Computational Materials Science最新文献

筛选
英文 中文
A GPT-assisted iterative method for extracting domain knowledge from a large volume of literature of electromagnetic wave absorbing materials with limited manually annotated data 在人工标注数据有限的情况下,从大量电磁波吸收材料文献中提取领域知识的 GPT 辅助迭代法
IF 3.1 3区 材料科学
Computational Materials Science Pub Date : 2024-10-05 DOI: 10.1016/j.commatsci.2024.113431
Dongbo Dai , Guangjie Zhang , Xiao Wei , Yudian Lin , Mengmeng Dai , Junjie Peng , Na Song , Zheng Tang , Shengzhou Li , Jiwei Liu , Yan Xu , Renchao Che , Huiran Zhang
{"title":"A GPT-assisted iterative method for extracting domain knowledge from a large volume of literature of electromagnetic wave absorbing materials with limited manually annotated data","authors":"Dongbo Dai ,&nbsp;Guangjie Zhang ,&nbsp;Xiao Wei ,&nbsp;Yudian Lin ,&nbsp;Mengmeng Dai ,&nbsp;Junjie Peng ,&nbsp;Na Song ,&nbsp;Zheng Tang ,&nbsp;Shengzhou Li ,&nbsp;Jiwei Liu ,&nbsp;Yan Xu ,&nbsp;Renchao Che ,&nbsp;Huiran Zhang","doi":"10.1016/j.commatsci.2024.113431","DOIUrl":"10.1016/j.commatsci.2024.113431","url":null,"abstract":"<div><div>Research on electromagnetic wave absorbing materials is an important part of materials science. Each year, a substantial amount of academic literature is published in this field, containing critical information. Rapid and effective knowledge extraction from these documents is key to accelerating field development, and automated knowledge extraction based on deep learning provides a solution to this challenge. However, deep learning models typically require extensive annotated data for training, which is time-consuming and expensive to obtain in highly specialized subfields. To address this issue, this paper presents a GPT-assisted iterative training method that uses only 30 manually annotated literature abstracts as a training set and ultimately achieves an F1 score of 82.94% for a named entity recognition model (NER). The effectiveness of this model is demonstrated by comparing with other large language models commonly used in materials science on a custom dataset. We constructed a knowledge extraction framework centered around the obtained NER model and collected literature on electromagnetic wave absorbing materials from the last decade. The extraction and application results demonstrate the practicality of our framework.</div></div>","PeriodicalId":10650,"journal":{"name":"Computational Materials Science","volume":"246 ","pages":"Article 113431"},"PeriodicalIF":3.1,"publicationDate":"2024-10-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142424696","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
The first principle calculations of band offsets among Cs2(Ti, Zr, Hf)X6 double halide perovskites Cs2(Ti, Zr, Hf)X6 双卤化物包晶带偏移的第一原理计算
IF 3.1 3区 材料科学
Computational Materials Science Pub Date : 2024-10-04 DOI: 10.1016/j.commatsci.2024.113436
Yongyut Laosiritaworn, Atchara Punya Jaroenjittichai
{"title":"The first principle calculations of band offsets among Cs2(Ti, Zr, Hf)X6 double halide perovskites","authors":"Yongyut Laosiritaworn,&nbsp;Atchara Punya Jaroenjittichai","doi":"10.1016/j.commatsci.2024.113436","DOIUrl":"10.1016/j.commatsci.2024.113436","url":null,"abstract":"<div><div>This study investigates the band offsets among Cs<sub>2</sub>(Ti, Zr, Hf)X<sub>6</sub> double halide perovskites. Valence band offsets (VBO) and conduction band offsets (CBO) were calculated using density functional theory (DFT) with the Perdew–Burke-Ernzerhof (PBE) functional and the hybrid Heyd–Scuseria–Ernzerhof (HSE) functional, employing the supercell technique. This technique offers greater accuracy and reliability by explicitly calculating the potential differences at the interface. A critical factor influencing the results is the contribution of the dipole potential (V<sub>D</sub>), which induces shifts in the VBO and CBO by approximately 0.2 to 0.6 eV relative to values predicted by the electron affinity rule. This discrepancy arises from the inclusion of interface-specific effects, such as charge redistribution and polarization. Additionally, the findings indicate that the energy band alignments among these compounds are type-I within the same group of halides, with nearly identical lattice constants for Zr- and Hf-based compounds due to their similar ionic radii. These results provide valuable insights for the design of heterostructures in electronic applications and highlight the potential of Cs<sub>2</sub>(Ti, Zr, Hf)X<sub>6</sub> compounds as efficient materials for solar cells, light-emitting diodes, and photodetectors.