OPEn HPC theRmomechanical tools for the development of eAtf fuels

进一步提高轻水核反应堆在新的运行条件下的安全性，这是由于它们与更多样化的能源结构相结合而引起的，这给燃料开发带来了许多新的挑战。这需要有效和经过验证的工具，使人们能够捕捉从标称到设计基础事故的各种操作条件下燃料元件行为的复杂性。OperaHPC提案的目标是开发使用高性能计算（HPC）的开放式工具，实现燃料元件的完整3D高保真热机械模拟，包括材料微观结构。这将有助于设计所谓的燃料元件数字双胞胎。这一发展包括一个雄心勃勃的基础研究计划，致力于利用从原子尺度到物质定律的多尺度实验和模拟来研究辐照燃料的非线性力学行为。这将产生燃料元件在桩内行为的详细描述和模拟所需的材料数据。开发的工具将根据最先进的1D/3D燃料性能代码进行评估，以进行验证、边界条件的定义以及与中子、热化学和热水力代码的耦合。验证和不确定度分析也将通过将三维模拟结果与辐照计划的实验数据进行比较来进行。由于应用了基于降阶和元模型（包括人工智能）的新方法，这些高级模拟的知识将被转移到工业燃料性能代码中。HPC工具最终将被应用于创新燃料元件概念的详细评估，包括（增强的）事故耐受性燃料，在几个轻水反应堆设计的瞬态条件下。

Increasing further the safety of light water nuclear reactors in the new operating conditions induced by their integration in a more variedenergy mix brings many new challenges for fuel development. This calls for effective and validated tools enabling one to capture thecomplexity of the behaviour of fuel elements under various operation conditions from nominal to design basis accident ones.. Theobjective of the OperaHPC proposal is to develop open tools using High Performance Computing (HPC) enabling a full 3D high-fidelitythermo-mechanical simulation of the fuel element including the material microstructure. This will contribute to the design of so-calledfuel element digital twins. This development includes an ambitious basic research program devoted to the investigation of non-linearmechanical behaviour of irradiated fuel using multiscale experiments and simulations from the atomic scale up to the material law. Thiswill yield the detailed description of the in-pile behaviour of the fuel element and the materials data necessary for the simulation. The toolsdeveloped will be assessed against state-of-the-art 1D/3D fuel performance codes for verification, definition of boundary conditions andcoupling with neutronic, thermochemical and thermohydraulic codes. Validation and uncertainty analyses will also be performed throughthe comparison of the results of the 3D simulations with the experimental data available from irradiation programs. The knowledge fromthese advanced simulations will be transferred to industrial fuel performance codes thanks to the application of new methods based onreduced order and meta models, including Artificial Intelligence. The HPC tools will finally be applied to the detailed evaluation ofinnovative fuel element concepts, including (enhanced) accident tolerant fuels, under transient conditions in several light water reactordesigns.