High costs are reshaping the development of fighter programs
 

BY BILL SWEETMAN

Fighters cost far too much and take too long to develop. Today's high-end fighters -- the Rafale,
Eurofighter Typhoon and F-22 -- were designed according to doctrines that trade cost against
effectiveness. They are survivable, multi-role aircraft which are far more than equivalent,
one-to-one, to the aircraft that they replace. However, all of them have reached the point of
indigestion for even superpower or multinational budgets.

Both the Typhoon and the F-22 will enter service
around 2005, more than 20 years since the specifications for the two aircraft were
written. The Typhoon will dominate its sponsors' defense budgets for a generation,
while the F-22's price tag has caused the number of aircraft built to diminish; the US Air
Force (USAF) will be lucky to get half the number of aircraft planned when engineering
and manufacturing development (EMD) started. Lower production rates drive the price
further upwards.

As new programs are delayed and production rates reduced, existing fighters must be
kept in service longer and require urgent and expensive upgrades, diverting more funds
from new programs. The result is a vicious spiral in which the average age of the fighter
force increases, until it is divided between a handful of new aircraft and a larger force of
older and less capable types.

This problem has been recognized since the 1970s. The USAF and US Navy (USN)
attempted to tackle it in the 1970s, by deploying a 'high-low mix' of fighters. More
costly, higher-performance aircraft, the F-15 and F-14, would be built for the most
difficult and critical missions, and would be backed up by two to three times as many
smaller, simpler F-16s and F/A-18s. This logical plan had two unintended
consequences. The first was that the 'high' aircraft became caught in the vicious spiral
of lower production rates and higher cost. The second was that the 'low' aircraft
competed with the high-end fighters when the time came to fund upgrades, follow-on
production and attrition replacements. The low-end aircraft prevailed in this competition
and eventually starved the high-end programs to death. One result is that the USN
currently has no heavy-payload, long-range strike aircraft apart from a small band of
F-14 Bombcats; another is the controversial replacement of the F-4G by the F-16C Block
50, rather than an F-15E, in the Wild Weasel mission.

Today, governments and industry are testing several new approaches to the problem of
fighter affordability. Most of them have yet to prove themselves, but the fighter
community can only hope that some of them succeed -- if only because octogenarian
fighters are even less of a practical prospect than octogenarian bombers (see feature
beginning on p56).

One way to reduce cost is to eliminate the massive investment required to design and
build a new aircraft, by improving an existing platform. The most ambitious program of
this kind is Lockheed Martin's effort to launch a range of new advanced-technology
options for the F-16, supported by export orders. This program has already achieved
some concrete successes, is close to accomplishing the company's strategic goal, and is
on the verge of a major breakthrough.

The F-16 product range today is aimed at three markets. The
entry-level version is a refurbished F-16A/B in the Mid-Life
Update (MLU) configuration, with thoroughly modernized, simple
avionics: this variant has now been selected by New Zealand,
which is planning to acquire 28 virtually new F-16A/Bs which
were originally built for Pakistan, but never delivered. The MLU
has a full-color cockpit, the Improved Data Modem (IDM)
datalink, a programmable electronic warfare (EW) management unit from TERMA
Elektronik and a readily upgradable mission computer. It is one of the first operational
fighters with a digital terrain system (DTS) for targeting, ground collision avoidance and
'quiet' low-level navigation. It is fully AMRAAM-compatible and is cleared to carry a
variety of sensor, reconnaissance and EW pods.

The mid-range F-16 is the Block 50/50+ variant. Despite the fact that the USAF allocated
little funding to the development of the F-16C/D since the early 1990s, the company was
able to keep the fighter contemporary by incorporating key MLU features -- notably, the
color cockpit, DTS and mission computer -- into the heavier, more powerful Block 50.
Offered an off-the-shelf upgrade based on this configuration, the USAF launched the
F-16 Common Configuration Implementation Program (CCIP) last June, awarding an
EMD contract to Lockheed Martin. Under the CCIP, the USAF will upgrade 700 Block 40
and Block 50 aircraft to a largely common configuration, reducing support costs. The
CCIP F-16 will feature the Link 16 datalink and provision for the AIM-9X missile and
associated Joint Helmet-Mounted Cueing System.