</div></div>","PeriodicalId":10650,"journal":{"name":"Computational Materials Science","volume":"246 ","pages":"Article 113436"},"PeriodicalIF":3.1,"publicationDate":"2024-10-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142424690","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
WC electron microscopy image segmentation based on improved watershed and Hu-moment edge matching algorithms 基于改进的分水岭和 Hu-moment 边缘匹配算法的 WC 电子显微镜图像分割技术
IF 3.1 3区 材料科学
Computational Materials Science Pub Date : 2024-10-03 DOI: 10.1016/j.commatsci.2024.113401
Yixuan Zhong , Yi Liu , Kai Liu , Teng Zhan , Shuli Liu , Yunlong Liang , Yuliang Hu , Mingfu Li , Gaopan Lei , Shiyu Zhou , Jingang Liu
{"title":"WC electron microscopy image segmentation based on improved watershed and Hu-moment edge matching algorithms","authors":"Yixuan Zhong ,&nbsp;Yi Liu ,&nbsp;Kai Liu ,&nbsp;Teng Zhan ,&nbsp;Shuli Liu ,&nbsp;Yunlong Liang ,&nbsp;Yuliang Hu ,&nbsp;Mingfu Li ,&nbsp;Gaopan Lei ,&nbsp;Shiyu Zhou ,&nbsp;Jingang Liu","doi":"10.1016/j.commatsci.2024.113401","DOIUrl":"10.1016/j.commatsci.2024.113401","url":null,"abstract":"<div><div>The particle size distribution of WC powder particles has a great influence on material properties. However, the traditional manual particle size analysis methods are both time-consuming and inaccurate, and the commonly used particle size detection methods belong to statistical indexes, which cannot reflect the real particle size. To address the above problems, this paper proposes an image segmentation method based on the improved watershed algorithm and the Hu-moment edge matching algorithm, which can realize accurate segmentation and particle size analysis of adherent particles in WC electron microscope images. First, an improved bilateral filtering and Otsu image coarse segmentation method is proposed to extract the target region of particles; then, an improved watershed algorithm based on the multi-threshold H-maxima transform is proposed to realize the segmentation of adherent particles; and a region merging correction based on the Hu-moment edge matching algorithm is proposed to avoid over-segmentation. We compare and analyze the performance of this method with manual segmentation and some other common segmentation methods. The experimental results show that the standard deviations of the particle sizes obtained by the method proposed in this paper are less than 3%, and the segmentation accuracy is greatly improved compared with other segmentation algorithms.</div></div>","PeriodicalId":10650,"journal":{"name":"Computational Materials Science","volume":"246 ","pages":"Article 113401"},"PeriodicalIF":3.1,"publicationDate":"2024-10-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142424691","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
Molecular simulation-based developer screening for molecular glass photoresists 基于分子模拟的分子玻璃光刻胶显影剂筛选
IF 3.1 3区 材料科学
Computational Materials Science Pub Date : 2024-10-03 DOI: 10.1016/j.commatsci.2024.113429
Peng Lian , Rongrong Peng , Tianjun Yu , Guoqiang Yang , Jinping Chen , Yi Li , Yi Zeng
{"title":"Molecular simulation-based developer screening for molecular glass photoresists","authors":"Peng Lian ,&nbsp;Rongrong Peng ,&nbsp;Tianjun Yu ,&nbsp;Guoqiang Yang ,&nbsp;Jinping Chen ,&nbsp;Yi Li ,&nbsp;Yi Zeng","doi":"10.1016/j.commatsci.2024.113429","DOIUrl":"10.1016/j.commatsci.2024.113429","url":null,"abstract":"<div><div>Screening of photoresist developers is critical for high-resolution lithography processes. Efficient estimation for photoresist solubility to facilitate the process of developer screening is of both fundamental and practical importance. In this study, we proposed a solubility prediction