Jean-Marc Duffal

Non-linear dynamics of a whole vehicle finite element model using a harmonic balance method

The aim of this paper is to apply the Harmonic Balance Method (HBM) to a finite element model of a complete vehicle (body, engine and engine mounts) to calculate the non-linear response of the assembly. The non-linear effects come from the amplitude-dependent stiffness of the engine mounts. First, the HBM is presented. A condensation process on the non-linear degrees-of-freedom is also proposed. This process reduces the original non-linear system by focusing only on the solution of the non-linear equations associated with the system’s non-linear components. Second, the engine mount stiffness dependency with amplitude is measured on a test bench to estimate a polynomial stiffness law. Finally, the numerical analysis is performed to analyse the non-linear response of the whole vehicle using the HBM algorithm with appropriate condensation located only on the non-linear coordinates of the system to minimise computer time.

H ∞ multi-objective and multi-model MIMO control design for broadband noise attenuation in an enclosure

The Active Noise Control (ANC) problem considered throughout this paper consists of attenuating the noise in one point of an enclosure, over a broadband frequency range, using a feedback control scheme. This paper proposes a general framework and a benchmark which allow a fair and quantitative comparison of achievable performances according to the number of actuator(s) and sensor(s) used for the control. At first, the experimental set up is described, then the identification of a low order MIMO acoustic model is presented. A multi-objective and multi-model control approach, is then proposed. The multi-objective synthesis allows to cope with design trade-off without conservatism, while the multimodel design enhances the robustness of the control. Finally some simulation and experimentation results are given which illustrate how the proposed methodology makes possible to compare achievable performances depending on the number of sensors and actuators used.

Robust stability analysis of brake squeal based on a parametric finite element model

Brake squeal is an instability phenomenon, which is severely dependent on many parameters. This study attempts to assess the effects of their variability on brake squeal behavior through FE computation. A detailed FE model of a commercial brake corner has been built up in order to predict its nominal squeal behavior. This analysis includes a non-linear preloading step to predict the system working-point and a complex eigenvalue analysis to assess its stability. A parametric study has been conducted in order to estimate the dependency with respect to the friction coefficient. The FE model has been parameterized to investigate the effect of variability. The process includes geometry simplifications to reduce CPU time, allowing far more configurations to be computed. Several parametric studies have been conducted to assess the effects of the friction coefficient, of the rotating direction, of the friction induced damping and of the hydraulic pressure. A numerical matrix test has been undertaken to synthesize the brake behavior in the wide variety of conditions it may encounter. Then, a full factorial design of experiments has been conducted with respect to the friction coefficient and the disc Young Modulus. This analysis shows biparametric coupling patterns and stability charts. Finally, it is possible to rank the parameters with respect to their influence and to assess the performance and the robustness of the system.

Parameter analysis of brake squeal using finite element method

Brake Squeal is a friction induced instability phenomenon known to be one of the most annoying noise for drivers. This paper focuses on the mode coupling aspect of brake squeal by means of a multi parametric analysis. The study is based on a Finite Element model of the whole brake corner. A complex eigenvalue analysis is undertaken, with a modal projection technique, to detect the stable and unstable modes. Following this process, the brake stability is assessed as a function of the friction coefficient. The results highlight accurately the modecoupling phenomenon also referred to as coalescence. Then, the emphasis is put on the disc Young modulus variability by launching a numerical design of experiment. Finally, the brake robustness is displayed as functions of the friction coefficient and of the disc Young modulus.