Optimising the engine and suspension of racing cars has always been a technologically intensive pursuit but the ability to log and then later analyse data has taken the sport to a new level.

Australian motorsport specialist MoTeC is at the forefront of racing car electronics, producing digital dashboards, engine management systems and data analysis software.

This month we’ll look at how racing car data is collected and then next month, examine MoTeC’s i2 data analysis software.

So what sort of data is collected from a racing car?

Engine

Collecting data on the engine status is made simpler because the engine management system’s Electronic Control Unit (ECU) already uses many sensors.

The outputs of these sensors can be used not only by the engine management system but also logged and then expressed in engineering units.

In addition, the ECU has available internally calculated data, such as injector duty cycle.

On naturally-aspirated race cars, load is normally calculated by the engine management ECU looking at engine speed and throttle position. (This is in contrast to road cars that most often use an airflow meter to directly measure the mass of ingested air.)

Fully configurable digital electronic dashboards are used to display sensed data and log it for later analysis. In addition, output alarms can be set when certain combinations of parameters are met. Data can also be scrolled through by the driver pushing a button.

Forced aspirated racing cars (that is, those with turbo or supercharged engines) use a Manifold Absolute Pressure (MAP) sensor that measures manifold pressure. When this is combined with measurement of engine speed, the ECU can again work out load. So when engine load is logged, the data is in the form of either throttle position and engine speed, or manifold pressure and engine speed.

While it might first appear that this is a complex way of logging engine load, in fact in racing car applications it is advantageous.

This is the case because engine load is most often used in conjunction with the logged air/fuel ratio to work out where in the load range the engine is running richer or leaner than desired.

Since the fuel injector outputs are determined from an ECU map of throttle angle (or MAP) versus engine speed, having available throttle position and rpm (or MAP and rpm) allows the engineer to quickly find the load site at which the problem is and then make the appropriate tuning change.

MAP sensors are calibrated in absolute pressure and are most commonly available in 1 Bar (suitable for naturally aspirated engines), 2 Bar (ie suitable for 1 Bar of boost) and 3 Bar (suitable for 2 Bar of boost) versions.

Interestingly enough, there is also available a 1.05 Bar version which takes into account the aerodynamic air pressure build-up possible in the airbox of a fast-moving car.

Most often used are Delco MAP sensors which start at $80. These conditioned sensors have a nominal 0-5V output and are widely used in production cars.

MoTeC's dashboard display and logger uses surface- mount components, a military spec connector and heavy-duty aluminium construction.

Coolant and oil temperatures are measured by NTC thermistors. The Bosch 023 and 026 sensors are commonly used – at $17 they are cheap, use a near universal 12 x 1.5mm thread and are 2-wire designs (ie, no chassis ground return) that use a standard fuel injector plug.

Intake air temperature sensors comprise a similar design but with the thermistor exposed to the passing airflow.

Intake air sensors can be used to sense air temperature in an intake runner just prior to entry into the engine (so measuring the temperature rise caused by the air compression of a supercharger or turbo, and intake manifold heat soak) or at the airfilter.

Previously, the air/fuel ratio was measured by a zirconia oxygen sensor such as the Bosch "four wire" design. Based very much on the technology of the oxygen sensors used in normal passenger cars, this device outputs a voltage of 0-1V, depending on mixture strength. However, the voltage is non-linear with respect to air/fuel ratio, with a sudden change in output around 450 – 550mV (corresponding to the air/fuel ratio passing through stoichiometric) and also varies with temperature. The Bosch unit has a slightly flatter response than garden variety oxygen sensors but still has severe limitations in accuracy, especially at the rich end of the automotive scale. Linearising it requires accurate temperature and voltage compensation.

An infrared receiver placed in the car watches for the output of a suitably coded trackside infrared transmitter. In this way, accurate lap times can be logged and also displayed on the in-car digital dash.

Replacing the Bosch "four wire" unit is the Bosch LSU probe. This probe works on a completely different principle and requires its own control circuit. In short, a zirconium-dioxide/ceramic measuring cell is used comprising a Nernst concentration cell and an oxygen pump cell, with a small diffusion gap positioned between them. Two porous platinum electrodes are placed within this gap – a Nernst measuring electrode and an oxygen pump electrode. The gap is connected to the exhaust gas via a small passage.

