The main functions of suspension system are to provide ride comfort for the passengers and vehicle handling (road holding). But, in many studies, full attention to the ride comfort leads to the determination of incorrect suspension system parameters as well as other problems such as rollover and reducing road-holding ability in the vehicle. The aim of this study is to present a method for the optimized design of the vehicle suspension system in order to improve the ride comfort, road holding, workspace and preventing rollover, considering a full vehicle model with 11-DOF. The most important feature of this study is that the prevention of rollover factor and all of suspension functions are considered simultaneously. In this research, in order to assess the ride comfort, the vertical acceleration values of seats that are caused by random road roughness are calculated by power spectral density of road in frequency domain. In the context of prevention of rollover, Fishhook manoeuvre is performed using a mathematical model for the roll motion, and then the dynamic behaviour of the variables is considered in rollover threshold. Then, the optimization problem is solved to minimize the vertical acceleration values and vehicle roll angle by considering the physical limitation and safety of the model. The results of the optimization show that the vertical acceleration, in frequency domain at the desired boundary values (as defined in ISO 2631), decreases and rollover resistance of the vehicle increases.

In this study, a new method to improve ride comfort, vehicle handling, and workspace was presented in multi-objective optimization using nonlinear asymmetrical dampers. The main aim of this research was to provide suitable passive suspension based on more efficiency and the low cost of the mentioned dampers. Using the model with five degrees of freedom, suspension system parameters were optimized under sinusoidal road excitation. The main functions of the suspension system were chosen as objective functions. In order to better illustrate the impact of each objective functions on the suspension parameters, at first two-objective and finally five-objective were considered in the optimization problem. The obtained results indicated that the optimized viscous coefficients for five-objective optimization lead to 3.58% increase in ride comfort, 0.74% in vehicle handling ability, and 2.20% in workspace changes for the average of forward and rear suspension.

Ignoring the possible impacts of uncertainties in vehicle components during the design phase can undermine the safety of passengers and the vehicle performance. The main function of a suspension system is to provide satisfactory ride comfort and road-holding with a sufficiently low probability of rollover. Despite many studies on the design of new suspension systems with inerters, the effect of uncertainties in vehicle weight and tire stiffness on the design of suspension with inerters has not received much attentions. This paper presents a new type of suspension with inerters and asymmetric dampers and investigates the dynamic behavior of a vehicle under variable vehicle speed. Moreover, the effect of uncertainties on the choice of acceptable values of inerters is evaluated. For this investigation, the authors developed a 9-DOF full vehicle model with roll and yaw motions under non-stationary random road excitations in the time and frequency domains and studied its dynamic response with different suspension models. The optimal design was performed using a multi-objective optimization algorithm called MOEA/D. The best model was then used to determine the effect of uncertainties on the choice of inerters. The optimization results show that using the optimized suspension with inerters and nonlinear dampers instead of conventional design improves the ride comfort by 0.16%, the vehicle road-holding by 3.54%, and the rollover probability by 44.73%. In the proposed model, by changing the values of vehicle parameters with uncertainty, the choice of inerters to have an acceptable performance would be variable.