PSI - Issue 28

Mihaela Iordachescu et al. / Procedia Structural Integrity 28 (2020) 39–44 Mihaela Iordachescu et al./ / Structural Integrity Procedia 00 (2019) 000–000

43

5

ammonium thiocyanate) at 50º C, while supporting a constant tensile load of 80% of the resistant capacity. The result of the test is the lifetime of the wire under these conditions; the standard (EHE, 2008) considers acceptable for prestressing use the steel-wire classes that exceed in a series of six SCC tests, the minimum and average life, respectively, of 1.5 and 4 hours.

Fig. 5. Load-time curves of: a) tensile tested – sheath-protected central strand-wire; b-d) stress corrosion tests under combination of constant load and increasing load at small strain rate in FIP environment – strand-wires with different service-damage level

In this work, three wire samples extracted from the broken strand of one of the bridge piles were SCC-CL tested in the FIP environment. Two of them came from peripheral wires of the strand sections protected and unprotected, respectively, by the polyethylene sheath, and the third, an unprotected central wire. Thus, the last two wire-samples had been permanently subjected to the pile interior environment, but the central wire-sample had the protection offered by the other six peripheral wires of the strand. The first wire-sample origin was part of one of the three peripheral strand-wires with a service-rupture that had a typical stress corrosion morphology, further discussed in the fourth section. Fig. 5b, Fig. 5c and Fig. 5d show the test results as load-time dependency. The curve depicted in Fig. 5a has been added as a reference and corresponds to a simple tensile test performed on the central wire from the sheathed strand section, not exposed to the pile environment. The SCC-CL tests illustrated in Fig. 5b and Fig. 5c were interrupted when they exceed to a significant extent the minimum required surviving time (EHE, 2008), at 48 and 24 hours respectively, though the wire-specimens were subsequently subjected to constant extension rate tensile testing (SCC-CERT) until failure, in the same environment. The strain rate in the SCC-CERT tests was three orders of magnitude lower than that used in the tensile test (1 µm/min instead of 1 mm/min) in order to guarantee that the lack of testing time did not inhibit the absorption and diffusion of hydrogen, essential for the wire embrittlement under the combined action of aggressive medium and applied load. As illustrated in Fig. 5c, the central wire, although exposed to the environment action, was not significantly altered in the 30 years of service, with its stress corrosion resistance being comparable with that of protected wire given in Fig. 5b. In contrast, the peripheral wire from the unprotected strand-section barely survived 0.5 hours in the FIP solution, when loaded to 80% of its resistant capacity (Fig. 5d). This poor resistance to stress corrosion cracking was well below the specifications of (EHE, 2008) and attributed to the existence of previous surface damage, which favored its rapid rupture by hydrogen uptake during SCC-CL testing. Thus, the SCC-CL results points to the surface state as a determining factor for the strand collapse, triggered by the assisted progressive cracking of peripheral wires. The failures of central and peripherical wires in the SCC-CERT tests (Fig. 5b and Fig. 5c) occur for loads higher than the 31 kN used in the SCC-CL, but less than the wire resistance capacity of 42 kN (Fig. 5a). The premature failure of wires was a consequence of the induced embrittlement by the FIP environment; while insufficient to cause their collapse in the SCC-CL, it was capable of anticipating the plastic exhaustion of the resistant ligament on subsequent tensioning at slow strain rate.