Made to measure

The current generation of Formula One cars are undoubtedly the most complex ever, thanks in the main to the much-touted new power units. The integration of these in the various chassis, as well as a never-ending knowledge wish list from chassis and aerodynamic engineers, means they are also the most instrumented. It is very much a case that if a component moves, or isn’t supposed to move, gets hot or cold or is under pressure, it will be monitored by a host of sensors – making each car effectively a lab on wheels.

For a start, the new power units, consisting as they do of two motor generators, an IC engine, battery storage and control electronics, need constant monitoring. This involves using a multitude of pressure, temperature, position and speed sensors that relay information both to the control electronics and via a telemetry link to trackside engineers. This information is vital for ensuring everything is working in unison, and with every team and engineer acknowledging that power unit control strategies are the key to performance in 2014, this fl ow of data is imperative in ensuring a car is always at its most competitive.

A classic Wheatstone Bridge strain gauge circuit, featuring three reference resistors (R1,2 & 3) and a strain gauging element (Rsg)

On top of this there are the data demands of the chassis engineers and aerodynamicists. At a basic level, engineers need to monitor the loads through components such as suspension uprights in order to ensure they are operating within their design limits. However, with the huge restrictions placed on testing and simulation resources, each track session is also a data gathering exercise. To this end, a team’s engineers will be keen to glean every scrap of information possible from the car on track in order to drive forward and verify their development programmes.

As with any racecar, packaging is a key design constraint and every extra sensor added needs to be integrated into the overall car package; inevitably, that adds weight. Therefore the challenge for teams and sensor manufacturers is to produce the smallest, lightest components possible, yet still ensuring they will survive under race conditions and provide reliable readings. So how has sensor technology progressed, and what have been the driving factors behind their development?

Strain gauges and load cells

Strain gauges are found predominantly in the suspension and aerodynamic components of a Formula One car. They are also used in pressure sensors, but for the moment we will concentrate on the former applications. For example, a suspension wishbone or upright will invariably be instrumented with a number of strain gauges in order to ascertain the loadings it is subject to, this data then being used to aid understanding of a car’s dynamic behaviour. The complexity of the part, and the axes along which it is loaded, will determine the orientation and number of gauges used, but first it is worth looking at how most such gauges operate.

A strain gauge is a sensor whose resistance varies as a force is applied across it. When an external force is applied to an object, stress and strain are the result. Stress is defined as the object’s internal resisting force, while strain is any displacement or deformation that may occur. By measuring these forces, the level of load on an object and the impact that load has on its form can be deduced.

Most strain gauges used in motorsport applications are of the bonded foil type, a design that dates back to before World War II. They consist of a thin foil resistor grid that is bonded using an epoxy resin directly to the surface of the object to be measured. As a load is applied to the object and it deforms, so does the foil resistor, with a resulting change in the resistance of the gauge. By measuring this change in resistance, and comparing it to a known value (obtained by calibrating the sensor under laboratory conditions) the strain on the object can be calculated.

These changes in resistance are very small, so it is normally necessary to use several gauges incorporated into a circuit in order to measure them effectively. This circuit is known as a Wheatstone bridge and is a divided bridge type circuit that allows measurement of static or dynamic resistance, with the output expressed in millivolts per volt input. In this type of circuit, both ‘active’ and ‘reference’ gauges can be used to tailor the sensor to a particular application. For example, a quarter-bridge circuit will incorporate one active gauge, which will be measuring the strain, and three reference gauges (resistors), with a highly accurate voltmeter measuring the differential between the reference gauges and the active component.

Unfortunately though, the materials used to create the strain gauging elements are also prone to having their resistance altered by other external factors, notably temperature. That means steps must be taken to compensate for such variances in order to obtain repeatable measurements. Differences in readings due to temperature variations are a result of two effects that influence the sensing element. First, the electrical resistance of the gauge is somewhat temperature-dependent; as a result, the resistance varies with temperature. The second factor is due to the differential thermal expansion between the gauge and the test part or substrate material to which the gauge is bonded.

One of the more common methods of compensating for temperature changes is to have a ‘dummy’ gauge on one arm of a bridge circuit. This is a strain gauging element, but it is isolated from any strain and provides a reference output that shows the change in resistance due to temperature. This value can then be used to calculate the resistance change in the second gauge that is due to strain.

Often, a part such as a tie-rod or wishbone will be subject to forces along a number of different axes, due to loads being applied from a variety of directions as the car accelerates, brakes and turns. In these situations, a number of strain gauges will be incorporated into the component, oriented to the load paths in order to isolate each specific loading.

As mentioned earlier, strain gauges can also be used to measure loads by being incorporated into a load cell. For example, most teams use load cells in the front wing mounting pylons to ascertain levels of downforce and downforce distribution across the wing. The gauges are bonded into the structure of the load cell, which is normally very stiff, and measure the deformation of the cell as load is applied. The cells can be tailored into functional components, for example a damper mounting, minimising the impact on overall vehicle packaging.

This rotary position sensor has been designed specifically to fit a Formula One car throttle pedal (Image: Gill Sensors)

It should be noted that all strain gauges need to incorporate a means of amplification in order to the turn the output signal, which is in the millivolt range, into one that is suitable for inputting into a data logger or ECU. Often, such amplifiers will also fulfil other roles, such as providing a stabilised voltage to the strain gauge. Some manufacturers also offer solutions such as dual-output amplifiers that provide two different levels of gain. For example, when a car is running along a straight, the load changes detected by the strain gauge will be fairly small, so to obtain detailed data the high-gain amplifier channel will be used. However, as the car runs over the kerbs, the loads vary dramatically, so engineers will instead use the data from the low-gain amplifier channel.

Torque

The ability to measure torque accurately is of vital importance in the 2014 generation of Formula One machines. The relationship between the accelerator pedal and the power unit is no longer a simple one, and it has essentially assumed the role of a torque controller. The car’s electronics will determine how that torque is delivered, governing the contribution of the IC engine and ERS to the force that reaches the rear wheels. In order for the phenomenally complex power unit control strategies to be implemented, accurate measurement of torque at various points in the system – for example the MGU-K and the driveshafts – is imperative.

Torque sensing technology has seen a number of developments over recent years, and there are various methods that can be used. In the past, there were two main methods for measuring the torque being applied to a shaft – using twist angle, and surface strain changes.