Search Results

Welding processes are complex in nature. They affect the mechanical properties of a weldment in and around the welding joint (in the heat affected zone: HAZ), causing deformation and inducing high level of residual stress and plastic strain which are detrimental to the weldment performance. After welding some materials soften while others harden in the heat affected zone, depending on the process heat input, the thickness of the material and its chemical composition. Traditionally, finite element (FE) performance analyses (crash, rupture, fatigue, static and dynamic tests) of weldments are performed without accounting for the effects of welding processes and as such the real performance of a weldment is not accurately predicted. On one hand, if base material properties are used to represent a weldment which hardens in the heat affected zone, the performance analysis results would be too conservative which would hinder/limit potential weight reduction strategies.

This paper introduces an approach that uses a statistical energy analysis (SEA) method for prediction of noise in the vehicle cabin from an electric motor sound source placed in the engine compartment. The study integrates three different physics, namely, electromagnetics, harmonics, and acoustics. A 2004 Prius permanent magnet synchronous motor with an interior permanent magnet was used for performing the integrated CAE analysis, as the motor’s design details were readily available. The Maxwell forces on the stator teeth were first calculated by an electromagnetic software package. These forces were then mapped into a finite element model of the motor stator to predict the velocity profiles on the stator frame. Velocity profiles were considered as boundary conditions to calculate sound pressure levels and the equivalent radiated sound power level in the acoustic environment.

Sound propagation through noise control treatment is governed by fluid, mechanical and geometric properties of the materials. The knowledge of material properties is important to improve the acoustical performance of the resulting noise control products. A method based on optimization together with genetic algorithm is used to estimate material properties of multi-layered treatments. Unlike previous inverse characterization approaches based on normal incidence performance metrics measured using standing wave impedance tubes, the current approach is based on random incidence performance metrics. Specially, the insertion loss ‘measured’ from two room transmission loss suite is utilized. To validate the applicability of the proposed method, numerically synthesized insertion loss computed from known material properties are used. In order to properly represent the ‘measured’ values, noise is added to the numerically synthesized insertion loss values.

Since Statistical Energy Analysis (SEA) is based on lumped parameters, acoustic responses predicted by SEA are spatially discontinuous. However, in many practical applications, the ability to predict spatially continuous energy flow is useful for guiding the design of systems with improved acoustical characteristics. A new approach, utilizing integral equations derived from energy flow concepts, is developed to predict the continuous variation of acoustic field such as sound pressure level in the interior of acoustic domains using structural response predicted by SEA. The computer code developed based on energy flow boundary integral equations is initially validated by analyzing sound propagation in a duct.

The identification of the propulsion noise of turbofan engines plays an important role in the design of low-noise aircraft. The noise generation mechanisms of a typical turbofan engine are very complicated and it is not practical, if not impossible, to identify these noise sources efficiently and accurately using numerical or experimental techniques alone. In addition, a major practical concern for the measurement of acoustic pressure inside the duct of a turbofan is the placement of microphones and their supporting frames which will change the flow conditions under normal operational conditions. The measurement of acoustic pressures on the surface of the duct using surface-mounted microphones eliminates this undesirable effect. In this paper, a generalized acoustical holography (GAH) method that is capable of estimating aeroacoustic sources using surface sound pressure is developed.

The identification/localization of propulsion noise in turbo machinery plays an important role in its design and in noise mitigation techniques. Near field acoustic holography (NAH) is the process by which all aspects of the sound field can be reconstructed based on sound pressure measurements in the near field domain. Identification of noise sources, particularly in turbo-machinery applications, efficiently and accurately is difficult due to complex noise generation mechanisms. Backward prediction of the sound field closer to the source than the measurement plane is typically an unstable “ill-posed” inverse problem due to the presence of measurement noise. Therefore regularized inversion techniques are typically implemented for noise source reconstruction. Another major source of ill-posedness in NAH inverse problems is a larger number of unknowns (sources) than available pressure measurements. A model reduction technique is proposed in this paper to address this issue.

As an alternative method to Statistical Energy Analysis (SEA), Energy Finite Element Method (EFEM) offers several unique advantages for vibro-acoustic analysis of structural-acoustic systems. In this paper, the theory of the energy finite element method is overviewed. The main developments of a recently available EFEM code are presented. This is followed by the investigation of several example problems using EFEM; (a) the acoustic pressure computation in an acoustical duct, (b) the sound transmission loss of an automotive dash, and (c) the vibro-acoustic analysis of a truck cab. The EFEM predictions are compared to the analytical solutions, SEA predictions or test data and good correlations are observed. Further, the advantages of EFEM in the solution of high and middle frequency vibro-acoustic problems are discussed.

A hybrid method is developed by combining energy finite element method (EFEM) and energy boundary element method (EBEM) to predict interior noise of structural-acoustic systems at high frequencies. In the hybrid EFEM/EBEM method, the structural domain of the system is modeled by structural finite elements, and the acoustic domain is modeled by acoustic boundary elements. The structural vibration response is computed from EFEM. The interior sound pressure level in the acoustic domain is recovered using EBEM. To validate the hybrid method, the interior noise levels in simplified airplane cabin and van models are computed and compared with that of EFEM only model. Good correlations are observed.

Energy Finite Element Method (EFEM) is an alternative method to currently practiced Statistical Energy Analysis (SEA) for the solution of high frequency vibro-acoustic problems. In this paper, the theory of the energy finite element method for interior noise prediction is reviewed first. This is followed by the investigation of two example problems using EFEM; (a) the interior noise of an airplane cabin and (b) the sound transmission loss of a dash. In both case EFEM results are compared to SEA predictions. The EFEM and SEA results using different mesh density are also investigated. Further, the advantages of EFEM in the solution of high frequency vibroacoustic problems are discussed.

