大型液体火箭姿控与跷振大回路耦合动力学建模与分析

大型液体火箭结构模态的空间化分布特征导致结构振动、姿态运动和推进系统液路脉动存在相互耦合,进而影响传统姿控回路的稳定性. 针对大型液体火箭, 充分考虑姿态控制系统对箭体姿态动力学和弹性振动的影响, 以及箭体结构弹性振动与推进系统的耦合作用(跷振(POGO)), 建立了姿控与跷振大回路耦合模型. 该模型包含了推进系统、结构系统与姿控系统之间的耦合因素, 可进行姿控-结构-推进大回路耦合机理研究. 该模型具有非奇异的优点, 可以直接用于频域分析和时域仿真. 基于该模型研究了我国某型号液体捆绑火箭推进系统参数——泵增益和蓄压器能量值对姿态运动与结构振动稳定性的影响. 研究得出, 泵增益和蓄压器能量值的变化不仅导致了结构振动的不稳定, 而且也导致了姿态运动的发散. 因此, 对于大型液体捆绑火箭, 推进系统与姿控系统之间存在不可忽略的耦合作用, 在设计姿控系统时, 有必要考虑推进系统对姿控系统稳定性的影响.      

高超声速流动中液滴蒸发对流动参数的影响研究

采用带有非平衡相变模型的数值方法对高超声速流动中液滴蒸发的影响进行了研究,重点探讨了液滴蒸发对斜激波下游区域流场和劈面参数的影响。研究结果表明:液滴在激波后区域的蒸发存在弛豫现象,蒸发弛豫过程与激波强度相关,并影响模型表面气动参数的均匀性;当偏转角从3°增大至16°时,研究发现特别是偏转角为10°时,激波强度刚好能使液滴在整个激波下游区域完全蒸发,此时蒸发引起的流场参数不均性影响最大,与不考虑蒸发时相比,液滴质量含量、压强、静温减小幅度分别达到27.2g/kg、401Pa、54K,该三项参数变化量的比分别达到96.8%、4.8%、14.2%。     

旋转固体火箭上面级章动不稳定动力学机理分析

针对上面级旋转固体火箭发动机出现的章动不稳定现象,即锥形运动发散,通过对变质量陀螺运动方程干扰力矩项的研究,得出了发动机内部流动质量产生的附加侧向扰动力矩是诱发上面级固体火箭锥形运动发散的动力学根源,并得到了其章动不稳定的发生判据。基于此判据,给出了通过提高火箭自旋转速和横向转动惯量来解决上面级旋转固体火箭发动机章动不稳定现象的方法;并得出较高的气动静稳定设计和低密度的高空环境,是尾翼稳定旋转火箭弹发生“掉弹”现象的根源。     

煤油液滴直径对两相旋转爆轰发动机流场的影响

为探究煤油液滴不同初始直径对气液两相旋转爆轰发动机流场的影响,假设初始注入的煤油液滴具有均匀直径,考虑雾化破碎、蒸发等过程,建立了非定常两相爆轰的Eulerian-Lagrangian模型,进行了液态煤油/高温空气爆轰的非预混二维数值模拟。结果表明：在初始液滴直径为1～70μm的工况范围,燃烧室内均形成了单个稳定传播的旋转爆轰波;全局当量比为1时,爆轰波前的空气区域大于液滴煤油的蒸气区域,导致波前燃料空气混合不均匀,波前均存在富油区和贫油区,两相速度差导致分离出的空气形成低温条带;当煤油液滴的初始直径较小时,波前的反应物混合过程主要受蒸发的影响,爆轰波可稳定传播;当直径减小至1μm时,煤油液滴在入口处即蒸发,旋转爆轰波表现为气相传播的特性,爆轰波结构平整;当煤油液滴的初始直径较大时,波前的反应物混合过程主要受液滴破碎的影响;对于相同的燃料质量流量,在不同初始煤油液滴直径工况下,煤油液滴最大的停留时间均占爆轰波传播时间尺度的80%以上;爆轰波前燃料预蒸发为气相的占比越高,爆轰波的传播速度越高;初始液滴直径为10～70μm的工况范围内,爆轰波的速度随初始直径的增大先升高后降低。     

