EXPERIMENTAL STUDY ON THE PERFORMANCE OF FRP TENDON ANCHORING DEVICES IN AXIAL TENSION

: The paper presents an experimental research on anchoring devices developed for the pre-tensioning of fiber reinforced polymer tendons in the stress range between 40 and 70% of tensile strength. The technique of testing, the criteria of assessing the load capacity and the influence of preloading applied on the anchor wedges on the initial tendons slippage are described. The optimal technical configurations of the anchoring device have been obtained, including parameters regarding the necessary prior anchorage of the tendon ends based on the strength properties of the tendon bar and the requirements to avoid slippage during the tensile tests.


Introduction
A variety of existing technical solutions for the pre-tensioning of composite Fiber Reinforced Polymer (FRP) tendons [1]- [15] shows that the problem of increasing the reliability of gripping and anchoring devices is of growing importance nowadays.
The most effective way to fix the tendon bar in the test sample is to place it send in the steel sleeve [2], tube [16] or tip [3] and polymerize it with an adhesive or glass-filled epoxy resin.This process provides appropriate conditions for minimal initial deformations in the area of the reinforcing tendon fixation, and results in maximum 136 I. V. АBRAMOV et al.
Pollack Periodica 15, 2020, 3 contact area and adhesion ratio.It is proposed in [17] to pour the gripping mechanism in a liquid form through a hole into the mould that contains the previously installed reinforcing rebar, and then press it with clamping elements.Paper [18] presents the results of experiments where fiberglass rods glued into the gripping parts were subjected to uniaxial tension.The author noted that 50% of the samples under tension have been damaged in the gripping part due to the breakdown of the adhesive layer between the sample and the steel sleeve, and the remaining 50% failed in the working part and were considered appraisable for defining the tensile strength.Paper [19] gives the results of tensile strength tests of several types of FRP rebars, carried out according to the Russian national standard [16].In this test the following specimen failure pattern were noted: for small diameters (up to 5 mm), the rabar failed at its connection with the anchor device.It was concluded that the proper use of an adhesive substance would require further studies especially on the strength and adhesive properties in the solid state, as well as on the precise centering of the FRP rebars in the sleeves before the polymerization process [20].
Modern systems of steel reinforcement prestressing use collet clamps [21], which have to be constructively adapted for tensioning FRP rebars.This is primarily due to the different physico-mechanical properties (mainly of the working surface hardness) of the steel collet and FRP, low crack resistance of the polymer composite matrix material, relatively low shear strength of FRP due to the absence of reinforcement in the transverse direction.
The above-mentioned factors add up to the formation of surface microcracks in the contact zone and the destruction of the polymer composite matrix material and longitudinal fibers, including the point where reinforcement bar exits from the collet [22], [23], [24].It is possible to solve this problem by either preventing the direct contact between the steel collet and the rod by using an intermediate sleeve [10], [25], or producing the collet using materials with physic-mechanical properties close to the FRP [1], [4]- [9], [13].
Article [26] shows that an increase in the load capacity of the tension system with an intermediate sleeve can be achieved by subjecting the inner surface of the sleeve to sandblasting, thus improving its frictional properties.The experiment confirmed that the friction coefficient and contact pressure are the main parameters affecting the strength of the cylinder-rod connection.
The second way to eliminate the stress concentration in the rod when using a collet clamp is to make the collet from a softer material.For example, a comparison of the experimental results on uniaxial tension using conventional collet grips of a tensile machine and plastic wedges adapted for testing FPR [27] showed a difference in ultimate tensile capacity.It determined the limiting characteristics of 30% associated with the destruction of a rebar in the area of steel collet.
Article [28] describes a condition for keeping the rod in the collet clamp under tension: the contact pressure should be lower at the loaded end of the rod and higher at the free end.This can be achieved by using collets of various stiffness along the length [5], [6], [7], [13], or using design features [12], that enable redistribution of contact pressures in the process of rebar tensioning.One of the design features is the angle of inclination of the cone collet and the body angle in the sleeve.It should be pointed out that recommendations of Russian [4], [29] and foreign authors [10], [12] differ.

PERFORMANCE OF FRP TENDON ANCHORING DEVICES IN AXIAL TENSION 137
Pollack Periodica 15, 2020, 3 Considering the information, mentioned above, the design of the tension device for FRP reinforcement is important, and requires not only the study of issues related to the theoretical justification of the design parameters of the device, but also the technology of manufacturing, subsequent experimental studies and recommendations for using this tension system for prestressing process of concrete structures.

