Mining structure-property linkage in nanoporous materials using an interpretative deep learning approach

Liu, Haomin; Shargh, Ali K.; Abdolrahim, Niaz

doi:10.1016/j.mtla.2021.101275

Cited by 7 publications

(4 citation statements)

References 49 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Uncovering of structure‐property (SP) relationship plays an important role in the design and development of new materials. [ 1–3 ] With the guidance of physical metallurgy and information strategy, we [ 4 ] have previously demonstrated that the microstructure information can improve the prediction of hardness of austenitic steels. Molkeri et al.…”

Section: Instructionmentioning

confidence: 99%

“…">InstructionUncovering of structure-property (SP) relationship plays an important role in the design and development of new materials. [1][2][3] With the guidance of physical metallurgy and information strategy, we [4] have previously demonstrated that the microstructure information can improve the prediction of hardness of austenitic steels. Molkeri et al [5] proved that incorporating microstructure knowledge into the materials design process results in better and faster solutions by proposing a novel microstructure-aware…”

mentioning

confidence: 99%

See 1 more Smart Citation

Symbolic Regression and Two‐Point Statistics Assisted Structure‐Property Linkage Based on Irregular‐Representative Volume Element

Chen

Zhao

et al. 2022

Advcd Theory and Sims

View full text Add to dashboard Cite

Quantifying the microstructure of materials is of significance in material development, especially for building the relationship between structure and property. To establish a remarkable structure‐property (SP) linkage, a novel concept referred to as irregular‐representative volume element (IRVE) based on panoramic image stitching technology (PIST) is proposed and a data‐driven scheme integrating irregular domain‐oriented two‐point statistics, principal component analysis (PCA), and symbolic regression based on genetic programming (GPSR) is constructed. Combining with advanced image processing and genetic programming technologies, this scheme improves the microstructure quantization framework. This scheme can not only be applied in different complex conditions for extracting the information of a material microstructure, but can also to embody details of microstructure from the perspective of large scale. IRVE is demonstrated to have both strong statistical representativeness and sufficient physical interpretation, which makes the scheme robust and reliable. Performing the scheme on an example of ferrite heat‐resistant steels, it shows a powerful ability in building an equational SP linkage with high precision (R = 0.91, RMSE = 13.17), the generalization ability of the linkage is also validated by an unseen steel (relative percentage error is 2.66%). The scheme has bright application prospects in predicting mechanical property and accelerating alloy design.

show abstract

Section: Instructionmentioning

confidence: 99%

mentioning

confidence: 99%

Symbolic Regression and Two‐Point Statistics Assisted Structure‐Property Linkage Based on Irregular‐Representative Volume Element

Chen

Zhao

et al. 2022

Advcd Theory and Sims

View full text Add to dashboard Cite

show abstract

“…Machine learning (ML) models are alternative promising tools that can explore the design space in a significantly faster pace in comparison with performing massive number of MD simulations to conduct the forward design approach. Different ML modes such as: support vector machine (SVM) 19 , random forest 20,21 , convolutional neural network (CNN) [22][23][24][25][26] , multi-layer perceptron (MLP) neural network 27 , attention-based transformer neural network 28 and graph-based neural networks 29 are adopted as surrogate forward models to relate the microstructures or microstructural features into mechanical properties in many applications. For instance, Yang et al 22 combined principal component analysis (PCA) and CNN to predict the stress-strain behavior of binary composites up to the failure point.…”

Section: Introductionmentioning

confidence: 99%

An interpretable deep learning approach for designing nanoporous silicon nitride membranes with tunable mechanical properties

Shargh

Abdolrahim

2023

npj Comput Mater

Self Cite

View full text Add to dashboard Cite

The high permeability and strong selectivity of nanoporous silicon nitride (NPN) membranes make them attractive in a broad range of applications. Despite their growing use, the strength of NPN membranes needs to be improved for further extending their biomedical applications. In this work, we implement a deep learning framework to design NPN membranes with improved or prescribed strength values. We examine the predictions of our framework using physics-based simulations. Our results confirm that the proposed framework is not only able to predict the strength of NPN membranes with a wide range of microstructures, but also can design NPN membranes with prescribed or improved strength. Our simulations further demonstrate that the microstructural heterogeneity that our framework suggests for the optimized design, lowers the stress concentration around the pores and leads to the strength improvement of NPN membranes as compared to conventional membranes with homogenous microstructures.

show abstract

“…There have been previous works on developing various machine learning and DL methods for diffraction data analysis 17,28 for different purposes such as pattern decomposition, cluster analysis [29][30][31][32] , crystal structure classification 33 , structure-property relationships 34 , and phase mapping [35][36][37] . Park et al 13 introduced convolutional neural network (CNN) models trained on simulated XRD patterns (synthetic data) for classifying crystal systems, and space groups.…”

Section: Introductionmentioning

confidence: 99%

Automated Classification of Big X-ray Diffraction Data Using Deep Learning Models

Abdolrahim

Salgado

Lerman

et al. 2023

Preprint

View full text Add to dashboard Cite

Current in-situ X-ray diffraction (XRD) techniques generate data over human analytical capabilities – leading to the loss of novel insights. Automated techniques require human intervention, and lack the performance and adaptability needed for material exploration. With the critical need for high-throughput automated XRD pattern analysis of novel materials, we developed a generalized deep learning model to classify a diverse set of materials’ crystal systems and space groups. In our approach, we generated training data with a holistic representation of patterns that emerge from varying experimental conditions and crystal properties. We then used an expedited learning technique to refine our model’s expertise to experimental conditions. Additionally, we used evaluation data to interpret our model’s decision-making and optimized model architecture to elicit classification based on Bragg’s Law. We evaluated our models on experimental data, novel materials, and altered cubic crystals, where we observed state-of-the-art performance and even greater advances in space group classification.

show abstract

Mining structure-property linkage in nanoporous materials using an interpretative deep learning approach

Cited by 7 publications

References 49 publications

Symbolic Regression and Two‐Point Statistics Assisted Structure‐Property Linkage Based on Irregular‐Representative Volume Element

Symbolic Regression and Two‐Point Statistics Assisted Structure‐Property Linkage Based on Irregular‐Representative Volume Element

An interpretable deep learning approach for designing nanoporous silicon nitride membranes with tunable mechanical properties

Automated Classification of Big X-ray Diffraction Data Using Deep Learning Models

Contact Info

Product

Resources

About