Chapter 11 Impact of Junction Properties on the Modal Behavior of Assembled Structures Jean-Baptiste Chassang, Adrien Pelat, Frédéric Ablitzer, Laurent Polac, and Charles Pezerat Abstract This communication is part of the study of the acoustic radiation of assembled structures and presents a preliminary work that consists in analyzing the links between the junction parameters and the dynamic behaviour of the structure, without considering its radiation. The study concerns a system of two beams connected by stiffeners that represent the junction and coupled compressional and flexural motions. The damping at the junction is described by a Kelvin-Voigt model. The coupled motions in each beam are developed by a modal decomposition based on the decoupled free/free beams. The model is solved in the frequency domain by a standard inversion method in the case of linear stiffenesses. The method is validated by comparison with reference results from finite element simulations obtained from Comsol. The analysis of the results aims to evaluate the phenomena of stiffening, damping and wave conversion induced by the junction that may significantly modify the vibroacoustic properties. Keywords Force response · Modal approach · Kelvin-Voigt model · Stick · SDM 11.1 Introduction Powertrains, regardless of their technology (thermal or electrical), are presented as mechanical assemblies with complex geometries. Attention is focused here on the oil sump assemblies that play a central role in the acoustic radiation of the whole structure. These assemblies involve interfaces and junctions of different natures (bolted, glued, crimped, clipped,. . . ) and having various mechanical properties. The faithful modeling of these junctions is an open scientific problem leading to research work. The description of the induced damping and stiffness added by these junctions are key points in this work. In many books and articles, the study is focused on modelling the bolted joint, from the rheological model to the constitutive model [1]. Internal studies at Renault have shown that accurate modelling of a junction does not necessarily have the expected impact on numerical obtained data. A hammer test is performed on an assembly of a cylindrical block and an oil pan connected by ten bolted joints and a frequency response is measured at the excitation point using an accelerometer. The experimental data are compared using a numerical model with a variant of the bolted joint modelling. In each of the three models, the contact areas of the 2 substructures are connected by “glue” elements (mechanical connection of a stiffness type) [2]. In the simplest model, screws are not represented. In a second, more sophisticated model, the screws are represented by equivalent point stiffness, refers as coupled model [3]. In a last model, the 3D geometry of the screws is represented in detail, calls as solid model [3]. The results obtained, Fig. 11.1, shows that these three models of increasing complexity lead to transfer functions very close but quite far from the experimental realization, especially at medium and high frequency. J.-B. Chassang ( ) Laboratoire de l’Université du Mans, LAUM – UMR 6613 CNRS, Le Mans Université, LE MANS, France RENAULT, Lardy, France e-mail: jean-baptiste.chassang@renault.com A. Pelat · F. Ablitzer · C. Pezerat Laboratoire de l’Université du Mans, LAUM – UMR 6613 CNRS, Le Mans Université, LE MANS, France e-mail: adrien.pelat@univ-lemans.fr; frederic.ablitzer@univ-lemans.fr; charles.pezerat@univ-lemans.fr L. Polac RENAULT, Lardy, France e-mail: laurent.polac@renault.fr © The Society for Experimental Mechanics, Inc. 2021 A. Linderholt et al. (eds.), Dynamic Substructures, Volume 4, Conference Proceedings of the Society for Experimental Mechanics Series, https://doi.org/10.1007/978-3-030-47630-4_11 127
RkJQdWJsaXNoZXIy MTMzNzEzMQ==