Transense and KERS Technology
Published on Sunday, 06 December 2009
A Kinetic Energy Recovery System (KERS) is not a new idea. Indeed the principle, often referred to as “regenerative braking”, by which kinetic energy of motion, normally lost when the brakes are applied, can be partially recovered, has been understood for more than 100 years:
On reaching the top of a long steep hill, if we do not want to coast, we convert the motors into dynamos, while running at full speed, and so change the kinetic energy of the descent into potential in our batteries. This twentieth century stage-coaching is one of the delights to which we are heirs, though horses are still used by those that prefer them.
From: A Journey in Other Worlds, by John Jacob Astor IV, published 1894
The Swiss Gyrobus, designed and manufactured by Oerlikon in the 1940s relied on stored flywheel energy to power the bus. Electrical regenerative braking was employed to add energy back to the flywheel thereby extending the vehicle range.
The braking torque created by allowing the normally driven wheels of the vehicle to back-drive the electric motor to add charge to the battery, or to spin up a flywheel, can also be sufficient for most car braking purposes except emergency braking or slowing the vehicle to a stop. So, while a standard braking system is still required, it can be lighter and the friction surfaces should last many tens of thousands of miles.
The idea of adding torque from a stored electrical energy reserve (a battery or ultra capacitor) in order to boost acceleration has been introduced with parallel hybrid systems, where either an internal combustion engine or an electric motor or a combination of both provides motive power, such as in the Toyota Prius.
In 2009, F1 motor racing entered the hybrid car world with an FIA inspired optional addition of KERS to provide a “push to pass” acceleration boost option which could be derived from either mechanical (flywheel) or battery plus electric motor sources. The concept had dual motives: to increase the excitement of F1 by increasing overtaking opportunities and to promote the green potential of energy saving technology by using the highly tuned and timely engineering skills of F1 teams and their suppliers.
However the FIA did not want a “free for all” approach which might result in significantly different performance from different designs leading to a lack of competitiveness, so it introduced strict rules:
Racecars can employ KERS to store up to 400kJ of energy per lap, to be reused via a 'boost' button which allows drivers to add up to 60 kW of power to their ic engine.
In practice, KERS (2009) enabled around 10% more power for almost 7 seconds per lap.
The specification reads quite simply, but in engineering terms it set some real measurement challenges:
Power is the product of KERS motor torque and speed (rpm). Both must be precisely known at every moment so that the FIA limit is not exceeded – not trivial when it is realised that engine speed in a low rotary inertia F1 engine varies with every degree of crankshaft rotation.
Energy is the product of power and time. So, when harvesting kinetic energy, KERS torque multiplied by rpm needs to be integrated over each lap time and must never exceed the FIA limit.
Systems in preparation during 2007–8 included both electrical and mechanical energy storage (flywheel) approaches. However in the event, although very promising in principle, neither of the two flywheel systems appeared on the track in 2009.
A typical electric KERS F1 configuration locates the electric motor in front of the ic engine, connected via a small shaft spur-geared to the front of the crankshaft. The shaft, turning at crankshaft speed, must be instrumented to measure torque, which can be in either sense: driving (positive) when the electric motor adds torque to the engine, and generating (negative) when the road wheels back-drive through the powertrain and engine to the electric motor which in turn creates a substantial braking effect on the car.
It was at this point in early 2008 that Transense Technologies were invited to propose a KERS F1 torque sensing system comprising hardware, interrogation electronics and application software.
A 2009 spec F1 engine running at up to 18,000 rpm will generate very high centripetal accelerations, acting inwards, on a sensor mounted on the shaft surface – approaching 4000g for the shaft in question. This means that a force equal to 4000 times the weight of the sensor acts outwards trying to tear it away from the shaft. In addition engine vibrations are intense, circa 100g, while operational temperatures range from 70 to 170°C.
Transense employed their surface acoustic wave (SAW), quartz substrate, dual and triple resonator sensors to opposing sides of the shaft within a hermetically sealed cavity. A non-contacting radio frequency (RF) rotary coupler provided two way signal connection between sensors and miniaturised interrogation electronics located within the electric motor power electronics and control box.
Software, which analyses the RF signals using a Discrete Fourier Transform (DFT) algorithm to generate independent torque and temperature signals with 3 kHz update rate, together with further code enabling the FIA to check that performance of the KERS torque system stayed within their requirements, was also provided by Transense.
Static torque measuring performance (linearity, repeatability, hysteresis, creep and rotation) was established initially at Transense laboratories near Oxford. Typically at 120°C, the combined errors were less than 1% of full scale.
A string of 3 F1 KERS shafts being calibrated within an environmental chamber over the full operational range of torque and temperature
Development proceeded in close co-operation with the electronics and engine suppliers to the F1 race team. Visits from FIA technical representatives were made to understand SAW technology and to ensure that company systems were in place to provide software version control and retrospective traceability.
Dynamic testing, at the engine manufacturer’s facility, were carried out on both electric and engine dynamometers. The acid test on the actual racecar followed.
Manufacturing and calibration of KERS shafts including bonding of quartz dies, gold wire interconnects, hermetic sealing and calibration to meet the 2009 season’s requirements were completed at Transense.
Clarence Pilgrim carries out gold wire bonding operations on an F1 KERS shaft in the Class 10,000 clean room
In a typical race weekend, F1 cars run 2 practice sessions, 1 – 3 qualifying sessions and the race itself, totalling up to 500 miles. As the 2009 season closes, Transense have supplied one race team with torque sensing systems throughout, which have enabled it to demonstrate the benefits of kinetic energy recovery and to make some exceptional passing manoeuvres!
Article contributed by Dr. Ray Lohr, October 2009