Ride-height sensitivity, rake & aero maps
On any car with an underbody, downforce, drag and balance depend strongly on dynamic front and rear ride heights. Aero maps capture this; the engineer's job is to place the car's dynamic platform in the right region of the map.
Ride-height sensitivity, rake & aero maps: L1 Foundation explanation
On an old car with a flat-bottom-free body, downforce came mostly from wings and hardly depended on ride height. On every modern race car a large part of the downforce comes from the underbody (floor, splitter, diffuser), and the underbody works by accelerating air through the gap between car and road. That gap is the ride height, so:
- Lower ride height → faster air under the car → lower pressure → more downforce (until the flow breaks down, see the ground-effect lesson).
- The front and rear ride heights do not have the same effect. Lowering the front usually moves the aero balance forward; lowering the rear moves it rearward.
Rake is the difference between rear and front ride height, often expressed as an angle:
A GT3 at 55 mm front / 75 mm rear on a 2.70 m wheelbase has 20 mm or ~0.42° rake. Increasing rake (front down or rear up) generally adds downforce by making the diffuser/floor more effective and moves the balance forward — up to a point where the diffuser or floor edge loses its seal.
The key idea: static ride height (in the garage) is not what the aero sees. At speed the downforce compresses springs and tyres; under braking the car pitches nose-down; on power it squats. What matters is the dynamic ride height on track, measured with laser sensors or estimated from damper potentiometers.
Typical numbers: a GT3 front ride height can drop 8–15 mm between pit-lane and the end of a fast straight; a formula car, much stiffer, maybe 5–10 mm at the front including tyre squash. Downforce changes of roughly 0.3–1.5 % per mm of front ride height near the working point are common on cars with significant underbody downforce (more close to the stall height).
An aero map is a table (from the wind tunnel, CFD or track tests) giving , and aero balance as functions of front and rear ride height. Your setup sheet's static heights are chosen so that the dynamic heights land where the map gives the best compromise.
Ride-height sensitivity, rake & aero maps: L2 Practitioner explanation
Reading an aero map. Typically three grids indexed by front RH (x) and rear RH (y): , and AB (Segers Fig. 13.22). Contours of usually rise towards low front / moderate rear, and AB contours run diagonally: lowering the front pushes balance forward, raising the rear pushes it forward too (more rake). Typical GT3 sensitivities near a normal working point (order of magnitude, car-dependent):
| Change | AB | |
|---|---|---|
| Front RH −1 mm | +0.5 to +1.5 % | +0.2 to +0.4 % fwd |
| Rear RH +1 mm | +0.2 to +0.8 % | +0.1 to +0.3 % fwd |
| Rear wing +1° | +1 to +2 % | −0.5 to −1 % (rearward) |
Note that a 2 mm front ride-height change can move the balance as much as a full wing click.
Choosing static ride heights.
- Decide the target dynamic platform at the fast corners (where aero matters) — e.g. front 40 mm, rear 60 mm at 220 km/h.
- Predict the compression: (wheel rate plus tyre rate in series). Example: 1300 N front aero load, front heave wheel rate 2×110 N/mm in series with 2×300 N/mm tyres → combined ~161 N/mm → ~8 mm.
- Static = dynamic + compression (+ fuel burn effect for endurance).
- Check the platform at low speed too: too high a static rear for rake at speed might make the slow-corner balance worse via mechanical effects (CG height, roll centres).
Rake changes are not "free". More rake raises the CG at the rear, changes the roll-centre heights and the kinematics, and moves weight distribution slightly. Always ask: was the lap-time effect aero or mechanical?
Fuel burn. In endurance races a full stint burns roughly 55–90 kg of fuel (LMP2 ~55 kg from a 75 L cell, GT3 up to ~85–90 kg). That raises the car by ~1–3 mm depending on axle stiffness and tank position (e.g. ~40 kg on an axle against ~180 N/mm combined wheel+tyre rate is ~2 mm), mainly at the axle nearest the tank. Expect the balance at speed to drift over a stint; the setup is usually a compromise between full and empty.
Plank / minimum height rules. Single-seaters and prototypes run a skid block whose wear is checked; GT series may have minimum ride height checks. The legal limit often caps how low you can run, which caps how far you can exploit the map.
Misconception: "lower is always faster." Below a certain height the floor stalls, the platform hits the ground (lost mechanical grip over bumps, plank wear), or the balance shifts so far forward that the car becomes undrivable at high speed.
Ride-height sensitivity, rake & aero maps: L3 Advanced explanation
Linearised map. Around the operating point :
Rewriting in heave and pitch is more useful because suspension modes are heave and pitch: heave mainly changes total load; pitch mainly changes balance. Front and rear heave/pitch stiffnesses therefore become aero tools.
