An Adaptive Computational Acoustics Model for Noise Control and Sound Radiation Analysis in Engineering Structures

Abstract

Engineering structure noise radiations have been a continuous menace in the aerospace, automotive, and industrial application whereby large acoustic emission volumes impact negatively on the performance of systems, system safety of operation, along with comfort to people. A typical model use of traditional computational acoustics is to use some form of static equation with solid material characteristics and boundary conditions, thereby restricting its applicability and usefulness in changing operating conditions. In order to circumvent these constraints, an adaptive computational acoustics model of noise control and sound radiations analysis of engineering structures is developed in the current paper. The proposed model combines structural vibration analysis, and acoustic wave propagation by a developed coupled finite element/boundary element formulation and adaptive parameter up-dating which implies consideration of the uncertainty of damping, boundary impedance, and material properties. The error minimization approach, which is based on feedback, ensures that model parameters are continuously varied and changed to correspond to differences between predicted acoustic responses and reference ones to make the model correct itself as the conditions change. Numerical studies conducted on exemplary structural/building models show that the adaptive framework is much more effective in predicting the sound pressure levels with high accuracy and convergence compared with the traditional, purely, vibro-acoustic models. Moreover, attainable noise reduction by the proposed approach is visible, specifically in the frequency bands dominated by resonance in which the parameter uncertainty plays the most significant role. The conclusions suggest that adaptive mechanisms can be directly incorporated into models of computational acoustics to increase their reliability and scalability, which means that the suggested approach is particularly applicable to the noise control of the future, to intelligent structural systems, and to the new digital twins-based analysis of acoustics.