocity is average missile speed
target from well before missile launch to kill assessment projected on a straight line between ship and intercept
after intercept. This limits ship fire power to one salvo point, (3) HAW and C3M/TH missile average velocity is
(one or two missiles) per useable director during this the same and (4) there is no difference between mono-
period. Second, the target range at time of missile static and bistatic target radar cross section at the illumi-
launch is limited by the reflected illumination signal nation frequency. Further the assumed hypothetical
strength at the output of the small, low gain, seeker homing time includes seeker acquisition time, kill as-
antenna in the nose of the missile. For other than large, sessment time and illuminator director reposition time.
relatively slow targets, this constraint often reduces the The Typhon C3M/TH guidance system promised
maximum intercept range to values less than can be sup- large increases in both firepower and maximum inter-
ported by the missile propulsion system. Figure l a cept range compared to HAW guidance. Also, in some
shows the geometry of a HAW intercept and Table 1 cases, improved intercept accuracy was indicated. With
close control command midcourse guidance, the missile
is tracked like another target, an optimum intercept tra-
jectory is continuously computed and guidance com-
mands are sent to the missile. At the appropriate time,
the target is illuminated and the missile seeker is com-
manded to acquire the target. Firepower is increased
because missiles can be launched against all tracked
targets as soon as they appear to be engageable with a
launching rate limited only by the agile beam radar il-
lumination channels and launcher launch rate.
MAX R A N G € PRODUCT= SF*HaG Maximum intercept range is increased for two rea-
, sons. First, since the ballistic-like trajectory is more
(6) C3M/TU
energy efficient than the HAW pursuit trajectory, max-
7; - T A R G E T AT MiSSlLE LAUN C U
X - T A R C E T AT START O F C3M/TH imum intercept range is increased and/or maneuverabil-
HQMING ity at intercept is greater for the same missile propulsion
T3- T A R G E T AT INTCRCEFT performance. Greater maneuverability at intercept can
9-SHIP ,
LOCATION SsH,
produce reduced miss distances. Second, illumination
ti,- M I S S I L E Ar START OF m w HOMING
4-MlSS1A.E A 7 START OF C 3 M / T H power is generally no longer a maximum range con-
HOMING straint. With semi-active homing for just the terminal
phase of the intercept, the range product at seeker ac-
Figure 1. Comparison of HAW and C3M/TH intercept quisition is much less. See Figure l b and Table 1. Since
geometry: required illumination power varies directly with the
Naval Engineers Journal, May 1988 63
TYPHON T O AEGIS INMAN

square of range product at the time of seeker acquisi- system would have had the same configuration but with
tion, the illumination power required for a Typhon type 10,800 elements at the Luneburg lens, 2,700 amplifiers
system can be a small fraction of that required for a and 10,200 horns in the array.
HAW system against the same target at the same in- There were two perceived advantages of this con-
tercept range. figuration over the conventional planar phased array.
First, a phase shifter was not required between the
AN/SPG-59 RADAR transmitter and the radiating element. This eliminated
phase shifter losses and, of even greater significance,
The Typhon Guidance System required a large num- eliminated the power handling limitations of a C-Band
ber of simultaneous fire control quality tracks and the phase shifter. Second, this approach inherently provides
use of one director per track would not be feasible for a the very large signal bandwidth of a time delay array,
shipboard system. Only an agile beam fire control radar and large signal bandwidth was required at C-Band to
can support such a system. The AN/SPG-59 Typhon counter weather clutter and ECM.
radar was to have been a very high power, broad band, Maintenance of equal time delay required that in-
C-Band radar using a spherical horn array transmitting dividual antenna element transmission electrical lengths
antenna, three Luneburg lens receiving antennas, and a had to be maintained to 5 mm or less through an
large suite of waveforms. A 10,OOO horn array antenna amplifier, two transmission line switches, a power
was planned for the cruiser and a 3,400 horn array divider, and a long length of C-Band waveguide with a
antenna was to be used in destroyer type ships. The number of twists and bends of many different con-
cruiser transmitter also would be three times as large figurations. This was not a trivial design task or an inex-
and powerful as the destroyer transmitter. At the time pensive system to fabricate. Further, the transmit-
of Typhon concept formulation, it was believed that the ter/antenna system per-element weight was at least two
then available planar phased array technology would hundred times that which can now be obtained from a
not support the Typhon radar requirements. These re- conventional phased array and transmitter. As the
quirements included operation in the C-Band of fre- development progressed, it became evident that even the
quencies, high power and very large signal bandwidth. destroyer-size system would be too large, heavy and
An innovative agile beam radar approach invented by costly to be practical for a DLG ship. Further there was
Dr. John Garrison of APL/JHU was selected [2]. This a growing concern about the system’s expected perfor-
approach used a relatively low power source to excite a mance, particularly in an ECM environment. These
large dielectric sphere in a spherical cavity formed by in- concerns and the desire by some in the Navy Depart-
ward looking horns. The dielectric sphere approximated ment to divert Typhon development funds to fix 3T
a Luneburg lens and served as both a beamformer and problems in the fleet led to program termination in
an analog phase taper “computer.” A Luneburg lens November 1963. The Secretary of Defence decision
has the property of converting a point source of energy guidance paper that terminated Typhon also established
on one side of the sphere into a plane wave across the funding for ASMS and provided for completion of at-
projected area of the sphere on the other side [3]. The sea testing of the prototype AN/SPG-59 radar being in-
receiving elements from the beamforming sphere were stalled in Norton Sound [I]. Testing began in June 196-4.
then switched through a bank of amplifiers and back System reliability was poor and computer programming
through switches t o the corresponding elements on the was incomplete. The biggest problem, however, was
spherical transmitting array. Figure 2 shows a schematic transmission path and signal processor losses which
of the proposed destroyer system which was also the were much higher than originally budgeted leading to
USS Norton Sound experimental system. The cruiser disappointing and completely unacceptable detection