and developer screening approach for the molecular glass photoresists based on the molecular simulation calculation of two-component solubility parameters. The values of the two-component solubility parameters (<em>δ<sub>Lennard-Jones</sub></em> and <em>δ<sub>Coulomb</sub></em>) for 60 solvents were calculated, and their correlation with experimental Hansen solubility parameters was investigated. Meanwhile, the parameters calculation methods of binary mixed solvents with different polarities were systematically investigated. Then, the accuracy of solubility prediction was verified by dissolution experiments and Hansen solubility parameters, revealing that the two-component solubility parameters could reasonably reflect the solubilities of neutral and ionic molecular glass photoresists in most solvents. Furthermore, developer screening schemes using both pure and mixed solvents were investigated based on the two-component solubility parameters, which was further confirmed by the practical lithography experiments. The current method provides a viable approach for characterizing the photoresist solubility and screening appropriate developers, which is beneficial for accelerating the development of photoresist materials.</div></div>","PeriodicalId":10650,"journal":{"name":"Computational Materials Science","volume":"246 ","pages":"Article 113429"},"PeriodicalIF":3.1,"publicationDate":"2024-10-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142424764","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
Analysis of microstructure uncertainty propagation in fibrous composites Empowered by Physics-Informed, semi-supervised machine learning 纤维复合材料微观结构不确定性传播分析 借助物理信息半监督式机器学习技术
IF 3.1 3区 材料科学
Computational Materials Science Pub Date : 2024-10-03 DOI: 10.1016/j.commatsci.2024.113423
Qianyu Zhou , Ryan S. Enos , Kai Zhou , Haotian Sun , Dianyun Zhang , Jiong Tang
{"title":"Analysis of microstructure uncertainty propagation in fibrous composites Empowered by Physics-Informed, semi-supervised machine learning","authors":"Qianyu Zhou ,&nbsp;Ryan S. Enos ,&nbsp;Kai Zhou ,&nbsp;Haotian Sun ,&nbsp;Dianyun Zhang ,&nbsp;Jiong Tang","doi":"10.1016/j.commatsci.2024.113423","DOIUrl":"10.1016/j.commatsci.2024.113423","url":null,"abstract":"<div><div>The advancements in multi-scale computational analysis of fiber reinforced composites have led to the possibility of predicting important material properties based on their microstructure characteristics. Nevertheless, major challenges remain. The fiber distributions feature inherent randomness, which naturally leads to variations in properties such as transverse strength. This in turn undermines the significance of deterministic analysis to guide manufacturing optimization. Direct Monte Carlo simulation for uncertainty analysis is computationally insurmountable, as a single run of finite element simulation is already costly. While several surrogate modeling techniques leveraging supervised learning have been explored, it is commonly recognized that the efficacy of these surrogate models hinges upon the size of training dataset. In this research we establish a semi-supervised learning framework that can produce highly accurate emulation results with much reduced size of labeled training dataset. A random fiber packing algorithm is employed to sample the representative volume element (RVE) images that are subsequently fed to the finite element analysis to generate the ground-truth labeled data used in the training of neural network. To reduce the ground-truth labeling cost while maintaining the deep learning capacity. we employ the pseudo labeling technique where the base model is initially trained on a small set of ground truth labeled data and then used to generate credible pseudo-labels for a larger pool of unlabeled data. The model is subsequently retrained on this augmented dataset with adjusted weights and biases to reflect the varying confidence in the label sources. This framework is successfully employed in the analysis of microstructure uncertainty propagation in fibrous composites. The proposed approach efficiently leverages patterns from both unlabeled and limited labeled samples to predict transverse strength for varied RVE samples, matching the efficacy of a fully supervised model trained with 1,000 ground truth labels while simultaneously slashing labeling efforts by 72%. This framework can be extended to uncertainty propagation analysis using microstructure characteristics of other materials.