交变电场作用下单液滴蒸发的分子动力学模拟

利用分子动力学方法研究了正弦形式的交变电场对三维悬浮水滴在超临界氮气环境下蒸发特性的影响, 主要考虑了电场幅值和频率对液滴蒸发寿命和液滴瞬时蒸发速率的影响. 其中水滴由8000个水分子组成, 环境气体由27000个氮气分子组成. 首先利用分子动力学方法模拟计算了不同状态下水的物性参数以及亚临界条件下匀强电场对液滴蒸发特性的影响, 从而验证了分子模型和蒸发模型的正确性. 接着模拟了在不同幅值和频率的交变电场作用下水滴在氮气环境下的蒸发过程, 结果表明, 相比于无电场或匀强电场, 交变电场能够更显著地促进水滴的蒸发. 在频率一定时, 随着电场幅值的增大, 液滴的蒸发速率不断升高, 蒸发寿命不断下降, 且液滴的瞬时蒸发速率、液滴温度、水分子的排列结构等参数都会产生频率为所加电场二倍的振荡特性, 且电场幅值越大, 振荡幅值也越大. 而在电场幅值一定时, 随着频率的增大, 液滴蒸发寿命和速率并不是单调变化的, 而是在频率$f=5$GHz时, 分别达到一个极大值和极小值, 文中从液滴能量和分子排列结构两个方面解释了产生了这一现象的原因.      

稀薄气体动力学: 进展与应用

简要回顾了稀薄气体动力学的发展历程; 重点介绍了该领域最近二三十年的主要进展, 这突出表现在分子模拟方法(DSMC 方法、信息保存方法等) 的迅速发展与成功应用; 概述了航天工业、真空技术、微机电系统等尖端技术中的稀薄气流问题, 以及最近几届国际稀薄气体动力学会议的主题. 在此基础上指出了学科前沿问题, 以及几个与实际应用有关的重大问题, 如过渡流区高超声速三维非平衡流场的精细预测和实验验证、热层大气的时空演化规律与探测、以气体为介质的微机电系统设计与优化、真空环境下原子水平的材料制备工艺的定量设计.     

黏弹性液滴热毛细迁移的对流不稳定

液滴在温度分布不均的固壁面上产生的热毛细迁移广泛存在于微流控、喷墨印刷等应用中, 对其流动进行稳定性分析对液滴迁移的精准控制具有重要意义. 本文采用线性稳定性理论研究了附壁黏弹性液滴在热毛细迁移中的对流不稳定性, 得到了不同Prandtl数($Pr$)下的临界Marangoni数($Ma_{\rm c})$与弹性数的函数关系, 并分析了临界模态的流场和能量机制. 研究发现: 流体弹性激发了更多不稳定模态, 小$Pr$的临界模态为斜波和流向波, 而中高$Pr$的临界模态为斜波和展向稳态模态. 强弹性使得$Ma_{\rm c}$显著下降, 而弱弹性略微增强了流动稳定性. 在中$Pr$下, $Ma_{\rm c}$随$Pr$的增大而增大. 对于斜波模态, 扰动温度的振幅可存在于流场中间区域, 而其他两种模态的温度振幅只存在于自由表面上, 并且在高$Pr$下的流线分布几乎是对称的. 能量分析表明: 随着弹性数增大, 基本流做功由正变负; 在小$Pr$中, 扰动应力做功既可能耗散能量又可能提供能量; 在高$Pr$中, 基本流做功可忽略不计. 对于同向流向波, 扰动速度和扰动应力做功在垂直方向上均存在多次振荡. 将液滴迁移与热毛细液层进行对比发现, 由于基本流和边界条件的不同, 两者在临界模态和扰动流场中均存在较大差异.      