-3600 HORNS 3qQO HQRN ARRAY

BEA/V FOR M E A T W S MI TTLR M € ” A

Figure 2. AN/SPG-59 antenna/transmitter concept.

64 Naval Engineers Journal, May 1988
INMAN TYPHON TO AEGIS

at the expense of ship system complexity and still more

demands on radar time and power for the com-
muncations link.

ASMS ASSESSMENT GROUP

ORGANIZATION

With the establishment of the ASMS program, the

Surface Missile Systems Project Office, PM3, requested
funded contract definition proposals from industry. In-
dustry, in general responded with proposed systems that
were much like Typhon but with a conventional planar
phased array antenna. At this point PM3, RAdm. E.T.
Reich, USN, decided to obtain wider Navy input into
Figure 3. USS Norton Sound (AVM-1) with experimental the ASMS program. He requested RAdm. F.S. With-
Typhon System. ington, USN, (Ret.), a former chief of the Bureau of
Ordnance, to lead an ASMS Assessment Group. This
range performance. In some tests with a small jet air- group, popularly known as the “Withington Commit-
craft, the target was never seen, as the minimum tee” was staffed with people drawn from OpNav, PM3,
operating range for the waveform being used exceeded BuWeps, BuShips, several of the Navy laboratories,
the maximum detection range on that target. Testing APL/JHU, the competing industrial contractors, the
continued into the spring of 1966. In June 1966 removal Bell Telephone Laboratories and Army air defense agen-
of the AN/SPG-59 from Norton Sound was ordered. cies. The group was charged with the task of establishing
Figure 3 shows USS Norton Sound (AVM-1) with the a baseline ASMS system configuration and making an as-
experimental Typhon system installed. The sphere on sessment of its effectiveness against the postulated threat.
the top of the tower is the transmission array and the Subgroups were formed to consider operational require-
small sphere at the base of the cone is one of the receiv- ments, assess effectiveness, and do top level concept
ing antennas. The tower is filled with transmitter equip- design for the ship system (less launcher), the launching
ment and waveguide. system and the missile. Ship installation, costs, reliability
and logistics support were also treated. The group was
TVM HOMING assembled in an “off campus” site in Silver Spring, Md.
in January 1965 with some members working until mid
Target homing illumination by the tracking radar was May.
another innovative feature of the Typhon system. The The marketing departments of some of the industrial
use of the radar for illumination has the advantage of contractors tried very hard to place marketing personnel
no separate directors to locate in the ship’s superstruc- in their slots in the group. PM3 prevented this by re-
ture and a potential for a large number of illumination questing system engineers, by name, from each com-
channels when required. There are three disadvantages pany. If that individual or a suitable alternate was not
to the use of the radar for target illumination. First, the available, that company would lose its slot. The desired
radar frequency is a compromise, higher than that talent was supplied.
desired for a powerful long range radar and lower than The ship system subgroup was further broken into
desirable for illumination. The seeker antenna, con- teams for radar, fire control, and command and control.
strained by missile diameter, has less gain and larger Critical sub-elements and components such as com-
angular resolution than that provided by an illumina- puters, displays, transmitter tubes and antenna phase
tion system at a more optimum frequency. In the case of shifters were treated by special assignments.
Typhon, compared to Tartar/Terrier, the relative
angular resolution was about 1.6 and the relative seeker OPERATION
antenna gain was about 0.4. Second, radar homing il-
lumination consumes a significant portion of the radars One problem was how to use the industry represen-
power and time. Power, and particularly time, are very tatives without them becoming nonproductive defenders
often major constraints on a multifunction radar’s of their parent companies proposal. One was assigned
search and tracking capability. A third problem with to the performance assessment organization and the rest
radar homing illumination is that the waveform must be were provided a separate room and assigned engineering
pulsed rather than CW, and a pulsed waveform is more trade studies with a requirement for a majority report
difficult t o process in the missile. In the case of Typhon, and minority reports, if any. These reports were useful to
the then current technology would not support pro- the whole Assessment Group during deliberations and
cessing in the missile so the seeker signals were relayed for preparing the Assessment Group report. Interaction
back to the ship for processing and the results were sent across the whole Assessment Group was required and ad-
back to the missile. This was called target-via-missile vanced by the distribution of weekly activity reports and
(TVM) homing. TVM solved the processing problem but frequent meetings.
Naval Engineers Journal, May 1988 65
 TYPHON TO AEGIS INMAN

