Accurate High-Performance Atomic Structure Calculations

该项目的目标是通过开发新的高性能软件，使GRASP2K计算模型的结果在给定的计算机资源下的精度提高10倍，该软件更高效、更易于维护，并且可以根据原子理论的未来发展进行修改。这需要：*1)重新设计程序，以遵守当前的软件工程设计原则，并在大型情况下使用高效的算法。*2)使用最先进的科学编程语言重铸程序，以实现高性能计算。*3)引入适当的模块风格，该风格预测了定义H.*历史的原子核模型、Breit校正和QED效应的未来变化，清楚地表明准确性可以对科学的进步产生巨大影响。一个例子是第谷·布拉赫(1546-1601)，他毕生致力于开发比以前精确十倍的工具来记录行星的位置。他的数据足够准确，足以让开普勒发现行星在椭圆轨道上运行，这为牛顿提供了建立万有引力平方反比理论所需的线索。在量子力学中，电子系统的状态由满足波动方程H W=E W的波函数W来描述。其中H是系统的哈密顿量，E是总能量。对于N个电子的原子，波动方程是3N个空间变量的偏微分方程组。使这一问题具有挑战性的是当两个电子之间的距离变为零时出现的奇点。系统的可观测性质是量子力学算符的期望值。因此，当H和W已知时，所有的原子性质都可以预测。对于轻原子，H通常是非相对论哈密顿量。对于重元素，H需要基于包含量子电动力学效应的完全相对论狄拉克理论，以及原子核的有限模型。超重元素的H是当前的研究课题。在原子物理学中，准确的计算结果需要与作为数值和不确定度报告的实验结果一致。对于任何给定的H，计算模型面临的挑战是奇点，即电子运动的相关性。*软件开发的测试用例将来自与国际同事合作完成的当前物理学研究课题。最大的挑战来自对重元素或高电离原子的计算。一个例子是目前正在考虑用于靶向癌症治疗的元素他汀类(N=85)。实验研究计划在瑞典进行。另一项关键测试将是铀(N=92)的光谱计算，但尚未报告可靠的结果。在寻找“稳定岛”的超重元素的情况下，该代码可能成为发展新物理理论的重要工具。

The goal of this project is to improve the accuracy of results from the GRASP2K computational model by a factor of 10 for given computer resources by developing new high-performance software that is more efficient, easier to maintain, and can be modified readily for future developments in atomic theory. This requires:******1) Redesigning programs to adhere to current software engineering principles of design and using efficient algorithms for large cases.******2) Recasting programs in the most advanced scientific programming language for high-performance computing.******3) Introducing a proper modular style that anticipates future changes in the model of the nucleus, the Breit correction, and QED effects that define H.******History shows clearly that accuracy can have tremendous impact on the advancement of science. An example is Tycho Brahe (1546-1601) who dedicated his life to developing tools for recording planetary positions ten times more accurately than before. His data was accurate enough for Kepler to discover that the planets moved in elliptic orbits which gave Newton the clues he needed to establish universal inverse-square gravitation theory.******In quantum mechanics, the state of an electronic system is described by a wave function W that satisfies the wave equation H W = E W. Here H is the Hamiltonian of the system and E the total energy. For an atom with N electrons, the wave equation is a partial differential equation with 3N space variables. What makes the problem challenging are the singularities that occur when the distance between the two electrons goes to zero. Observable properties of the system are expectation values of quantum mechanical operators. Thus, when H and W are known, all atomic properties can be predicted. For light atoms, H often is the non-relativistic Hamiltonian. For heavy elements H needs to be based on fully relativistic Dirac theory that includes quantum electrodynamic effects, and a finite model for the nucleus. H for superheavy elements is a current research topic. In atomic physics an accurate computed result needs to agree with an experimental result reported as a value and an uncertainty. For any given H, a challenge for the computational model are the singularities, i.e. correlation in the motion of the electrons.*** ***Test cases for the development of the software will be drawn from current research topics in physics, done in collaboration with international colleagues. The biggest challenges are presented by calculations for heavy elements or highly ionized atoms. An example is the element Astatine (N=85) that is currently being considered for use in targeted cancer therapy. Experimental studies are planned in Sweden. Another critical test would be spectrum calculations for Uranium (N=92) where reliable results have not been reported. In the case of superheavy elements for the search of “islands of stability”, the code could be an important tool for the development of new physics theory.