Incorporation of machine learning in additive manufacturing: a review

Raza, Ali; Deen, Kashif Mairaj; Jaafreh, Russlan; Hamad, Kotiba; Haider, Ali; Haider, Waseem

doi:10.1007/s00170-022-09916-4

Cited by 12 publications

(1 citation statement)

References 75 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Given the complexity, few studies have developed process-structure-property relations through physicbased [115][116][117] or data-driven models. [118][119][120][121] AM has a greater capacity to make components with complicated geometries, more operational flexibility, and shorter production times than traditional processes. However, severe problems related to the mechanical properties and surface quality of AM processes also exist.…”

Section: Future Research Pathsmentioning

confidence: 99%

A Review on Vibration Characteristics of Additively Manufactured Metal Alloys

Azher,

Shah,

Bakhtari

et al. 2024

Adv Eng Mater

View full text Add to dashboard Cite

The advent of additive manufacturing (AM) has dramatically shifted the manufacturing sector conceptualization, design, and creation of products. AM is poised to revolutionize goods production by establishing a new manufacturing paradigm, from reducing lead times to enabling rapid prototyping. AM can facilitate the production of complicated geometries and create functioning components with distinctive features for aerospace and automotive applications. However, defects such as pores, voids, interfaces, and inclusions can impair the quality and functionality of AM components. Vibration Analysis (VA) has become a popular tool for the dynamic qualification and testing of products and non‐destructive testing, but the literature lacks a comprehensive review of VA applied to AM. Hence, this paper summarizes recent advances in the application of VA for identifying and characterizing flaws in metal alloys, including titanium, aluminum, and nickel‐based alloys produced by AM. The review also includes studies on defects, such as porosity, cracks, and inclusions, and their effect on VA. The article concludes with a discussion of the limitations of VA for defect characterization and future research directions. Overall, VA is a promising non‐destructive testing method for quality assurance in AM and offers insights on overcoming the difficulties for further development and application of this technology.This article is protected by copyright. All rights reserved.

show abstract

Section: Future Research Pathsmentioning

confidence: 99%

A Review on Vibration Characteristics of Additively Manufactured Metal Alloys

Azher,

Shah,

Bakhtari

et al. 2024

Adv Eng Mater

View full text Add to dashboard Cite

show abstract

Machine learning in polymer additive manufacturing: a review

Nikooharf,

Shirinbayan,

Arabkoohi

et al. 2024

Int J Mater Form

View full text Add to dashboard Cite

Additive manufacturing (AM) has emerged as a commonly utilized technique in the manufacturing process of a wide range of materials. Recent advances in AM technology provide precise control over processing parameters, enabling the creation of complex geometries and enhancing the quality of the final product. Moreover, Machine Learning (ML) has become widely used to make systems work better by using materials and processes more intelligently and controlling their resulting properties. In industrial settings, implementing ML not only reduces the lead time of manufacturing processes but also enhances the quality and properties of produced parts through optimization of process parameters. Also, ML techniques have facilitated the advancement of cyber manufacturing in AM systems, thereby revolutionizing Industry 4.0. The current review explores the application of ML techniques across different aspects of AM including material and technology selection, optimization and control of process parameters, defect detection, and evaluation of properties results in the printed objects, as well as integration with Industry 4.0 paradigms. The progressive phases of utilizing ML in the context of AM, including data gathering, data preparation, feature engineering, model selection, training, and validation, have been discussed. Finally, certain challenges associated with the use of ML in the AM and some of the best-practice solutions have been presented.

show abstract