At that time, the usual approach to operations and in some cases, even between technical and program-
analysis was to seek a maximum performance against matic issues. From previous Typhon program work,
the whole threat spectrum with more or less equal there was no question but what ASMS would use
weighting of the threats. The Withington Committee C3M/TH guidance. Further, the terminal homing would
performance assessment subgroup under John Long of be semiactive as the technology could not support active
APL/JHU favored the use of the criterion of minimum homing, and passive homing was considered to be un-
probability of failure against any reasonable threat in suitable as a primary homing mode. Because of the
the threat spectrum. The argument was that an intelli- Typhon experience, only planar phased arrays were con-
gent enemy would exploit a system weakness so that an sidered for the agile beam radar.
equal weighting of all threats was not a good measure of The following technical issues are discussed in the
effectiveness. Long’s analysis frequently gave measures order in which they had to be resolved and in order of
of effectiveness that were quite different from the equal priority. This is followed with the same approach for
weighting method and were very useful in making programmatic issues.
system decisions.
To focus the ship system subgroup’s assessment of HOMINGILLUMINATION
system configurations, four top-level system perfor- Two issues influenced the selection of a homing illum-
mance charactistics - “cornerstones” - were listed in ination concept. One was the choice between pulsed and
order of priority and concepts tested against them. The CW illumination and the other was the choice between
cornerstones were: illumination from the radar and the use of separate dish
Reaction Time illuminators slaved to the multi-function radar. Pulsed
Fire Power
Immunity to Weather Clutter and ECM illumination, would permit, but not require, illumina-
Availability tion from the radar while CW illumination would rule
The emerging low altitude cruise missile threat was out the use of the radar. Seeker signal processing and
the reason for the priority interest in reducing reaction overall system complexity, particularly if the radar was
used for illumination, was the main argument against
time. The 1958 Cornell Aeronautical Laboratory “Proj-
ect Flame” Report had emphasized antiship missiles as pulsed illumination. The use of the radar for illumina-
major threat elements in the 1960-1970 era [l]. The con- tion would avoid the problem of locating illuminators in
the superstructure and avoid the “illuminator slaving
cern about a coordinated saturation attack by ASMs or
SSMs was the reason for the firepower concern. ECM is risk.” While it should be obvious that there would be no
real risk in open loop slaving of an illuminator with a
always a concern, but in that time period, weather clutter
beam width many times greater than the tracking ac-
was the most prevalent “jamming” source. Even though
curacy of the radar, this was a concern of many in the
listed in fourth place, the goal was essentially continuous
availability and, in fact, low availability in the SMS Navy era of the ASMS Assessment Group and it continued to
be a concern well into the Aegis development program.
had been a major reason for the establishment of PM3.
However, the strongest argument advanced for pulsed
The Aegis program manager, RAdm. W.E. Meyer,
illumination was the concern that a CW seeker, lacking
USN, much later and very appropriately, added missile
range as the fifth “cornerstone” for the Aegis develop- range resolution, could be made to intercept the mid-
point between two targets that were more than two
ment program.
System and subsystem selection was treated as a deci- times the warhead lethal radius apart. This effect can,
with certain simplifying assumptions, be demonstrated
sion tree with all reasonable branches explored. In
analytically and was demonstrated with a simple static
general, decision was sought by group consensus or at
simulation using an actual Tartar seeker. A more so-
least by majority opinion obtained in conference.
phisticated review of this problem indicated that the
An OpNav committee was established to review the
probability of two real-world targets achieving, and
progress of the ASMS Assessment Group. This commit-
maintaining during the intercept, the exact spatial and
tee was chaired by the vice chief of naval operations and
dynamic configuration required for this effect to be a
included the deputy chiefs of naval operations and some
problem was vanishingly small. Incidentally, it is be-
of their staff. The committee received several in-depth
lieved that this effect has never been observed in the
briefings during the life of the Assessment Group. Dr.
field. In the interests of development risk reduction and
Eugene Fubini, DDR&E, had been a strong advocate of
system simplicity, CW illumination was selected. In
Typhon termination and he also requested several in-
closing out the radar illumination issue, the selection of
depth briefings. These briefings were a source of
CW illumination had a major impact on radar frequen-
guidance and provided senior decision makers with an
cy selection, the USN/USA commonality issue and the
understanding of the system and Assessment Group
question of evolutionary missile development.
issues. This understanding led to valuable support for
A pulsed illumination selection would surely have
the new ASMS program.
added considerable cost to the development program as
KEY ASMS ISSUES well as cost and complexity to the production system. It
is also conceivable that this single decision could have
INTRODUCTION influenced a chain of decisions that would have led to an
As can be expected in any large system design, there “Aegis” system with much less capability and, perhaps
was considerable interaction between technical issues, to another aborted development effort.
66 Naval Engineers Journal, May 1988
INMAN TYPHON TO AEGIS

