非结构化网格下近岸波生流数值模拟

波浪破碎产生的近岸流是近岸海域关键的水动力因素之一。基于近岸波浪的椭圆型缓坡方程和二维近岸波生流方程,建立了非结构化网格下近岸波浪破碎形成的近岸流数值模型。数值模型中,在空间上采用有限体积法进行数值离散,在时间上采用欧拉向前格式数值离散。数值计算结果表明,该数值模型可以有效地模拟近岸波浪破碎产生的近岸流。     

Sub-seasonal variability of Luzon Strait Transport in a high resolution global model

The Luzon Strait is the main impact pathway of the Kuroshio on the circulation in South China Sea (SCS). Based on the analysis of the 1997–2007 altimeter data and 2005–2006 output data from a high resolution global HYCOM model, the total Luzon Strait Transport (LST) has remarkable subseasonal oscillations with a typical period of 90 to 120 days, and an average value of 1.9 Sv into SCS. Further spectrum analysis shows that the temporal variability of the LST at different depth is remarkable different. In the upper layer (0–300 m), westward inflow has significant seasonal and subseasonal variability. In the bottom layer (below 1 200 m), eastward outflow exhibits remarkable seasonal variability, while subseasonal variability is also clear. In the intermediate layer, the westward inflow is slightly bigger than the eastward outflow, and both of them have obvious seasonal and subseasonal variability. Because the seasonal variation of westward inflow and eastward outflow is opposite, the total transport of intermediate layer exhibits significant 50–150 days variation, without obvious seasonal signals. The westward Rossby waves with a period of 90 to 120 days in the Western Pacific have very clear correlationship with the Luzon Strait Transport, this indicates that the interaction between these westward Rossby waves and Kuroshio might be the possible mechanism of the subseasonal variation of the LST.     

波浪能利用的发展与前景

文章首先介绍了波浪能的定义、成因和我国波浪能资源的分布情况。在此基础上,指出波浪能的主要利用方式是发电,并分析了国内外利用波浪能发电的发展进程。从波浪能的发电原理、转换装置及技术上的突破等方面,论述了波浪能发电的研究开发情况。最后,指出波浪能利用中存在的一些问题并对其发展前景进行了展望。     

预修正快速傅里叶变换方法在波浪绕射和辐射问题中应用

在波浪对海上建筑物绕射和辐射问题的计算中,面元法被广泛使用,但由于传统面元法的存储量和计算量均为未知量的平方量级,很难满足大范围多未知量问题的计算需要。采用预修正快速傅里叶变换方法(pFFT方法),使计算量与存储量都降低到未知量的线性量级。以淹没圆球与漂浮圆柱两个典型算例为基础,通过不同未知量时pFFT方法与传统面元法的计算量与存储量的对比,以及pFFT方法自身各步骤计算时间的对比,研究了不同网格方案的选取对pFFT方法计算量和存储量的影响,推荐根据未知量个数采用计算时间最小化原则选取pFFT网格参数。     

新型沉箱防波堤兼作岸式OWC波能装置的设计及稳定性研究

防波堤与岸式OWC波能装置一直是被独立研究,从未产生交叉,但二者设施布置的条件显然具有关联性。本文根据二者关联的特点,设计出沉箱防波堤兼作岸式OWC波能装置,并通过物理模型试验验证了其可行性。     

城市地下空间资源探测:面波与初至波层析成像联合探测济南泉域近地表速度结构

济南地区地质结构复杂,地下空间开发尚处于起步阶段,未来轨道交通规划和地下空间开发面临复杂的地质问题,如何保护泉水,避免地下空间的不当开发影响泉水断流,以及查明济南主城区地下地质结构与地质资源,是急需解决的关键问题.本文利用多道面波分析方法采集的数据,综合初至波层析成像与面波频散分析技术方法,建立了纵波速度、横波速度和纵横波速度比值模型,开展趵突泉敏感区东部边界地质结构探测工作,获得了测区浅层0～80 m地层结构特征,推测在泉水敏感区边界存在东西向径流通道,为济南中心轨道交通建设和其他地下空间资源开发利用提供了新的探测技术方法以及有力支撑.     

