As automated or autonomous driving is becoming more widespread, it can be assumed that steer-by-wire systems will prevail. With them, the driving experience can be improved and driving safety can be increased. In addition, OEMs have more flexibility in designing the interior of the vehicle as well as in the engine compartment by eliminating the mechanical components. In this article, Thyssenkrupp describes the advantages and the current state of development of steer-by-wire systems.
HIGHLY AUTOMATED DRIVING PROMOTES STEER-BY-WIRE DEVELOPMENT
Steer-by-Wire (SbW) systems are a further development of today’s electric power steering systems. There is no continuous mechanical link between the steering wheel and the road wheels. The steering command is transmitted purely electrically. An actuator on the steering wheel produces the necessary steering torque (feedback actuator) to create an authentic steering feel for the driver. Another actuator on the steering gear steers the road wheels (steering actuator).
As a result of the increasing trend towards highly and fully automated driving, SbW is currently experiencing a further development surge. Higher levels of autonomous driving require a redundant power supply, communication and steering rack. The road towards a full SbW system is therefore a lot shorter than in the past in regards to cost and technology.
ADVANTAGES OF STEER-BY-WIRE SYSTEMS
Above all, SbW will change vehicle architectures. The elimination of mechanical components means OEMs will have greater flexibility when designing the car interior and engine compartment. Theoretically, the car can be steered from anywhere inside the vehicle with SbW. Also there is no need to design and produce different versions for left- and right-hand drive. As another positive effect, with no direct mechanical link to the road, the vehicle interior is decoupled acoustically. Vibrations and noise are no longer transmitted straight into the cockpit and elaborate sound-proofing measures are no longer needed.
A possible next step in this context would be to retract the steering wheel or to stow it in the cockpit because with SbW the steering wheel does not necessarily have to rotate. This allows the driver to use the vacated space for other things. Also, the suspension can be optimised because traction and braking forces no longer influence steering feel. In addition, SbW paves the way for additional safety-enhancing driver assistance systems. The elimination of the intermediate steering shaft significantly reduces the risk of injury to the driver from mechanical steering components, enhancing passive safety. Of key importance for driving feel are the new degrees of freedom in steering design provided by SbW. These are opened up by the elimination of the fixed steering ratio, permitting an individually selectable and optimised steering feel.
FAULT-TOLERANT REDUNDANCY CONCEPTS
Safety is the top priority in the development of SbW systems. A vehicle with SbW must be safely manoeuvrable even when signal transmission is interrupted or a sub-system fails. The challenges in development lie in creating intelligent solutions for these necessary redundancies at an acceptable extra cost. The redundancies can be created inside or outside the steering system. In principle, there are two ways to design SbW systems either as “fail safe”, which means that in the event of a fault a mechanical or hydraulic link between the wheels and the steering wheel is restored, or as “fail operational” without a mechanical link as a fall-back.
One fail operational solution, for example, would be to add a second steering motor or a steering motor with two separate windings, (1). The traction motor of an electric vehicle, for example, or the brakes could also be used to steer the front wheels in a controlled manner (torque vectoring). By varying the traction or braking forces to the wheels, they could perform a steering function if required, so the vehicle would remain safely steerable .
NEW DEGREES OF FREEDOM IN CONFIGURING THE STEERING FEEL
The many new possibilities offered by SbW solutions are set against a number of technical challenges. One of these is how to enable a realistic steering feel in accordance with the OEM’s steering DNA. This includes the necessary feedback from the road so that the driver can sense the vehicle’s response on the steering wheel at all times. This steering feel differs from OEM to OEM as part of their brand signature. It therefore has to be adapted to any specific requirements.
A base steering feel can already be achieved with a basic torque build-up and a damping component to give the driver a reliable feel for the vehicle in normal driving situations. To tune any such base steering feel, a software-based tool was developed for converting fingerprint measurements of a conventional EPS vehicle into an identical steering feel for SbW in a very short time.
Such a tool is very helpful in the development phase in transferring the steering feel of different OEMs to a SbW prototype. To achieve optimum results, various vehicle measurements at different speeds and lateral accelerations are necessary with the tool. Based on these, the tool delivers a steering feel that corresponds 80 % with that of the measured vehicle. This way, the very time-consuming steering feel tuning can be shortened considerably. Actual steering feel configuration in the vehicle can be limited to fine tuning. (2) shows the result of tuning with the tool. It can be seen that the steering feel generated from the measurements corresponds with the actual steering feel to a large extent. However, to cover all driving situations far more functions are needed.
Road feedback is a particularly critical point, when it comes to steering feel in SbW vehicles. With no direct link between the steering wheel and the steering gear/ tyres, most road bumps are not transmitted to the steering wheel in SbW systems. Some OEMs regard this as acceptable or even desirable, but for others it is important that road conditions can be felt clearly through the steering wheel. A function has also been developed for this that can reproduce steering wheel torque proportional to road conditions.
