來自荷蘭格羅寧根大學的Francesco Maresca教授和博士生張磊，提出了一種提取裂紋尖端信息的主動學習方法，這一方法可應用于不同機器學習框架和不同材料，能夠以近第一性原理精度預測原子尺度下裂紋尖端的變形行為。

作者基于主動學習構建了DFT精度的高斯機器學習勢函數，給出了低溫下單晶鐵的脆性斷裂主導的裂紋擴展機制。他們的研究表明了機器學習勢函數可以通過特殊設計的數據庫構型來增加精度，且主動學習的效率要遠大于人工加入相關構型。

Fig. 3 Main steps of the active learning procedure.

該研究與相關實驗測得的斷裂韌性的對比揭示了即使在低溫條件下（77K），位錯活動對斷裂韌性的影響仍不可忽略，為多尺度方法模擬工程材料的斷裂韌性提供了新方法。相關論文近期發布于npj?Computational Materials?9:?217?(2023)。手機閱讀原文，請點擊本文底部左下角“閱讀原文”，進入后亦可下載全文PDF文件。

Editorial Summary

The prediction of atomistic fracture mechanisms in body-centred cubic (bcc) iron is essential for understanding its semi-brittle nature. Current research suggests that the interplay between thermal activation of screw dislocations and crack-tip dislocation emission dominates the brittle-to-ductile transition in BCC metals. Due to the limitation of computational power, DFT is not able to simulate the atomic-scale crack-tip extension. Classical molecular dynamics (MD) with different EAM interatomic potentials yield different crack-tip deformation mechanisms, which contradicts each other.

This study proposes an active learning approach to extract crack-tip configurations, applicable across various machine learning frameworks and materials, enabling first-principles accuracy in predicting atomic-scale crack-tip deformation mechanism.?

Professor Francesco Maresca and PhD student Lei Zhang at the University of Groningen developed a Gaussian approximation potential with near DFT accuracy, revealing that brittle fracture is the dominating mechanism in single-crystal iron with pre-existing cracks at low temperatures. The research demonstrates that the accuracy of machine learning potentials can be improved through specially designed database configurations, and active learning is significantly efficient than manually adding relevant configurations. By comparing the MD predicted fracture toughness with experiments, the study showed that even at low temperatures (77K), the influence of dislocations on fracture toughness cannot be ignored.?

The research emphasizes the importance of multiscale simulations in predicting fracture toughness, offering new insights for engineering materials using multiscale simulation methods.This?article was recently?published in?npj?Computational Materials?9:?217?(2023).

原文Abstract及其翻譯

Atomistic fracture in bcc iron revealed by active learning of Gaussian approximation potential (基于主動學習的高斯近似勢函數揭示BCC鐵在原子尺度下的斷裂機制)

Lei Zhang,?Gábor Csányi,?Erik van der Giessen?& Francesco Maresca?

Abstract: ?

The prediction of atomistic fracture mechanisms in body-centred cubic (bcc) iron is essential for understanding its semi-brittle nature. Existing atomistic simulations of the crack-tip under mode-I loading based on empirical interatomic potentials yield contradicting predictions and artificial mechanisms. To enable fracture prediction with quantum accuracy, we develop a Gaussian approximation potential (GAP) using an active learning strategy by extending a density functional theory (DFT) database of ferromagnetic bcc iron. We apply the active learning algorithm and obtain a Fe GAP model with a converged model uncertainty over a broad range of stress intensity factors (SIFs) and for four crack systems. The learning efficiency of the approach is analysed, and the predicted critical SIFs are compared with Griffith and Rice theories. The simulations reveal that cleavage along the original crack plane is the atomistic fracture mechanism for {100} and {110} crack planes at T?=?0?K, thus settling a long-standing issue. Our work also highlights the need for a multiscale approach to predicting fracture and intrinsic ductility, whereby finite temperature, finite loading rate effects and pre-existing defects (e.g., nanovoids, dislocations) should be taken explicitly into account.

摘要

原子尺度下準確預測鐵的斷裂機制有助于人們深刻理解bcc鐵的脆韌轉變行為，現有的幾種經典勢函數在預測原子尺度斷裂機制時結論不盡相同甚至有些相互矛盾。為了解決該問題，本研究開發了基于高斯近似機器學習勢函數的主動學習框架，在該框架基礎上拓展了bcc鐵的第一性原理計算數據庫。提出的新機器學習勢函數模型能夠預測鐵裂紋尖端原子尺度斷裂機制。主動學習框架確保了機器學習迭代過程中模型不確定性的收斂性，給出了預制裂紋在{100}和{110}面單晶bcc鐵在低溫下裂紋尖端斷裂機制為脆性斷裂本質。本工作為多尺度模擬預測斷裂提供了理論指導，強調了多尺度模擬（考慮溫度和預存在缺陷，如納米孔，位錯等）在斷裂韌性預測中的必要性。該研究可為高斷裂韌性的Fe基工程材料設計提供理論性指導。