The use of machine learning algorithms in the simulation of multi-layer acoustic palliatives. 2024-01-2928

Acoustic palliatives used in the automotive industry have evolved from simple felt and heavy layer combinations into highly complex formulations and combinations to account for higher performance targets, lower weight and inevitably cost constraints. Achieving Customer performance compliance usually involves a time-consuming exercise of material characterisation and measurement. Ideally this should be carried out via simulation, but as material mixtures and compositions become more complex, the ability to accurately simulate their acoustic performance is becoming increasingly difficult. Historically, Biot parameters and their associated TMM models have been used to simulate the acoustic performance of multi-layer material compositions. However, these simulations are not able to account for real-world complexities such as manufacturing imperfections or inter-layer gluing effects. The assumptions made by simulation models, such as a perfectly diffuse field, are rarely true in actual measurements, further increasing the uncertainty when comparing measurement against simulation. There already exists widely accepted methods for obtaining the Biot parameters for single-layer materials. Typically, a multi-layer simulation considers each individual layer in isolation rather than its interactions with the rest of the composition after heating, compression, or gluing. The current trend towards sustainability is also adding restrictions to the types of materials that can be used. With the above considerations in mind, and due to the availability of large measured material databases, the goal of the authors is to design a machine learning algorithm that will directly calculate the overall acoustic performance without needing to consider each layer individually. This paper examines each stage of the design of the algorithm and its expected efficiency.