Mechanism of limestone degradation under alkaline wet-dry cycles using mel-frequency cepstral coefficient

• Published in: Engineering geology Vol. 357; p. 108355
• Main Authors: Chen, Yucheng, Xie, Qiang, Zhang, Hexing, Tu, Zhengnan, Lei, Xianghuan, Fu, Xiang
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.10.2025
• Subjects: Alkaline environment; Failure warning method; Limestone; Mel-frequency cepstral coefficients; Wet-dry cycles; Limestone; Mel-frequency cepstral coefficients; Wet-dry cycles; Alkaline environment; Failure warning method
• ISSN: 0013-7952
• DOI: 10.1016/j.enggeo.2025.108355

Periodic reservoir water level fluctuations subject rocks in the water-level fluctuation zone to repeated wet-dry cycles, leading to progressive mechanical degradation. The pH of the water plays a critical role in influencing rock behavior under these conditions. This study investigates the mechanisms of strength degradation in jointed limestone subjected to neutral and alkaline wet-dry cycles, reflecting the weakly alkaline conditions characteristic of the Yangtze River. Nuclear Magnetic Resonance (NMR) and X-ray Diffraction (XRD) analyses were conducted to characterize microstructural changes. Acoustic Emission (AE) signals were processed using Mel-frequency cepstral coefficients (MFCC), Gaussian Mixture Models (GMM), and gray-level co-occurrence matrix (GLCM) techniques.The results show that the peak strength of the sample decrease by 18.66 % in an alkaline environment, whereas in a neutral environment, the reduction is 9.7 %. The alkaline environment also increase the likelihood of surface spalling during failure due to dedolomitization reaction. Under both neutral and alkaline conditions, the MFCC spectrum entropy values exhibit an initial increase followed by a decrease with loading time. In contrast, the alkaline environment results in a significant increase in MFCC spectrum contrast, from 16.32 to 56.26, indicating a higher degree of brittle failure in the sample. Compared with the traditional b-value method, the average prediction lead time of the MFCC entropy-based method is 10.78 s earlier, which demonstrates superior overall prediction performance. This study clarifies the degradation mechanism of rock strength in alkaline environments and provides a scientific basis for mitigating engineering geological hazards in water-level fluctuation zone. •Alkaline condition enhanced mesopore-macropore transformation and surface spalling.•Alkaline condition raised MFCC contrast, indicating more brittle failure in sample.•A novel MFCC-GLCM method enables staged microcrack and failure prediction.