FREQUENCY
RADAR OPERATING trol radars which devote the entire time of the radar to
one or two targets. However, the use of pulse doppler
The choice of radar operating frequency was the sin- MTI for the ASMS search function where a large num-
gle most important ASMS issue that had to be resolved. ber of targets had to be handled was not considered to
With the selection of CW illumination, this function be practical. In that era, canceller MTI for microwave
was removed as a potential radar requirement and the radars at sea had not enjoyed much success. Accord-
frequency selection could focus on search and fire con- ingly, it was a goal that the ASMS baseline radar must
trol capabilities and ship installation considerations. have useful performance without any MTI. It was rec-
X-Band and above could not provide the required ognized, however, that a MTI function might be added
range. Anything below S-Band would require antennas if the additional performance seemed worth the addi-
that were two large for shipboard use. This left the tional cost.
choice between C-Band and S-Band. C-Band promised Frequency selection discussions were long and
the following advantages: spirited. The differences of opinion derived from the
differences in point of view of the partisans on each
Better detection performanmce against low elevation side. Personnel from BuShips and its associated labora-
targets tories, in general, had a search radar point of view
- Lighter antenna which stressed long detection ranges and weather clutter
- Propagation factor
immunity. BuWeps associated personnel, in general,
ECM Immunity
had a fire control radar point of view with concerns
- Greater frequency agility bandwidth
- Greater signal bandwidth
about low elevation target detection and tracking. All
Ship Installation Ease ASMS Assessment Group members were aware of the
- Smaller, lighter antennas growing ECM threat. Also, all members were sensitive
Technology Base to ship installation concerns due to the recent Typhon
- Operational shipboard fire control radars cancellation and experience with the large AN/SPS-
- Typhon 32/33 systems in USS Long Beach and USS Enterprise.
Originally, an uneasy consensus in favor of C-Band was
S-Band promised the following advantages: reached. RAdm. Withington approved a switch to
S-Band during the final weeks of the Assessment Group
Better detection performance against long range high deliberations. The final selection of S-Band was based on
altitude targets the realization that, for a multifunction (search and fire
- Larger power aperture product
- Less internal losses and less atmospheric attenua-
control tracking) radar, no other radar capability has
tion meaning until the target is detected.
- Better receiver noise figure From what we now know, if C-Band had been se-
Better small target detection and tracking lected and essentially the same performance as specified
- Less weather clutter for Aegis was demanded of the system, we could have
- Better MTI performance had another failed development effort. If C-Band had
ECM Immunity been selected and performance demands scaled back, we
- Higher power could now have a capable AAW missile system suitable
- Lower antenna sidelobes
- Better MTI performance
for local area defense but with development and pro-
Technology Base duction costs no less than those for Aegis.
- Operational shipboard height finding radars in- The following is a discussion of the technical aspects
cluding the AN/SPS-33 phased array radar of the frequency selection decision.
- BuShips sponsored planar phased array advanced
development work. Antenna Size and Weight

It was believed that the antenna would dominate the For a 1.5 degree broadside beamwidth, a C-Band ar-
radar cost and that there would be no fundamental cost ray is about 8 feet in diameter and an S-Band array is
difference between equal power C-Band and S-Band about 14 feet in diameter. Weight however will not scale
antenna systems. For equal detection range performance, directly as the cube of the operating frequency due to
if in fact that were possible at all, the S-Band system the constant weight of the phase shifter control and
would have a cost advantage because of the cost of pro- cooling functions. The S-Band array was expected to
ducing and handling the additional transmitter power re- weigh between two and three times as much as the
quired by a C-Band system. C-Band array. The smaller C-Band antenna would per-
The radar was required to have an air intercept con- mit either a smaller ship or a higher antenna location in
trol (AIC) capability as it was desired to reduce the size a ship of the same size.
of the guard zone between SAM and AIC operations. A
detection range capability roughly the same as the Propagation Factor
AN/SPS-48 AIC radar was established as a require-
ment. For a shipboard system, C-Band has one significant
In the early 1960s, pulse doppler moving target indi- operational advantage against small, fast, low flying
cation (MTI) had been successfully used with fire con- targets. The C-Band antenna beam can, in general, be
Naval Engineers Journal, May 1988 67
TYPHON T O AEGIS INMAN

