Search

The current work aims at bridging the gap between the current vehicle handling characteristics and the future demands of higher vehicle handling performance, required to guarantee higher safety and facilitate the application of autonomous driving, platooning and automated highways systems. For this task a state 'of the art vehicle chassis control system known as "Wheel Speed Distribution Control" (WSDC) has been proposed. WSDC in principle relies on controlling the vehicle driven wheel speeds to enforce better vehicle handling performance. The WSDC system capacity has been investigated using numerical simulation. Tberefore, an innovative vehicle handling simulation model has been developed from first principles. It employs the Magic Formula (MF) tyre model for combined slip, has 23 degrees of freedom and includes more than 60 vehicle handling parameters. The vehicle handling model has been developed using the novel Cartesian Geometric Translation (CGf) technique which employs geometry, trigonometry, Cartesian coordinates and finite difference approximation in the time domain to facilitate development of high speed models. The model has been built using the BASIC'O programming code in the DOS'O environment and optimised to meet the novel Model Predictive Control (MPC) based feedforward WSDC yaw rate controller requirements, such as small code size (less than 35 kb) and processing speed faster than real time. The simulation results validated the WSDC principles as it showed the capacity of WSDC to enforce the desired yaw rates, with acceptable driven wheel longitudinal forces. To put WSDC into practice,an original hardware" Wheel Speed Distribution Differential" (WSDD) design has been developed and optimised for lower speed, torque, power,production and maintenance requirements.It has the capacity to precisely differentiate the driven wheels speed under the influence of a DC motor with relatively small power requirements. It has linear speed and torque characteristics which facilitate its control. It also has been developed to allow many beneficial differential modes. The simulation results of the whole WSDC system have clearly demonstrated that it can in fact achieve its development target of feasibly enhancing vehicle handling performance.

This thesis investigates the principle of integrated vehicle dynamics control through proposing a new control configuration to coordinate active steering subsystems and dynamic stability control (DSC) subsystems. The active steering subsystems include Active Front Steering (AFS) and Active Rear Steering (ARS); the dynamic stability control subsystems include driveline based, brake based and driveline plus brake based DSC subsystems. A nonlinear vehicle handling model is developed for this study, incorporating the load transfer effects and nonlinear tyre characteristics. This model consists of 8 degrees of freedom that include longitudinal, lateral and yaw motions of the vehicle and body roll motion relative to the chassis about the roll axis as well as the rotational dynamics of four wheels. The lateral vehicle dynamics are analysed for the entire handling region and two distinct control objectives are defined, i.e. steerability and stability which correspond to yaw rate tracking and sideslip motion bounding, respectively. Active steering subsystem controllers and dynamic stability subsystem controller are designed by using the Sliding Mode Control (SMC) technique and phase-plane method, respectively. The former is used as the steerability controller to track the reference yaw rate and the latter serves as the stability controller to bound the sideslip motion of the vehicle. Both stand-alone controllers are evaluated over a range of different handling regimes. The stand-alone steerability controllers are found to be very effective in improving vehicle steering response up to the handling limit and the stand-alone stability controller is found to be capable of performing the task of maintaining vehicle stability at the operating points where the active steering subsystems cannot. Based on the two independently developed stand-alone controllers, a novel rule based integration scheme for AFS and driveline plus brake based DSC is proposed to optimise the overall vehicle performance by minimising interactions between the two subsystems and extending functionalities of individual subsystems. The proposed integrated control system is assessed by comparing it to corresponding combined control. Through the simulation work conducted under critical driving conditions, the proposed integrated control system is found to lead to a trade-off between stability and limit steerability, improved vehicle stability and reduced influence on the longitudinal vehicle dynamics.

