Data interpretation for detecting voids is straightforward. Voids visually manifest as relatively darker areas in the radiograph
of a concrete sample. However, other phenomena can result in relatively darker images. For instance, relative to the lighter shadow projected by reinforcing
bars, the rest of the plate will appear darker. These situations are easily identified since the rebars project a well-defined shape. Another, more subtle
effect is due to the intensity of the radiation falling onto the plate being not only a function of the
attenuation of the materials it traverses but also of the inverse of the square of the gamma-ray path length d. If d is comparable
to the dimensions (x,z) of the plate, the radiation intensity will be noticeably higher, producing a darker zone near the point at which d is smallest.
That is the point (x0,z0) at which the line from the source perpendicular to the plate intersects the plate. In these
cases, the photographic density D(x,z) on the plate will diminish regularly as (x,z) is
farther from (x0,z0), exhibiting circles of equal photographic density.
Figure 5 shows an example where these two effects combine to make identifying a void difficult. The radiograph in figure 5 corresponds to a concrete
structure with several rebars (numbered vertical light gray bands) and a few stirrups (horizontal light gray lines). The small white rectangle with black dots,
the label above it and the thin vertical metallic segment are external reference elements.
The source-to-plate distance in this measurement was only 22 cm. In order to determine whether the darker area at the center of the radiograph corresponds
to a void or not, the photographic density along the dotted line was measured (red data points in the lower plot) and compared with the expected (calculated) photographic density.
The green and magenta lines on the graph in figure 5 are the calculated photographic densities with and without steel bars. The good agreement between the
measured data and the calculations indicates that the dark areas extending from near the center towards the borders of the radiograph are not due to a void but rather to
the variation of the gamma-ray path length among the different points.