As I write this there are only just over two weeks to go before the first of the Formula One winter test days at Jerez, and while many a race enthusiast might be concerned about the changes in the sound the new turbocharged V6 power units will make, as an engineer I am thinking about things that are arguably perhaps much more fundamental.
So at a time while many may be looking at what has been brought in for 2014, as an engineer I am looking more to what has been left out.
Perhaps the most fundamental characteristic of any engine – be it diesel, gasoline, two- or four-stroke – is its combustion system, and while the Formula One power unit regulations for 2013 specified only port injection fuelling systems, for 2014 this restriction is left out. So while previously in the formula direct injection has been specifically excluded, starting in March this year it must surely be the only way to go. Two little letters, DI – such a fundamental change to the sport but totally hidden from the rest of the world!
In motorsport, however, direct injection is nothing new. Winning the world championships in 1954 and ’55, the Mercedes Grand Prix racer used a mechanical system to inject the fuel through the side of the cylinder in its straight eight-cylinder M196. Reintroduced into automotive technology in 1996 by Mitsubishi using electronic means, in 2001 Audi was next and brought its twin-turbo 3.6 litre V8 to the technology, winning Le Mans that year. By this time of course, many other manufacturers began to see the advantages of increased power (up to 5%) and better fuel consumption (up to15%). But like many things in life, such improvements don’t come easily, and while the potential benefits of improved cylinder filling and better mixture preparation are highly attractive, to achieve them takes a lot of painstaking development work.
In a port-injected engine the fuel can be injected over much of the four-stroke cycle, with injection ending as the intake valve closes. At, say, 10,000 rpm this injection period may be something of the order of 12 ms. In a DI unit, however, injection can’t even start until the inlet valve opens, and then only when there is no chance of the fuel escaping through the exhaust valve during valve overlap. Start of injection will therefore be at or close to exhaust valve closing for that cylinder. After that, the fuel has to be given sufficient time to evaporate to create a combustible mixture before it can be fired by the spark. At this same 10,000 rpm, the time for this to happen falls to somewhere nearer 1.6 ms.
This in itself is not a problem, since under the new regulations fuel rail pressures are considerably higher (up to 500 bar). However, getting the correct air-fuel mixture at the correct time in the engine cycle sequenced to the position of the piston, and ensuring that this mixture burns quickly and completely, takes a great deal of understanding of the airflow in the cylinder. This, as well as the progression of the flame front across the combustion chamber, requires much more knowledge of the in-cylinder flows at any particular instant than with our port-injected engine. Little wonder then that in the past two years or more, countless hours of CFD work will have been undertaken to model both air and fuel mixing to get to where we are now. Increasing the engine speed to the maximum of that now allowed (15,000 rpm) reduces this time for evaporation and mixing still further, and you now begin to see the nub of the problem.
By the time you read this, how well individual engine manufacturers are faring may be all too visible, but there is one thing of which you can be certain. If, as of 28 January 2014, we can say the starting pistol has been fired, there will be a lot more CFD activity in darkened rooms around the world until the first race in March. The Formula One season may not have officially started yet, but the race to extract maximum performance out of 100 kg of the fuel allowed certainly has. As of that date Formula One suddenly got interesting again!
Fig. 1 - Such little time!
Written by John Coxon