entire cfm history article

\

<r<

•? ty f >

Customers and products THE CFM56 STORY

Total: 270 users worldwide AB Airlines ABSA Cargo Aer Lingus Aeroflot Aeropostale Aerostar Leasing Air 2000 Air Afrique Air Algerie Air Austral Air Belgium Air Berlin Air Calin Air Canada Air Charter Air China Air Europa Air France Air Guizhou Air Holland Charter BV Air Jamaica Air Lanka Air Liberie Air Madagascar Air Malawi Air Malta Air Mauritius Air Nauru Air New Zealand Air Nippon Air One (Italy) Air Pacific Air Tahiti Nui Air Tanzania Air Toulouse International Air Transat Air Vanuatu Airasia Airtours Intl Alaska Airlines Alitalia All Nippon America West American Airlines Ansett Australia Ansett Worldwide

AOF Leasing AOM French Airlines Arab Leasing Aramco ASERCA Asiana Austrian Airlines Balkan Airlines Bavaria Flug Blue Panorama Airlines Boeing Business Jet Boullioun Aviation Braathens Britannia Airways British Airways British Midland Burlington Air Express CAAC Camaroon Airline Canada 3000 Canadian Airlines Intl Canarias Regional Air Carlisle Aircraft CASC Cathay Pacific Chilean Air Force China Airlines China Eastern China Eastern Shanxi China Northwest China Southern China Southwest China United China Xinhua Airlines China Xinjiang Airlines China Yunnan CNAC - Zhejiang Airlines Color Air Condor Continental COPA Panama Corsair Country Heights Golf

Resort Croatia Airlines Cronus Airlines

CSA Czech Airlines Delta Air Lines Deutsche BA DHL Airways Eastwind Airlines Easyjet Edelweiss Air EgyptAir El Al Emery Estonian Air Euralair Eurowings Falcon Cargo Fast Air Finnair Fischer Air Florida West Int'l Airways Flying Colours Airlines French Air Force Frontier Futura Intl Gabonese Government Garuda GATX GB Airways GECAS German ia GO GPA Group Guangxi Airlines GulfAir Hainan Airlines Hapag-Lloyd Heller Financial Iberia Iberworld Icelandair ILFC ISRAir Istanbul Airlines Itochu Airlease ITT Flight Operations JAL Japan TransOcean Air JAT

Jet Airways Jet Aviation Kenya KLM Korean Airlines Kuwait LAB LAPA Lauda Air Lithuanian Airlines LOT Lotus Air Lufthansa Luxair Macedonian Airlines Maersk Maersk Air UK Magic Carpet Avia Malaysia Malev MAS Air Cargo Mid East Jet Monarch Myanmar Airways NASA North American Airlines Northwest Nouvelair Tunisie Novair Sweden Novel Leasing Olympic Oman Air Pakistan Pakistan - Government

VIP Pegasus Pembroke Captial Peru - Government VIP Philippine Airlines Polynesian Premiair A/S Princess Airlines Privatair Pro Air Qantas Rio-Sul

Royal Air Cambodge Royal Air Force Royal Air Maroc Royal Jordanian Royal Saudi Air Force Royal Thai Air Force Ryan International

Airlines Ryanair Sabena Sabre Airways Sahara India Airlines SALE SAS SATA

Shaheen Air Int'l Shandong Airlines Shangahi Airlines Shantou Airlines Shenzhen Airlines Silk Air Singapore Airlines Skyservice Sobelair Solomon South Pacific Air

Charters Southwest (US) Star Europe Sterling European Sun Express Sunrock Aircraft Swissair TACA TAESA TAG Aviation Tampa Airlines TAP Tarom

Tea Switzerland Thai Airways

International Tombo Aviation Services Transaero Airlines Transavia Transbrasil

Travel Service Airlines Tunisair Turkish Air Force Turkish Airlines Turkmenistan US Air Force US Navy USAF National Guard USAF Reserve Uganda Airways Ukraine International

Airlines Unident D United Airlines UPS US Airways Varig VASP Vietnam Airlines VIP - Amiri Flight Abu Dhabi VIP - AS Bugshan VIP - Brunei VIP - Dallah Al Baraka VIP - Dubai VIP - Egyptian VIP - Italian AF VIP - Prince Aziz VIP - Qatar VIP - ROCAF VIP - Saudi Oger VIP - Saudi Royal Fit VIP - Trans Aeros

Presidencial Virgin Atlantic Virgin Express Virgin Express France Virgin Express Irelane Viva Volare Winair (US) Wuhan Airlines Xiamen Airlines Zhongyuan Airlines Zhuhai Airlines

As or 31 March, 1999

CFM56 product line CFM56-2 Current rating is 22,000-24,000lb (98-107 kN) thrust Engine entered commercial service in April 1982 Commercial Applications: DC-8 Super 71, 72, and 73 Commercial aircraft in service: 110

Engine entered military service in June 1984. Military applications: Boeing KC-135R, C-135FR, and

KE-3A tankers, E-3A Airborne Warning and Control System aircraft,

RC-135 reconnaissance and E-6A submarine communications

aircraft. Military aircraft in service: 463.

CFM56-3 Current rating is 18,500-23,500lb (82-104 kN) thrust Engine entered revenue service in December 1984 Applications: Boeing 737-300, 737-400, and 737-500 Aircraft in service/on order: 1,987

Current rating is 25,000-26,500lb (111-118 kN) thrust Engine entered revenue service in April 1988 Applications: Airbus Industrie A319 and A320. Aircraft in service/on order: 522

Current rating is 22,000-32,000lb (98-142 kN) thrust. Engine entered revenue service in April 1994 Applications: Airbus Industrie A319, A320, and A321. Aircraft in service/on order: 610

Current rating is 31,200-34,00lb (139-151 kN) thrust. Engine entered revenue service in March 1993. Application: Airbus Industrie A340. Aircraft in service/on order: 202

CFM56-7 Current rating is 18,500 to 27,3001b (82-121 kN) thrust Engine entered revenue service in January 1998 Applications: Boeing Next-Generation 737-600, 737-

700, 737-800, and 737-900 Aircraft in service/on order: 1,140

As of 30 April, 1999

FLIGHT INTERNATIONAL May 19 - 25 1999

-is

5>

CFM56-7B iareth Burgess

f ss (when compared to CFM56-3 engine This page should hold a cutaway poster of , h 0 J d f a n b ,ades

K M

the CFM56. If yours is missing or dam- 'm c o n 1 p r e s s o r aims

aged please return the whole issue to Kim Hearn and we will replace it. red with low-emission double annular combustor (optii


I "We need to invest in this technology, to demonstrate its feasibility, and get it ready for us to move'''' - Bill Clapper

keep it simple, reliable and affordable - the hallmarks of its success to date.

Major studies are focused on a high stage-loading high-pressure compressor incorporating advanced three-dimensional computational fluid dynamics design. The aim is to cut the high-pressure compressor stages to six, from the current nine. Another major departure is to the plan to use a close-coupled, counter-rotating, four-stage low pressure turbine. This uses technology pioneered on the GE36 propfan and GE's YF120 advanced fighter engine, and requires studies of counter-rotating differential bearings as well as complementary high pressure turbine advances.

To test the low speed compressor development, CFMI devised a one-of-a-kind test device which resembles a huge, vertically stacked core. This allows easy access to the compressor and enables a larger number of different configura

tions to be tested in a shorter time. Testing ofthe high-stage-loaded high pressure compressor in 1998 validated the new design and demonstrated the improved aerodynamic stability of forward swept blades. Tests of advanced high pressure turbine designs also proved the validity of a new convergent-divergent blade, and a 3D-aerodynamically designed first stage high pressure turbine vane. The tests also yielded a 0.5% fuel burn improvement and up to a 10°C increase in exhaust gas turbine margins. The low-pressure turbine changes, meanwhile, resulted in a 2 % fuel consumption improvement and a higher bypass ratio of 9 (compared to 5.5 on the original design), with only four stages and 20% fewer blades and vanes.

Another maj or focus is on the development of a swept, wide chord fan. The superior aero

dynamics of the swept design are expected to generate higher thrust, a 1 % fuel consumption reduction, lower noise and up to 70kg (1501b) weight saving. "As you go up in size beyond the -7, you have to do something like this," says Alike Benzakein. Crosswind tests are planned this year using a CFM56-7 to provide the core power. Fatigue and initial containment tests were completed in 1998, and a full-size blade release is scheduled for later this year.

Intimately tied to the new fan blade studies is the development of a load reduction device (LRD), or fan decoupler. The Snecma developed innovation is designed to sever instantly the link between the low-pressure system and the rest ofthe engine in the event of a fan blade detaching. As well as increasing safety by reducing fan imbalance, the system offers potential weight savings of up to 90kg because less structural reinforcement will be needed. The LRD works by using shear bolts in the number one bearing support assembly. These are designed to break when loads exceed preset limits, which would be reached almost instantly following a blade failure.

"It's a totally new approach to the blade out problem," says Clapper, who expects the feature to be one of the first from the TECH56 programme to migrate into production. The LRD has passed a successful proof-of-concept on a GE90-size fan rig in the UK. "It worked like a charm," says Mike Benzakein. "We blew one blade off and the unbalanced loads did not go into the rest ofthe engine. Then we ran 4h of windmilling and it did not induce any unusual vibrations." Snecma has also run the LRD on a CFM56-5C, and "it worked very well", he says.

Building on its DAC experience, CFMI has also begun studies of an advanced twin-annular pre-swirl (TAPS) combustor, which is expected to produce half the NOx ofthe conventional SAC at pressure ratios between 24 and 32. The TAPS combustor consists of several sectors, each comprising five combustor cups. The careful design ofthe individual cups and associ-

Specially developed low speed compressor test tool

ated flame tubes is meant to optimise combustion patterns and - using different sectors to perform the same role as the DAC - is aimed at significant improvements in emissions. Noise is also being tackled with tests of a new chevron-shaped jet pipe exhaust. This was expected to reduce acoustic energy by more than 50% and jet noise by up to 3.5dB.

The first full year of Project TECH56 saw component tests of the low speed compressor, high pressure turbine vane and TAPS mixer. Design ofthe high-pressure compressor rig was also completed and assembly has begun, along with that of the high-pressure turbine aerodynamic test rig. Tests of the swept fan blades have also begun, and due to continue into 2000, the final year ofthe initiative. This is expected to culminate in tests of new high-pressure seals, the running ofthe high pressure compressor rig, the TAPS demonstration and dual spool tests.

Baseline reference data for TECHS 6 goals are taken from a specially instrumented CFM56-7B


Into the CFMI's advanced TECH56 development initiative is the key to the company's future as it enters the 21st century

Advanced 3D aerodynamic design is apparent in these TECHS6 high-pressure stators on test

In late 1997, CFMI began marshalling its engineering experts to plot a course into the below for others. CFMI is also aiming for up to opment of a new baseline engine to cover die 21st century. The charter of the resulting 7% lower fuel consumption compared to the 20,000-40,0001bthrustrange(89-178kN).The TECH56 initiative was simple: maintain the CFM56-5B/P, as much as 50% lower emissions upward extension of die tlirust range from that

dominance of the CFM56 family and fend off than late 1990s regulations, and up to 20% covered by the CFM56 represents an interest-competition from IAE as well as Pratt & lower maintenance costs. ing new development. Only the abandoned Whitney. The three-year effort, launched in Although tie V2500 had begun to make sig- CFMXX project, which had been aimed at the early 1998, is aimed at "identifying, determin- nificant in-roads into CFMI's Airbus territory growth A340 models before being superseded ing the feasibility, and validating engine tech- by the late 1990s, the main challenge seems to by GE's new CF6 derivative, had ventured into nology for future market requirements". be P&Ws planned PW8000. This is a geared this higher thrust territory.

"We are the leader and we want to remain die turbofan based on die core of die new PW6000 CFMI executive Bill Clapper says: "We see a leader. We don't want to be caught like Pratt & under development for die Airbus A318. Its period of a couple of years where we need to Whitney was with die JT8D," says CFMIpres- main perceived advantages of much lower invest in diis technology, to demonstrate its fea-ident Gerard Laviec. Key drivers for the next noise, emissions and fuel consumption have sibility and get it ready for us to move", generation CFM engine include a 15-20% spurred CFMI into action. Although no major changes are envisaged lower cost of ownership compared to current The idea was to generate a new menu of tech- "before 2004/5", Clapper adds that "the tech-models and cumulative noise levels 20dB below nologies that could be used either to update die nology to do this is a very major step". Against Stage 3 for the higher thrust versions and 2 7dB later generation engines, or support the devel- this background, CFMI's underlining goal is to

future


Boeing's 737 had helped secure CFMI's emerging dominance of the mid-thrust market like no other programme. The prospect of a successor to the 737 there

fore drove the engine maker to new limits to ensure success on the Next Generation family, which was launched in 1993.

In June thatyear, CFMI was formally selected over IAE by Boeing. The outcome of die fiercely contested engine competition had by no means been a "slam dunk" for CFMI. In the end, the decision was based on a mix of factors. Boeing had suffered the anguish, and cost, of a three-way engine certification effort on the 747-400 and was re-living the experience with the 777. It therefore favoured a sole source.

IAE, on the other hand, had elected not to pursue the programme beyond the point where it was financially viable to sustain the risk. Boeing had sought support for the nonrecurring development costs of the new aircraft and, given CFMI's exclusivity on the "Classic" 737, IAE figured the battle for market share was simply not worth the expense. CFMI, on the other hand, had the weight of its installed base fully behind it and, through its corporate family connections to major leasing customers such as GE Capital, was able to offer a higher level of financial backing, reportedly worth "several hundred million dollars", to support its bid.

