Process modeling in laser powder bed fusion towards defect detection and quality control via machine learning: The state-of-the-art and research challenges

2022 ◽  
Vol 73 ◽  
pp. 961-984
Author(s):  
Peng Wang ◽  
Yiran Yang ◽  
Narges Shayesteh Moghaddam
Keyword(s):  
Machine Learning ◽  
Quality Control ◽  
Defect Detection ◽  
Process Modeling ◽  
State Of The Art ◽  
The State ◽  
Powder Bed Fusion ◽  
Laser Powder Bed Fusion ◽  
Powder Bed ◽  
Research Challenges

scholarly journals Laser Powder Bed Fusion Parameter Selection via Machine-Learning-Augmented Process Modeling

JOM ◽  
2020 ◽  
Vol 72 (12) ◽  
pp. 4393-4403
Author(s):  
Sandeep Srinivasan ◽  
Brennan Swick ◽  
Michael A. Groeber
Keyword(s):  
Machine Learning ◽  
Process Modeling ◽  
Parameter Selection ◽  
Powder Bed Fusion ◽  
Laser Powder Bed Fusion ◽  
Powder Bed

scholarly journals In Situ Monitoring Systems of The SLM Process: On the Need to Develop Machine Learning Models for Data Processing

Crystals ◽  
2020 ◽  
Vol 10 (6) ◽  
pp. 524
Author(s):  
Pinku Yadav ◽  
Olivier Rigo ◽  
Corinne Arvieu ◽  
Emilie Le Guen ◽  
Eric Lacoste
Keyword(s):  
Machine Learning ◽  
Defect Detection ◽  
In Situ Monitoring ◽  
Fusion Process ◽  
Monitoring Systems ◽  
Powder Bed Fusion ◽  
Laser Powder Bed Fusion ◽  
Powder Bed ◽  
In Situ Sensing

In recent years, technological advancements have led to the industrialization of the laser powder bed fusion process. Despite all of the advancements, quality assurance, reliability, and lack of repeatability of the laser powder bed fusion process still hinder risk-averse industries from adopting it wholeheartedly. The process-induced defects or drifts can have a detrimental effect on the quality of the final part, which could lead to catastrophic failure of the finished part. It led to the development of in situ monitoring systems to effectively monitor the process signatures during printing. Nevertheless, post-processing of the in situ data and defect detection in an automated fashion are major challenges. Nowadays, many studies have been focused on incorporating machine learning approaches to solve this problem and develop a feedback control loop system to monitor the process in real-time. In our study, we review the types of process defects that can be monitored via process signatures captured by in situ sensing devices and recent advancements in the field of data analytics for easy and automated defect detection. We also discuss the working principles of the most common in situ sensing sensors to have a better understanding of the process. Commercially available in situ monitoring devices on laser powder bed fusion systems are also reviewed. This review is inspired by the work of Grasso and Colosimo, which presented an overall review of powder bed fusion technology.

A Physics-Informed Two-Level Machine Learning Model for Predicting Melt-Pool Size in Laser Powder Bed Fusion

2021 ◽  
Author(s):  
Yong Ren ◽  
Qian Wang ◽  
Panagiotis (Pan) Michaleris
Keyword(s):  
Machine Learning ◽  
Initial Temperature ◽  
Pool Size ◽  
Powder Bed Fusion ◽  
Laser Powder Bed Fusion ◽  
Powder Bed ◽  
Melt Pool ◽  
Input Feature ◽  
Melt Pool Size ◽  
Melt Pool Characteristics

Abstract Laser powder bed fusion (L-PBF) additive manufacturing (AM) is one type of metal-based AM process that is capable of producing high-value complex components with a fine geometric resolution. As melt-pool characteristics such as melt-pool size and dimensions are highly correlated with porosity and defects in the fabricated parts, it is crucial to predict how process parameters would affect the melt-pool size and dimensions during the build process to ensure the build quality. This paper presents a two-level machine learning (ML) model to predict the melt-pool size during the scanning of a multi-track build. To account for the effect of thermal history on melt-pool size, a so-called (pre-scan) initial temperature is predicted at the lower-level of the modeling architecture, and then used as a physics-informed input feature at the upper-level for the prediction of melt-pool size. Simulated data sets generated from the Autodesk's Netfabb Simulation are used for model training and validation. Through numerical simulations, the proposed two-level ML model has demonstrated a high prediction performance and its prediction accuracy improves significantly compared to a naive one-level ML without using the initial temperature as an input feature.

