Mercedes vs Ferrari

Their performances are similar, but the outstanding Formula 1 cars of 2017 followed very different routes to the same destination

The competitive timbre of the 2017 season has been very different from that of the last few years, in that the title contenders have come from opposing teams. The race-to-race variance in the duel that has raged between Lewis Hamilton and Sebastian Vettel has only been amplified by the sharply contrasting traits of their respective cars, the Mercedes W08 and Ferrari SF70H.

They behave just as differently as Hamilton and Vettel themselves, but their variation from each other’s performance is greater. Any fluctuations we’ve seen from one track to the next have been far more to do with cars than drivers. Two machines with totally opposing solutions to the new regulations of 2017, they came from different circumstances, philosophies and specialisations and work in very different ways. Intriguingly, they were also designed around two different interpretations of a key part of the technical regulations. 

If two leading cars were ever going to be so different in concept and behaviour, it was in 2017, the year of a major redrawing of the technical regulations, with wider, more aerodynamically powerful cars running on fatter tyres. With the parameters redrawn and a virtual infinity of solutions, even F1 software cannot define the optimum ones in advance, especially when the black art of tyre performance is mixed in. So the two teams came at the problem from opposing sides and delivered two very different machines. Only as the regulations remain unchanged will we see more signs of a technical convergence.

As a generalisation, the Ferrari’s speed has proven more robust, more easily accessible, whereas the Mercedes has enjoyed arguably higher peaks of performance (especially in final qualifying where an engine mode advantage of about 0.15sec has frequently been decisive), but been more inconsistent across different types of track, tyre compounds, track temperatures and so on. 

Here in detail are the factors that have been driving those differences. 


Mercedes – uniquely – continued with its low-rake concept while every other team has been following Red Bull’s lead of introducing ever more rake into their cars. With the SF70H Ferrari went further towards Red Bull’s extremes, typically running at about 1.55 degrees (Red Bull 1.58, Mercedes 0.9), nose down/tail up (ie a greater rear ride height than front).

Historically, Red Bull had led the way on exhaust blowing, sending it further down the high-rake path that everyone but Mercedes has since followed. Mercedes, by contrast, had a fairly basic exhaust-blowing system but was an early pioneer of FRICS (front-to-rear interconnected suspension). Both technologies have long been banned, but they sent the teams in aerodynamic development directions that probably led them to their differing concepts. Ferrari was a late adapter to both.

Almost all the outward differences we see between W08 and SF70H (long wheelbase vs short/intricate aero detail under the nose on the Merc but around the sidepods on the Ferrari) are just consequences of that divergence in rake philosophy, one that under these regulations dominates all else. The major long-lead hard points of the car – monocoque length, wheelbase, gearbox case length – are determined by the implications of pursuing low- or high-rake aero philosophies. Hence two fundamentally different cars.


The principle of running the car with rake is that it gives a greater expansion area to the underfloor. Using the principle of ground effect, the air enters the under-floor through the small gap between the floor’s leading edge and the ground. Because the area behind that gap opens out (even more so with rake), it lowers the air pressure there. Air rushes towards any low pressure to fill the vacuum, and so the low pressure beneath the floor sucks the oncoming air through the gap faster. The faster the air moves, the more downforce it creates.

This has an effect on the front of the car (the air being sucked hard through the gap pulls it back over the front wing and other aero devices faster) and also at the back. The faster the air flows beneath the floor and through the diffuser, the faster the air exiting the diffuser can be directed to the underside of the rear wing, increasing the speed of the oncoming air onto the wing (again because of the pressure differential created, this time between the wing underside and its top surface).

But high rake isn’t all upside. When the ride heights increase as the car slows and the downforce acting upon the suspension reduces, the rear ride height increases more relative to the front than on a low-rake car – ie a high-rake car becomes even higher rake at low speeds. This is almost the opposite of the ideal, in that as the downforce bleeds off with the reducing speed so the gap between diffuser and the ground increases to the point where the air seal is broken (ie the airflow is no longer fast enough to maintain a seal around a gap that is getting bigger). So the reduction in downforce in itself brings about a yet-bigger reduction in downforce as the diffuser stalls. It’s therefore difficult to maintain good low-speed aero grip at the rear of a high-rake car (which is what led Red Bull to revisit exhaust-blown diffusers a few years ago, reviving a concept first used by Renault on the RE30 of 1983). The lower the speed at which you can prevent the diffuser stalling, the greater rake you can run. Diffuser stall sets the limitation upon how far you can go with rake – hence why teams cannot just put ever-more rake on their cars and watch them go faster. They won’t, not without hugely intricate working of the rear bodywork, especially around the tyres.