The CCIP is the baseline for the Block 50 versions now offered for export, but Lockheed
Martin offers some extra airframe options. In 1998, the company supplied Singapore
with F-16Ds that feature the long, square dorsal spine first seen on Israel's two-seat
Block 40 aircraft. The spine accommodates electronic equipment, the inflight refueling
receptacle and chaff/flare dispensers, releasing more space within the fuselage for fuel.
The result is a two-seat aircraft with almost the same internal fuel capacity as the
single-seat aircraft -- for this reason, the extended spine is usually associated with a
missionized rear cockpit designed for a weapon system operator, not a trainee.

Greece has now become the first export customer for the Block 50+, which combines
CCIP-type avionics with the conformal fuel tank (CFT) modification which Lockheed
Martin originally proposed to Israel, and tested in 1994­95. With the overwing
conformal tanks and 600 USG underwing tanks, the F-16 is claimed to have a hi-lo-hi
mission radius of 1,170km, carrying two 900kg bombs, two AIM-9s, two AIM-120s and
LANTIRN and ECM pods. The new tanks provide more options in stores carriage: the
CFTs make it possible to achieve a respectable combat radius with a four-HARM load
(tested in 1997). Greece's two-seaters will have the enlarged spine, and are likely to be
fitted with the Litton-integrated ASPIS (Advanced Self-Protection Integrated Suite) EW
system, already ordered for Greek Block 50s.

The F-16 is available with a more diverse range of EW options than any other fighter,
allowing the customer to find the most affordable solution to the national requirement.
ASPIS represents a high-end solution: based on the Litton Applied Technologies
ALR-93 radar warning receiver and management unit, it incorporates the Raytheon
ALQ-187 active jammer, the Litton/DaimlerChrysler AAR-60 missile warning system and
the Marconi ALE-47 dispenser. Likewise, Turkey has adopted a comprehensive suite for
its Block 50s, in the form of the ALQ-178(V)5. This system, descended from the Loral
Rapport line, is produced and supported domestically by Microwave Electronics
Systems Inc (MiKES), which was founded by Loral, and has been co-developed by
MiKES and Lockheed Martin. Korea, meanwhile, has selected the ITT/ Northrop
Grumman ALQ-165 jammer and Lockheed Martin ALR-56M RWR. Elta and Elisra offer
F-16 EW packages, as does Thomson-CSF Detexis.

'Pay-as-you-go' F-16

The USAF has stuck with the Raytheon ALQ-184 pod as its standard jamming solution
for the Block 40/50. After one of its F-16s was shot down by an SA-6 over Bosnia in
June 1995, however, the USAF launched a quick-reaction program to equip its entire
fleet with the Raytheon ALE-50 towed monopulse decoy. Modified outer wing pylons
carry two ALE-50 launch tubes on each side.

The importance of this evolutionary development process is that it has delivered a
much improved aircraft on a pay-as-you-go basis. The US has not had to spend a lot of
money on F-16 development since the Block 40 and Block 50, in the late 1980s, but now
has a significantly better aircraft, capable of challenging newer European rivals.

Additional orders from Egypt and Greece will sustain F-16 production into the early
2000s. This takes Lockheed Martin a long way towards its strategic goal: to keep the
F-16 line warm until the Joint Strike Fighter (JSF) replaces it. As this issue closed for
press, too, Boeing announced that it was closing its F-15 line, suggesting that the
pending Israeli fighter order will also go to Fort Worth; and the F-16 remains in
competition with the Typhoon in Norway. However, F-16 development does not stop
with the Block 50+. Under contract to the United Arab Emirates (UAE), Lockheed
Martin plans to develop further improvements to the basic platform, to the point where
it will emulate the JSF in almost every respect except stealth.

Lockheed Martin is wary of using the term 'Block 60' and it is hard to pin down the UAE
configuration with absolute precision. However, there are a number of features that
appear to distinguish the UAE aircraft from the Greek Block 50+. The most important
and expensive of these is the Northrop Grumman sensor suite. Based on the
APG-68(V)5, it is an integrated system that blends the Internal FLIR and Targeting
System (IFTS) with an Agile Beam Radar (ABR). The ABR is an active-array system,
drawing on F-22 technology. The IFTS has two stabilized turrets, above and below the
nose. A wide-field-of-view navigation and target acquisition FLIR is installed in the
upper turret, while the lower unit houses a targeting FLIR with a built-in laser
designator/ranger. The two systems share a common optical path and use a
third-generation mid-wave infrared detector. The IFTS eliminates the drag of external
targeting pods and uses the signal processing hardware as the radar.