scholarly journals Review of Laser Powder Bed Fusion of Gamma-Prime-Strengthened Nickel-Based Superalloys

Metals ◽  
2020 ◽  
Vol 10 (8) ◽  
pp. 996
Author(s):  
Olutayo Adegoke ◽  
Joel Andersson ◽  
Håkan Brodin ◽  
Robert Pederson
Keyword(s):  
Defect Formation ◽  
State Of The Art ◽  
Powder Bed Fusion ◽  
Laser Powder Bed Fusion ◽  
Cast Material ◽  
Creep Properties ◽  
Powder Bed ◽  
Gamma Prime ◽  
Cracking Mechanisms ◽  
Nonequilibrium Solidification

This paper reviews state of the art laser powder bed fusion (L-PBF) manufacturing of γ′ nickel-based superalloys. L-PBF resembles welding; therefore, weld-cracking mechanisms, such as solidification, liquation, strain age, and ductility-dip cracking, may occur during L-PBF manufacturing. Spherical pores and lack-of-fusion voids are other defects that may occur in γ′-strengthened nickel-based superalloys manufactured with L-PBF. There is a correlation between defect formation and the process parameters used in the L-PBF process. Prerequisites for solidification cracking include nonequilibrium solidification due to segregating elements, the presence of liquid film between cells, a wide critical temperature range, and the presence of thermal or residual stress. These prerequisites are present in L-PBF processes. The phases found in L-PBF-manufactured γ′-strengthened superalloys closely resemble those of the equivalent cast materials, where γ, γ′, and γ/γ′ eutectic and carbides are typically present in the microstructure. Additionally, the sizes of the γ′ particles are small in as-built L-PBF materials because of the high cooling rate. Furthermore, the creep performance of L-PBF-manufactured materials is inferior to that of cast material because of the presence of defects and the small grain size in the L-PBF materials; however, some vertically built L-PBF materials have demonstrated creep properties that are close to those of cast materials.

Quality assurance in metal powder bed fusion via deep-learning-based image classification

2019 ◽  
Vol 26 (2) ◽  
pp. 259-266 ◽  
Author(s):  
Maximilian Hugo Kunkel ◽  
Andreas Gebhardt ◽  
Khumbulani Mpofu ◽  
Stephan Kallweit
Keyword(s):  
Machine Learning ◽  
Mechanical Properties ◽  
Quality Control ◽  
Metal Powder ◽  
Monitoring Data ◽  
Powder Bed Fusion ◽  
Destructive Testing ◽  
Deep Convolutional Neural Networks ◽  
Content Type ◽  
Powder Bed

Purpose This paper aims to establish a standardized, quick, reliable and cost-efficient method of quality control (QC) in metal powder bed fusion (PBFM) based on process monitoring data. Design/methodology/approach Based on destructive testing results that emerged from a statistical investigation on powder bed fusion process exceeding reproducibility of mechanical properties, it was investigated if the generated monitoring data from a concept laser machine allows reliable deductions on resulting mechanical properties of the manufactured specimens. Findings The application of machine learning on generated melt pool images, under-recognition of destructive testing results, enables enhanced pattern recognition. The generated computational model successfully classified 9,280 unseen layer images by 98.9 per cent accuracy. This finding offers an automated approach to quality control within PBFM. Originality/value To the authors knowledge, it is the first time that machine learning has been applied for the purpose of QC in additive manufacturing. The ability of deep convolutional neural networks to recognize patterns, which are imperceptible to the human eye, shows high potential to facilitate activities of QC and to minimize QC-related costs and throughput times. The achieved processing speed for image analyses also points a way for future developments of self-corrective PBFM systems.

Sign in / Sign up

Export Citation Format

Share Document