King of Speed: Cosworth's CA 2.4l V8
From 1906 through to 2006, Grand Prix engine speeds rose ever higher, from less than 2000 rpm to ultimately a mind-boggling 20,000 rpm. Then the rule-maker abruptly halted the march of progress with a 19,000 rpm rev limit for 2007, subsequently reduced to the current stifling 18,000 rpm – plus, to add insult to injury, a moratorium on development. While engine evolution is back in 2014, the emphasis henceforth will be on fuel efficiency rather than outright, untrammelled performance. The glorious Century of Speed is over. The first Formula One engine to attain 20,000 rpm on track was the Cosworth CA of 2006, and it is generally agreed that no rival surpassed it as the benchmark before rev limiting was cruelly imposed. These days the naturally aspirated 2.4 litre V8 CA lives on, powering Marussia in its close fight against Renault-engined Caterham for the honour of top dog of the young teams in Formula One. Remarkably, although the CA is still on active duty, Cosworth has permitted RET full access to the innermost secrets of Grand Prix racing’s all-time engine speed champion.
Operating speed and horsepower steadily climbed during much of the 3.0 litre V10 era proceeding the switch to V8s mandated for 2006. It is widely agreed that BMW reached 19,000 rpm first, in 2002. However, engine mileage requirements were lengthened in 2004 and 2005, which had the effect of pegging the ongoing crankshaft speed rise. Representative of the top 2005 V10s was the Toyota, which ran to a maximum of 19,200 rpm and produced an estimated 930 bhp midseason. All of the 2005 V10s exceeded 900 bhp but it is not thought that any exceeded 950 bhp, with the possible exception of the Honda at the end of the season. Cosworth’s 2005 V10 was the TJ, which had its redline at 19,000 rpm. Indeed, it had taken a relatively long time for the Northampton virtuoso to rise above 18,000 rpm with its V10s. Nevertheless, with the CA it took the uncompromising approach of targeting 20,000 rpm from the outset. This was the first time it had produced such a high-speed V8 – the previous fastest running of the type had been the XF CART engine, which ran to 16,250 in qualifying back in 2002 (immediately before the switch to a Cosworth-supplied spec engine for that series).
By regulation, the CA retained the per-cylinder displacement of the existing 3.0 litre TJ, which had a 95 mm bore. As Cosworth’s technical director Bruce Wood remarks, “To go faster you just have to keep making the bore bigger, the stroke shorter and sort out your valves. “While developing the TJ we did tests on our single-cylinder rig of 96, 97 and 98 mm bores – it was all about higher speed. We were considering a bigger bore, bigger valves and a compound valve angle before the mandatory switch to V8s was brought in. We had thereby established that combustion was OK with the 98 mm bore [the maximum permitted in 2006] so there was no reason not to move to it.”
However, we should note that, back in 1999, Cosworth had evaluated the potential of a V12 derivative of its contemporary V10. That V10 ran to 17,500 rpm, and with its smaller cylinders the V12 was ambitiously targeted at 20,000 rpm. It was concluded though that the benefit of the 2500 rpm increase would be negated by an associated increase in frictional losses. That loss of performance didn’t happen when the CA progressed 1000 rpm above TJ operating speed, and an article in RET 20 (February 2007 – ‘Cosworth Insight/Formula One and Frictional Losses’) explores the reasons why.
In essence, by 2005 Cosworth had a much better understanding of how to mitigate the increase in friction implicit in running faster, and it had even pioneered the use of a DLC coating for the piston skirt. The CA consequently had the most extreme stroke-to-bore ratio that Cosworth had ever used – 0.406:1 compared to 0.557:1 for the TJ. At its 19,000 rpm redline the TJ saw a mean piston speed (MPS) of 33.4 m/s (109.5 ft/s) whereas at 20,000 rpm the 6% bigger bore CA saw an MPS of 26.4 m/s (86.7 ft/s). That is a figure that many would nevertheless still consider worryingly high, but it is comparable to that of Cosworth’s 14,000 rpm, 3.5 litre HB V8 of more than a decade earlier.
The upshot was that the CA’s crankpins were actually subject to less loading than those of the TJ, despite the implicit increase in reciprocating weight due to the larger-diameter piston. The CA Data 2006 sidebar on page 48 sho ws that the CA’s maximum piston acceleration was 9581 g at peak power speed (PPS) of 19,000 rpm, rising to 10,616 g at 20,000 rpm. Note that the load on a crankpin reached a very substantial 5937 kg at that unprecedented speed.
With intake valve area varying (through development) from 32.9% to 37.2% of bore area, the TJ saw intake valve headbased mean gas velocity at PPS in the region of 87.34 to 98.75 m/s, which is extremely high by the standard of naturally aspirated Formula One race engines. By contrast, the figure for the 2006 CA (with intake valve area as 35.5% of its 6% larger bore area) is a more conventional 70.7 m/s at a PPS of 19,000 rpm.
It follows that, from the point of view of both the associated mechanical stresses and of getting the charge into the cylinder, the CA’s significant increase in bore size was advantageous. The real challenge was to burn the mixture effectively using such a large bore. As the stroke-to-bore ratio decreases, it becomes ever harder to obtain an adequate compression ratio, while clearly the flame travel is lengthened, plus the time for combustion shortens as rpm rises to take advantage of the bore increase.