The new F-16 will have a revised avionics architecture, making greater use of
commercial technology. This will probably be able to fuse information from the ABR
and IFTS with the digital terrain system and the fighter's EW system. It is the
last-named item that has been a sticking point in negotiations between the UAE and the
US government.

Apparently, the UAE is seeking a more advanced suite than the ALQ-165, and favors a
system based on a version of the ITT ALQ-211 Suite of RF Countermeasures, under
development for the AH-64D Apache. It also includes Litton Amecom's high- precision
ALR-100 radar-warning receiver, with a four-quadrant interferometer array. The problem
is that such a system may well be superior to any in-service US equipment; it will, for
example, give the UAE aircraft the ability to fire a HARM in range-known mode without
any off-board cueing. Lockheed Martin still hopes that the releasability issues will be
resolved in time to sign a contract with the UAE this year.

Other features of the new F-16 are a more powerful engine (either the P&W
F100-PW-229A or the GE F110-GE-129EFE) and a revised cockpit, with larger primary
color displays and fewer, smaller standby instruments. Both engine makers have touted
design changes that increase the life of the engine and reduce maintenance costs.
Lockheed has worked very hard to reduce the cost of the F-16 platform, and LMTAS
has become converted to the doctrines of 'lean production' (see sidebar). The mile-long
Fort Worth plant, which formerly produced thousands of B-24s and hundreds of B-36s,
is now half empty, but remarkably quiet and clean.

Gripen and Typhoon

As it is nearly ubiquitous in fighter competitions, the F-16 sets the benchmark for rivals
such as Rafale, Mirage 2000, Typhoon and Gripen. These programs, too, have embraced
lean production and other cost initiatives in an attempt to dislodge the F-16 from its
position. Sales successes by the Gripen in South Africa and the Typhoon in Greece in
the past year are undeniably significant. For many years, Sweden's fighter programs
have been admired worldwide in almost exactly inverse proportion to their commercial
success, while previous European multinational fighter programs have flopped on the
export market: not even the Tornado's greatest fans could deny any linkage between the
Saudi sales and US restrictions on F-15 exports. The Gripen and Typhoon deals,
however, appear to have been as straightforward as can be expected in this business.

            While the Typhoon is nobody's idea of an inexpensive aircraft, British
            Aerospace has developed a widely acknowledged world-class
            capability in lean manufacturing, and the fighter's long development
            has given BAe and the other Eurofighter partners plenty of time to
            refine the design and squeeze out unnecessary costs. Fortunately for
            Eurofighter, the kind of manufacturing and information technology
that supports lean manufacturing also makes it easier to spread production work around
multiple sites. Typhoon, therefore, does not carry the kind of production cost penalty
that afflicted early multinational military aircraft.

Eurofighter's approach to Typhoon has been to stress effectiveness as well as cost.
The team is adding systems such as DTS, and proposing new options such as
conformal fuel tanks, to make the fighter more versatile. BAe, meanwhile, has moved
into the low-cost fighter market by teaming with Industry Group JAS on an exportable
Gripen.

Sweden has been among the first countries to introduce many of the most important
technologies and philosophies in fighter aircraft. The JAS 39 Gripen continues this
trend, having been designed specifically to survive in the toughest environment for a
fighter -- an era of restricted budgets. Smaller than the Viggen, which it will replace, the
Gripen is built in one basic version, with pods for special missions, and is designed to
require minimal support. It is the first fighter to be designed on the assumption that its
primary weapons will be precision-guided, reducing the need for a heavy weapon load.

There is more to the Gripen, however, than small size. Brigadier General Mats Hellstrand,
director of plans and programs for the Swedish Air Force, says that the Gripen ushers in
a "fourth-generation air force" with a strong emphasis on information warfare: "denying
and destroying enemy information processes while we protect our own". Connected by
secure datalinks to other Gripens, and to the air force's 'information aircraft' -- the S 102B
SIGINT platform, based on the Gulfstream IVB, and the S 100B AEW&C aircraft -- and
able to switch targets and missions in flight, the Gripen can prevail without relying on
size and raw firepower.

