[1]徐彪,林也平,郭龙,等.核能材料辐照损伤的长时间动力学模拟新进展[J].中国材料进展,2026,45(11):070-79.
XU Biao,LIN Yeping,GUO Long,et al.Recent Progress on Long?time Dynamics Simulation of Radiation Damage in Nuclear Materials[J].MATERIALS CHINA,2026,45(11):070-79.
点击复制
核能材料辐照损伤的长时间动力学模拟新进展()
中国材料进展[ISSN:1674-3962/CN:61-1473/TG]
- 卷:
-
45
- 期数:
-
2026年11
- 页码:
-
070-79
- 栏目:
-
- 出版日期:
-
2026-10-31
文章信息/Info
- Title:
-
Recent Progress on Long?time Dynamics Simulation of Radiation Damage in Nuclear Materials
- 作者:
-
徐彪; 林也平; 郭龙; 邓辉球
-
湖南大学物理与微电子科学学院，湖南 长沙 410082
- Author(s):
-
XU Biao; LIN Yeping; GUO Long; DENG Huiqiu
-
School of Physics and Electronics, Hunan University, Changsha 410082, China
-
- 关键词:
-
辐照损伤; 缺陷演化; 长时间动力学; 跨尺度模拟; 机器学习; 动力学蒙特卡洛
- Keywords:
-
radiation damage; defect evolution; long?time dynamics; multiscale simulation; machine learning; kinetic Monte?Carlo
- 文献标志码:
-
A
- 摘要:
-
核能材料在反应堆服役期间长期处于高通量中子辐照、高温、腐蚀冷却介质耦合的极端工况下，辐照诱发的空位、间隙原子、位错环、堆垛层错四面体、空洞等微观缺陷会在飞秒至年跨尺度时间维度持续演化，是引发材料辐照肿胀、硬化、脆化等性能劣化的核心内因。传统分子动力学可提供原子级细节但受限于纳秒模拟时长，连续介质方法可实现长时间微结构演化模拟却丢失了原子级化学与结构精细信息，两者形成的"时间鸿沟"是辐照损伤模拟的核心挑战。本文以"空间分辨率—时间跨度—化学复杂度"三维能力空间为分析框架，以近五年代表性进展为重点，同时回顾相关经典方法，梳理加速分子动力学、动力学蒙特卡洛、机器学习辅助动力学蒙特卡洛、缺陷速率驱动长时间动力学及团簇动力学等技术体系，并比较其信息保留、适用边界与验证条件。本文重点讨论复杂固溶体合金中局域化学环境对动力学事件的影响、级联后缺陷的长时间演化及其与实验可观测量的衔接，以及位错环形貌定量判据在缺陷识别与模型校验中的应用。最后指出跨尺度信息传递保真度验证与模拟结果不确定性量化是当前领域研究短板。
- Abstract:
-
Nuclear materials are subjected to extreme conditions during reactor service, including prolonged exposure to high-flux neutron radiation, high temperatures, and corrosive cooling media. Irradiation-induced defects such as vacancies, interstitials, dislocation loops, stacking fault tetrahedra and voids,evolve continuously over the time-scales ranging from femtoseconds to years. These defects constitute the primary key cause of material performance degradation such as irradiation swelling, hardening and embrittlement.Classical molecular dynamics can provide atomistic?scale details, yet it is constrained by nanosecond?scale simulation durations. Continuum?based methods enable long?term microstructural evolution simulations but sacrifice atomistic?level chemical and structural fine?scale information. The resulting “time?scale gap” represents a core challenge in radiation?damage simulation. Taking the three?dimensional capability space of “spatial resolution?–?time span?–?chemical complexity” as the framework, this paper systematically reviews various multiscale simulation approaches for radiation?damage developed in recent years, covering the technical systems: accelerated molecular dynamics, kinetic Monte?Carlo (KMC), machine?learning?accelerated KMC, defect?rate?driven long?time dynamics and cluster dynamics. The applicable conditions,performance strengths and limitations of each method for bridging the time?scale gap are evaluated. This work highlights key advances of machine?learning?accelerated KMC in overcoming the combinatorial?explosion bottleneck for complex solid?solution alloys, the cross?scale linking capability of the defect rates-based long-time dynamics for connecting displacement cascades and long?term microstructural evolution, as well as the application of quantitative morphological criteria for dislocation loops in defect identification and model validation. Finally, this paper identifies two existing research bottlenecks in the field: fidelity verification of cross?scale information transfer and quantification of simulation uncertainty. Establishing standardized benchmarking systems is an indispensable pathway to advance the field from methodological innovation toward practical engineering?oriented predictive tools.
更新日期/Last Update:
2026-09-28