Features - Technical
OCTOBER 6, 1998
BY PETER WRIGHT
Fortunately the sight of Ron Dennis and Jean Todt (or it could be Adrian Newey and Ross Brawn), sitting on the pit wall in front of radio control consoles, fighting for the Championship is unlikely to ever occur. Instead, ever since driver-aids were banned, the driver and his technical contribution towards developing the current Formula1 cars, engines, tyres, and complex systems, has become more important. We are entering the era of the driver as "test pilot", epitomised by Williams' decision to select Alex Zanardi to lead it into the next millennium. Re-forming in the wake of Adrian Newey's departure to McLaren and the loss of Renault, Patrick Head has recognised the importance of the driver's input in regaining competitiveness, particularly while they work with BMW to develop their first non-turbo Formula1 engine. Selecting Zanardi was based as much on his reputation for a technically intelligent approach to development and testing, as his proven aggressive driving and race craft.
Tom Wolfe, in his best seller "The Right Stuff", tells the story of how NASA made the mistake of believing that they could dispense with humans to pilot space craft. The predictability of events is not so high that it is possible to do without the unsurpassed analytical qualities and adaptability of the trained human brain. Roland Beamont, one of the UK's best known test pilots, describes in his book "Testing Years" how aircraft have changed during his professional lifetime from "....Ésimple biplanes and early monoplanes of the 1930s to the complex, highly automated and densely equipped machines of the 1970s with computer-based stability-augmentation and control systems which tend to isolate the pilot from direct control of the aeroplane. Yet while he is still required to command the aircraft, the end-product of whatever control system is employed, be it manual through cable or control rod direct to the control surfaces or by dynamic hand-controller input to a fully-computerised electronic control and stability-augmentation system, must be aircraft response matched to the human range of response capabilities. In other words it has to 'handle'." Add 20 years to the dates and change "aircraft" to "racing car" and he could be describing Formula1. Beamont adds: "...an intensive effort has been made to quantify 'handling' criteria so as to be able to express in numbers the design qualities needed for pilot-acceptability in new designs. Somewhat surprisingly in a technological age this effort has been only partially successful. While much is now known of the limits of control forces and response and damping rates acceptable to pilots, the translation of this knowledge into engineering fact seldom achieves first-off conditions of control harmony in a new type which are completely acceptable to pilots as a good standard. Hence one of the basic needs for 'flight development'." The situation in motor racing is not so different today.
Prior to the development of data systems suitable for running on a racing car, the only conduit for information about the car on the track was through the driver. A few drivers displayed great ability in analysing the responses of engine, chassis and tyres, conveying this information to their engineers and contributing to the development of solutions to the numerous problems of making a car go fast. Jack Brabham, Bruce McLaren, Graham Hill and Jackie Stewart spring to mind as particularly good examples. It is probably no coincidence that all of these drivers started their own GP teams. Brabham sometimes short-circuited the communication system by jumping out of the car and grabbing some spanners to make a set-up change before driving out onto the circuit again. Words were not necessary!
As the flow of data from chassis, engine and transmission has swelled into the multi-Mbits/sec. range and almost everything that changes is measured and recorded for analysis, the role of the driver in reporting what car and systems are doing has changed. Some engineers have considered that the driver as a source of data and other information is redundant. It is true that sensors can measure many things far more accurately than a human can and that many parameters change dynamically at a rate undetectable by the human sensory system but, to conclude that the data system can tell you everything is far too optimistic. Take understeer, for example.
When a car corners at, say 3g, the lateral force generated by the tyres is around 17,500N, of which 7,500N comes from the front tyres, and 10,000N from the rears. Because the car bounces up and down on it's relatively un-damped tyres, these forces vary by as much as ±30% at between 5 and 10 cycles per second i.e. ±2,250N at the front and ±3,000 at the rear. Understeer makes itself apparent to the driver when the front tyres can generate only 99.5%, or less, of the force needed to balance the rear - a shortfall of less than 40N. If the engineer could run alongside the car and push on the side of the nose with his hand, producing just 4Kgf, he would stop his driver complaining of understeer.