High rake also tends to increase the drag resistance of the car, lowering its terminal straight-line speeds.

This all means that the underlying trait of such a car will be that its centre of aerodynamic pressure will move forwards as speed reduces (initial understeer, then neutrality as the speed falls – exactly as Vettel likes it). This trait in theory is a good match with the general conflicting demands from an F1 car for high-speed rear stability but a responsive low-speed direction change.


All other things being equal, an optimised low-rake car will not produce as much total downforce as an optimised high-rake car. But there are ways of ameliorating that – and the Mercedes aero team opted for the most obvious: increasing the length between front axle and leading edge of the floor. In this way the aerodynamic power of the floor’s leading edge can be boosted by accelerating the air harder before it gets there, thus at least partly compensating for the lower expansion angle of the floor. In that increased length, there is an intricate array of guide vanes below the nose artificially to boost the speed at which the airflow hits the floor’s leading edge. So although the air isn’t being sucked as hard from behind by the underfloor (because its lower angle means less expansion and less pressure differential), it’s being accelerated towards it by other means.

This arrangement has given a powerful effect on the front of the car, but been less effective at maintaining high-speed flow beneath the floor and out to the rear. To combat that, the rear axle has been sited well back so as to increase the total floor area. The idea is that the greater floor area can compensate for the lower specific pressure (per square inch) to give a similar total.

In underlying traits such a car will tend to have a very strong front end at high speeds, but the centre of pressure will move rearwards as the speed falls (initial positive front, progressing towards understeer as downforce bleeds off more at front than rear), ie the opposite of the Ferrari. This is not ideal in that it narrows the set-up window. Rear entry stability is essential in high-speed corners and so the chosen set-up needed to take the edges off that will make the car less responsive in slow turns. So long as the circuit in question doesn’t have too big a spread of corner speeds, a happy medium can usually be found. But, occasionally, that spread can leave the team without the tools to keep it balanced at all points on the track.

The requirement of plenty of length between front axle and floor and that of increased floor length – both measures to compensate for the low-rake floor’s reduced expansion angle – has entailed a very long wheelbase on the Mercedes (at 3738mm it’s just short of the longest wheelbase available in a current Ford Transit, 143mm longer than the Ferrari, 200mm longer than the Red Bull).

A longer car means greater weight and the W08 could not even get down to the minimum limit at the start of the season (it’s believed to have been running 8kg over until Spain). This in turn means the static weight distribution is frozen and ballast cannot be used better to match the weight distribution to the demands of a particular track, further narrowing the set-up window. The Ferrari, incidentally, can run with up to 7kg of ballast.


Generically, the new wider, fatter-tyred cars of 2017 made the heavy hybrid cars even heavier – and the minimum weight limit had to be increased to 722kg. But the W08, with its extreme layout born of trying to maximise the downforce-inducing surfaces so as to overcome the increased power of high-rake from the new regs, was particularly heavy.

The wider 2017 cars (and the complication of fatter tyres at big steering angles) have not been friendly to the low-rake concept. It made high-rake more powerful. But also in terms of balance, it makes for a greater balance change on a low-rake car between low/high speed as ride heights change and at big steering angles (because it loses more front relative to rear than before).

1) The wider floors were more aerodynamically powerful – but by a greater amount for high-rake than low (the specific pressure per square inch was now multiplied by a greater surface area in both cases, but the high-rake car had greater specific pressure).

2) The wider floor made for a greater balance change on a low-rake car across varying speeds and at big steering angles. It loses more front downforce relative to rear than before as speed falls.

Mercedes, having developed its cars for years around low rake, was reluctant to begin at the bottom of the mountain on a whole new philosophy that would require years of catching up. So it went extreme in layout to compensate. The reason it felt comfortable doing this was it had a new tool that was going to get around the traditional downsides of low rake.


For several years Mercedes had a very tasty R&D project that it believed would allow it to have its low-rake cake and eat it. It would allow the lower drag and other advantages of a low-rake car, but without the concomitant smaller balance range. It would allow it a powerful front floor even as the speed and downforce bled off, compensating for the lack of expansion angle in the floor.

