High resolution imaging for inspection of Laser Beam Melting systems
06 May 2013-pp 707-712
TL;DR: This work presents a high resolution imaging system for inspection of LBM systems which can be easily integrated into existing machines and shows that the system can detect topological flaws and is able to inspect the surface quality of built layers.
Abstract: Laser Beam Melting (LBM) allows the fabrication of three-dimensional parts from metallic powder with almost unlimited geometrical complexity and very good mechanical properties. LBM works iteratively: a thin powder layer is deposited onto the build platform which is then melted by a laser according to the desired part geometry. Today, the potential of LBM in application areas such as aerospace or medicine has not yet been exploited due to the lack of process stability and quality management. For that reason, we present a high resolution imaging system for inspection of LBM systems which can be easily integrated into existing machines. A container file stores calibration images and all layer images of one build process (powder and melt result) with corresponding metadata (acquisition and process parameters) for documentation and further analysis. We evaluate the resolving power of our imaging system and show that it is able to inspect the process result on a microscopic scale. Sample images of a part built with varied process parameters are provided, which show that our system can detect topological flaws and is able to inspect the surface quality of built layers. The results can be used for flaw detection and parameter optimization, for example in material qualification.
Citations
More filters
TL;DR: In this paper, a review of the literature and the commercial tools for insitu monitoring of powder bed fusion (PBF) processes is presented, focusing on the development of automated defect detection rules and the study of process control strategies.
Abstract: Despite continuous technological enhancements of metal Additive Manufacturing (AM) systems, the lack of process repeatability and stability still represents a barrier for the industrial breakthrough. The most relevant metal AM applications currently involve industrial sectors (e.g., aerospace and bio-medical) where defects avoidance is fundamental. Because of this, there is the need to develop novel in-situ monitoring tools able to keep under control the stability of the process on a layer-by-layer basis, and to detect the onset of defects as soon as possible. On the one hand, AM systems must be equipped with in-situ sensing devices able to measure relevant quantities during the process, a.k.a. process signatures. On the other hand, in-process data analytics and statistical monitoring techniques are required to detect and localize the defects in an automated way. This paper reviews the literature and the commercial tools for insitu monitoring of Powder Bed Fusion (PBF) processes. It explores the different categories of defects and their main causes, the most relevant process signatures and the in-situ sensing approaches proposed so far. Particular attention is devoted to the development of automated defect detection rules and the study of process control strategies, which represent two critical fields for the development of future smart PBF systems.
505 citations
Cites background from "High resolution imaging for inspect..."
...Other distortions affect critical features like thin walls, overhang surfaces and acute corners (zur Jacobsmühlen et al., 2013; Grasso et al., 2016)....
[...]
...The analysis of the surface pattern reveals irregularities that may be caused by different defects, including balling, porosity and other kinds of distortions (Kleszczynski et al., 2012 – 2014; zur Jacobsmühlen et al., 2013)....
[...]
...…of the slice characteristics allows detecting super-elevated edges and other surface irregularities that may have a strong impact on the wear of the recoating system and the consequent defect propagation within the building area (Kleszczynski et al., 2012 – 2014; zur Jacobsmühlen et al., 2013)....
[...]
TL;DR: In this paper, an overview over laser-based additive manufacturing with comments on the main steps necessary to build parts to introduce the complexity of the whole process chain is presented. But despite good sales of AM machines, there are still several challenges hindering a broad economic use of AM.
415 citations
TL;DR: A computer vision algorithm is used to automatically detect and classify anomalies that occur during the powder spreading stage of the process, which has the potential to become a component of a real-time control system in an LPBF machine.
Abstract: Despite the rapid adoption of laser powder bed fusion (LPBF) Additive Manufacturing by industry, current processes remain largely open-loop, with limited real-time monitoring capabilities. While some machines offer powder bed visualization during builds, they lack automated analysis capability. This work presents an approach for in-situ monitoring and analysis of powder bed images with the potential to become a component of a real-time control system in an LPBF machine. Specifically, a computer vision algorithm is used to automatically detect and classify anomalies that occur during the powder spreading stage of the process. Anomaly detection and classification are implemented using an unsupervised machine learning algorithm, operating on a moderately-sized training database of image patches. The performance of the final algorithm is evaluated, and its usefulness as a standalone software package is demonstrated with several case studies.
273 citations
TL;DR: In this paper, a convolutional neural network (CNN) was used for autonomous detection and classification of spreading anomalies in a laser powder bed fusion additive manufacturing (LPDAM) system.
Abstract: In-situ detection of processing defects is a critical challenge for Laser Powder Bed Fusion Additive Manufacturing. Many of these defects are related to interactions between the recoater blade, which spreads the powder, and the powder bed. This work leverages Deep Learning, specifically a Convolutional Neural Network (CNN), for autonomous detection and classification of many of these spreading anomalies. Importantly, the input layer of the CNN is modified to enable the algorithm to learn both the appearance of the powder bed anomalies as well as key contextual information at multiple size scales. These modifications to the CNN architecture are shown to improve the flexibility and overall classification accuracy of the algorithm while mitigating many human biases. A case study is used to demonstrate the utility of the presented methodology and the overall performance is shown to be superior to that of methodologies previously reported by the authors.
165 citations
References
More filters
01 Jan 2001
TL;DR: This book is referred to read because it is an inspiring book to give you more chance to get experiences and also thoughts and it will show the best book collections and completed collections.
Abstract: Downloading the book in this website lists can give you more advantages. It will show you the best book collections and completed collections. So many books can be found in this website. So, this is not only this multiple view geometry in computer vision. However, this book is referred to read because it is an inspiring book to give you more chance to get experiences and also thoughts. This is simple, read the soft file of the book and you get it.
14,282 citations
TL;DR: In this article, the authors describe which tools are and will be available to fulfil those requirements from the perspective of a laser machine manufacturer and set the focus on melt pool analysis and control.
242 citations
01 Jan 2000
TL;DR: The influence of several variables, including color misregistration, edge location estimation, data-record length and image noise on the measured MTF are addressed.
Abstract: The development and adoption of standards for the evaluation of digital camera resolution has helped foster the widespread use of slanted-edge-based analysis. In addition, the form of these evaluation methods suggests their use in imaging system analysis and design. The standards-specific methods and algorithms, however, are not intended for direct MTF evaluation, but if care is taken to avoid bias and minimize random error, the methods can successfully be used for this purpose. In this paper the influence of several variables are discussed. Specifically, the effect of color misregistration, edge location estimation, data-record length and image noise on the measured MTF are addressed.
205 citations
23 Jun 1999
TL;DR: The algorithm is ideally suited for real-time applications with a fixed amount of available processing time, since voting and line detection is interleaved and the most salient features are likely to be detected first.
Abstract: We present a novel Hough Transform algorithm referred to as Progressive Probabilistic Hough Transform (PPHT). Unlike the Probabilistic HT where Standard HT is performed on a pre-selected fraction of input points, PPHT minimises the amount of computation needed to detect lines by exploiting the difference an the fraction of votes needed to detect reliably lines with different numbers of supporting points. The fraction of points used for voting need not be specified ad hoc or using a priori knowledge, as in the probabilistic HT; it is a function of the inherent complexity of the input data. The algorithm is ideally suited for real-time applications with a fixed amount of available processing time, since voting and line detection is interleaved. The most salient features are likely to be detected first. Experiments show that in many circumstances PPHT has advantages over the Standard HT.
195 citations
TL;DR: In this paper, the design of an optical system capable of monitoring high scanning velocities and melt pool dynamics is introduced as a first step to the integration of a monitoring and control module into a SLM-machine.
190 citations