连铸过程中湍流传输和两相区凝固的数值模拟

综述了在连铸数值模拟过程中，为评估流体流动对铸坯热量传递和凝固过程影响所开发的有效导热模型和耦合模型。根据文献和用商业CFD软件计算的结果分析了在连铸湍流传输和两相区凝固的数值模拟过程中影响模型预测精确性的关键因素，包括有效导热模型中有效导热系数的确定、数值模拟模型的选取、湍流模型的应用、连铸坯凝固过程中湍流向层流的过渡、两相区流动中多孔介质的渗透率取值、固相分率与温度和成分的关系等，并给出选取合适模型参数的原则。     

多元合金两相区凝固过程的数值仿真研究

基于微观枝晶计算域内的溶质质量守恒关系，推导出适用于枝晶凝固方式的二元合金微观偏析半解析数学模型，并根据简单加合原理对模型进行了多元化扩展，应用扩展后的微观偏析模型针对浇铸钢种（视为Fe—C多元合金）的两相区凝固过程进行了数值仿真计算，通过仿真获得了连铸凝固传热计算中所需要的钢种的非平衡凝固路径及实际固相线温度，研究结果表明，本文所建立的多元合金两相区凝固计算微观偏析数学模型及其仿真程序具有较好的合理性和广泛的适用性，可以方便地在连铸过程静态及动态仿真计算中加以耦合运用。     

板坯连铸结晶器中三维凝固壳厚度分布的数值模拟及实验验证

建立了连铸结晶器中三维凝固壳厚度分布的计算模型和计算方法，提出了非耦合计算时流动计算域与凝固传热计算域的衔接问题，以及铸坯表面换热系数确定的方法。针对２５０ｍｍ×１３００ｍｍ板坯连铸实际工况，用所建立的模型放处理方法数值模拟了其结晶器中的三维温度场和三维凝固壳厚度分布。同时用实测的凝固壳厚度分布数据验证了本计算听模型、边界条件和计算方法。本采用的方法可满足工程精度、并较简便实用。     

薄板坯连铸二冷传热数学模型的研究

依据凝固传热理论建立数学模型，探讨了网格尺寸和时间步长对数值计算结果的影响。编制传热计算程序，分析不同拉速下铸坯的凝固状况，运用冶金准则对铸机原有水表进行了合理评价。     

气淬渣滴冷却过程数值模拟研究

液态钢渣气淬后将经历复杂的冷却凝固过程。采用Fluent软件对渣滴冷却过程动力学进行了数值模拟,并与数值计算结果相比较。研究结果表明:渣滴冷却经历液相冷却、潜热释放和固相冷却三个阶段;在没有考虑过冷度的情况下,Fluent软件模拟出液态冷却耗时较短,而潜热释放和固相冷却耗时较长;Fluent数值模拟能更形象地展示渣滴凝固形壳过程,数值计算更能准确地计算液相冷却、潜热释放和固相冷却等各阶段所消耗的时间;综合来看,Fluent数值模拟和数值计算两种方法均可准确预测出渣滴由1 723 K冷却至1 073 K所需要的时间,从而为液态钢渣气淬工艺提供理论基础。     

连铸喷淋结晶器的传热数学模型

针对铸坯在结晶器内的凝固特性，建立了喷淋冷却结晶顺的传热数学模型。并数值计算了普通循环水冷结晶器与喷淋冷却结晶器的冶金参数，讨论分析，比较了两种冷却方式的冶金效果。     

液芯钢锭数学模型

本文建立了描述钢锭冷凝、加热，轧制全过程的数学模型。其特点是建立了凝固率变化和凝固前沿推移的连续模型；考虑了相邻钢锭对辐射传热的遮蔽作用；研究了钢液中的结晶沉降过程，提出并定义了焓平均温度，且将它作为钢锭热状态的判别标志；在加热模型的计算中．由给定热负荷计算温度和由给定温度计算热负荷均获得了收敛的结果。在此基础上，还获得了计算焓平均温度和凝固率的回归公式，作出了多种传搁时间图表，得出了确定最佳人炉焓平均温度和最佳侍搁时间的方法。根据本模型能为生产实际制定液芯轧制和加热工艺规程。     

AH36钢两相区非平衡凝固路径的数值预测

应用多元合金两相区凝固微观偏析数学模型，对AH36钢在具体冷却速率条件下的非平衡凝固路径进行了数值计算，获得了相应的局部固相分数与局部温度之间的变化关系，为该钢种对应的连铸浇铸过程仿真及其它凝固过程仿真提供了必要而准确的耦合数据，具有重要的理论意义和广泛的适用性。     