Technically, CFMI had a huge mountain to climb. Boeing's performance targets for the aircraft demanded a radical approach. The new 737 not only had to fly faster, higher and further, but it had to achieve these targets with up to 15 % lower maintenance costs, improved reliability, better fuel consumption and 1

tests at Villaroche, paving the way for flight tests early the next year on GE's newly arrived 747 testbed.

By this stage, Boeing had launched all three initial variants of the Next Generation family, and CFMI was committed to full scale development of engines across the full power range. By December 1995, four development engines were on test and early indications showed a 4% thrust margin at hot day conditions up to a maximum thrust rating of 26,3001b. In addition, SFC and exhaust gas temperature margins were better than predicted.

But already the first problems were starting to show. Fan blade-out and medium birdstrike tests showed changes were needed to the Boeing-designed nacelle and duct. The solid titanium blades weighed 35% more than the CFM56-3 blade set, and debris penetrated the casing in some areas. Subsequent concerns over the containment margin and inlet structural stiffness led to a strengthening of the intake. An additional containment collar, extending forward by 3 3 0mm, was bonded on to the existing nacelle fairing to provide extra protection and increased stiffening. Some areas of the containment ring were also thickened and the inlet gearbox casing was strengthened. Cracks in the

ower emissions. The biggest challenge in designing

the new engine was how to increase airflow and substantially reduce SFC The without increasing fan diameter, and therefore compromising the entire underwing installation. The answer was a 1.55m (61in)-diameter wide chord fan, the first of its type used on any CFMI engine. First defined as the CFM56-3XS, the engine was redesignated the CFM5-7B on launch and was rated across a thrust range from 18,5001b (80kN) to 26,3001b. The engine incorporated the CFM56-5B core and low-pressure turbine, in addition to an advanced FADEC, the existing CFM56-3 gearbox, and a new low-pressure compressor.

The CFM56-7B demonstrator engine ran for the first time at Villaroche on 19 September, 1994. The engine ran at 5,540rpm, or roughly 3 % over the red-line speed, on its first test. The success of the initial tests with the wide chord fan gave CFMI high hopes of a smooth development run, but in some ways they were to be disappointed. On 28 April, 1995, the first full-up CFM56-7 engine began

wide chard fan was the biggest external change to the CFMS6-1

exhaust nozzle necessitated a material change from titanium to Inconel.

Inlet manufacturer Rohr (now part of BFGoodrich), worked hard to tackle the problems, most of which began to appear after blade-out, spin pit tests. Stall detection and recovery logic in the FADEC software was also altered after the birdstrike tests, which caused thrust to reduce to 70% before the post-strike stall could be cleared. FAA regulations called for the engine to meet a 7 5 % target.

These changes put an additional strain on the flight test phase, which began at Mojave on 16 January, 1996. By now, the Next Generation 737 orderbook was building at a record rate, and CFMI knew it was vital to minimise the need for any major changes late on. "We want to be right as early as we can because we don't want to be in a position where we have to retrofit a huge amount of aircraft just because we didn't test something right in January 1996," said CFM56

programme manager Bruce Hughes at the time. "When we go into production on this it's going to be like taking a drink from a fire hose," he added. The tests went well, indicating cruise SFC to be around 0.6% better than expected. By the end of tests, SFC was estimated to be around 8% lower than that of the CFM56-3.

More problems were to come, however. The first occurred when the extended containment system was tested. The containment worked well, but around six blades detached instead of the one to two expected. It seemed as if disaster had struck, but a closer look at the test video revealed the problems lay with the axial retainers in the dovetail assembly at the base of each root. These were strengthened to provide around 40% more load capability and subjected to a full whirl test the following August.

More trouble was in store. The tests showed the revised retainer was now too strong, and redistributed the excessive loads into the blade shank. Another blade retainer system redesign was devised which was strong enough to hold the heavier blades, yet flexible enough to prevent the loads travelling into the wrong part of the engine. To the collective relief of Boeing and CFMI, the redesign was validated in November and the engine certificated at the end of 1996,

around two months behind schedule.

"Compliant" CFM56-7 engines had meanwhile been shipped to Seattle for installation on the first 737-700.This finally took to the air for the first time on 9 February, 1997, culminating three years of engine development work involving more than 3,000h and 6,500 cycles in flight and ground tests. The CFM56-7 powered 737-700 was certificated by the FAA on 7 November and entered service with Southwest in January 1998. By this stage, orders had grown to

more than 800, of which 10 were for the newly available 737-900 stretch. Later in 1998, the first -700s, -800s and -600s entered service in Europe, including the first DAC-equipped CFM56-7-powered 737-600s for launch customer SAS. Apart from a few isolated, but notable, incidents, the engine performed well in service.

While tackling problems with a handful of bearing failures, and the hydromechanical control unit, CFMI also concentrated on building up the production rate. By April 1999, orders for the Next Generation 737 stood at 1,13 0 and CFMI has ramped up production to supply 611 CFM5 6-7s by the end of the year. Including the equally dramatic build-up in business at Airbus, the combined tally means CFM56 production will reach a record 1,137 during 1999, guaranteeing that the CFM56 is well on its way to becoming the best selling commercial jet engine in history.


THE CFM56 STORY Next generation - CFM56-7

CFMI ha fight to maintain its exclusivity whe decided to develop the next generatio

^*?^LJ

Lease Finance was for up to 16 aircraft to be powered by the V2SOO-A5 - a sign of the increasingly fierce competition between CFMI and IAE for Airbus business. At the end of July, however, CFMI secured its place on the A3 21 when it was selected by Alitalia and Iberia for a combined total of 28 firm aircraft.

As the A3 21 programme firmed up, it became obvious that the weight and payload requirements for the aircraft would require a higher baseline thrust. The -5B was therefore retargeted for certification at 31,5001b with offerings at two main ratings-a-5Bl at 30,0001b and a-5B2 at 31,0001b. CFMI also saw the chance to sharpen its competitive edge against die V2500, and made the new engine available for the A3 20 from late 1994 onwards. On die smaller aircraft, the engine was to be derated to 2 8,5001b dirust, witli a common A320/A321nacelle and pylon, offering significant take-off performance and Iife-on-wing advantages.

Following the official launch of die A3 21 by Airbus on 24 November, 1989, sales campaigns began in earnest. One of these was for the key

compared to a single annular combustor (SAC). The three-week, 42.25h tests focused on performance and operability characteristics.

Meanwhile, flight tests of die baseline engine on die A321 had began with a "flawless" initial sortie from Hamburg to Toulouse on 28 May, 1993. The engine itself was also certificated that same day by bodi die FAA and DGAC, at three ratings. Fourdi and fifth ratings had by now been defined for the A319, a shortened A3 20 then close to launch by Airbus. The new variants, rated at 22,0001b and 23,0001b (later increased to 23,4001b) were dubbed the -5B4 and -5B6, respectively, and provided CFMI with complete coverage of the entire Airbus narrowbody range.

The A319 was duly launched the following week by Airbus at the Paris air show, with ILFC selecting the CFM56 to power the initial batch of six aircraft. By this stage, die -5Aand -5B had been selected to power 650 firm and option A319s, A320s and A32Is, while a further 300 CFM56-powered A3 20s were in service.

Following a 176 flight, 400h test effort, the

ed by basing its next move on the -5B core. By early 1995, these plans had crystallised into the -5BX advanced technology engine, which CFMI hoped would serve as the basis for bodi die Next Generation 737 series and later A320 family applications.

Component rig tests were under way by mid-1995 and, by the end of die year, evaluations had begun of a new high pressure turbine, hi addition, die -5BX included an improved high pressure compressor and a modified low pressure turbine. Newly honed three-dimensional aerodynamic design techniques were brought to bear for the first time, resulting in reduced losses at die turbine blade tip and root, with consequent efficiency improvements in the high pressure turbine. Collectively, these resulted in a potential 3 % SFC reduction, vital armour in the constant war with IAE and the V2 500

By mid-1995, die -5BX programme was renamed the CFM56-5B/P and certification was targeted for March the following year. Component tests had cleared the way for the start of ground tests in July 1995 and a 40h com-

I "A crucial factor in the selection [of the CFM56-5] was the ability to meet tough new environmental laws sweeping through Europe."

joint procurement of a DC-9 replacement by Swissair and Austrian Airlines. Acrucial factor in the selection was the ability to meet tough new environmental laws sweeping into Europe. Although the high bypass of the CFM56 series ensured ample noise margin, the engine maker knew it could reduce emissions by changing the basic architecture of the combustor.

The resulting double-annular combustor (DAC) was aimed specifically at cutting emissions -particularly oxides of nitrogen (NOx), and proved pivotal in CFMFs subsequent victory over IAE, which was announced in December 1991. Austrian ordered up to 26 A320/321s and Swissair up to 52. The combined value of die deal was worth more than $1.05 billion.

The DAC traced its origins to work performed on both sides of the Atlantic by GE and Snecma. GE's DAC technology stemmed from NASA's Experimental Clean Combustor Programme demonstration engine of the 1970s, while Snecma had run a full DAC combustor in the early 1980s. GE had continued DAC development with work on a compact combustor for NASA5 Energy Efficient Engine After rig tests in 1992, the first full-up DAC ground tests were conducted in Ohio in March 1993. The tests were crucial, as the DAC had to meetits advertised target of reducing emissions of NOx and odier pollutants by as much as 3 5 %

The June 1993 launch of the A319 opened fresh territory for the CFM56.

CFM56-5B was certificated on the A321 in February 1994. Flight tests of the DAC-equipped engine began the following month on GE's 707 testbed in California, paving the way for a three-month flight test effort on the A3 2 0. This began in August 1994, by which time data from the Mojave tests had begun to show the DAC was more than capable of meeting the targets. So much so, in fact, that NOx emissions were subsequently expected to be cut by as much as 45%. In early January 1995 the DAC equipped engine was certified by the DGAC, clearing the way for entry into service with Swissair later that month.

Despite the marketing boost offered by die DAC, and die overall improvements offered by die -5B development, the competitive pressure from IAE continued to mount. CFMI respond-

pliance flight test programme on the A3 20 in early 1996. In August 1995, CFMI and Airbus signed an agreement to offer the -5B/P engine on the A321-200. This was a heavyweight version of the stretch which would eventually reach a maximum take-off weight of 90,000kg (196,2001b), an increase of 6,000kg over the baseline aircraft. Later that month, the -5B flew for the first time on the A319 at Hamburg, Germany. By now, the family was offered

at six ratings ranging from the 22,0001b -5B5 to the 32,0001b -5B3, and was selling well. The engine had logged more than 43,000h widi no in-flight shutdowns, and had a 99.93 % despatch reliability. The only significant problem to occur was the discovery of cracks in some DAC units during inspection by Swissair. This was later corrected. Flight tests of the DAC on die -SB-powered A319 continued at Toulouse, where Airbus tested a mix of engines widi bodi types of combustor. Tests of a -5A-powered A319 were also undertaken for a separate certification programme. The tests passed off smoothly, with Swissair taking the first DAC-equipped CFM56-5B-powered A319 in April 1996. The SAC version of the -5B/P entered service on the A3 20 with Air Jamaica and on the A3 21 with Alitalia later that year.

FLIGHT INTERNATIONAL 19 - 25 May 1999

IHE CFM56 STORY Airbus power - the CFM56-5

The environmentally friendly DAC-equipped CFM56-5 clinched a major order from Swissair

ground runs on 2 7 December. The end of 1989 saw CFMI marking its best

sales year to date, booking 808 firm orders worth $2.8 billion. Total order commitments for 1989 reached 3,100, taking the potential value of the year's business to $12 billion. In all, commitments had been received for more than 9,000 engines since the first order in 1979.

The new A340 engine began flight tests on GE's 707 testbed atMojave on 29 August, 1990, achieving its full 31,2001b thrust rating on the initial sortie. As flight tests continued, CFMI and Airbus worked on plans forthe-5C4, which were solidified the following December when the two signed an agreement to develop the 34,0001b-thrust version.

Compared to the -5C3, which was essentially aFADEC-only change, the-5C4 incorporated several modifications to increase thrust. To reduce exhaust gas temperature and raise low pressure compressor pressure ratio, the fan hub was modified, while new materials were used to raise the temperature capability of the high and low pressure turbine blades and vanes.

The CFM5 6-5 C was certificated at two takeoff ratings - the 31,5001b -5C2 and 32,5001b -5C3 - on 31 December, 1991, shortly after

flight tests had begun on the first A340. The engine was easily distinguished from previous CFM56s by its Rohr-designed long duct nacelle. It was the first use of forced mixing on any CFM56, with 80% mixing effectiveness being measured in flight tests. The first -5C2-powered A340 entered service in February 1993 and, three months later, flight tests began on the first -5C4. The first -5C3-powered aircraft entered service inMay 1994, five months before certification of the -5C4.

While preparations were under way for introduction of the -5C4 powered A340 into service in March 1995, CFMI and Airbus began flight testing a "thrust bump" development aimed at the 2 711 A340-3 00E versions ordered by Singapore Airlines and others. This option, available to the crew as a push-button selection, provided up to 4% more thrust at take-off, and entered service in May 1996.

MORE APPLICATIONS With development of a more powerful CFM56 for the A340, CFMI saw a chance to spread the cost by proposing the same powerplant for another new airframe being studied by Airbus, the A321. First mooted publicly in April 1988,

this was an A3 20 stretched with two fuselage plugs to seat between 185 and 220 passengers. This brought it into the thrust bracket of the CFM56-5C2.

