Towards an instant structure-property prediction quality control tool for additive manufactured steel using a crystal plasticity trained deep learning surrogate

Tu, Yuhui; Liu, Zhongzhou; Carneiro, Luiz; Ryan, Caitríona M.; Parnell, Andrew C.; Leen, Seán B.; Harrison, Noel M.

doi:10.1016/j.matdes.2021.110345

Cited by 25 publications

(10 citation statements)

References 75 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…23 Structure-property models are usually developed to capture the microstructure-sensitivity of PBF-LB material. 24,25 Specifically, full-field crystal plasticity (CP) model, for example, CP finite element [26][27][28] and CP fast Fourier transform [29][30][31] models, are widely used to predict mechanical properties based on inherited microstructure, for example, from the PBF-LB process. However, the visible intragranular and intergranular mechanical fields from full-field CP also results in very high computational overhead and again, for a typically small-scale representative volume element of simulated material.…”

Section: Introductionmentioning

confidence: 99%

Process-structure-property modeling for postbuild heat treatment of powder bed fusion Ti-6Al-4V

Liu

Yang

Chai

et al. 2023

Proceedings of the Institution of Mechanical Engineers, Part L:

Self Cite

View full text Add to dashboard Cite

Postbuild heat treatment is an important component in optimized manufacturing processing for laser beam powder bed fusion (PBF-LB) Ti-6Al-4V. The development of predictive modeling, based on the understanding of the relationships between process parameters, microstructure evolution, and mechanical properties, is a potentially key ingredient in this optimization process. In this paper, a process-structure-property (PSP) model is developed to predict the effect of postbuild heat treatment on yield strength, which is a key tensile property for PBF-LB Ti-6Al-4V. The process-structure part is developed with a focus on the prediction of solid-state phase transformation, especially dissolution of martensite during the heating phase. Subsequent tensile properties are quantified by a microstructure-sensitive yield strength model based on the predicted microstructure variables. The integrated PSP model is validated via experimentally measured phase fraction, α lath width and monotonic tensile tests on PBF-LB Ti-6Al-4V with different heat treatment temperatures, for identification of optimal process parameters.

show abstract

Section: Introductionmentioning

confidence: 99%

Process-structure-property modeling for postbuild heat treatment of powder bed fusion Ti-6Al-4V

Liu

Yang

Chai

et al. 2023

Proceedings of the Institution of Mechanical Engineers, Part L:

Self Cite

View full text Add to dashboard Cite

show abstract

“…Figure depicts the multimodal deep-learning CNN framework proposed herein. CNN was utilized as it has been demonstrated to be successful for image feature extraction and the labor-intensive process of feature engineering. , Compared to the single input of CNN utilized in most of the previous studies, a multimodal input comprising the morphology (SEM image), composition (EDX image), and time data of the laser-textured surface was generated. As shown in Figure , the morphology and composition are coupled by overlapping; thus, six-channel coupling images are obtained.…”

Section: Multimodal Cnn Frameworkmentioning

confidence: 99%

Multimodal Deep-Learning Framework for Accurate Prediction of Wettability Evolution of Laser-Textured Surfaces

Zhang

Yang

Zhao

et al. 2023

ACS Appl. Mater. Interfaces

View full text Add to dashboard Cite

The lengthy process through which laser-textured surfaces transform from hydrophilic to hydrophobic severely restricts their practical applications. Accurately predicting the wettability evolution curve is crucial; however, developing a reliable prediction model remains challenging. Herein, a data-driven multimodal deep-learning framework was developed, in which multimodal data of micro/nanostructure morphology images, composition distribution images, and time information are effectively coupled and fed into a convolutional neural network (CNN). Rich data input and in-depth data mining make the framework more robust, achieving accurate prediction of the wettability evolution curves of various typical micro/nanostructures. Additionally, accurate prediction of input images with varying magnifications and untrained laser-textured surfaces demonstrates the generalizability of the multimodal CNN framework. The visualization results of the convolution layer confirmed the rationality of the information learned by the model. Additionally, the proposed multimodal CNN framework was successfully utilized to investigate the optimization process. Further, a laser-textured surface with a shorter evolution period and a larger final contact angle was realized. The proposed multimodal CNN framework offers an efficient and cost-effective method for predicting the wettability evolution curves and exploring the optimization processes, enhancing the application potential of laser micro/nanofabrication of superhydrophobic surfaces.

show abstract

“…Employing deep learning methods for material data analysis has demonstrated higher accuracy in predicting material mechanical responses than traditional machine learning methods, including support vector machines (SVM), decision trees, and back propagation (BP) neural networks, effectively reducing the computational and simulation time required to obtain material properties. In addition, Tu et al [12] . employed crystal plasticity finite element (CPFE) simulation results as a dataset to develop deep neural networks (DNN) and establish the relationships between different microstructures and strengths in stainless steels or multiphase materials.…”

Section: Introductionmentioning

confidence: 99%

Employing deep learning in non‐parametric inverse visualization of elastic–plastic mechanisms in dual‐phase steels

Han,

Wang,

Zhang

et al. 2024

Materials Genome Engineering Advances

View full text Add to dashboard Cite

Enhancing the interpretability of machine learning methods for predicting material properties is a key, yet complex topic in materials science. This study proposes an interpretable convolutional neural network (CNN) to establish the relationship between the microstructural evolution and mechanical properties of non‐uniform and nonlinear multisystem dual‐phase steel materials and achieve an inverse analysis of the elastic‐plastic mechanism. This study demonstrates that the developed CNN model achieves an accuracy of 94% in predicting the stress‐strain curves of dual‐phase steel microstructures with different compositions and processes, with the mean absolute error not exceeding 50 MPa, representing merely 5.26% of the average tensile strength of dual‐phase steels in the dataset. The reverse visualization results of the CNN model indicate that, during tensile deformation, the grain boundaries maintain deformation coordination within the grains by impeding dislocation slip. This results in a significant stress concentration at the grain boundaries, with stresses at the boundaries being higher than those borne by the martensitic phase and minimal stresses in the ferrite phase. Moreover, compared with traditional crystal plasticity models, the CNN model exhibits a substantial improvement in computational efficiency. This method provides a generic plan for improving the interpretability of machine learning methods for predicting material properties and can be easily applied to other alloy systems.

show abstract

Towards an instant structure-property prediction quality control tool for additive manufactured steel using a crystal plasticity trained deep learning surrogate

Cited by 25 publications

References 75 publications

Process-structure-property modeling for postbuild heat treatment of powder bed fusion Ti-6Al-4V

Process-structure-property modeling for postbuild heat treatment of powder bed fusion Ti-6Al-4V

Multimodal Deep-Learning Framework for Accurate Prediction of Wettability Evolution of Laser-Textured Surfaces

Employing deep learning in non‐parametric inverse visualization of elastic–plastic mechanisms in dual‐phase steels

Contact Info

Product

Resources

About