【作者】 肖毅；

【导师】 邵学军；

【作者基本信息】 清华大学 ， 水利工程， 2013， 博士

【摘要】 冲积河道的自我调整是多因素相互作用的复杂响应过程，河型成因机理及转化判别研究是河床演变学及河流动力学的基本问题之一。目前，河型成因理论及转化判别准则研究中得到的众多成果，尚有待于采用数理逻辑建立统一的体系和定量表达式。本文选用基于正交曲线坐标系的平面二维水沙动力学模型作为研究河型转化控制因素的手段，对基于水动力学模型、泥沙输运模型以及河道崩岸模块的平面二维数值模型进行如下改进。1)考虑了植被对流动阻力的影响：在前人研究成果的基础上，引入植被压力源项，修正水流动量守恒方程；2)改进了泥沙输运模型：进一步考虑泥沙分选、河床粗化、弯道二次流及河床结构的影响，并根据近底床沙质量守恒方程提出简化床沙级配调整计算方法；3)完善了河道崩岸模块：在已有粘性土崩岸的基础上，针对河湾形态对崩岸的影响提出新的非粘性土崩岸模拟方法，并增加简化混合土体崩岸模块，使其能模拟边滩形成，提高了计算效率。利用本文改进的二维水沙数值模型，分别模拟了Yen弯道水槽泥沙冲淤试验、Seal水槽下游细化试验及Friedkin室内弯曲小河塑造试验。模拟结果与试验观测吻合良好，表明可以采用该模型作为定量研究河型转化影响因素的手段。以该模型为基础，通过数值试验，分别模拟地球自转的柯氏力、水力比降、流量以及边岸抗冲强度对概化河道演变过程的影响；定量得到不同控制因素对概化河道形成及河型转化的影响程度。对模拟结果进行归纳，总结出了影响河型转化的主要控制因素。根据所得到的河型转化主控因素，选取适当的控制变量与状态参量，基于尖点突变模式，推求得到了河道状态的平衡方程式，并绘制出了3维坐标下的平衡曲面图。依据该平衡方程，选取相应的临界状态参数，对控制参平面进行2维投影，得到了河型判别准则的表达式。从河道稳定性的控制变量与状态参数出发，推导得到河流稳定性判别指标的表达式。采用100多条中小型天然河流、室内小河试验以及概化河道的数值模拟进行验证，结果表明：基于尖点突变模式所建立的河型判别准则及河道状态判别指标，可判定河段所处河型及状态，并对其调整方向作出预测。研究成果对于河道整治工程有一定的参考价值。更多还原

【Abstract】 Adjustments to alluvial channels involve a large number of variables whoseinterdependence is not always clear because the roles of a single variable cannot easilybe isolated. It is still a matter of debate and continues to attract close attention fromhydrologists, geomorphologists, and engineers. Based on different criteria, variousstudies have provided a great deal of information on the response of channelmorphology to controlling variables and classifications of natural rivers, although noneis particularly quantitative. With the rapid developments of numerical and mathematicsmethods in fluid mechanics, multiple-mathematics models have become important toolsfor hydraulic engineering and river dynamics research. The general purpose of thisstudy is to find out the main control factors of channel pattern transformation by usingan improved2D numerical model; and then based on the cusp catastrophic theory,establish a threshold to classify the channel patterns, to describe the stability of riverchannels, and to predict the transformation of channel patterns by selecting suitableparameters derived from the main control factors.This study adopts a hydrodynamic model to study the control factors on thetransformation of channel patterns. An imrpoved2-D depth-averaged model forhydrodynamic, sediment transport and river morphological adjustment is presented inthis paper, which is based on orthogonal curvilinear grid system. The hydrodynamicsubmodel takes into account the impact of vegetation with a vegetation stress term inthe flow momentum conservation equation. The sediment transport submodel considersnon-uniform sediment, bed surface armoring, impact of secondary flow on the directionof bed-load transport, and transverse slope of river bed; the grain size distribution issimulated according to the sediment mass conversation equation for bed surface. Basedon the original one for cohesive bank erosion, the bank failure submodel adds anon-cohesive bank erosion model considering the influence of river bend and a simplebank erosion model with mixture sediment, so that it is capable to simulate theevolution of sand bars with different bank material.The extended2D numerical model is applied to the experiment on downstream finingby Seal, a180°bend with a constant radius under unsteady flow conditions, and toFriedkin’s laboratory meander channels. The results are in acceptable agreement with measurements, confirming the two dimensional model’s potential in predicting theformation of river meandering and improving understanding of patterning processes.After that, some numerical experiments are performed to disscuss the influence ofCoriolis force on river meandering, simulate different channel patterns with variousdifferent factors by the improved2D numerical model, and the dominant control factorson the transformation of alluvial channels are identified.According to the results of numerical experiments, equations of equilibrium state andthe transformation of channel patterns are established based on the model ofcusp-catastrophe surface by selecting suitable parameters. The stability of channelpatterns can be identified by such a model in a direct way with quantified index, whichis a cusp catastrophe surface in a translated three dimensional coordinate, and the2Dprojection of the cusp catastrophe surface can be used to classify alluvial channelpatterns, and discriminant functions are obtained from the model to distinguish thechannel patterns. Predictions based on this model are consistent with field observationsinvolving about100natural rivers of small or medium sizes. The results indicate thatthis method may be applied to study the regime of natural rivers and to assist decisionmaking in river engineering.更多还原