Just when CFMI thought it had found a good fit, it got bad news. Initial figures for the A321-100 revealed a relatively lightweight design, which gave the -5C a disadvantage against the competing V2 500 offering. The -5 A, although briefly considered, was not powerful enough, while the -5C was simply too heavy. The solution turned out to be a new version: the CFM56-5B.

The -5B combined the core of the CFM56-5 C2 with an improved version of the 1.7 3 m fan and four-stage low pressure mrbine of the basic -5A1, enabling it to be enclosed within a nacelle virtually unchanged from that of the -5 A-pow-ered A320. The Airbus supervisory board authorised the marketing go-ahead of the A3 21 in May 1989 and the following month, at the Paris air show, CFMI and Airbus signed an agreement to develop the -5B. The engine was to have an initial rating of 29,0001b thrust, with growth potential to 31,0001b.

The announcement came on 10 June, one day after the first firm commitments for the aircraft were unveiled by Airbus. Unfortunately for CFMI, the order placed by International


were facing the SuperFan and we explained that fundamental reliability hinges on the core. We had enjoyed outstanding reliability on the -3 and - 5, and what Airbus was buying in our offering was the prospect of a ver/ reliable machine from the word go, versus a paper offering with potential problems. We also predicted that the net performance delta was not as large as they promised because, although the SFC of a very high bypass ratio engine is lower, the installed drag is higher," says Homan.

The decision to grow to higher thrust, initially to compete with IAE and later to improve the performance of the A340, was not exactly "rocket science", he adds. "We had done studies to see what could be done and we felt the engine had it in it to grow. There were no great decisions as to whether we should do it or not. The

The uprated -5C4 gave Airbus the option to increase the A340-300 take-offweight to 271,000kg

aircraft needed it and we went ahead and did it," says Homan.

The emerging -5C design was dominated by the larger fan and related changes to the low and high-pressure systems. The low-pressure compressor was increased by one stage to four stages to increase overall pressure ratio. The high-pressure compressor included several changes to increase the airflow and efficiency at high rotational speed. This was mainly aimed at coping with the higher pressure levels and included a new fourth-stage bleed design. The high-pressure turbine featured new materials and some aerodynamic improvements for higher efficiency. The low pressure turbine was similarly modified with improved materials and an extra stage, making five in all. The engine also

had an improved FADEC and a forced exhaust mixer in a long-duct nacelle, which CFMI provided as a complete unit for the first time.

The plan called for certification in October 1991 at 32,5001b thrust, though the initial service rating was planned to be 31,2001b. Airbus had already seen the need for growth, and CFMI promised to develop a 34,0001b-plus version for service from mid-1996 onwards. This would require higher temperature materials in the nozzles and vanes in stages one and two of the high and low pressure turbines.

By the end of November 1989, plans were finalised for a lOOh test programme at Villaroche to be followed by a further test phase at Peebles. In all, the company expected more than 5,000h of tests, which began with the first


THE CFM56 STORY Airbus power - the CFM56-5

was cautious of being wooed by the larger engines into growing the aircraft too large. CFMI successfully persuaded Airbus that it was better to base the design around engines already matched for the 27,500-30,0001b thrust range, rather titan operate bigger and more expensive engines at a de-rated setting. Ironically, its argument was supported by IAE, which again emerged as a competitor with the V2 500.

Following an Airbus supervisory board meeting in January 1986, theTAl 1 officially became the four-engined A340. A parallel twinjet project, formerly die TA9 and renamed the A3 3 0, shared virtually identical characteristics with die exception of the engines. The A3 30, with a range of 10,730km (5,800nm) and capacity for 308 passengers, was to be offered with GE CF6-80C2 or P&W PW4000 engines. The longer range A3 40 was to be capable of carrying 261 passengers over 12,400km.

Throughout 1986, CFMI and IAE worked with Airbus to agree the thrust requirements. In October, CFMI and Airbus signed a memorandum of understanding to develop a "throttle-push" version of die CFM56-5A1, dubbed the -SS2 and rated at 28,6001b. "The commonality with the CFM56-5A1 engine on the A320 will offer airlines unique investment savings," Welsch said at die time. The agreement called for four engines to be delivered in June 1990 for an A3 40 flight test aircraft, with a first flight expected in the second quarter of 1991. Certification of the -5S2 was set for 1990, and aircraft certification for 1992.

In reality, events of the next few years were to be quite different, largely due to the intervention of IAE. The new engine consortium was having a hard time and was desperate to break dirough on to the A340. It had suffered development problems with die V2500-A1 for the A320, including technical issues with the high pressure compressor. IAE had needed to bring

something new to the market to attract business away from CFMI, and had therefore applied "leading edge" technology such as titanium rotors, blades and casings to achieve better performance. Development had proved substantially more difficult than expected.

During die build up to die A3 40 competition, IAE's best offering was a 2 7,5001b-dirust development of the V2 500-A1. Bodi Airbus and IAE knew this was inadequate, even diough it was only slightly below the dirust offered by CFMI. Once more, IAE was forced into attempting to leapfrog CFMI with something new. This time it came up widi a radical geared fan concept dubbed die "SuperFan".

The proposed engine featured a large, ducted fan driven via a gear by die core of a V2500. The SuperFan offered a substantial increase in thrust, to over 30,0001b, as well as dramatically reduced fuel consumption. IAE told Airbus die engine would have an SFC 8-17% lower than that of the V2500 - already proving to be noticeably more fuel efficient than die CFM56.

The SuperFan changed everything. Now Airbus had an engine with which its new aircraft could outperform its closest rival, the planned McDonnell Douglas MD-11. It had begun to be obvious to Airbus diat the US tri-jet could easily beat die A340'spayload and range. Worse still, die A3 40 engine choice at diat stage meant diat, from the outset, there was virtually no room for growth. Based on the SuperFan, Airbus went back to die drawing board.

The A3 40 was redefined into two basic options: a 262-seat -200 with a range of 14,500km; and a 4.3m-stretched version, die -300, capable of carrying 295 over 13,000km. The new-lookA340 immediately began attracting interest from airlines and, widiin weeks of die redesign, Lufthansa became die first to commit to the A3 40 when, on 15 January, 1987, its supervisory board approved die purchase of up to 30 SuperFan-powered aircraft.

The second commitment came on 31 March, when Northwest signed a letter of intent for up to 30 A340s. But no choice of engines was announced and, just a week later, die reason was revealed to a stunned industry.

In early April, an Airbus team headed by president Jean Pierson flew to P&Ws headquarters in East Hartford, Connecticut, believing it was to sign a definitive agreement on die SuperFan. Instead, to die shock of die Airbus team, P&W informed diem die SuperFan was dead and that IAE had been premature in its offer to develop the engine. Dumbfounded, Pierson asked what P&Ws suggested solution was - diere is no record of the response he or his team made to its proposal that die PW2000 should be offered in place of die SuperFan.

CFMI, meanwhile, been working furiously on a revised proposal of its own. "Airbus was floundering around looking for an engine, so I

CFM56-S reliability was 99.8% by 1999

gave Jean Pierson a ring and said: 'Look, we will agree to build an engine widi a bigger fan diameter'. So we cut a deal widi Pierson on the A340 and made the French Government very pleased because it saved die programme," recalls Rowe. The deal was linked to a proposed agreement giving GE semi-exclusivity on the A330. "We told diem we would build it if they agreed to let us put the CF6 on die first 100 aircraft, and if diey put the new engine [later to become the -5C1] on the stretched A320," says Rowe. Indie event, CFMI and Airbus signed an agreement on an "improved engine" for the A3 40 and, at the same time, concluded a separate deal to develop a uprated version of the CF6-80C2 for theA330.

"We decided 68in [fan diameter] was no longer adequate and we jumped to 72.3in. We


which culminated in European certification on 26 February, 1998.

One of the most nerve-wracking moments for CFMI in February 1987, when the -5 passed the medium bird ingestion test at Snecma's Corbeil site in France. With fan speed set at 4,900rpm, slightly higher than required for the engine's 25,0001b thrust rating, seven birds weighing more than 0.7kg (1.51b) each were fired into the fan in less than Is. To CFMI's relief, the engine continued running without barely a murmur and maintained 98% thrust, compared to the 75% certification target. With critical birdstrike and blade-off tests completed, 5,000h of test time and three years of development behind it, the CFM56-5A1 received joint USAFrench certification on 27 August, 1987.

On 28 March, 1988, the first CFM56-pow-ered A320 was delivered to Air France. Recognising the significance of the event, Jean Billien, then president and chief executive of CFMI said: "It has taken us 14 years to reach this historic milestone in the CFM56 programme. When Snecma and GE reached agreement to develop the' 10t' engine, the initial studies were concentrated on 150-seat aircraft from Dassault and Aerospatiale, with similar studies by Boeing. Such an application was considered key to our decision to proceed with the development of the CFM56. In 1978, when the Airbus partners decided to proceed with the 200-seat A310 ratlier than a 150-seat aircraft, we had to seek out our success elsewhere. Needless to say, that success has been achieved at a level far beyond our initial projections".

Billien was right. Just a week before the A320 delivery ceremony, CFM5 6 orders had reached 3,900 from almost 100 operators. In the previous year alone, CFMI had taken orders for 93 6 engines and the orderbook was valued at about S10 billion.

ENTER THE A340 With four-engined airliner programmes all but non-existent by die 1980s, it was with more tJian passing interest that CFMI watched Airbus finalise its plans for a project dubbed die TA11. This had emerged from die A300B11 growtli study of the 1970s and, by 1980, had been refined into a 220-passenger, DC-10/L-1011 replacement called die TA (twin aisle) 11. Configurations were considered with three Rolls-Royce RB211-535s or Pratt & Whitney JT10D-2 32s and, although these came to nodi-ing, die option of larger engines was a worry to CFMI, which had been provisionally baselined on the design from the start of the A300B11 studies in the early 1970s.

By 1983 the threat emerged once more as bodi R-R and P& W offered derated versions of the RB211 -5 3 5 and P W2 000, respectively. The TA11 was offered with two fuselage lengths, seating either 2 3 0 or 2 70 passengers, but Airbus

CFMI spread its wings farther on the A340



/tmw^''

board and came back to Airbus with a significantly improved design, the CFM56-5. Rated at 22,000-26,5001b thrust, it retained the FADEC and fan diameter of die -4, but little else. The changes produced an overall SFC improvement of 8%, achieved mainly through a totally new fan. Designed by Snecma using three-dimen-sional computational fluid dynamics techniques, die fan offered improved efficiency well into the climb. For a short-to-medium range aircraft like the A320, this made all the difference and, combined with a new high efficiency booster design, brought the CFMS6 back into line with the V2500.

Other changes were made within the engine's core, in design of the ninth-stage high pressure compressor and associated bleed. Clearances, seals and liners were also optimised, as were the control schedules for the variable stator vanes. The combustor was tweaked, with the discharge flowpath being opened slightly to match the new turbines, staged combustion to match the reduced fuel/air ratios of the -5, and reduced pressure losses. The high and low pressure turbines were also modified significantly to

By 1999 the CFM56 had been sleeted to power •more than 1,950 A320 "family"aircraft

improve the flowpath, but retained the same number of stages as the CFM56-2/3 (one plus four). This latter effort, together with improved clearance control in both turbines, ended up as a major part of the -5 A performance improvement package.

As CFMI began to work with Airbus, it became obvious that the Toulouse staff preferred to work with GE engineers rather than their French counterparts. "GE had a very good team at Airbus, which did not want to work with Snecma at first," recalls Rowe. "We had to convince Snecma that GE should be the leader for the Airbus programmes, and that, for some reason, Airbus preferred it that way. We eventually got through that and now we have an equally balanced team."

Aside from this early difficulty, the programme progressed well. Extensive component tests in 1985 revealed "4-5% better fuel burn performance than we've offered," said Ron Welsch, then the newly appointed vice-president of marketing and advanced applications for CFMI. The first complete -5 was due to begin tests in early 1986, with service entry planned for the first quarter of 1988. By this stage, 160 engines were on firm order for the A3 20.

The -5 FADEC had been tried out successfully on a CFM56-2 test engine at Evendale in late 1985, and CFMI predicted a 2% performance improvement over a comparable engine with standard hydromechanical control. With tests going better than expected, CFMI advanced its schedule and began full-scale tests of a prototype CFM56-5 with FADEC in November 1985 at Villaroche. The first production representative -5 began tests injanuary 1986 and, within a week of its first run, had exceeded its full nominal rating of 25,0001b thrust. By this stage, three test engines had accumulated more than 150h on test and Jacques Renvier, then Snecma's chief engineer on the CFM56, said results were "most encouraging. The engine has demonstrated its ability to start smoothly and rapidly. Vibration levels are low at all operating levels".

The engine first took to the air on 2 6 August, 1986, mounted under the wing of GE's 707 test-bed. The lh 23min flight from Mojave tested the overall operability of the engine and its FADEC system to an altitude of 37,000ft (11,290m) and a top speed of Mach 0.85.

The first two A3 20 flight test engines were being shipped to Toulouse when, in October 1986, CFMI received some of the best news it had heard in a long time. Northwest Airlines had ordered 100 A3 20s, and gave every indication that it would chose the CFM56 - which it later did. The order consisted of an initial commitment for 10, but by Farnborough 1990 the airline had confirmed the entire 100.

Powered by the CFM56-5A1, as it was now called, the A3 20 made its first flight on 22 February, 1987, and Airbus embarked on a rigorous four-aircraft, l,200h test programme


added fuel economy, idle speeds will be reduced for descent/taxi from CFM56-2 levels, and a full authority digital electronic control [FADEC] will be installed," CFMI said at the time. Target service-entry date was then early 1988, with engine tests due to begin in mid-1984. Certification at 25,0001b thrust (1 lOkN) was scheduled for October 1986.

