F1 Engine Power Secrets

By Ian Bamsey
appeared in June 2000 RACER magazine

 

     A time traveler from 50 years ago would find today's Formula 1 cars radically different, but would be equally surprised at the relative lack of change in engine technology. Turbochargers have come and gone and there hasn't been a switch to two-stroke or rotary, scotch yoke engines - let alone to gas turbines or something not even invented in 1950. The good old four-stroke internal combustion engine powered the very first Grand Prix car in 1906 and lives on in a form instantly recognizable by any time-traveling engineers from 1950.

     What would impress them is the performance of the current breed of 3.0-liter naturally aspirated engine, that pumps out in excess of 260hp per liter, when in 1950, 80hp per liter was a competitive figure. With careful attention to breathing and the development of highly potent fuel-including the generous use of nitromethane-the 2.5-liter Vanwall of 1957 attained 120hp per liter as a "flash" dyno reading. Its more representative 115hp per liter race output-on a marginally less eye-watering (but still drag racing style) mix-was still then an all-time high for a naturally aspirated Formula 1 engine.

     The following season exotic fuel was outlawed and, at a stroke, power outputs fell below 110hp per liter. However, Fl being Fl, they were soon creeping up again. Ten years on, development was such that, running on "pump" gasoline, 1967's Cosworth DFV produced 133hp per liter. Astonishingly, the current Cosworth 3.0-liter, Formula 1 engine, fed comparable fuel and likewise naturally aspirated, produces about twice that!

tn_dfv.gif (3920 bytes)
Click on picture for a larger view of the DFV engine

     How can performance have doubled in 33 years, when there's been no major change in engine concept? Indeed, the DFV of 1967 set the pattern for the contemporary engine with its "over-square" cylinder dimensions, its pent-roof shape combustion chamber, its narrow included valve angle, its four valves per cylinder operated by double overhead camshafts and its clean porting. The DFV was not radically different from previous engines, Its significance was that its detail design took advantage of the potential of four, rather than two, valves per cylinder for enhanced breathing and burning-the fundamentals of effective combustion.

     The power output of the four-stroke internal combustion engine is a function of the torque seen at its flywheel and the speed at which that flywheel spins. Power per liter per 1000rpm as measured on the dyno is the so-called brake mean effective pressure (bmep) that indicates torque. The Vanwall of 1957 produced its 120hp per liter at 7300rpm, at that engine speed giving 16.44hp per liter per 1000rpm-a nitro- boosted peak power bmep reading of 213psi (14.7 bar). The DFV of 1967 produced its 133hp per liter at 8500rpm- 15.65hp per liter per 1000rpm, a peak power bmep reading of 203psi (14.0 bar). This fall in specific torque recognizes the far less potent fuel used in the DFV.

     The increase in performance of the current 3.0-liter engines is purely a function of faster flywheel (crankshaft) speed. To attain 800hp, a current Fl engine must turn in the region of 17,000rpm, which means its peak power bmep roughly equals that of the 1967 DFV. In fact, as crankshaft speed rises, the tendency is for bmep to fall-the combustion event has to take place in a correspondingly shorter time and frictional and other losses increase disproportionately.

     On the other side of the coin, the fact that the current Fl engine can match the peak power bmep of the DFV of 1967 running at half its speed is a tribute to considerable development devoted to overcoming the inevitable losses that occur with rising speed. It should be noted that some internal losses quadruple with the doubling of running speed. Measures to counteract these losses include a drastic reduction in bearing sizes, the development of high-performance coatings for the bearings, the piston and liner and so forth, and a conceptual revision of the oiling system. For example, the contemporary engine has its crankcase divided into separate chambers to reduce losses to windage.

     Another major development since '67 has been the advent of engine management systems. Contrary to popular belief, the precision of computer-timed ignition and fuel injection does not automatically increase maximum power-at least in the case of a pure race engine-but it does help keep everything running on cue as crankshaft speed rises.

