Local thermal history can significantly vary in parts during metal Additive Manufacturing (AM), leading to local defects. However, the sequential layer-by-layer nature of AM facilitates in-situ part voxelmetric observations that can be used to detect and correct these defects for part qualification and quality control. The challenge is to relate this local radiometric data with local defect information to estimate process error likelihood in future builds. This paper uses a Short-Wave Infrared (SWIR) camera to record the temperature history for parts manufactured with Laser Powder Bed Fusion (LPBF) processes. The porosity from a cylindrical specimen is measured by ex-situ micro-computed tomography (μCT). Specimen data from the SWIR camera, combined with the μCT data, are used to generate thermal feature-based porosity probability maps. The porosity predictions made by various SWIR thermal feature-porosity probability maps of a specimen with a complex geometry are scored against the true porosity obtained via μCT. The receiver operating characteristic curves constructed from the predictions for the complex sample demonstrate the porosity probability mapping methodology’s potential for in-situ based porosity detection.

在金属增材制造 (AM) 期间，零件的局部热历史可能会发生显着变化，从而导致局部缺陷。然而，AM 的逐层顺序特性促进了原位零件体素观察，可用于检测和纠正这些缺陷，以进行零件鉴定和质量控制。挑战在于将此局部辐射测量数据与局部缺陷信息相关联，以估计未来构建中的过程错误可能性。本文使用短波红外 (SWIR) 相机记录使用激光粉末床融合 (LPBF) 工艺制造的零件的温度历史。圆柱形样品的孔隙率通过非原位微型计算机断层扫描 (μCT) 测量。来自 SWIR 相机的样本数据与 μCT 数据相结合，用于生成基于热特征的孔隙率概率图。由具有复杂几何形状的试样的各种 SWIR 热特征-孔隙率概率图做出的孔隙率预测与通过 μCT 获得的真实孔隙率进行评分。从复杂样品的预测中构建的接收器操作特性曲线证明了孔隙率映射方法在原位孔隙率检测方面的潜力。