Failure Projection of Corroding Bridge Post-Tensioned Tendons Considering Partitioned Attack

Post-tensioning (PT) has evolved to become an important technology for affecting integrity of large, increasingly sophisticated reinforced concrete structures. In the case of bridges, however, tendon failures resulting from wire/strand corrosion have been reported as early as two years post construction. In response to this, a recent study introduced, evaluated, and employed an analytical modeling approach that projects timing of such failures, given statistics which characterize the distribution of wire corrosion rate. These efforts all considered that corrosion penetration is normally distributed across the entire population of wires comprising all tendons. However, it has been reported that corrosion, resultant wire and strand fractures, and tendon failures can be confined to a specific location on a bridge structure as a result of variations in material properties or construction improprieties (or both). Also, the distribution of corrosion rates can differ within individual tendons because of, first, variations in grout structure and composition and, second, presence of voids and free water. The present research extends these previous efforts and addresses such situations; that is, those where the corrosion rate distribution is spatially variable. The results are discussed within the context of better assuring structural integrity for PT bridges.

Theoretical study of the prestressing strand's stress by computer modeling methods

The modern construction industry widely uses reinforced concrete structures, where high-strength prestressing strands are used. Key parameters determining strength and relaxation resistance are a steel microstructure and internal stresses. The aim of the work was a computer research of a stage-by-stage formation of internal stresses during production of prestressing strands of structure 1х7(1+6), 12.5 mm diameter, 1770 MPa strength grade, made of pearlitic steel, as well as study of various modes of mechanical and thermal treatment (MTT) influence on their distribution. To study the effect of every strand manufacturing operation on internal stresses of its wires, the authors developed three models: stranding and reducing a 7-wire strand; straightening of a laid strand, stranding and MTT of a 7-wire strand. It was shown that absolute values of residual stresses and their distribution in a wire used for strands of a specified structure significantly influence performance properties of strands. The use of MTT makes it possible to control in a wide range a redistribution of residual stresses in steel resulting from drawing and strand laying processes. It was established that during drawing of up to 80% degree, compressive stresses of 1100-1200 MPa degree are generated in the central layers of wire. The residual stresses on the wire surface accounted for 450-500 MPa and were tension in nature. The tension within a range of 70 kN to 82 kN combined with a temperature range of 360-380°С contributes to a two-fold decrease in residual stresses both in the central and surface layers of wire. When increasing temperature up to 400°С and maintaining the tension, it is possible to achieve maximum balance of residual stresses. Stranding stresses, whose high values entail failure of lay length and geometry of the studied strand may be fully eliminated only at tension of 82 kN and temperature of 400°С. Otherwise, stranding stresses result in opening of strands.

Incorporation of Stress Time Dependence into Failure Projection Modeling of Corroding Bridge Post-Tensioning Tendons

Post-tensioning (PT) has evolved to become an important technology for designing long span bridge structures. However, tendon failures resulting from wire/strand corrosion have been reported as early as 2 y post construction. In response to this, a recent study introduced and evaluated an analytical modeling approach that projects corrosion-induced wire and strand fractures and tendon failures, given statistics that characterize wire corrosion rate. This past modeling effort assumed that tensile stress in tendons was constant with time at 63% of the guaranteed ultimate tensile strength (GUTS); however, in actuality this stress decreases with time over an assumed 10,000 d (approximately 27 y) from an initial value of about 70% of GUTS to a long-term value in the range 60% to 63% of ultimate at mid-span for a simply supported beam as a consequence of long-term concrete creep and shrinkage and strand relaxation. The present study builds upon this model by incorporating this time dependence of tendon stress into the failure projection modeling. Results are discussed within the context of better understanding bridge tendon integrity issues and corrosion related failure concerns.