Search Results

Several of the authors have recently developed procedures to efficiently evaluate experimentally the relative contributions of various wind noise paths and sources. These procedures are described and, as a case study, results are provided for the noise in the interior of a production automobile subjected to wind tunnel airflow. The present measurements and analysis indicate that for the tested vehicle significant contributions to interior noise are provided by underbody and wheel well flows, radiation from the roof and seal aspiration. A significant tone associated with vortex shedding from the radio antenna was also noted.

A CAE procedure for modeling the aerodynamic excitation of greenhouse surface vibration and its reradiation as noise is described. The procedure begins with a description of the steady flow over the surfaces. This is used as a basis for estimating the spatially varying unsteady pressure loading. The approach is semi-empirical, utilizing normalized pressure data collected through wind tunnel testing of production vehicles. The unsteady pressures are utilized within a normal mode analysis to predict vibration of the greenhouse panels. Interior noise associated with the panel vibration is estimated from a statistical energy analysis model. We show that contributions of multiple surfaces can be significant.

Power train noise reaches the interior through structureborne paths and through airborne transmission of engine casing noise. To determine transfer functions from vibration to interior noise a shaker was attached at the engine attachment points, with the engine removed. A simple engine noise simulator, with loudspeaker cones on its faces, was placed in the engine compartment to measure airborne transfer functions to interior noise. Empirical noise estimates, based on the incoherent sum of contributions for individual source terms times the appropriate transfer function, compared remarkably well with measured levels obtained from dynomometer tests. Airborne transmission dominates above 1.5kHz. At lower frequencies engine casing radiation and vibration contributions are comparable.

Statistical Energy Analysis (SEA) models of passenger vehicles may involve 400-500 subsystems to characterize the vehicle structure and sound package. Full sized models are often difficult to describe and some results may be awkward to present. A more intuitively comprehensible model size would involve 30-50 subsystems, which is typical of an experimental SEA approach. A methodology is described for processing full SEA model results in condensed form including major structural components and acoustic spaces. The approach is based on the experimental SEA methodology in which power injection and response predictions are carried out for 30-50 condensed subsystems. Full matrix inversion and weak coupling assumptions were investigated. Differences between condensed and full SEA model predictions are observed when the subsystem connectivity is limited.