基于经验模态分解的应力松弛曲线波动性分析

Application of Empirical Mode Decomposition in the Fluctuation Analysis of Stress Relaxation Curves

应力松弛试验是研究土体时间效应特性的重要手段,但其受土体本身的复杂性、试验周期长和外界因素的影响及测量系统的限制,获得的结果常具有复杂性和波动性。为深入分析应力松弛试验结果与提高分析结论的可靠性,本文应用经验模态分解理论,分析了非饱和石灰改良膨胀土应力松弛试验的实测时间序列,以探求松弛曲线波动性产生的主因。研究得出应力松弛曲线波动性以中长期波动为主,特征时间尺度为1天的特征模态函数分量的波动贡献率达17.4%,结合实测温差和偏应力数据分析得出昼夜温差是引起实例松弛曲线波动的主要因素,试验采样间隔对曲线波动性也存在一定影响。研究结果为应力松弛试验的数据后续处理和试验方法的改进提供了重要依据。

Stress relaxation test is a major method used to investigate the time-dependent behaviour of soil mass.Stress relaxation curves of soil often show fluctuation and complexity since they involve limitations of the testing system and influences of soil nature,long test period,and external factors. To further investigate measured time series and improve the analysis reliability,the empirical mode decomposition（ EMD） method is introduced therefore to decompose measured stress relaxation curves of the lime-treated expansive clay into time series in multiple time scales to analyze the reason of fluctuation of curves obtained from stress relaxation experiments by geotechnical digital system（ GDS） for unsaturated soils. Results show that the mid-long term fluctuations are dominant in stress relaxation curves. The variance contribution ratio of intrinsic mode function（ IMF） of 1day is 17. 4%,which is resulted from the temperature difference between day and night and can be a major influence factor of fluctuation combined with measured data of temperature and deviatoric stress. And the sampling interval also contributes to the fluctuation.These results confirm the effects of the temperature difference between day and night on the stress relaxation,and provide an important basis for the subsequent process of testing results and the improvement of the testing method.