Although die first commitment for the A3 2 0 came in June 1981 when Air France signed a letter of intent for 50 aircraft, it was not until March 1984 that the complex financing for the project was in hand and the aircraft was officially launched. Airbus ordered 80 CFM56-4s, with deliveries to customers to start in 198 8. "It puts us in die enviable position of being first on anodier new aircraft," CFMI president Jacques Chausse said at the time. However, die competition again forced CFMI to change tack.

This time, the opposition came in the form of International Aero Engines (IAE), which had been created to develop the similarly sized V2 500. Although CFMI had time in hand and a year's head start on the V2500, it realised that the later development timescale gave its competitor an inevitable technical advantage. IAE played its cards on the V2500's fuel consumption, performance retention and noise reduc

tion benefits to their maximum. The effect was almost immediate, as recalled by CFMI vice-president Frank Homan. "We soon judged that the CFM5 6-4 was not going to be good enough. It did not have enough power or SFC [specific fuel consumption] performance."

Shortly after the V2500 was offered on the A3 2 0, in September 1984, Pan Am signed a let

ter of intentforup to 50 aircraft. Significantly, it did not make an engine selection and only two months later Cyprus Airways became the first airline to choose the V2500 on the A320. The followingjanuary, Pan Am announced its intent to follow suit and select the V2 500. The writing was clearly on the wall for CFMI.

The company went back to the drawing

CFMPs development of the CFMS6-Sfor the A320 had later benefits for the A3 40 -powered by the -SC, the first example of which is pictured on GE's 707 testbed



Airbus debut The CFM56 proved to be a pivotal powerplant for a new family of Airbus products

For many years, Snecma and the French Government shared the dream of a new generation "French" airliner powered by a domestically developed engine. It was

through the CFM56-powered Airbus A320, therefore, that their combined aspirations came closest to fruition.

Airbus essentially inherited the A3 20 programme through its partners' efforts on the Joint European Transport (JET) programme. By 1980, this had metamorphosed into Airbus' SA (single aisle) 1, SA 2 and SA 3 designs and, by early 1981, the project had assumed an Airbus identity as the A3 2 0. The European consortium chose to focus initially on the SA 1, as the A3 2 0-100, and the SA2, as the A320-200.

The two offerings were perfectly sized for the CFM56. The smaller -100 was to be 36m

(118.25ft) long, with capacity for 130-140 passengers. The -200 was to be 39.3m long, with seating for 150-160. Both were competing with Boeing's newly announced 737-300 and a proposed 150-seater frornMcDonnell Douglas and Fokker, the MDF-100 - a successor to the F29.

"Airbus decided it wanted to start a twin, and of course we thought the ideal engine was the CFM56-2," says Brian Rowe. "They wanted more performance than we could give, and Rolls-Royce was offering the RJ500, which they claimed would offer up to 7% better fuel performance. The thing that really upset me was that there was no way it could be as good as that, but it forced us to redesign a new engine for our bid." The new engine was the CFM56-4.

The -4 featured a new 1.73m-diameter fan and matching' low-pressure compressor. "For


topped 1,980 - the -300 version alone accounting for 1,107

record levels and the orderbook full, an incident near New Orleans, Louisiana, suddenly grabbed the headlines. On24May,a737-300of El Salvador airline TACA, with 41 passengers and crew on board, suffered a double engine flame-out while descending through severe hail and rain. The engines refused to relight and the crew pulled off a spectacular dead-stick landing on a flat, grassy area beside a bayou. All aboard were safe, and the aircraft was virtually undamaged, but it had been a close thing. A similar incident had occured the previous August, when an Air Europe aircraft descending

tion. It even flew a CFM56-3 on GE's 707 test-bed behind a KC-135 tanker which squirted water, instead of fuel, from its flying boom directly into the intake of the CFM56. These trails were completed in early 1989 and the solution was a sensor that automatically triggered the combustion chamber burners into continuous ignition. The flow path and spinner shapes were also optimised to divert more hail and rain away from the core.

Another problem which CFMI tackled the following year followed thejanuary 1989 crash of a British Midland 7 3 7-400 at Kegworth in the

the modification, Rossignol said: "The bill is extremely large...but we will pay for it. It is still difficult to tell what happened; the reasons for the failure are still not clear." CFMI did work out, however, that the coupled mode which caused the destructive blade flutter occurred between "the fan wheel and disc at mid-altitude around 2 5,000ft".

The revised blades were fitted to all -3B2 and -3C1 engines from mid-1990. At that time, almost 6,900 CFM56 engines were on order, of which 3,750 were -3 models. Of these, nearly 1,800 had been delivered, representing a signif-

First of the many. Sales of CFMS 6-powered 727s exceeded 3,100 by 1999

between Skiathos and Salonica got caught in heavy hail with both engines at flight idle. The engines flamed out, but the crew managed to relight them having lost 3,000ft in altitude.

The FAA issued an airworthiness directive instructing crews not to reduce thrust setting below 45 % Nj (low pressure spool speed) in severe hail and rain or sleet. CFMI began an extensive series of tests to determine engine performance with high levels of water inges-

UK. Vibration had been detected in one of the engines and this was later traced to the separation of a fan blade outer panel. This led to a power level restriction on -3C1 engines until a cause had been discovered, and hampered some operations until modifications were devised and approved. The modification involved recam-bering the blades by increasing the part-span shroud angle from 27° to 37°.

Speaking at the time of the introduction of

icant modification programme for CFMI. Despite the problems, the CFM56-3 was an

extraordinarily successful engine, becoming one of the fastest selling commercial power-plants in history by 1989. Within five years of launch, the engine had been selected by 72 airlines to power more than 1,100 aircraft on firm order.

Few would have guessed, even among the most optimistic CFMI enthusiasts, that this record could be topped within a decade.


THE CFMS6 STORY Boeing breakthrough - the CFM56-3

CFMI aired early problems with the CFM56-5, such as susceptibility to hail and ice ingestion, restoring confidence in the 131. Sales of the "Classic" later

The CFM56-3 passed the redesign hurdle, however, and received joint FAA/DGAC certification on 12 January, 1984. The milestone coincided with the appointment of Jacque Chausse as president and chief executive, succeeding Rossignol. Chausse had been general manager of FAMAT (Fabrications Mecaniques de l'Atlantique), a start-up GE-Snecma satellite manufacturing plant, but had spent most of his

l,300h of flight tests. The first aircraft for the two launch customers, USAir and Southwest, were handed over on 28 and 30 November, respectively. Southwest innaugurated services with die type between Dallas and Houston on 7 December, while USAir introduced the 737-3 00 on to its Pittsburgh-Harrisburg route on 18 December. The most important new chapter in the history of the CFM56 had begun.

from Piedmont Airlines. The -3 B1 turbofan, by this time, was exhibiting an impressive 99.9% dispatch reliability rate and acheiving an engine-caused shop visit rate of just 0.043 per 1,000 engine hours.

The 737-400 featured a 3.05m fuselage stretch and offered a series of higher optional maximum take-off weights (MTOW) up to 150,0001b. Aircraft above 142,5001b MTOW

I "They were used to hearing about the early JT9D. They thought it would suck up garbage cans and flame out in crosswinds... it turned out to be a real workhorse" - Borge Boeskov

Snecma career in manufacturing and production positions at the Paris and Corbeil sites.

On 24 February, 1984, the 737-300 made its maiden flight from Boeing's Renton site and flew for 2h 56min before landing at the nearby Boeing Field. Jim iVIcRoberts, chief test pilot for the 737-300, said the aircraft was "quiet, smart and lovely to fly" and commented that it handled in a similar way to the -200.

While flight tests continued, CFMI worked on the certification of die higher thrust -3B2 model rated at 22,0001b. This incorporated adjustments to the low-pressure turbine speeds and modifications of the power management and main engine controls. The first customer for the higher thrust engine was Pakistan International Airlines, one of the batch of new-customers that had helped firm orders creep up to 124 since roll-out. Some of the new business was attracted by the early performance figures for die new airframe-engine combination, which was showing 21-25% fuel consumption advantage over the JT8D-powered 73 7-200. In addition, CFMI was promoting the expected reliability of the new engine based on the figures coming in from the CFM5 6-powered DC-8 fleet. Chausse commented at the time that "after more than two-and-a-half years of revenue service, the CFM56-2 has flown more than 1 million flight hours and is setting precedents for remarkable reliability, with a dispatch reliability rate of 99.9%".

The -3B2 variant was certified by the FAA and DGAC in July 1984 as flight testing continued on the 737-300. The aircraft received type certification on 14 November, after almost

By 1985 Boeing's studies into an all-new 150-seater, the 7-7, had been dropped in favour of another derivative of the 737. This stretched version was to be called the 737-400 and, according to Boeing at the time, was an easy development to complement the -300 and "...tide the airlines over until the 1992 aircraft [the 7J7]". However, the 7J7 was also dropped eventually and the 737-400 effectively took its place in the Boeing line-up.

Later thatyear, CFMI and Boeing got a foretaste of the phenomenal success that was to come with the CFM56-powered 737, when United ordered 110 -300s. Later that month, CFMI celebrated the delivery of the 1,000th engine - an event that many at both GE and Snecma thought on many occassions would never happen. Chausse said at the time: "To see the 1,000th CFM56 engine leave the factory gives me and all my associates in GE and in Snecma a real sense of achievement. Yet, it is perhaps more exciting to realise diat the 1,000di engine milestone is still far from the midpoint of the production programme. CFAII currently has booked close to 2,300 engine orders, with a strong sense of optimism tiiat diere are many more to come."

Significantly, the 1,000th engine was a CFM56-3, production of a record 325 of which was planned in 1986. The milestone coincided with logging of the 2 millionth engine flight hour, more than 1.8 million of which had been amassed on the CFM56-2. Further increases in production and hours became certain when Boeing launched launched the 737-400 on 4 June, 1986, with an order and options for 55

required strengthening to the overwing body structure, keel beam and wing leading edge, and beefed-up wheels and brakes. These resulted in a requirement for more engine thrust, so CFMI quickly adapted a new derivative, the -3 C1 rated at 2 3,5001b dirust. The engine later became the standard 737 "Classic" powerplant on die -3 00, -400 and shorter -500, which was launched on 20 May, 1987, on the back of orders and options for 73 aircraft from four airlines.

In 1988, with production ramping up to


the -2 blades. The -3 also featured a new low-pressure compressor, still with three stages, and side-mounted the gearbox on die fan casing to maximise ground clearance. This emphasised the squat appearance of the engine from the front, leading some to compare it to a hamster with bulging cheeks.

Several improvements were also introduced into the core and the low-pressure turbine. These included sealed high-pressure compressor dovetails, dual "cone" fuel nozzles, new materials for the high-pressure turbine blades to reduce cooling flow requirements, and reduced low-pressure turbine case cooling. Alost of these improvements were introduced into production CFM56-2s in early 1984.

By early March 1981, two key airlines were ready to launch the new 737. USAir announced its intention to buy 10 on 5 March, and 13 days

was increased to 41.66ft. Borge Boeskov, now president of Boeing Business Jets, was Boeing's Southwest representative at the time and recalls die airline's suspicion of the new engine.

SOUTHWEST SUSPICION "They were used to hearing about the early JT9D. They thought it would suck up garbage cans and flame out in crosswinds. We warned them: 'This is not the same'. But the main reason they bought it was noise at Love Field [the airline's Dallas base]. They knew noise was going to kill them, and Herb Kelleher had to do something. They expected real problems with the engine, and we were cautious. So when they finally got it they said: 'What the hell are you guys on about?' It turned out to be a real workhorse."

Assembly of the CFM56-3 quickly accelerat-

I "Boeing's Joe Sutter and his team came down... They said: We'd h and they unrolled these drawings of something they called the 737-

The CFM56 transformed the sales success of two generations of the Boeing 131 family

I "Boeing's Joe Sutter and his team came down... They said: We'd like to show you something' -and they unrolled these drawings of something they called the 737-300" - Dick Smith

later Southwest Airlines unveiled plans to place orders and options for 20. Crucially, both also selected the CFM56 over the RJ500, marking the start of die US-French engine's legendary association with the Boeing twinjet. The RJS00 remained an option for some time, but quietly dropped from the scene by mid-1981, when the CFM56 had achieved defacto exclusivity.

Boeing formally launched the 737-300 programme on 26 March, 1981 with expectations of first deliveries in late 1984. In line with USAir and Southwest's requirements, the -300 was configured to seat between 121 and 149, and was stretched by 2.6m to accommodate the extra seats. The wing was strengthened to carry the heavier CFM56s and fitted with a revised leading-edge slat outboard of die nacelles, modified trailing-edge flaps and flaptrack fairings. An additional ground spoiler and increased dorsal fin area improved lateral stability and engine-out performance. Wingtip extensions increased span to 310m, while die tailplane span

ed and, on 31 March, 1982, the initial test engine was fired up at Evendale for the first time. During tests, the engine exceeded its takeoff rating of 2 0,0001b thrust by 5 %, opening the immediate prospect of further growth within the family.

Ultimately this growth capability would lead to development of three basic CFMS6-3 models; a 20,0001b -3B1; a 22,0001b -3B2; and a 2 3,5001b -3C1, which could be derated to either ofdiese thrust ratings. Rossignol, president and chief executive of CFMI since taking over from Malroux the previous October, said prophetically at the time: "We believe initial tests set the pace for continued success throughout the CFM56-3 programme."

The second of the five engines in the development programme was assembled by Snecma at Villaroche, and was delivered to the Saclay test site for the start oftest runs in June 1982. By this stage the first engine had amassed almost 1 SOh of running time and was producing "excel

lent overall performance," said CFMI. The third development engine was scheduled to begin tests in July 1982 and confidence in the programme was growing.

