晃荡（英文）

本文列举了诸多工程领域中的液体共振运动现象,详细探讨了船舱中伴有剧烈流动的晃荡问题.描述了基于理论分析的非线性多模态方法,该方法便于波动稳定性分区、多分支解和物理稳定性的研究.强调了方形舱、垂向圆柱舱以及球形舱内伴有旋转和混沌(不规则波动)的三维流动的重要性.晃荡引起的砰击涉及到各种各样的内流条件,这些条件随液体深度与舱体长度之比而变化.针对棱柱状LNG舱,讨论了许多与流体力学和热力学参数、影响砰击载荷效应的水弹性以及模型实验缩尺比的物理现象.     

Sloshing

本文列举了诸多工程领域中的液体共振运动现象,详细探讨了船舱中伴有剧烈流动的晃荡问题.描述了基于理论分析的非线性多模态方法,该方法便于波动稳定性分区、多分支解和物理稳定性的研究.强调了方形舱、垂向圆柱舱以及球形舱内伴有旋转和混沌（不规则波动）的三维流动的重要性.晃荡引起的砰击涉及到各种各样的内流条件,这些条件随液体深度与舱体长度之比而变化.针对棱柱状LNG舱,讨论了许多与流体力学和热力学参数、影响砰击载荷效应的水弹性以及模型实验缩尺比的物理现象.     

MPS与GPU结合数值模拟LNG液舱晃荡

液舱晃荡是一种在外部激励作用下部分装载的液舱内液体的波动现象,它会对液舱结构强度和运输船舶稳性产生危害.移动粒子半隐式法(moving particle semi-implicit,MPS)是一种典型的无网格粒子类方法,可以有效地模拟剧烈的液舱晃荡问题.但MPS方法存在计算效率低的缺点,难以模拟大规模三维问题,而GPU并行加速技术已广泛应用于科学计算领域.因此,本文将MPS方法与GPU并行加速技术相结合,采用CUDA程序语言编写,自主开发了MPSGPU-SJTU求解器,对三维液化天然气(liquefiednatural gas, LNG)型液舱晃荡进行了数值模拟.通过三种不同粒子间距的数值模拟,验证了求解器的收敛性,其中最大计算粒子数达到了200多万.与其他研究结果相比,MPSGPU-SJTU求解器能够准确地预测壁面砰击压力,并且捕捉晃荡过程中自由面的大幅度变形和强非线性破碎现象.相比CPU求解器的计算时间,GPU并行加速技术可以大幅度地减小计算时长,提高MPS方法的计算效率.本文将LNG型液舱与方型液舱的晃荡进行对比,结果表明在高充液率下LNG型液舱可以有效地减小晃荡幅值和壁面砰击压力.但在中低充液率下,LNG型液舱则会加剧晃荡,自由面呈现明显的三维特征.本文还进一步研究了水和LNG两种不同介质的液舱晃荡现象,数值模拟结果表明二者的流场基本相似,砰击压力则正比于液体密度.      

MPS与GPU结合数值模拟LNG液舱晃荡

液舱晃荡是一种在外部激励作用下部分装载的液舱内液体的波动现象,它会对液舱结构强度和运输船舶稳性产生危害.移动粒子半隐式法(moving particle semi-implicit, MPS)是一种典型的无网格粒子类方法,可以有效地模拟剧烈的液舱晃荡问题.但MPS方法存在计算效率低的缺点,难以模拟大规模三维问题,而GPU并行加速技术已广泛应用于科学计算领域.因此,本文将MPS方法与GPU并行加速技术相结合,采用CUDA程序语言编写,自主开发了MPSGPU-SJTU求解器,对三维液化天然气(liquefied natural gas, LNG)型液舱晃荡进行了数值模拟.通过三种不同粒子间距的数值模拟,验证了求解器的收敛性,其中最大计算粒子数达到了200多万.与其他研究结果相比,MPSGPU-SJTU求解器能够准确地预测壁面砰击压力,并且捕捉晃荡过程中自由面的大幅度变形和强非线性破碎现象.相比CPU求解器的计算时间,GPU并行加速技术可以大幅度地减小计算时长,提高MPS方法的计算效率.本文将LNG型液舱与方型液舱的晃荡进行对比,结果表明在高充液率下LNG型液舱可以有效地减小晃荡幅值和壁面砰击压力.但在中低充液率下,LNG型液舱则会加剧晃荡,自由面呈现明显的三维特征.本文还进一步研究了水和LNG两种不同介质的液舱晃荡现象,数值模拟结果表明二者的流场基本相似,砰击压力则正比于液体密度.     

