Browsing by Subject "powder bed fusion additive manufacturing"

Now showing 1 - 2 of 2

Machine Learning for Defect Detection for PBFAM Using High Resolution Layerwise Imaging Coupled with Post-Build CT Scans
(University of Texas at Austin, 2017) Petrich, Jan; Gobert, Christian; Phoha, Shashi; Nassar, Abdalla R.; Reutzel, Edward W.
This paper develops a methodology based on machine learning to detect defects during Powder Bed Fusion Additive Manufacturing (PBFAM) processes using data from high resolution images. The methodology is validated experimentally using both a support vector machine (SVM) and a neural network (NN) for binary classification. High resolution images are collected each layer of the build, and the ground truth labels necessary for supervised machine learning are obtained from a 3D computed tomography (CT) scan. CT data is processed using image processing tools—extended to 3D—in order to extract xyz position of voids within the component. Anomaly locations are subsequently transferred from the CT domain into the image domain using an affine transformation. Multi-dimensional features are extracted from the images using data surrounding both anomaly and nominal locations. Using cross-validation strategies for machine learning and testing, accuracies of close to 90% could be achieved when using a neural network for in-situ anomaly detection.
Selection and Installation of High Resolution Imaging to Monitor the PBFAM Process, and Synchronization to Post-Build 3D Computed Tomography
(University of Texas at Austin, 2017) Morgan, Jacob P.; Morgan, John P. Jr.; Natale, Donald J.; Smith, Robert W.M.; Mitchell, Wesley F.; Dunbar, Alexander J.; Reutzel, Edward W.
Industrial applications of PBFAM continue to expand, and there is a growing interest in the use of sensors to monitor the build process. Sensor data collected during the build process provides insight into process physics and may also lead to a reduction in overall fabrication time and cost by offering an alternative to extensive post-build nondestructive inspection for quality control. Ultimately, sensor data may serve as feedback for real-time control systems that automatically repair flaws before they are buried by subsequent layers. In this work, high resolution images are explored as a means of monitoring the PBFAM build process inside a 3D Systems ProX320. Key design considerations for camera selection and integration are discussed. Methods and algorithms are developed to calibrate and map layer-wise imagery to laser scan vectors. Images are stacked and exported to standardized 3D data formats to enable easy inspection and comparison to post-build 3D computed tomography (CT) volumes.