Milton Keynes was a new town put onto green fields in the England midlands in the 1960s. The industrial estate where Red Bull Racing has its HQ is so generic it could be anywhere in the world. But this is where the Aston Martin Valkyrie is undergoing advanced testing, despite the fact it doesn’t physically exist yet.
The Valkyrie is best thought of as Adrian Newey’s fever dream: realisation of the Red Bull F1 designer’s long-held desire to build a road-legal car with performance to rival a full-fat endurance racer. One hundred and fifty are going to be made, and despite a pricetag of USD$3.2m, all have already been sold. We’ve already told you about its Cosworth-developed atmo V12 engine, the contribution that the hybrid system will add, and even reported Aston is planning to bring a very small number to Australia and New Zealand. N
ow I’ve been invited to see both how it is being developed, and how ridiculously fast it is going to be.
Like all Formula 1 outfits, Red Bull Racing does a huge amount of simulator work, with restrictions on real-world testing meaning most development work is virtual. Mark Webber once admitted he spent far longer in a digital Red Bull than a real one back when he was pedalling for the team – he also used to leave his 911 parked right next to the main entrance. But now RBR’s huge expertise has been put to use on a road car project with the Valkyrie, allowing work to start on it well ahead of the creation of the first physical prototype.
My guide for the day is Chris Goodwin, Aston’s chief test pilot. Having spent most of his working life at McLaren, Goodwin defected in 2017 to head up development of AM’s three mid-engined cars. With the Valkyrie barely more than a set of performance targets at the time, that meant a huge amount of early work was done on the simulators. Goodwin has driven 8500km in the virtual Valkyrie since starting, with the software model steadily evolving and becoming more advanced as he’s done so.
“I’ve been using simulators for 19-odd years and thought that I’d been at the cutting edge,” he says, “but what hit me here is the rate of development we’ve been able to drive forward using what are outwardly the same tools.”
Because this isn’t just pretend. Simulator data is harvested – RBR has extracted around 700GB so far – and this is both analysed and shared with suppliers. So Goodwin’s laps of a virtual Silverstone track have been used for the dyno test sessions of the Cosworth V12, with the real behaviour of the engine then sent back to update the software model’s map.
The Valkyrie’s Bosch stability-control system will soon be running from live data, getting the same sensor inputs it would from a real Valkyrie and then outputting directly into the sim. It means that a huge amount of basic work has been done, but also that the team has been able to check for conflicts between the Valkyrie’s active aero and chassis systems.
“Without simulation, you wouldn’t be able to do a project like this given all the time in the world,” says Goodwin. “It would just be unaffordable; it would take years and years.”
Simulators are expensive so productivity has to be high. For each session, a series of different scenarios have been created to test for certain things. Goodwin will then go through these back-to-back – he says he can spend up to five hours at a time in the simulator – but without the need to physically change settings or parts, it is up to 10 times more productive than real-world testing.
“We still have to do that later on, of course,” he says, “but this process means we will hit the ground running with a car we just need to tune. The functionality of all those complex systems will already be proven.”
Goodwin admits the sim work has dramatically cut down on his travel to testing venues. “Gianfranco in the Rive Del Sole hotel in Nardo hasn’t seen me for months, he probably thinks I’m dead,” he jokes. “It’s the first winter for years I’ve spent in England."
Right, it’s my turn. The dynamic simulator being used for Valkyrie work started out doing race testing, as proved by the use of an F1 car tub wearing a slightly battered Red Bull livery. This is mounted to a raised platform, so I need to climb a ladder to get in. I also need to take my shoes off, Goodwin saying they won’t fit in the tight confines of the footwell.
Getting in is a squeeze – let’s say I’m not F1-driver shaped – but once installed, I’m looking at a wraparound screen with three projectors beaming images onto it, including a pair of Michelin-branded tyres and a single-seat perspective. I’m also holding a full-on F1-grade steering wheel of brain-melting complexity.
Fortunately, I can ignore most of the controls. The gear-change paddles work and there’s a push-to-talk button to communicate with the control room behind; there’s also an F1-style clutch paddle, but Goodwin says I don’t need to use it.
I’m going to be driving around Spa, and before experiencing the Valkyrie, I’m being sent out in what’s meant to be a generic supercar to represent where the segment is. “It’s a bit McLaren, a bit Ferrari,” Goodwin says. It will give a feel for both track and simulator, with my start position just after Blanchimont and pointing towards the chicane. There’s a vibration as the simulator becomes live, and a voice in my ear from the control technician tells me I can start.
I’ve spent too much of my life playing various racing games, but the big-boy simulator feels deeply strange. There’s a sense of acceleration and cornering loads, but it isn’t the motion of a real car: short, sharp motions rather than sustained loads. The brake pedal has little resistance and anything more than gentle pressure has the front wheels locking – there’s no ABS or stability control running yet.
The engine note is entirely generic, and despite the huge amount of processing power I know is driving everything, the graphics are surprisingly average, and some way off the crisply rendered look of Forza 7.
I find it really hard to judge speed and distance, braking far too early but then carrying over-optimistic velocities into corners. On my second lap, I attempt Eau Rouge at such a ridiculous pace that I somehow manage to break through the track at Raidillon and end up hanging in space. “You’ve crashed the programme,” says the voice in my ear, “it’s going to take a couple of minutes to reset it.”
Eventually I cobble together a cautious lap without any egregious mistakes, although I’m still certain I’m braking far too cautiously. Goodwin climbs onto the platform for a pep talk and I admit that it doesn’t feel like a real car to me.
“That’s because it’s a tool, not a game,” he says. “It’s a motion platform, but the movement is only enough to inform an experienced simulator driver. It feels completely different to the way a finished car will, all simulators do to be honest, but there’s a parallel universe of simulator behaviour and real car behaviour. Recognising the translation between them is what comes from experience.”
Round two, and I’m in the Valkyrie – although facing an unchanged F1-style viewpoint. Goodwin has said I’m driving a model that’s already a fair bit out of date, the newer version being quicker and more accurate. As before, there’s no ABS or stability control, but there is now a V12 engine note that is sounding much angrier by the time the gear-change lights come on.
It’s also massively quicker, but it also seems considerably easier to drive. Spa’s straights feel immediately shorter and – although I’m still finding it hard to judge braking – cornering forces are also much higher. The system is modelling aero downforce, and as speed increases this starts to add its contribution. In slower corners I need to be gentle with the throttle given the lack of stability control, but as speeds rise, the Valkyrie starts to feel unstickable.
It’s not, of course. I leave the track several times – another Eau Rouge incident deserves a place on the sim team’s blooper reel – but I do manage one almost clean lap and confirm that this virtual Valkyrie can be driven flat from Stavelot to and through the very fast Blanchimont left-hander without undue drama; something even full-on GT3 race cars struggle to do.
If reality matches the simulation – as Red Bull insists it will – the biggest limiting factor for the Valkyrie's performance is going to be bravery and neck muscles.
I’m not given any times, as Aston doesn’t want me to try and extrapolate Valkyrie performance even at my pace. But I do get one number from Goodwin, who says my fastest lap in the Valkyrie was 25 seconds quicker than the best one in the generic model. Obviously I improved slightly over time, but that still seems an impossible difference. Goodwin insists it’s not.
“I talk to customers who have paid good money for this car, and they don’t know what they’ve got,” he says.
“There will only be an upside to it, because I don’t think people can really have any idea what the performance level of this car will be until they actually experience it, nor will they be understanding quite how user-friendly it’s going to be.”