Preliminary Brake pad viscoelastic properties mediate friction force and contribute to brake squeal. Elastic constants/engineering constants are often used as inputs to noise/vibration models (C.A.E. models) developed to help the investigation of the noise/vibration aspects of the braking action. C.A.E. model is built using the matrix of engineering constants made by elastic moduli (E), shear moduli (G) and Poisson’s ratio (V). Research and /or Engineering Questions/Objective The objective of this activity is to understand how these engineering constants data given as inputs for the C.A.E. model influence the outputs: namely the Eigenfrequencies spectra of brake pad. Methodology In a first step, the material characterization of the reference brake pad is done using the ETEK 3000 instrument (IMS) : from 2 samples cut in the friction material part of the pad, it calculates the complete matrix of engineering constants. Then, the C.A.E model is built and to check its validity, FRF measurements are performed. In a second time, each independent variable is modified following a DOE (Design Of Experiments) and its influence on the NVH simulation is analyzed looking at the Eigenfrequencies values. From the entire matrix of engineering constants, considering the transversely isotropy (xy plan) of the brake pad, from 9 properties only 5 are independent: Ex ; Ez ; Gxz ; νxz and νxy ; density is added as fundamental data for C.A.E. construction model. Results The first studied material is an ECE type. Firstly, the DOE is about Ex ; Gxz ; νxy values . Using Minitab software, the DOE is made by 12 runs corresponding to modified ETEK values. Only 2 present C.A.E. simulations results satisfying (they have all frequencies in [-3% ; +3%] range of reference frequencies). Looking in details at how each Eigenfrequency varies with engineering constants, it appears that in the range [X-45%X ; X ; X+45%X] (with X=measured modulus of the reference material), Gxz and νxy have minor influence, while for Ex a variation in intervals [0.82Ex ; 0.88Ex] Ս [1.13Ex ; 1.16Ex] leads to the first frequency shift of between 3% and 5% of the reference value ; and for a variation in intervals [0.45Ex ; 0.82Ex] Ս [1.16Ex ; 1.55Ex] the first frequency shift is beyond 5% of the reference value, considered as relevant difference for the NVH pad behavior. A second DOE is made considering Ez, νxz and density (a variation of 5% for the latter is arbitrary chosen): no influence of these values in the considered range [-45% ; +45%] on the Eigenfrequencies spectra is measured. Limitations of this study It is important to understand if these results are specific of this material/format or can be generalized. Thus, other samples have been analyzed using the same DOE approach with Ex ; Gxz ; νxy values with different pad shapes, presence or not of UL and different materials. These analysis are still in progress. What does the paper offer that is new in the field in comparison to other works of the author? So far, no study investigated the influence of engineering constants in C.A.E. models, and it is still difficult to judge the quality of ETEK outputs for a material: which difference have to be considered as relevant and meaningful of a material modification and which instead could be simply due to aleatory difference form pad to pad, and have no impact on friction behaviour. Conclusions/next steps At this point of the study the main result obtained is the major influence Ex parameter has on Eigenfrequencies, while the other engineering constants have limited impact on them. To figure out how this statement could help tailoring the pad properties, we should understand how formulation and process affect engineering constants. To proceed, a material will be treated with different thermal treatments to have an idea of how process influences viscoelastic values; and in the same way from a reference material some modifications of formulation will be done to study the influence of formulation on viscoelastic values.

Dr. Eng. Florence Vivier, R&D Engineer, ITT Motion Technologies; Ing. Stefano Barale, Project Manager, ITT Italia, S.r.l; Ing. Filippo Gedda, Systems Engineer, ITT Italia, S.r.l; Dr. Diego Pellerej, R&D Laboratory Manager, ITT Italia, S.r.l; Ing. Andrea Della Rovere, R&D Test Tech Supervisor, ITT Italia, S.r.l