Correlating Porosity and Photodiode Response in LPBF Manufactured Samples using Spatial Statistics

The detection of porosity within laser powder bed fusion (LPBF) manufactured components is, and always will be an important step for quality assurance and certification.Computed tomography is the gold standard for non-destructive inspection, but the time, monetary investment, and data-handling challenges can be prohibitive, especially for larger components.The utilisation of various in-situ monitoring sensors might prove to be a fast, and cost-effective method for detecting porosity within laser powder bed fusion printed parts but further work is needed to correlate the response of these sensors with different types of defects that appear during the laser powder bed fusion process.In this work, photodiode data collected during the printing of a coupon with a thermal constriction is voxelised into an efficient format, and then the standard deviation of the data contained within each voxel is calculated to provide a measure of the process stability in 3D space.This spatial standard deviation is then shown to correlate well with pores identified using a high-resolution computed tomography scan of the coupon, showing that photodiode data can be used for the detection of certain types of porosity in laser powder bed fusion printed parts.