</div></div>","PeriodicalId":10650,"journal":{"name":"Computational Materials Science","volume":"246 ","pages":"Article 113423"},"PeriodicalIF":3.1,"publicationDate":"2024-10-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142424678","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
Phase transformation mechanism in irradiation-induced superlattice formation 辐照诱导超晶格形成过程中的相变机制
IF 3.1 3区 材料科学
Computational Materials Science Pub Date : 2024-10-03 DOI: 10.1016/j.commatsci.2024.113418
Larry K. Aagesen , Yongfeng Zhang , Chao Jiang , Jian Gan
{"title":"Phase transformation mechanism in irradiation-induced superlattice formation","authors":"Larry K. Aagesen ,&nbsp;Yongfeng Zhang ,&nbsp;Chao Jiang ,&nbsp;Jian Gan","doi":"10.1016/j.commatsci.2024.113418","DOIUrl":"10.1016/j.commatsci.2024.113418","url":null,"abstract":"<div><div>Atomic kinetic Monte Carlo simulations were used to model void superlattice formation under irradiation in molybdenum, driven by anisotropic diffusion of self-interstitial atoms. A change in the phase transformation mechanism from nucleation and growth to spinodal decomposition occurred with increasing dose rate, with both mechanisms leading to superlattice formation. Analysis of a rate-theory based analytical model showed that an observed change in the kinetics of vacancy accumulation, the appearance of a region of positive second derivative in the plot of average vacancy concentration versus time, was caused by the onset of spinodal instability. The analytical model showed that for molybdenum and several other metals where void superlattice formation is commonly observed, the phase transformation likely occurs by nucleation and growth. However, nickel may offer the possibility of experimental observation of the transition between phase transformation mechanisms.</div></div>","PeriodicalId":10650,"journal":{"name":"Computational Materials Science","volume":"246 ","pages":"Article 113418"},"PeriodicalIF":3.1,"publicationDate":"2024-10-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142424774","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
Construction of Al–Si interatomic potential based on Bayesian active learning 基于贝叶斯主动学习的铝硅原子间势构建
IF 3.1 3区 材料科学
Computational Materials Science Pub Date : 2024-10-02 DOI: 10.1016/j.commatsci.2024.113422
Xuedong Liu , Yan Zhang , Hui Xu
{"title":"Construction of Al–Si interatomic potential based on Bayesian active learning","authors":"Xuedong Liu ,&nbsp;Yan Zhang ,&nbsp;Hui Xu","doi":"10.1016/j.commatsci.2024.113422","DOIUrl":"10.1016/j.commatsci.2024.113422","url":null,"abstract":"<div><div>Nanoscale simulations for optimizing the performance and processing of Al–Si alloys are currently facing two major obstacles: the scarcity of high-quality semi-empirical potentials tailored to complex alloy systems, and the prohibitively high computational cost associated with ab initio molecular dynamics simulations. In order to enhance simulation efficiency and accuracy of the Al–Si alloys’ microstructural evolution, this study employs a dynamic active learning technique, FLARE, to develop a non-parametric machine learning potential that combines the high accuracy of density functional theory (DFT) with the efficiency of classical molecular dynamics (MD). Without relying on extensive initial ab initio molecular dynamics data or existing databases, collection of the necessary data is progressively made during the active learning process, thereby constructing a potential capable of accurately simulating the structure and dynamics of high-temperature Al–Si alloys. By comparing with experimental measurements and ab initio molecular dynamics calculations, the high accuracy and computational efficiency of this potential is demonstrated in predicting energy, force, structure, and dynamic properties. The results provide novel theoretical insights for optimizing Al–Si alloy processing and underscore the usefulness of active learning methods in constructing high-accuracy potentials.</div></div>","PeriodicalId":10650,"journal":{"name":"Computational Materials Science","volume":"246 ","pages":"Article 113422"},"PeriodicalIF":3.1,"publicationDate":"2024-10-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142424771","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