7a/ \

Figure 4. Radar antenna pattern break-up due to constructive

and destructive combination of direct energy with reflected
Figure 5. Example of propagation factor curves for S-Band
energy from the sea.
and C-Band radars with otherwise identical characteristics.
brought down closer to the horizon. An antenna over
the reflecting sea will have a null at the horizon even if is a valuable ECCM capability. With a conventional
the antenna beam is pointed directly at the horizon. phased array, the instantaneous signal bandwidth in
Over the spherical earth, the angular width of the null is percent is approximately equal to the broadside beam-
approximately inversely proportional to the frequency width in degrees. Any larger bandwidth input signal will
and antenna height [4]. The C-Band system can have a be spread over a larger beamwidth in space [ 5 ] . This
narrower null both because of the frequency and be- limits the signal bandwidth for a 1.5 degree beamwidth
cause, in general, the smaller, lighter antenna can be antenna to about 85 MHz at C-Band and 45 MHz at
higher in the superstructure. In Figure 4, which illus- S-Band. Signal bandwidth gives range resolution which
trates this effect, the vertical angle has been expanded reduces the effects of volumetric clutter such as rain and
for clarity. With a state 0 sea, the actual lobe peaks will chaff. The advantage is not enough to counteract the in-
be almost twice the field strength at the same range com- creased rain clutter at C-Band but can give the C-Band
pared to the free space pattern. With some sea surface radar a significant advantage against chaff. Exploita-
roughness, the actual peaks will be less and the nulls not tion of signal bandwidth, however, is expensive.
complete. The radar detection range equation includes a
“propagation factor” t o account for this low elevation Power and Losses
phenomenon [4]. Figure 5 is a plot of S-Band and S-Band, in general, permits operation at higher power
C-Band propagation factors for a representative situa- with less system losses and greater receiving system sen-
tion. Also shown in Figure 5 , is a typical sensitivity time sitivity compared to C-Band. S-Band can also be ex-
control (STC) detection limit established t o control pected to provide better antenna side lobe ratios. The
radar clutter from nearby waves, water droplets, in- receiving system sensitivity advantage has no value in a
sects, etc. The intersection of the detection limit line and jamming environment, but the higher power and re-
the propagation factor curve represents the maximum duced transmission losses along with reduced at-
detection range on a low elevation target of a given size. mospheric attenuation make S-Band more suitable for a
In the Figure 5 example, the detection range difference long range radar. To a first approximation, the difficul-
is about 1.4 nm which would give an 8.5 second advan- ty of microwave power generation and handling varies
tage to the C-Band system for a M 0.9 target. With a as the square of the frequency. This gives the S-Band
higher C-Band antenna, the advantage would be system a 4.9 d b advantage.
greater. With a MTI capability, the STC threshold can In a non ECM environment, receiving system sensitiv-
be lowered and there is less relative advantage for the ity is directly proportional to antenna area. This gives
C-Band system. Figure 5 uses the 413 earth radius con-
vention to account for atmospheric refraction as was
standard in 1965. It is now customary t o use a 3/2 earth Table 2. S-Band Radar Detection Advantages
radius convention for propagation at sea to account for Over a Comparable C-Band Radar
the surface evaporation duct that is almost always pres- Environment
ent at sea. The use of a 3/2 earth radius extends both the
S-Band and C-Band predicted detection ranges. Element Non ECM ECM

Transmitter power 4.9 db 4.9 db

Frequency Agility and Signal Bandwidths Antenna Area 4.9 -0-
Internal losses 3.0 1.5
Since a tuning bandwidth of 10% was postulated, the Atmospherics losses (200nm) 1.o 0.5
C-Band system would be able to operate over a band of Receiver Noise figure 1 .O -0-
Antenna slidelobe ratio -0- 2.0
about 560 MHz while the S-Band system would be
limited t o about 310 MHz. Frequency agility bandwidth System Total 14.8 d b 8.9 db