In fact the major worry at the time was not the engine, but die aircraft as a whole. No further orders were clinched in 1981 and only one additional sale, an order for five from UK charter airline Orion Airways, was achieved in the whole of 1982. The situation hardly changed throughout 1983, with only 25 firm new sales being announced during the year bringing total firm orders to 50 - an uncomfortably slow start for a new Boeing programme.

The CFM56-3 took to the air for the first time on 25 February, 1983, mounted in the left inboard position on GE's newly acquired 707 testbed. The flight, from Mojave, California, lasted 49min with the engine perfoming "excel

lently" throughout, according to Rossignol. An engine at Villaroche, meanwhile, began an exhaustive, 150h block test introduced as a joint FAA/DGAC certification requirement. The block test was an endurance evaluation at red-line temperatures (conditions not to be exceeded during the most rigorous engine operation), and the engine completed the test successfully by mid-year.

BIRDSTRIKE PROBLEM A sixth engine successfully passed hailstone ingestion certification tests at Saclay on 27 July and went on to perform medium bird ingestion tests. It was here that problems cropped up. The bird tests produced bad results with significant damage to the blades and retention devices as well as die containment system. The resulting redesign of the blade retention configuration and containment collar, and associated verification tests, put a big dent in the development budget as well slipping the certification schedule from September 1983 to early 1984.


Snecma's Ravaud is credited with the original idea of re-enginging die Boeing 737, while GE's Neumann was "very sceptical", recalls Rossignol. "But Ravaud and T Wilson were really good friends. Wilson began to come around to Ravaud's idea, which he worked out could be amortised with only 2 00 aircraft, since tJiere were no changes to die aircraft itself, only around the engine installation."

The timing of die unsolicited offer could not have been better for Boeing, which was in die midst of studies to develop new 727 and 737 derivatives. The most promising of these was a study aircraft called the 737-300, details of which were revealed to an unsuspecting CFMI early in 1980. "That was a stunner. We didn't expect it all," says Smith. "Boeing's Joe Sutter and his team came down and we went through the all the programme with them. Then they said: 'We'd like to show you something'. And diey unrolled these drawings of something they called the 737-300."

LOOKING GOOD The economics and performance of the CFM56-powered 737 looked right from the start, particularly to airlines like USAir, which had specifically asked Boeing to look at a 737 growth model. According to Smidi, the biggest problem was persuading everyone that the higher bypass ratio engine could fit under die 737's low wing. "A lot of people said a 60in fan would hit die ground. But Boeing solved diat by flattening die inlet." The curiously contoured nacelle of the CFM56-3 remains a hallmark of the 737 and, despite the unusual shape of the lower half of the inlet, was found to have a beneficial effect on engine performance. To ensure adequate clearance, the powerplant was canti-levered forward of the wing, rather than being mounted underneath.

By the 1980 Farnborough show, Boeing was publicly discussing the737-300 which, by then, had attracted die interest of Rolls-Royce. The engine maker was proposing a powerplant called the RJ500. This was under joint study widi Japanese Aero Engines (a combination of the three Japanese engine manufacturers), and had evolved from the earlier RB43 2. Unlike the CFM56, the RJS00 was a paper engine and never lived long. It did provide, however, part of the genetic code for an engine that was later to emerge as the first serious competitor to the CFMS6. In March 1983, Rolls-Royce and the Japanese group stopped work on the RJ500 and formed International Aero Engines (IAE), in partnership with Pratt & Whitney, MTU and Fiat, to develop the V2500.

CFMI designated the new variant of its engine the CFM56-3 and decided to incorporate a totally new fan. The blades were scaled-down versions of the recently developed CF6-80A fan, rather than clipped derivatives of

High bypass power boosted the 737's performance


150/160-seat airliner. Rather than stretch the F2 8, the Fokker study group opted for a larger, six-abreast fuselage cross-section, and came up with a design that bears close resemblance to the A3 2 0 of later years. The main difference was the familiar Fokker T-tail, which was taken straight from the F28 and grafted on to the fuselage of what was dubbed the F29. Although talks were held with all the US airframers, these came to nothing and by 1980 Fokker was in serious negotiations with the Japanese.

Rolls-Royce, which had earlier proposed the RB401 for the raft of regional and short-haul aircraft then under study, defined a scaled-up version rated at 21,0001b thrust (95kN) for the F29, called the RB432. Due to the potential involvement of the Japanese, Rolls-Royce signed an agreement to develop the engine with

On Boeing's 737-300, CFM1 finally achieved its original target - a twin-engined application

Ishikawajima-Harima, Kawasaki and Mitsubishi Heavy Industries. The F29 proposal also fell right into the thrust bracket of the CFMS6. However, the smaller scale and low-slung, underwing, engine location on the proposed twinjet meant that the CFM56-2 was too large to give adequate ground clearance.

With no othersuitable projects apparentlyso close to reality, CFMI knew it had to go all-out to meet the F29 requirement. This was only possible if it reduced the CFM56's fan diameter - an expensive gamble given the relatively fragile market base on which the future of the engine depended at the time. "It was a big decision, and it was led all the way by Brian Rowe," recalls Smith. Rowe was by then group executive in charge of GE Aircraft Engines.

Under the codename DR-18, several studies were carried out into fans as small as 1.39m (5 Sin) and 1.5m. The 1.5m fan produced better overall results, and was eventually selected when it became apparent that the Fokker studies were heading nowhere.

The small fan theoretically reduced bypass ratio from 6 to around 5.1, while airflow-dropped from 375kg/s (8271b/s) on the CFM56-2 to 296kg/s on the study engine. The smaller overall size of the engine, the fan and its containment collar also reduced weight by almost 170kg, while thrust was estimated at around 20,0001b. "Well, nothing really happened at Fokker, but we used the estimated performance data and sent it unsolicited to Boeing - they didn't ask for it," says Smith.


THE CFM56 STORY Boeing breakthrough - the CFM56-3

Twinjet Transition

With two four-engined transports under its belt, the CFM56 was still looking for its original niche - a twinjet. Then along came Boeing with the 737-300

By late 1979, against die odds and all its predictions, CFMI found itself on board two four-engined, long-range transports rather than any short-to-medium range

twins. "That's the crazy part," says Dick Smith. "We were always looking for applications on an aircraft that would replace the 727, and everything else was incremental. So what we suddenly seemed to be ending up with was just the incremental."

As it turned out, the "incremental" market was pivotal both to the survival of the CFM56 and to the long-term profitability of the programme. Although the 707-700 initiative failed to prolongthe life of Boeing's first jet-powered airliner, it provided a valuable bridgehead to military programmes based on 707 derivatives. In the springof 1980, the CFM56-powered 707 completed a 15-day demonstration tour of 19 US Air Force and Air National Guard bases, and Pentagon planners were shown that much of the basic design work required for the KC-13S re-engining project had already been accomplished, and paid for. The tour followed the award to Boeing, on 22 January, 1980, of a $13.65 million study contract to evaluate the KC-13 5RE, as it was initially called.

Beginning with the KC-135R re-engining effort, the CFM56-2 engine quickly became the lead powerplant for all military versions. The USAF awarded Boeing a S12 9 million contract on 31 March, 1981, to re-engine one KC-13 5 A with CFM56s, and on 29 December, 1981, President Ronald Reagan signed the 1982 US defence budget, which authorised re-engining of the first batch of nine tankers.

Today, 428 KC-135s have been re-engined with the F108 (the military designation for the CFM56-2), while other applications have included the E-3 Airborne Warning and Control System (AWACS) aircraft of several international operators, the US Navy's E-6 strategic communications fleet and the USAF's RC-135 Rivet Joint electronic intelligence air

craft. CFiVlI also hopes to re-engine the USAF's remaining 112 KC-135Etankers,its33-strong AWACS force and up to 13 707-based Northrop Grumman E-8 Joint Surveillance Target Attack Radar System aircraft.

CFMI's critical transition to twinjets, and the applications for which its engine was always designed, came not through any direct links widi Boeing, however, but rather through studies for Fokker-VFW

The Netherlands manufacturer, under its new chairman Frans Swarttouw, had embarked on a major market study in the late 1970s to evaluate future airline requirements. Not surprisingly, in common with just about every other commercial aircraft manufacturer, Fokker came to the conclusion that die biggest demand was for a new short-to-medium range


John Marsden AMRAeS

© Reed Business Information 1999

1 Removable leading edge 2 Pylon upper spar 3 Lower spar

Front engine-mounting trunnion and torque [inks Pylon/engine truss Rear engine-mounting support and links Fire seal

8 Front wingspar/pylon fitting 9 Rear pylon/wing attachment 10 Rear fairing 11 Inspection panels 12 Stage five bleed air

to airframe 13 High pressure (HP)

compressor discharge bleed air

14 Bleed air for active clearance control of HP turbine casing

15 Low pressure turbine casing cooled by bypass flow through piccolo tubes

16 Bleed air cooler 17 Ram air for cooler 18 Ram air exit 19 Air start supply line 20 Electric junction box 21 Acoustically lined intake 22 Hot air de-iced leading edge 23 Oil tank and accessories 24 Thrust reverser cascades 25 Thrust reverser blocker door 26 Translating cowl 27 Engine vent-air outlet 28 Bleed valve actuator 29 Variable-stator actuator

fore came as a pleasant surprise to CFMI. In May 1979, Grumman Aerospace was con

tracted to make die pylons and nacelles, which were interchangeable rattier man being "handed" (left or right) like the original JT3 Ds. MDC agreed to provide engineering support to Cammacorp for the conversion work, to be carried out at die company's Tulsa, Oklahoma, site. In the end, conversions were also carried out in Atlanta as well as Canada and France.

Meanwhile, preparations continued at Boeing for the first flight of the "707-700", which took place on 27 November, 1979. The flight capped a tumultuous month for CFMI which, on 8 November, had received simultaneous certification for the engine from the FAA and French DGAC.

The DC-8 activity ramped up rapidly diroughout 1980, with more airlines adding to the orderbook. In May 1980, CFMI initiated production plans for the programme, which had been renamed the DC-8 "Series 70". The first DC-8-61 was delivered to Tulsa by United on 30 September, 1980, and made its first flight just under a year later, on 15 August, 1981. An MDC test crew, headed by chief test pilot Phil Battaglia, took the aircraft aloft for more than 5h. "I've logged a lot of DC-8 hours, and I'd call

I UI can hear the vultures flying around this building.Ifwe don H get the United order tomorrow, we're dropping the whole thing," Rossignal quoting Ravaul

airline had decided to go for the CFM56, and notthecompetingJT8D-200.TheCFM56was finally "safe", just two weeks before the internal deadline set by GE and Snecma effectively to freeze the whole programme. "We needed a minimum of 75 aircraft, and I think we settled for 60. We estimated the total [DC-8] market was around 150, though ultimately we did 110 aircraft," says Smith.

United ordered 29 DC-8-61 retrofits, Delta 13 and Flying Tigers a total of 18, 16 of which were -63F freighters. Other big customers were to follow in the next two years.

It emerged just how close United had come to going for the P&W option. The airline had

initially favoured the JT8D conversion, which was priced at $980,000 per engine, compared to $1.5 million for the CFM56. Flying Tigers had performed the bulk of the engineering assessment and, on hearing about United's preference for the JT8D, asked it to reconsider. It told United that the CFM56-powered DC-8 would not require "cut back" procedures on climb out to meet the noise requirements of some local airports, whereas the JT8D version probably would. It also stressed that the long term fuel savings of the CFM56 would eventually more dian compensate for the higher conversion cost. United's $400 million order to re-engine it 29-strong DC-8-61 fleetwith CFM56-ls there-

CFM estimated it needed to carry out a minimum of IS DCS retrofits - it eventually completed 110 commercial service.

this one of the best flights in my experience -particularly since it was a first flight," said Battaglia, who added: "This was a full shakedown of a new series aircraft."

Some production "inefficiencies" slowed die pace of the Series 70 programme, and CFMI found itself in the thick of organising alternative conversion sites to get back on schedule. One site chosen was Delta's overhaul and maintenance base at Atlanta, Georgia, where 44 of the 110 aircraft were eventually modified. It was from Atlanta that a Delta DC-8-71 made the first commercial flight of the CFM56-powered derivative on 24 April, 1982, operating a scheduled service to Savannah, Georgia. After almost 11 years, the CFM56 engine had finally entered


THE CFM56 STORY Taking flight - the CFM56-2

beginning, having been a negotiator of the preliminary Snecma/GE agreement in 1971, and later Snecma's CFM56 programme manager. Malroux's last job at Snecma before moving to CFMI was as marketing general manager for commercial engines. CFM56 programme managerjean Bagneux took over this task, while he was replaced, in turn, by Jacques Rossignol. The same transition phase also saw Dick Smith replace Jack Hope.

By die time the new team was in place testing had built up to almost 3,000h on seven development engines. The two flight test engines had, by this stage, built 172h in 83 flights and five other development engines were under construction with 10 more released for production. Some of these were allocated to the 707 flight test effort and were due to be shipped to Seattle in the first quarter of 1979.

It was a chance get-together at a remote site in Wyoming that September, however, which was to prove vital to the long term success of the CFM56. At that year's "Conquistadors" meeting - an informal gathering of industry leaders, the concept of DC-8 re-engining was mooted for the first time. "It was an idea cooked up in an offsite meeting with Jackson McGowen and Brian Rowe," says Smith. "They basically said: 'Hey, how about re-engining DC-8s?'"