On the other side, the Nernst cell is connected to the atmosphere by a reference air passage. By applying a pump voltage across the electrodes, oxygen is pumped from the exhaust gas into or out of the diffusion gap.

The sensor controller varies this voltage so that the composition of the gas in the diffusion gap remains at stoichiometric.

If the exhaust gas is lean, the pump cell pumps the oxygen to the outside (positive pump current). If the exhaust gas is rich, the oxygen is pumped from the exhaust gas into the diffusion gap (negative pump current). The pump current therefore reflects the actual air/fuel ratio. Again, linearising is required.

Other than the most recent M400/600/800 series MoTeC engine management systems and the PLM air/fuel ratio meter, no MoTeC logging device can accept a signal directly from the LSU sensor. Instead they read the data from the ECU or PLM via a CAN bus communication, while the PLM also has a configurable analog output voltage that can be read by the data logger.

These expansion units allow a greater number of inputs to be logged by the digital dash or the engine management ECU. The E816 has an additional 18 analog voltage inputs and eight PWM outputs, while the E888 has eight analog voltage inputs, eight K-type compensated thermocouple inputs, four digital inputs and eight PWM outputs.

Exhaust gas temperature is measured with K-type thermocouples. Again, an interface device is needed, this time to amplify and cold junction compensate the signal.

One example of such an interface is the $1045 MoTeC E888 input/output expander. Amongst other inputs and outputs, this unit can accept eight K-type thermocouple inputs and then communicate this data to the engine management ECU or digital dash logger by means of a CAN bus connection.

Exhaust gas temperature is most often measured at individual exhaust outlets near the engine, so explaining the requirement for eight probes in many race car applications. These give a guide to cylinder-to-cylinder mixture consistency and are most commonly used in drag racing.

Two types of sensor are used in these applications. The first is the traditional Bourdon tube based potentiometer, as exemplified by the large canister VDO units used as oil gauge pressure sensors on countless road cars.

However, the accuracy of these sensors in race car applications is suspect: when tested on the bench, a light finger tap can sometimes change the measured output by 5 psi!

Manifold Absolute Pressure (MAP) sensors made by Delco are used in conjunction with RPM and intake air temperature to measure load. These sensors are available in 1, 1.05, 2.3 and 4 Bar versions

Individual cylinder exhaust gas temperatures are often measureed and logged to indicate cylinder-to-cylinder mixture consistency

Replacing these are Texas Instruments sensors that use a load cell backed by a diaphragm. Available up to 2000 psi maximum pressures, these sensors have a conditioned 0-5V output and are available in gauge and absolute pressure configurations.

Throttle position sensors comprise rotary potentiometers mounted on the throttle shaft. They are available in a wide range of physical designs to match various shafts but a common one accepts a D-shaped shaft. As we will see next month, knowing what the driver is doing with the throttle is a vital component in race car data analysis.

Bosch LSU sensors are used to sense the oxygen concentration of the exhaust gas and from this, work out the actual air/fuel ratio. These new sensors replace the older zirconia design and are faster and have higher accuracy over a wider measuring range

Linear potentiometers are used to sense damper movement. Data interpret- ation software allows damper speeds to be calculated from this displacement data, allowing optimal bump and rebound settings of the dampers to be set.

Engine Speed

Engine speed is sensed from the crankshaft position sensor. This normally comprises an inductive sensor mounted on flywheel, although in engines not specifically built for racing but instead adapted from road cars, the sensor can alternatively be optical or use a Hall Effect device.

Other sensors that are sometimes uses on the engine include infra-red thermometers measuring block temperature and pressure sensors in the coolant system, the latter used primarily to sense a catastrophic loss of coolant.

Chassis and Suspension

Suspension data requires the installation of specific sensors. Where the behaviour of individual wheels needs to be monitored, this involves four sets of sensors.

Hall Effect sensors are used to sense engine speed, a parameter used by the engine management ECU and also logged for later analysis.

Load-cell- based pressure sensors are used to measure oil, brake and fuel pressures. In some cars, even the coolant pressure is measured!

Damper Movement

Damper movement is sensed by linear potentiometers. These are available with different stroke lengths (for example: 75, 100, 125, 150 and 200mm) and are mounted such that they move over as much of their range as possible as the suspension moves from full bump to full droop.