脊状表面减阻机理研究

针对脊状表面流场的特点,通过实验测量和数值模拟的方法对脊状表面微观流场进行了深入研究,获得了脊状表面湍流边界层的时均速度分布曲线、湍流度分布曲线和微观流场结构.为了得到脊状结构对壁面物性的影响,对脊状表面进行了疏水性测试,获得了液滴在脊状表面上的表观接触角,并通过水洞试验验证了脊状表面的减阻效果.研究表明,与光滑表面相比,脊状表面微观流场结构中存在"二次涡",近壁区的黏性底层厚度比平板的要厚得多,湍流度显著降低,且脊状表面表现出明显的疏水性.由此提出了基于壁面隔离效应、增大湍流阻尼效应和改变壁面物性效应的减阻机理.     

预处理方法在低速弹箭流场计算中的应用

基于Weiss-Smith预处理矩阵和全局截断预处理参数,采用有限体积方法对雷诺平均Navier-Stokes方程进行离散。对流项离散采用二阶线性重构和AUSM ＋-up格式,时间推进方法采用多重网格下的LU-SGS方法。结合M PI消息传递方法,建立了一套计算低速流动的并行数值方法。计算了低速椭球体的流场和气动力,压力系数和切应力系数计算结果与文献实验结果对比吻合度较好。生成了末敏弹的流场计算网格,对绕末敏弹流场进行了数值模拟。对多重网格下多进程的加速比和并行效率进行了测试,显示了程序良好的并行效率。计算的气动力结果与实验结果吻合。综合结果表明：本文的数值方法能够用于低速弹箭流场和气动力计算,为新型弹箭的设计和定型提供保证。     

离散型织构表面液滴的铺展及其接触线的力学特性分析

针对离散型织构表面上液滴的铺展过程,采用数值模拟和润湿性实验相结合的方法,引入织构润湿因子θ*,得到了不同类型的离散型织构对固体表面润湿性的影响,在此基础之上分析了液滴铺展过程中接触线的力学特性,以期从微观界面力学的角度解释微织构对液滴铺展过程的促进作用.研究表明:离散型织构增大了液滴铺展过程中的固-液接触面积,位于铺展前沿的液体分子部分浸润织构内部,导致液面曲率和液滴内部的拉普拉斯压力增大,相邻离散型织构间的液体获得了额外的驱动力和能量,铺展速度加快,平衡接触角减小;槽状离散型织构对表面润湿性的影响程度最大,液滴在其上铺展过程具有各向异性特性.另外数值仿真分析表明,接触线的钉扎效应与固体表面粗糙度的大小和微织构类型密切相关,表面粗糙度越大,钉扎效应越明显,其中槽状织构对接触线的钉扎作用还具有方向性.      

Heat transfer in droplet-laden turbulent channel flow with phase transition in the presence of a thin film of water

We present results of a numerical study of turbulent droplet-laden channel flow with phase transition. Previous studies of the same system did not take into account the presence of gravity. Here, we do so introducing a thin film of water at the bottom wall and permitting droplets to fall into and merge with it. We treat the carrier phase with the Eulerian approach. Each droplet is considered separately in the Lagrangian formulation, adopting the point–particle approximation. We maintain the film thickness constant by draining water from the bottom wall to compensate for (a) the droplets that fall onto the film and (b) evaporation/condensation. We also maintain on average the total mass of water in the channel by inserting new droplets at the top wall to compensate for the water that has been drained from the bottom wall. We analyze the behavior of the statistically averaged gas and droplet quantities focusing on the heat exchange between the two phases. We increase (a) the initial droplet diameter keeping the same initial droplet volume fraction and (b) the initial number of droplets in the channel keeping their diameter the same. In both parameter studies we find that droplets grow less than in the reference case. In case (a) this is explained by the larger velocity with which they travel to the bottom wall and in case (b) by the lower rate of condensation of vapor due to the presence of neighboring droplets.     

