The Ultimate Dogfight?

Nonetheless, the Mirage cannot continue indefinitely, and next month we are due to learn whether the RAAF is to be allowed to purchase 75 F-16s or 75 F-18s. This is therefore a very appropriate time to recall why a new fighter is to be procured, and to examine the various arguments for and against the two remaining contenders.

Roles and Missions

The Air Staff Target (AST) for a Mirage replacement was first issued in September 1970 and graduated to the status of a formal Requirement (ASA) on November 1971. Thinking on the operational employment of TFF was refined in the Central Studies Establishment’s TAWS (Tactical Air Weapons Study) and by the Air Staff’s own “Fighter Development and the Tactical Fighter Force” study of 1974, which led to the issuing of an RFP to manufacturers in November of 1976.

None of this work was made public, but Defence Minister Killen, writing in “Aircraft” of December 1975, stated that “There is a very great need for informed debate on defence matters in Australia”. Thus encouraged, the Strategic and Defence Studies Centre of the Australian National University in Canberra organised a conference in November 1976 on “The Future of Tactical Air Power in the Defence of Australia”.

This conference is a useful jumping-off point for any TFF discussion since it took the form of a series of theses by the staff of SDSC, an independent study group that is specifically concerned with the general philosophical, technological, strategic, and industrial aspects of Australian defence.

What emerged was a strong consensus in favour of choosing equipment on the basis of deterrent effect, giving priority to the roles of a maritime strike against an invasion force, offensive counter-air operations, and interdiction, with STOL performance to allow dispersal or the use of bombed runways. Air defence was seen as a secondary role, with air superiority only desirable in a local context to assist the Army in containing an enemy beachhead.

Although the conference had no official standing with DoD, the SDSC view of TFF is noteworthy as being apparently quite logical, yet running completely counter to RAAF policy. Aside from the interdiction of sea lanes (in the defence of which the RAAF can play only a limited part), the only conceivable major non-nuclear threat to Australia is invasion over a long sea route. If such an invasion can be deterred by the presence of an anti-shipping strike force, or – failing that – destroyed in passage, then Australia would undoubtedly survive.

Even in an extreme worst-case scenario with Soviet forces occupying Indonesia, the air threat to northern Australia would be severely limited. Looking ahead to the next century, if the Soviet Navy really does develop a force of super-carriers and imitate the US Navy in the use of air power at sea, then the need for air defence for Australian cities might become very real. However, that possibility is a generation away. In the interim, cruise missiles launched by ships or submarines might form a limited threat, but this possibility hardly justifies the purchase of air defence fighters.

Notwithstanding the evident logic of a maritime strike aircraft purchase, the then-GAS (Jim Rowland) opposed the idea of choosing an aircraft on the basis of a single scenario on the grounds that it represented a “reaction-type defence philosophy” that would severely restrict RAAF options.

Lower down the chain, RAAF officers voiced the opinion that as long as they can control the sky, no invasion can hope to succeed. An enemy might gain a limited foothold in the north, but without air cover his resupply would be desperate, and in due course, he would be thrown back in the sea. In effect, they appear to be arguing that the role of the RAAF is not to win the war single-handed but to act in concert with the Navy and Army, each Service doing the job it does best.

Rightly or wrongly, the RAAF is convinced that the new fighter must be chosen for its capability in the mission for which fighters are best employed, namely gaining control of the air (both by air defence and offensive fighter sweeps) and that air-to-surface missions can be performed by this same aircraft once air superiority has been achieved. The TFF missions in descending order of priority are thus: air defence, air superiority, all-weather tactical interdiction, and close air support.

Provided that the other Services do not drag their heels in procuring advanced surface-to-surface weapon systems, then RAAF’s emphasis on control of the sky may be quite justified. However, looking back at the Service’s record over the last quarter-century, it is worth noting that the Sabre was virtually useless in the air-to-surface mission, and that the Mirage Ill is heavily biased toward air defence. It is therefore to be hoped that whatever aircraft replaces the Mirage will provide a major advance in warlord-radius performance in the ground attack mission, despite air-to-air performance taking priority environment.

