Search Results

Middle and high frequency vibro-acoustic simulation of complex shape insulators requires using 3D poroelastic finite elements. This can be applied to either the whole part (up to 2500 Hz maximum) or through singly curved pre-computed Insertion Losses (up to 5000 Hz maximum) to be introduced in large SEA or energy-based models. Indeed, a dependence of the Insertion Loss slopes of noise treatments following the curvature is observed both experimentally and numerically. Beyond frequency range limitations, poroelastic finite element simulations following all curvatures and thickness 3D maps typically take too much time of up to a few hours each. A cylindrical Transfer Matrix Method spectral approach significantly reduces the time for the calculation of singly curved Insertion Losses up to 10 kHz to only a few minutes. This simplifies enormously the SEA modeling effort enabling easier, more precise fully trimmed vehicle middle and high frequency vibro-acoustic simulations.

The purpose of the SONVERT project is to create a link between the acoustical sources of a car and the environment in terms of traffic and architecture. Based on well validated approaches, it introduces the notion of a “macro-source” which integrates the major acoustic sources: engine, tires and exhaust, taking into account the low and high frequency aspects, from measurements made on real vehicles. The macro-source is then integrated into an original approach dealing with outdoor propagation. The proposed method can consequently be seen as a first step toward a global approach for the study of traffic noise in real conditions.

In the high frequency range, the SEA method has been applied to air borne path with success to predict both internal and external sound environment. Nevertheless, structure-borne prediction is still at issue -especially for cars, in the range 200 to 2000 Hz- as results are widely dependant on subsystem partition and validity of various assumptions required by SEA. Experimental SEA test technique (ESEA), applied to car bodies, has brought to the fore that SEA power balanced equations could robustly describe structure-borne noise. To make ESEA predictive, the database of measured FRF is simply replaced and enlarged by synthesized data generated from a finite element (FE) model and a selected observation grid of nodes. This technique, called Virtual SEA (VSEA), has been presented at SAE/NVC 2003.

Statistical Energy Analysis (SEA) method is mainly used in automotive industry to build-up comprehensive models of the vibratory energy transmission throughout the body-in-white and to perform estimates of the acoustic performance on the full-trimmed vehicle. If the latter use is given rather satisfactory predictive results, the former is requiring more expertise and backward experience from testing. The main cause of dispersion is due to the use of simple analytical formulae to describe complex structural behavior in current SEA software and to the choice of a SEA portioning into subsystems that makes the model accuracy dependant on the user's choice. Nevertheless, SEA models are very attractive as they are small, easy to optimize and include the main dynamics features. A technique for improving accuracy of SEA models, especially for structure-borne-sound by interfacing them with Finite Element (FE) modeling is presented.

Statistical Energy Analysis (SEA) is used to predict wide-bandwidth noise and vibration. That prediction may rely on parameters derived from theory or from test, which essentially means that there are two distinct approaches, analytical SEA and test-based SEA. The latter is the focus of this paper. Both theory and practice are reviewed, so that the current status of the method can be established. This review also provides some insight on what information can be extracted from the experiment, how the measurements must be conducted and how the results must be interpreted. Another important aspect of test-based SEA is its interaction with the more widely used analytical SEA method. It is demonstrated that both methods are complementary and that the analytical and test-based parameters can either be compared or mixed in a “hybrid” SEA,model. Benefits of the combined use of the methods are discussed. The discussions are supported by results obtained for automotive applications.