Sensor technology: temperature, pressure, displacement, acceleration, strain, speed and GPS
Choose sensors by asking what, what range, what environment, what accuracy and what budget. Know the working principle and failure modes of the sensors on your car so you can trust – or distrust – their data.
Sensor technology: temperature, pressure, displacement, acceleration, strain, speed and GPS: L1 Foundation explanation
Selection questions (Segers): What must be measured (temperature, pressure, flow, displacement, velocity, acceleration, force)? What range? What environment (heat, vibration, fluids, EMI)? What accuracy is really needed? What budget?
The usual suspects on a race car:
| Quantity | Typical sensor | Notes |
|---|---|---|
| Fluid temperature | NTC thermistor, RTD (PT100/PT1000) | accurate, limited range |
| Exhaust temperature | K-type thermocouple | to ~1250 °C |
| Tyre / brake surface temp | Infrared (single-point or 8–16-channel arrays) | surface only, emissivity |
| Oil, fuel, brake, MAP | Piezo-resistive pressure | absolute / gauge / differential |
| Damper, steering, throttle | Linear / rotary potentiometer, Hall non-contact | wear on wipers |
| Ride height | Laser triangulation | filter road texture |
| Chassis acceleration | MEMS capacitive IMU | measures DC (steady g) |
| Vibration, knock | Piezoelectric accelerometer | AC only, high bandwidth |
| Wheel / engine speed | Hall-effect or VR on trigger wheel | low speed resolution limited by tooth count |
| Loads | Strain gauges (Wheatstone bridge) | temperature compensation |
| Position, speed | GPS / GNSS + IMU | 1–3 m position, 0.1–0.2 % speed |
Sensor technology: temperature, pressure, displacement, acceleration, strain, speed and GPS: L2 Practitioner explanation
Temperature.
- Thermocouples generate a small voltage from the junction of two metals; they need cold-junction compensation and special extension wire. Grounded junctions respond faster than ungrounded; exposed junctions are fastest (air temperature) but only in clean, dry environments.
- Thermistors change resistance strongly with temperature (NTC: resistance falls) – very sensitive and accurate, non-linear, range limited to a few hundred °C. Typical for water/oil.
- RTDs (platinum) are standardised and interchangeable, wider range than thermistors.
- Infrared sensors read radiated energy: fast (ms), non-contact, work on rotating targets. Limitations: surface temperature only (not carcass), need correct emissivity (rubber ~0.95; polished metal is very low and unreliable), lens contamination, field of view (spot size grows with distance – a sensor too far away averages tread and rim).
Pressure. Piezo-resistive silicon diaphragms with bridge resistors. Absolute (vacuum reference: MAP, barometric), gauge (atmospheric reference: oil, fuel, brake), differential (two ports: pitot, aero pressure taps, filter Δp). Typical ranges: oil 0–10 bar, fuel 0–10 bar for low-pressure supply (200–350 bar on direct-injection rails), brake 0–200 bar (some up to 250), aero differential ±5–10 kPa.
Displacement. Linear potentiometers for dampers (50–150 mm stroke), rotary for steering and throttle, string pots for awkward packaging. Non-contact (Hall/magnetostrictive) for life. Laser ride-height sensors work by triangulation; resolution is best at short range; filter out asphalt texture and expect dropouts over kerbs, painted lines and water.
Acceleration. Capacitive MEMS sensors respond down to DC – needed for steady lateral g; typical range ±5 g on the chassis. Piezoelectric sensors have no DC response but huge bandwidth and range (±50–500 g) – used for knock and upright vibration. Their usable band ends well below the resonance , and they lose the content below ~0.5–3 Hz, so for hub acceleration used in contact-patch force estimates prefer DC-coupled high-g MEMS or piezoresistive sensors.
Strain and load. Strain gauges change resistance with strain: for metal foil. Strains are tiny (10⁻⁴–10⁻³), so gauges sit in a Wheatstone bridge with an amplifier; full bridges cancel temperature and bending effects. Pushrod and wishbone load cells give suspension forces; driveshaft torque uses telemetry or magneto-elastic bands.
