The elastic properties of a friction material play a significant role in the vibro-acoustic noise, vibration and harshness behaviour of braking systems on a structural scale. Common testing, such as pad compressibility testing and modal testing, allows identification of the global static and dynamic elastic properties respectively, which generally provide data for finite element numerical models. The present investigation presents static and dynamic testing methodology and focuses on the distribution of the elastic properties of friction materials on a pad scale. Besides trying to determine a correlation with bench testing, the added value in updating friction material models with distributed elastic properties for squeal simulation models is evaluated. For investigation of the distribution of the friction material properties of brake pads, quasi-square friction samples of friction materials were collected from several locations of the pad. The experimental meshing provides data on the inner radius and the outer radius of the brake pads as well as on the inlet and the outlet of the pad. For the purpose of this study, six different materials were tested. In addition to four known prototype friction material formulations obtained using the same production process, two commercial friction materials were investigated. The static stiffness is evaluated with a laboratory compression machine set-up whereas the dynamic stiffness is identified from laboratory ultrasonic investigations. The results and coherence of the static and dynamic stiffness distributions obtained by both approaches are compared. The trends with the noise, vibration and harshness bench testing results are also discussed for selected pads. From ultrasonic testing, it is found that the dynamic stiffness distribution can take several forms (flat, concave or convex) which can be different for the in-plane stiffness and the normal stiffness of the pad. It is mainly noticeable along the tangential direction of the pad, reaching a variation of up to 16% between middle of the pad and the edges. The stiffness can also vary by up to 8% along the radial direction of the pad. The measurements performed during this study confirm also that these stiffness distribution patterns are strongly dependent on the formulation friction of the material and on the production process itself. The above-mentioned stiffness distributions were implemented in finite element brake pad models to estimate the potential frequency shift arising from these considerations. The impact of taking into account such information was shown to be limited for the free–free pad modal properties. Nevertheless, the influence of such distributions on the complex eigenvalue analysis of complete finite element brake models was observed and is discussed. Even if the used pad models take into account to some extent the static and dynamic stiffness distributions for complex eigenvalue analysis, they still do not agree with the non-linear material models regarding the stress and the frequency, as shown experimentally. Even if an additional effort is made to determine these stiffness distributions which provide more detailed information for the brake pad finite element models, this physical degree of freedom could drive, to a certain extent, the noise, vibration and harshness responses of the brake systems. A more mature understanding of the influences of the formulations and the process on these distributions and their control at the pad design stage could add a complementary strategy for reduction in the noise, vibration and harshness issues in brake systems.
Read full abstract