This function is based mainly on estimated steering rack force and measured wheel speed. The quality of rack force estimation is of key importance for road feedback. It depends largely on the mechanical friction in the steering gear. Forces within the friction band of the steering gear cannot be estimated. This problem has been known since EPS were first introduced and is regarded as acceptable. Suitable filtering of the estimated rack force allows the various frequencies of road bumps to be transmitted to the driver or suppressed. The advantage over conventional EPS steering systems is that the amplitude of the road feedback can be varied almost at will. This enables OEMs to adapt the intensity of road feedback to the selected driving mode. For example, in comfort mode, road bumps can be suppressed, while in other driving modes, road feedback is transmitted more intensely to the steering wheel.
VIRTUAL STEERING RATIO
The steering ratio of SbW systems is entirely virtual. Its limits are theoretically set only by the tuning of the vehicle response functions and also by the vehicle chassis. As with conventional steering systems, the virtual steering ratio can be varied via the steering rack position. There are no mechanical limits and it is possible to simulate far more distinctive rack profiles. The steering ratio can also be adapted in line with vehicle speed to significantly enhance the manoeuvrability and controllability of the vehicle. This previously required a complex system of superimposing gear units; the virtual solution is considerably easier to implement. Tests have shown that the virtual steering ratio has no negative impact on vehicle response during accelerating or braking during cornering. The driver will barely notice the ratio change as long as it is kept within certain limits and adapted to the behaviour of the vehicle.
For the steering feel, the virtual steering ratio also requires a variable end stop. The challenge here is to design the high end stop torque with a high resolution and make it variable over numerous turns of the steering wheel without the need for an oversized feedback actuator. For a clearly defined end stop, a static steering torque of over 40 Nm should be reached, (3). To realise this, Thyssenkrupp collected a wide range of requirement profiles and examined various mechanical and mechatronic approaches.
The virtual steering ratio should not be seen only as an amplification factor; it can also be used to apply a steering angle offset. This is particularly useful for steering stabilisation, which can be achieved not only through Electronic Stability Control (ESP) but also through the superimposition of an additional steering angle. Thyssenkrupp has tested and implemented different models of such a steering stability control.
The steering stability controller compares the vehicle yaw rate with model figures and intervenes in the event of deviations. This is done in only fractions of a second and can trigger a highly aggressive vehicle response. As an alternative, Thyssenkrupp has therefore also developed a steering stability assistant. This differs in its control structure in that it does not correct the static yaw rate error. The driver at the steering wheel is still expected to respond but the assistant provides most of the stabilisation via a superimposed steering angle.
The steering stability controller and assistant are active safety systems that can intervene almost unnoticed to enhance the stability of SbW vehicles; for example, in the event of over-steering μ-split braking or side wind and tyre pressure compensation. In all these situations even an experienced driver will react more slowly than the controller, which stabilises more quickly with less steering angle and less over-shoot. This is because of the significantly lower response time (latency) as shown in the “human versus controller” comparison in skid pad tests, (4). In autonomous driving situations, the steering stability controller can improve steering behaviour and vehicle response.
It is even conceivable to follow a trajectory based on the target yaw rate. The steering stability controller can, however, be disadvantageous when the driver no longer feels a connection with the wheels. This can be restored with the right steering feel information from the feedback actuator. Unlike with mechanical superimposed steering systems the lack of a mechanical link in SbW solutions eliminates troublesome torque imbalance in the steering feel, as with current active steering systems the dynamic torques are transmitted to the steering wheel.
SUMMARY & OUTLOOK
Today, Thyssenkrupp is able to develop and analyse various SbW concepts taking into account the parameters of functional safety, actuators, sensors, control and steering functionalities. The aim of the development approach is to highlight the great innovation potential of SbW solutions and to improve the driving experience compared with today’s steering systems. In addition, the company is developing tailored solutions for SbW concepts at component and system level and addressing their functional integration into the vehicle as a whole. SbW systems are now being developed to production readiness in collaboration with OEMs.
Alongside the various concepts for redundancy and functional safety, further development work is focusing on the various ways of creating the required steering feel and on improvements to driver assistance and stability functions. System design and component selection are likewise being optimised to reduce costs. Other forms of human-machine interfaces are also currently being researched as an alternative to the steering wheel as well as vehicles without any sort of steering wheel for highly and fully automated driving.
If SbW concepts are seen not just as a new type of steering system but in the context of the overall car system, with numerous technical and functional synergies, the advantages and possibilities of SbW are enormous. It is highly probable that as automated and autonomous driving becomes more widespread, SbW systems will become established in the future. The timeline for this depends on many factors. But based on current estimates, this technological transition could take on a concrete form in five to 10 years.
 Polmans, K.; Stracke, S: Torque vectoring as redundant steering for automated driving or steer-by-wire. 5th International Munich Chassis Symposium, Munich, 2014
DIPL-ING KRISTOF POLMANS, MSC MULT is Head of Technology & Innovation at Thyssenkrupp Steering in Eschen (Liechtenstein).
DIPL-ING ANDREAS MITTERRUTZNER is Development Engineer for Steer-by-Wire at Thyssenkrupp Steering in Eschen (Liechtenstein).
MARCO DÄHLER, M. ENG. is Development Engineer for Steer-by-Wire at Thyssenkrupp Steering in Eschen (Liechtenstein).
DIPL-ING YANNICK THOMA is Project Leader Innovation Management at Thyssenkrupp Steering in Eschen (Liechtenstein).