通过变粒岩单轴压缩下声发射试验,采集了应力一应变曲线与声发射数据。联合应力曲线,总结了声发射积累计数和积累能量的演化特征。利用Welch谱估计法与小波能谱系数法分别对试件声发射信号功率谱密度及能量分布特征进行了分析。研究表明,声发射积累能量比振铃积累计数能更好地预测岩石破坏;在试件破坏前或产生较大裂纹时,声发射信号峰值频率会有所下降,其能量集中频段也在扩大;在[78kHz,156kHz]频段能谱系数持续增大且其值超过50%的声发射信号出现频繁的现象,可作为试件破坏失稳的前兆信息。结合岩石声发射信号时频特性,可为预测岩体破坏提供一种新思路。
Through the acoustic emission(AE) test of leptynite with uniaxial loading, the stress-strain curves and data of AE were obtained. Combined with the stress curve, the evolutionary characters of AE cumulative count and energy were analyzed. Power spectrum and energy distribution of rock AE signals were also analyzed by using Welch power spectrum estimation and wavelet energy spectrum coefficient method. The results show that cumulative energy is better than cumulative count in predicting rock failure. Before the failure or fissure occurred on specimen, peak frequency of AE signal declined with frequency range of energy concentration expanding. The phenomenon of energy spectrum coefficient in the frequency range of [78 kHz, 156 kHz] increasing as precursory information for assessing instability and with over 50% AE signals cropping up frequently, can be used damage characteristics of AE signal, can provide an effective technological of specimen, which, together with time-frequency means for predicting rock failure.