One of several issues that complicate the choice of a Mirage replacement is that at some time during the service life of this new aircraft, both the F-111 and the Canberra will be phased out of the RAAF inventory (the latter supposedly in 1983). It follows that either an additional batch of Mirage replacements should be purchased in due course to supersede these long-range aircraft, or that the Mirage replacement will operate alongside a new all-weather strike/interdiction/reconnaissance aircraft.

If a single type is to replace all existing RAAF combat aircraft, then the choice will need to be raised toward a larger size than simple Mirage replacement considerations would suggest, in order to achieve good warlord-radius performance and then have internal space for all-weather attack avionics.

On the other hand, if two different types are to be operated, then commonality considerations (especially in regard to the engines) may well play an important role in the selection of aircraft. This matter of replacing the F-111 could well be the crucial issue in TFF selection since the retention of an offensive capability is seen by the RAAF as the principal hope of deterring a potential aggressor.

By the same token, the Mirage replacement issue cannot be divorced from the question of whether the RAN is to be funded to replace HMAS Melbourne with ships capable of operating high-performance fixed-wing aircraft. In the absence of organic naval air cover, the RAAF would presumably be committed to providing air defence over an extensive area of sea. This might bias the choice towards a larger aircraft, and emphasis on overwater operation would certainly favour a twin-engined aircraft.

The alternative and highly unlikely course of giving the RAN a land-based fighter for maritime strike and offshore air defence (analogous to the West German Bundesmarine use of Tornado and F-104G) would have the same effect on the choice of Mirage replacement since DoD would clearly insist on inter-Service commonality.

One Donk Or Two?

Before embarking on a specific comparison of the F-16 and F-18, it may be useful to put the general debate between single- and twin-engined combat aircraft into perspective. The idea that two engines are safer than one assumes that a twin-engined aircraft will generally survive the failure or destruction of one engine and that the associated saving of aircraft more than offsets the consequences of having twice as many engine shut-downs and twice as many engines for the enemy to hit.

Whether it is fair to assume that such aircraft generally survive the loss of a single engine depends on the individual type of aircraft. For example, engines mounted closely together risk interactions from such hazards as fires, burst rotors, etc. Conversely, widely-spaced engines pose asymmetric handling problems, as in the Canberra. A compromise on spacing (eg F-14, F-15) is clearly preferable from an attrition standpoint. The problem is more complicated in the case of a ground attack aircraft that is underpowered on one engine in the approach configuration: it may simply not have sufficient residual thrust for an overshoot.

Since intuition on one-vs-two does not appear to work, one must look for statistical evidence. However, once again there are problems in making a fair comparison. For example, a comparison of current USAF accident statistics for the T-38 on training flights over the Texas desert and early German Luftwaffe figures for the F-104G on low-level ground attack missions over Europe in winter would obviously show a sensational advantage for the twin-engined aircraft, but it would prove nothing.

The first rule in such comparisons is that attrition rate figures must be averaged over a vast number of flying hours if they are to have any statistical significance. Hence you look at US data or Soviet data, or you forget it!

Secondly, the data must compare aircraft used in the same role. The fighter that spends its life in the wild blue yonder of air defence has a far better chance of ending up as a gate-guard than a ground attack aircraft that flies at 600 knots “in the weeds”, regardless of the number of engines.

Thirdly, it is vitally important that the data should relate to the same timescale. With each decade there is a dramatic fall in accident rates for combat aircraft in general: early attrition of the Meteor and Canberra (for example) would nowadays be regarded as suicidal. This trend is explained largely by better-engineered aircraft, using experience gained from generations of jet fighters to achieve higher standards of reliability and more comprehensive emergency systems, but it is probably also true to say that operators are becoming increasingly more professional and more aware of the cost of replacing an aircraft and of the public reaction to the loss of life. In addition, the data would relate to a fixed point on the learning curve.