Like the 707, the DC-8 faced the threat of upcoming noise legislation, yet the long life expectancy of the model coupled with its popularity with several major operators like Delta and United made it vital that a solution was found. The US Federal Aviation Administration was preparing a rule that all four-engined jets had to meet Stage 2 noise limits by 1985. In addition, rising fuel costs were making it hard for JT3D-powered aircraft like the DC-8 to be competitive against new generation widebod-ies equipped with high bypass engines.

"I went out to Los Angeles to meet McGowen in early 1978 and see what kind of aircraft it would make, and what kind of market there might be," recalls Smith. McGowen, a former president of Douglas Aircraft, had formed Cammacorp, a consultancy to manage lease programmes, and could draw on the resources of an army of former Douglas engineers who had worked on the DC-8. "We basically decided to go out and test the water. We met with CFMI, which authorised a full-blown marketing effort and to proceed with installation studies," says Smith.

Throughout spring 1978, Smith and McGowen went out to sell the concept. The main targets were operators of DC-8-61, -62 and -63 models rather than earlier variants. The higher capacity of these models, in terms of range and payload, meant they could compete favourably with newer aircraft like the A310, 757 and 767, whereas the economics of the earlier Series 50 were marginal.

The long fatigue life of the rugged DC-8 was another factor in its favour. Douglas designed it

for a life of 100,000h and, by the time Smith and McGowen were visiting airlines, most had amassed between 35,OOOh and 40,000h. Douglas deliberately designed a structure free from susceptibility to fatigue cracking, and spent an unusually high amount on airframe and component testing. Structural audits on several DC-8 "Super60s" revealed thatremaininguseful life was about 50,000h, meaning that each airframe potentially had up to 20 years of life left.

"The three key airlines were United, Delta and Flying Tigers. They all had a noise problem, and they all had to make a decision," says Smith. By this time, CFMI faced competition from Pratt & Whitney which had seen an opening for its re-fanned JT8D-2 00. "That engine did not have the noise or performance benefits of the CFM, but it was a couple of hundred thousand [dollars] cheaper and did not require the development of an in-flight thrust revers-er," he says. This was needed for the DC-8 because it had no spoilers. Snecma tookrespon-sibility for designing the thrust reversers, which were fitted to the inboard engines only, and for making sure "they did not come unglued".

BUILDING SUPPORT While CFMI and Cammacorp drummed up interest in the DC-8 project, flight tests restarted on the Caravelle. This time the tests, starting in July 1978, evaluated two different types of nacelle. The first six flights tested a confluent flow, or long duct nacelle, while a further 15 tested a more conventional, separate flow nacelle design. Tests by now had shown the baseline design was capable of higher thrust and

the rating of the initial CFM56 variant was notched up to 24,0001b, from 22,0001b. Test hours on the Caravelle and on ground test engines continued to build and, by the start of 1979, had passed the 4,000h mark.

As CFMI entered the new year, its management faced the bitter fact that time was running out. It had been five years since the company had been formed, and almost eight since the concept had been initiated, yet still not a single production contract had been clinched. The imminent go-ahead of the 757 and 767 programmes had all but torpedoed the 707-700 effort, though CFMI still hoped it might bring in business from the USAF. In reality, however, everything hinged on the DC-8 deal.

A decision was due in March 1979. Shortly before this the company held a board meeting in Lynn."Rowe said if we didn't get the DC-8 order we'd just have to put the CFM56 'on the back burner'. He didn't mean cancel it, but it would have meant not spending any more money on it either. Wth two engine companies behind it, neither would cancel it, but effectively it was a life or death decision after five years," says Smith.

Rossignol, who was vice-president engineering at the time, had spent a year with Douglas on an abortive attempt to re-engine a modified DC-9 and knew the seriousness of clinching the DC-8 work. "It was a matter of life or death. I remember Rene Ravaud calling me late at night and saying: 'I can hear the vultures flying around this building. If we don't get the United order tomorrow, we're dropping the whole thing'."

After an agonising wait, Ed Wood took a call on 29 March from United's Dick Ferris. The


• the DC-8 re-engining effort (far right)

Boeing. Smith says: "Noise regulations were beginning to kick-in then, and they had to meet Stage 2. There were no hushkits at that time, and so the CFM56 appeared to be the right solution."

Boeing and CFMI announced on 10 February, 1977, an agreement to "conduct a development and flight test programme leading towards full certification of a CFM56-powered

and sharing the income. "It got to be a big issue, but we just couldn't get into it," remembers Hope. The problem was illustrated by the hard time the partners had in even deciding where to meet. "We said: 'Let's meetin Caracas'. Butthey said: 'No way, they kidnap people there. Then they suggested Martinique, and we said: 'Hell no, that's French'. We needed somewhere neutral, and we finally met to sort out the issue in

YC-15, which had managed a maximum altitude of 40,000ft with the CFM56, and a maximum speed of M0.78.

With flight tests under way, and what was believed to be a pivotal evaluation on the 707 coming up, the tension at CFMI seemed to be easing. In the first few months of 1977 it had made more concrete progress than at any other time since the first meeting at Paris in 1971.

I "I estimated we'd incur a 20% delay to the programme just learning to work together. But at the end of it, I would be able to tell if they were smiling or frowning over the telephone" -Jack Hope

707. The aim of the programme is to develop additional commercial and military markets for the 707 series." By this time, sales of the 707 had slowed to a trickle with military airframes dominating what was left of the line at Renton. The hope was that the CFMS6 could generate interest in a new version, the 707-700, which would have a range of 5,200nm (9,640km), or roughly 10% further than the 707-320Cwithafdll pay-load. Boeing also sketched out plans to offer retrofit kits.

With the prospect of production contracts suddenly looming, CFMI began tackling the complex issue of sorting out production value,

Barbados in the fall of 1976." The 707 deal helped make February 1977 the

busiest month yet for CFMI. Six days after the Boeing plan was announced, the CFM56 flew for the first time under the wing of one of the two YC-15 test aircraft. The 2h 4min flight began at MDC's Long Beach site in California and ended at the flight development site in Yuma, Arizona. To CFMI's relief, the operation of the engine was "trouble-free" and, throughout the test programme, aircrew reported that the engine was easy to use, reliable and responsive. By May, more than 75 test hours had been accumulated on 29 flights and the aircraft was being prepared for flying demonstrations at the Paris air show in June that year.

FLYING IN FRANCE The engine also took to the skies over Europe on 17 March, 1977, when it powdered Snecma's SE210 Caravelle testbed on a 3h 7min sortie from Merignac. With the large CFM5 6 mounted on the right side of the aircraft, and the standard Rolls-Royce Avon on the other, the Caravelle looked somewhat lop-sided. The initial test phase was completed on 10 May after a flight time of 36h 35min had been accumulated over 12 flights. The tests confirmed performance of the engine over a flight envelope up to 45,000ft(13,700m)andspeedsuptoMach0.82, and covered air starts, power management including full power take-off and fixed throttle climb to cruise altitude, lubrication and fuel system verification, and engine vibration surveys. The Caravelle tests helped demonstrate the engine over a wider performance range than the

Then came bad news - on 6 July, 1977, B-l bomber production was cancelled by the US Government.

Even as the Caravelle testbed effort was being completed, US President Jimmy Carter had indicated that cancellation of the bomber was a possibility. There were several reasons, not least of which was the political expedience of scrapping the B-1 in the midst of strategic arms limitation negotiations with the Soviet Union. Added to this, a Pentagon review had recommended a cut in procurement to 150 amid revelations that unit cost was expected to be around $100 million. B-l opponents included defence secretary Harold Brown, who argued that air-launched cruise missiles were a more cost-effective alternative. Brown later heaped further misery on CFMI when he killed the AMST programme in 1978, and with it any hopes of powering the YC-15. "My biggest surprise was when Carter cancelled the B-l, and with it the F101," recalls Hope. Many CFMI veterans feel that the CFM56 programme once more survived by the narrowest of margins and that, if the B-l cancellation had come much earlier, the whole story could have been different. In some ways, however, the sheer determination of CFMI to survive was never more evident than at this point. "Our agreement with Snecma said even if this happened, we'd keep going...and that's what we did," says Hope.

The cancellation, unsurprisingly, turned up the pressure even more on the CFMI team. From September 1977 this was headed byjean-Claude Malroux as chief executive. Malroux had been associated with the project since the


THE CFM56 STORY Taking flighl - the CFM56-2

The CFM56 took to the skies in a flurry of activity in early 1977. Trials on a Caravelle, YC-15 and 707 culminated in the programme's first breakthrough

Hope recalls the swift build-up in understanding between the GE and Snecma teams working on the CFM56. "We had to deal with a 6h time difference and the language, and I had originally estimated that we'd incur a 2 0 % delay to the programme just learning to work together. But at the end of it, I would be able to tell if they were smiling or frowning over the telephone," he says.

As engine test hours began to build through 1975, CFMI negotiated its first contract. On 18 November, it announced mat an agreement had been reached with McDonnell Douglas (MDC)

of the normal Pratt & Whitney JT8D-17, and would be used to validate the compatibility of a higher thrust engine widt the aircraft. It was also designed to test the effect of temperature on flaps - a critical parameter in a concept that depended heavily on externally blown flaps.

TESTING MILESTONE As CFMI began building the YC-15 test engine, the company marked a significant milestone when it passed the 1,000h test mark at the end of 1975. This was 2 OOh more than expected at this stage, and boosted confidence that certification

GE's new CFM56 programme general manager in 1977, recalls the tough selling job. "In 1976, Douglas Aircraft [MDC] and others were trying to create some interest in a new generic twin. Unfortunately there was not a lot of interest, but it made us realise that the biggest thing was just selling the concept of CFMI itself. I don't think there had been a joint venture commercial engine programme like this before, other than the Rolls-Snecma Olympus, which was different anyway. The Olympus project was useful to us because it reallv helped Snecma's credibility with the airlines."

Flight trialson the 707 paved the way for the hugely successful KC-135R programme

and the US Air Force to flight test a CFM56 on the YC-15 prototype Advanced Medium Short take-off and landing Transport (AMST). This was a USAF programme aimed at developing a jet-powered replacement for the Lockheed Martin C-130 and pitted the four-engined YC-15 against Boeing's YC-14 twin design.

The deal with MDC called for CFM56 flight tests to be carried out "after the present flight test objectives have been satisfied", but was funded within the company's original YC-15 contract. A single engine was to be fitted in the aircraft's left outboard (Nol) position, in place

would be achieved in 1978. By the end of 1975 four test engines were running, three of them at test sites in Villaroche and Saclay in France.

As development work continued into 1976, progress on test engines remained steady but the only contract in hand was for a single engine on a USAF/MDC test aircraft. The prospect of racing through certification with no large scale application began to concern the parent companies. The certification target quietly slipped into 1979 as marketeers chased every potential programme.

Richard (Dick) Smith, who was appointed as

Although the studies with MDC came to nothing, the CFM56 was beginning to attract the interest of Boeing. "It was the first really significant thing that started to happen with this engine," says Smith. Boeing's president, Thornton Arnold (T) Wlson, took the initiative and contacted Neumann and Ravaud to discuss re-engining the 707. "It was T Wilson's idea. He saw something there and didn't have the complete support ofhis organisation, even though it was done in a short time and under budget," says Smith.

It was mainly the determination of Wilson that gained CFMI its first vital foothold at


ground ensuing global economic depression had a dramatic effect on civil aviation, slowing traffic growth and almost killing many airlines and manufacturers.

Although government approval on both sides of trie Atlantic was needed for the formation of a joint subsidiary company, GE and Snecma signed a definitive project agreement in late January 1974. The agreement built on ground rules for die programme established at an earlier meeting in Fountainbleu, France, in late

1973. For the signing, Ravaud and his entire team flew first to Lynn, followed by a trip to GE's Aircraft Engine Group (AEG) plant at Evendale, near Cincinatti, Ohio.

In a statement to staff on die eve of Ravaud's arrival, Neumann said "trie reason for our meeting is an historic one for AEG and its employees. Mr Ravaud and I plan to sign a formal agreement leading to die establishment of a new company-called CFM International, to manage the development of the CFM56 advanced aircraft engine to meet die requirements of the world's airlines in the late 1970s and the 1980s".

Ever confident, Neumann predicted a great future for the engine. "All in all, in a time when few major new engine programme requirements loom on the horizon, the CFM56 represents an excellent opportunity for AEG to team up with a great partner to produce what we know will be an engine the airlines will want in substantial quantities later in diis decade."

TEAM BUILDING At the same time, co-chairmen Neumann and Ravaud were drawing together their first hand-picked team. In keeping with die original agreement, all chairman and chief executive positions were to be filled by Snecma personnel, beginning with Jean Sollier. One of Neumann's first calls was to Jack Hope, whom he offered the job of general manager of GE's CFM56 "programme department", while his counterpart was named as Jean-Claude Malroux, also general manager for Snecma's CF6-50 contribution. Hope recalls Neumann saying: "Look, you got us die [export] licence, now come back and run die programme."

Hope did not regret the move and remembers the strength of Neumann's and Ravaud's joint leadership during die difficult early years. "They were die real secret to this programme. In essence, I got anything I required for this programme, and they supported me all the way. Neumann would see things in an easier and clearer way dian anyone else I knew. We'd have this giant business plan, about 3/4in diick, and he'd just glance at it and put it aside. Then he'd say: 'Look, the world's going to need this engine. We are just going to go and do it'."