关于钢绽凝固、模壁挠曲和锭—模气隙形成的实验和数理研究

测量30吋×35吋×101吋上大下小钢锭模内凝固的10吨钢锭凝固过程中模壁温室、模壁挠曲及锭——模间的气隙厚度。运用测出的模子内表面温度(可高达1000℃)计算了模锭界面随着时间变化的热流和导热系数。这些导热系数旋即输入双维传热凝固模型。运用此模型研究钢锭表面质量问题,确定镇静时间,研究改善钢锭内部质量并且对不同的保温帽操作及钢锭模设计进行比较。 将凝固模型运用于不同宽度、厚度的钢锭而得到了凝固关系式,用此关系式计算出的各种尺寸的钢锭的完全凝固时间与文献资料相符。此关系式同时考虑钢锭的宽度和厚度,而不像标准的凝固公式那样,只根据凝固的厚度进行计算。 测量表明:浇注15—20分钟以后,钢锭的宽面与窄面均已形成稳定的锭—模间气隙。气隙的形成非常明显地随模子内壁温度的变化而波动。最大的模壁挠曲量和气隙厚度分别为1.0和0.75厘米,而且是出现在浇注以后的3—4小时。测出的模子宽面和窄面的模壁挠曲量与根据为设计最佳模型而建立的钢锭模双维弹塑性热应力数学模型计算出的数值是一样的。     

水平轴离心铸造套筒凝固过程温度场数值计算

本文通过分析水平轴离心铸造长／径比小于５的圆筒型铸件的凝固过程，建立了表征铸件凝固过程的二维传热模型，并采用交替隐式差分的方法对离心铸造４５＾＃碳钢单材质套筒凝固过程温度场进行了数值计算。结果表明：套筒凝固过程表现为沿径向从靠近模具有铸件外表面层向内表面层、轴向从靠近端盖的铸件层向套筒长度中心部位顺序凝固的特点。模具初温对铸件的凝固速度也有重要的影响。提出减少套筒内表面形成的型腔对外界冷空气的卷吸     

矩形坯连铸机凝固末端位置的模拟计算

根据连铸矩形坯凝固传热特点，运用直接差分法建立了二维非稳态矩形坯凝固传热模型，对首钢矩形坯进行了凝固末端位置的模拟计算，得出了3种断面的液芯长度，找了搅拌器的合适位．置。     

矩形坯连铸凝固传热的数学模型

根据连铸矩形坯凝固传热特点，在上海浦东钢铁有限公司１号连铸机二冷系统改造中，动用直接差分法建立了二维非稳态矩形坯凝固传热数学模型，已应用于连铸凝固过程的模拟计算，在分析拉速、浇注温度等在数对钢水凝固过程的影响后，为提高拉速找到了理论依据。     

气隙对连铸坯凝固影响的有限元数值模拟

以连铸坯凝固传热有限元数学模型研究了铸坯角部形状和气隙对坯壳凝固行为的影响研究结果表明：采用有限元法离散铸坯圆角能更合理地反映铸坯角部的几何和换热条件，从而可以更准确地研究角部换热条件对铸坯凝固行为的影响。角部气隙显著地降低了坯壳表面的换热，使铸坯偏角区成为热节区。此热节区是铸坯凹陷、皮下裂纹等缺陷和漏钢事故发生的诱因。     

Modeling of multidimensional solidification of an alloy

The solidification of an alloy and pure metal is distinct because of the morphological differences between the respective interfaces. For certain conditions of imposed heat extraction, this dis-tinction can lead to large differences in the times required for complete solidification of the alloy over the pure metal. These conditions are examined in this paper. To do this, a powerful numerical technique to model alloy solidification is introduced which allows for the precise integration of the Scheil equation. The model takes into account the nonlinearity of the fraction liquid with temperature as well as the interface nonlinearity in the heat flow. As a test for the model, accurate temperature profiles from a previously published experiment are numerically simulated. The numerical results are noted to closely match experimental values, and as a con-sequence, contact heat transfer values are tabulated. One- and two-dimensional solidification of aluminum-copper alloys are now simulated for different cooling conditions (i.e., varying Biot numbers). The results obtained indicate the essential differences for alloy solidification at low and high Biot numbers and highlight the importance of properly accounting for the mushy zone.     

Numerical simulation of the infiltration of fibrous preforms by a pure metal

The numerical simulation of the metal-matrix composites (MMCs) elaboration by an injection process is presented. The equations governing the heat and mass transfers through a porous medium are applied to the metal injection process. The bidimensional numerical model is described based on a finite volume formulation. It is shown that two types of metal solidification appear during the injection: (1) a frontal solidification related to the heat balance between the fibrous preform and the metal and (2) a lateral solidification related to the heat losses toward the mold walls. Numerical tests validate the numerical model in the case of theoretical mono-dimensional geometries when analytical solutions exist. The model is then applied to bidimensional geometries. The preform is progressively closed until the channel is obstructed so that injection is achieved to the “impregnation depth.≓ A systematic study determines the influence on the impregnation depth of different parameters, such as thermal properties of fibers and mold walls, injection flow rate, and preform geometry. Finally, the results of the experimental injection of biphenyl resin through the SAFFIL* preform are discussed and compared with the numerical results.     