Fourthly, statistics for comparison should relate to single and twin-engined aircraft using the same engine. It follows that a fair comparison is rarely possible: to be genuinely meaningful, the two sets of attrition data should relate to aircraft that are US fighters, employed in the same role, in the same time frame, and using the same engine!

The first major attempt to produce a valid generalised comparison of attrition rates emerged as an article entitled “One Hole Or Two?” by Lt-Col Robert G. Dilger, which was published in “USAF Fighter Weapons Review” of Spring 1975. Although this journal has restricted circulation, the article was duplicated and distributed by several companies interested in promoting single-engined aircraft.

Dilger’s analysis concluded firstly that USAF fighter peacetime accident rates had decreased by approximately 50 per cent each decade, ie from a rate per 100,000 flying hours of around 40 in the mid-1950s to 20 in the mid-1960s and 10 in the mid-1970s. One might therefore extrapolate to an average rate of five in the mid-1980s and 2. 5 in the mid-1990s. The report went on to compare mission risks and showed that Tactical Air Command had a 60 per cent higher attrition rate than that for Air Defense Command.

Comparing accident rates per 100,000 flying hours for all USAF fighters in the period 1969/73, Dilger produced an average figure of 12. 2 for single-engined aircraft and 12. 3 for twins. He went on to examine engine-caused accidents, which over the same period averaged out at 2. 6 for single-engined aircraft and 0. 8 for twins, implying that there should be an overall difference of less than 20 per cent. However, it appeared from his set of statistics that other types of accidents more than compensated for the advantage of twins in engine caused accidents, giving an overall slight advantage in peacetime attrition to the single-engined aircraft. Turning to wartime attrition, Dilger examined USAF and USN statistics for day strike missions against North Vietnam and concluded that the single-engined A-4, A-7, and F-105 had an average attrition rate approximately half that for the twin-engined A-6 and F-4. He summarised his findings by saying that neither peacetime nor wartime data supported the intuitive view that a twin-engined aircraft is safer than a single-engined one.

Appearing. at the time of the great YF-17 vs’ YF-16 contest for “The Sale of the Century”, the Dilger Report (which was as well-known in its day as the present Bazzocchi Report on military flying training) naturally led to an instant rebuttal from Northrop.

The company’s main objection to his peacetime statistics was that it represented a random collection of “apples and oranges”, mixing data on aircraft of widely differing performance, in different missions, and at different stages of their learning curves. Northrop’s own analysis from the same sources (USAF Accident Bulletins and USN Aircraft Accident Statistics) showed that 32 per cent of recorded engine failures resulted in a major accident for singles, compared to only four per cent in the case of twins. Of these accidents, 94 per cent resulted in total loss of the aircraft for singles, and 60 per cent for twins.

In the last 10 years (ie prior to 1975), approximately 40 per cent of single-engine aircraft attrition was engine-related, compared to only 1 5 per cent for twins. Compared on the basis of similar missions, time periods, and experience levels, major accident rates for singles were 1. 5 to 2. 0 times higher than for twins. This overview is supported by comparing J57-engined aircraft at the one million hour mark, taking the F-101, F-100, and F-8 (disembarked) in the tactical mission, and the F-101 and F-102 in the air defence mission.

Turning to Dilger’s combat statistics, Northrop argued that the sample sizes on which his attrition figures were based were too small to be statistically significant. Whereas Dilger’s analysis related to only 100,000 sorties, Northrop looked at over one million, and concluded that the total loss rate was 50 per cent higher for singles, engine-caused losses were 400 per cent for singles, and that the likelihood of an aircraft loss – given a hit in the engine compartment – was 25 per cent for singles and only four per cent for twins. It is noteworthy that over 50 F-4 pilots returned safely from combat with one engine out. The company also emphasises that from an all-aspect average, the projected area of a twin-engined powerplant is no more than for the equivalent single.

Looking into the future, Northrop projects for the mid-1980s a loss rate per 100,000 hours of 10 for singles and six for twins, assuming no difference in non-engine- Erelated attrition. It is pointed out that historically pilot fatalities number approximately 40 per cent of destroyed aircraft.

