【作者】 王金照；

【导师】 过增元； 陈民；

【作者基本信息】 清华大学 ， 工程热物理， 2002， 硕士

【摘要】 深过冷液态金属快速凝固技术是一种非常有前途的高性能材料制备方法,深过冷液态金属热物性的研究在该制备技术起着重要的作用。本文是在已有液态金属热物性研究的基础上,采用分子动力学方法对深过冷金属热物性开展进一步的研究。主要的研究成果如下:1:针对势能模型对液态金属结构和比热预测的影响进行了模拟和理论分析,结果表明液态金属结构对势能模型不敏感,不同的势能模型模拟结果差别很小。液态金属比热对势能模型的要求比较高,不同的势能模型得到的比热有明显的差别。要准确地模拟液态金属比热,模拟中采用的势能模型应反映出Rose状态方程所体现的系统能量与原子间距的关系。2:采用约束方法的等温等压系综对常压下液态金属的密度和比热进行了模拟,与前人模拟结果不同的是,模拟得到的密度和比热与已有实验数据点吻合的很好,其差别分别在4%和5%以内。这表明采用合适的势能模型以及准确的模拟方法,分子动力学模拟可以准确地预测等压条件下的液态金属性质。3:对银的快速冷却过程进行了分子动力学模拟,并在模拟中分别从热力学性质、结构特性和动力学特性等方面观察到银的玻璃态转变,这一结果表明在很高的冷却速率下,金属银也可以发生玻璃化转变,其玻璃化温度在1000K左右。4:在集群计算机系统上实现了4,000,000个分子的并行计算,并对算法进行了优化,得到了具有较高并行效率和较好可扩展性的算法,为深过冷液态金属快速凝固的进一步研究打下了基础。更多还原

【Abstract】 Rapid solidification of supercooled liquid metals offers a promising way forpreparing the high performance materials. The thermal properties of undercooledliquid metals play an important role in the prediction and control of this solidificationprocess. Based on the available research of the undercooled liquid thermal properties,this thesis adopts molecular dynamics method to investigate the thermal properties ofundercooled liquid metal. The main results can be summarized as following aspects.The effect of different potential models on the structure and heat capacity ofliquid metal are simulated and analyzed. The results show that the prediction of liquidmetal structure is comparatively insensitive to different versions of the EAM model,while the prediction of specific heat demands a more rigorous potential than thestructure predictions do. The potential model which embodies the atomseparation-energy relation in the Rose’ equation can predict the heat capacityaccurately. Johnson’s potential model can predict the structure and heat capacity ofliquid metals accurately, and is simpler than other equations. Therefore, Johnson’spotential model is fit for the simulation of the liquid metal’s thermal properties.The density and heat capacity of liquid metal under constant pressure aresimulated by the constraint isobar-isothermal ensemble. Different with the availablesimulation results, the simulated density and heat capacity are well agree with theavailable experimental results, the difference is less than 4% for density and 5% forspecific. This results show the undercooled liquid metal’s thermal properties can bepredicted accurately by the molecular dynamics simulations which adopt the fitfulpotential model and right simulation method.The rapid cooling process of silver are carried out by molecular dynamicssimulation. The glass transition is found by the observation of the thermal properties,structure parameter and dynamics characteristic. These results show that the glasstransition of silver can be fulfilled in the quick cooling process.Parallel molecule dynamics simulations with 4,000,000 molecules on the clustercomputer system are realized and optimized. The optimized calculation has highparallel efficiency and wide expansibility, which pave a firm foundation to the furtherresearch on rapid solidification of the undercooled liquid metal.更多还原