Filtern
Dokumenttyp
- Vortrag (9)
- Beitrag zu einem Tagungsband (7)
- Zeitschriftenartikel (5)
- Posterpräsentation (2)
- Corrigendum (1)
Schlagworte
- Thermography (13)
- Laser Powder Bed Fusion (9)
- In-situ Monitoring (7)
- Defect detection (5)
- Laser powder bed fusion (5)
- Machine Learning (5)
- In-situ monitoring (4)
- Additive Manufacturing (3)
- Additive manufacturing (3)
- Additive manufacturing (AM) (3)
Organisationseinheit der BAM
Paper des Monats
- ja (1)
Eingeladener Vortrag
- nein (9)
Laser powder bed fusion is used to create near net shape metal parts with a high degree of freedom in geometry design. When it comes to the production of safety critical components, a strict quality assurance is mandatory. An alternative to cost-intensive non-destructive testing of the produced parts is the utilization of in-situ process monitoring techniques. The formation of defects is linked to deviations of the local thermal history of the part from standard conditions. Therefore, one of the most promising monitoring techniques in additive manufacturing is thermography. In this study, features extracted from thermographic data are utilized to investigate the thermal history of cylindrical metal parts. The influence of process parameters, part geometry and scan strategy on the local heat distribution and on the resulting part porosity are presented. The suitability of the extracted features for in-situ process monitoring is discussed.
Additive manufacturing (AM) of metals and in particular laser powder bed fusion (LPBF) enables a degree of freedom in design unparalleled by conventional subtractive methods. To ensure that the designed precision is matched by the produced LPBF parts, a full understanding of the interaction between the laser and the feedstock powder is needed. It has been shown that the laser also melts subjacent layers of material underneath. This effect plays a key role when designing small cavities or overhanging structures, because, in these cases, the material underneath is feed-stock powder. In this study, we quantify the extension of the melt pool during laser illumination of powder layers and the defect spatial distribution in a cylindrical specimen. During the LPBF process, several layers were intentionally not exposed to the laser beam at various locations, while the build process was monitored by thermography and optical tomography. The cylinder was finally scanned by X-ray computed tomography (XCT). To correlate the positions of the unmolten layers in the part, a staircase was manufactured around the cylinder for easier registration. The results show that healing among layers occurs if a scan strategy is applied, where the orientation of the hatches is changed for each subsequent layer. They also show that small pores and surface roughness of solidified material below a thick layer of unmolten material (>200 µm) serve as seeding points for larger voids. The orientation of the first two layers fully exposed after a thick layer of unmolten powder shapes the orientation of these voids, created by a lack of fusion.
The detection of internal irregularities is crucial for quality assessment in metal-based additive manufacturing (AM) technologies such as laser powder bed fusion (L-PBF). The utilization of in-process thermography as an in situ monitoring tool in combination with post-process X-ray micro computed tomography (XCT) as a reference technique has shown great potential for this aim. Due to the small irregularity dimensions, a precise registration of the datasets is necessary as a
requirement for correlation. In this study, the registration of thermography and XCT reference datasets of a cylindric specimen containing keyhole pores is carried out for the development of a porosity prediction model. The considered datasets show variations in shape, data type and dimensionality, especially due to shrinkage and material elevation effects present in the manufactured part. Since the resulting deformations are challenging for registration, a novel preprocessing methodology is introduced that involves an adaptive volume adjustment algorithm which is based on the porosity distribution in the specimen. Thus, the implementation of a simple three-dimensional image-to-image registration is enabled. The results demonstrate the influence of the part deformation on the resulting porosity location and the importance of registration in terms of irregularity prediction.