For the data system to detect this shortfall in tyre lateral force, via its accelerometers, rate gyros and load cells, it must be able to resolve 0.5% of a load that has a ±30% signal to noise ratio. Even the best post processing analysis will not achieve this, but the driver will go on about the understeer all day until it is tuned out.
If the data system cannot measure the understeer force deficit, it cannot resolve where it comes from - inner or outer tyre, tyre load (weight transfer, aerodynamics, roll stiffness or damping), tyre camber, tyre temperature or some other characteristic. However, if the driver understands the suspension, tyres and aerodynamics and is adequately tuned into the car, he is uniquely placed - in the cockpit with his hands and feet on the controls - to assist with the process of identifying the source of the problem and hence reduce the number of trial and error test runs needed.
As a young driver progresses through the lower formulae towards Formula1, instead of spending 75% of his time racing and 25% of it testing, he will spend 25% racing and 75% testing. While the prime objective is still to drive faster than his competitors, the majority of his job will involve driving to a prescribed test schedule and observing and communicating everything that happens while the car is on the track. Most testing falls into one of four categories:
1. Reliability/durability testing.
2. System characteristics development and calibration.
3. Modelling and laboratory test validation.
4. Tuning and set-up.
Each category requires a different approach from the driver, but before he can attempt any of this he must learn how to operate the extraordinarily complex machine that a Formula1 car has become. Almost all the controls except throttle and brake, plus the displays, have ended up on the steering wheel. This is not only a very densely packed control panel, but it rotates as well! A dozen buttons and rotary switches, three or four paddle switches behind the wheel, RPM lights, warning lights and an LCD screen, on which various menus can be displayed and selections made, are packed into the centre boss and spokes. Not only must the driver memorise all the functions, including the latest updates, but he must practise precise operation while travelling at 280 kph between corners, without missing his braking point. Having mastered this ultimate play-station he can turn his mind to the test engineers' tasks.
For reliability and durability testing the driver must drive as if in a race. A race distance, without competition but driving at competitive speeds, is something that many drivers find not only hard, but also boring. However, if the car is not driven hard enough the test may not be representative and may fail to identify a particular weakness. The data stream will give the engineers most of the information they require about temperatures, pressures etc and performance, but subtle changes in characteristics - vibration for instance - can often only be picked up by a driver in tune with his car. Precise repetition of lap times marks a good test driver, and this is particularly important when testing tyre durability.
The development of new or improved systems is an area where the technical driver can make a major contribution, contrary to what many engineers believe. The system designers and R&D engineers will have developed the system in the laboratory, simulating and rig testing it until they believe they know all its characteristics and quirks. Perhaps what they forget is that uncertainty increases exponentially with the complexity of a system. While the performance of some systems can be determined completely by data measurements, if they affect handling then the driver is automatically involved - as Roland Beamont states so clearly.
A good example of how a driver can influence systems development is the work Ayrton Senna carried out with Honda on throttle progression. Senna was extremely focused on the importance of driveability as a major influence on his ability to maintain the rear tyres right at the point of spinning and hence at maximum grip. When Honda introduced their 3.5l NA V10 engine and the broad, flat torque curve of the turbos was lost, driver and engineers concentrated on the development of an extremely complex mechanical throttle linkage until Senna had tamed the characteristics of the V10 to his liking. This focus on a characteristic that is at the centre of controlling the car led Honda to pioneer drive-by-wire on Formula1 cars, taking them beyond the limitations of a purely mechanical system and enabling the even peakier torque characteristics of their powerful V12 to be exploited.