An integrated experimental/numerical technique based on nearfield acoustical holography (NAH) is developed to identify aeroacoustic noise sources associated with turbomachinery noise sources. An efficient enhanced surface potential formulation that is applicable for the resolution of complex noise source identification is developed for this purpose. Numerical examples are presented to validate the applicability of the presently developed formulation for the identification of aeroacoustic noise sources.

A new Energy Finite Element Formulation was developed for interior noise prediction that includes not only ‘the indirect transmission path associated but also the direct transmission path. The formulation was subsequently extended to model noise control treatments by incorporating appropriate modifications to structural-acoustic and acoustic-acoustic joint matrices. The formulations developed are implemented and the resulting computer program was validated by comparing the predictions from the present development to the results from alternative methods.

The relationship between the acoustical performance and macroscopic material properties of elastic porous materials, based on modified Biot's theory, is used to develop sensitivity relationship among acoustic and macroscopic material properties. The sensitivity information are normalized (scaled) and subsequently used to improve and tune the acoustical performance of elastic porous materials. Example problems are used to demonstrate the applicability of the technique.

In this paper, an analytic method has been developed to effectively evaluate the acoustic performance of a vehicle when a traditional sound package component is replaced by a lightweight sound package component. This method avoids the expensive full vehicle tests and statistical energy analysis (SEA) model simulations, and can be used to evaluate the results quickly with acceptable level of accuracy. The developed method is verified by comparing the results with those obtained from a chassis dynamometer test and a full vehicle SEA model simulation. Good correlation is observed between the test and the model. Some parametric studies regarding the performance of lightweight sound packages are also carried out using the method developed in this paper.

Formulation of a fast algorithm for evaluating the Nearfield Acoustic Holography (NAH) transfer functions based on boundary integral equations is presented. The new algorithm overcomes some of the well known numerical difficulties exhibited in the traditional Boundary Element Method (BEM) based NAH transfer function formulations such as excessive computational cost, non-uniqueness problem, and regularization of hypersingular integrals. The formulation is based on deriving transfer functions between acoustic field pressure and surface layer potentials using modified Helmholtz/Kirchhoff integral equations. Example problems are analyzed using both traditional and new algorithms. Solution time and NAH reconstruction results are compared to validate and demonstrate the effectiveness of the new algorithm. Finally, the acoustic holography reconstruction is applied for an automotive powerplant.

An application involving noise source reconstruction on a full automotive powerplant including the engine, manifolds and the transmission is considered herein, to demonstrate the versatility of modern generalized acoustical holography. The complex source geometry necessitates measurements on non-conforming surfaces. The acoustic pressures were experimentally acquired at three different engine excitations. Accelerometers were mounted at select locations on the powerplant in order to study the accuracy of the reconstructed vibrations from acoustical holography. Through a series of synthetically generated holograms with added random noise, it is conclusively demonstrated that the error margins in the reconstructed vibrations on the powerplant are consistent with errors in reconstructed vibrations from numerically synthesized holograms of a similar Signal to Noise Ratio (SNR).

Biot's theory provides a framework for the numerical modeling of propagating stress waves in elastic porous materials. A finite element method technique based on the adaptation of Biot's theory [1, 2] to acoustic porous material that is applicable for the solution of complex systems consisting of porous, fluid and structural media is described. Acoustic indicators such as absorption coefficient and transmission loss are calculated for flat samples and these results are compared to known solutions. Finally the transmission loss of a complex dash system is computed and contrasted with the corresponding planar multi-layer results.

The acoustical performance of a vehicle dash system is determined by computing the sound transmission loss (STL) in the diffuse field. The determination of the sound transmission loss of the dash system is usually performed in the test laboratory with the anechoic chamber on the transmission side of the dash. Computational modeling of such system poses challenge due to the complexity of the material properties as well as geometry. In this paper, the transmission loss of a vehicle dash is calculated by using Finite element method. The dash system considered is a multilayer system consisting of elastic porous and elastic structural materials. The effects of different design parameters on the dash STL are investigated.

A general numerical formulation based on the Boundary Element Method (BEM) for computing radiated sound power sensitivity is presented in this paper. The total radiated sound power is computed using surface acoustic pressure and velocity information. Explicit analytical differentiation of the sound power with respect to acoustic normal velocities is performed on the boundary integral equations to obtain sound power sensitivity information. The formulation is applicable to structures with arbitrary geometries, free edges, openings, and multiple connections. Acoustic absorption materials applied on the structure surface can also be modeled as impedance boundary conditions. The developed formulation is validated and its application is demonstrated.

In this paper, the development of a generic SEA template of a four-door passenger car is presented. Validation of the templates is given by comparing the results with those obtained from the detailed SEA model and tests. The derivation of SEA model based on validated templates can reduce modeling effort substantially. Additionally, it also improves the quality of the model since this model is extracted from validated template model. Morphing the developed templates to a new production vehicle was also demonstrated to illustrate a morphing procedure, which can save large SEA model developing time.

Acoustical performance of many NVH systems is evaluated by computing the radiated noise of such systems. Although the computation of relevant acoustical quantities such as the radiation efficiency and sound power can be performed by using approximate method such as the Rayleigh integral method, in most instances rigorous methods such as the boundary element method is preferred. Unfortunately the boundary element method for radiation problems fails at the eigenfrequencies associated with the interior cavity. Additionally, in most instances the acoustical performance is needed to be evaluated over a range of frequencies and in a traditional boundary element system this evaluation is performed independently at each discrete frequency. In this presentation, issues associated with the accuracy and efficiency of these solutions are addressed by investigating the solution of progressively more complex industrial problems.