Die zerstörungsfreie Prüfung von metallischen Bauteilen hergestellt mit additiver Fertigung (Additive Manufacturing - AM) gewinnt zunehmend an industrieller Bedeutung. Grund dafür ist die Feststellung von Qualität, Reproduzierbarkeit und damit auch Sicherheit für Bauteile, die mittels AM gefertigt wurden. Jedoch wird noch immer ex-situ geprüft, wobei Defekte (z.B. Poren, Risse etc.) erst nach Prozessabschluss entdeckt werden. Übersteigen Anzahl und/oder Abmessung die vorgegebenen Grenzwerte für diese Defekte, so kommt es zu Ausschuss, was angesichts sehr langer Bauprozessdauern äußerst unrentabel ist. Eine Schwierigkeit ist dabei, dass manche Defekte sich erst zeitverzögert zum eigentlichen Materialauftrag bilden, z.B. durch thermische Spannungen oder Schmelzbadaktivitäten. Dementsprechend sind reine Monitoringansätze zur Detektion ggf. nicht ausreichend.
Daher wird in dieser Arbeit ein Verfahren zur aktiven Thermografie an dem AM-Prozess Laser Powder Bed Fusion (L-PBF) untersucht. Das Bauteil wird mit Hilfe des defokussierten Prozesslasers bei geringer Laserleistung zwischen den einzelnen gefertigten Lagen unabhängig vom eigentlichen Bauprozess erwärmt. Die entstehende Wärmesignatur wird ort- und zeitaufgelöst durch eine Infrarotkamera erfasst. Durch diese der Lagenfertigung nachgelagerte Prüfung werden auch zum Bauprozess zeitversetzte Defektbildungen nachweisbar.
In dieser Arbeit finden die Untersuchungen als Proof-of-Concept, losgelöst vom AM-Prozess, an einem typischen metallischen Testkörper statt. Dieser besitzt eine Nut als oberflächlichen Defekt. Die durchgeführten Messungen finden an einer eigens entwickelten L-PBF-Forschungsanlage innerhalb der Prozesskammer statt. Damit wird ein neuartiger Ansatz zur aktiven Thermografie für L-PBF erforscht, der eine größere Bandbreite an Defektarten auffindbar macht. Der Ansatz wird validiert und Genauigkeit sowie Auflösungsvermögen geprüft. Eine Anwendung am AM-Prozess wird damit direkt forciert und die dafür benötigten Zusammenhänge werden präsentiert.
In this study, we present an enhanced deep learning framework for the prediction of porosity based on thermographic in-situ monitoring data of laser powder bed fusion processes. The manufacturing of two cuboid specimens from Haynes 282 (Ni-based alloy) powder was monitored by a short-wave infrared camera. We use thermogram feature data and x-ray computed tomography data to train a convolutional neural network classifier. The classifier is used to perform a multi-class prediction of the spatially resolved porosity level in small sub-volumes of the specimen bulk.
The prediction of porosity is a crucial task for metal based additive manufacturing techniques such as laser powder bed fusion. Short wave infrared thermography as an in-situ monitoring tool enables the measurement of the surface radiosity during the laser exposure. Based on the thermogram data, the thermal history of the component can be reconstructed which is closely related to the resulting mechanical properties and to the formation of porosity in the part. In this study, we present a novel framework for the local prediction of porosity based on extracted features from thermogram data. The framework consists of a data pre-processing workflow and a supervised deep learning classifier architecture. The data pre-processing
workflow generates samples from thermogram feature data by including feature information from multiple subsequent layers.
Thereby, the prediction of the occurrence of complex process phenomena such as keyhole pores is enabled. A custom convolutional neural network model is used for classification. Themodel is trained and tested on a dataset from thermographic in-situ monitoring of the manufacturing of an AISI 316L stainless steel test component. The impact of the pre-processing parameters and the local void distribution on the classification performance is studied in detail. The presented model achieves an accuracy of 0.96 and an f1-Score of 0.86 for predicting keyhole porosity in small sub-volumes with a dimension of (700 × 700 × 50) μm3. Furthermore, we show that pre-processing parameters such as the porosity threshold for sample
labeling and the number of included subsequent layers are influential for the model performance. Moreover, the model prediction is shown to be sensitive to local porosity changes although it is trained on binary labeled data that disregards the actual sample porosity.
Great complexity characterizes Additive Manufacturing (AM) of metallic components via laser powder bed fusion (PBF-LB/M). Due to this, defects in the printed components (like cracks and pores) are still common. Monitoring methods are commercially used, but the relationship between process data and defect formation is not well understood yet. Furthermore, defects and deformations might develop with a temporal delay to the laser energy input. The component’s actual quality is consequently only determinable after the finished process.