However, it is also worth noting that the Northrop data shows the A-4 loss rate in Vietnam is not significantly different from that of the F-4 and A-6, despite the fact that the engine-caused loss rate for the A-4 was four times the twin-engine average. The overall single-engine average was spoiled by the F-8, F-105, and the F-100. It is also noteworthy that the figures for loss following a hit in the engine compartment omit SAM strikes: in this case, the number of engines was not a significant factor! What this mass of information really amounts to is largely subjective. In the writer’s view, the average single-engined aircraft probably has a peacetime attrition rate in the region of 50 per cent higher than that of the equivalent twin, but (as) Dilger concluded) both rates will be very low by the Mirage replacement service timescale.

To get some feel for the magnitude of TFF losses, one might guesstimate that by the mid-1990s (ie halfway through the life of the type) the peacetime attrition rate per 100,000 flying hours should be in the region of 3. 33 for an average twin and 50 per cent more, ie 5. 00, for the corresponding single. These numbers are somewhat higher than Dilger predicted, but it is debatable whether the rate of improvement he found can be maintained.

If, out of the 75 aircraft to be bought by the RAAF, a force of 48 is operated at an annual rate of 250 hours per aircraft, then over a 20-year life the fleet will accumulate 240,000 flying hours. Accepting the above rates of attrition, some 12 single-engined aircraft or eight twins would be lost.

The difference of four aircraft in peacetime attrition hardly appears significant in comparison with other factors such as cost, performance, and equipment standard. If it is assumed that wartime attrition would mainly be due to missiles of a high kill probability, then in principle there appears to be no great attrition argument in favour of twin-engined fighters for the 1985-2005 time frame …

However, before leaving this generalised discussion of loss rates, there are two points worth noting. Firstly, whatever the statistical pointers, it would be the height of irresponsibility to buy any aircraft (single or twin) unless one was satisfied that a specific engine type would be reasonably reliable by the time it was introduced into service in Australia. Analyses of past records are useful guides only if future equipment corresponds to the state of the art for its generation.

Secondly, having an air force is not the same as having a licence to kill, hence the lives of the pilots are really the decisive factor, rather than the cost of attrited aircraft. If local operating conditions are such that loss of an aircraft would normally carry a high risk to the pilot or to people on the ground, then a twin-engined aircraft is justified, however slight the statistical margin in its favour. This consideration was undoubtedly a significant factor in Canada’s choice of the F-18 since the CAF is very much concerned with air defence over the Arctic. Conversely, as long as the RAAF’s primary consideration for TFF is operations either close to the coast or over comparatively sparsely-populated yet survivable land, then there is no special case for a twin-engined fighter in Australia.

The Question Of Cost

When US Secretary of the Air Force John L. McLucas announced the selection of the F-16 as the winner of the Air Combat Fighter competition on January 13, 1975, he stated that the flyaway cost of the aircraft was expected to be eight per cent less expensive than that for a production model of the Northrop YF-17.

Since that time both aircraft have changed considerably. and it is probably fair to say that the YF-17 has grown proportionally more in weight and complexity in meeting US Navy requirements, on its way to becoming the McDonnell Douglas F-18A. The cost differential would accordingly be expected to have increased, and this appears to be borne out by the latest US budget data available.

Quoting from the US Government Systems Acquisition Report, the average unit flyaway cost in FY75 terms was -estimated at $US4. 69m for the F-16 and $US6. 82m for the F-18. This suggests that the F-18 is approximately 45 per cent more expensive than the F-16.

At first sight, a 45 per cent price differential may appear surprisingly high in the context of two fighters that originated in the same competition, but this is not out of keeping with their difference in size. Take-off weight in air superiority configuration is 23,357lb (10,952kg) for the F-16, and 35,400lb (16,055kg) for the F-18, making the latter 52 per cent heavier. Maximum take-off weight is 35,400lb (16:055kg) for the F-16 and 50,000lb (22,675kg) for the F-18, making the latter 41 per cent heavier. Allowing for the greater complexity associated with two engines, a flyaway price differential of 45 per cent does not appear too unreasonable.