In this thesis an integrated powertrain control for gearshifts on twin clutch transmissions is developed. First, a detailed model of an automotive powertrain featuring a twin clutch transmission is developed in Matlab/Simulink®. This model includes detailed friction models for the twin clutch that enable an investigation into the effects of different friction materials on the performance of the gearshift controller. The transmission model also includes detailed models of the synchronisers and thus allows a simulation of synchroniser-to-synchroniser shifts. A simplified phenomenological model, derived from a more complex non-linear model, is employed to model the hydraulic actuation of clutches and synchroniser. The thesis finds that the dependency of the friction coefficient on the sliding speed has an important influence on the gearshift quality and the performance of gearshift controller, while the absolute level of the friction coefficient is less important. Based on this powertrain model the key problems of gearshifts on twin clutch transmissions were identified and a control that overcomes these problems was developed. The first stage was to devise a gearshift control algorithm that handles single clutch-to-clutch shifts without a oneway (freewheeler-, overrunning-) clutch. This basic gearshift control algorithm featured a control of clutch slip for the engine torque transfer and a control of engine speed through engine torque manipulation (plus clutch pressure manipulation for downshifts). In a second stage, an optional transmission output torque control was developed that could be integrated in the basic control. The thesis shows that these control strategies are superior, in terms of shift quality, to conventional gearshift controls as used on planetary-type transmissions and are also robust against variations in the powertrain parameters (including friction coefficient) and sensor noise. The control strategies developed for single clutch-to-clutch shifts were extended to handle double and other multiple gearshifts that take place in the same transmission half. The thesis also investigates the other main part of gearshifts on twin clutch transmissions, the gear pre-selection. The thesis shows that, on power-on gearshifts, the torque reactions at the transmission output due to the gear pre-selection with conventional hydraulically actuated synchronisers can be effectively compensated for by a simple manipulation of engine torque.

This thesis investigates the principle of co-ordination of chassis subsystems by proposing a new control structure for co-ordinating active steering technologies and a brake based directional stability controller. A non-linear vehicle handling model was developed for this study using the Mattab and Simulink tools. This consists of a 4 degree of freedom (d. o. f) lumped-parameter model that includes longitudinal, lateral, yaw and roll motions with quasi-static longitudinal load transfer effects including non-linear suspension and tyre descriptions. The non-linear vehicle dynamics are discussed for the whole operating regime and two specific driving tasks are identified, steerability and stability. In the context of the vehicle states these are yaw rate control and side slip angle bounding respectively. Linear active steering controllers for front, rear and four wheel steering are designed and evaluated in the context of the vehicle handling problem throughout the non-linear operating regime to assist the driver in the two driving tasks previously defined. It is shown through the analysis of the vehicle dynamics in the Chapter 3 that linear controllers can be used to significantly improve the handling behaviour of a non-linear vehicle when only one active input is considered, however when controlling two active inputs, non-linear multivariable approach is required to deal with the strongly coupled nature of the vehicle handling with respect to front and rear steering inputs. A brake based stability system that reflects the state of the art is implemented. The work then proposes a novel co-ordination controller structure for coordination of an active steering controllers and a brake based stability controller for improving to vehicle handling control. The controller was assessed both in steady state and transient tests selected to simulate real world driving manoeuvres over the whole non-linear vehicle handling regime. The co-ordination controller is found to lead to a trade-off between stability and limit cornering performance. The proposed structure improves vehicle stability and reduces interactions in the longitudinal vehicle motion. A detailed discussion of the implications of a coordinated control approach showing it to be a powerful tool providing, the interactions can be conveniently related vehicle handling task and that an appropriate measure of vehicle performance is available. The limitations of the approach are discussed. The most significant limitations being a) the difficulty in proving the optimalty of a heuristic control structure, b) the difficult in assessing the controller behaviour and its interaction with a real driver and c) the likely complexity of the rule base for coordinating more than 2 or 3 systems or describing more complex interactions than were observed here.

