【作者】 许年春；

【导师】 吴德伦； 赵明阶；

【作者基本信息】 重庆大学 ， 岩土工程， 2007， 博士

【摘要】 节理对岩体强度和稳定性有重要的影响,其影响程度直接与节理的几何参数和力学参数相关,包括节理的走向、倾角、位置、厚度、刚度、强度等。如何确定隐伏节理的特征参数,一直以来是岩体工程界关心和亟待解决的关键技术问题。目前岩体工程波动法测试中在对测试信号的解释和分析时往往只单一考虑波速、振幅衰减或频谱特点,使得节理波动探测成本高或是精度低,且获得的参数少。因此通过综合、深入分析节理反射信号,反演节理多个特征参数的研究无疑具有重要的理论意义和工程实用价值。垂直节理面入射应力波在节理内往返共线传播,向波源点发出多次回波,引起波源点的振动变化,根据正弦波形函数和Snell定律推导出各次回波的振幅与节理特征的关系式,据此建立节理多参数反演反射法理论模型。为接收到最理想的垂直节理面回波信号,需事先采用反射波反演出节理产状。从地表发出的应力波,经节理面反射后回到地表面,根据解析几何中点、线、面的关系,推导出节理面反射波走时,得出地表反射波等走时圆公式,从而建立起由测点反射波走时反演节理产状的理论模型。室内模型试验中,砂浆模拟岩石,湿砂层模拟节理,测试面(模型上表面)与节理面斜交,中心布置发射探头,接收探头四周布置,受面波、直达波干扰,无法直观判断出接收信号中反射波初至时刻,通过复信号分析可得到信号的瞬时信息,由瞬时频率能有效判断反射波初至时刻,得到五个测点处反射波走时。在MATLAB平台上编制程序,由测点走时搜寻出等走时圆圆心,进而得到节理面真倾向及倾角值,反演结果与模型设计值相差较小。另一个模型试验中测试面与节理面平行,采用垂直反射法测试(接、发探头并排微小间距摆置)。对测试信号作复信号分析,瞬时频率能清楚地显示出了各次回波的初至时刻;频谱分析表明测试信号中含有两个主频,分别对应发射探头的发射主频与测试面自振频率(采用ANSYS有限元程序作半空间体模态分析加以验证);采用Littlewood-Paley小波改造的带通滤波器分别滤出两个主频信号,滤出的信号均能反映出接收信号随各次回波的传至的变化规律。为从测试信号中提取出零次回波,采用Duhamel积分计算出探头发射信号激励下测试面的受迫振动,该受迫振动沿表面传播形成面波、直达波,借助于Laplace变换和Hankel变换可模拟出表面振动,受迫振动向下传播形成节理入射波,以该受迫振动的波形作为零次回波的波形,以测试信号消除零次回波后瞬时频率标准差最小作为评判标准,在MATLAB平台上编程搜寻出局部最小值,确定出零次回波的振幅值。提取零次回波的方法不适用于提取一次、二次回波,一方面是因为后至回波振幅减小,另一方面是因为应力波穿过节理层时衰减滤波,波形发生了变化。通过Hilbert变换可得到测试信号(一维)的平面迹线(二维),测试信号中只含有两个频率成分,视为双频信号,双频信号的平面迹线相当于两连杆绕平行轴转动时外杆杆端迹线,利用理论力学中运动合成和瞬心知识,推导出双频信号平面迹线步长平方满足余弦线性复合函数,而迹线转角可视为线性变化,据此可对回波到来之前的测试信号平面迹线作延长,得到一定长度的信号延展段,信号延展段与实测段作差可得到一次、二次回波的第一个波峰值。在得到各次回波第一个波峰值后,代入已建立的节理多参数反演理论模型,求出模型实验中节理的厚度、品质因子、波速等,较好地实现了多参数反演。更多还原

【Abstract】 The joint has great effect on the strength and stability of rock mass, the effect degree directly relates to joint’s geometry and mechanics parameters, including the joint’s strike, dip angle, location, thickness, stiffness, strength et al. How to determine an underlying joint’s characteristic parameters has been a key technical subject which concerned by rock mass engineers. Present rock mass engineering wave testing always only consider wave velocity or amplitude attenuation or spectrum characteristics when the tested signal be complained and analyzed, as a result, the cost is high or precision is low when the technique be used to detect a joint, furthermore, only one or two parameters can be acquired. So a research on multi-parameter inversion for joint through comprehensive and further analysis of signal reflected by joint indubitably has important theoretic meaning and engineering practical value.Through the analysis of stress wave’s propagating process in the vertical direction at the joint, the expression of a series of echoes generated at the joint is achieved, then the combined vibration at the initial wave source is determined, finally a inverse model to detect the joint’s characters based on signal process of the vibration is given.Based on the analytic geometry theory, a formula to calculate the reflected wave’s traveling time can be deduced. The formula establishes a direct relationship between a joint’s real dip angle, real dip direction and the traveling time of testing spots. The reflected wave’s isotraveling time circle described by the formula has a clear conception and a definite meaning. Through complex signal analysis, a signal’s instantaneous amplitude, phase and frequency can be presented. A simulating jointed rock masses, mortar to simulate rock and wet soil layer to simulate joint, be made and the complex signal analysis be applied to analyze the vibration signal at the testing spots on the upper rock surface. Despite the Rayleigh wave’s serious interference, the initial arriving moment of the joint reflected wave can still be judged from instantaneous frequency efficiently. Using the complex signal analysis, the traveling time of several testing spots distributing around the source of vibration be get, a program based on MATLAB be written to search the center of the isotraveling time circle, then the joint’s attitude be inversed in the end.In the other simulating joint model the testing plane is parallel to the joint, the receiving probe be set close to the sending probe to receive the vertical reflected signal. Applying complex signal analysis to analyze the signal, the initial arriving moment of every echo be showed clearly by instantaneous frequency figure. The signal’s amplitude spectrum reveals that there are two main frequency, the higher is the probe signal’s dominant frequency and the lower is the testing plane’s free vibration frequency. The band-pass filter which be transformed from Littlewood-Paley wavelet can filter the two frequency signal separately, and the filtered signals display the change of the signal’s amplitude that corresponding with the each echo’s arrival.In order to separate initial echo from the tested signal, Duhamel integral be using to calculating the testing plane’s forced vibration motivated by the probe’s emitted signal. The forced vibration induces direct wave and Rayleigh wave which propagate along surface, with the help of LT and HT, the surface vibration be simulated. The forced vibration transmits downwards and induces incident wave for the joint, so the initial echo’s waveform is same as the forced vibration’s. A new signal can be get if initial echo be subtracted from the tested signal, set the new signal’s instantaneous frequency’s standard deviation as a assessment function, a program be designed on MATLAB to search the minimum, then the echo’s amplitude be obtained.Through Hilbert transform, a signal’s plane trace can be get. There are two frequencies in the tested signal, so the signal’s plane trace may be looked as the track of the out pole’s end when two poles rotate around parallel axles. Using the knowledge of motion synthesis in theory mechanics, a formula to describing the change of plane trace’s step length and angle for two-frequency be get. So we may prolongate the tested signal’s plane trace from the echo’s initial arriving moment, and get a certain length of the signal’s prolongation. If we subtract the prolongation from the real tested signal, the amplitude of 1st and 2nd echo be obtained.After the amplitude of every echo be obtained, we can use the inverse model to acquire the joint’s thickness, quality factor, P-wave velocity for the simulating model.更多还原