On the other hand, there are many items of equipment (eg the ejection seat) that are unchanged when the weight of the aircraft is increased. In addition, the GE objective was to produce in the F404 of the F-18 (the production derivative of the YJ101 in the YF-17 an engine that was so fundamentally uncomplicated that two could be purchased for the price of a single P&W F100 in the F-16. Such considerations would support the idea of a much lower cost differential than 45 per cent. Conversely, the R&D levy on the F-18 is reportedly almost twice that on the F-16, ie $US1.2m compared to $US0.64m.

In trying to obtain some feel for the likely order of price difference, it is worth noting that GD estimated that in the Canadian case, depending on assumptions regarding the location of assembly and the payment of local sales taxes and the R&D levy, some 29-34 per cent more F-16s could be purchased for the fixed funding available.

The idea of a basic 30 per cent difference in cost is probably not too far off the mark. However, it is worth emphasising that no one will know the actual cost for Australia until the last aircraft has been charged for, since the price of FMS aircraft varies with the rate of production. In order to minimise the effect on the defence budget, deliveries of the 75 RAAF aircraft will be spread over a six-year period, from 1984 to 1990. During this time the ratio of production rates for the F-16 and F-18 will naturally fluctuate, and the ratio of unit prices will vary accordingly. At this stage the best estimate may be that the cost differential between the two will probably be in the range of 20-40 per cent.

The F-16 Case

The case for the F-16 therefore starts off with the fact that, in buying a fixed number of aircraft, this option would save the taxpayer money. Even 20 per cent of two billion dollars represents a lot of hospitals, schools, highways, etc. If retained within the Air Force slice of the budget, it adds up to a lot of airfield improvements in the north, radar systems, and ordnance stocks. Spilling out to the outer services, it would represent a major expansion in Army missile defences, or in the Navy’s ability to protect vital trade routes.

The second major point in favour of the F-16 is that while meeting RAAF air defence and air-surface mission requirements, it excels in air superiority, the one role in which it is absolutely vital to out-class enemy equipment. The F-16 is small and hence difficult to see and to hit, it is highly manoeuvrable (9g, rather than the 7. 5g of the F-18), with a high-g cockpit. In the 1974 playoff for the USAF Air Combat Fighter it out-performed its rival in agility, acceleration, and turn rate, especially in the transonic and supersonic regimes, and there is no published data to suggest that the situation has changed since.

Other advantages of the F-16 include the fact that it is much further down its learning curve (over 1000 of the series would have been delivered to six operators before the first aircraft for Australia) and that its structure contains less advanced composite materials, which still have little practical serviced experience. In addition, the air intake is set well ahead of the nose-gear, whereas the intakes of the F-18 are well back under the wing, where in land-based operations they may scoop up spray and FOO from the nosewheel.

The main criticism levelled at the F-16 has related to teething problems with its F-100 engine. There can be little doubt that if this had begun life in the F-16, rather than the twin-engined F-15, then there would have been a significant number of aircraft lost. In 1977 alone there were some 26 F-15s that made single-engine landings! However, major teething problems arise with the best of engines. The Rolls-Royce Merlin that won the Battle of Britain and powered the night bombers that laid waste Hitler’s Reich was a disaster throughout its early testing. The GAO (General Accounting Office) Report on the F-16 in 1978 included various estimates of the likely loss rate due to engine problems, all of which were markedly higher than the USAF goal of 1/100,000hr. However, it is worth noting that no operational stall-stagnations (the major problem) have occurred on modified engines in the F-16, that stallstagnation is a high altitude phenomenon, and that if the engine failed to relight, the F-16 can be landed dead-stick from a height of 7000ft at the high-key position.