The geometry of the CA’s combustion chamber was assisted by introducing a compound valve angle; its operation was assisted by fuel pressure. In simple terms, as the time for mixture preparation went down (with increasing rpm) Cosworth had found it needed to exploit higher and higher fuel pressure. As Wood explains, “The ability these days to phase fuel delivery, to have good fuel preparation, has aided combustion, and the ability to combust the fuel well has enabled us to increase the bore size; that in tur n has enabled us to run the engine faster.
“Nevertheless, when we first tried bigger bores in the V10 days, we didn’t manage to make them work successfully, because we couldn’t get the combustion right. The necessary mixture preparation was enabled by running higher and higher fuel pressures. By regulation we run at 100 bar now, whereas for a while the CA was running at 200 bar on the dyno. We never raced it at that level (due to the 100 bar rule), but that was what we were developing. We did find performance from it, because the mixture preparation was enhanced, from the higher pressure.”
As engine speed increases, so does the associated vibration. When the V8s replaced the V10s, most Formula One engine manufacturers reported problems. A 90º V10 is inherently better balanced than a 90º V8 with a flat-plane crankshaft; the V8 is balanced vertically but not horizontally. Wood reports that when the CA first ran, even with all of Cosworth’s experience of V8 engines, this one, running more than 3000 rpm faster than any previous example, took the company by surprise in some respects.
“When we first started running the CA, the scavenge pumps, which are held onto the sump with horizontal bolts, would fall off. Those are 8 mm cap screws, the heads of which snapped off because of the unbalanced force – which is why our scavenge pumps are now secured by Multiphase bolts!”
Cosworth numbers V8 engine cylinders one through to four down the right-hand bank, then five through to eight down the left-hand bank (therefore front left is cylinder five). On this basis, traditionally it has used the firing order 1-8-3-6-4-5-2-7. That goes right back to the days of the DFV, its pioneering 3.0 litre Formula One 90º V8, introduced in 1967. However, in the late 1990s it switched the order to 1-5-2-6-4-8-3-7. That was found beneficial in terms of performance, at the cost of more severe crankshaft torsionals, and is the firing order used by the CA.
“In Formula One it is a case of using the best performance firing order and fixing the torsionals,” remarks Wood. “That’s why we have lots of dampers on the CA!
“In terms of the torsional vibration inside the engine, we knew what we were up against, which is why the CA has far more damping devices in it than our previous V10 engines. We have a ‘compliant’ geartrain that has been in our Formula One engines for years, then in addition [to two dampened compound gears] the CA has compliant quill drives within each of its two auxiliary drives, a big viscous damper on the back of the crankshaft, viscous dampers on the back of each camshaft and friction quill dampers in the front of each camshaft. That means in total it has 13 dampers – 14 when fitted with KERS.”
These are all friction- or viscous-type dampers – Cosworth hasn’t embraced the pendulum-type damper used by some other recent Formula One engines. While there is that damper on the back of the crankshaft, the clutch is in the gearbox. “If you have the flywheel on the back of the crank, you lower the frequency of the crank,” explains Wood. “Moving the flywheel/clutch assembly to the gearbox beneficially raises the frequency of the crankshaft, even if the teams would prefer it on the engine.”
For engine life, in 2006 the CA had to match that of the TJ, which was around 800-900 km (500-560 miles) since it had to run two consecutive Grand Prix race meetings between rebuilds. A DFV would have had a similar life, essentially because in the late 1960s and ’70s relatively impecunious Formula One teams expected to get a couple of race meetings out of an engine. In the late Nineties and early Noughties there were still no rules on engine mileage, but manufacturer involvement was now at its peak, and engines were changed between qualifying and the race, often running no more than the 350-mile race distance, less if a dedicated qualifying engine.
Having debuted in 2006, by the time the CA returned to service in 2010 there was a limit of eight engines per driver for the 19-race season. Over the second half of the 2006 season (September onwards) Cosworth had permitted the use of the CA to 1150 km, allocated over a number of different track operation modes. This compares with about 2200 km (1370 miles) from 2010 onwards in the mandatory 18,000 rpm guise. In other words, the required engine mileage had become more than double that for which it was originally designed. Wood points out that a significant difference from the days of the DFV is the amount of endurance testing that each new component has to undergo before it is raced. Back in the days of the DFV, developments that increased performance as measured on the dyno tended to be introduced to the race programme without extensive durability testing (partly due to a lack of ability in those days to simulate race track conditions accurately). These days, anything new routinely has to undergo a pair of 3200 km endurance tests under Monza race conditions, as accurately simulated on a transient dyno. A first example is run to 3200 km, then a second example is put through the same test. “Nowadays we have a religious dedication to endurance testing,” emphasises Wood. “We went into 2010 having conducted a total of 32,000 km of endurance testing, on top of our performance development work.”
The heart of the matter
Wood reports that the design of the CA’s piston was key to allowing it to attain 20,000 rpm from the outset. Rival manufacturers similarly discovered that it wasn’t the likes of valve control, combustion or mechanical stress that were ultimately the most difficult hurdle to overcome to reach that level of engine speed; it was the integrity of the heavily loaded 98 mm bore piston, which by new-for-2006 regulation had to be aluminium alloy. ...This article is an exclusive extract of "King of speed", an article written by Ian Bamsey and published in Race Engine Technology Issue 73. If you wish to read more, you can buy the issue at highpowermedia.com and put 'f1technical' as voucher code to benefit a 10% reduction on your purchase price.