Performance parameters for evaluating pore detection ability of computed tomography systems

Abstract

Industrial x-ray computed tomography (CT) has become an important tool for detecting and characterizing pores in workpieces produced through additive manufacturing (AM). However, a procedure to quantify the ability of a CT system to reliably detect pores, ideally in a non-destructive manner and before the actual analysis process, is still being researched. Previous approaches can either only be carried out at a great expense and destructively, or have not yet been validated in the actual case of pore detection. This work presents a potential reference object and corresponding performance parameters, the metrological structural resolution and the grey-scale resolution. To investigate their suitability for predicting the ability to detect pores, both the reference object, and special pore-containing AM samples were examined using the same CT settings. Linking the performance parameters with the pore detection rate of the AM samples showed that structural resolution, but also image sharpness, are suitable parameters.