# Setup change priority and the tool hierarchy

> Not all setup changes are equal: they differ in time to make, predictability, scope and side effects. A disciplined hierarchy – tyre pressures → ARB → wing → springs → dampers → geometry – plus correct magnitude and one change at a time keeps a test session productive.

Area: Setup & Driver Feedback · Starting level: L1 Foundation · Updated 2026-11-16
Source page: https://race-engineer-academy.vercel.app/lesson/setup-change-priority

## L1 · Foundation

In a 60-minute practice session you get maybe five or six runs. Every change must earn its place. A practical **hierarchy** of tools, from first to last resort:

1. **Tyre pressures** – the tyre is the only contact with the road. If the tyres are outside their pressure/temperature window, no other change will give a reliable reading. Quick to change, fully reversible.
2. **Anti-roll bars** – change roll stiffness distribution (balance) without changing ride height or heave platform. Often adjustable in seconds (blade, lever or drop-link position), sometimes from the cockpit.
3. **Wing / aero balance** – front wing flap, rear wing angle, Gurney. Changes high-speed balance; costs or gains drag.
4. **Springs** – change ride, roll and pitch together; affect ride height and aero platform. Takes 15–30 min with ride-height reset.
5. **Dampers** – shape the transient behaviour; clicks are quick but effects are subtle and phase-specific, and easy to mis-evaluate.
6. **Geometry** – camber, toe, roll centre, anti-dive/squat, caster. Requires alignment (setup pad), long time, and has many coupled effects.

Why this order? Earlier tools are **faster**, **more predictable** and **less coupled**. Pressures and ARBs mostly change one thing (grip level/window and balance); springs change balance *and* ride height *and* aero; geometry changes almost everything.

Three rules sit alongside the hierarchy:

- **One change at a time** – otherwise you cannot attribute the effect.
- **Big enough to feel** – a change smaller than the driver's and the data's noise is wasted track time.
- **Know the baseline** – always compare to a known reference run on similar tyres and fuel.

## L2 · Practitioner

**Magnitudes that drivers can feel and data can resolve** (typical for GT3 / formula; adjust for car):

| Tool | "Felt" change | Notes |
|---|---|---|
| Hot pressure | 0.05–0.1 bar | Aim at target hot pressure, change cold to suit; ~0.1 bar per 10 °C gas temperature |
| ARB | 1–2 positions (~10–15% axle roll stiffness) | Balance change ~0.5–1% LLTD per step on many cars |
| Front flap / rear wing | 1–2 clicks (~0.5–1.5% aero balance) | Check drag cost on straight speed |
| Springs | one step (~10–15% rate) | Re-set ride heights; affects aero |
| Dampers | 2–4 clicks LS | Single clicks rarely felt; check histograms |
| Camber / toe | 0.25–0.5° / 0.5–1 mm per side | Look at tyre temps and steering trace |

**Bracketing**: when you don't know the direction, try a deliberately large step (e.g. ARB full soft vs full hard), learn the sensitivity, then interpolate. It's faster than creeping one click at a time.

**One change at a time** – with two exceptions: (a) **compensating pairs** – e.g. stiffer front spring plus a lower front ride-height setting to keep the platform – are one *intent*; and (b) when chasing a known baseline (returning to a known setup). Document such packages explicitly.

**Interactions to watch**:
- Springs → ride height → aero balance. A stiffer rear spring adds rear roll stiffness and rear LLTD (mechanical oversteer) **and** raises rear dynamic ride height at speed (it compresses less under downforce). On a rake-sensitive car the extra rake moves aero balance *forward*, so the high-speed oversteer shift is larger than LLTD alone predicts – the two effects add, they do not cancel.
- ARB → inside-wheel load → traction. A stiffer rear ARB gives more rotation mid-corner but more inside-rear wheelspin on exit.
- Wing → drag → gearing. Adding rear wing may require checking top-gear limiter at the end of the longest straight.
- Pressures → everything. Hot pressures drift with ambient and track temperature; re-check them after every run before blaming a setup change.

The hierarchy is not dogma: if the problem is clearly a pure entry issue, brake bias or diff coast (fast, phase-specific) beat ARBs. "Fast, predictable, targeted" is the real rule; the list is a default order.

## L3 · Advanced

The hierarchy is really an optimisation of **information gained per minute of track time**, subject to risk. For each candidate change estimate:

- **Expected effect size** $\Delta$ on the metric you care about (lap time, balance metric), relative to the noise $\sigma$ of that metric.
- **Cost**: garage time, out-lap/in-lap, tyre life, risk of a reset error (springs/geometry require corner weights and alignment; errors there contaminate subsequent runs).
- **Coupling**: number of secondary effects (springs move platform and aero; geometry moves camber, roll centre, bump steer simultaneously).
- **Reversibility**: can you return to baseline exactly?

Pressures and ARBs score well on all four. Dampers score badly on effect-size-to-noise: a 2-click change might move lap time by less than the lap-to-lap scatter, and the driver adapts within a lap. That's why damper work is best done with **objective metrics** (histograms, gated roll/pitch velocity, tyre load variation) or on a 7-post rig, rather than by subjective A/B alone.

