SI2-SSE: Development of Cassandra, A General, Efficient and Parallel Monte Carlo Multiscale Modeling Software Platform for Materials Research

材料的性质是组成这些材料的原子之间相互作用的结果。现在，即使对于自然界中尚未存在的材料，也可以通过使用先进的计算方法来研究材料的组成原子如何相互作用及其环境，从而非常准确地预测这些性质。这一领域依赖于复杂的软件包的存在，使研究人员能够进行这些模拟。有两种通用的方法来模拟块体材料：分子动力学和蒙特卡罗，每一种都适用于某些问题。有许多可用的分子动力学软件包，但几乎没有通用的蒙特卡罗程序。该项目旨在开发一个高效的、通用的开源蒙特卡罗代码Cassandra。为此，PI小组开发的学术蒙特卡罗代码将得到扩展和增强。该代码将能够模拟散装和异构环境中的任何类型的分子。该代码将包含广泛的高级功能，使其对一系列问题有用。通过向研究界提供通用代码，并建立一种机制，使用户可以添加自己的功能并扩展代码，该项目将对研究界产生广泛的影响。它将使非专家能够使用蒙特卡罗模拟来研究新问题。它将使经验丰富的分子建模人员能够利用并为单个优化和验证的代码贡献功能，从而减少为单个应用开发专用代码所需的时间和费用。由于该代码将用于教学和研讨会，因此将向教育工作者提供材料，以便在讲授热力学、分子建模和统计力学等课程时在课堂上使用该代码。

The properties of materials are the result of the interactions between the atoms that make up these materials. These properties can now be predicted with great accuracy, even for materials that have not yet existed in nature, by using advanced computational methods to study how the constituent atoms of the materials interact with one another and their environment. This field relies upon the existence of sophisticated software packages that enable researchers to conduct these simulations. There are two general approaches for simulating bulk materials: molecular dynamics and Monte Carlo, each of which is appropriate for certain problems. There are many molecular dynamics software packages available but almost no general purpose Monte Carlo codes. This project seeks to develop an efficient, general-purpose open source Monte Carlo code called Cassandra.To do this, the academic Monte Carlo code developed in the PI's group will be extended and enhanced. The code will be capable of simulating any type of molecule in bulk and heterogeneous environments. The code will contain a wide range of advanced features, making it useful for a range of problems. By providing a general purpose code to the research community and establishing a mechanism whereby users can add their own features and extend the code, this project will have a broad impact on the research community. It will enable non-experts to use Monte Carlo simulations to study new problems. It will enable experienced molecular modelers to utilize and contribute features to a single optimized and validated code, thereby alleviating the time and expense associated with developing specialized codes for individual applications. Because the code will be used in teaching and workshops, materials will be made available to educators to use the code in the classroom when teaching courses such as thermodynamics, molecular modeling and statistical mechanics.