Abstract

The paper outlines the process of developing a computer model to predict particle distribution during centrifugal casting using the ANSYS FLUENT 16.0 software module. There are several mathematical models developed to predict the volume distribution of particles. The majority of the models is based on a steady state assumption: these are models describing criteria for repelling particles by a propagating crystallization front and the models calculating critical velocities of absorbing particles by growing dendritic crystals. Some models are aimed at describing the dynamic state of the system or at determining the criterion of capturing non-metallic inclusions by hardening front during centrifugal metal casting. The description of the process for deriving the model, the schematic diagram and geometry, is given. The model preprocessor considers the following phenomena: the two-phase flow, the energy equation, the laminar flow, the introduction of discrete phases (hardening particles), and the melting/crystallizing phenomena. The model considers the interaction between two liquid phases: air and steel melt. The interphase interaction is described using the surface tension equation. Steel grade 12Х18Н10T is used as the base metal. The particles of tungsten carbide, boron carbide and yttrium oxide are introduced as dispersion. The physics and chemical parameters of these substances are considered by the simulator. The process of simulating particle distribution during centrifugal casting using the computational cluster is described using Skif-Ural, which is one of the top 500 most powerful computers in the world. In addition to graphical display, the simulation generates data for coordinates of each particle with the time increment of 0.00001 sec which allows predicting the exact location of each particle at each moment of casting. The results show that the process of centrifugal casting involving introduction of dispersed particles allows obtaining dispersed and hardened metal materials having forecasted distribution of heat-resisting particles.

中文翻译：

---

金属棒上模拟增强分散颗粒的分布

摘要

本文概述了使用ANSYS FLUENT 16.0软件模块开发计算机模型以预测离心铸造过程中颗粒分布的过程。有几种数学模型可以预测颗粒的体积分布。大多数模型基于稳态假设：这些模型描述了通过传播的结晶前沿排斥颗粒的标准，并且这些模型通过生长树枝状晶体来计算吸收颗粒的临界速度。一些模型旨在描述系统的动态状态或确定通过在离心金属铸造过程中硬化前端来捕获非金属夹杂物的标准。给出了推导模型的过程，示意图和几何形状的描述。模型预处理器考虑以下现象：两相流，能量方程，层流，离散相（硬化颗粒）的引入以及熔化/结晶现象。该模型考虑了两个液相之间的相互作用：空气和钢水。使用表面张力方程式描述相间相互作用。钢级12Х18Н10T被用作基础金属。引入碳化钨，碳化硼和氧化钇的颗粒作为分散体。这些物质的物理和化学参数由模拟器考虑。使用Skif-Ural描述了使用计算集群在离心铸造过程中模拟颗粒分布的过程，它是世界上功能最强大的500台计算机之一。除了图形显示之外，模拟会以0.00001秒的时间增量生成每个粒子坐标的数据，从而可以预测每个铸造时刻的每个粒子的确切位置。结果表明，涉及引入分散颗粒的离心铸造过程允许获得具有预测的耐热颗粒分布的分散和硬化的金属材料。