Data-Driven Microstructure and Microhardness Design in Additive Manufacturing Using a Self-Organizing Map

Gan, Zhengtao; Li, Hengyang; Wolff, Sarah; Bennett, Jennifer L.; Hyatt, Gregory; Wagner, Gregory J.; Cao, Jian; Liu, Wing Kam

doi:10.1016/j.eng.2019.03.014

Cited by 54 publications

(24 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, Goh et al [52] showed that several other methods such as genetic algorithm, k-means, random forest, reinforcement learning, support vector regression and ensemble methods have also been used to address problems related to process optimization and quality control. In addition to these supervised learning techniques, a self-organizing feature map in the sense of unsupervised learning has recently been used to discover process-structure-properties (PSP) relationships in large and high-dimensional AM process datasets [53]. Overall, ML algorithms seem to be generally superior to conventional optimization methods such as polynomial regression, Taguchi or analysis of variance (ANOVA) due to their ability to establish nonlinear relationships between input and output variables [52].…”

Section: Existing Work On Process Optimization and Limitationsmentioning

confidence: 99%

Improving Build Quality in Laser Powder Bed Fusion Using High Dynamic Range Imaging and Model-Based Reinforcement Learning

et al. 2021

View full text Add to dashboard Cite

In laser-based additive manufacturing (AM) of metal parts from powder bed, information about actual part quality obtained during build is essential for cost-efficient production and high product quality. Reliable and effective monitoring strategies for laser powder bed fusion (LPBF) therefore remain in high demand and are the subject of current research. To address this demand, a novel analysis approach using high dynamic range (HDR) optical imaging in combination with convolutional neural networks (CNN) is proposed for spatially resolved and layer-wise prediction of the surface roughness of LPBF parts. In a further step, the predicted surface roughness maps are used as a feedback signal for a reinforcement learning technique that employs a dynamics model to subsequently identify optimal process parameters under varying and uncertain conditions. The proposed approach ultimately combines the estimation of the local surface roughness based on image texture and model-based reinforcement learning to an in-situ optimization framework for LPBF processes. In addition, the relationship between the layer surface roughness of the part and the overall part density is discussed on the basis of experimental data, which also indicate the applicability of the proposed method in industrial environments. This preliminary study is a first step towards highly adaptive and intelligent machines in the field of automated laser powder bed fusion with the primary goals of reducing production costs and improving the environmental fingerprint as well as print quality.

show abstract

Section: Existing Work On Process Optimization and Limitationsmentioning

confidence: 99%

Improving Build Quality in Laser Powder Bed Fusion Using High Dynamic Range Imaging and Model-Based Reinforcement Learning

et al. 2021

View full text Add to dashboard Cite

show abstract

“…Due to its complex nature, ANN algorithms have been used to determine the PSP relationships for many AM techniques. Gan et al [ 93 ] attempted to use a self-organizing map (SOM) to identify the PSP relationship of the directed energy deposition process for Inconel 718. Multiple objective optimizations of the process parameters can be achieved from the large and high-dimensional dataset, which is obtained from simulation and validated with experimental results, with the help of visualized SOM, as shown in Figure 13 .…”

Section: Applications Of Ann In 3d Printingmentioning

confidence: 99%

“… An illustration of the workflow normally used in current numerical studies ( top row ) and experimental studies ( bottom row ), accompanied by a description of how the ANN technique can be incorporated to discover useful process–structure–property relationships of certain materials [ 93 ]; published under the open access Creative Commons license with Elsevier. …”

Section: Figurementioning

confidence: 99%

Artificial Neural Network Algorithms for 3D Printing

et al. 2020

View full text Add to dashboard Cite

Additive manufacturing with an emphasis on 3D printing has recently become popular due to its exceptional advantages over conventional manufacturing processes. However, 3D printing process parameters are challenging to optimize, as they influence the properties and usage time of printed parts. Therefore, it is a complex task to develop a correlation between process parameters and printed parts’ properties via traditional optimization methods. A machine-learning technique was recently validated to carry out intricate pattern identification and develop a deterministic relationship, eliminating the need to develop and solve physical models. In machine learning, artificial neural network (ANN) is the most widely utilized model, owing to its capability to solve large datasets and strong computational supremacy. This study compiles the advancement of ANN in several aspects of 3D printing. Challenges while applying ANN in 3D printing and their potential solutions are indicated. Finally, upcoming trends for the application of ANN in 3D printing are projected.

show abstract

“…In 2019, Gan et al published a machine learning approach that resulted in a full processing-microstructure-property linkage for DED of Inconel 718 [388]. This work is unique in that it leveraged diverse experimental and computational approaches.…”

Section: Machine Learned Processing-structure Relationshipsmentioning

confidence: 99%

Survey of Modeling and Simulation Techniques for Advanced Manufacturing Technologies Volume I – Predicting Initial Microstructures

Nicolas

Chakraborty

Paulson

et al. 2020

View full text Add to dashboard Cite

This report summarizes the current state of modeling and simulation methods for predicting the initial structure and properties of material in components assembled using advanced manufacturing technologies (AMTs). The report is the first volume in a two-volume series. This first volume focuses on predicting initial microstructures of AM material. The second volume discusses predicting material properties given the initial microstructure. The focus is on technologies of particular relevance to the design and manufacture of nuclear reactor components. The purpose of the report is to help develop the technical knowledge base to support regulatory decisions that will be needed to assess nuclear components manufactured with AMTs as these components are installed at nuclear power plants (NPPs). The report develops a list of AMTs of particular interest to the NRC and summarizes the key microstructural features and corresponding processing parameters relevant to each technology. Further sections describe available physically-based and data-driven prediction methods and survey widely available software tools. The report concludes with a summary of gaps that may be of particular interest to the NRC when evaluating modeling and simulation methods for AMTs. v

show abstract

Data-Driven Microstructure and Microhardness Design in Additive Manufacturing Using a Self-Organizing Map

Cited by 54 publications

References 24 publications

Improving Build Quality in Laser Powder Bed Fusion Using High Dynamic Range Imaging and Model-Based Reinforcement Learning

Improving Build Quality in Laser Powder Bed Fusion Using High Dynamic Range Imaging and Model-Based Reinforcement Learning

Artificial Neural Network Algorithms for 3D Printing

Survey of Modeling and Simulation Techniques for Advanced Manufacturing Technologies Volume I – Predicting Initial Microstructures

Contact Info

Product

Resources

About