R&D departments generate theories which they develop and test in the laboratory, and as computer models. As with any simulation, it can only include what is known, the final proof comes from track testing. Aero. maps, calculated from wind tunnel data, have to be checked; computer control software must be validated in a real environment; vehicle dynamics modelling requires input information as well as validation; tyre models and rig test results are renowned for giving only a part of the tyre's characteristics. R&D track tests require the driver to drive precise profiles - speed, RPM and power settings, steering and brake inputs etc - so that the performance of various parts of the system can be calibrated and the input-response characteristics logged and analysed. An understanding by the driver of what the boffins are trying to do helps enormously in this type of testing and increases the driver's contribution to the precision and feedback involved. It is not most driver's favourite type of testing, but involvement can alleviate the tedium of a day's driving off the limit.
Tuning is the systematic testing of the car's sensitivity to various set-ups, including engine characteristics, aerodynamics, tyres, suspension geometries and spring/dampers, brakes, differentials etc. It is the hardest type of testing to be totally objective about and it requires repeated baseline tests and a constant driver effort. His feedback is just as important as data. A full and practised understanding of how to set-up the car for a given circuit and conditions pays off during the limited practice periods at GP's. The ability to communicate to the race engineers about the car's strengths and weaknesses will have a major bearing on grid position and race results.
It is also during this type of testing that the base data on which race strategies are founded is gathered by both driver and engineers. The effect of fuel quantity, tyre degradation, top speed versus downforce trade-off, and wet and intermediate tyre performance are all best learnt about away from a race weekend, leaving conclusions to be checked out at the event itself.
The complexity of today's Formula1 car is in some ways greater than the jet fighters of the 1970's (the aircraft had limited electronics and computers) and the understanding of its aerodynamics and vehicle dynamics somewhat similar. Test pilots must have an engineering degree as a minimum qualification, followed by extensive training in order to become qualified to test-fly and develop aircraft. Much of that training is spent in the classroom and a fair portion of the flight training is spent in variable-stability aircraft, in which the instructor can expose the pupils to all sorts of undesirable flight characteristics so that they learn to identify them correctly.
To become a top racing driver today requires a youth spent Karting. If a potential driver is not winning Junior Championships by the age of 10, he has no chance. School education must suffer and the ability to combine university with a serious F3 campaign is virtually impossible. Some drivers accumulate technical experience and knowledge during a long career driving a wide variety of racing cars. Jackie Stewart's skills, gained in winning three World Championships, were used extensively by Ford and Goodyear after he retired, to develop and test drive their road products. Nicki Lauda started his own airline and flew as a pilot for Lauda Air. When one of his aircraft, a Boeing 767, crashed for no apparent reason while climbing out of Bangkok, he spent many hours in the simulator trying different flap and spoiler configurations and failure modes, in an attempt to determine how control could have been lost. This highly technical, simulated test-flight work contributed to the discovery that one thrust reverser could be inadvertently deployed in flight and would cause the pilot to lose control.
Now however, Formula1 teams want their drivers technically knowledgeable and experienced as test drivers before they join the team. Though they are not yet sending young drivers to university on scholarship, they have invested in teams in the junior formulae, from Karting to F3000, to bring up potential stars in the techniques of Formula1. The opportunity to learn from the experienced race engineers and to practice setting-up cars in a methodical and systematic fashion is going to be a long-term benefit to both drivers and teams.
Of all the drivers I have worked with, Alex Zanardi is the one who was most able and most prepared to get involved with the development of the car. With his technical background, he quickly grasped the concept behind test programmes and was able to carry them out to precise instructions. His feedback was detailed and thorough and engineers were sometimes stretched to note it all down as he delivered a point-by-point description of some engine or chassis characteristic. It is also clear that he has made a major contribution to the development of the Reynard-Honda-Firestone Champcar, raced by the Ganassi team in the last three years. I shall watch with interest to see how he gets on with Williams and BMW, and whether he can help get Williams back to the front of the grid. If they succeed, the era of the techno-driver will have started.