基于时域解耦算法的多液舱浮式结构物运动模拟

对于带有多个晃荡液舱的浮式结构物,浮体的运动、外场水动力以及各舱内的液体晃荡力会实时相互决定,发生复杂的耦合作用.为准确模拟多液舱浮式结构物的运动,本文引入一种有效的时域解耦算法.该方法以模态分解法为基础,通过对浮式结构物所受外域水动力和各液舱内非线性晃荡力进行模态分解,最终形成时域解耦运动方程,无需迭代求解过程即可显式计算浮式结构物的瞬时加速度.该方法可避免传统迭代求解方法在迭代次数、截断误差和收敛特性等方面的不足,减少解耦过程的计算耗时.本文进一步结合边界元数值方法,分别对单液舱浮式结构物和多液舱浮式结构物的工况开展数值模拟研究.通过与单液舱浮式结构物的实验结果对比,验证了本文时域解耦算法的有效性.本文详细分析了晃荡力对单液舱浮式结构物运动的影响,发现存在一个共振影响区间:当外场波浪频率在该区间之外时,可以在时域计算结果中观察到稳定的浮体运动;在比该区间更低频的波况下,液舱晃荡力与外场波浪力相位相反甚至可以相互抵消,此时晃荡液舱的存在可以减弱浮体运动;在比该区间更高频的波况下,液舱内晃荡力与外场波浪力可以具有相同相位,此时晃荡液舱的存在会加剧浮体的运动.本文进一步研究了四液舱浮式结构物在波浪中的纵荡、垂荡和纵摇运动情况,发现非线性液舱晃荡可对纵荡和纵摇运动产生影响,但对垂荡运动影响很小.     

基于时域解耦算法的多液舱浮式结构物运动模拟

对于带有多个晃荡液舱的浮式结构物, 浮体的运动、外场水动力以及各舱内的液体晃荡力会实时相互决定, 发生复杂的耦合作用. 为准确模拟多液舱浮式结构物的运动, 本文引入一种有效的时域解耦算法. 该方法以模态分解法为基础, 通过对浮式结构物所受外域水动力和各液舱内非线性晃荡力进行模态分解, 最终形成时域解耦运动方程, 无需迭代求解过程即可显式计算浮式结构物的瞬时加速度. 该方法可避免传统迭代求解方法在迭代次数、截断误差和收敛特性等方面的不足, 减少解耦过程的计算耗时. 本文进一步结合边界元数值方法, 分别对单液舱浮式结构物和多液舱浮式结构物的工况开展数值模拟研究. 通过与单液舱浮式结构物的实验结果对比, 验证了本文时域解耦算法的有效性. 本文详细分析了晃荡力对单液舱浮式结构物运动的影响, 发现存在一个共振影响区间: 当外场波浪频率在该区间之外时, 可以在时域计算结果中观察到稳定的浮体运动; 在比该区间更低频的波况下, 液舱晃荡力与外场波浪力相位相反甚至可以相互抵消, 此时晃荡液舱的存在可以减弱浮体运动; 在比该区间更高频的波况下, 液舱内晃荡力与外场波浪力可以具有相同相位, 此时晃荡液舱的存在会加剧浮体的运动. 本文进一步研究了四液舱浮式结构物在波浪中的纵荡、垂荡和纵摇运动情况, 发现非线性液舱晃荡可对纵荡和纵摇运动产生影响, 但对垂荡运动影响很小.      

基于SPH法的二维矩形液舱晃荡研究

液体晃荡是一种复杂的流体运动现象,自由液面的存在使得该现象具有很强的非线性和随机性。针对二维矩形液舱在不同振幅水平激励下的纵荡问题,应用SPH法对其进行了数值研究。首先计算了小振幅激励下的纵荡,计算结果分别与线性理论解、文献VOF法结果及文献SPH法结果作了比较分析,以验证所建数值模型的合理性;然后计算了液舱在大振幅水平激励下的纵荡,着重分析了不同振幅下液体晃荡的速度向量图、液面波动时程、压强波动时程、动量波动时程以及波动的频谱图,并将计算所得液面波动结果与小振幅激励下的液面波动结果作了比较。分析结果表明,在大振幅水平激励下,液面波动的波峰值较小振幅下的结果有较为明显的增大,而波谷值则无过大的变化,总体波动幅值比小振幅下的结果大;随着激励幅值的增大,液面波动幅值呈现明显增大的趋势,压强的整体波动幅值也呈增大趋势,动量波动的均值亦有明显增大;波动能量随着激励幅值的增大而增大并向第一阶频率区域集中。SPH法对处理液体大幅晃荡这种具有自由表面大变形的问题有十分优越的特性。     