This work focuses on the potential benefits that can be gained from the use of Individual Wheel Control on a large off-road vehicle. The vehicle concerned is a theoretical, six-wheel drive, off-road Hybrid-Electric Vehicle based on an existing conventional six-wheel-drive Combat Support Vehicle developed by QinetiQ (formerly DERA). The proposed vehicle utilises Individual Wheel Control through the use of six, in-wheel, Hub Mounted Electric Drives. A novel intelligent mobility control system is developed to fully exploit the capability that this configuration offers. Initially, simplified vehicle models are developed to design and test the mobility control components. The control systems developed are Traction Control, Anti-lock Braking and Direct Yaw-moment Control. These controllers are developed individually, with the aim of improving vehicle stability and handling behaviour. Once tested, they are combined into a single system on a basic non-linear handling model, where the controller co-ordination scheme is demonstrated. Preliminary testing shows the full controller to reduce driver workload by offering predictable vehicle handling and improved vehicle stability. An eighteen-degree of freedom vehicle model is then developed, incorporating the vehicle suspension and load transfer characteristics, based on the conventional vehicle. Field test data taken from the existing vehicle trials is used to partially validate the on-road handling behaviour of the vehicle model. On this model, the full mobility controller is tuned to offer optimal performance for a large range of driving conditions and an extension of the controller to limit side-slip at high lateral accelerations is introduced. The controller is then tested on and off-road against a fixed torque distribution system and also the conventional vehicle equipped with various differentials. By exploiting the high torque capability and controllability of the electric drive, the potential of Individual Wheel Control is demonstrated, along with the benefits offered by the hybridelectric drivetrain with respect to mobility. Through the simulation work conducted, the major benefits of Individual Wheel Control are shown to be: improved stability and manoeuvrability; more predictable vehicle behaviour leading to reduced driver workload; accurate yaw rate tracking and increased safety at handling limits.

This thesis presents the results of a research project which sought to find links between driver subjective ratings and objective measures of vehicle handling. The experimental data used in this project has been made available from a previous research project. The experimental data was collected using a prototype vehicle which was used in 16 different configurations. Objective data was collected based around the ISO defined steady state, step input, and frequency response tests. Subjective assessments were collected from eight trained test drivers using a numerical rating scale to a questionnaire covering various aspects of vehicle handling. Analysis of the subjective assessments has been done to identify any shortcomings that may affect any subsequent analysis. From the literature review, an approach that claims to relate four simple objective metrics to subjective measures of vehicle handling has been developed in two new ways. Firstly, the proposal was tested [1] with the large amount of subjective data available to see if good levels of correlation could be found between the proposed metrics and driver subjective ratings to specific handling questions. Secondly, the method was extended to include further simple metrics to try and improve links between the subjective and objective data [2]. Non-linear relationships in the correlation of subjective vs. objective data have been investigated for the first time [3] using non-linear genetic algorithms, which, in addition have not previously been used to correlate driver subjective ratings with objective measures that describe vehicle handling. From the results, it has been possible to specify ranges of preferred values of objective metrics in order to produce a subjectively satisfying vehicle. Finally, the work discusses how the results obtained can be used by engineers to aid the vehicle design and development process.

Racing teams use numerous computational tools (CAD, FEA, CFD) to aid in the design of racing cars and the development of their performance. Computer simulation of racing car handling through Lap Time Simulation (LTS) packages complements these tools. It also allows teams to examine the effect of different vehicle parameter setups to optimise vehicle performance. In similarity with the automotive industry, time is limited and rapid development of new ideas and technology is essential. Thus, the use of a more sophisticated computer simulation would allow a team to gain a significant advantage over their competitors. As LTS are computationally intensive,previous packages have simulateda full lap using a quasi-static method which splits the path of the vehicle into segments. An analysis is then made of the vehicle at each segment point using the external forces acting on the vehicle. Due to the constant acceleration(i.e. steady state) assumption across each segment, this method does not take into account the effect of roll, pitch and yaw inertia as well as damping and tyre lag effects. Another aspect that is not accounted for is the variation in the fastest effective vehicle path along the circuit (i.e.racing line) due to change in driver control inputs or vehicle parameters. The overall aim of this thesis is to develop a transient LTS methodology, which adopts a strategy to vary the racing line taken in order to address the problems found with the existing quasi-static LTS packages. In parallel an investigation of the accuracy of vehicle models in relationship to racing car performance has been developed. The thesis begins with a study of racing car modelling techniques and a review of current LTS packages. A description is then given of the collection of vehicle handlingd ataf rom an actualr acingc ar (alongw ith attaining a vehicle parametesr et) and the measured results displayed and discussed. The creation of two vehicle models, a simple and sophisticated version, is detailed and the measured results are compared to the simulated results of each vehicle model. It was found that the simple vehicle model does not fully represent the actual vehicle's lateral dynamic behaviour, although its steady state response was deemed to be accurate. The sophisticated vehicle model was seen to not only accurately predict the full range of lateral dynamic behaviour of the actual vehicle, but also the actual vehicle's longitudinal and combined lateral and longitudinal dynamic behaviour. To further investigate LTS techniques, a comparison study was made between various simulation approaches which indicated that the transient approach, although more complicated and time consuming, allows for more accurate tuning of a greater number of vehicle parameters. Finally, the creation of two simulation packages has been detailed and case studies are presented to provide further insight into the look and feel of the packages. The first package is a quasi-static approach based LTS package, where a case study is made into the sensitivity of overall lap time to a range of vehicle parameters. The second is a transient approach based simulation package which optimises the driver controls,varying the racing line taken by the vehicle and ensuring the manoeuvre is completed in the quickest time for that vehicle parameter set. This final Manoeuvre Time Minimisation package fulfils the overall aim of the thesis and a case study is made into the effect of front damping value on manoeuvre completion time.