面向盐下成像的基于波前构建法的高斯波束偏移

高斯波束偏移是目前工业上应用最广泛的偏移成像技术之一.但是当处理盐下成像问题时,由于盐体上界面起伏剧烈且盐体与围岩速度差异较大,只有小角度入射的射线才能穿过盐体,从而导致了盐下区域照明较差.通常,采用加密初始射线的方式来缓解这个问题,但这会导致计算时间的剧烈增加.由此,本文提出了基于波前构建法的高斯波束偏移,以保持波前传播过程中射线密度近似恒定.相较于加密初始射线的方式,本文提出的方法是自适应的,对于提高射线的覆盖范围更加有效.此外,本文采用高精度的波前初始化射线追踪技术作为波前构建的方法.Sigsbee2A模型下的应用实例表明,相较于常规高斯波束偏移加密初始射线的方式,本文提出的方法不仅能够获得更好的盐下照明效果,且对计算效率的影响也更小.     

滇中地区的岩石圈结构及其动力学意义

虽然地球化学和地质学的研究确信峨眉山大火成岩省的形成与二叠纪地幔柱活动有关,但从二叠纪至今,峨眉山大火成岩省所在的华南板块至少向北漂移了3000 km,以致大火成岩省与地幔柱的空间位置不再对应．本研究使用川滇地区75个宽频固定台记录的远震事件计算P、S波接收函数,并利用参考地球模型把单个接收函数从时间域变换到深度域,最后从叠加道上读取地壳厚度和岩石圈厚度．结果表明,Moho面深度从云南南部的33 km向北逐渐增厚,在青藏高原东部可达约66km,滇中地区地壳厚度变化剧烈,在攀枝花附近明显增厚．在金沙江—红河断裂西侧的印支块体下方,岩石圈厚度仅为70～80km,但在滇中地区从约70 km变化约到120 km,最薄处发生于攀枝花附近,仅约为70 km．另外,滇中地区部分台站的S波接收函数显示Moho面的Sp转换相一致性较差,而P波接收函数则表明在Moho面附近存在一个正极性的转换界面．综合已有的研究结果,我们推测Moho面附近这一正极性界面是地幔柱岩浆底侵的结果．     

基于CUDA的高阶旋转交错网格有限差分法的弹性波正演模拟

应用高阶旋转交错网格的有限差分法,求解了基于一阶的速度-应力弹性波方程,推导了时间二阶精度空间2M阶精度的高阶有限差分离散格式.为了提高弹性波正演模拟效率,论文探讨了常规的CPU串行计算、基于GPU平台的节点并行计算、基于GPU平台的分量、节点同时并行的3种正演模拟计算模式,给出了不同算法的流程示意图.通过对比3种算法的耗时及后两种算法的加速比,可清晰判定,基于GPU平台的节点并行算法较CPU平台的串行算法有明显加速,最大加速比可达到60倍,而基于分量、节点同时并行算法的最大加速比最大可到180左右,加速效果更为明显.最后应用Marmousi模型的15000次时间步迭代的正演计算,基于分量、节点GPU并行算法大约是节点并行计算耗时的1/3,说明了该并行设计能有效提高弹性波正演模拟的效率,节约计算耗时.     

Loss of energy dissipation capacity from the deadzone in linear and nonlinear viscous damping devices

In a viscous damping device under cyclic loading, after the piston reaches a peak stroke, the reserve movement that follows may sometimes experience a short period of delayed or significantly reduced device force output. A similar delay or reduced device force output may also occur at the damper’s initial stroke as it moves away from its neutral position. This phenomenon is referred to as the effect of “deadzone”. The deadzone can cause a loss of energy dissipation capacity and less efficient vibration control. It is prominent in small amplitude vibrations. Although there are many potential causes of deadzone such as environmental factors, construction, material aging, and manufacture quality, in this paper, its general effect in linear and nonlinear viscous damping devices is analyzed. Based on classical dynamics and damping theory, a simple model is developed to capture the effect of deadzone in terms of the loss of energy dissipation capacity. The model provides several methods to estimate the loss of energy dissipation within the deadzone in linear and sublinear viscous fluid dampers. An empirical equation of loss of energy dissipation capacity versus deadzone size is formulated, and the equivalent reduction of effective damping in SDOF systems has been obtained. A laboratory experimental evaluation is carried out to verify the effect of deadzone and its numerical approximation. Based on the analysis, a modification is suggested to the corresponding formulas in FEMA 356 for calculation of equivalent damping if a deadzone is to be considered.