Journal of Acoustics and Vibration Research

Feature-Optimised Machine Learning Framework for Intelligent CNC Grinding Tool Monitoring

Tuan A.Z. Rahman — 2026-06-17

Tool Condition Monitoring (TCM) is vital in CNC grinding to ensure dimensional precision, surface quality, and process efficiency. This study introduces a machine learning-based TCM framework that integrates vibration signal analysis, feature extraction, and metaheuristic optimization. Vibration data from industrial CNC grinding were processed to extract time- and frequency-domain features, which were then applied to tool wear prediction using Support Vector Regression (SVR). To improve accuracy and robustness, the Marine Predators Algorithm (MPA) was employed for hyperparameter tuning. The MPA-optimized SVR achieved the lowest mean squared error (0.0028), outperforming autoregressive (0.0088) and power-law (0.0084) models. These findings demonstrate the potential of combining vibration analysis with metaheuristic-optimized machine learning for intelligent, scalable, and real-time TCM. The proposed framework aligns with Industry 4.0 objectives by reducing downtime, improving reliability, and enabling adaptive monitoring in precision manufacturing.

Acoustic Modeling and Simulation of an Adjustable Concert Hall with Variable Reverberation Time for Classical and Angklung Music

Devina Tambunan — 2026-06-17

The acoustic environment of a concert hall is acoustically specially designed for musical performances, typically in a single configuration for general use. While this may be effective for certain genres, it can be less suitable for others. Since traditional music is usually performed with no standardized acoustic, large-capacity halls specifically designed for traditional music, particularly angklung, remain scarce, let alone one capable of accommodating both classical and traditional music in a multipurpose setting. This study explores adjustable design capable of accommodating multiple genres through variable reverberation times. Through modeling and simulation using the room mapping module in EASE 5 software, four hall configurations were developed for angklung and three classical music periods by achieving their respective target RT at 1 kHz. Additional parameters evaluated were C₈₀, T_s, LF, SPL, and G. Adjustability was achieved by varying the ceiling height and materials of the ceiling and rear wall. All four configurations successfully fulfilled the acoustic criteria for their respective genre: 2.0 s for Romantic, 1.7 s for Classical, 1.6 s for Opera, and 1.3 s for angklung. The results of this study confirm that an adjustable concert hall design with variable geometry and surface materials is a versatile solution that can effectively accommodate a wide range of musical genres, from classical to Indonesian traditional music.

Analysis on Sound Reduction and Transmission Loss Using Compress Acoustic Chamber

Zainal Fitri Zainal Abidin — 2026-06-17

Noise pollution has increasingly affected public health due to its impact on productivity, hearing loss and overall quality of life. Acoustic insulation is an effective and practical solution to reduce unwanted noise in various environments. However, most acoustic chambers used to evaluate insulation performance are large, costly and require specialized facilities. Therefore, this study aimed to design, construct and validate a reduced size acoustic transmission chamber according to ISO 10140-5 guidelines. The chamber was used to measure sound insulation properties of commonly available materials such as polyester felt and expanded polystyrene (EPS) foam. Testing methods involved recording Sound Pressure Level (SPL), Sound Transmission Loss (STL) and transmission coefficient across frequencies from 100 Hz to 5000 Hz. The study concluded that scaled-down acoustic chambers can reliably evaluate acoustic materials and confirmed that thickness, density and internal material structure significantly influence sound insulation effectiveness.

Identification of Suitable Excitation Nodes for Assembled Structures using Modal Participation Factors

M.S.A. Mohd Kahar — 2026-06-17

Accurate excitation point selection is essential in experimental modal analysis (EMA) to ensure sufficient excitation of structural vibration modes and the acquisition of quality Frequency Response Functions (FRFs). However, excitation nodes are commonly selected based on geometric accessibility and engineering judgement, which may result in incomplete dynamic characteristics. This study presents a systematic methodology for identifying suitable excitation nodes for a simplified aircraft structure. FE modal analysis is first performed using MSC NASTRAN SOL103 to obtain the eigenvalues and eigenvectors of the structure. The extracted modal parameters are subsequently used to calculate the Modal Participation Factors (MPFs) and effective modal masses for all candidate excitation nodes. Based on the obtained MPFs, excitation nodes are ranked according to their capability to excite multiple structural modes within the investigated frequency range. The effectiveness of the selected excitation locations is then validated using driving point FRF analysis through MSC NASTRAN SOL111. The results show that nodes with large MPF and effective modal mass values produce significantly stronger FRF amplitudes and clearer resonance peaks across multiple vibration modes. The strong agreement between the MPF based excitation ranking and the corresponding FRF responses demonstrates the effectiveness of the proposed methodology for EMA and structural dynamic investigations.

Identification of Nonlinearities in Mechanical Structure Joints: A Systematic Review

M.S.M. Sani — 2026-06-17

The identification of nonlinearities in mechanical structure joints is essential for improving predictive maintenance, structural health monitoring, and design optimization of engineering systems. Mechanical joints frequently exhibit complex nonlinear behaviors, including friction, backlash, hysteresis, and contact-induced effects, which can significantly influence the dynamic performance and reliability of structures. Although numerous studies have investigated these nonlinear phenomena, a comprehensive synthesis of identification techniques specifically addressing joint nonlinearities remains limited. This systematic review aims to bridge this gap by critically examining contemporary methodologies used to detect, characterize, and quantify nonlinear behavior in mechanical joints. The review focuses on three primary objectives: (i) classifying existing identification approaches, (ii) evaluating their effectiveness, applicability, and limitations, and (iii) identifying current challenges and future research opportunities. A systematic literature search was conducted across major engineering databases, encompassing experimental, numerical, and hybrid techniques for nonlinearity identification. Key methods, including nonlinear system identification, parameter estimation, model updating, and data-driven approaches, were comparatively analyzed. The findings indicate that frequency-domain and time-domain methods remain the most widely adopted techniques, while recent advances increasingly integrate machine learning and artificial intelligence to improve identification accuracy and robustness. Furthermore, the review highlights the growing importance of multi-physics and hybrid modeling frameworks for capturing complex joint behaviors under varying operational conditions. By consolidating and critically evaluating existing knowledge, this study provides a structured reference for selecting suitable identification methods and offers insights to support the development of more reliable mechanical systems, thereby contributing to advancements in structural health monitoring, maintenance, and engineering design.