Uncertainty in phase diagram simulations with interatomic potentials

原子间势的原子模拟在计算化学、物理和材料科学中得到了广泛的应用。目前许多重要的过程是量子力学方法所不能及的，如密度泛函理论；只有经验电位才能达到必要的微观结构长度尺度和扩展的时间尺度。目前，在复杂原子模拟的输出上设置有意义的误差条几乎是不可能的。该博士项目将通过与ICAMS（德国波鸿）和OpenKIM项目（美国明尼苏达州）的合作伙伴合作，将模拟结果与潜在的形式和参数联系起来，解决这一挑战。

Atomistic simulations with interatomic potentials are very widely used throughout computational chemistry, physics and materials science. Currently many important processes are beyond the reach of quantum mechanical methods such as density functional theory; only empirical potentials can reach the necessary microstructural length scales and extended time scales. Currently it is almost impossible to put meaningful error bars on the output of complex atomistic simulations. This PhD project will address this challenge by relating simulation outcomes to the form and parameters of the potential, in collaboration with partners at ICAMS (Bochum, Germany) and the OpenKIM project (U. Minnesota, USA)