Self-consistent platform. Downforce compresses the suspension, which changes downforce. With axle stiffnesses , (wheel+tyre, per axle) the equilibrium at speed solves
Iterate (fixed-point) on the map. Because the feedback is positive (lower → more load → lower), so stiffness requirement rises with — the static analysis of the porpoising problem.
Pitch sensitivity & braking. Aero load decreases as the car slows, while pitch from braking is largest at the start of the zone. The CoP at turn-in is therefore a function of both: an engineer should evaluate over the braking zone. Front packers / bump rubbers with a known engage point make the front RH at max decel almost speed-independent, which stabilises the balance at entry — a key reason modern cars run on packers.
Building track-based maps. With laser ride heights (preferably at all four corners, corrected to the axle lines, Segers §13.4.2) and load cells or calibrated spring/damper forces, collect steady-speed points across a range of static settings. Fit a low-order polynomial surface by least squares. Keep only points with g, g, and correct for longitudinal load transfer (Segers §13.5, Eq. 13.18–13.19) and tyre radius (Segers §13.4.1). Compare with the manufacturer's map — offsets of 5–10 % between tunnel and track are normal; the gradients are what you most need to trust.
Ride-height estimation errors. Damper potentiometers miss tyre deflection — on a stiff formula car most of the RH change at speed can be tyre squash. A dynamic tyre-radius lookup (load, speed) or lasers are needed; laser data needs filtering for track texture and kerbs (e.g. median filter, then 10–20 Hz low pass).
Ride-height sensitivity, rake & aero maps: L4 Expert explanation
Dimensionality. Real aero maps are at least 5D: front RH, rear RH, roll, yaw (body slip + wind) and steer, sometimes also DRS/flap state and wheel rotation. A 2D (h_F, h_R) map is a slice at zero roll/yaw. In mid-corner a GT3 may run 1.5–2° roll and 2–4° yaw; the corner downforce can be 5–15 % below the straight-line map value and the balance can move by several percent. Lap simulation correlation work therefore weights the yaw/roll slices heavily for high-speed corners.
Hysteresis and history. Floors with vortex-sealed edges or diffusers close to stall show hysteresis: the downforce on the way down (lowering) differs from that on the way up. Static tunnel maps (stepped, steady) can miss this; on track it appears as scatter in vs RH plots that correlates with the sign of ride-height velocity. Expert workflow: colour scatter by to separate the branches.
Tunnel–track gaps. Rolling-road tunnels typically run at 50–60 % scale for formula cars, ~40–50 m/s, with deformable-tyre models of limited fidelity. Reynolds effects, tyre squish and sidewall bulge (which changes the floor-edge flow), heat exchangers and exhaust flows produce systematic offsets. Teams build correction factors per region of the map rather than a single global scale.
Using the map in setup optimisation. Static heights, heave springs, packers and rake are chosen jointly by an optimiser running a quasi-steady or transient lap sim where every time step evaluates the map at the predicted dynamic platform. Constraints: plank wear limit, minimum RH over the worst bump (from a ride model with a measured track-PSD), tyre load variation, and drivability (AB spread across corner phases). The best solution is rarely at the map's peak — it is at the peak of usable downforce: the region where gradients are moderate so that bumps, fuel burn and tyre wear do not throw the car across a cliff in the map.
Key formulas
- rear ride height (mm)
- front ride height (mm)
- WB
- wheelbase (mm)
- aero load on axle (N)
- axle heave wheel rate (N/mm)
- axle tyre vertical rate (N/mm)
Trackside takeaways
The aero sees dynamic ride heights, not the setup-sheet static values.
Lowering the front (or raising the rear) usually adds downforce and moves balance forward — until the floor stalls.
A 1–2 mm front ride-height change can be worth as much balance as a wing click.
Think in heave (total load) and pitch (balance) rather than individual axle heights.
Fuel burn shifts ride height and therefore balance over an endurance stint.
Aim for usable downforce: a region of the map with moderate gradients, not the peak next to a cliff.
On your car
GT3 runs significant rake (often 10–25 mm static). Ride heights are free within homologation, sometimes with BoP minimum ride heights. Front ride height is the most powerful high-speed balance tool after the wing.
Check yourself
A GT3 runs 54 mm front / 74 mm rear on a 2.70 m wheelbase. What is the rake angle?
- Segers – Analysis Techniques for Racecar Data Acquisition — §13.4 Ride Height Measurement
- Segers – Analysis Techniques for Racecar Data Acquisition — §13.7–13.8 aero maps from constant-velocity tests
- Milliken & Milliken – Race Car Vehicle Dynamics — Ch. 15 Applied Aerodynamics
- Milliken & Milliken – Race Car Vehicle Dynamics — Ch. 16 Ride and Roll Rates