Speed. Hall-effect sensors on a toothed wheel are robust to dirt; resolution at low speed depends on tooth count and logger timing.
GPS. Standalone position ±2–3 m; IMU-fused ±1–2 m; RTK/dual-frequency systems reach centimetres. Speed accuracy ~0.1–0.2 % – far better than wheel speed for distance. Dual-antenna systems also give heading, and with an IMU, body slip angle.
Sensor technology: temperature, pressure, displacement, acceleration, strain, speed and GPS: L3 Advanced explanation
Installation is half the measurement. Most sensor failures in racing are not the sensor itself but: connectors (vibration fretting, water), looms rubbing through, sensor mounts flexing, and EMI from ignition coils or hybrid inverters. Shielded twisted-pair wiring, proper grounding (single-point), and strain relief prevent the majority of "noisy channel" issues.
Temperature lag. A sheathed thermocouple or thermistor has a time constant of seconds. A "slow" oil temperature trace is physics, not a fault – but it means you can't correlate a temperature spike with a 0.5 s event. IR sensors on brake discs need the right wavelength and emissivity correction; disc temperatures on carbon brakes read differently from cast iron.
Damper pots and motion ratio. If a pot is mounted parallel to the damper, it measures damper travel directly. If mounted on a rocker or at an angle, correct with the geometry. With MR = damper travel / wheel travel, wheel travel = pot travel / MR(x) – divide, do not multiply – and use the MR-versus-travel curve from the kinematics (it often changes 5–15 % over the stroke on rocker suspensions), checked by a wheel-lift test: jack the hub in 5 mm steps and log the pot. Pot wipers wear and produce dropouts/noise at the most-used position (static ride height) – exactly where you look most.
Ride-height lasers on ground-effect cars. Mount positions differ from the aero reference points; you must transform measured heights to the aero-map reference (front/rear axle or skid heights) using pitch and roll. Errors of 1–2 mm are significant when the aero map gradient is steep near the floor.
Load-cell calibration. Pushrod load cells need calibration in the installed condition (bending and temperature). Pushrod force is not wheel load: convert it through the installation ratio and angle, and add the unsprung weight, which goes straight to the ground. Validate on the scales: Σ(pushrod force × installation ratio) + Σ unsprung weight = car weight, with each corner within ~1–2 % of the corner-weight reading.
GPS limitations. Multipath near pit buildings and grandstands, tree cover and satellite geometry cause jumps; position comparisons between different days can be offset by metres (satellite drift) – compare driving lines within a session or use track-limit references. Software blends IMU data through short outages.
Key formulas
- GF
- gauge factor (≈2 for foil gauges)
- strain (–)
- bridge excitation voltage (V)
- k
- stiffness of sensing element (N/m)
- m
- seismic mass (kg)
- air density (kg/m³)
- V
- airspeed (m/s)
Trackside takeaways
IR measures surface temperature only and depends on emissivity and spot size.
MEMS accelerometers for chassis g (DC); piezo accelerometers for vibration and knock (AC).
Strain gauges need a bridge, amplification and temperature compensation.
GPS speed is excellent for distance; GPS position drifts between days.
Most "sensor failures" are wiring, connectors and mounts – install like it matters.
On your car
Factory packages include IR tyre sensors and TPMS; teams add damper pots and ride-height lasers for testing. Watch TPMS data for slow punctures and mandatory minimum pressure compliance.
Check yourself
Why is a piezoelectric accelerometer a poor choice for measuring steady lateral acceleration in a long corner?
- Segers – Analysis Techniques for Racecar Data Acquisition — §19.5–19.16 Temperature, pressure, displacement, acceleration, speed sensors, strain gages, torque, pitot, oxygen, GPS, laser, SAW
- Segers – Analysis Techniques for Racecar Data Acquisition — §14.7.3 Driving Line Analysis Using GPS · cross-day GPS position offsets