Features - Technical
NOVEMBER 6, 1998
A Christmas Reverie
BY PETER WRIGHT
1. The car must meet the FIA definition of a land-born automobile
2. It must fit within a box measuring 5 metres long (to prevent gridlock at hairpins and for cars to fit existing Pit complexes) x 2 metres wide (to prevent a very wide car blocking those behind) x 1 metre high (to fit under bridges etc.).
3. It must be open-wheeled and open-cockpit so that it is recognisable as a Formula1 car.
4. Fuel must meet Health and Safety regulations i.e. commercially available motor fuel - petrol, diesel or LPG.
At this point I woke up in a panic, realising that there were just 26 months to the first race. What sort of car would one build? How much power? How much downforce? How many wheels? 2-WD or 4-WD (or more)? Before these questions can be answered, there are some fundamental performance issues that must be addressed. Firstly, what are the human limits? With sustained cornering and braking levels over 5g possible, will a driver be able to drive at these levels for two hours? The first thing to be done is to visit the Institute of Aviation Medicine at Farnborough and book their centrifuge for a programme to assess the task-performance of the drivers under high lateral g. At the same time a literature search on human performance at high g-levels should be conducted. After all, building a car capable of 7 or 8g might be a complete waste of effort.
Secondly, I would visit the tyre company (or companies) to find out what they plan to offer. Sustained high g's will put a lot of energy into the tyres, causing them to overheat. There will be a practical limit, depending on the size of the tyres and the number per car - 4 big ones or 6 or 8 small ones. If there is just one tyre company involved, they will probably go for a tyre size that they are already equipped to make, rather than try and satisfy every variation envisaged by the Technical Directors. The energy put into the tyres is a function of the mass of the car, the g-levels generated at any given speed and the slip angles and slip ratios that the tyres run at. G-levels and speeds need to be as high as possible at all times; peak slip angles and slip ratios cannot be too low or the driver (or control system) will not be able to measure and control them. This leaves mass, which must be as low as possible. The challenge is to choose a specification that gives the maximum performance that the driver can endure and that weighs as little as possible, so that he can use the performance for as long as possible before the tyres give up. It is no good building a car with every "bell and whistle" that the FIA has banned in the past and, as a consequence weighs in at 1000 kilograms with enough fuel to race. It is the mass consideration that will dictate how much power to install in the car and how to transmit it to the road.
The realistic engine options are: unlimited capacity NA piston engine; unlimited capacity and boost turbo-charged piston engine; or gas turbine. The best NA engine, the Illmor designed Mercedes, produces 800 bhp and weighs 107kilograms bare, or approximately 150 kilograms with cooling and oil systems, exhausts and the six speed gear cluster necessary for a piston engine. The best turbos would probably produce 1000bhp for a similar weight, (the reputed 1200bhp produced by the BMW engine in the 1985 Brabham BT54, used "rocket fuel" that would not pass today's Health and Safety regulations and the engine lasted only a few laps). Production turbo-shaft engines, built to power helicopters, generate around 1500bhp for a weight of 150kg.Thus the three options can be rated:
NA piston 5.3
Turbo-charged piston 6.7
But of course it is not that simple.
The power-to-weight advantage of the turbo shaft has to be weighed against the problems of throttle response inherent in a gas-turbine engine. These engines are designed to run at a specific speed where they are most efficient, and do not have to change speed very quickly. They are also capable of increased power, accompanied by a reduction in the life of the temperature sensitive turbine blades. The figure quoted is for engines designed to run for over 5,000 hours and the potential for a much greater power-to-weight ratio is available.
For these, and other reasons that will become clear later, I would go for a turbo-shaft engine and set about solving the response problem. Thus the next visit to be planned would be to the gas-turbine experts: Rolls Royce, Pratt and Whitney Canada, Allison, General Electric etc to see what they can offer and how to best address the problems of response. This assumes of course that they would be willing to create racing divisions and put their best engineers into them!
My guess is that an engine capable of around 1500 bhp would be the optimum, but that it must be easily de-rated for the race in order to keep fuel mass requirements within the tyre capabilities. Race power of around 1000bhp, with qualifying and overtaking power of up to 500bhp more should be practical from a small, light turboshaft engine, enabling the mass of the car to be kept below today's figure of 600kilograms with driver. Fuel consumption will depend on grip levels and how long the engine can be kept on full throttle. At 1000bhp race setting, the fuel consumption is likely to be to 50% more than today's cars, with their 800 bhp and low grip.
To deliver 1500bhp to the ground would need either 4-WD or a great deal of low speed grip. Today's cars achieve maximum acceleration (1.4g) at a speed of around 120kph, where aerodynamic download allows full throttle to be applied. However, if fan-car technology can generate download equivalent to the weight of the car at low speed, then things improve rapidly. As the acceleration increases, more weight is transferred onto the rear axle and traction improves. A quick calculation indicates that at 80kph, with tyres generating a m=2, and the fans doubling the static weight on the tyres, two rear tyres would accelerate the car at 3g. The lowest speed corners (2nd gear) are taken currently at around 50kph and 2g lateral acceleration. With fans, this would rise to 4g and the speed to 70kph. The power required to accelerate at 3g out of these corners would be around 575bhp, including overcoming drag. By the time the car has reached 100kph (less than 0.3secs later) it will have risen to nearly 900bhp. With downforce increasing with the square of the speed, the full 1500bhp would be useable above around 120kph.
