Why the air under a racing car is climbing out of a valley

Why the exit condition sets how much downforce a floor can produce.

The air under a racing car is climbing out of a valley because the underbody is shaped low in the middle and rising toward the back, so any air that has been sped up and squeezed low beneath the car has to slow back down and climb its way back out through that rising section before it rejoins the air behind the car, and how well it manages that climb decides how much of the floor's potential downforce actually gets produced.

Adrian Newey's How to Build a Car had come out the year before I joined the team, and it made aerodynamics look less like a dark art and more like a sequence of decisions somebody had genuine reasons for, which is exactly the way I wanted to start thinking about the floor of our own car.

Speeding up under the floor, slowing down at the back

Air arriving at the front of a car's floor is moving at roughly the car's own speed relative to the ground, and if the floor then narrows the gap between itself and the road, that air has to speed up to keep the same volume moving through the smaller gap, which lowers its pressure and produces downforce pulling the floor toward the road. That speeding up cannot continue forever, though, because the floor eventually has to rejoin the air flowing past the rest of the car, at roughly the same pressure that air is at, which means the fast, low-pressure air under the middle of the floor has to slow back down again before it leaves the car, exactly as a valley has to rise back up to the level of the surrounding land eventually. The section of floor that manages this slowing down, usually a rearward diffuser rising away from the road, is doing just as much work as the narrow section further forward, because if the air cannot climb back out of the valley smoothly, it detaches from the diffuser's surface instead, collapsing the low pressure that the front of the floor worked to create in the first place. Because the whole system is connected, from the narrow entry to the rising exit, the floor has to be understood and shaped as one continuous journey for the air rather than as a set of separate sections each optimised for its own local job.

Driving through a dip in the road

Driving over a dip in a road, or riding the equivalent dip in a fairground attraction, gives an intuitive feel for exactly this exit problem. Dropping into the dip at speed presses the body down into the seat as the path curves away underneath, the same low-pressure, accelerated feeling that air experiences being squeezed under the front of a floor. Climbing back out of the dip afterwards is a different challenge entirely, because the vehicle needs enough momentum to comfortably crest the rise on the far side, and a vehicle without enough of it struggles to make the climb cleanly, losing speed or even losing contact with the road surface partway up. Air under a diffuser faces the same test on the way out of its own valley, and air that entered the dip too slowly, or is being asked to climb too steep a rise, simply cannot make the exit cleanly. A gentle, gradual dip and rise lets a vehicle carry its momentum through comfortably at almost any reasonable speed, while a sharper one demands a narrower, more specific range of speeds to cross without trouble, exactly the same sensitivity a diffuser's own angle introduces to how forgiving the whole floor is across a range of driving conditions.

Separation and blocked exits

This valley picture describes the pressure and speed changes happening to the air itself, but it does not on its own explain why a diffuser that rises too steeply causes the air to separate from its surface rather than simply slowing down smoothly as intended, which is a separate question about how quickly a flow can be asked to climb against rising pressure before it gives up following the surface at all. It also assumes there is somewhere for the air to actually go once it exits the valley, an assumption that stops holding if the area behind the car is blocked or disturbed by other bodywork, since a diffuser exiting into disturbed air behind it cannot complete its climb as cleanly as one exiting into clean air. For the same reason, the bodywork immediately behind a diffuser's exit, sometimes barely considered part of the aerodynamic package at all, can meaningfully change how well the whole floor performs, simply by changing what kind of air the climbing flow finds waiting for it at the top.

Why floor length and diffuser angle are decided early

Because the exit condition matters just as much as the entry, a diffuser's rise has to be shaped carefully rather than simply made as steep as possible to squeeze out more downforce, since a diffuser angled too aggressively asks the air to climb faster than it is able to follow, and the resulting separation can collapse the low pressure across the whole floor rather than just the rearmost section. The same reasoning makes the length of a car's floor and the space available behind it for the diffuser to rise through first-order decisions on any new design, well before individual details of the floor's shape are settled, because a floor with no room to let its air climb back out properly can never generate as much downforce as one with a longer, gentler valley to work with. Two cars with almost identical floors underneath can produce downforce figures that differ by more than the floors themselves would suggest once the rear bodywork above the diffuser differs, since one design may simply be giving its air a cleaner, more gradual path out of the valley than the other.