**Interactions** are where experience pays. Examples:
- **Roll stiffness and LLTD vs roll-centre change**: moving the rear roll centre up shifts geometric load transfer rearward instantly (no roll needed) while ARBs work through roll. The result at steady state can be similar, but the transient differs.
- **ARB and spring at the same axle**: both add roll stiffness, but springs also add heave stiffness – ARBs are the "pure balance" tool, springs the "platform" tool. If you want balance only, use the bar.
- **Pressure and camber**: changing camber changes the contact pressure distribution, so temperature profile and optimal pressure move together. Change pressure first, then camber, then re-target pressure.

**Change decision tree** used by many engineers: (1) Are tyres in window? (2) Is the problem phase/speed-specific? Choose phase-specific tools. (3) Global balance? ARB or wing. (4) Platform/compliance problem (bumps, kerbs, bottoming)? Springs, packers, ride height, HS damping. (5) Only when the above are exhausted, or when data show a kinematic cause (e.g. bump steer, camber loss), touch geometry.

## L4 · Expert

For the lead engineer, the hierarchy is embedded in a **run plan** that is itself a design of experiments. Considerations:

**Sensitivity maps.** Before the event, use lap simulation or DIL to estimate the gradient of lap time and of balance metrics per click of each tool, $\partial t / \partial x_i$, and the cross-terms $\partial^2 t / \partial x_i \partial x_j$ for strongly coupled pairs (rear ARB × diff power lock, rake × wing). Large cross-terms mean "one change at a time" will give a misleading local optimum; you then test **small factorial designs** (2×2) on those pairs.

**Non-linearity and the tyre window.** Setup response is not linear in the tyre: if a change moves the tyre across its temperature window (e.g. more rear roll stiffness overheats the outside rear in a long fast corner), the sign of the lap-time effect can flip within a run. Evaluate on representative tyre state and log temperatures/pressures with the run.

**Driver adaptation.** Drivers adapt in 2–4 laps, which biases short A/B runs toward the incumbent setup on the first laps and toward the new one later. Plan ABA (or ABBA) sequences so that drift (track evolution, fuel, tyre) and adaptation cancel to first order.

**Risk-weighted value.** Setup changes with big potential but high reset risk (geometry, springs before qualifying) belong in long FP sessions or the simulator, not in the 20-min session before qualifying. The 'quali-safe' set is pressures, ARBs, wing, bias, diff/TC maps.

**Homologation and BoP context.** In GT3/LMDh many items are homologated or range-limited; the hierarchy then compresses to the free items. In parc-fermé formats (F2/F3 after qualifying in many events) the decision of *which* tools to use before the freeze is strategic: carry robustness (e.g. damper settings that tolerate fuel burn and tyre deg) rather than a one-lap optimum.

## Key formulas

**Effect size vs noise (decide if a change is worth testing)**

$$z = \frac{\Delta}{\sigma\sqrt{2/n}}$$

- \Delta: expected change of the metric (e.g. lap time, s)
- \sigma: lap-to-lap standard deviation of the metric
- n: laps per configuration

## Trackside takeaways

- Default order: tyre pressures → ARB → wing → springs → dampers → geometry (fast, predictable, uncoupled first).
- Phase-specific problems: use phase-specific tools (bias, diff, TC) before global ones.
- Make changes big enough to be felt and measured (ARB 1–2 steps, wing 1–2 clicks, dampers 2–4 clicks).
- One change at a time; compensating pairs count as one intent and must be documented.
- ARB = balance tool; springs = platform tool (they also move ride height and aero).
- Bracket when the direction is unknown, then interpolate.

## By car category

### GT3

Many GT3 cars have bars adjustable in seconds and wing in homologated clicks; springs come in fixed homologated sets – so pressures, ARB, wing and electronics are the in-session tools.

### One-make Cup

With a spec setup range (e.g. Porsche Cup: ARB holes, wing positions, camber/toe, ride height), hierarchy collapses to pressures → ARB → wing → alignment.

### LMP2

ORECA 07: heave elements and packers are key platform tools; changes to them alter aero balance as much as mechanical balance – treat as "platform + aero" changes.

### Formula 4

Tatuus F4: limited adjustability (ARB, wing, ride height, camber/toe, dampers on some series) – pressure and ARB work deliver most of the gains.

## References

- Milliken & Milliken, Race Car Vehicle Dynamics (SAE 1995): Ch. 12 Chassis Set-up (order of adjustments)
- Milliken & Milliken, Race Car Vehicle Dynamics (SAE 1995): Ch. 11 Testing and Development (test planning)
- Segers, Analysis Techniques for Racecar Data Acquisition, 2nd ed. (SAE 2014): §9.3 Using Roll Gradients as a Setup Tool
- Segers, Analysis Techniques for Racecar Data Acquisition, 2nd ed. (SAE 2014): §9.5 Using the Roll Ratio as a Setup Tool