Adding a pair of wheels at the rear - as per the Williams 6-wheeler - would not help matters much as the total load on the 4 wheels would be about the same as on just 2. They could, of course, be smaller in diameter giving a benefit in drag, but whether they would justify the increased weight and complexity of the transmission and second suspension is doubtful. Driving the front wheels is an option, but requires a drive-shaft to pass through the fuel tank and past the driver to a final-drive unit in front of his feet. The added weight and complexity have not yet been shown to be a net benefit on a single seater racing car, as any torque distributed to the front wheels causes understeer in a corner. Indeed, it should be noted that 4-WD has only been successful in competition where the engine has been mounted in a non-optimal position i.e. front-engine, FWD (Audi) or rear-engine (behind axle), RWD (Porsche 959).
Electric drive, with a sophisticated control system to distribute torque to each of the four wheels, may one day be THE solution. However, such systems remain heavy and inefficient, even though they are capable of storing regenerative-braking energy. The weight would really hurt the tyres.
So, I am going to go for 2-WD, low weight, lots of low speed downforce, and traction control. The time spent in a straight line when full throttle cannot be applied is very short when accelerating at over 3g.
A turbo-shaft engine consists of a gas-generator section (a multi-stage compressor driven by a multi-stage turbine and the combustion chambers) and a power turbine. Provided the gas generator section is spooled-up to the operating RPM, there is sufficient energy being generated in the hot gases for the power turbine to generate a very high torque and this can be modulated reasonably fast. The power turbine acts somewhat like a torque-converter. (This characteristic is very different from a piston engine, which relies on the output RPM being high enough to generate adequate airflow through the engine to generate torque, and so requires a multi-step gearbox to maintain the engine in its narrow operating speed range.) If the gas-generator is allowed to slow down, throttle response is very slow. However, the driveability of a turbo-shaft engine will never be as good as an NA piston engine. Discussions with engine manufacturers would reveal just how responsive a small (by aviation standards) turbo-shaft engine could be if variable geometry turbine-inlet guide vanes were used along with Full Authority Digital Engine Control (FADEC). One thing is certain, it will not be responsive enough, so another approach is needed as well. Because a turbo-shaft produces high torque at low speed, a multi-step gearbox or CVT is not needed. A clutch would be necessary for starts, using a launch control system to control the initial delivery of torque. Otherwise the transmission would consist of a step-down gear (turbine RPM is 20,000+) and a final bevel drive. I would use the rear brakes, greatly enlarged, to modulate and control the torque to the rear wheels and so provide the driver with the response that he would want.
Freed of the 13inch wheel diameter regulation, the tyre companies would probably design tyres for wheels of around 17 or 18in, and a tyre diameter of around 28ins (as per current GT cars and F1 cars before they were limited). Rear tyre width could be as wide as 20ins and fronts around 16ins. This would leave plenty of space for BIG brakes - carbon-carbon, of course - controlled and applied by a full brake-by-wire system. With this system it would be possible to provide high-response torque control, coupled with a lower response throttling of the engine to prevent the brakes over heating. The driver would apply power early on exiting a corner, and the brake system, coupled with yaw control, would deliver torque to the rear wheels at the optimum level for cornering and acceleration. This is no more than traction control and anti-spin systems on road cars do. The differential would be a simple, open differential, as the brake system would obviate the need for any form of limited slip.
Producing high levels of aerodynamic downforce at low speed is essential for the success of this configuration. As mentioned earlier, I would employ fan-car technology but not in the same way as used on the Brabham BT46. When Brabham brought out their fan-car in 1978, it displayed superior grip at low speeds compared to the Lotus T79 ground effect car; however, above 160kph the Lotus was superior. Lotus explored a combination of venturis ground effect and fans, before all fan systems were banned. I would dust off this technology and use it, as model testing showed great potential. The arrangement is configured just like a normal, fully skirted, venturis car, but has twin fans mounted in the rear of the diffusers to draw air under the car at low speeds. As speed increases the motion of the car takes over driving the air under it and the fans do less and less work, endowing the car with the best of both worlds.
I would drive the fans from the gas-generator section of the engine, as this runs at constant speed and would benefit from being loaded-up in order to keep it ready to deliver power to the wheels (the hydraulic pump and generator would also be driven by the gas-generator turbine). The fans would absorb around 100bhp at low speeds, when there is surplus power from the engine. The fans blades would be variable pitch to enable the downforce to be regulated and to unload the fans completely at high speed for maximum power. Aerospatiale's helicopters use variable-pitch, ducted-fan tail-rotors, called Fenestrons, so I would go and see them to learn how to design a light weight unit.
