PSI - Issue 64

Isabella Mazzatura et al. / Procedia Structural Integrity 64 (2024) 114–121 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

117

4

3. Results In the following, the results are given firstly in relation to the influence of the tests’ execution modalities and then to the cover thickness. Finally, the overall accuracy of the NDT is obtained considering the main output of the analysis. 3.1. Influence of the tests’ execution modalities The errors are calculated considering as reference line the base of the walls (0 location in Fig. 3a and d), therefore a positive error means an underestimation of the actual location of the duct (lower location than the actual one), while a negative error means an overestimation (upper location than the actual one). Table 1 shows the main statistical parameters of the location error of the ducts. Both the mean and the median, and both the standard deviation and the interquartile range are given. Fig. 4 shows the histograms of the error considering a bins’ amplitude of 1 cm. The distributions are approximately normal for every configuration, and also the statistical parameters are similar for every configuration.

Table 1. Statistical parameters of the location error for each Configuration. Configuration Wall Concrete Cover Mean (cm)

St. Dev. (cm)

Median (cm)

IQR (cm)

I

All All All All

All All All All

0.56 0.87 0.55 0.52

1.34 1.56 1.67 1.55

0.5 1.0 0.5 0.5

2.0 2.0 2.5 2.5

II

III IV

a) d) Fig. 4. Histograms of the errors of a) Configuration I, b) Configuration II, c) Configuration III, and d) Configuration IV. b) c)

The potential influence of the test’s execution is assessed by relating the errors and the average height at which the duct is located. As Fig. 5 shows, all the configurations result in a very similar regression line. The maximum accuracy of the instrument is at a height between 131 cm and 148 cm, and a clear trend can be identified. Therefore, the relative height of the duct into the specimen seems to be a factor of influence even stronger than the modality in which the test is executed.