A correlation defect set was selected to allow results from the flow loop and pull rig to be directly compared. These defects also allow comparisons between successive passes using the flow loop, thus enabling the evaluation of sensor, magnetizer, and amplifier variation for long duration inspections.

Rather than repeating all of the defects in the pull rig defect set, 31 defects were chosen to calibrate major effects of length, width, depth, and surface angle, as shown in the following table. Similar to the secondary defect set for the pull rig, the flow loop correlation defect set is based on a two-thirds fraction of three primary variables (length, width, and depth), repeated at three surface angles with only one plan roundness. The more severe defects (all lengths at 80 percent depth; and 6-inch lengths at 50 percent depth) were not included. Otherwise, the defect set is the same as the common defects in the primary and secondary pull rig defect sets.

The location for the flow loop correlation defects, near the replaceable straight section, was selected to provide a constant tool velocity. The correlation defects were installed directly west of the replaceable section, near one of the control defects. The results from the correlation set are directly comparable to those for defects in the replaceable section. This area is above ground and relatively accessible for installation of defects should defects be added to the correlation set.

The flow loop correlation set was installed in two defect rows. Installing the correlation defects in an axial row ensures that the same set of sensors records results from all correlation defects.