The main, flat section of the venturis would be as long as possible, positioned to provide an aerodynamic balance, and would probably need a blown slot of some sort to ensure attached flow in the aerodynamically aggressive diffuser. As the engine does not need cooling radiators, the sidepods would provide the intake air to the engine as well as air for the slot. They would also provide side impact protection, but could probably be lower than current sidepods. Sliding skirts, curved to run continuously around the rear wheels (as per the Lotus T80), would be developments of those banned in 1981. They would be linked to the suspension, as on the Chaparral, or even actively controlled. A transverse flap beneath the fan ducts would seal the diffuser at low speeds, and blow back as the fan-flow was replaced by car motion.
The car might not need wings, but if wind tunnel testing showed a benefit, small chord wings, front and rear, with the full 2 metres span would be fitted. They would be very efficient, due to their high aspect ratio, and would either be variable incidence or have movable flaps, all under the control of the central computer. The aerodynamic centre of pressure would be adjusted with these wings and drag minimised at high speed. I doubt that the front wing would need to be mounted low, near the ground, as there would be more than enough, downforce and pitch sensitivity should be avoided. Top speed would be of the order of 450kph in 1500bhp, qualifying trim.
Active suspension, with full load and damping control i.e. the Lotus fully active system, would be necessary to maintain the ride height accurately and provide optimum suspension for the stiff tyres. The system would also be used to limit the total downforce at high speeds - if the car sat too low to the ground at 450kph, the local air velocity under the car could be in the transonic region which would start to introduce some interesting aerodynamic problems due to compressibility.
Brakes, as described above, would be large and full brake-by-wire to give anti-lock, thus providing optimum fore-and-aft and across-car balance automatically, and to provide traction control and Yaw control. Calipers would not need to be as stiff as with a non-power system and so beryllium could probably be avoided. I would not fit 4-wheel steer as, with the powerful rear brake system, there would be little tyre capacity left to control the car by steering the rear wheels. The driver would thus control the tyre slip angles through the steering and setting up the car slip angle - a technique he is well used to - and the Yaw control system would control tyre slip ratios. Trying to mix the two systems would probably confuse the driver, if not the engineers as well.
Other systems for consideration would include: tyre pressure control - to optimise for high and low speeds; launch control for starts; Differential GPS - so that the car knew where it was at all times to within a few centimetres, and settings could be changed appropriately (it could even ensure no speeding in the pit lane); high speed refuelling - 40 litres per second would be needed to refuel the car in the time it currently takes to change the tyres, compared to the current 12l/sec.
Wheels would eventually be CFRP as would many other parts of the car, including all gear-cases. Carbon-carbon would be used for the turbine exhausts.
The CFRP monocoque would be pretty conventional, except for the need to extend it rearwards to mount the engine. Gas-turbine engines expand with the high temperatures involved, and could not carry chassis loads. The fuel tank size would depend on a very careful assessment of the number of stops for fuel and tyres needed in a race. It would probably be of a similar size to today's cars and there would be more pit stops.
The performance of such a Formula1 car would be right at the physical limits of drivers. It would accelerate from standstill at 3g, corner at 4+g in hairpin corners and at 6+g in fast corners. Fast corners would not be taken flat, because of the excess of power, but the car would still accelerate at 2g out of a 280kph corner. Braking would be at the 6+g level at high speed, falling to 4+g at turn in to a hairpin, and would need to be with a top speed around 450kph. Just how long the drivers could stand racing at these performance levels we will never know, but they would probably need the regular relief of the frequent 4-second pit stops. Driving the car would be itself simpler than today - there would just be a steering wheel, without gear-change paddles, or adjustment knobs, a go-pedal and a stop-pedal.
The car I have proposed does not look so different from today's cars, and I have no doubt other dreamers would come up with vastly different arrangements: 6 wheelers (4 at the rear); highly turbo-charged engines; active everything, and it would be fascinating to see where unrestricted aerodynamics led to. The cars would become so fast, the driver fitness training so demanding and the gravel beds so large that in the end the FIA would have to restrict everything again!
Engine 1000+bhp turbo-shaft; 1500bhp for qualifying and overtaking. Variable turbine inlet guide vanes. FADEC.
Weight 600kg including driver
Dimensions Wheelbase 3.0m
Track F: 1.59m R: 1.49m
Transmission Single speed. Open differential.
Brakes Brake-by-wire. Carbon-carbon.
Aerodynamics Twin fan blown venturis. Full length sliding skirts, activated by suspension. 2.0m span wings F&R, with active incidence.
Wheels & tyres F: 16" x 28" tyres on 18" CFRP wheels
R: 20" x 28" tyres on 18" CFRP wheels
Control systems Drive-by-wire linked to FADEC engine controller
Traction control/Yaw control/Brake-by-wire with ABS
Variable incidence wings
Fan pitch control
Launch control clutch
Variable tyre pressures
DGPS and 2-way telemetry link to pits