This thesis presents results from a research project seeking to correlate subjective and objective measures of automobile handling. An underlying goal of the work was to demonstrate how a relatively simple lumped parameter model, suitable for effective use at the early stages of vehicle design, could be used to predict both the objective responses and subjective feel of the car. The work associated with the project was centred around sixteen configurations of a prototype saloon car. Objective evaluation included ISO defined steady state, step input, and frequency response testing. Subjective assessments were conducted by eight trained test drivers who supplied feedback in the form of numerical ratings on a questionnaire covering various aspects of handling. Examination of the two sets of data highlighted aspects of handling for which driver ratings correlated with objective data. It was also possible to quantify the average effect each objective response parameter had on driver ratings and thus to identify responses which most strongly influence subjective ratings. In addition a lumped parameter model allowing for lateral, yaw and roll degrees of freedom was validated against the experimental data. This validation demonstrated that the model was capable of accurate steady state and transient predictions both in the linear and non-linear range. The work concludes with a brief discussion about how the validated model, combined with the knowledge gained from the correlation work, could be used by engineers to streamline the design and development process.

One of the most important source of noise and vibrations associated with vehicles is the vibration of driveline systems. Such phenomena are subjectively associated with customer complaints. In this study the torsional vibrations of driveline systems were investigated using discretised and lumped mass models of the system. In the literature, many of the problems associated with torsional vibrations and refinement in drivelines have been tackled through relatively simple, lumped mass models combinedw ith experimentalm easurements. However, some problems remain particularly where instabilities occur or complex coupling with other vehicle vibration modes exists. The review of previous work showed that although it is important to understand the dynamic behaviour of the individual driveline components; for example, engine, clutch, gearbox, etc., the whole system must be analysed together because of all the coupling which occurs. The main source of excitation for torsional vibration of the driveline system is the engine fluctuating torque. A computer program using MATLAB subroutines was developed to obtain this fluctuation torque for different engine parameters for subsequent use in the modelling. A substructure approach, using stiffness coupling technique with combined use of residual flexibility and modal synthesis, was used to analyse free and forced vibrations of the system, as a linear system. A computer program using MATLAB subroutines was designed to facilitate application of this technique. Good agreement between results for the overall system model and substructure model was found even for a few considered modes. This substructure technique offers significant computational advantages over other methods. The effect of non-linear sources in the driveline system such as backlash, non-linear spring stifffiess, Hooke's joint and angularity of the propeller shaft on the system torsional vibrations was investigated. The effect of backlash in the driveline system was significant and, as expected, vibration levels increased as backlash increased. Hooke's joints caused an additional complex source of excitation but their significance is dictated by the details of the particular driveline design. The modelling showed that instabilities commonly referred to a shunt or shuffle could occur during clutch engagement. The stick slip frictional properties of the clutch were crucial in this behaviour and the relative importance of various design features was quantified. A mathematical model including torsional motion of the driveline system and other vehicle body motions was developed to analyse the ways in which the driveline couples with other dynamic components. Two running conditions were considered; steady state running and transient during clutch engagement. It was shown that the complete system was capable of self-excited oscillations under certain conditions during normal running as well as the instability which could occur during clutch engagement. This comprehensive model represents an important contribution of the work in this area of research in two ways. First, it clarifies understanding of the dynamic coupling between the rotational and translational components of the whole vehicle system. Second, it provides design information to tackle instability problems and to lead to reductions in overall vibration levels.