To overcome this drawback, thermographic in-situ testing is introduced. The defocused process laser is utilized for nondestructive testing performed layer by layer throughout the build process. The results of the defect detection via infrared cameras are shown for a research PBF-LB/M machine.
This creates the basis for a shift from in-situ monitoring towards in-situ testing during the AM process. Defects are detected immediately inside the process chamber, and the actual component quality is determined.
Laser powder bed fusion is one of the most promising additive manufacturing techniques for printing complex-shaped metal components. However, the formation of subsurface porosity poses a significant risk to the service lifetime of the printed parts. In-situ monitoring offers the possibility to detect porosity already during manufacturing. Thereby, process feedback control or a manual process interruption to cut financial losses is enabled.
Short-wave infrared thermography can monitor the thermal history of manufactured parts which is closely connected to the probability of porosity formation. Artificial intelligence methods are increasingly used for porosity prediction from the obtained large amounts of complex monitoring data. In this study, we aim to identify the potential and the challenges of deep-learning-assisted porosity prediction based on thermographic in-situ monitoring.
Therefore, the porosity prediction task is studied in detail using an exemplary dataset from the manufacturing of two Haynes282 cuboid components. Our trained 1D convolutional neural network model shows high performance (R² score of 0.90) for the prediction of local porosity in discrete sub-volumes with dimensions of (700 x 700 x 40) μm³.
It could be demonstrated that the regressor correctly predicts layer-wise porosity changes but presumably has limited capability to predict differences in local porosity. Furthermore, there is a need to study the significance of the used thermogram feature inputs to streamline the model and to adjust the monitoring hardware. Moreover, we identified multiple sources of data uncertainty resulting from the in-situ monitoring setup, the registration with the ground truth X-ray-computed tomography data and the used pre-processing workflow that might influence the model’s performance detrimentally.
Defects are still common in metal components built with Additive Manufacturing (AM). Process monitoring methods for laser powder bed fusion (PBF-LB/M) are used in industry, but relationships between monitoring data and defect formation are not fully understood yet. Additionally, defects and deformations may develop with a time delay to the laser energy input. Thus, currently, the component quality is only determinable after the finished process.
Here, active laser thermography, a nondestructive testing method, is adapted to PBF-LB/M, using the defocused process laser as heat source. The testing can be performed layer by layer throughout the manufacturing process. We study our proposed testing method along experiments carried out on a custom research PBF-LB/M machine using infrared (IR) cameras.
Our work enables a shift from post-process testing of components towards in-situ testing during the AM process. The actual component quality is evaluated in the process chamber and defects can be detected between layers.
Quality assessment of components produced by metal based additive manufacturing (AM) technologies such as laser powder bed fusion is rising in importance due to the increased use of AM in industrial production. Here, the presence of internal porosity was identified as a limiting factor for the final component quality. The utilization of thermography as an in-situ monitoring technique allows the determination of the part’s thermal history which was found to be connected to the porosity formation [1]. Combining the local thermal information derived from thermography with the porosity information obtained by x-ray micro computed tomography, machine learning algorithms can be utilized to predict the porosity distribution in the part. In this study, a first approach for the prediction of keyhole porosity in a cylindric specimen from AISI 316L stainless steel is presented. It is based on data augmentation using the “SmoteR” algorithm [2] to cure the dataset imbalance and a 1-dimensional convolutional neural network.
