PSI - Issue 64

A. Cagnoni et al. / Procedia Structural Integrity 64 (2024) 951–958 953 Alessandro Cagnoni, Pierluigi Colombi, Marco A. Pisani, Tommaso D’Antino / Structural Integrity Procedia 00 (2019) 000–000 3 Shi et al. (2022) designed basalt FRP wedges. In all these studies with non-metallic wedges, specimens failed due to local fiber rupture in correspondence of the anchor zone. Concerning the clamp anchors, Ye and Guo (2011) conducted an experimental campaign to study the effects of the bonded length, pre-tightening forces, and number of clamping bolts. In most of the tests conducted, failure occurred due to the slippage of the rod within the anchor. Burningham et al. (2014) designed an unibody clamp after a series of tests performed to optimize the number and the diameter of bolts, the bonded length, and the thickness. Specimens failed when reaching their ultimate tensile strength. 2.2. Bonded anchors Bonded anchors are the most frequently studied and utilized anchoring system due to their (relatively) low manufacturing cost, good stress distribution, and high mechanical performance. Typically, bonded anchors consist of a steel pipe filled with an inorganic resin or cement-based grout to embed the tendon. A specific type of bonded anchor is the composite anchor, which incorporates mechanical parts (e.g., wedges) and filler material. The load-carrying capacity of bonded anchors depends on the friction and chemical adhesion. Compared to mechanical systems, bond anchors require a higher installation and curing time and a longer embedded length. Additionally, they pose difficulties if the system needs to be modified during its service life. The performance of bonded anchors depends on the geometry of components, properties of filler material, and embedded length (Mei et al. (2020)). Different studies have been conducted to investigate the behavior of various geometries and filler materials. Zhang and Benmokrane (2004) designed an anchorage system made of a cement based filler for single or multiple rod applications. They discovered that the tensile behavior of single and multiple rods differs because of the interaction between the rods. Saeed et al. (2020) conducted an experimental campaign to evaluate the behavior of their anchor system based on expansive grout. The campaign comprised tensile, cyclic, and sustained loading tests. The authors observed that a large cross-sectional area of tendons limited grout deformation and increased the load-slip stiffness and tensile strength. Jia et al. (2022) performed tensile tests on various filler materials, including epoxy resin, epoxy resin with added quartz sand, and ultra-high performance concrete dry-mix. The epoxy-based fillers exhibited a good performance, while the specimens filled with concrete failed prematurely due the tendon slippage. Cai et al. (2015) and Mei et al. (2020) designed two different bonded anchors combining the mechanical wedge-barrel anchor with an epoxy-resin filler. Composite systems exhibited a more complex stress distribution along the bonded length compared to the bonded anchors without mechanical parts. The anchorage efficiency η of the systems mentioned above is compared in Table 1, where η is defined as the ratio between the tensile capacity of the entire system ( F u ) and that of the tendon only ( F u,FRP ). The results showed that, in general, bonded anchors work better than mechanical anchors. Indeed, excluding the results obtained by Jia et al. (2022) with a high concrete mix, the anchor efficiency of bonded anchors was always near or higher than 100%. Good results were obtained with wedge-barrel anchors made of steel, while low performances were obtained by anchors not made with metal alloys. 3. Stress level applied to the tendons Among different types of polymeric composites, CFRP tendons exhibit the highest resistance to fatigue and creep phenomena. However, the American standard ACI 440.4R-04 (ACI Committee 440 (2004)) limits the maximum stress applicable to tendons to avoid ruptures resulting from creep and fatigue phenomenon. Specifically, the suggested limits are 0.65 and 0.60 of the design tensile strength ( f pu ) at jacking and immediately following the load transfer, respectively. The Canadian standard CSA S806-02 (Canadian Standards Association (CSA) (2002)) provides limits equal to 0.70 f pu at jacking and 0.60 f pu at the load transfer. Upon analyzing the available studies in literature, the stress limits provided by the two standards seem overly restrictive. 2.3. Comparison