"We have had secure datalinks in the Swedish Air Force for more than a decade,"
Hellstrand comments, "but with AMRAAM and the datalink on the JAS 39, we may see
a change in tactics." Hellstrand believes that, with secure communications, groups of
fighters will spread out more widely to cover a greater area with their missiles.

The combination of intra-flight datalinks and AMRAAM is very new in service, with
the Gripen and the F-16 MLU leading the way, but is showing every sign of being a true
force multiplier: another key to affordability. With datalinks, fighters can use their radars
more sparingly. One fighter in a group may stand back and search for targets, sending
information via datalink to other fighters. The adversaries focus on the 'illuminator' and
may not see the other attackers until it is too late. "The Belgians went up against some
USAF F-15s earlier this year, and didn't say that they were taking MLUs," commented
John Engels, Lockheed Martin avionics team leader on the MLU program. The F-16
pilots set up 'silent' AMRAAM attacks using the MLU's IDM, and gave the Eagles an
unpleasant shock. Any system that evens the odds between a larger and smaller fighter
is a step forward in affordability.

While the managers of current programs are working hard to drive affordability into
their products, one fighter program has been focused on costs from the outset: the
Pentagon's potentially massive JSF project. A cardinal principle in the JSF program is
'cost as an independent variable' (CAIV), meaning that the JSF's price can and will be
kept within its target range, and that the required performance will be adjusted as
necessary to achieve those goals.

The crux of the JSF requirement is the USAF's demand for a US$28 million flyaway cost,
somewhat more than an F-16. For this, the customer wants F-16-like flight performance,
considerably greater range, first-day stealth with internal weapons, sensor fusion and a
fully integrated internal visionics suite. What could be more reasonable?

The current phase of the JSF program has two tracks, both of which support CAIV. In
one track, Boeing and Lockheed Martin are designing their Preferred Weapon System
Concept (PWSC) aircraft, while the JSF Program Office is writing its requirement. This
process has run on an annual cycle, through a series of Joint Interim Requirements
Documents (JIRDs). As each JIRD evolves from a draft to a final document, the
contractors evolve their PWSC designs to meet it, and to take advantage of lessons
from JSF technology programs, while remaining inside the cost limitations. The JIRD
also changes as the Pentagon conducts more analysis and simulation of the way that
the changing performance projections affect the entire air campaign. The JIRD process
will culminate next year in a final Joint Operational Requirements Document (JORD), and
the final iteration of the PWSC design will be the contractors' EMD proposals.

The other track in JSF comprises the Concept Demonstration
Aircraft (CDA) programs being conducted by the two contractors.
At this point, 18 months or more after the CDA designs were
frozen, changes in the PWSC no longer affect the CDA; however,
lessons from the CDA are still being taken into account in the
PWSC design. The main function of the CDA is to demonstrate
that the contractors understand the key airframe technologies in the PWSC: of these, by
far the most important is the short take-off/vertical landing mode, with carrier approach
and landing qualities running second. The difference between the CDA and a Dem/Val
or fly-off competition is that the goal is not to show that one or other of the contractors'
approaches to the JSF mission offers better performance, but to confirm that the
Pentagon can choose the less costly solution with acceptable risk.

The result, irrespective of which team wins, should be a low-risk approach to a
thoroughly 'scrubbed' requirement, and a customer who is fully aware of how much it
will cost to meet any change in requirements. CAIV, it is hoped, will deter the growth in
requirements which underlies many cost increases.

As the PWSC designs have evolved, Boeing and Lockheed Martin approaches to the
JSF have diverged. Boeing still offers a highly common design -- 85­90% common
across all three variants, the company says. The primary difference between the short
take-off/vertical landing aircraft and the other two variants is the lift module, located
behind the engine. The rear nozzle is identical across all three versions, and the USN
variant's wing is only slightly larger than that of the USAF aircraft.

CTOL cost control

David Brower, Boeing's director of affordability for the JSF team, argues that the direct
lift solution that the company has chosen is inherently lower-cost than the
fan-augmented Lockheed Martin system. "The requirement for bring-back weight [the
load with which the JSF can land vertically] is a good match for high performance and
survivability. It's a three-point solution, and you don't have to make the aircraft too big,
to lug around equipment that you only use for 90 seconds."