Notwithstanding these assurances, the RAAF would obviously want cast-iron evidence that the engine modifications have had the desired effects, just as the Service would want to be satisfied with all the safety aspects of flying a basically unstable aircraft that relies on FBW controls.

The F-18 Case

The case for the F-18 is basically that it provides more capability in the air defence and air-to-surface roles, offsetting its higher price. Being a much larger aircraft than the F-16, it is proportionally less penalised by heavy, high-drag warlords, especially the two Sparrow missiles which it carries semi recessed in the lower fuselage. Having a later radar with a high PRF facility, it detects targets at longer range in the head-Gn aspect (although GD claims the F-16 radar gives better range than the F-18 when the target is tail-on).

The other advantages of the F-18 include the fact that it has spare volume for avionics growth and additional equipment, that it has natural stability, mechanical back-up for the FBW system, carefree engine handling, and effectively no AOA restrictions. In terms of introducing new technologies to the Australian industry, it makes more use of advanced structural materials than the F-16, but on the other hand its engines are technologically simpler.

The latest GAO report on the F-18 criticises growth in cost and weight, and a degradation in performance, viz in acceleration time, combat ceiling, radius of action, and landing speed. Excessive nose wheel unstick speed at take-off has been cured by the unusual procedure of Boeing in the vertical tails tor rotation. Although the F-18 is almost certainly more of a multi-role aircraft than the F-16, the report makes it clear that the US Navy has no requirement for all-weather air-to-surface attack capability. However, the main thrust of this report is to ensure that development is completed before full production is authorised.

Conclusions

The RAAF is in a uniquely difficult position in aiming to purchase a moderately priced all-can-do combat aircraft. Larger air forces avoid the problem by using several types of aircraft. Small air forces try to form alliances (as in NATO) and divide the missions between them, so that each service needs only one fighter type. Australia is arguably not wealthy enough for several different categories of fighters and is too remote to make use of an alliance in this respect.

Further complications arise from the fact that the RAAF – rather than being allowed to buy the well-established, highly capable F-15 – has to choose between the new, downmarket F-16 and F-18, neither of which has yet completed development to the standard required for Australia. Neither type is dirt cheap, yet both would require retrospective modification (variable-geometry intakes) if they were ever to achieve the top-end performance that the Mirage provided 20 years earlier.

Comparing these two aircraft, the F-16 would provide a really substantial economy of funding and would excel in air superiority, in which role it is vitally important to excel. The F-18 would provide a better capability in air defence and air-to-surface attack. It also has two engines, although the advantage of this appears significant only if the regular overwater operation were to be called for. The conclusions to be drawn are again subjective. In the writer’s view, the decisive considerations may very well lie outside the TFF contest. In the event that the RAN is to be denied Melbourne replacements, ie, ships able to carry high-performance fixed-wing aircraft, then the balance must be tipped in favour of the twin-engined F-18 to provide extended air cover for naval units and convoys within flight-refuelled reach of land bases.

However, if the RAN can provide its own air defence from flat-tops, then the critical consideration becomes the RAAF replacement for the F-111. It currently seems most likely that, if it is financially possible, the service will replace this aircraft with, whatever the USAF develops to supplement its own F-111s in Europe. If the Americans choose a derivative of the F-15 for this role, then it would make considerable sense for the RAAF to have the F-16 to replace the Mirage, in order to provide engine commonality on the F100.

If a mix of F-16 (Mirage replacement) and F-15 (F-111 replacement) is seen as too expensive for Australia, then the F-18 might be· accepted as a compromise, replacing both Mirage and F-111. However, if the USAF develops the Northrop F-18L to supplement its F-111s, then it would make sense for the RAAF to have this aircraft in both the fighter and strike roles, regardless of the fact that it is now officially out of the running for TFF.

To summarise, the F-16 and F-18 are significantly different in cost and capability, and both involve a certain degree of risk at this comparatively early stage of their development. The choice between them cannot rationally be made without knowledge of how Melbourne is to be replaced and without some indication from the USAF on what aircraft may be expected to be available to replace the RAAF F-111 later this century.