PSI - Issue 62

Isabella Mazzatura et al. / Procedia Structural Integrity 62 (2024) 369–376 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

373

5

765 769

818 822

28% 27%

23% 22%

In following figures, the results are plotted in terms of relative errors (Fig. 2a-2b, Fig. 3a-b) and the comparison between stresses measured by the X-ray diffraction method (with and without residual stresses) for each load step for new and extracted wire are shown (Fig. 3a-b). Fig. 2 shows the percentage histograms of the errors. In particular, Fig. 2a considers separately the errors in the estimation of the tension for the new wire and the extracted one, comparing the actual tension _ with the measured one (thus no residual stresses are considered). In case of the extracted wire, the errors are higher and spreader than in the case of the new one, the 60% of the estimations resulted in errors higher than 30%. In Fig. 2b the measurements executed both in the new and the extracted wire are compared, including, or neglecting the influence of the residual stresses. Thus, relative errors are computed both for , and ( + 0 ) . By correcting the measure with the residual stresses, the relative errors are almost constant (note that all the values between 10% and 20% are relative to the new wire, and all the values within 20% and 30% are relative to the extracted wire). Fig. 3a) and Fig. 3b) show the correlation between the errors evaluated with eq. (3) considering (a) or not (b) the residual stresses and the effective acting stresses _ (black dotted line), for each load step. If from the Fig. 3a, seems that there is a tendency between the error (evaluated without residual stress) and the applied acting stresses, in Fig. 3b, (where residual stresses are included for the calculation of the error), it becomes almost constant. In conclusion, the results of the laboratory tests show that if the residual stresses are considered in the estimation of the actual stresses, a constant relative error is derived independently on the level of applied tension.

All Not considering res. stresses Considering res. stresses

Titolo del grafico

New wire

Extracted wire

60%

60%

Only extracted wire

Only new wire

50%

50%

40%

40%

30%

30%

20%

20%

10%

10%

0%

0%

[10%;20%) [20%;30%) [30%;40%)

[>40%]

[10%;20%) [20%;30%) [30%;40%)

[>40%]

Error

Error

a)

b)

a)

b)

Fig. 3. Relative errors for each load step (and thus for a certain level of tension ): a) neglecting residual stresses; b) considering the residual stress.

Fig. 2. Histograms of errors: a) discriminating the new and the extracted sample and neglecting residual stresses; b) considering/neglecting residual stresses.

3.3. Second phase – execution of the tests

A second phase of tests was executed given the preliminary results obtained in the first phase. Through the previous testing campaign, it was possible to state the influence of the residual stresses in the measurement of the acting tensions through the X-ray diffraction technique. If it was possible to estimate a priori the entity of residual stresses of the tendons, a sound estimation of the actual acting stressing, and therefore of the prestressing, would be obtained. As previously mentioned, the entity of the residual stresses strongly depends on the modalities by which the stabilizing phase is performed, in terms of temperature and imposed load (Caballero et al., 2011; Ruiz-Hervias et al., 2006). The main objective of the second phase of the laboratory test through the X-ray diffraction method, was to investigate the variation of residual stresses depending on:

• the localization of the measure ’s points along the element . • the geometry of the wire (diameter and cross-section). • the year of production and the influence of the grout removal.

A total of 15 samples (both wires and strands, both new and extracted from existing structures) were selected. On each a minimum of 3 measuring points were individuated (one every 20 cm) along the sample, depending on its total length and a minimum of 3 measures for each measure’s point were performed (Fig. 4). The description of the tested