Prediction of Minimum Sound Emission Requirements of an Electric/Hybrid Vehicle 2023-01-1099
Electric and Hybrid vehicles have standards for emitting enough noise to reduce danger and risk to pedestrians when operating at low speeds. Simulation can help to support development and deployment of these systems while avoiding a time-consuming, test-based approach to design these AVAS (Acoustic Vehicle Alerting System) warning systems. Traditionally, deterministic simulation methods such as Finite Element Method (FEM) and Boundary Element Method (BEM) are used at low frequencies and statistical, energy-based methods such as Statistical Energy Analysis (SEA) are used at high frequencies. The deterministic methods are accurate, but computationally inefficient, particularly when the frequency increases. SEA is computationally efficient but does not capture well the physics of exterior acoustic propagation. An alternative method commonly used in room acoustics, based on geometrical or ray acoustics, is “Ray Tracing” and can be used for sound field prediction. Ray Tracing generally estimates sound fields with an acceptable degree of accuracy through the full range of frequencies and can predict temporal and spatial distribution of sound fields with suitable accuracy for auralization. The image-source method is a ray acoustics method that guarantees finding all reflection paths up to a given order and can include diffraction effects.
In this paper, a systematic approach is discussed to evaluate the potential for Ray Tracing to predict sound fields for AVAS design. Ray Tracing models with different combinations of parameters are compared to a BEM model as a reference. An accelerated BEM method known as H-matrix is also included in the comparison as a reference and to evaluate the potential for accelerated BEM to be used when more accuracy than Ray Tracing is needed but standard BEM is challenging to use because of model size and frequency range. Some conclusions about these methods for this application and others, such as Pass-By Noise prediction, are presented.