Numerical-experimental comparison of the Levit paddle shift dynamic behaviour

This project addresses the vibration transmissibility of the leverage system of the paddle shift device at the steering wheel. The device under study is called Levit and is generally composed by a main body and by two pads which are the main subject of the analysis. The leverage system activation is influenced by the presence of magnets located on the main case which introduce some non-linearities. The main goal of this work is to develop and to tune a linearised dynamics analysis on the paddle shift with experimental tests. The behaviour of the device is experimentally tested in Valeo laboratory, then the data are post processed with Siemens LMS Test.Lab17 and Matlab software. The experimental results are compared with the ones provided through the numerical model. This last one is already created and will be improved with this work. The numerical model and the relative data are elaborated in Lupos, a parametric FEM code developed in Matlab. Two different experiments are conducted on the Levit in order to detect its natural frequencies and to identify the corresponding mode shapes. These two analyses consist in a sine sweep vibration and a roving hammer test. Furthermore, another experiment is carried out, focusing on the actuation and release of the pads to analyse more in details the reaction to excitation as in vehicle arrangement. The sweep vibration test is applied along all the three axes of orientation of the Levit. For each axis three different displacements (0.005, 0.010 and 0.012 mm) are set to evaluate the natural frequencies in several excitation conditions. It is worth noting that the results from y axis at 0.005 mm are the more interesting. Moreover, it is chosen to refer to y axis since it has the same direction of the steering column, thus it becomes the most excited axis when the device is correctly assembled on the vehicle. The second experiment conducted to study the linearized dynamic behaviour involves the usage of an impact hammer. The Levit pads have been hammered few times in perpendicular way with respect to the paddle surface. The smaller amount of data recovered from this experiment allows it to be easier used in the numerical model. Then, few parametric analyses are set in order to develop and improve the numerical model. In fact, the purpose is to bring numerical results closer to the experimental ones modifying one model parameter at time. Thanks to these updating actions, very good results are achieved in terms of natural frequencies and similitude of mode shapes between experimental and numerical fields. The last experiment has the objective to simulate the Levit behaviour in the vehicle layout. It consists in many repetitive actuations and releases of one pad at time; both pads are tested. Excellent repetition of results is observed, due to the instrumentation used and the setting of automatic cycle. From the observed performance it is visible that during the release phase there are oscillations on both pads, the one which is excited and the other one that receives its dynamics. This is also influenced by the nonlinear behaviour proper of the magnets inside the device, which will be investigated in a future work.