Development of FBMINC model for particle diffusion in fluids

An accurate particle tracking method using FBMINC （new fractional Brownian motion） is outlined. It generates non-Fickian diffusion rather than Fickian diffusion as traditional particle tracking model does. The FBMINC model is based on fractional Brownian motion （fl3m） which is generalization of regular Brownian motion. The two models of fBms （FBM model and FBMINC model） were explored and the differences of the two models are compared from the three aspects： the standard deviation of each step, the small memory and the effect of the number of particles in the cloud. The results show the FBMINC model is a better model as it produces more accurate statistics. The effect of simple shear dispersion for both Brownian and fBm was investigated. The power law scaling of fBm shear dispersion was correctly identified. In addition, a scaling coettieient was found numerically. The FBMINC model is then used for producing both superdiffusive and subdiffusive particle paths, therefore, the non-Fickian diffusion of soil particle clouds can be modelled. The particle clouds represent soil contaminant are released in an idealised coastal bay and the fBm particle tracking method is used for simulation the particle cloud spreading in the bay. There is a noticeable increase in the spreading rate of the cloud. In addition, owning to the spreading rate of the cloud, a noticeable part of it has escaped the bay area and transported downstream. The variation of the Hurst exponent can lead to an area of the flow being affected by a contaminant cloud which is not picked up by the regular Brownian motion models. The purpose of this paper is to bring soil transport engineers a new angle on soil particle transport research in fluids. Using FBM1NC particle tracking model allows more flexibility in simulation of diffusion in soil contaminant spread in coastal bay or ocean surface.     

参数化地震子波估计模型定阶方法研究综述

参数化模型地震子波提取中模型阶数的确定对子波提取的精度至关重要,笔者针对因果地震子波MA、ARMA模型,总结了目前比较成熟的四类定阶方法：相关分析法、信息量准则法、线性代数法及基于优化算法的模型定阶方法,分析比较发现这些算法都存在不同程度的缺陷,只能在各自的适用范围内得到较好的定阶结果,针对算法在子波建模应用中的不足提出了改进建议;同时分析比较了基于高阶累积量的非因果系统模型定阶的各种方法,并对非因果地震子波模型定阶进行了展望.     

Liquefaction macrophenomena in the great Wenchuan earthquake

On May 12, 2008 at 14:28, a catastrophic magnitude M 8.0 earthquake struck the Sichuan Province of China.The epicenter was located at Wenchuan (31.00°N, 103.40°E). Liquefaction macrophenomena and corresponding destruction was observed throughout a vast area of 500 km long and 200 km wide following the earthquake. This paper illustrates the geographic distribution of the liquefaction and the relationship between liquefaction behavior and seismic intensity, and summarizes the liquefaction macrophenomena, including sandboils and waterspouts, ground subsidence, ground fissures etc., and relevant liquefaction features. A brief summary of the structural damage caused by liquefaction is presented and discussed. Based on comparisons with liquefaction phenomena observed in the 1976 Tangshan and 1975 Haicheng earthquakes, preliminary analyses were performed, which revealed some new features of liquefaction behavior and associated issues arising from this event. The site investigation indicated that the spatial non-uniformity of liquefaction distribution was obvious and most of the liquefied sites were located in regions of seismic intensity Ⅷ. However, liquefaction phenomena at ten different sites in regions of seismic intensity Ⅵ were also observed for the first time in China mainland. Sandboils and waterspouts ranged from centimeters to tens of meters, with most between 1 m to 3 m. Dramatically high water/sand ejections,e.g., more than 10 m, were observed at four different sites. The sand ejections included silty sand, fine sand, medium sand,course sand and gravel, but the ejected sand amount was less than that in the 1976 Tangshan earthquake. Possible liquefaction of natural gravel soils was observed for the first time in China mainland.     