[1] C.S. Lough et al., Local prediction of Laser Powder Bed Fusion porosity by short-wave infrared thermal feature porosity probability maps. Journal of Materials Processing Technology, 302, p. 117473 (2022)
https://dx.doi.org/10.1016/j.imatprotec.2021.117473
[2] L. Torgo et al., SMOTE for Regression. Progress in Artificial Intelligence, Chapter 33, p. 378-289 (2013)
https://dx.doi.org/10.1007/978-3-642-40669-0_33
The manufacturing of metal parts for the use in safety-relevant applications by Laser Powder Bed Fusion (L-PBF) demands a quality assurance of both part and process. Thermography is a nondestructive testing method that allows the in-situ determination of the thermal history of the produced part which is connected to the mechanical properties and the formation of defects [1]. A wide range of commercial thermographic camera systems working in different spectral ranges is available on the market. The understanding of the applicability of these cameras for qualitative and quantitative in-situ measurements in L-PBF is of vital importance [2]. In this study, the building process of a cylindrical specimen (Inconel 718) is monitored by three camera systems simultaniously. These camera systems are sensitive in various spectral bandwidths providing information in different temperature ranges. The performance of each camera system is explored in the context of the extraction of image features for the detection of defects. It is shown that the high temporal and thermal process dynamics are limiting factors on this matter. The combination of different spectral camera systems promises the potential of an improved defect detection by data fusion.
Online quality control of security relevant parts manufactured by Laser Powder Bed Fusion (LPBF) remains to be a challenge due to the highly complex process conditions. Furthermore, the influence of characteristic scan strategy parameters is not sufficiently clarified yet due to the commonly used method of single-track investigations. In this contribution, this topic is addressed by observing large 316L volume sections using in-situ melt pool monitoring by thermography in high temporal and spatial resolution. In detail, the influence of the scan angle on the melt pool geometry is investigated on. Characteristic melt pool features are extracted from the image data and analyzed using statistical methods data for altering scan angles. The results show significant changes in the melt pool dimensions and temperature distribution over the scan angle rotation. A first explanation approach is presented that connects the observed changes to phenomena of beam attenuation by metal vapor plume.
Lichtbogenbasierte Schweißverfahren wie das Metallschutzgasschweißen (MSG) zählen zu den Standardverfahren der Fügetechnik und werden in vielen Industriebereichen automatisiert unter Verwendung von Industrierobotern eingesetzt. Dabei können Schweißnahtabweichungen auftreten, die aus Änderungen der Prozessrandbedingungen und der hohen Prozessdynamik resultieren. Hier ist die Kontrolle von Schmelzbad- und Schweißnahtgeometrie für die Sicherung der Nahtqualität bedeutsam. Durch den Einsatz optischer Sensorsysteme können mit hoher zeitlicher Auflösung in-situ Informationen des Prozesszustands ermittelt werden. Dabei stellen die rauen Prozessbedingungen und die hohe Strahlungsintensität des Lichtbogens eine Herausforderung für die optischen Komponenten dar. Forschungsarbeiten am Institut für Schweißtechnik und Fügetechnik der RWTH Aachen haben gezeigt, dass durch den Einsatz einer HDR-Kamera in Kombination mit einer strukturierten Laserbelichtung gezielt geometrische Informationen des Lichtbogens und des Schmelzbads aus den Prozessaufnahmen gewonnen werden können. In dieser Arbeit wird eine parallele Schweißnaht- und Schmelzbadbeobachtung durchgeführt, wobei geometrische Informationen durch die Anwendung von Bildverarbeitungsalgorithmen extrahiert werden. Dabei wird ein nachlaufendes Sensorsystem eingesetzt, welches aus einer HDR-Kamera und einer Laserbeleuchtung besteht. Es werden diffraktive optische Elemente (DOE) zur Erzeugung von verschiedenen Laserprojektionsmustern verwendet, um sowohl eine unidirektionale als auch eine multidirektionale Prozessbeobachtung durchführen zu können. Aus den geometrischen Informationen werden Kenngrößen berechnet, anhand derer der Prozesszustand beurteilt werden kann. Es zeigt sich, dass anhand der Kenngrößenverläufe Abweichungen in der Schweiß- und Schmelzbadgeometrie und Positionierungsfehler des Roboters identifiziert werden können.