The problem o f disc brake squeal has been examined by developing a finite element model of the coupled pad-disc system , conducting complex eigenvalue analysis and associating unstable modes with potential squeal problem areas. A key issue in this process is the representation of the contact pressure distribution at the frictional interface between the disc and the pad. Non-linear contact analysis using the finite element model of the pad revealed that contact is only partial at the pad-disc interface and that the contact pressure distribution depends on the friction coefficient, Young’s modulus of the friction material and the way the applied pressure is distributed on the pad backplate. A new method is proposed in which interface contact stiffness is related to brake line pressure using a statistical approach based on the measured surface properties of the interface. Complex eigenvalue analysis of the coupled pad-disc system has shown that unstable modes exist within different ranges of contact stiffness thereby providing an explanation of the effect of varying line pressure on squeal. The two most unstable modes from the analysis show good correlation with experimental squeal results. The coupled model is then used for parametric studies the results of which indicate that high coefficient of friction and uniform contact pressure distribution increase instability whilst a trailing edge biased pressure distribution and a high support stiffness at the pad backplate reduce it. Limiting the disc symmetry by introducing equispaced slots was shown to be effective in reducing instabilities involving diametral modes of the disc with the same order of symmetry only Other modes were stabilised by increasing the rigidity of the pad. The overall results suggest that either the pad or the disc can be mainly responsible for the instability depending on the mode thus unifying the different approaches to disc brake squeal and enabling the most appropriate component to be targeted for squeal abatement purposes.

First, techniques for modelling the vehicle and road surface are discussed, and the standard linear and nonlinear analysis methods are reviewed. Then, using the quarter car model and a single idealised road surface, a brief analysis of the passive and active suspensions, including full and limited state feedback schemes, is presented. The performance in terms of ride comfort, road holding ability and suspension travel for both systems is established, providing a yardstick against which the controllable damper systems can be compared. Three suspensions fitted with controllable dampers are then analysed. In order of increasing complexity these are; a three-state adaptive system, a two-state switcliable system, and a continuously variable system. After a performance comparison of the ideal system, the practical limitations present in real hardware are included in the damper model. Their effect on performance is quantified and realistic response targets are set. The model is then extended to a two dimensional “bicycle” model, which enables control laws to be generated which take into account the correlation between front and rear wheel inputs. Using these laws to drive the active and continuously variable damper system, the advantages of a correlated law are identified. The accelerations and suspension displacements of a passively suspended production vehicle are measured during actual runs over three roads of varying roughness. These are used to estimate the surface roughness properties of the roads. Using this range of typical conditions, the idea of system adaptation is then considered. The performance of each controllable damper system lies between that of the active and conventional passive systems. The continuously variable system generally offers the best ride comfort, but worthwhile improvements are also possible with the two-state switchable system. The adaptive system offers only marginal improvements for the conditions considered.

A successful model for off-road tyres must be reliable, efficient and capable of reproducing and predicting the main system phenomena. Mathematical models are proposed for longitudinal, lateral and combined lateral and longitudinal force generation characteristics of off-road tyres. For a better understanding of off-road tyre behaviour, the study of the interaction between the tyre forces and those generated by the deformed soil is very important. Details of the force system in the contact patch, therefore, are used as a basis for developing models for the prediction of the tractive performance and comering characteristics of off-road tyres. Previous work of the relevant literature pertaining to tyre behaviour is reviewed to provide the reader with background information on off-road tyre characteristics. Various models for off-road tyres of differing degrees of complexity but which all incorporate the key features of off-road tyre problems are then developed. Previous methods of analysing the combined lateral and longitudinal forces generated by off-road tyres on deformable surfaces are investigated. A modified version of a previous model is then proposed which is based on a different and original method for investigating tyre behaviour in the contact region. An entirely new model is then developed which is based on a modification of the "multi-spoke" tyre model used for road vehicle studies. Predicted results are compared with those obtained from other models and with reported experimental data. The usefulness of such models is in problems involving the steering, braking and handling behaviour of off-road vehicles. Hence, the models are formulated so that they can be applied to such vehicle problems by enabling lateral and longitudinal forces on the tyre to be predicted from any combined conditions of wheelslip, wheelskid and slip angle. The proposed models provide an improved qualitative description of behaviour in the contact region. Although slightly more complex than previous models, the computational load is nevertheless sufficiently small that the tyre models can conveniently be incorporated in off-road vehicle handling models. Suggestions for future recommendations are discussed with particular reference to improving the predictive models and for a possible extension of the study to generate more detailed practical results for tyre forces under controlled experimental conditions.

