Vehicle Dynamics: Passenger Comfort Evaluation

The thesis project originates from the need to evaluate the working principles and to validate an existing numerical model, developed in a multibody dynamics environment, which can simulate the ride comfort as perceived by customers on board of passenger cars and possibly commercial vehicles.?? Such model represents the interaction of the vehicle-seat-passenger system and has been developed by CRF. This model aims to incorporate in vehicle simulation the characteristics of seats (viscoelastic characteristic that varies with different models) and human response (nonlinear, depending on a high number of variables).?? The final goal of the activity performed by the candidate is then summarized as: •??Learn the definition of the numerical model and how it is implemented in the multibody analysis software •??Verify that input parameters can be effectively tuned in order to correlate the simulation result with the real phenomenon •??Correlate and validate the model on a series of experimental measurements, taken on different test tracks. The whole activity is much more articulated. In the first period, the activity was done in collaboration with the vehicle testing department, analyzing experimental data of three vehicles recorded on test tracks specialized for comfort. The three vehicles belongs to the C-SUV segment and are equipped with Diesel engines (4 in-line cylinders, 1.6 to 2.0 l displacement), manual transmission, curb weight within 1400 and 1600 kg. Test tracks are located in the Balocco Proving Ground and are designed to excite all the vehicle natural frequencies that are relevant to ride comfort, in the range 0.5-25 Hz; the available data was recorded on the tracks: •??Cleat: an obstacle on the road such as a small rail, acting on left wheels only (asymmetric) or on both sides (symmetric), representing an impulse input signal. •??Highway: a smooth tarmac road with a random, low amplitude, steady state force input •??Uneven road: a road with non-smooth or damaged tarmac, with high input amplitude •??Cobblestone: random input force at high frequency All the analyses and calculation of relevant indicator parameters was done on VDArco, MATLAB and Microsoft Excel. Signals acquired were both for standard ride comfort evaluation and additional ones for the evaluation of seat comfort. The second part of the activity was dedicated to the correlation of a full vehicle multibody model with the experimental data of one of the three vehicles in the dataset, done on MB-SHARC software a custom version of ADAMS/Car by MSC. Model correlation is the basis of simulation, since a good model will predict accurately any possible effect of modifications to be applied to the vehicle. For comfort model, identification of parameters starts from wheels, then suspensions and finally vehicle body inertial properties. The last part concerns the study, application and improvement of the human-seat interface model, already implemented in MB-SHARC. The modifications introduced, permitted a good correlation on one mission (Uneven road) and promising results on the others, proving that now the model seems to be almost ready for use by the Virtual Validation team of FCA. The thesis work of the candidate will be continued by a fellow master thesis student with the aim of a complete validation of the model, on all tracks and on rear seats.