Lockheed Martin's approach is rather different. "Commonality is a means towards
affordability," says Lockheed Martin deputy JSF program manager Harry Blot. "We
have about 80% commonality between the CTOL and STOVL aircraft, and 70% between
the STOVL and CTOL versions. It has not affected affordability."

In the latest iteration of the Lockheed Martin JSF, Design 230-3, the company has
exploited the flexibility of its wing/tail configuration to provide the carrier-capable (CV)
version with a far larger wing than the other JSF variants. The Design 230-3C CV
version has an F-15-sized 55.7m2 wing, 45% bigger than the 38.3m2 wing of the 230-3A
conventional take-off and landing (CTOL) version and the 230-3B short take-off, vertical
landing (STOVL) variant. Moreover, Lockheed Martin's STOVL system adds a complete
propulsive lift unit, with inlets and a nozzle, behind the cockpit, and the STOVL version
has a different aft nozzle. This approach reflects the fact that the way to keep the total
acquisition cost for the JSF low is to reduce the cost of the CTOL aircraft, because that
variant will account for most of the JSFs built. This is reflected in the customer's unit
cost targets: as noted above, the USAF price is US$28 million, but the US Marine Corps
will pay up to US$35 million for the STOVL aircraft. Lockheed Martin keeps the basic
CTOL configuration small and simple, and adds components to meet the CV and STOVL
requirements. Both competitors are laying great stress on the use of new design and
manufacturing techniques to reduce costs in development, production and support.
Essentially, the goal is to make a complete transition to database-driven manufacturing
and assembly. Every part of the aircraft is modeled in three dimensions in the computer,
and every machining, composite lay-up and assembly operation is defined and
simulated on computers before it is attempted physically.

                To consider the impact of this process on cost, it is important to
                look at the traditional way of building aircraft. Airframes are
                large, complex and built with very tight tolerances. In a
                traditional approach, a preliminary design is laboriously
                translated into thousands of two-dimensional drawings, which
                are used to fabricate parts. As discrepancies are found during
assembly, the drawings and parts are revised, consuming more time and money. The
process is complex and error-prone; parts constantly fail to pass inspection and are
returned for rework.

To assemble the aircraft with high precision, components and sub-assemblies are
loaded into heavy, rigid fixed tools or jigs (in a large aircraft, these rest on 23mm2 steel
columns) which hold every part in exactly the right place. Once the parts are brought
together, fasteners are installed with hand-held tools. Installing production tooling is
enormously expensive and it is very difficult to make changes after this is done.

Computer-aided design has improved the first part of this process -- the transition from
a preliminary design to part design -- and has generally improved the first-time fit of
components, but heavy tooling has remained the rule and parts are still hand- fastened
together.

In the 1980s, many production experts looked at the automotive industry as a model for
aircraft assembly, envisioning highly automated factories with extensive use of robots.
Today's new techniques, however, recognize that aircraft manufacture is a different
animal. Even high production rates have two to three fewer zeroes left of the decimal
point than most mass-production operations. In aircraft production, it is vital to reduce
the effort and investment involved in establishing or changing the line.

Boeing, in particular, is stressing the fact that its X-32A/B CDA aircraft are not just
flight-test aircraft, but demonstrators for the new production technology which it would
use in EMD. One database -- a 'single source of data' is the watchword -- controls work
in Seattle (where the X-32 was designed, and where most of the titanium and aluminium
parts are machined); in St. Louis, where the forward fuselages are assembled, and in
Palmdale, where the aircraft are being built.

Boeing has demonstrated tools that take complex parts from the database to reality in
three steps. An automatic 'control code generator' translates the electronically defined
part into commands which drive a five-axis machine tool or an automated composite
tape-layer. Then, the operation is simulated, allowing engineers to watch the process
happening and make any needed improvements. Only then is the physical part
produced.

Some of the techniques used in production are derived from the F-22 (Boeing builds the
fighter's wings and aft fuselage) or other current production programs. Wing spars, for
example, are produced by resin transfer moulding (RTM) in which carbonfibers are
placed in a mould and hot liquid resin is injected. In other areas, new techniques have
been used. High-speed machining uses milling heads which spin at very high speeds,
reducing stress on the material, and is used to fabricate doors and other parts which
would otherwise be hand-assembled from sheet metal skins and stiffeners. The X-32
nosewheel door is made from two high-speed machined skins which include tongues
and grooves, so that they snap together and are assembled without a fixed tool. The
complex inlet duct, which curves and transitions from a 'smile' at the front to the circular
engine face in a very short distance, was produced in one piece.