Various types of actively controlled suspension systems for automotive or agricultural applications are theoretically studied on the basis of the well known quarter car model subjected to realistic inputs chosen to represent road surface/forward vehicle speed combinations for a range of different conditions. The vehicle response is evaluated through the performance criteria (the ride comfort, dynamic tyre load and suspension working space) and power requirements (the power input to actuator, power dissipated in damper or actuator and power fluctuation in spring and tyre). The range of suspension systems includes fully active, semi-active, slow-active, single state feedback active, two state switchable damper, continuously variable damper and conventional passive. Computer programmes relating to the general dynamic modelling of ground vehicle suspension systems (generation of the artificial road, random process analysis and human response criteria) are designed. Computer programmes relating specifically to vehicle ride (linear or non linear), derivation of responses and power calculations for linear or non linear models, as well as performance criteria and optimal control of vehicle suspension are also designed. The switchable damper system which involves continuously switching between two discrete settings is of considerable interest because such dampers are currently available. It is shown to offer worthwhile improvements over passive systems in terms of ride performance if a simple control law is followed. Linear optimal control theory is used to obtain the optimal feedback gains of the fully active and slow-active suspension systems. The behaviour of the fully active linear control systems and the possibility of improving their performance by using a non-linear control law is investigated. The performances of the four and two state feedback slow-active systems, using an actuator with limited frequency response up to 3 Hz, are similar. In terms of the power demand, there is little difference between the fully active and slow-active systems, configured with a conventional passive spring in parallel with the actuator, and their ride performances are also similar. The behaviour of the semi-active systems are evaluated with a control law based exactly on the optimal control of the fully active system, except that no power input is available. A method of comparing performance and power requirements is based on the practical viewpoint that the suspension designer is essentially allocated a given amount of working space and must optimise the suspension within this constraint. Hence, all the competing systems are compared on the basis of equal workspace contours. Conclusions and suggestions for further work are discussed with particular reference to the relationship between the predictive models and their practical usefulness in assisting the designer of advanced suspension systems for on and off - road vehicles.

This work is concerned with a theoretical investigation into the ride behaviour of actively suspended vehicles. It is based on the concept of developing control laws for active suspension systems fitted to full vehicle models and then comparing the performances of competing systems using a fair basis for comparison. A review of published work was revealed the need for further work on active control laws for full vehicle models. Theoretical techniques necessary to generate random road profiles and various methods for evaluating vehicle ride response are explained. This is followed by an evaluation of the performance of the passive suspension systems. Techniques for obtaining control laws for an active suspension implemented at all four wheel stations of a vehicle are outlined. It is shown that the classical control law based on ignoring the correlation between the road inputs can be replaced by one which involves limited state feedback and includes the effect of the wheelbase time delay. Performance of this system is better than the system which ignores the time delay and almost as good as that of the full state feedback active system with time delay. In addition, it is much more attractive in terms of its practical implementation. The general strategy of keeping seat accelerations, dynamic tyre load under control and the fore/aft and lateral dynamic tyre load transfers is used, within the realistic design constraint of a specified amount of available working space. The effect of road surface representation, including the cross correlation and the measurement errors in deriving the control laws is examined. Finally the performances of all the active suspensions considered are compared with those obtained from passive systems and conclusions and recommendations for future work are made.