68 Naval Engineers Journal, May 1988

INMAN TYPHON TO AEGIS

the S-Band system another 4.9 db advantage. Antenna RADARTRANSMITTER

TUBE
sidelobes are, in part, caused by physical errors as
measured in wave lengths. This gives the S-Band system Tube Type
another advantage in an ECM environment. Table 2
gives representative S-Band advantages in clear and off- Historically, the success or failure of a radar develop-
target ECM environments. The system totals represent S- ment was often directly related to transmitter tube per-
Band range advantages of 2.3 to 1.0 and 1.7 to 1.0 for formance. At best, the power tube could be relied upon
non ECM and ECM environments, respectively. to furnish several unpleasant surprises during the
radar’s development and production. The policy of the
Clutter Immunity ASMS Assessment Group was to pick, for the baseline
design, the most promising tube type that was mature
The other important S-Band advantage is the relative enough to minimize development risk, then provide a
immunity t o weather clutter compared to C-Band. backup. For ASMS, tube selection included two issues,
Weather clutter is caused by water droplets in the at- tube size and tube type. Candidate S-Band tube types in-
mosphere. The radar cross section of a given volume of cluded traveling-wave-tubes (TWT) as well as crossed-
droplets varies as the fourth power of the frequency and field-amplifiers (CFA), both linear and reentrant,
is a function of precipitation rate. The ratio for rain backward wave and forward wave. Since the modulator
radar cross section at C-Band compared with S-Band is is a major element in any radar transmitter which must
therefore about 10 t o 1. Radar signal bandwidth can be amplify a suite of waveforms, an RF initiated tube that
used to reduce the ratio of clutter signal t o target signal does not require a modulator was a new attractive op-
but even when this is considered, the advantage of tion. One RF initiated tube in a very early stage of
S-Band over C-Band is still about 6 to 1. Figure 6 shows development appeared to be the ideal choice but it was
the apparent radar cross section of weather clutter in a rejected because of development immaturity. This was a
1.5 degree by 1.5 degree beam from S and C-Band fortunate decision as this tube was never able to
radars both with signal bandwidths of 16 MHz from demonstrate many of its earlier performance claims. A
two 8 MHz coherent channels. Typical nose aspect more conservative design RF initiated reentrant CFA re-
radar cross sections of a small jet fighter and an ASM quiring only a low-power extinguishing modulator was
are shown for comparison. selected with its cathode modulated version as a backup.
At the time of the ASMS Assessment Group, it was Early in the engineering development, the backup was
believed that there was no difference between S-Band required.
and C-Band for sea clutter. More recent work has sug-
gested a slight advantage for S-Band even against sea Tube Size
clutter.
Continuous availability requires that the transmitter
contain more than one output tube. A small number of
large tubes is usually better than a large number of small
tubes for minimum production cost and weight. The op-
timum is somewhere between 2 and 8 tubes. (For ease of
combining and switching, candidate configurations
usually include a tube number that is an integer power
of two). A configuration with 32 tubes was selected
because it facilitated a phased array approach that had
growth potential to a broad band true time delay anten-
na. When this antenna approach was abandoned during
engineering development, the transmitter development
was too far along to change. The resulting AN/SPY-
lA/IB/ID transmitters have been a most satisfactory
system element with a good availability record and a
great deal of installation flexibility, but are probably
somewhat heavier and more costly to produce than one
with a more optimum tube size.

MISSILELAUNCHER

In the ASMS Assessment Group era it was widely

believed that tube missile launchers offered many ad-
vantages over the pointing launchers then being used.
0.1 ’ 2
/ . . . . . A , ,.
5 10 ..
1

_.
I . so. 1 . . A
,“O Advantages included reaction time, fire power, avail-
RANGE - N M
ability, mechanical simplicity, and cost. It therefore
Figure 6. Comparison of apparent radar cross sections of rain came as a surprise when the Launcher Subgroup con-
at S-Band and C-Band. cluded that the baseline system should use a dual arm
Naval Engineers Journal, May 1988 69
TYPHON TO AEGIS INMAN

pointing launcher. The reason was the belief that resup-

ply and strike-down at sea would require a heavy, cost-
ly, complex and awkward strike-down mechanism
which would negate many of the claimed advantages for
tube launchers. This decision had some small but accep-
table system performance reduction impact but
facilitated ASMS commonality with the Tartar missile.
The later development and deployment of a multi-
missile type vertical launching system (VLS) has provid-
ed some of the earlier claimed advantages. Perhaps of
greater importance, the current vertical launcher has
greatly facilitated AAW/ASW/strike warfare missile
load-out flexibility. In retrospect, the original decision
for pointing launchers was a good one at that time.