Numerical simulation of the infiltration of fibrous preforms by a pure metal

The numerical simulation of the metal-matrix composites (MMCs) elaboration by an injection process is presented. The equations governing the heat and mass transfers through a porous medium are applied to the metal injection process. The bidimensional numerical model is described based on a finite volume formulation. It is shown that two types of metal solidification appear during the injection: (1) a frontal solidification related to the heat balance between the fibrous preform and the metal and (2) a lateral solidification related to the heat losses toward the mold walls. Numerical tests validate the numerical model in the case of theoretical mono-dimensional geometries when analytical solutions exist. The model is then applied to bidimensional geometries. The preform is progressively closed until the channel is obstructed so that injection is achieved to the “impregnation depth.” A systematic study determines the influence on the impregnation depth of different parameters, such as thermal properties of fibers and mold walls, injection flow rate, and preform geometry. Finally, the results of the experimental injection of biphenyl resin through the SAFFIL* preform are discussed and compared with the numerical results.     

Numerical Analysis of Coupled Fluid Flow,Heat Transfer and Macroscopic Solidification in the Thin Slab Funnel Shape Mold with a New Type EMBr

As an expansion of our former work, a series of numerical simulations of fluid flow, heat transfer, and macrosolidification under the control of both the conventional and the new type electromagnetic brake (EMBr) were carried out. In the present work, the physical models and parameters remain the same as those measures in our former research, and the solution processes of the flow field and electromagnetic field are similar. The specific heat method is used to treat the latent heat of solidification, and the Darcy’s Law-type porous media treatment is used to account for the effect of phase change on convection. User-defined functions (UDFs) are compiled for the following two aspects: One is the Darcy source term, and the other is the thermal conductivity, which varies in mushy zones. From the comparisons of two kinds of EMBr produced by two type magnets separately, the new type EMBr gives its superiorities of the effects on the flow field, heat transfer, and solidification with lower cost of electrical energy.     

钢锭宏观偏析数值预测

本文在改进准三维凝固传熟数值模拟技术的基础上,建立了新的钢锭宏观偏析预测数模。实际测试表明:数模可准确预测钢锭的逆V偏析分布、沉积锥区域及碳的宏观偏析以及钢锭缩孔形状。     

A numerical and experimental study of the solidification rate in a twin-belt caster

A numerical and experimental study was carried out to investigate the solidification process in a twin-belt (Hazelett) caster. The numerical model considers a generalized energy equation that is valid for the solid, liquid, and mushy zones in the cast. Ak-ε turbulence model is used to calculate the turbulent viscosity in the melt pool. The process variables considered are the belt speed, strip thickness, nozzle width, and heat removal rates at the belt-cast interface. From the computed flow and temperature fields, the local cooling rates in the cast and trajectories of inclusions were computed. The cooling rate calculations were used to predict the dendrite arm spacing in the cast. The inclusion trajectories agree with earlier findings on the distribution of inclusion particles for near horizontally cast surfaces. This article also reports the results of an experimental study of the measurement of heat flux values at the belt-cast interface during the solidification of steel and aluminum on a water-cooled surface. High heat fluxes encountered during the solidification process warranted the use of a custom-made heat flux gage. The heat flux data for the belt surface were used as a boundary condition for the numerical model. Objectives of the measurements also included obtaining an estimate of the heat-transfer coefficient distribution at the water-cooled side of the caster belt. Y.G. KIM, formerly Graduate Student, Materials Engineering Department, Drexel University.     

Sensitivity of Thermophysical Material Properties on Solidification Simulation of Al-Si Binary Alloys

The challenges in the numerical simulation of the solidification of binary alloys are not only in the complexity of the algorithms themselves, but also in the validity of the data used to define the material properties of the various phases to obtain a valid simulation. The effect of material properties on the numerical simulations was investigated in the present study wherein the Al-3 wt pct Si hypoeutectic binary alloy was solidified such that the solidification front traveled against the gravity vector (upward solidification). Numerical simulations were carried out with a new algorithm that was developed to include the effect of undercooling of the liquid temperature prior to the solidification event. The effect of specific heat of solid, density of solid, solute diffusivity coefficient of liquid, and thermal conductivity of solid on transient temperature distribution and solidification start time at mushy zone/liquid interface was investigated. It was found that specific heat and thermal conductivity of the solid could not be assumed as constants, whereas most properties in the liquid phase could be assumed as constants for the temperature range used in the study and the experiments used for validation (low initial melt superheat temperature). These properties were enumerated and quantified. The results of the numerical simulations using the optimum set of material properties were validated by experiments.