基于ARMA模型非因果空间预测滤波(英文)

常规频域预测滤波方法是建立在自回归(autoregressive,AR)模型基础上的,这导致滤波过程中前后假设的不一致,即首先利用源噪声的假设计算误差剖面,却又将其作为可加噪声而从原始剖面中减去来得到有效信号。本文通过建立自回归-滑动平均(autoregres sive/moving-average,ARMA)模型,首先求解非因果预测误差滤波算子,然后利用自反褶积形式投影滤波过程估计可加噪声,进而达到去除随机噪声目的。此过程有效避免了基于AR模型产生的不一致性。在此基础上,将一维ARMA模型扩展到二维空间域,实现了基于二维ARMA模型频域非因果空间预测滤波在三维地震资料随机噪声衰减中的应用。模型试验与实际资料处理表明该方法在很好保留反射信息同时,压制随机噪声更加彻底,明显优于常规频域预测去噪方法。     

公路隧道局部塌方洞段的围岩稳定性评价

结合某公路隧道工程,以现场地质调查和监测资料为基础,在结构面发育且可能发生大规模塌方的围岩洞段,根据现场实际统计的优势结构面产状,采用关键块体理论、非连续变形方法（DDA）首先确定围岩主要的失稳破坏模式,然后运用连续介质分析程序FLAC,在DDA方法确定的围岩实际变形破坏塌方形态的基础上进行计算分析,评价围岩的整体稳定性及加固措施的有效性。分析表明,围岩破坏的主要模式为受结构面控制的局部块体的失稳,工程开挖后围岩不会发生大变形,这与实际情况基本一致,为安全施工提供了保证。采用非连续变形方法和连续介质计算相结合的分析方法对隧道工程塌方稳定性进行研究,取得了较好的效果,研究方法和经验可为类似工程参考并借鉴。     

高山峡谷地区大型地下洞室群非平稳人工地震动拟合

基于近场地震记录频谱分析,结合地震危险性分析得到的基岩场地反应谱,进行了高山峡谷地区大型地下洞室群的人工地震动加速度时程的合成。该方法使用三角级数迭加法,利用攀枝花地震动记录进行傅里叶变换,得到离散点的傅里叶幅值谱,选择两个卓越频率作为控制傅里叶谱的峰值控制点,通过反复迭代生成了初始人工地震波,根据高山峡谷地形进行了深度修正、基线校正和高频滤波,最终生成了可以用于FLAC和ABAQUS等数值分析软件的人工地震动加速度-时程。数值分析表明,此方法拟合的人工地震波可以较好地反映不同频谱对地下洞室群强震响应的影响。     

平原地区水文地质结构条件模拟及其应用:以华北平原为例

为了分析储水介质非均质性问题,采用了基于马尔柯夫地层序列分析的分布条件模拟方法,用转移概率矩阵的谱分析函数替代传统地质统计学中的变差函数,利用协同指示克里格方法和条件模拟建立结构模型。以华北平原为例,论述了模拟结构建立过程及其在水流模型中的应用。通过对模拟结果与实测数据的分析比较,认为模拟结果能够客观反映整个研究区的沉积规律,以岩性赋水文地质参数的方法更加客观准确地解决储水介质非均质性问题。分析了该方法的优、缺点,并提出了该方法深入研究的方向与思路。     

ROP-ENCS:一个城市游憩空间形态研究的类型化框架

从结构主义视野出发,在剖析城市游憩空间的概念内涵的基础上,分析构成城市游憩系统的主要结构要素和衡量城市游憩空间形态的主要指标,通过二元对立关系刻画并在此基础上相互叠加,进而构建一个ROP-ENCS的类型化框架.该框架对于进一步的理论建构以及城市游憩系统规划都具有一定的参考价值.     

多孔介质中非水相流体运移的数值模拟

针对多孔介质中水、气和非水相流体(NAPLs)的多相流动特点,建立了非水相流体(NAPLs)污染物迁移模型,分析了非水相流体在土壤非饱和区和地下水系统中的运移规律。通过有限元数值解对轻非水相流体和重非水相流体在土壤系统中的迁移过程进行模拟,得到了污染物的时空分布特征和污染范围。计算结果表明,数值模拟方法能够合理地描述非水相流体的运移过程和污染特征。土体渗透性和污染物残余饱和度是其重要影响因素。