This was the hydraulically-operated front heave damper. Introduced on the W07 last year, but only at a selected few tracks, what they learned from that was incorporated into the design of the W08 – a car conceived to take full advantage of the technology. Generically, heave dampers resist the vertical forces on the cars to give a more stable aero platform. When such a component was actuated hydraulically, asymmetric valving would allow it to store the energy, allowing the car to dive instantly low at the front when braked but retain that low ride height into and through the corner despite the downforce loads bleeding off it. The valving would allow the front of the car to rise back up only slowly after being braked – for long enough to keep it low into and through the corner. That got around the tricky balance change inherent in the low-rake concept, but without the disadvantages of high rake. It did much of what the banned FRICS technology had formerly done and was essentially a mechanical computer that controlled ride height for aerodynamic purposes. It allowed a low-rake car to have many of the high-rake advantages. The only downside was that it added even more weight to an already heavy car.

It was Ferrari who made the query to the FIA, not long before the start of the season. It outlined how it believed such a system would work and queried whether it would contravene article 3.15 regarding whether such a component would be ‘wholly incidental to the main purpose of the suspension system’ or ‘have been contrived to directly affect the aerodynamic performance of the car’.

The FIA’s Charlie Whiting then issued a technical directive which informed all teams that, in his opinion, such a system would be considered against the regulations (effectively banning it). Mercedes had been running the system in the first pre-season Barcelona test. It was removed mid-week – and the performance of the car was suddenly much less impressive. Former Mercedes engineering chief Paddy Lowe said of Ferrari’s query: “Ferrari has been sending missiles for a while but they hit the aircraft carrier with that one.”

A lot of intense work was done in the weeks between the tech directive being issued and the start of the season in Melbourne, to bring the W08 back into the ballpark – sufficiently so to allow Hamilton to take pole. But the fact remained that the technology around which the car had been conceived was no longer allowed.

Given a clean sheet of paper in the knowledge of the hydraulic heave spring ban, it seems improbable that Mercedes would have conceived such a dimensionally extreme car. It might not even have conceived a low-rake car. That hydraulic-heave spring ban was actually a devastating blow to Mercedes. Without it, we might well have had another season of silver domination.


The differing traits of the two cars has had a major impact upon the outcome of many of the season’s races. Not being able to vary the weight distribution narrowed the Merc’s set-up window and meant it often couldn’t get both tyre compounds working. Usually with too much of a rearwards bias, the fronts could take too long to reach working temperature, by which time the rears could already be too hot. Playing with the set-up to alleviate that tended to mean it worked only with one tyre compound, taking it out the range of the other.

The Merc tended to suffer higher rear tyre degradation, especially when the track temperature was high. The first signs of this came in Melbourne as the pole-winning Hamilton could not shake off Vettel’s Ferrari, which combined greater race day pace with better tyre usage and went on to win.

The Mercedes was often only winning races through the track position advantage that pole position brought, as in China – and a crucial part of that was the greater power boost of its engine in Q3. At the power-sensitive Bahrain track, pole-winning Bottas was 0.5sec faster than Vettel but the Mercedes did not work the super-soft well and Vettel was again able to undercut ahead to victory.

Mercedes had greater aero efficiency, the Ferrari greater total downforce. At slow circuits where aero efficiency was not an issue and it was all about putting on as much downforce as possible, the Ferrari proved to have significantly more ‘dirty downforce’. Low-grip surfaces, slow corners and high temperatures all adversely affected the Mercedes relative to the Ferrari. At Monaco it had all three and was no match for Ferrari, even in qualifying. A similar pattern would be seen in Singapore.

A big spread of corner speeds would expose the Merc’s narrower set-up window, making it impossible to balance over the whole lap (something the hydraulic heave spring would have solved) and leading both drivers to comment occasionally that the front and rear of the car felt unconnected. A big spread of corner speeds contributed considerably to the W08’s difficulties at Hungaroring and Sepang.

Silverstone was an interesting case study in that the FIA had insisted Ferrari’s floor front be stiffened and the ducts on the floor’s edges were no longer functional. It left the team running out of set-up options to balance the initial understeer inherent in its aero philosophy – contributing to both cars running out of front tyre tread late in the race. For that track, the Merc’s positive front end was perfect and it duly dominated. Ferrari was able to reconfigure its floor by Hungary, giving it once more a wide range of set-ups.

James Allison, Mercedes technical director: “If you look at the season as a whole, it has been divided into three types of experience. There have been a few races where we have come out and crushed everything in front of us. There have been a few where we’ve had the other end of that deal, where we have definitely come off second-best. And then a whole lot in the middle where it has been pretty much a 50/50 slugging match.

“It’s not always easy to get the best from the car. It’s been a challenge this year to achieve the results we have, but nevertheless we have achieved some pretty decent outcomes with it, so it’s not been a bad machine for us. However, we would like a car that is easier to throw at the racetrack and easier to guarantee that every time we’ll get every last little bit from it. We hope that next year we’ll make something with a slightly sweeter temperament.”