Comparison of field measured long-term prestressing force loss to design code equations for PSC girder

Prestressed concrete (PSC) is a structural system in which compressive forces are intentionally applied through high-strength tendons to counteract tensile stresses and prevent cracking. Although PSC has been widely adopted in large-scale constructions such as bridges, buildings, and tanks due to its long service life, the accurate long-term prediction of prestress (PS) losses remains a critical issue, as time-dependent reductions in prestressing force—caused by concrete creep, drying shrinkage, anchorage slip, and steel relaxation—can compromise structural integrity and durability. Despite its importance, most previous studies have relied on short-term laboratory tests or numerical simulations, and long-term field data—especially for ultra-high-strength tendons—remain extremely limited. In this study, PS losses were monitored continuously over approximately 10 years in post-tensioned PSC beams embedded with steel tendons of three different tensile strengths: 1860 MPa, 2160 MPa, and 2400 MPa. The measured strain data were compared against prediction equations provided in major design codes to assess their applicability to modern high-strength tendons. The experimental results demonstrated that tendons with higher tensile strengths (2160 MPa and 2400 MPa) exhibited distinct long-term PS loss characteristics, including smaller loss rates and different seasonal and spatial variation patterns, compared to lower-strength tendons. The existing code-based prediction equations, originally developed for 1860 MPa tendons, significantly underestimated the long-term losses in high-strength tendons, highlighting the need for model refinement. Based on the long-term measurements, the study proposes modified PS loss prediction equations that better reflect the time-dependent behavior of ultra-high-strength tendons. These findings not only offer rare empirical data for validation but also provide critical insight for improving code-based PS loss predictions, with the potential to influence future revisions of design provisions for PSC structures.