The electronic and optical properties of group III-V semiconductors: Arsenides and Antimonides III-V 族半导体的电子和光学特性:砷化物和锑化物
IF 3.1 3区 材料科学
Computational Materials Science Pub Date : 2024-10-01 DOI: 10.1016/j.commatsci.2024.113381
Ruixin Gong , Lianqing Zhu , Qingsong Feng , Lidan Lu , Bingfeng Liu , Yuhao Chen , Yuanbo Zhang , Shiya Zhang , Yang Chen , Zhiying Liu
{"title":"The electronic and optical properties of group III-V semiconductors: Arsenides and Antimonides","authors":"Ruixin Gong ,&nbsp;Lianqing Zhu ,&nbsp;Qingsong Feng ,&nbsp;Lidan Lu ,&nbsp;Bingfeng Liu ,&nbsp;Yuhao Chen ,&nbsp;Yuanbo Zhang ,&nbsp;Shiya Zhang ,&nbsp;Yang Chen ,&nbsp;Zhiying Liu","doi":"10.1016/j.commatsci.2024.113381","DOIUrl":"10.1016/j.commatsci.2024.113381","url":null,"abstract":"<div><div>Investigating the structural, electronic, and optical properties of zinc-blende III-V semiconductors, particularly arsenides, and antimonides, which are crucial for optoelectronic devices such as transistors, infrared detectors, and quantum technologies due to their wide range of direct bandgaps. In this work, we have employed a first-principles approach integrating G<sub>0</sub>W<sub>0</sub> with the HSE06 hybrid functional and spin–orbit coupling (SOC) to study their fundamental properties. Traditional Density Functional Theory (DFT) methods, particularly those using Generalized Gradient Approximation (GGA) PBE functionals, tend to underestimate bandgaps, leading to discrepancies with experimental results. To address this, our study corrects the bandgap underestimation and refines the calculation of optical constants, including the dielectric function, refractive index, extinction coefficient, and absorption coefficient. Moreover, the optimized lattice constants and electronic properties derived from our computational model strongly correlate with experimental data, demonstrating the model’s reliability in predicting material properties. The findings suggest that our methods can be applied to arsenides and antimonides, offering a pathway to designing materials with optoelectronic properties involving III-V compounds and their complex heterostructures for advanced device applications.</div></div>","PeriodicalId":10650,"journal":{"name":"Computational Materials Science","volume":"246 ","pages":"Article 113381"},"PeriodicalIF":3.1,"publicationDate":"2024-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142358011","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
Ab initio molecular dynamics simulations on the combustion mechanism of Al/Fe2O3 nanothermite at various temperatures 不同温度下 Al/Fe2O3 纳米热敏电阻燃烧机理的 Ab initio 分子动力学模拟
IF 3.1 3区 材料科学
Computational Materials Science Pub Date : 2024-10-01 DOI: 10.1016/j.commatsci.2024.113427
Chuang Xue , Pin Gao , Guixiang Wang , Xuedong Gong
{"title":"Ab initio molecular dynamics simulations on the combustion mechanism of Al/Fe2O3 nanothermite at various temperatures","authors":"Chuang Xue ,&nbsp;Pin Gao ,&nbsp;Guixiang Wang ,&nbsp;Xuedong Gong","doi":"10.1016/j.commatsci.2024.113427","DOIUrl":"10.1016/j.commatsci.2024.113427","url":null,"abstract":"<div><div>Acquiring a comprehensive understanding of the combustion mechanism of nanothermites is crucial for elucidating phenomena and enhancing performance. The ab initio molecular dynamics method was employed to investigate the reaction process of the Al/Fe<sub>2</sub>O<sub>3</sub> nanothermite at various temperatures. The complex combustion behavior and reaction mechanism were qualitatively described in terms of dynamic morphologies, potential energy, atomic density distribution, etc. Results suggest