Experimental study on prestressed concrete bridge girders strenghthened with external post-tensioning

An increasing number of existing bridges demand for retrofit interventions to restore or increase their load-bearing capacity considering increased traffic loads. In the case of pre- stressed concrete (PC) bridges, lower residual prestress levels can produce greater deform- ability and cracking of girders. Among strengthening techniques, external post-tensioning (EPT) deserves higher consideration due to beneficial effects such as upward deflection improving bridge serviceability. This paper presents an experimental investigation on four 1:5 scaled PC girders, including two benchmark specimens with different internal prestress levels and two strengthened specimens having the same level of internal prestress of their as-built counterparts and an EPT system. Assuming for the two EPT girders the same target prestress force as a combined outcome of both internal and external post-tensioning, the strengthened girders developed the same cracking load regardless of their initial prestress. In the post-cracking stage, a negligible residual displacement was recorded upon unloading due to a recentering action of the EPT system in its elastic range. Beyond cracking and ac- cording to increased girders flexibility, external tendons experienced an increase in peak stress, thus significantly contributing to bending moment capacity. In failure conditions, the two EPT girders resisted a similar peak load that was about 50% higher than as-built configuration, confirming the effectiveness of the system in achieving target performance levels of non conforming PC girders at both service and ultimate conditions.

During the last few years existing bridges received increasing attention in order to assess their residual service life and safety levels, thus preventing future losses due to damage, deterioration and extreme loading. In the case of prestressed concrete (PC) bridges, poor service conditions can arise from low level of residual prestress. Even if a number of tech- niques and strengthening systems have been investigated over the last two decades, limited studies have been carried out on internally prestressed girders with low residual prestress strengthened with external post-tensioning systems (EPT) (Aparicio et al., 2002; Leicht et al., 2022).

In order to shed light on advantages of EPT compared to different strengthening solutions, this paper presents an experimental investigation on four 1:5 scaled PC bridge girders, including two benchmark specimens having different internal prestress levels and two strengthened specimens with EPT. The substandard PC girder with low residual internal prestress demonstrated higher capacity levels compared to the benchmark girder with high- er prestress after EPT had been installed on both specimens.

Experimental Program

Design of experimental tests

A set of post-tensioned PC T-girders were investigated in their as built-configuration incor- porating defects (particularly, imperfect grouting and low residual prestress) and damaged tendons, in the framework of an experimental program carried out at the University of Naples Federico II (Losanno et al., 2023; Losanno et al., 2024). The different specimens had two different prestress levels, namely high (HP) and low (LP) prestress.

Within this study two additional specimens are presented to investigate the influence of EPT on the flexural behavior of two girders having HP and LP, respectively, without defective grouting nor damage. The corresponding specimens with EPT will be named as HP- EPT and LP-EPT, with internal tendons following a parabolic longitudinal profile as ob- served in real bridges.

The two EPT specimens were cast on the same day and an average cubic strength Rc = 34 MPa was obtained after 28 days of curing in accordance with experimental values measured on existing PC bridges (Miluccio et al., 2021). Prestressing steel of internal and external tendons was characterized by Young’s modulus Ep = 203,400 MPa, conventional yield strength fp,1 = 1782 MPa, and ultimate strength fp,t = 1969 MPa.

A multi-linear longitudinal profile was assumed for the EPT with two mono-strand tendons (one per side of the web girder, having the same geometrical and mechanical properties of the internal tendons), in order to provide maximum eccentricity at midspan and zero eccen- tricity at supports (Figure 1a). Two end plates for live and dead anchors were installed, while two deviators were provided at one third of the girder length on each side (Figures 1b and 1c). The EPT system was designed to increase the cracking load of the LP specimen and compare the overall performance of LP-EPT and HP-EPT girders under the same level of total prestressing action.

External post-tensioning

The experimental tests were carried out at the Department of Structures for Engineering and Architecture of the University of Naples Federico II, Italy, according to a four-point bend- ing configuration (Figure 2) with 850-mm-long central loading region. The EPT system was installed after the specimen had been positioned under the actuator just before starting the test program.

To provide both HP-EPT and LP-EPT specimens with a total jacking prestress force of around 400 kN as a combination of internal and external actions, the EPT jacking force (PEPT,i) was set equal to 100 kN (jacking force of internal tendons equal to PINT,i = 300 kN) for HP-EPT specimen and 250 kN (PINT,i = 150 kN) for LP-EPT specimen, corresponding to a single tendon load of 50 kN and 125 kN, respectively.

For each tendon, a load cell at the live anchor and two strain gauges were installed for mon- itoring axial load variations during jacking operations and external loading. One strain gauge was installed at midspan and the other beyond the deviator on the side of the live anchor. Midspan deflection of the girder de was measured through a wire potentiometer at midspan plus two vertical linear variable displacement transducers (LVDTs) at supports to measure any settlement. The external load Fe applied to the specimen was measured through a load cell between the actuator and the loading frame. Three horizontal LVDTs were installed at midspan to get a local measure of experimental curvature, i.e. at the centroid of the top slab, at the centroid of the gross section and at 90 mm from the section intrados.

...Continua a leggere la memoria nel PDF in allegato (in lingua inglese).