These sensors cost about $400 each but they are fully rebuildable, something often required as their vulnerable positioning results in frequent damage in racing incidents. Finding space for the sensors and mounting them so that no bending loads are placed on them can be difficult; however, the logging software can be easily configured to show actual suspension deflection even when the sensor is angled from the vertical or is subjected to a non-linear motion ratio.

The temperature of the oil within the dampers is sensed indirectly, either by the use of stick-on thermocouples or, less commonly, by infra-red temperature sensors.

One, two and three axis accelero-meters are used to sense accelerations. These sensors are conditioned with a 0-5V linear output and can be specified to have maximum acceleration of 10g. (In Top Fuel drag cars the previous 4g maxima were being exceeded in longitudinal acceleration!) However, in circuit racing cars, two-axis accelerometers with a maximum acceleration of 4g are more normally used.

Fully programmable engine management units like this MoTeC design incorporate memory for data logging. Engine sensor data is already available to the unit and suspension data can be communicated to it from the digital dashboard by CAN bus.

Cost varies from $360 for a single axis 4g accelerometer to $688 for a 3-axis 4g sensor.

As we will cover next month, the outputs of this sensor can be used by the data analysis software to automatically construct a track map.

The accelerometer is normally mounted at the roll and pitch centre of the car. However, two accelerometers can be individually mounted on the front and rear axle lines and when their outputs are compared to steering angle, be used to assess the magnitude of oversteer and understeer.

Yaw is sensed by a Bosch yaw sensor, as normally fitted to the Subaru STi model WRX that uses an active centre differential as part of its four wheel drive system. In addition to a yaw rate signal output, this sensor also contains a lateral accelerometer. Cost is $1014.

A dual axis accelerometer is used in most data-logged racing cars to sense lateral and longitudinal acceleration. The unit is designed to work up to 4g and outputs a conditioned 0-5V signal.

Tyre and brake temperatures are monitored by infrared thermometers aimed appropriately. In the case of Le Mans racing cars, no less than three infrared sensors are used per tyre – quite a cost at $480 each sensor! Tyre temperatures are amongst the most useful of data in setting-up a car for optimal lap times as the temperature distribution shows how hard each tyre – and each part of the tyre – is working. The infrared thermometers have a conditioned 0-5V output and are available in 100°, 200° and 1000° Celsius ranges – the latter being used to measure brake temps.

Steering angle is normally sensed by a multi-turn rotary potentiometer driven by a toothed rubber belt from a pulley mounted on the steering shaft.

In road cars adapted for racing, the ABS system is usually disconnected. In these cases, one of the inductive wheel speed sensors can then be used for measuring road speed. The logging software is configured for the AC voltage levels of the sensor and the frequency/speed relationship. In purpose-built race cars, a new inductive sensor is fitted behind a wheel.

Lap Time

Car racing is about going faster than anyone else and so lap speed is a critically important parameter.

MoTeC use a trackside mounted infrared transmitting beacon and a car mounted receiver. A configurable frequency signal is emitted by the beacon and the car’s system is programmed to respond to only this signal.

Lap times are logged and also displayed to the driver in terms of laps to go or lap number. In addition, split times can be gained by the use of extra trackside beacons programmed appropriately.

Logging and Displaying
the Data

Given the number of channels and the frequency at which many are collected, most teams choose to use in-car logging rather than real time telemetry. (Telemetry is still used but for slow-changing factors like fuel levels and monitoring engine health.)

It is useful if the device that stores the data can also display some of it for the driver and so a common approach is to use a customisable digital dashboard that can perform both functions. MoTeC’s Advanced Data Logger (ADL2) is such a unit.

The ADL2 can read 28 analog voltage inputs, 12 digital inputs and two Bosch ‘four wire’ air/fuel ratio sensor inputs.

And if even more logging capability is required, another 22 inputs can be added by means of an expansion unit! The unit will also accept data communicated to it in RS232 (eg, from a GPS unit) and CAN formats. A 16Mb internal memory is incorporated and the microprocessor is 32-bit. The fully configurable backlit LCD can display any of these inputs, shown in user-selectable engineering units.

Conclusion

As we’ve seen, literally anything that can be sensed on a race car is capable of being logged. However, all the information in the world is of little use if no sense can be made of it.

Next month, we’ll take a look at the MoTeC i2 data analysis software which has mind-boggling capabilities – not only can it display the data in many different forms but it can also make mathematical calculations based on that data and then display those calculations in relation to the collected data!