Numerical and experimental analysis of temperature compensation for eddy current testing during laser powder bed fusion

Abstract The quality of parts produced via laser powder bed fusion (PBF-LB/M) can be affected by microstructural defects inherent to the process. In-situ monitoring technologies are crucial for ensuring consistent component quality. Eddy current testing (ECT) offers a viable method for real-time monitoring during the PBF-LB/M build cycle. However, complex temperature fields arising during manufacturing significantly impact electrical conductivity, posing a challenge for ECT accuracy. Thus, precise temperature distribution prediction is essential for reliable flaw detection. This study develops a Finite Element (FE) model to predict the temperature field in multi-part build cycles. Scan vectors are grouped into clusters based on their timestamps, enabling the homogenization of thermal loads from multiple scan vectors. This approach allows for optimizing the trade-off between modeling accuracy and computational efficiency in a multi-part build. The study assesses the prediction accuracy required for ECT and investigates the appropriate level of homogenization needed to achieve it. Model predictions are validated through comparisons with thermography images and thermocouple measurements conducted during the manufacturing of 316L steel components.