|
Studies have shown that road accidents are caused mostly by rollover increasing the number of total deaths due to traffic accidents both in Europe and in Turkey. Therefore, rollover occurences must be decreased via improved vehicle dynamics. This is one of the most significant active safety systems that should be designed carefully to improve vehicle dynamics.
In this study, by taking a mid-size bus as a case study on heavy vehicles, a virtual model was created using Adams/Car software program. In order to verify the virtual model, a standard tilt table test was applied to the physical vehicle. As the physical tilt table test can not be continued until vehicle rollover actually occurs, tilt table test was performed with verified vehicle model using Adams/Car software. To perform sensitivity analysis studies the roll-over dynamics of the vehicle, three different front anti-roll bars and two different front leaf springs were used during the tests. In conclusion, according to the results of the analysis, the best front leaf spring and front anti roll-bar are selected increasing the safety margin from the roll angle threshold.
Five different geometrical design dimensions have been selected as optimization parameters in order to increase suspension roll center height and suspension roll stiffness using Adams/Car software. In the simulation, the roll angle that is measured when vehicle pass over the bump has been selected as the system output. Statistical Taguchi metod was used to as the main optimization metod. In the result of optimization of the suspension, geometric dimensions found to be optimal are presented in the thesis.
In order to obtain a dynamical model for roll-over behavior during manuvers, a four-wheel vehicle model for lateral dynamics and Pacejka mathematical model for road-wheel friction have been applied. Furthermore, mathematical models for yaw and roll dynamics have been formed and the full vehicle model is completed by combining sub-systems in Matlab/Simulink programme. The signals of the yaw,roll angle and roll rate have been selected as outputs of the simulation in Simulink model.
Because the steering wheel angle is the system input of the Simulink model, the steering wheel angle and vehicle speed have been set to such sequence of values so as to cause to the vehicle rollover on purpose. Such an established maneuver having dangerous dynamics excited is known as J-Turn maneuver and it was selected for simulation on the vehicle model. Afterthat, the controller design has been performed in order to reduce rollover threshold and system was transformed in to the state-space matrix format. Dynamic load transfer ratio was selected as the system output and linear quadratic regulator and pole placement controller design are applied in order to take control of the vehicle.
In these simulations, the rollover threshold and the dynamic load transfer ratio have been selected as the system outputs.
In the simulations without the controllers, a dynamic load transfer ratio of larger than one is observed duringthe J-turn maneuver. However, it is found that this value can be decreased as low as "0.85" if the designed controllers are included in the loop. |
