ARTICLE | doi:10.20944/preprints201901.0169.v1
Subject: Engineering, Industrial & Manufacturing Engineering Keywords: electron beam melting; in-process monitoring; quality control; electronic imaging; spatial resolution
Online: 17 January 2019 (02:53:17 CET)
Electron Beam Melting (EBM) is an increasingly used Additive Manufacturing (AM) technique employed by many industrial sectors, including the medical device and aerospace industries. In-process EBM monitoring for quality assurance purposes has been a popular research area. Electronic imaging has recently been investigated as one of the in-process EBM data collection methods, alongside thermal/ optical imaging techniques. Despite certain capabilities of an electronic imaging system have been investigated, experiments are yet to be carried out to benchmark one of the most important features of any imaging systems – spatial resolution. This article addresses this knowledge gap by: (1) proposing an indicator for the estimation of spatial resolution which includes the Backscattered Electrons (BSE) information depth, (2) estimating the achievable spatial resolution when electronic imaging is carried out inside an Arcam A1 EBM machine, and (3) presenting an experimental method to conduct edge resolution evaluation with the EBM machine. Analyses of experimental results indicated that the spatial resolution was of the order of 0.3mm-0.4mm when electronic imaging was carried out at room temperature. It is believed that by disseminating an analysis and experimental method to estimate and quantify spatial resolution, this study has contributed to the on-going quality assessment research in the field of in-process monitoring of the EBM process.
ARTICLE | doi:10.20944/preprints201901.0098.v1
Subject: Engineering, Industrial & Manufacturing Engineering Keywords: Additive Manufacturing; Electron Beam Melting; In-Process Monitoring; Quality Control; Electronic Imaging
Online: 10 January 2019 (11:52:10 CET)
Electron Beam Melting (EBM) is an increasingly used Additive Manufacturing (AM) technique employed by many industrial sectors, including the medical device and aerospace industries. The application of this technology is, however, challenged by the lack of process monitoring and control system that underpins process repeatability and part quality reproducibility. An electronic imaging system prototype has been developed to serve as an EBM monitoring technique, the capabilities of which have been verified at room temperature and at 320+10°C. Nevertheless, in order to fully assess the applicability of this technique, the image quality needs to be investigated at a range of elevated temperatures to fully understand the influence of thermal noise due to heat. In this paper, electronic imaging pilot trials at elevated temperatures, ranging from room temperature to , were carried out. Image quality measure Q of the digital electron images was evaluated, and the influence of temperature was investigated. In this study, raw electronic images generated at higher temperatures had greater Q values, i.e. better global image quality. It has been demonstrated that, for temperatures between , the influence of temperature on electronic image quality was not adversely affecting the visual clarity of image features. It is envisaged that the prototype has significant potential to contribute to in-process EBM monitoring in many manufacturing sectors.
ARTICLE | doi:10.20944/preprints201809.0346.v1
Subject: Engineering, Mechanical Engineering Keywords: SLM, Process Control, Semi-supervised Machine Learning, Randomised Singular Value Decomposition
Online: 18 September 2018 (11:21:58 CEST)
Risk-averse areas such as the medical, aerospace and energy sectors have been somewhat slow towards accepting and applying Additive Manufacturing (AM) in many of their value chains. This is partly because there are still signicant uncertainties concerning the quality of AM builds. This paper introduces a machine learning algorithm for the automatic detection of faults in AM products. The approach is semi-supervised in that, during training, it is able to use data from both builds where the resulting components were certied and builds where the quality of the resulting components is unknown. This makes the approach cost ecient, particularly in scenarios where part certication is costly and time consuming. The study specically analyses Selective Laser Melting (SLM) builds. Key features are extracted from large sets of photodiode data, obtained during the building of 49 tensile test bars. Ultimate tensile strength (UTS) tests were then used to categorise each bar as `faulty' or `acceptable'. A fully supervised approach identied faulty specimens with a 77% success rate while the semi-supervised approach was able to consistently achieve similar results, despite being trained on a fraction of the available certication data. The results show that semi-supervised learning is a promising approach for the automatic certication of AM builds that can be implemented at a fraction of the cost currently required.