TOPSIDE
PHYSICAL
CONFIGURATION Figure 7. USS Norton Sound (AVM-I) with experimental
Aegis system.
Topside physical configuration considerations pro-
duced issues with three system elements, the phased ar-
ray antenna, the illuminators and the deckhouse. added for a total of six. As discussed later, this was not
Antenna requirements included adequate tracking ac- a valid approach to system sizing when the illuminators
curacy to support missile guidance as well as search were large, heavy and expensive.
coverage from horizon to zenith at all azimuths from a Some time before the Aegis engineering development
ship with a 15 degree roll and a 5 degree pitch. proposals were solicited, the modular deckhouse ele-
Because of uncertainties as to exactly what tracking ment, called Deckhouse Mark 20, was added to the
accuracy was required and what could be obtained from system. The plan was to assemble the whole Aegis AAW
a phased array system, a conservative position was system except the launcher and CIC displays into a pair
adopted and the largest antenna within search time con- of two level deckhouses. At that time a group in the Of-
straints was selected. To d o this the smallest beamwidth fice of the Secretary of Defense (OSD) was pushing
that would permit hemispheric coverage to a given range modular warship construction and it is believed that the
and in a given time was selected. Then the largest ele- Deckhouse Mark 20 was added to the system partially in
ment spacing that would provide the required antenna response to OSD desires. This method of installing a
scan angles was selected. Trade studies suggested that a major shipboard system had some attractions. In theory
five antenna system might be better than a four antenna the deckhouse system could be assembled and tested as a
system but the advantage was slight and it was feared “turnkey” operation then landed on the ship at pre-
that five-fold symmetry would cause major installation pared locations and power, external signal cables and
and coordinate conversion problems. Another trade other ship services connected. The disadvantages of this
study suggested that there could be an advantage to us- approach, however, far outweighed any advantage.
ing ship roll stabilizers with a resulting reduction in Figure 7 shows a deckhouse installed on Norton Sound.
antenna elements, but again the indicated advantage The loaded deckhouses for the CGN-38 Aegis Weap-
was small and roll stabilizaiton was not proposed as a on System would have weighed over 200 long tons each.
ship requirement. Work with the AN/SPY-1 system has This weight was made up of four elements:
confirmed that the antenna size is on the upper end of a
band of appropriate antenna sizes for a system like The antennas which must be high in the superstruc-
Aegis. Figure 7 shows USS Norton Sound with the ex- ture.
perimental EDM-1 Aegis System. Compare this S-Band The RF part of the transmitters which should be near
planar array with the C-Band and spherical array in the antennas but which, at some slight cost in system
performance, can be lowered by one or two decks.
Figure 3. The AN/SPY-1 requires 4 arrays each with
Heavy system components that did not need to be in-
over 4,000 active elements for complete hemispheric stalled anywhere near the antennas.
coverage. The AN/SPG-59 used one array of 3,400 A stand alone deckhouse structure having a structural
elements for hemispheric coverage but only 900 were weight considerably greater than the weight of struc-
used at any one time. ture in a functionally equivalent conventional ship
Because of the reduced illumination power require- design.
ment for a C3M/TH system, the selected illuminators
were low power with a small dish size. It was expected The problem was exacerbated by the forward deckhouse
that this would provide a very light, low cost unit that which had to be over the bridge and higher than it
would be relatively easy t o locate in the superstructure. needed to be while adding structural problems to the
Studies indicated that, in general, two illuminators bridge. In the case of the DD-963 hull, which was used
available and bearing could support missile guidance. early in the engineering development program to focus
To provide availability, two more were added and to simplification studies, the off-center uptakes caused
compensate for line of sight blockage, another two were more difficulties, particularly for the after deckhouse.
70 Naval Engineers Journal, May 1988
INMAN TYPHON T O AEGIS

