PSI - Issue 84

Michele D’Amato et al. / Procedia Structural Integrity 84 (2026) 1175–1182

1179

3.3. Prestressing steel

In case of prestressing steel, a sample consisting of 3918 records is recorded for all structural elements examined. In particular, these data come from design nominal values (2.8% – 111 records), from acceptance certificates (96.7% – 3789 records) and in situ tests (0.5% – 18 records). The analysis of the distribution of different types of elements, such as strands, braids, wires, cables, and bars, highlights how strands are the most frequently used elements in all structural elements examined, namely beams (43%) (Fig. 3c), cross-beams (37%), slabs (67%), and structural elements (not available) (55%). These strands typically have a design nominal tensile strength of f ptk = 1860 MPa. The analysis of the overall sample of 3918 records confirms that strands are the most frequent prestressing elements, with a percentage of 55%. By partitioning the sample by data source, strands result as the most prevalent both in design nominal values (43%, mainly used for beams) and in acceptance certificates (55%). Conversely, in the case of in situ tests, the most recurrent values refer to cables, which reach 72%. For this material as well, it is of interest to evaluate the average tensile strength and the corresponding CV by analyzing the data from acceptance certificates and in situ tests across all elements. The resulting average values are consistent with the prestressing elements examined, showing high levels of uniformity, with CVs below 4% for acceptance certificates and under 3% for in situ tests.

a

b

c

Fig. 3. Class breakdown obtained for beams (a) reinforcing steel; (b) concrete; (c) prestressing steel. (D’Amato et al., 2025)

4. Statistical Analysis Goodness-of-fit tests allow for the assessment of the degree of agreement between an observed data sample and a theoretical probability distribution, measuring the level of consistency or discrepancy (D’Agostino et al., 1986). In the present study, goodness-of-fit tests are employed to evaluate whether the empirical distributions of material strength variables can be adequately represented by assumed theoretical models. Statistical hypothesis testing determines if the null hypothesis (H 0 ) holds or should be rejected. A significance level α = 1% is adopted to verify if a variable is Statistically Significant (SS) or Not Statistically Significant (NSS). This threshold defines the probability of