Description of the test method for anchoring devices
Three series of tests of devices for tensioning of glass-composite reinforcement Glass Fiber Reinforced Polymer (GFRP) were carried out.Reinforcing rods had a diameter of 8 mm, a length of 520 mm, and three different surface treatments: -set 1, -the surface treatment is formed by winding the continuous fiber on the power rod; -set 2, -the surface treatment is formed by pressing out threads of homogeneous material in the FRP rod; -set 3, -an surface treatment is formed by application of sand coating.
The appearance of the investigated samples of the GFRP rebars is presented in Fig. 1.The geometric and mechanical characteristics that were obtained from previously performed tests to comply with the requirements of GOST 31938 -2012 [16] are shown in Table I for GFRP rebars produced by different manufacturers.

Fig. 1. FRP rods appearance
Before testing, the ends of the reinforcement tendon were fixed by initial pressing of anchor wedges with F1 and F2 forces in mandrels №1 and №2 respectively (see Fig. 2), and then testing beacons (Fig. 3) were glued to the rod close to the ends of the cylinders for the determination of slippage.
Next, the cylinder with a reinforcing rod is fastened in the gripping devices of the testing machine Instron 5982.The testing configuration is presented in Fig. 3.The tension force Fp was set from 0.4 to 0.7 of the ultimate load at axial tension F ult , the speed of the plate movement was set to 10 mm/min.

PERFORMANCE OF FRP TENDON ANCHORING DEVICES IN AXIAL TENSION 139
Pollack Periodica 15, 2020, 3 After reaching a given force and holding for 60 seconds, the load was removed.The gaps ∆h1, ∆h2 were visually assessed and measured with a ruler when it was necessary.
In some cases, it was noted that the slippage of the rebar in the mandrel corresponded to an abrupt change in the tension force, which made it possible to determine the initial moment of slippage and the corresponding force Fss according to the loading diagram (Fig. 4, curve 2).
For each test, parameters regarding pre-fixing of the reinforcement and the tension characteristics were registered.

Analysis of test results
Based on the method presented in the previous section, tests on different types of anchoring devices were carried out.Typical tension diagrams of reinforcement sample sets 1, 2 and 3 are presented in Fig. 5, respectively (the initial pressing force applied on anchor wedges was in the range of 41 kN and 42 kN).
Diagrams show that the slippage of the ribbed reinforcing bars of set 1 is characterized by a sharp drop in the tension force (by 35-40%) and a lower value of the initial sliding moment force.It can be explained by the specific characteristics of manufacturing and the weaker frictional properties of the tendons surface compared to other samples sets.
The test results of samples of reinforcement bars of set 1, set 2 and set 3 are presented in Fig. 6.They include obtained values of forces at the moment of initial sliding depending on the forces of preliminary pressing of anchor wedges and their trend line.Force F corresponds to the minimum of forces F1 and F2 for each test.
Tests carried out on the device for tensioning the GFRP rods with anchor wedges with a length of 70 mm have shown that the maximum holding capacity is different when GFRP rods of a different profile are exposed to the short-term tensile load.The composite rod is fixed according to the regulatory requirements of 0.45σ in accordance with SP 295.1325800.2017'Concrete structures reinforced with polymer composite reinforcement design rules' by pre-fixing the ends of the rod and by providing compression of the polymer anchor wedges by pressing them depending on the type of FRP rod with forces as follows:

Conclusions
Results of experimental tests carried out on the developed device for placing and fixing reinforcing composite rebars and tendons have confirmed appropriate working capacity and successful avoidance of slippage during the tests in the required range of tension forces.

Fig. 4 .
Fig. 4. Axial tension diagrams of rebar: 1 -without slippage; 2 -with slippage of the rod in the gripping device

Fig. 6 .
Fig. 6.Experimentally obtained values of force at the initial moment of sliding depending on the force of preliminary pressing of anchor wedges with a length of l = 70 mm ( -series 1, -series 2, -series 3) for GFRP rods of set 1, the pressing force is 50 kN, which provides the force of the initial moment of slipping F ss = 33.3kN, or 48% of the ultimate load F = 70.32kN; for GFRP rods of set 2, the pressing force is 40 kN, which provides the force of the initial moment of slipping F ss = 31.2kN, or 51% of the ultimate load F= 60.95 kN; for GFRP rods of set 3, the pressing force is 55 kN, which provides the force of the initial moment of slipping F ss = 46.6 kN, or 72% of the ultimate load F = 65.35kN.

Table I
Geometric and mechanical characteristics of the tested GFRP rebars