Evaluation of temperature gradients within combusting droplets in linear stream using two colors laser-induced fluorescence

The scope of this paper concerns the heating process of fuel droplets injected in a hot gaseous environment. The two colors laser-induced fluorescence technique allows measuring the temperature distribution within a droplet by scanning the droplet volume by a sufficiently small probe volume compared to the droplet volume itself. The temperature field is reconstructed using two approaches which have been compared. One is based on a geometrical optics model and the other is based on the 3D calculation of the internal excitation field within the droplet, using the generalized Lorenz-Mie theory. Experimental results have been obtained on a combusting monodisperse ethanol droplet stream (diameter around 200 m).     

Quadrature method of moments for modeling multi-component spray vaporization

This article puts forward the quadrature method of moments (QMoM) for modeling droplet composition during the spray vaporization process. This method is implemented for solving the Continuous Thermodynamic Model (CTM) of multi-component droplet vaporization, an advantageous alternative to the classical Discrete Component Model (DCM) when the droplet is formed of a great number of components. The CTM approach consists in modeling the droplet’s composition using a probability density function (PDF). This method was first tried out for vaporizing droplets by Hallett, who assumed a Gamma-function for the PDF. However, Harstadt et al. underlined some problems in the case of vapor condensation on the droplet surface, since the Gamma-PDF model presumes the PDF’s mathematical form. The QMoM which does not require this hypothesis is studied in this article, according to Lage’s research dealing with QMoM application to phase equilibria. The numerical features of QMoM are investigated in detail, and then the method is implemented for the difficult test case of vapor condensation. The results are analyzed to illustrate the application of QMoM to multi-component droplet vaporization modeling and to provide a better understanding of the QMoM main advantages and limitations.     

气相中双液滴正碰过程数值模拟研究

液滴碰撞现象普遍存在于动力装置燃烧室喷嘴的下游区域,影响燃料的雾化性能。为了揭示相同直径的双液滴中心碰撞机理,求解了轴对称坐标系下的N-S方程,采用VOF(Volume of Fluid)方法捕捉液滴碰撞过程中气液自由表面的演化规律。利用Qian等提供的实验结果对计算模型进行数值校验,验证了模型的准确性。在此基础上,研究了环境压强对液滴碰撞反弹后不同结果(分离和融合)的影响,分析了环境压强和Weber数对液滴碰撞分离的影响。结果表明,液滴在碰撞反弹后的状态(分离或融合)是由液滴间气膜压强与环境气动阻力共同作用的结果,环境压强对液滴碰撞分离过程基本没有影响;Weber数越大,碰撞过程中变形的幅度越大。     

On the calculation of wall temperatures in the post dryout heat transfer region

A non-equilibrium post dryout heat transfer model for calculating the wall temperature distribution in vertical upflows is presented in this study. The model is based upon the three path heat transfer formulation developed by MIT researchers (Laverty & Rohsenow 1964, Forslund & Rohsenow 1968, Hynek et al. 1969 and Plummer et al. 1974) that involves heat transfer from wall to vapor, from wall to droplets in contact with the wall and from vapor to liquid droplets in the vapor core. Downstream gradients for the bulk vapor temperature, vapor quality, droplet size and vapor velocities are identical to those used by Hynek et al. (1969) and Plummer et al. (1974). Conditions at the dryout location are calculated using a modified version of a technique developed by Hynek et al. (1969).A procedure for determining an average droplet diameter based on a size distribution is introduced. Migration of droplets through the boundary layer and droplet deposition flux are predicted with the model of Gani? & Rohsenow (1979). Heat transfer from the wall to the impinging liquid droplets is calculated with a correlation by Holman & McGinnis (1969). Mechanisms contributing to wall to droplet heat transfer are identified as (a) droplet-wall contact, (b) intensive droplet evaporation inside the boundary layer, and (c) destruction of the boundary layer due to droplet migration to, and rebound from, the hot surface. The significance of the average droplet size and size distribution is demonstrated through its control over the free stream evaporation and droplet deposition rates.Predicted uniform heat flux wall temperature profiles for water, nitrogen and freon 12 are in good agreement with the data of Era et al. (1966), Bennett et al. (1967), Forslund & Rohsenow (1968), Ling et al. (1971), Groeneveld (1972) and Janssen & Kervinen (1975).     

