PSI - Issue 84

Silvia Manarin et al. / Procedia Structural Integrity 84 (2026) 231–239

235

Figure 1 Boxplot of ϑ values for single wire measurements.

This first phase of the experimental campaign was dedicated to the assessment of X-ray diffraction (XRD) measurements on individual prestressing wires. The results demonstrate an overall good agreement between the stresses determined by X-ray diffraction (XRD) and the reference stresses imposed by mechanical loading. For the cleaned wires, the θ values were close to 1, with limited dispersion, as indicated by the low coefficients of variation ( CoV ). This outcome confirms the critical role of surface preparation in enhancing measurement repeatability and reducing uncertainty. In contrast, the measurements performed in the “as-received” condition exhibited slightly higher variability, which can reasonably be attributed to surface irregularities, residual oxide layers, or minor defects that locally modify the diffraction conditions and, consequently, the accuracy of the stress assessment. With respect to the comparison of the different peak-analysis algorithms, the GNR PVII and Parabola methods systematically produced results with lower scatter than those obtained using the Centroid method, particularly for the cleaned surface condition. The Centroid algorithm appeared more susceptible to surface-related artefacts, resulting in moderately higher CoV values in the as-received state. 4.2 7-Wire Strand Tests Following the analysis of the single-wire testing campaign, the configuration with an X-ray incidence angle of α = 40◦ and acquisition times of 60 s and 120 s was selected for more detailed investigation. This choice was based on the outcomes of the preliminary phase, which demonstrated that α = 40◦ offers an appropriate compromise between signal quality and operational feasibility for field applications. Subsequently, it was decided to adopt the 60 s acquisition time for all further testing phases. This decision was driven by the requirement to reflect the practical constraints anticipated during in-situ measurements on actual bridge structures, where reduced acquisition durations are essential to ensure operational feasibility and limit the overall time of on-site interventions. Maintaining a moderate acquisition time is particularly critical for the success of field-testing campaigns, especially when multiple measurement locations must be investigated under stringent time and accessibility limitations. With regard to the number of measurement points, the configuration with 11 points was selected. Increasing the number of acquired points improves the robustness of the statistical analysis and facilitates the identification and exclusion of potential outliers during post processing, without jeopardizing the reliability or repeatability of the results. This strategy led to a refinement of the experimental plan, reducing the number of configurations investigated while maintaining a high level of data quality and practical applicability. The focus was thus directed toward configurations that are most representative of real field conditions, optimizing the balance between technical requirements and operational constraints.