Optimised Braking and Turns

Optimised Braking and Turns

“Optimised braking and turns” is how Professor David Cebon, from Cambridge University, characterises the work he has been doing with his colleague, Dr Graeme Morrison, and in collaboration with brake manufacturer Haldex.

Optimised Braking and Turns

By monitoring vehicle dynamics, the system can decide where intervention is needed. Sensors on the truck and trailer feed back data to the controller. The driver’s steering input is also monitored.


As a truck goes round a corner, lateral forces are the same on all the wheels. When the brakes are applied, the stopping forces on the wheels goes up but the lateral forces holding the truck in line reduce. The control actions of the system are based on the position the prime mover and trailer should be in, in terms of speed, articulation angle, steering wheel angle, and what the sensors on board observe to be the case. Inputs are selected and timed to return the truck to its ideal position.


The system then uses this data to work out how the vehicle actually is performing and what it should be doing at that speed with that steering angle. It then compares the two situations and creates what David calls “slip attenuation vectors,” which are how much the slip should be adjusted on each axle to keep the truck in line while maximising braking force.


The system is being asked to brake hard, to which it has to add the attenuation factors to keep it on the road. This ‘Attenuated Slip Demand’ (ASD) decided where and when to reduce slip on each axle.


The latest phase of the development has seen tests of the new system taking place on the MIRA test track in the UK. Testing on the track saw a loaded truck fitted with a steering robot fitted and controlled by a GPS, to ensure the truck followed the exactly same path every time it went through the course. The surface of the track is particularly slippery and tests the ability of any system to keep the semi in a line, whilst going round the 100-metre radius corner.


Each of the valves on the actuator was fitted with a micro-controller, received data on wheel speed and brake chamber pressure from sensors located at the corresponding wheel. A central computer communicated with the local controllers via CANbus. From this hub, the local controllers receive either a vehicle speed and demand wheel-slip signal in slip control mode, or a demand brake chamber pressure signal if operating in pressure control mode.


The driver got to a set speed and then shifted the gearbox into neutral just before the brakes were applied. When testing the initial electronic braking system (EBS) results, the driver applied full brakes until the vehicle came to complete rest.


To test the ASD, a test engineer in the passenger seat sent the signal for an emergency stop from a laptop to the controller, releasing the brakes once the vehicle came to a complete stop.


For the tests the trucks are going at exactly the same speed on the track in exactly the same position. What was immediately obvious to the testers was, when brakes are applied on the corner, the ASD-braked semi holds the line well, but the EBS truck and trailer get much closer to the edge of the road.


More importantly, at the end of the curve the ASD combination is running straight and true on the right track, while the EBS is slowing the truck down, running at an angle to the correct path.


Comparing the effort of the steering robot, it was making multiple turns of the steering wheel with EBS, but there are only small steering inputs with the ASD. In all of the tests the ASD truck was stopping in considerably shorter distances. Using EBS, the steering angle reaches 16o as a maximum, veering from understeer to oversteer and back again. Meanwhile there was only one major steering input on the ASD trial and it only went as far as 6o.


When measuring the path the front steer axle should follow, the EBS truck got up to 400mm off the target track, while the ASD only veered 100mm from the ideal line.


Articulation angle changes dramatically in the different tests, with the worst-case scenario seeing the articulation angle between the prime mover and trailer 11o from where it should be, while the ASD-fitted vehicle kept close to the correct articulation angle throughout the exercise.


“Effectively EBS is inefficient, in terms of both braking and cornering performance,” says David. “A highly efficient braking system linked with slip control can improve the cornering performance, on the sorts of manoeuvres on which you could get into trouble.


“Attenuated slip demand is a simple modification could almost completely eliminate the path error associated with EBS and reduce the driver-steering effort required. Normally, a driver could not keep the truck on the road in this manoeuvre, but could with attenuated slip demand.


“It also improved jack-knife performance and also had a seven per cent better braking distance. We didn’t get the full 16 percent we had gained by changing the valves but kept seven per cent and stability. In a straight line we get the full benefit of the slip control’s improved braking at 16 per cent, but only seven per cent better in the turn. We would really like to test the system, much faster on a nice icy surface, but we couldn’t.”


The next step in development will see the design of a device to measure tyre-to-road surface friction values to aid the ASD in its calculations. This is available in prototype, but is expected to be available when this kind of braking improvement comes to the market.


“I started working on this project over ten years ago,” says David. “What I wanted to do was combine braking and steering. We have done a lot of work on steering. When we looked at the combination, we realised the brakes were going on and off, so the steering angle had to go back and forth as the brakes went on and off. The conclusion was there was no way we were going to do combined steering and braking, so we better do something else.


“We spent a lot of time optimising the brake actuators, so they don’t use any more power than those fitted to current trucks. We are at least five years away from this kind of system being available to the truck-buying public.”