科研之友
微信
新浪微博
Facebook
分享链接
摘要
Ultrasonic wave propagation in heterogeneous porous cores under laboratory studies is an extremely complex process involved with strong scattering by microscale heterogeneous structures. The resulting coda waves, as an index to measure scattering attenuation, are recorded as continuous waveforms in the tail portion of wavetrains. Because of the contamination of reflections from the side ends and reverberations between the sample surfaces, it is difficult to extract pure coda waves from ultrasonic measurements for the estimation of the P- and S-coda attenuation quality factors. Comparisons of numerical and experimental ultrasonic wave propagation in heterogeneous porous cores can give important insight into understanding the effect of boundary reflections on the P-and S-codas in the laboratory experiment. It challenges numerical modeling techniques by three major issues: the creation of a digital core model to map heterogeneous rock properties in detail, the perfect simulation with a controllable and accurate absorbing boundary, and overcoming the numerical dispersions resulting from high-frequency propagation and strong heterogeneity in material. A rotated staggered-grid finite-difference method of Blot's poroelastic equations is presented with an unsplit convolutional perfectly matched layer (CPML) absorbing boundary to simulate poroelastic wave propagation in isotropic and fluid-saturated porous media. The contamination of boundary reflections on coda waves is controlled by the CPML absorbing coefficients for the comparison between numerical and experimental ultrasonic waveforms. Numerical examples with a digital porous core demonstrate that the boundary reflections contaminate coda waves seriously, causing much larger coda quality factors and thus underestimating scattering attenuation.
- 出版日期2014-5
- 单位中国科学院地质与地球物理研究所; 中国科学院地质与地球物理研究所兰州油气资源研究中心