the initial reaction of Al/Fe<sub>2</sub>O<sub>3</sub> is initiated by the migration of interfacial O atoms and the dissociation of Fe-O bonds, subsequently leading to the formation of alumina at the interface, which impedes the further progression of the thermite reaction. The increase in temperature enhances atomic diffusion and provides sufficient energy for the reaction. The interfacial metal Al layer undergoes melting and diffuses into the iron oxide layer, while vacancies generated during the reaction process sustain the continuous migration of internal oxygen atoms. At 2000 K and 3200 K, the initial structures completely collapse, facilitating the inward propagation of the thermite reaction, which subsequently results in the formation of alumina, iron clusters, and intermetallic compounds (AlFe, AlFe<sub>3</sub>, and Al<sub>6</sub>Fe). These findings offer significant insights into the combustion reaction mechanisms of nanothermites.</div></div>","PeriodicalId":10650,"journal":{"name":"Computational Materials Science","volume":"246 ","pages":"Article 113427"},"PeriodicalIF":3.1,"publicationDate":"2024-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142358085","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
Deep operator network surrogate for phase-field modeling of metal grain growth during solidification 用于凝固过程中金属晶粒生长相场建模的深度算子网络代用工具
IF 3.1 3区 材料科学
Computational Materials Science Pub Date : 2024-10-01 DOI: 10.1016/j.commatsci.2024.113417
Danielle Ciesielski, Yulan Li, Shenyang Hu, Ethan King, Jordan Corbey, Panos Stinis
{"title":"Deep operator network surrogate for phase-field modeling of metal grain growth during solidification","authors":"Danielle Ciesielski,&nbsp;Yulan Li,&nbsp;Shenyang Hu,&nbsp;Ethan King,&nbsp;Jordan Corbey,&nbsp;Panos Stinis","doi":"10.1016/j.commatsci.2024.113417","DOIUrl":"10.1016/j.commatsci.2024.113417","url":null,"abstract":"<div><div>A deep operator network (DeepONet) has been constructed that generates accurate representations of phase-field model simulations for evolving two dimensional metal grain morphology growing from melt. These representations serve as lower resolution, computationally efficient stand-ins for quick parameter space exploration of solutions to the Allen–Cahn equations that dictate the phase-field model simulations. The experimental target for the phase-field model is a uranium casting system cooling a 434 <span><math><mi>g</mi></math></span> uranium charge from a maximum temperature of 1400 °C at an average rate of 30 °<span><math><mfrac><mrow><mi>C</mi></mrow><mrow><mtext>min</mtext></mrow></mfrac></math></span>, traversing the crystallographic phases of the pure metal. Experimental parameters inform the phase-field model, whose higher resolution computational model solutions are used to train the DeepONet in a given parameter space with the aim of developing a faster, more efficient method for predicting the solidifying metal’s microstructure at different potential experimental values. The final DeepONet generates high accuracy, lower resolution predictions with cumulative relative approximation error over all timesteps of less than 0.5%, while ensuring solutions remain within physically feasible ranges. These relative error values are comparable with other state-of-the-art DeepONet models for microstructure evolution, while significantly reducing the amount of training data required. Training a convolutional neural network simultaneously with the DeepONet, enforcing realistic values at the complex metal grain boundaries, and mathematically encoding boundary conditions into the structure of the DeepONet improved prediction accuracy and computational efficiency over a standard DeepONet model.</div></div>","PeriodicalId":10650,"journal":{"name":"Computational Materials Science","volume":"246 ","pages":"Article 113417"},"PeriodicalIF":3.1,"publicationDate":"2024-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142424772","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
0
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
相关产品
×
本文献相关产品
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信