Hydraulic, fuel and oxygen lines have been designed and installed with the help of a
'smart router'. Told where the lines are supposed to go, and what loads they carry, the
router checks the 3-D database for interferences, and defines the lines in terms of
diameter, wall thickness and bend radius. Working from a database of parts, the router
defines and installs connectors, clamps and attachments. Every one of the 95 lines in
the forward fuselage fit first time, according to Boeing.

Laser theodolites

In final assembly, Boeing has dispensed with conventional tooling, using a technique
which the Phantom Works has been exploring since the early 1990s. Instead of using
massive fixed tools, designed to retain their tolerances for the entire production run,
Boeing is assembling major X-32 components using adjustable fixtures which stay in
place only long enough for one assembly operation. What makes this possible is that
the location of the part is no longer defined by the fixed tool. Instead, assembly is
controlled by computer-driven laser theodolites.

Working from reference points that are marked on each part during assembly, the
theodolites (produced by Leica) determine the exact location of each part and illuminate
fastening and drilling points with a laser spot. For instance, the wing-to-body junction
requires 130 brackets to be installed on the lower surface of the wing. The laser was
used to mark all of them, with a tolerance of 0.01cm. As soon as an aircraft assembly is
complete, the adjustable tooling is removed and the part becomes its own tool: the
process is exact enough that errors will not be compounded during further assembly
operations.

Precision has other advantages. Fastener holes that traditionally would be drilled after
assembly -- to ensure that the holes through two parts would match -- are now
pre-drilled or pilot-drilled when the parts are fabricated. Skin components are delivered
pre-drilled and trimmed to size. ("In the old days," comments John Priday, leader of the
assembly team at Palmdale, "you added a half inch on every side and prayed that you
had enough.") Ducts and tube sections arrive at the assembly plant with both ends
ready to fit to the next section, and they fit. When the forward fuselage for the X-32A
arrived in Palmdale from St. Louis, "we had it mated to the mid-body, laser-tracked and
the holes drilled in six hours," remarks Priday. "It was April 1 and everyone thought we
were kidding."

The new processes are showing large, measurable improvements.
"When we started on the mid-fuselage, we expected that it would
take 50% of the labor that went into the YF-22 aft fuselage," says
Brower. "It took half of that. We replanned the process for the
second fuselage and it took less time than the first." The X-32 has
taken a quarter as many tools as the YF-22, and has had 90% fewer
discrepancies in assembly. "We are very comfortable," says Brower, "that we can meet
or beat the target flyaway cost."

The JSF office is using competition to impose the same kind of cost discipline on the
fighter's engine; both Pratt & Whitney and a GE/Rolls-Royce Allison team will provide
engines for production aircraft, and the cost goal, remarks one Allison engineer, is "not
even close" to the F119-PW-100 on the F-22. While Pratt & Whitney is reducing the
price of its engine through redesign (some of which will save money on F-22 engines),
GE is working towards a common core for the JSF engine and the follow-on to the
best-selling CFM56 commercial turbofan.

Whether the JSF truly represents the next century remains to be seen. The only
certainty about the program's long-term future is that there is a Presidential election
between now and the start of EMD, and that no new administration will rubber-stamp a
plan that will dominate defense budgets for two terms. A hawkish administration may
question whether one aircraft can be the answer to all the services' needs; a
parsimonious one may look at the nature of current and likely conflicts and adversaries,
and ask whether an advanced F-16 does not meet many of the users' needs at lower
cost.

It can also be argued that emerging technologies -- such as small precision-guided
weapons, tailless aircraft design, and advanced engines with fixed exhausts -- promise
much less costly solutions to a fighter requirement than JSF, which represents no great
advance over the F-22 in weaponry, aircraft design or stealth. Even more radical
approaches -- such as space-based weapons and UCAVs ­-- may also compete for
funding. Whatever is built, though, it will have a price tag before preliminary design
begins. The days of an open government wallet are over.