晃荡液舱内水动冲击力的分段仿射模型

针对舱内晃荡液体与舱壁的相互作用,对舱内水动冲击力的等效力学模型进行了研究。基于混合系统理论,建立了强非线性液体晃荡的等效摆分段仿射模型,重点对矩形液舱的简化等效力学模型进行了分析。利用计算流体动力学软件Flow3D对矩形液舱内的强非线性液体晃荡进行了数值仿真。理论分析表明：分段仿射模型更符合刚性碰撞的假定,可以更有效地描述等效摆和舱壁碰撞时的速度跃变。仿真结果的对比表明：受到激励时,等效摆分段仿射模型所产生的力与Flow3D计算的结果比较接近,利用此模型可以恰当地描述强非线性液体晃荡。     

MPS方法数值模拟楔形体入水问题

入水问题是船舶海洋工程中典型的流动问题。当船舶在恶劣海况中航行或海洋平台遭遇恶劣天气时,结构物和水体之间往往会出现剧烈的砰击作用。砰击发生时,伴随着结构物湿表面的变化、自由液面的翻卷和破碎等强非线性现象。本文采用本课题组自主开发的基于移动粒子半隐式法MPS(Moving Particle Semi-Implicit Method)的求解器MLParticle-SJTU对二维楔形体入水问题进行了数值模拟。选取斜升角为30°的楔形体模拟其入水过程,研究了不同粒子布置方式对于计算结果如垂向水动力和自由面变化的影响,并与MLM砰击模型(Modified Logvinovich Model)的结果进行了比较,吻合较好。在此基础上,进一步研究了不同斜升角对计算结果(垂向水动力和自由面变化)的影响,其中流动分离前的垂向水动力与MLM结果相近,表明了MPS方法能有效地模拟入水问题。     

基于SPH法的返回舱入水载荷模拟分析

返回舱水上着落前期面临较大砰击,为研究过程中受到的砰击载荷,基于光滑粒子流体动力学(SPH)气-液两相流算法,首先通过模拟平板和楔形体两种算例模型的入水过程,并与相关文献的试验结果进行对比,验证算法的有效性。在此基础上,对返回舱的入水过程进行模拟。结果显示,两种算例模型的计算结果与相关文献试验结果吻合良好。返回舱入水速度和倾角对砰击有较大影响并且过程中存在二次砰击现象。砰击载荷随入水速度增大而增加。第一次砰击载荷峰值随倾角增大而减小,第二次砰击载荷峰值随倾角增大先增后减。结果表明,SPH气-液两相流算法能够较好地模拟返回舱入水过程。     

微重力条件下部分充液贮箱气液界面波动特性的数值模拟

微重力环境中部分充液贮箱内气液界面和气、液两相介质在残余重力或加速度干扰下的运动特征是先进空间流体管理技术的基础. 本文针对空间贮箱常用构型和实际尺寸, 基于邦德数相似准则设计了3个缩比模型, 数值模拟了原型贮箱和缩比模型中加速度变化引起的贮箱内气液两相流动及气液界面上界面波的传播. 数值模拟结果验证了原型和模型间的运动相似性, 发现在满足邦德数相似准则的前提下, 系统还近似满足韦伯数相似准则, 或等价地, 近似满足弗劳德数相似准则. 此外, 数值模拟结果也表明原型和模型间的运动存在细微偏差, 这主要源于黏性耗散作用的差异. 由韦伯数相似准则可知, 缩比加大, 贮箱尺寸减小, 重力突变后由表面张力释放出来的驱动力增强, 相同韦伯数下流动速度增大, 黏性耗散作用随之增强, 本文的数值模拟结果证实了该结论. 相关结果可以用于指导空间贮箱流体管理技术的地面模拟试验的方案设计等.      

Modelling of solids distribution in stirred tanks: analysis of simulation strategies and comparison with experimental data

The predictive capabilities of CFD techniques as applied to solid–liquid stirred vessels are investigated. The distribution of solid particles was simulated in three baffled stirred tanks agitated with single and multiple impellers. Suspensions of glass beads of different diameters and average concentration up to 6?vol. % in water were studied. The simulations of solid–liquid suspensions in the stirred vessels were performed by using fully predictive approaches. Eulerian multiphase models were adopted for modelling the solid–liquid flow, coupled with three different extensions of the standard k-? model to the case of multiphase flows. The simulated particle axial concentration profiles are compared with experimental data and critically discussed. The most successful simulation strategy and one possible implementation are described.     