As the organisation gradually came together, so did die parts of the first engine. Excitement began to build as the fan, low pressure compressor, engine frames, gearboxes and low-pressure turbine modules arrived from Snecma to be fitted around die GE core. Finally, on 20 June, 1974, almost exactly three years after the first meeting between Ravaud and Neumann at Paris, "engine number one" was fired into action for die first time in a test cell at Evendale. The tests went so well that, within the first lOh of operation, test engineers rammed forward the dirotdes and pushed the engine to its maximum thrust rating of more than 22,0001b (98kN).

After initial test runs, the engine was moved to GE's outdoor test site at Peebles, around

160km (100 miles) from Evendale. Here, noise and emissions tests were started as assembly of the second test engine got under way. At this stage, Snecma employees worked under strict scrutiny, on die surface at least. They were not allowed to see any design details of the core and, incredibly, were not allowed access to this part of die engine until September 1978.

Recalling a situation in July 1974, shortly after die first run, Hope says: "We called a meeting widi Snecma and all the government people involved in licensing. We still had security, so when we got down to die IDR [instrumentation data acquisition room] there were two people sitting in two rows behind the monitors. Because of die security restrictions, we had to officially keep the Snecma people from the screens, so we put up yellow 'do not cross line' tape across die backs of all die seats. Well, the Snecma engineers just walked along and, of course, they could see everything. We satisfied everybody diat we were keeping our side of the agreement, but it made the French laugh."

The meeting in July, and subsequent get-togethers diat summer, paved the way for formal creation of the joint subsidiary company, CFM International, in September 1974. To some, it seemed everything had changed in the intervening 18 mondis since the Presidential go-ahead. The first engine was running, and the CFMI organisational chart was filling up with real people. Ravaud and Neumann served on the board of directors and took an active part in the management of the company, which had its headquarters in Paris.

The number two CFM56 was completed in Ohio around late November 1974, and shipped to Snecma's Villaroche test site in France in December. By this time, the initial performance of die engine was encouraging, with early performance guarantees met or bettered, including a 2.2% margin in specific fuel consumption recorded during die first 200h of testing.

Initial tests included comparative assessments of two nacelle configurations, a long duct design with a confluent flow exhaust, and a short duct with a separate flow design. Other tests included icing, a full range of transient "bursts and chops", crosswind tests up to 50kt at 90% engine speed, and full fan mapping (altering the size of both fan and core engine exhaust nozzles to vary fan and booster pressure ratios).

The handover of engine number two marked the real start of the close working relationship between die two engine makers. Before it was crated for its trip to France, die second power-plant was fired up in Evendale to assess its basic operating characteristics. These were dien compared with die first set of data which arrived back from die first tests at Villaroche in early 1975, and was used to correlate test cell measurements between the various test sites. The latest satellite communications technology was used to transmit test data between France and the USA


THE CFM56 STORY Taking flight - the CFM56-2

CFMI's agreement with Boeing to re-engine a 707 testbed was a key breakthrough

Presidential approval was only the first step towards getting the CFM56 programme under way in the USA and France

Getting oii the Even though General Electric and Snecma

had Presidential approval for the CFM56, the way forward under the revised structure and guidelines was far from clear.

The programme planners were encouraged by the engineering work which, under die original responsibility division, had continued, despite the hiatus. GE had focused on core work through the F101 programme while Snecma developed the low-pressure system. The French engine maker had pushed ahead with

acoustic and aerodynamic tests of the proposed fan design, while continuing to refine die configuration of full-scale low pressure system components.

In September 1973, four months after the Iceland summit, the US Government gave GE die official go-ahead for the project. A month later, and events could have overtaken the fledgling undertaking. In October 1973 the Arab-Israeli War began. This sparked an oil embargo, OPEC price iijcreases and an energy crisis. The

m FLIGHT INTERNATIONAL 19 - 25 May 1999

The memo was a near-fatal blow to the fledgling CFM56 programme. Neumann was furious and argued that not only was the USA throwing away the chance of gaining massive share in the future engine market, but that the performance goals of the CFMS6 simply could not be met with technology from the CF6 core. Talks were deadlocked and the programme had hit a brick wall.

It was then that Weiss came up with "a stupidly simple idea". It not only allowed the CFM56 programme to happen, but ironically provided the structural blueprint which allowed International Aero Engines to develop the competing V2500 turbofan several years later. "I think I was jogging when it came to me. I thought why not just switch it all around and put the sensitive part of the programme in Ohio, where we can protect it," he says.

A CHANCE TO TALK Weiss' simple idea coincided with a fortuitous visit to Paris to discuss trade issues with the Atlantic Council of the USA. Although the symposium had nothing to do with aviation, it gave Weiss a chance to talk to Snecma about the alternative plan which was forming. Since the September rejection, Weiss and several other members of the Options Group had begun to rethink the whole concept; "Most likely because we thought latent in the venture was a thoroughly promising idea." In the latter stages of GE's desperate pitch to get the go-ahead, it had discussed the market potential for the CFM56, which showed a possible requirement for 2,000 engines to re-engine Strategic Air Command's Boeing KC-13 5 tanker fleet. On top of that, GE judged the market to power possible new versions of the Boeing 737 at 2,000 engines.

Through Snecma's US representative, Don Agger, Weiss suggested a meeting in Paris to try and salvage the plan. Once there, he abandoned an evening session with the Atlantic Council to meet with Snecma's Jean Crepan and a colleague atMaxim's restaurant. "While expounding back and forth about test locations we consumed quantities of napkins sketching parts, test cells, all the while being amused by strolling violins. The maitre d'hotel was less amused as we made aviation history by increasing his napkin laundry," recalls Weiss. The Snecma executives liked what they heard and, the following night, Weiss met with Ravaud himself.

"We had never met before and I just loved him. He definitely had what it takes and together we agreed on what we needed to do. I said: 'Here is what we are going to do...we'll build the first engines in Ohio. At this stage I hadn't even told GE about the plan," Weiss recalls. While Ravaud took up the proposed change in structure with Snecma and GE, Weiss returned to Washington to drum up political support.

Ravaud and Neumann stand before their creation: the CFMS6

It soon became obvious this was not going to be as easy as first envisioned. "In an utter tirade, the NASA Administrator called the project a giveaway of US technology and was doubtful thatthe CFM56 could meetits low-noise goal," says Weiss. "The Treasury representative concluded that the joint venture would be ruinous to US aerospace trade and destructive of our jet engine industry." As if that was not bad enough, with apparently no warning Kissinger issued a National Security Directive stating that the US Government would not condone further approaches and that the President's denial was firm. This directive later "evaporated" on reflection, says Weiss.

Meanwhile, Hope rallied to the cause and tackled the technology transfer issue head-on. He recalls: "I took a cardboard box full of parts, nozzles, blades and so on, and went out to visit the CIA. I then drew a big chart about six feet wide and eight feet long, which started with the J47 engine and came right up to date. Then I took out all the blades and other pieces, and drew arrows from the chart to the relevant pieces. All of a sudden I was the number one show at the CIA. They were particularly worried about the blade technology, and the impingement cooled blades in the Fl 01. They were amazed GE had already licensed the technology to Germany. So they ended up asking 'what are we protecting?'."

Weiss plodded on with the political solution, adding a financial incentive to attract Treasury support. Together with the DoD, the Treasury Department had pointed out that F101 development costs were close to $100 million, and that any deal with Snecma based on this technology should involve some kind of payback. The Treasury first suggested that recoupment over the life of the programme should be around $250 million, to compensate for the long term harm to the US economy it claimed was bound to result.

Documents declassified in October last year reveal that GE initially proposed typical royalty payments of around $12,500 per engine based on a commercial market estimate of 4,500 units. Of these, GE predicted 2,700 would be on aircraft sold abroad and 1,800 on those sold in the USA. The B-l programme, at this stage,

called for 1,169 engines. The total market, therefore, came to 5,669 engines, giving an overall payoff of around $55 million to the US Government, versus the $48 millionit was originally expecting based on the B-l programme alone. Based on market predictions of 4,000 engines, Weiss later calculated a more realistic royalty of $80 million, payable at $20,000 per engine. GE accepted this and on 11 May, 1973, submitted a revised licence application. The document also included a provision that the French Foreign Ministry would agree to oppose any increase in

import duties levied on US aerospace products sold to the European Community - another idea from Weiss. At this point, the CFM56 plan still hung in the balance. The Pentagon dropped its technology transfer objections, but raised a financial objection, saying the project "would be of no net economic benefit to the USA" since it called for the exchange of US market access merely for French capital.

It looked as if the stalemate would carry on forever when Weiss, and the CFM56, had a stroke of luck. "By chance, I happened to meet Treasury Secretary Shultz coming out of a White House meeting and quickly told him of the $80 million royalty: he said that was just fine by him, that the project was an excellent idea and to go ahead. That eveningl told his assistant about the Secretary's decision at a White House softball game."

ELEVENTH-HOUR RESCUE On 18 May, 1973, the GE-Snecma Options Group recommended release of the F101 and approval of the revised scheme, pending final agreement on the royalty. Wth victory in sight, GE and Snecma officials were about to crack open the champagne when they heard, to their dismay, that the Defense Department had changed its mind and, once more, was vehemently objecting.

Just as the whole process seemed about to be painfully repeated, rescue arrived in the form of an unexpected letter to Nixon from Senator Barry Goldwater. The letter strongly advised that the project be given the go-ahead and expressed hopes that the Administration was not about to "do something stupid" and deny the licence.

Timing was critical at this point. The Options Group recommendation came just two weeks before a long-scheduled summit meeting with Pompidou in Iceland. As soon as Ravaud heard the group's decision, he pulled as many strings as he could get hold of, and made sure the CFM56 was on Pompidou's agenda. "This was a deliberate and artful move to force a US decision - without it nothing might ever be settled," recalls Weiss. Twelve days later, on 30 May, 1973, the CFM56 project was officially given the go-ahead by Nixon and Pompidou.


THE CFM56 STORY Presidential beginnings

these amazing restaurants in France. So when he was in the USA Neumann wondered what to do. We had this little diner across the street from his office in Lynn and they all agreed to go there with die French for lunch. To Ravaud's amazement they rolled a red carpet right across the street. They even got the diner a one-day liquor licence, the menu was all in French and even the signs. Ravaud loved that."

FORMAL PARTNERS In November 1971, Snecma formally selected GE as its partner to design and develop the CFMS6. The following month, the French Government approved Snecma's recommendation to launch the engine in co-operation with GE, while the US engine manufacturer made what it thought was a routine export licence application to the Department of State's Office of Munitions Control.

To GE's shock, the Department of Defense

and, the CFM56 project was grounded before it could even enter the design phase.

GE appealed the decision and the Council of International Economic Policy, which advised on issues of trade and technology, was brought in to find a solution. Headed by Flanigan, the Council recruited one of its top economists, Gus Weiss, to tackle the affair. "On April 24 [1972], Pete asked me to look into 'some sort of problem' about a jet engine export and to do so by forming a team of government people to review the problem and provide recommendations within a month," says Weiss. The team consisted of "six or eight government aviation experts whose minds had not yet been ossified by years in the civil service," says Weiss, who is credited with being instrumental to the survival of the CFM56 programme.

"GE just would not give up on it. We talked to Neumann and everyone else, and it turned out the real problem was they were not protecting

pleted B-l hardware sent to France would be unclassified," he says.

Hope, on the other hand, was one of the government people turning the idea down. "The big thing we needed to protect was how to do the business. The biggest issue I had was mat the French were designated as system integrators. GE had a technical licence for one year and it said they could use CF6 technology with the French. I pointed out that they were not really doing that by using the F101. The USAF propulsion technology lab was particularly against it, so we disallowed a new one-year extension of die licence as it stood. We said we would allow it, but only with certain restrictions," adds Hope.

On 17 July, 1972, the Options Group set up to discuss the problem reported to Flanigan, Henry Kissinger and George Shultz. According to Weiss, it advised that, "should F101 export be approved, Snecma would obtain substantial

Spinning a military core off into a commercial engine promised to save the USAF money

(DoD) objected to the approval because of "possible compromises of sensitive technology" due to use of die F101 core. The DoD believed that revealing the technology in the cooling, metallurgy and design of the turbine and compressor constituted a "national security problem". Through the US Central Intelligence Agency (CIA), and its British counterpart, the DoD was aware diat secrets of the Olympus 593 engine had been leaked to the Soviet Union. They expressed concern that the technology of the F101 would follow a similar route. GE was perplexed, as the B-l engine was not a classified project. In addition, elements of the core were no more advanced than other parts already licensed for production by other foreign partners. Despite these facts, die judgement stood

the engine bv having it bolted together in Paris. It just seemed to be due to this that the entire idea was being turned down," says Weiss. He thought Neumann was "a phenomenon. On and on he went: that the CFM56 was a fabulous idea, that the B-l engine had been promised by the Defense Department, that military uses would be found, that the project would be a huge success, and that the US Government couldn't be so shortsighted as to turn it down".

Neumann tried to reassure Weiss that there would be no technological leaks. "Snecma would not physically observe the inner workings of the B-1 engine, but would be given interface data from which to build its own sections of the powerplant. Specially sensitive performance data was to be excluded, and all the corn-

access to the US airframe market, but we judged this hardly sufficient reason to stop the project. It was the loss of engineering art and know-how in the integration and first tests in France that was of genuine concern, and there was no way in GE's plan that this risk could be controlled. We recommended against licence approval, which was accepted".

The result was National Security Decision Memorandum 189, issued by the White House on 19 September, 1972. It said: "The President has approved extension of the CF6-50A data licence for one year. He has disapproved GE's export application for one XF101 engine core. In so doing, he has emphasised that the data to be exchanged under the extended CF6-50A licence will be exclusive of data embodied in XE101 engine core technology."