Lichtbogenbasierte Schweißverfahren wie das Metallschutzgasschweißen (MSG) zählen zu den Standardverfahren der Fügetechnik und werden in vielen Industriebereichen automatisiert unter Verwendung von Industrierobotern eingesetzt. Dabei können Schweißnahtabweichungen auftreten, die aus Änderungen der Prozessrandbedingungen und der hohen Prozessdynamik resultieren. Hier ist die Kontrolle von Schmelzbad- und Schweißnahtgeometrie für die Sicherung der Nahtqualität bedeutsam. Durch den Einsatz optischer Sensorsysteme können mit hoher zeitlicher Auflösung in-situ Informationen des Prozesszustands ermittelt werden. Dabei stellen die rauen Prozessbedingungen und die hohe Strahlungsintensität des Lichtbogens eine Herausforderung für die optischen Komponenten dar. Forschungsarbeiten am Institut für Schweißtechnik und Fügetechnik der RWTH Aachen haben gezeigt, dass durch den Einsatz einer HDR-Kamera in Kombination mit einer strukturierten Laserbelichtung gezielt geometrische Informationen des Lichtbogens und des Schmelzbads aus den Prozessaufnahmen gewonnen werden können. In dieser Arbeit wird eine parallele Schweißnaht- und Schmelzbadbeobachtung durchgeführt, wobei geometrische Informationen durch die Anwendung von Bildverarbeitungsalgorithmen extrahiert werden. Dabei wird ein nachlaufendes Sensorsystem eingesetzt, welches aus einer HDR-Kamera und einer Laserbeleuchtung besteht. Es werden diffraktive optische Elemente (DOE) zur Erzeugung von verschiedenen Laserprojektionsmustern verwendet, um sowohl eine unidirektionale als auch eine multidirektionale Prozessbeobachtung durchführen zu können. Aus den geometrischen Informationen werden Kenngrößen berechnet, anhand derer der Prozesszustand beurteilt werden kann. Es zeigt sich, dass anhand der Kenngrößenverläufe Abweichungen in der Schweiß- und Schmelzbadgeometrie und Positionierungsfehler des Roboters identifiziert werden können.
Laser powder bed fusion is used to create near net shape metal parts with a high degree of freedom in geometry design. When it comes to the production of safety critical components, a strict quality assurance is mandatory. An alternative to cost-intensive non-destructive testing of the produced parts is the utilization of in-situ process monitoring techniques. The formation of defects is linked to deviations of the local thermal history of the part from standard conditions. Therefore, one of the most promising monitoring techniques in additive manufacturing is thermography. In this study, features extracted from thermographic data are utilized to investigate the thermal history of cylindrical metal parts. The influence of process parameters, part geometry and scan strategy on the local heat distribution and on the resulting part porosity are presented. The suitability of the extracted features for in-situ process monitoring is discussed.
Avoiding the formation of defects such as keyhole pores is a major challenge for the production of metal parts by Laser Powder Bed Fusion (LPBF). The use of in-situ monitoring by thermographic cameras is a promising approach to detect defects, however the data is hard to analyze by conventional algorithms. Therefore, we investigate the use of Machine Learning (ML) in this study, as it is a suitable tool to model complex processes with many influencing factors. A ML model for defect prediction is created based on features extracted from process thermograms. The porosity information calculated from an x-ray Micro Computed Tomography (µCT) scan is used as reference. Physical characteristics of the keyhole pore formation are incorporated into the model to increase the prediction accuracy. Based on the prediction result, the quality of the input data is inferred and future demands on in-situ monitoring of LPBF processes are derived.
The appearance of irregularities such as keyhole porosity is a major challenge for the production of metal parts by laser powder bed fusion (PBF-LB/M). The utilization of thermographic in-situ monitoring is a promising approach to extract the thermal history which is closely related to the formation of irregularities. In this study, we investigate the utilization of convolutional neural networks to predict keyhole porosity based on thermographic features. Here, the porosity information calculated from an x-ray micro computed tomography scan is used as reference. Feature engineering is performed to enable the model to learn the complex physical characteristics of the porosity formation. The model is examined with regard to the choice of hyperparameters, the significance of thermal features and characteristics of the data acquisition. Based on the results, future demands on irregularity prediction in PBF-LB/M are derived.
