Anti-lock why not?
Page 49
If you've noticed an error in this article please click here to report it so we can fix it.
by P. A. C. Brockington, MIMechE
THE DEVELOPMENT of a form of anti-lock system which is commercially acceptable may be regarded as a top priority requirement for artics in particular and lorries in general. It's time it came in—after all, it's 15 years since an anti-lock type of brake was first successfully applied to aircraft. This system has proved a boon in obviating tyre bursts, eliminating flats caused by locked wheels, and enabling planes to be braked to a standstill in a relatively short distance.
Just why has there been this delay? This is a pertinent question, particularly so in view of the fact that several systems for road vehicles have proved their worth experimentally.
One of the most important contributions to this subject was a recent SAE paper on the evolution of the Sure-Track brake system for private cars. The point is that it relates to experiments sponsored by a manufacturer—Ford of America—in conjunction with a maker of auxiliary equipment, the Kelsey-Hayes company.
The authors of the paper are Mr. R. H. Madison of Ford and Mr. Hugh E. Riordan of Kelsey-Hayes. They indicate the complexity of the problems of mass production application by the impresssive list of the names of additional conderns that were represented in the team of technicians who developed the system. These included the Hydro-Aire Division of the Crane company, Texas Instruments Inc., the Brown Engineering Division of the Teledyne company, Cornell Aeronautical Laboratory, Battelle Memorial Institute, Rensselaer Polytechnic Institute, Booz Allen and Hamilton and various consultants and specialists. In its latest form the system is applicable to the rear wheels of a car, the brake actuators of which are of the vacuum-servo/hydraulic type.
Electro-magnetic sensors in each wheel produce a.c. voltage pulses at a rate that is proportional to wheel speed. An electronic control module receives the pulses and signals the brake actuator when locking is imminent. The brakes are then released and are re-applied when the speed of the wheels is increased to a given rpm.
There is nothing very complicated in the system. But details of development work indicate the importance and difficulty of adapting the vehicle to the system as well as the system to the vehicle. The authors refer to the "massive effort" that was required in analysis, computer simulation and vehicle testing to enable the techniques evolved to be applied to the system.
Possibly of greater importance with regard to applications covering a variety of vehicle types, is the fact that incorporation of the sensor units in the back axle assembly was a major exercise. And operation of the system initially produced serious forms of vibration in the suspension.
The authors mentioned in particular the necessity to allow for "split-coefficient" conditions arising from variations in lateral road friction. If the variation is substantial it is necessary that the wheel , on the high-friction surface is not braked to an extent sufficient to cause instability and that the control system reacts to this requirement. It is also noted that if stopping distances on high-friction surfaces are to be acceptable, both wheels have to be driven at a speed giving a relatively high slip. Allowance has to be made for unbalanced brakes by providing sufficient brake pressure to ensure that the under-braked wheel operates with adequate slip, and it is essential that the control system adjusts braking to cater for this condition.
Sensor units Perhaps the most significant aspect of the paper with regard to development problems relates (as indicated) to the difficulties encountered in evolving suitable sensor units. For example, a unit that necessitated teeth being cut on the brake-drum flange was unsatisfactory because of its vulnerability to damage. A system based on a rotating toothed ring inside the drum and a stationary magnetic pick-up operated satisfactorily in experimental form. However, a unit designed for production was unsatisfactory because its performance was affected by brake tolerances and deflections.
When a single sensor was driven by the transmission (in place of the speedometer) a good signal was generated but the backlash and vibration between the 'sensor and rear tyres created "electrical noise" which the control module sometimes interpreted as skid signals. Various forms of small wheel sensor were then tried and each type was thoroughly tested in service conditions and in the laboratory.
As development proceeded the system became increasingly sophisticated, but many design modifications had to be incorporated in the units before misinterpretation of signals by the control module was eliminated. Difficulties that had to be overcome included the inability of the control module on occasions to identify the skid signal because of the small difference between the velocity signal and the noise signal generated by axle deflections.
Over 15,000 road tests of vehicles equipped with the production type of Sure Track system have been conducted in addition to laboratory tests and computer analyses. It is claimed that the system (applied to the driving axle only) provides for the control of yaw stability within close limits under all conditions, and that stopping distances are better or not more than normal distances regardless of the coefficient of friction of the road surface or the speed at the time the brakes are applied.
The most important lesson to be learnt from this paper is obviously that catering for deflections of the back axle is essential in the case of a system having sensors located in the axle or wheels (which is normal practice) or an even greater range of deflections if one sensor is driven from the transmission. The range of deflections of a commercial-vehicle driving axle or transmission are far greater than the range of a typical American car and it can be assumed that the difficulties in evolving a system for trucks and buses are correspondingly more complex.
Possibly the major problem arises from the difficulty of eliminating response by the control module to signals that are derived from sources other than the sensors. It is known that the producers of anti-lock systems in the UK are fully aware of the types of problem mentioned. Vehicle manufacturers who decide that anti-locking brakes are a "must", may well have to cope with the major problem of redesigning chassis components to give freedom from unwanted signals.