The disadvantages of the deckhouse concept were dem- ter immunity was of less importance to SAM-D. SAM-
onstrated when NavSec made an Aegis DD-963 design D did not have an AIC mission so the maximum detec-
conversion feasibility study using the Deckhouse Mark tion range requirement was much less than for ASMS.
20 concept. The deckhouse concept was finally aban- Finally, the defended volume was only a little more than
doned for the CG-47 design, and many of the originally a quarter of the hemisphere. This greatly relieved radar
claimed advantages were recovered at the production power and time budget limitations compared to ASMS
test site where systems could be assembled and groomed which had to cover the whole hemisphere.
prior to ship installation. SAM-D mobility was a major requirement for air de-
fense operations. While firing on the move was not a re-
EVOLUTIONARY
DEVELOPMENT
AND quirement, a battery had to be able to transition from
COMMONALITY
WITH OTHERNAVYSYSTEMS moving to full AAW readiness in a very short time. This
meant that the radar and missile guidance functions had
The Typhon system was t o be all new. The ASMS to be mounted in a single tracked vehicle with height,
baseline system that came from the Assessment Group width and weight constraints. The smaller C-Band
was new except for computers and displays. The antenna which was also used for illumination was ideal
launcher was to be new but developed t o serve both for this land mobile system. Compared to the Army, the
ASMS and Tartar. By the start of engineering develop- Navy had a broader set of requirements and a much
ment, the missile was a derivative of Standard Missile larger “vehicle” to accommodate the system. The Navy
and the illuminators were derived from the Tartar Department strongly opposed the common system and
system. In fact, two of the six ASMS Tartar derived il- OSD finally dropped the common system development
luminators were to have retained an active radar track- initiative when proposal cost estimates seemed to indi-
ing capability in addition to the capability of being slaved cate that the two systems could be separately developed
to the multifunction radar. These programmatic for a total cost less than the cost for developing a com-
changes had the advantage of directing funding and em- mon system.
phasis to the radar and computer based control
elements, the heart of the system. A further advantage
was the benefits to other Navy programs from the Aegis SUMMARY AND CONCLUSIONS
effort. The disadvantages were the additional project
management problems and the considerable increase in The Typhon system concept introduced in the late
production cost and weight of the illumination system. 1950s featured close control command midcourse with
This was somewhat relieved by deleting the two tracking semiactive terminal homing missile guidance and system
illuminators. In retrospect it is believed that, on support by an agile beam radar. Typhon promised
balance, the evolutionary development and commonali- major shipboard surface-to-air missile system perfor-
ty decisions were good for the Aegis program as well as mance improvements compared to the Tartar/Terrier
other Navy programs. systems. The promised performance improvements and
more have now been realized with the Aegis system
ASMWSAM-D COMMONALITY using the Typhon concept.
The decision to use the radar for semi-active target
After the completion of the ASMS assessment illumination caused the Typhon radar development
studies, OSD concluded that the development of ASMS failure by demanding more from the radar than the then
and the Army’s new AAW missile system, SAM-D, current technology could provide.
should be merged and a common system developed. The choice of S-Band over C-Band for the ASMS
SAM-D evolved into the current Patriot system. Since baseline radar system was the correct decision for a top-
the threat targets of the two systems were the same, the of-the-line cruiser/destroyer shipboard missile AAW
merger concept had a certain appeal to OSD. The SAM- system. The choice of C-Band could have led to another
D baseline system was much like a simplified Typhon development failure or a less capable system, more suit-
system with a single face phased array radar. The radar able for smaller (FF) ships.
operated in the C-Band and provided target illumina- The ASMS radar transmitter tube size decision, while
tion for missile guidance. A commonality study was correct for the then current circumstances, has resulted
ordered and a joint Army/Navy group was formed to in a AN/SPY-IA/B/D transmitter that is now some-
do this study. The study group quickly came to realize what heavier and more expensive than optimum. The
that the system differences were driven by very real dif- advantages for reoptimizing at this time are not worth
ferences in the operational requirements and the dif- the cost.
ferent Army and Navy environments. The ASMS decision to follow an evolutionary devel-
With the Army operational concept, low altitude air opment approach and some forced commonality with
defense was provided by light AAW systems distributed existing systems for the launcher, missile, illuminator
in the battlefield. The SAM-D primary mission was to and computer/display systems was good for both the
provide a high altitude “umbrella” for those light AAW new Aegis system development and the existing systems.
systems that have little or no capability against the The vertical launching system was another step in the
higher altitude threat. The targets at higher altitudes evolution and it is hoped that a new missile family and,
were usually well above the weather clutter so that clut- perhaps a light weight, low cost director are in the future.
Naval Engineers Journal, May 1988 71
TYPHON TO AEGIS INMAN

The Deckhouse Mark 20 was not a good ship installa- Early review of a major combat system development
tion concept as the weapon system design independence by a broad spectrum, service dominated, assessment
from the ship design process and the production testing group was of great value to the ASMS program. Time
advantages are not worth the cost in metacentric height. delay, cost and program manager frustration not with-
Any major shipboard combat system development standing, such a review is probably worthwhile for any
effort involving large topside elements such as antennas, major program during the concept design phase.
launchers and directors should not proceed without
strong and continuous input from the naval architec- REFERENCES
tural community. By the same token warship designs
need a strong and continuous input from the combat [ I ] Madsen, Adelaide: A n Aegis History (unclassified section
system community. The establishment of NavSea has on Typhon), NavSea, May 1986.
[2] The Johns Hopkins University Applied Physics Labora-
facilitated this exchange.
tory: The First Forty Years, 1983.
“Obvious” inter-service weapon system development [3] Skolnik, Merrill I: Introduction to Radar Systems,
duplication should be carefully reviewed. Profoundly McGraw Hill Book Company, Inc. 1961. pp 291-294.
different system approaches can be the direct result of [4] Nathanson, Fred E., Radar Design Principles, McGraw
valid differences in operational requirements or operat- Hill Book Company, Inc. 1969.
ing environment. On the other hand, techniques, com- [5] Cheston, The0 C., IEEE Spectrum, November 1968, pp
ponents and subsystems may be directly transferable. 102-1 1 1.

12 Naval Engineers Journal, May 1988