The formation of irregularities such as keyhole porosity pose a major challenge to the manufacturing of metal parts by laser powder bed fusion (PBF-LB/M). In-situ thermography as a process monitoring technique shows promising potential in this manner since it is able to extract the thermal history of the part which is closely related to the formation of irregularities. In this study, we investigate the utilization of machine learning algorithms to detect keyhole porosity on the base of thermographic features. Here, as a referential technique, x-ray micro computed tomography is utilized to determine the part's porosity. An enhanced preprocessing workflow inspired by the physics of the keyhole irregularity formation is presented in combination with a customized model architecture. Furthermore, experiments were performed to clarify the role of important parameters of the preprocessing workflow for the task of defect detection . Based on the results, future demands on irregularity prediction in PBF-LB/M are derived.
In this study, we present an enhanced deep learning framework for the prediction of porosity based on thermographic in-situ monitoring data of laser powder bed fusion processes. The manufacturing of two cuboid specimens from Haynes 282 (Ni-based alloy) powder was monitored by a short-wave infrared camera. We use thermogram feature data and x-ray computed tomography data to train a convolutional neural network classifier. The classifier is used to perform a multi-class prediction of the spatially resolved porosity level in small sub-volumes of the specimen bulk.
For the wide acceptance of the use of additive manufacturing (AM), it is required to provide reliable testing methods to ensure the safety of the additively manufactured parts. A possible solution could be the deployment of in-situ monitoring during the build process. However, for laser powder bed fusion using metal powders (PBF-LB/M) only a few in-situ monitoring techniques are commercially available (optical tomography, melt pool monitoring), which have not been researched to an extent that allows to guarantee the adherence to strict quality and safety standards.
In this contribution, we present results of a study of PBF-LB/M printed parts made of the nickel-based superalloy Haynes 282. The formation of defects was provoked by local variations of the process parameters and monitored by thermography, optical tomography and melt pool monitoring. Afterwards, the defects were characterized by computed tomography (CT) to identify the detection limits of the used in-situ techniques.
Metallbasierte additive Fertigungsverfahren werden zunehmend industriell zur Anfertigung von komplex geformten Komponenten eingesetzt. In diesem Zusammenhang ist das Laser-Pulverbettschweißen von Metall (PBF-LB/M) ist ein weitläufig genutztes Verfahren. Im PBF-LB/M-Prozess werden lagenweise aufgetragene Metallpulverschichten selektiv mittels eines Lasers aufgeschmolzen. Die Entstehung von internen Fehlstellen (bspw. Porosität, Lunker oder Risse) während des Fertigungsvorgangs stellt ein ernstzunehmendes Risiko für die Bauteilsicherheit und somit für die weitere industrielle Etablierung des Verfahrens dar. Die Entstehung von Fehlstellen hängt eng mit lokalen Änderungen der thermischen Historie des Bauteils zusammen. Mit Hilfe von thermografischen Kameras zur Prozessüberwachung kann die thermische Historie bereits während der Fertigung erfasst werden. Damit eröffnet sich die Möglichkeit, die Entstehung von Fehlstellen anhand der thermografischen Daten vorherzusagen und somit potenziell Kosten für eine nachgelagerte Qualitätssicherung einzusparen.
In diesem Beitrag soll die Modellierung der Fehlstellenvorhersage anhand thermografischer Prozessdaten diskutiert werden. Hierbei liegt ein Schwerpunkt auf der Fragestellung, mit welcher Genauigkeit unterschiedliche Formen von Fehlstellen, im speziellen Anbindungsfehler und Keyhole-Porosität, auf lokaler Bauteilebene vorhergesagt werden können. Weiterhin werden verschiedenen Modelltypen aus dem Bereich des Maschinellen Lernens auf ihre Eignung für die Fehlstellenvorhersage verglichen. Ein weiterer zentraler Aspekt in diesem Zusammenhang ist die Untersuchung der Eingangsdaten des Modells auf ihre Relevanz für das Vorhersageergebnis.