圆柱形弹性贮液容器的自振特性分析

贮液容器,无论是由钢材制成的或是由钢筋混凝土建造的。在工业或工程上都有着广泛的应用,如石油化工工业中的储油罐,给水工程中的水即为典型例子。在核动力工程中的压水堆或快增残堆,所用的主容器都可看成为一种特殊的贮液容器。它与一般容器很大的区别在于除贮液外,在堆内还设置一些构件。这对结构分析计算速写为不小的困难。本文根据剪切悬壁梁理论且考虑液动压力效应,建立了弹性圆柱形贮液容器的自振方程及相关特征值矩阵,     

Selection of working fluids for electrohydrodynamic experiments in terrestrial conditions and microgravity

Electrohydrodyamic (EHD) heat transfer enhancement and flow control methods are becoming increasingly popular in engineering science and applications both in terrestrial and low gravity applications. The correct choice of the working fluid is essential for the design and performance of EHD hardware and can pose challenge because some working fluids with favorable EHD properties can be unstable or hazardous. In this paper key properties and criteria for the selection of working fluids for single-phase (liquid) as well as gas–liquid and vapor–liquid two-phase electrohydrodynamic experiments and applications are discussed. Key physical and electrical properties as well as environmental and safety issues are reviewed for the sample fluids PF-5052, FC-72, R141b, cyclohexane and pure water. Microgravity experiments impose additional demands on the selection of the working fluids. Some of these demands are addressed by contrasting bubble dimensions and shapes at detachment, estimated using a simple thermodynamic model, in terrestrial and microgravity conditions with and without electric fields. Data are obtained using a simplified analytical model and verified experimentally.     

A coupled BE–FE study for evaluation of seismically isolated cylindrical liquid storage tanks considering fluid–structure interaction

The effects of base-isolation on the seismic response of cylindrical vertical flexible liquid storage tanks subjected to horizontal seismic ground motion are presented in this paper. The whole system consists of two main parts: the base-isolation component, and the fluid–structure interaction subsystem. Dynamic analysis of liquid storage tank is achieved through the use of finite shell elements for the structure and internal boundary elements for the liquid region. The boundary element equations are employed to obtain an equivalent liquid mass matrix which is then coupled with the shell structure mass matrix, resulting in the coupled equations of motion. Finally, the coupled equations of motion are connected with the base-isolation system to observe the whole system behavior. A bilinear hysteretic element is used to illustrate the base-isolation system. The analysis is performed in the time domain in which, hydrodynamic interaction is taken into account. It is shown that the seismic response of the isolated tank could be significantly reduced compared with the fixed-foundation tanks. Parametric studies are carried out to study the effects of different system parameters on the effectiveness of the base-isolation system. These parameters are the tank geometry aspect ratio (height to radius), the flexibility of the isolation system, the liquid surface displacement variations, and the tank wall flexibility. It is observed that the seismic isolation is more effective in slender tanks in comparison with broad tanks. Furthermore, it is shown that the isolation efficiency is partially more significant in rigid tanks. It is also noticeable that flexible isolators considerably reduce the seismic response in comparison with stiff isolators. Despite the foregoing advantages, liquid surface displacement increases due to seismic isolation, especially in slender tanks.     

Local water slamming impact on sandwich composite hulls

The local water slamming refers to the impact of a part of a ship hull on stationary water for a short duration during which high local pressures occur on the hull. We simulate slamming impact of rigid and deformable hull bottom panels by using the coupled Lagrangian and Eulerian formulation included in the commercial software LS-DYNA. We use the Lagrangian formulation to describe plane-strain deformations of the hull panel and consider geometric nonlinearities. The Eulerian formulation is used to analyze deformations of the water. Deformations of the hull panel and of the water are coupled through the hydrodynamic pressure exerted by water on the hull, and the velocity of particles on the hull wetted surface affecting deformations of the water. The continuity of surface tractions and the inter-penetrability of water into the hull are satisfied by using a penalty method. The computer code is verified by showing that the computed pressure distributions for water slamming on rigid panels agree well with those reported in the literature. The pressure distributions computed for deformable panels are found to differ from those obtained by using a plate theory and Wagner's slamming impact theory. We have also delineated jet flows near the edges of the wetted hull, and studied delamination induced in a sandwich composite panel due to the hydroelastic pressure.     

Flow visualisation experiments on free draining of a rotating column of liquid using nets and tufts

The characteristics of the air-core vortex formed during free draining of liquids from tanks strongly depend on the size and shape of the drain ports used. In order to understand the role of the drain-port in influencing the air-core vortex, the flow currents induced while draining a rotating column of liquid with different drain-port diameters is assessed. The flow currents are monitored using nets and tufts. It is shown that for small drain ports, for which an air-core vortex is spontaneously formed, the rotational flow currents are concentrated along the central core. With larger drain ports, the rotational flow currents are smaller and are diffused over larger radii leading to weak air-core vortex.