Development of Evaluation Technology for Tire External Noise and Investigation of Generation Mechanism (2nd Report)

Kotaro Mune･Ryuichi Tanimoto (Sekiso)

In our previous report, we presented a mapping of external noise generated by tires and methods to separate the noise radiated from the front and rear wheels. In the present study, we investigated the mechanism by which tire noise spreads outside the vehicle and proposed a countermeasure to confine the noise in the space behind the bumper before it spreads. Using this technique in an actual vehicle test, we improved the sound flow around the tires and successfully reduced external noise.

Analysis of suspension and tire modal energy contributions to whole structure resonance

Masato Hashioka･Masashi Komada･Hiroaki Mizuno (Toyota Motor)･Yuichi Matsumura (GIFU UNIVERSITY)

To reduce road noise, it is necessary to clarify the contribution of each subsystem to the whole-system resonance, which is a strong vibration coupling between the tire and suspension. In this study, mode energy contribution analysis is performed on the transfer function of only the tire and suspension coupling points to extract important eigenmodes of each subsystem and understand the mechanism of vibration coupling. Furthermore, guidelines for modifying subsystems to control the blocked force of the whole-system resonance are presented for assigning effective performance targets to subsystems.

A Machine Learning Approach to Estimate Tire Block Force Spectrum for Road Noise Simulation

Yonghun Kim･Taeyoung Kim･HyunSeok Kang (Hankook Tire)

To predict vehicle road noise using the Frequency Based Sub-structuring (FBS) method or to operate a virtual noise simulator, accurate tire block forces are essential. Prior to manufacturing a physical tire, these forces are commonly estimated through finite-element (FE) analysis; however, FE simulations typically require several days, making it difficult to rapidly assess the impact of design modifications on road-noise performance. To overcome this limitation, this study proposes a data-driven modeling framework capable of predicting the tire block-force spectrum directly from design specifications. The methodology, including feature construction, model training, and validation procedures, is described in detail, and the applicability of the proposed approach is demonstrated through prediction results.

Efficient Machine Learning Optimisation of Gear Micro Geometry and Comparison with Manually Designed Gears

Andrew Wild･Simon Terry･Paul Langlois (SMT)･Jaekwon Lee･Shinya Sudo (Hino Motors)

Optimisation of gear micro geometry guided by machine-learnt surrogate models is compared to a past engineer-led optimisation and proven to be a highly efficient technique. Subject to engineer-defined objectives, the optimiser identifies high-performing micro geometries with no further human intervention, with all predictions automatically validated by a hybrid Hertzian and FE-based loaded tooth contact analysis. Going beyond the scope of the original manual optimisation, the optimiser is used to identify micro geometries that are torque robust or retain similar performance to the specification without using different micro geometries on the two input gears, thereby theoretically reducing manufacturing costs.