     Of course, the biggest challenge has been holding everything together as reciprocating and rotating parts are worked ever faster, and generate increasingly fierce loadings. Even at "only" 12,000rpm there are seven tons going up a con rod, which responds by growing longer, then 12 tons going down it, which unavoidably shortens it somewhat!

     For a long time, the biggest headache of all was keeping control of the valves given the less than ideal characteristics of steel coil valve springs. Using lightweight titanium rather than steel valves helped, but titanium does not make a suitable spring. The breakthrough came with pneumatic valve actuation, which offers precision of control, even at 17,000rpm, and consequently is now universal in Fl.

     These so-called air springs opened the door to today's crankshaft speeds from a V10 engine, with its 40 large valves. The 32-valve DFV V8 had an 85-67mm bore, while today's Vl0s have bore sizes in the region of 92-96mm, with correspondingly larger valves (albeit titanium rather than the steel employed in 1967).

tn_valvesystem.jpg (5377 bytes)
Click on picture for a larger view of the pneumatic valve system

     Although the piston is larger in diameter, it is smaller in depth and significantly lighter, thanks partly to materials development. Indeed, it is materials development that has made possible the recent push to a peak-power speed of 17,000rpm, even though the crankshaft is still steel and the con rods are still titanium.

     Painstaking development-together with computer aided design-has produced crankshafts and con rods than can handle far higher loading despite employing significantly less material. The same can be said of pistons. Although an air spring engine employing only "traditional" materials has reached a peak-power speed of at least 16,000rpm, most crankshaft speed gains come from using alternatives to traditional aluminum alloy pistons.

     First in this field was Ilmor, producer of the Mercedes V10s used by McLaren in recent seasons. Since 1998, Ilmor has manufactured pistons from an aluminum-beryllium alloy, thereby reducing their weight by a third, possibly more, and gaining enhanced thermal conductivity. The cost of this alloy, and the fact that fine beryllium dust particles arguably constitute a health hazard, has led to an effective ban on its use, imposed by the FIA. Under pressure from McLaren and Mercedes, however, this ruling, for which Ferrari lobbied hard, has been postponed to the end of the current season.

     Rather than specifically outlawing aluminum-beryllium (as requested by Ferrari) the new ruling for 2001 prohibits the use of any metallic material with a specific modulus of elasticity in excess of 40 Gpa/ (gm/cc). This leaves the door open for a newly introduced Metal Matrix Composite (MMC) material developed for Formula 1 piston manufacture by liner and piston supplier Perfect Bore.

     This aluminum and ceramic alloy offers a weight-saving approaching that of aluminum-beryllium, together with excellent thermal characteristics. Unlike aluminum- beryllium, says Perfect Bore, it has a lot of potential for inlet valve as well as piston manufacture, promising significant gains over titanium valves.

     Another application for Perfect Bore's latest MMC is the cylinder liner. Aluminum- beryllium has been used to produce lightweight wet liners, but the latest trend is the use of a super-thin dry liner within what is effectively a linerless block.

     A key feature of 1967's DFV was its packaging-very compact by the standard of the day-a significant benefit for the chassis designer, but compared to a current Vl0, it looks huge. Today's clutch is much smaller in diameter (4.5in. now vs. 7.5in. then), the crankshaft is set significantly lower and the whole package is more tightly knit. While the cylinder count is the ultimate constraint, with some lateral thinking last season Cosworth reduced the size and weight of its engine by dispensing with traditional wet liners.

     The linerless block permits a beneficial reduction in cylinder bore spacing, but there may be some advantage in retaining some form of (dry) liner. Not only can this be replaced-rather than the entire block-in the event of internal damage, it permits a wider selection of cylinder wall coating, which helps reduce friction.

     With the help of ongoing materials development, how much faster can the 3.0 liter Formula 1 engine run? And can it get any smaller than the current Jaguar-badged Cosworth V10? If you don't have access to a time machine, you'll just have to wait and see!

Home