PSI - Issue 84

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

236

In this phase, 201 measurements were performed on the outer wire of a 7-wire strand (nominal cross-section of a single wire: 15.2 mm 2 ), adopting the optimized acquisition parameters identified during the single wire testing phase. All measurements in this phase were performed on cleaned wires to ensure optimal surface conditions and minimize variability due to surface irregularities. Despite the increased geometric complexity of the strand and the application of higher stress levels (800, 1000, and 1200 MPa), the XRD measurements exhibited relatively stable behavior. The following Table 2 reports the mean value, standard deviation ( SD θ ), and coefficient of variation ( CoV ) for each combination of parameters. The θ values recorded for the strand tests were consistently greater than 1, indicating a systematic underestimation of the actual applied stress by the XRD system. Several factors may have contributed to this behavior. The geometric complexity of the strand can influence diffraction conditions and hinder precise measurement alignment. In addition, internal residual stresses—particularly those associated with the helical arrangement of the outer wires—may affect the strain state detected by the XRD system. Finally, local curvature and non-uniform load distribution along the wire surface introduce further challenges, making accurate stress evaluation more difficult. Nevertheless, the dispersion of the measurements, as indicated by the Coefficient of Variation ( CoV ), remained within acceptable bounds, confirming that under controlled surface conditions, the XRD system provides good repeatability even when applied to complex geometries. The systematic deviation observed, however, suggests that dedicated calibration factors or correction models may be required to improve the accuracy of XRD measurements on full-scale prestressing strands. A comparison between the two testing phases shows that the XRD system achieves higher accuracy and lower scatter on single clean wires than on strand wires. Measurements on the 7-wire strand tend to underestimate the applied stress, although repeatability remains acceptable. This behavior reflects the added complexity of strand geometry and the potential presence of residual stresses, both of which make highly accurate measurements more challenging.

Table 2 Results in terms of mean value of μ θ , standard deviation ( SD θ ) and coefficient of variation ( CoV ) for the 7-wire configurations.

Algorithm Cleaning Angle (°)

N. Points

Time (s)

N_measure

μθ

SD_θ 0.14 0.18 0.16 0.09 0.12 0.16 0.14 0.12 0.18 0.19 0.18 0.12

CoV 0.09 0.12 0.06 0.08 0.11 0.09 0.08 0.12 0.13 0.12 0.08 0.1

CENTROID 1 CENTROID 1 CENTROID 1 CENTROID 1

40 40 40 40 40 40 40 40 40 40 40 40

9 9

60

15 15 33 15 15 33 15 15 33 4 4

1.49 1.49 1.56 1.55 1.42 1.46 1.53 1.53 1.55 1.52 1.54 1.56

120

11 11

60

120

GNR PVII GNR PVII GNR PVII GNR PVII

1 1 1 1

9 9

60

120

11 11

60

120

PARABOLA 1 PARABOLA 1 PARABOLA 1 PARABOLA 1

9 9

60

120

11 11

60

120

4