Skip to Main Content
Radiography is an effective method for nondestructive inspection of thick metal components. However, complete diagnosis of internal defects is a difficult task that needs reliable computer assistance. In particular, fully 3-D reconstruction would be a very valuable tool for the expert, but it is not so clear whether digitized radiographs can be made suitable for reconstruction. We show that this is indeed the case by studying the radiographic image formation process in the context of primary loop piping inspection in pressurized water nuclear reactors. In a more general way, our findings apply to the testing of thick metal pecimens with simple, known geometry. Based on justified simplifications and approximations, we demonstrate that it is possible to process the raw data to closely fit the conventional line-integral projection model. More specifically, we provide a full processing procedure that includes (i) a criterion for subsampling the data without loss of pertinent information, (ii) a novel field-flattening algorithm, and (iii) a calibration method that requires minimal knowledge about the data acquisition parameters. The actual 3-D reconstruction issue is addressed in another paper whose results further validate the present work.