Vapor core turbulence in annular two-phase flow

 This paper reports a new technique to measure vapor turbulence in two-phase flows using hot-film anemometry. Continuous vapor turbulence measurements along with local void fraction, droplet frequency, droplet velocity and droplet diameter were measured in a thin, vertical duct. By first eliminating the portion of the output voltage signal resulting from the interaction of dispersed liquid droplets with the HFA sensor, the discrete voltage samples associated with the vapor phase were separately analyzed. The data revealed that, over the range of liquid droplet sizes and concentrations encountered, the presence of the droplet field acts to enhance vapor turbulence. In addition, there is evidence that vapor turbulence is significantly influenced by the wall-bounded liquid film. The present results are qualitatively consistent with the limited data available in the open literature. Received: 17 August 1998/Accepted: 12 April 1999     

An experimental method for the investigation of droplet oscillations in a gaseous medium

An experimental method for the investigation of droplet oscillations in a gaseous medium is presented. The droplets are produced using vibrating orifice droplet generators. Experiments are carried out with droplets in the diameter range from 91 m to 288 m using propanol-2, water, and n-hexadecane; the gaseous host medium is air. Oscillatory motions of the fundamental mode n = 2 and of the first higher order mode n = 3 occur during the disintegration of the liquid jet produced by the droplet generator. The periodical production of the droplets allows the observation and evaluation of each phase of the motion under quasi-steady conditions. Surface energies are determined from the droplet shapes on photos. The periods of the oscillations are found to be very close to the prediction of the linear theory.     

Effects of fuel properties on the burning characteristics of collision-merged alkane/water droplets

The combustion characteristics of freely falling droplets, individually generated by the merging of colliding alkane and water droplets, were experimentally investigated. The outcome of the collision droplets was firstly studied and then the subsequent burning processes such as the flame appearance, ignition and burning behaviors were recorded, through either visual observation or microphotography with the aid of stroboscopic lightening. If the merged droplets were exhibited in an insertive manner, while the water droplet inserted into the alkane droplet, these yield the burning behaviors prior to the end of flame were very much similar to that of pure alkane. The burning was ended with droplet extinction for lower-C alkane, and with either droplet “flash vaporization” or extinction for hexadecane. And if the merged droplets were in adhesive manner, for hexadecane with large water content, they either could not be ignited for the large merged droplets, or be ignited with a much prolonged ignition delay, followed by a soot-reducing flame and an ending of droplet extinction for the small merged droplets. “Homogeneous” explosion was not observed in any of the tests, and “heterogeneous” explosion, induced by trapped air bubbles, occasionally occurred for merged droplets with C-atom in alkane is higher than dodecane. And the sudden disappearance of droplet definitely decreased the burning time and thus enhanced the burning intensity. Besides, the fuel mass consumption rates were increased, even in the cases that having droplet extinction, because of the enlargement of the surface area due to the stuffing of water droplet.     

Spherical expansion of vapor during evaporation of a droplet

The problem of the evaporation of a spherical particle is solved by a numerical finnite-difference method for the stationary and nonstationary cases on the basis of the generalized Krook kinetic equation [1]. Evaporation into a vacuum and into a flooded space are considered taking into account the reduction in size and cooling of the droplet. The minimum mass outflow is determined for stationary evaporation into a vacuum at small Knudsen numbers. The results are compared with those of other authors for both the spherical and plane problems. Most previous studies have used different approximations which reduce either to linearizing the problem [2, 3] or to use of the Hertz-Knudsen equation [4]. The inaccurate procedure of matching free molecular and diffusive flows at some distance from the surface of the droplet [5] is completely unsuitable in the absence of a neutral gas. Equations for the rate of growth of a droplet in a slightly supercooled vapor were obtained in [6] from a solution of the ellipsoidal kinetic model by the method of (expansion of) moments.Translated from Izvestiya Akademii Nauk SSSR, Mekhanika Zhidkosti i Gaza, No. 2, pp. 97–102, March–April, 1976.