以分子動力模擬探討高熵合金之疊差能與力學性質

Abstract

高熵合金優秀的機械性質一直以來都是重要的研究課題，其中能維持fcc單相的Cantor alloy系統更因擁有高強度、高延展性及可擴展性而受到許多人的關注。本研究透過疊差能建立了Cantor alloy變形機制及其對力學性質影響的連結，使得合金調配有更堅實的理論基礎，並針對改進力學性質提出配比建議。透過力學機制與機械性質間的連結，降低了高熵合金配比研發的複雜性，加速高熵合金研發的進程。 藉由分子動力模擬及OVITO視覺化軟體，拉伸試驗、疊差能模擬及過程中缺陷的演進得以被視覺化，機械性質和變形機制也可以有效的分類。疊差、內部疊差、外部疊差、孿晶和hcp transformation induced plasticity (TRIP)透過自行開發的缺陷分類演算法得以分類並視覺化，變形機制的探討也因此有了比較堅實的物理基礎。 本研究結果吻合文獻提出，內部疊差能對延展性有一最佳的區間，可以最佳化合金的延展性。也由此分類出四種不同延展性表現的合金類別，對於其中的缺陷演進、變形路徑可以有更細微的觀察與探討。並分類出三種變形路徑，分別為gliding induced twining (GIT)、bundled twin growth (BTG)和bulk hcp TRIP；疊差能較高者會透過疊差的滑移延展孿晶，而較低者通常能透過差排滑移延續孿晶的成長。Hcp TRIP則能進一步提供更多滑移空間，因而增加合金的延展性。在不同機制的交互影響下，即使變形路徑不同，還是有可能會有高延展性，其關鍵在於孿晶的生長與長度。而同時擁有GIT、BTG變形路徑組合的配比通常能獲得高強度與高延展性。最後，本研究也透過貝式最佳化成功發現了更高強度的合金配比。

Reasons for the great mechanical attributes of high entropy alloys (HEAs) have been an important question to answer. In addition, Cantor alloys with fcc single phase have great expansibility in mechanical performance among other HEAs. In this work, connection between mechanical performance and deformation mechanisms was built, providing more solid background for compositional tuning as well as guidance for compositional tuning. Through the linkage between mechanical performance and deformation mechanisms, the design complexity of HEA is much lower, thus boosts the progress of HEA studies. Molecular dynamics simulations of tensile and stacking fault energy were conducted for observation of micro mechanisms. Stacking fault, intrinsic stacking fault, extrinsic stacking faults, twin and hcp transformation induced plasticity (TRIP) were classified and visualized through OVITO, of which defect evolutions and deformation maps were concluded accordingly. An optimum region of intrinsic/extrinsic stacking fault energy for higher ductility was obtained, where was later classified into four ductile types. Deformation evolution also were studied with respect to three different deformation mechanisms: gliding induced twinning (GIT), bundled twin growth (BTG) and bulk hcp TRIP. Growth and propagation of twins was found crucial for prolongation of HEAs, and a positive relation between strength and stacking energy was inducted. Better strength and ductility could be accomplished by controlling stacking fault energy while combining GIT and BTG. Bayesian optimization also was utilized for compositional predictions, of which compositions with much higher strength and product of ultimate tensile strength and total elongation (PSE) was found.