I applications and ahead of its time, and the Olympus engine forthe Concorde. Like P& W, though to a far lesser extent, R-R also harboured ambitions for narrowbody powerplants of its own. The UK manufacturer had enjoyed some success with the Spey, and planned further developments using the same core, which would eventually emerge as the Tay.

GE, on the other hand, had seen the vast potential of the JT8D replacement market and was as keen as Snecma to diversify further into commercial propulsion. In 1971, the company's only commercial engine in production was the CF6. Conscious of the yawning gap in its civil product line, GE had launched studies in 1968 of a new-technology turbofan in the 22,000-25,0001b thrust arena. Dubbed the GE 13, the project was focused on high performance coupled with reduced complexity, lower weight and fewer parts. To achieve these targets, GE needed a more advanced core than that of the CF6-50, and looked to the US Air Force's Advanced Turbine Engine Gas Generator (ATEGG) programme to provide the answer.

Two advanced cores were under development by GE for the ATEGG effort, one of which (the GEl/9) had a compact nine-stage high pressure compressor and a single-stage high pressure turbine. The shorter compressor meant the entire engine could be supported on fewer bearings. This, in turn meant the number of support frames, oil sumps and oil supply systems could be reduced from three to two - dramatically reducing complexity and cost.

At around the same time as the GE13 study was under way, GE was bidding hard against P& W to win the engine contract for the USAF's next bomber, the Advanced Manned Strategic Environmental pressures boosted the cause of the CFM56

ve the inner workings of the B-l engine... Specially sensitive performance data leted B-l hardware sent to France would be unclassified" - Gus Weiss

prominent, Snecma advertised the M56 as Europe's future "quiet engine". The turbofan was designed with a high bypass ratio of 4:1 and a fan duct flow of 260kg/s (5761b/s). Overall mass flow was 327kg/s, while overall pressure ratio was 18:1. The engine had a six-stage low pressure compressor, five-stage high pressure compressor, single-stage high and low pressure turbines and what the company described as a "three-stage fan drive turbine".

POTENTIAL RECOGNISED Snecma clearly had its work cut out developing theM56 in the proposed timescale, and Ravaud instantly recognised the potential time saving of GE's expertise in advanced cores. Neumann saw that the GE13 (Fl 01) core was a perfect fit for the M56, and the two sketched out a plan to combine the projects. They agreed to assemble technical teams and meet a month later at GE's Lynn plant in Massachusetts.

Progress was rapid. A team of Snecma engineers arrived in the USA armed with blueprints of the proposed M56. Their GE counterparts spread out drawings of the GE13 and the two groups compared data. Neumann and Ravaud, meanwhile, began work on the structure of the

Aircraft. Its proposed engine was based on a modified version of the GEl/9 ATEGG core, and had the same nine-stage compressor and single-stage turbine. In 1970, GE won the contract to power the new B-1A bomber to be built by Rockwell North American. Its engine, now designated the F101-101, was rated at more than 3 0,0001b thrust with maximum afterburner. The military contract, therefore, gave a boost to the GE 13 study by providing a perfectly sized, fully funded advanced core.

By the time Neumann and Ravaud first talked, Snecma was already well advanced with its M56 plan. The first bench test run was scheduled for the end of 197 3, with service entry due in 1977. With environmental issues already

joint development programme, establishing a rough cost basis for sharing out the work equally. They estimated that the cost of the core would be crudely equal to that of the fan, low pressure turbine and gearbox, so this formed the baseline for the division of responsibility. GE would make the core, while Snecma would make the fan module, low pressure turbine and accessory drive gearboxes.

They agreed that both companies would assemble complete engines. Due to the constant currency fluctuations, both sides agreed not to ask each other for a cost accounting and would simply pay their own bills. Snecma was to take overall project leadership and put the complete engines together as the system integrator in

France, while GE was to have responsibility for systems engineering. Marketing was initially given to Snecma, but GE later resumed marketing to specific airlines with which it had closer relationships.

Lastly, the two agreed on a name for the new engine. As with the rest of the simple plan, the name combined elements from both companies. From GE it took the CF or "Commercial Fan" of the CF6, while from Snecma it adopted the M56. The result was the CFM56.

Jean Sollier, the first general manager of CFM International, recalls his input at this stage: "I was the one who selected the name CFMI, and I designed the CFMI logo. That was easy... Snecma s logo was within a red rectangle and GE's a circle. I tried combining the two and the result was something which looked a bit like the Olympic Games insignia. Well, we did become champions, didn't we?"

With their plan in hand, Neumann and Ravaud headed off to explain the scheme to their respective company boards. "We agreed to make a real partnership of our joint efforts, and to split the responsibility evenly between our two companies," Neumann recalls in his memoirs. "We created a small management

company, CFM International, registered both in France and the United States. We Americans silently blessed the Lord that He made the French so clever that they could conduct all business with us in English."

From the outset, the strength of the relationship was echoed by Neumann's respect for Ravaud, a former naval designer and Ingenieur General of the French Army. "Ravaud, a big, tough soldier and a charming person, had lost his right arm in France shortly after D-Day 1944, during the battle for the liberation of Brest harbour. There could not be a more courageous, intelligent and enthusiastic leader: Ravaud was much admired by French presidents and members of their cabi

nets. He and I clicked from the very first moment we met," he recalls.

The often humourous relationship between Neumann and Ravaud helped sustain CFM through the tough times ahead. Neumann recalls: "I sent him a cable to Paris that we had succeeded in a certain negotiation 'even if it cost us an arm and a leg'. Too late I realised what message I had sent to one-armed Ravaud, and made every effort to retrieve this cable. But it was too late. I received a cable the next morning: ' IT IS BETTER TO LOSE ONE'S ARM THAN ONE'S HEAD. RAVAUD'."

Hope recalls Neumann and Ravaud were "always trying to outdo each other with little tricks. Ravaud would always take Neumann to


THE CFM56 STORY Presidential beginnings

The B-l 's F101 engine provided the CFMS6 core

tory. Ravaud raised the topic of Snecma's new " 1 Ot" study engine, the M56. Would it be possible, he asked, for GE to join forces with Snecma on this new project?

Neumann was extremely interested. Like everyone in the engine business, he knew Snecma was looking for a partner to help it develop the M56 - a new commercial engine in the 20,0001b-thrust (89kN) range. The French Government-supported engine project had been conceived as a running mate for the newly created Airbus Industrie. The engine would not only allow Snecma to diversify away from its reliance on French military aircraft, but would also give it a chance to compete to power a raft of proposed new narrowbody twins.

Foremost of these was a possible follow-on to the Pratt & Whitney JT8D-powered Dassault

Mercure, while others included multi-national short take-off and landing concepts and 150-seater projects that would later evolve into Europlane, and the Joint European Transport (JET). The JET study, undertaken by partners in France, Germany, the Netherlands and the UK, was based around the CFM56 and later gave birth to the Airbus SA (single-aisle) family, which - 10 years after that first Paris meeting between Neumann and Ravaud - ultimately became the A3 20.

Although die French plan was full of ambition, it was tempered by reality. Snecma knew that, to develop a successful new commercial powerplant, it would need to form a partnership with one of the "big three", GE, P& Wor Rolls-Royce. It had a relationship with GE, but initially began partnership talks with P&W, which already held an 11 % stake in tire French engine

I ''Snecma would not physically obst was to be excluded, and all the com

maker. However, it soon became obvious that P&W was not very interested. This was hardly surprising given that the M56 was a proposed successor to its ownJT8D, which was becoming established as the fastest-selling civil power-plant ever. Furthermore, P&W had plans to re-fan die JT8D, and saw plenty of life left in its engine. This re-fanned version would eventually emerge as the JT8D-200, powering the McDonnell Douglas MD-80.

R-R was more interested in the concept, but was struggling to climb out of bankruptcy following RB2 11 development problems, and was in no position to take on new projects. Besides, it was still fully engaged with Snecma on the M45, a lightweight regional turbofan short on


Presidential beginnings THE CFM56 STORY

Presidential beginnings

World events and the Presidents of France and the USA played key roles in the launch of the CFM56 by General Electric and Snecma

The birth certificate of the CFM56

President Richard Nixon scanned his briefing notes as "Air Force One" carried him high over the North Atlantic towards Iceland and a vital meeting with French

President George Pompidou. The US president had a lot on his mind. Two

of his closest advisors had just been forced to resign, and the storm of Watergate was fast approaching. Despite this domestic nightmare, Nixon knew he needed to focus on reaching a lasting and stable peace in South-East Asia. It was May 1973 and a ceasefire agreement to end the Vietnam War had been signed in Paris the previous January, but only four months later the political fall-out was far from certain. Now Nixon was counting on Pompidou's firm stance on Sino-Soviet relations to help stabilise the strategic situation.

It is probably not surprising, therefore, that Nixon was apparently perplexed by one document put before him during the flight from Andrews AFB, Maryland. Handing him the unusual memorandum was none other than his National Security Advisor, Dr Henry Kissinger - whose shuttle diplomacy with Le Due Tho had led to the ceasefire in Laos. Marked "Confidential", the memo was from Peter Flanigan, Assistant to the President and Director of the Council on International Economic Policy.

Entitled Revised proposal for export of the B-l bomber engine to France, it gave Nixon the choice of either approving or disapproving a proposed joint venture between General Electric and Snecma, the French engine maker. The two wanted to make a new " 1 Ot" commercial turbo-fan using the core of the Rockwell B-l bomber's engine, GE's F101, but had been blocked by US security concerns. The memo gave Nixon three possible options. Flanigan put an "X" in the "approve" section of option two, and in his recommendation to the President noted that "messrs Kissinger and Shultz [treasury secretary George Shultzj and Ash concur".

Option two was that the US Government should grant an export licence "subject to an appropriate agreement concerning the physical security of the engine technology". It also

stressed that the USA should "seek assurances from the French that they will not seek EC [European Commission] tariffs against US aircraft imports". In his recommendation to Nixon, Flanigan also noted that "the French have indicated President Pompidou's personal interest in this engine venture and their need to have a US decision for purposes of their internal civil aviation planning".

Option one was a simple yes or no to GE's licence request, and option three was that no decision should be made at the Iceland summit. "This is the Department of Defense position," Flanigan duly noted.

Nixon scanned the sheet with irritation. "He couldn't have cared less about this," recalls Jack Hope, the first GE-appointed general manager of what was to be called the CFM56. Hope later got to know Nixon through his work as a technical consultant to the President's Office of Science and Technology. "He'd send me notes saying'what the hell does S.. .N.. .E.. .C.. .M... A stand for?' He would say: 'Here I am interested in world peace, and the French are interested in a goddam airplane motor'."

Wth a flourish, Nixon quickly signed option two. To all intents and purposes, the CFMS6 project had just been bom at 3 5,000ft (10,600m).

FRENCH AMBITION Nixon's signature ended the first act of a drama that had begun two years earlier at the 1971 Paris air show. On the last day of the show, then GE Aircraft Engines vice president Gerhard Neumann decided to introduce himself to the newly installed president and director general of Snecma, Rene Ravaud. The two companies had, by this stage, become involved on the CF6-50 engine for the Airbus A300, and Neumann had not had a chance to pay his respects to his French counterpart.

The two hit it off immediately and their rapport was to have a profound effect on the history of aviation. Neumann travelled with Ravaud to Snecma's nearby Paris headquarters for a short discussion before his flight home. After covering the success of their technical collaboration on the CF6, the talks entered new terri-


w 0 $ A> IEAOI Am o mmt-

Contents and Introduction • THE CFM56 STORY I A; l£>\

CFM56: Power and the glory

CFM International will mark the handover of the

10,000th CFM56 engine at the 1999 Paris air show.

This impressive milestone is all the more remarkable

for a programme that, just 20 years ago, was within

days of cancellation because of an empty orderbook.

From a mere concept, CFMI has overcome hurdle after

hurdle to become the world's most successful commercial jet

engine programme. By early 1999, almost 12,900 engines were

on firm order and total commitments extended the overall tally to

almost. More than 3,500 CFM56-powered aircraft are in service

and, to meet surging demand, production is expected to peak at

around 1,130 units in 1999 - an all-time record. Cumulative gross

income for the programme is close to $39 billion.

Two generations of Boeing 737s have enjoyed unparalled

sales success thanks, in large part, to the CFM56. The same

engine has, paradoxically, formed the foundation for the equally

dramatic sales success of the Airbus A320 family, while powering

the consortium's long-range A340. The engine has transformed

the performance, endurance and longevity of the US Air Force's

Boeing KC-135 tanker fleet, as well as a host of other 707-based

military aircraft operated by several other nations.

While the phenomenal success of the engine is obvious,

the deeper impact of the pioneering joint venture can only be

guessed at. The international structure created by General

Electric and Snecma to allow the formation of CFMI provided the

blueprint for several successive cross-border collaborations.

CFMI therefore became a vital role model for the industry just as

it faced the inevitable wave of globalisation.. Now, as it faces the

challenges of 2000 and beyond, CFMI is protecting the future

with its aggressive "TECH56" research and development effort.

With an impressive 25-year legacy behind it, the company seems

to be in good shape to deal with the coming century.


papter % . ' • Presidential beginnn

Gelling o f f " L11 LHii LCU' 1UJ

l l l l l a

jexf general^! the ful

written

edited

'&*«£• ''«,VVt't M i l i i l j

contributions Julian rvtoxon

design James Mason?'

production "f Graeme Osborn

LmajHaiim

cutaway ^ " H i m Hall . : * B ' . : • — '

.1.-1 fil- '' '-artwork §§£) iGareth EBfifless

Publishec e'ed Business-lnfjPrmation Ltd,

Sutton, Surrey, SM2£A5

RBI Ltd;^999 .

Quadrant, AJ*. \

L1