Modeling of the Ultra-Precision Machining Process Using New Combined Molecular Dynamics/Monte Carlo (MD/MC) Simulation

这笔赠款为开发原子水平的加工模拟技术提供资金，称为分子动力学(MD)模拟。将考虑以下三个重要的模拟领域，它们将对我们对切割过程的理解产生重大影响。它们是：(1)在常规切割速度下的加工模拟，由于传统的MD模拟涉及较长的加工时间，以前从未尝试过；(2)用金刚石工具加工半导体材料(如硅、锗)的模拟。此外，在这一类别下还包括：(1)使用改进的嵌入式原子方法(MEAM)用钻石工具加工铁的模拟，以研究磨损的化学性质；以及(3)加工bcc(体心立方)和hcp(六角密排)材料(以及目前正在建模的fcc(面心立方体)金属)的模拟；(3)在分布式计算环境(或Beowulf集群)中使用并行处理，以显著减少每次运行的计算时间，以便可以考虑大尺寸工件(多达100万个原子)或更低的切割速度。混合分子动力学/蒙特卡罗(MD/MC)方法能够解决在常规切割速度下的加工问题。在MC模拟中，时间(或切割速度)不是一个显式变量，因为它与一系列平衡状态有关。然而，它是通过切割过程中的温度间接参与的。如果一个人事先知道在常规切割速度下的温度分布，那么这个信息可以被用作MC运动的输入。根据这项拨款建议的工作将能够在纳米级确定半导体材料的机械性能，以应用于微型机电系统(MEMS)，并用于各种材料(包括金属和半导体材料)的超精密加工。值得注意的是，实验技术需要非常昂贵的高精度、高硬度的温控环境下的机床和昂贵的单晶金刚石工具。仿真可以提供足够的信息，使得只需要进行几次测试来验证MD仿真结果。新的混合MD/MC方法还可以使用更大尺寸的工件(最多一百万个原子)和接近传统切割速度的切割速度。

This grant provides funding for the development of techniques for the simulation of machining at the atomic level, known as, molecular dynamics (MD) simulation. The following three important areas of simulation that would have a significant impact on our understanding of the cutting process will be considered. They are: (1) simulations of machining at conventional cutting speeds, never before attempted due to long processing times involved with conventional MD simulations, (2) simulations of machining of semiconductor materials, such as silicon, germanium with a diamond tool. Also, included under this category are the simulations of machining of iron with a diamond tool to investigate the chemical nature of wear and simulations of machining of bcc (body centered cubic) and hcp (hexagonal close packed) materials (in addition to fcc (face centered cubic) metals currently being modeled), using the Modified Embedded Atom Method (MEAM), and (3) use of parallel processing in a distributed computing environment (or Beowulf cluster) to significantly reduce the computational time per run so that large size work pieces (up to 1 million atoms) or lower cutting speeds can be considered. The hybrid Molecular Dynamics/Monte Carlo (MD/MC) approach enables addressing of the machining problem at conventional cutting speeds. In MC simulations, time (or the cutting velocity) is not an explicit variable as one is concerned with a series of equilibrium states. However, it is involved indirectly through the temperature in the cutting process. If one knows the temperature distribution at conventional cutting speeds a priori, then this information can be used as an input to the MC moves. The work proposed under this grant will enable determination of mechanical properties of semiconductor materials at nanoscale for application to microelectromechanical systems (MEMS), and for ultraprecision machining of a wide range of materials (both metals and semiconductor materials). It may be noted that experimental techniques require very expensive high precision, high rigidity machine tools in a temperature controlled environment and costly single crystal diamond tools. The simulations can provide adequate information such that only a few tests to verify the MD simulation results are necessary. The new hybrid MD/MC approach also enables use of larger size workpieces (up to a million atoms) and cutting speeds close to conventional.