Als Datengrundlage für die durchgeführten Untersuchungen dienen die Fertigungsprozesse von zwei identischen Haynes-282-Bauteilen (Nickel-Basislegierung), welche mit Hilfe einer im kurzwelligen Infrarotbereich arbeitenden Thermografiekamera überwacht wurden. Das Bauteildesign umfasste lokale Bereiche, in denen mit Hilfe einer Parametervariation die Entstehung von Fehlstellen forciert wurde. Um die Position und Größe der entstandenen Defekte zu quantifizieren, wurden beide Bauteile nach erfolgter Fertigung mittels Computertomografie (CT) geprüft. Im Rahmen der Datenvorbereitung für die Modellierung erfolgte eine Reduzierung der erhobenen Thermogramme zu physikalisch-interpretierbaren Merkmalen (bspw. Schmelzbadfläche oder Zeit-über-Schwellwert). Weiterhin erfolgte eine Registrierung der thermografischen Daten mit den Fehlstellen-Referenzdaten der CT, um eine exakte örtliche Überlagerung von thermischer Information und lokalem Fehlstellenbild zu erzielen. Zur Ermöglichung einer lokalen Fehlstellenvorhersage wurden die thermografischen Daten schichtweise in kleinteiligen Volumina angeordnet, welche als Eingangsgröße für die genutzten ML-Algorithmen dienten.
Die Ergebnisse der Untersuchungen zeigen, dass sich die Porosität auf Bauteilschichtebene mit einer hohen Genauigkeit vorhersagen lässt. Eine Vorhersage der Porosität auf lokaler Bauteilebene erweist sich noch als herausfordernd. Die erprobten ML-Algorithmen zeigen vergleichbare Ergebnisse, obwohl ihnen unterschiedliche Modellierungsannahmen zugrunde liegen und sie variierende Komplexität aufweisen. Mit Hilfe der erzielten Erkenntnisse eröffnet sich die Möglichkeit, Rückschlüsse auf die gewählte Prozessüberwachungshardware und Datenvorverarbeitung zu ziehen und somit langfristig die Leistungsfähigkeit von Modellen zur Fehlstellenvorhersage zu verbessern.
Laser powder bed fusion is one of the most promising additive manufacturing techniques for printing complex-shaped metal components. However, the formation of subsurface porosity poses a significant risk to the service lifetime of the printed parts. In-situ monitoring offers the possibility to detect porosity already during manufacturing. Thereby, process feedback control or a manual process interruption to cut financial losses is enabled.
Short-wave infrared thermography can monitor the thermal history of manufactured parts which is closely connected to the probability of porosity formation. Artificial intelligence methods are increasingly used for porosity prediction from the obtained large amounts of complex monitoring data. In this study, we aim to identify the potential and the challenges of deep-learning-assisted porosity prediction based on thermographic in-situ monitoring.
Therefore, the porosity prediction task is studied in detail using an exemplary dataset from the manufacturing of two Haynes282 cuboid components. Our trained 1D convolutional neural network model shows high performance (R2 score of 0.90) for the prediction of local porosity in discrete sub-volumes with dimensions of (700 x 700 x 40) μm³.
It could be demonstrated that the regressor correctly predicts layer-wise porosity changes but presumably has limited capability to predict differences in local porosity. Furthermore, there is a need to study the significance of the used thermogram feature inputs to streamline the model and to adjust the monitoring hardware. Moreover, we identified multiple sources of data uncertainty resulting from the in-situ monitoring setup, the registration with the ground truth X-ray-computed tomography data and the used pre-processing workflow that might influence the model’s performance detrimentally.
Since laser powder bed fusion (PBF-LB/M) is prone to the formation of defects during the building process, a fundamental requirement for widespread application is to find ways to assure safety and reliability of the additively manufactured parts. A possible solution for this problem lies in the usage of in-situ thermographic monitoring for defect detection. In this contribution we investigate possibilities and limitations of the VIS/NIR wavelength range for defect detection. A VIS/NIR camera can be based on conventional silicon-based sensors which typically have much higher spatial and temporal resolution in the same price range but are more limited in the detectable temperature range than infrared sensors designed for longer wavelengths. To investigate the influence, we compared the thermographic signatures during the creation of artificially provoked defects by local parameter variations in test specimens made of a nickel alloy (UNS N07208) for two different wavelength ranges (~980 nm and ~1600 nm).