Using dFL to Detect Power Deviations and Porosity in Additive Manufacturing скачать в хорошем качестве

Using dFL to Detect Power Deviations and Porosity in Additive Manufacturing

High-frequency LPBF sensors record everything—laser commands, melt pool signals, X-ray CT—but turning that into real process intelligence is tough. In this demo, we load a NIST Laser Powder Bed Fusion dataset into dFL by Sophelio and show how to go from raw layer logs to an autolabeled anomaly dataset. You’ll see how to: Visualize command vs actual laser position and power for machine fidelity Use custom graphs like a 2D laser path map, XCT layer-trend, and histogram heatmap Build Python-based autolabelers for power deviation and XCT porosity Run bulk autolabeling across 250+ layers and export clean labels for QA or ML If you work in metal additive manufacturing, process monitoring, or LPBF research, this walkthrough is a practical pattern for scaling from one-off plots to a fully labeled dataset. Learn more about dFL: https://dfl.sophelio.io