Impact resistance capacity and degradation law of epoxy-coated steel strand under the impact load

As the main load-bearing component of the bridge structure, the cable is damaged by the impact from time to time, so it is very important to clarify its impact resistance capacity. Based on the method of the drop-weight test, this article mainly studies the degradation law of the mechanical properties of epoxy-coated steel strand (ECSS) under the impact load. Through the impact test of ECSS under different working conditions, the influence law of prestress, impact energy, initial defects and other factors on the impact resistance of the steel strand was revealed. Then, the difference of the impact resistance of ECSS and ordinary steel strand (OSS) was verified under the same impact conditions. Finally, the failure mechanism and the residual-bearing capacity of ECSS were analyzed through the secondary impact test. The results show that the initial prestress is helpful to improve its impact resistance when it is less than 0.11 fu, and exceeding this value will accelerate its failure process. The effect of impact height on the strain at the impact point of ECSS is significantly greater than that near the anchor end. ECSS has better energy absorption characteristics than OSS. The impact resistance of ECSS with initial defects is very sensitive to the impact energy. The influence on outer strand in the secondary impact is significantly higher than the central strand, and its residual tensile capacity still has 0.85 fu.


Introduction
Steel strand is extensively used in various structures such as bridges, stadiums, dams and ports. As the steel strand is the main load-bearing component of the structure, any breakage or local damage to it will have a huge impact on the whole structure [1,2]. Compared with ordinary steel strand (OSS), epoxy-coated steel strand (ECSS) cables have good durability and can be used as prestressed tendons or stay cables to improve the overall bearing capacity of the structure [3,4]. In engineering design, the main function of epoxy coating is to alleviate the corrosion problem of OSS to a certain extent and prolong its service life [5]. ECSSs are generally applicable to the scenarios where OSSs can be used, including internal cables, external cables, stay cables, rock anchors, etc. With the increasing use of the stranded cables in different structural forms and complex environments, the correlational study is gradually deepened [6,7]. When the steel strand is impacted, it will have an adverse impact on itself and its connected structure, and may even cause the collapse of the whole structure. This will be seriously threatening the safety of the structure. Therefore, the problem of the impact has attracted the attention of many scholars.
There is no unified assessment standard for the damage status of the defective or the damaged ECSS, so the bearing capacity of the damaged members cannot be determined. Therefore, their safety performance also cannot be accurately evaluated [8]. At present, in terms of external impact conditions, the state of damage of ECSS is related to the quality of impact object, prestress and impact energy; for ECSS itself, its impact resistance state is related to its material strength, structural form and other factors [9,10]. For OSS, proper prestress can increase its bending stiffness, but the impact resistance of ECSS after application of prestress has not been deeply discussed [11]. When the steel strand cable is coated with epoxy resin, its nonlinear response after impact is more complex than that of OSS, and the influence of material parameters and collaborative work on impact resistance should be taken into account [12,13]. In addition, in the actual engineering, the steel strand is usually subjected to different degrees of wear or corrosion, and may work with certain defects. Therefore, the impact resistance of defective steel strand also needs to be studied.
At present, the research on the impact of the strand is mainly focused on the OSS in the building structure. However, the impact resistance of ECSS needs to be further studied. For example, Judge conducted an experimental study on the damage mode of the steel strand which was impacted by the debris simulation bomb at different rates [14]. The influence law of the fragment velocity on the impact performance of the steel strand was compared and analyzed by finite element method. The local damage range, fragment penetration depth and monofilament failure mode of the components after the impact can be demonstrated through the simulation. This study mainly focused on the failure form and did not reveal the relationship of the stress-time of the steel strand. Ghelli and Minak have carried out low-speed and high-speed impact tests on carbon fiber-reinforced plastic (CFRP) laminates and studied the similarities and differences of the impact damage of the cable under the two loading modes [15]. It is found that the influence of size and shape of the specimen on its low-speed impact performance is more obvious than that of high-speed impact. Bonneric et al. have clarified the static, fatigue and anti-corrosion properties of the steel strands with the coating type and filling type through theoretical and experimental research. The result shows that the breaking strength of the steel strand with the filling type is 3-4% higher than that of its coating type, indicating that steel strand with the filling type has higher safety performance. In addition, because the inner part of the filament is filled with epoxy resin to form an antifriction layer, the loss caused by the friction between the steel strands can be reduced, so the anti-fatigue performance is more superior [16]. Xiang et al. conducted an experimental study on the transverse impact resistance of CFRP strand and established an improved energy profile diagram based on nonlinear geometric stiffness and impact energy. It can predict the impact-bearing capacity and the deflection of CFRP strand under single and multiple impacts. The results showed that the two impacts on the strand resulted in a 37% reduction in transverse resistance, a 30% reduction in maximum tension, a 39% reduction in transverse stiffness and a 20% reduction in energy dissipation capacity [17]. Although the materials of CFRP and steel are different, their theories and test methods can be used for reference to lay a certain foundation for the study of the impact resistance of ECSS.
Based on the above theory and method, this research carried out impact tests on ECSS with different working conditions using the drop-weight tests. The impact resistance of ECSS was studied by changing the impact energy, prestress and other factors. The relationship of strain-time under various working conditions was compared, and the stress concentration position and failure mode were analyzed. Then, the impact effects of ECSS with initial notch and without initial notch were compared to explore the influence law of initial defects on the impact resistance of the strands. By comparing the deformation and failure modes of ECSS and smooth steel strand, the advantages and disadvantages of the impact resistance of the two types of strands were determined. Finally, the secondary impact test of ECSS was carried out to reveal its impact failure characteristics and the degradation law of the residual-bearing capacity. This research can provide the key data and the design methods for the application of ECSS in the architecture and bridge engineering.

Experimental program
In order to study the impact resistance of ECSS, the impact test is mainly carried out on the steel strand through the impact test device as shown in Figure 1. The test device is mainly composed of an operating system, driving system, host frame, buffer device, lifting system, hammer lock and release equipment, hammer structure, specimen support and safety protection device. The entire test system can ensure that the hammer provides sufficient impact force. The main parameters of the impact device are shown in Table 1.
The test object of this program is 1 × 7 wires of ECSS with a nominal diameter of 15.2 mm and a total length of 1,800 mm. The twist distance (150 mm), ultimate strength (1,860 MPa), elastic modulus (200 GPa) and ultimate elongation (3.5%) were provided by the manufacturer through test and calculation. The layout of the measuring point is determined according to the test purpose and the position of wires in the cross section. In order to prevent damage caused by the friction between the strain gauge at the position of the test point and the fixed reaction steel plate, and maximize the acquisition of strain data near the end, the strain gauge is posted at about 50 mm at the end. The strain gauge at the impact position is posted at a distance of 50 mm away from the impact point to prevent the hammer from damaging the strain gauges. The details of the cable size, strain sensor layout, anchoring and tensioning methods are shown in Figure 2 [18].
In the test, the impact height of the drop hammer is controlled by the operating system, and the initial prestress is applied to the steel strand by the hydraulic jack on one side of the anchor end. The initial defect of the steel strand is generated by circular cutting at 50 mm away from the impact point (4 notches) and the anchor end (4 notches) through an electric saw to form a V-shaped notch with a width and depth of 1 mm, as shown in Figure 3. In this article, the impact resistance of ECSS was systematically studied through test and simulation. The main factors include surface treatment form of the steel strand, initial defects, impact height, prestress and other factors, as shown in Table 2. 3 Results and discussion

Effect of initial prestress on impact resistance
By changing the initial prestress, the variations of strain with time of ECSS at the anchor end and the impact point were analyzed, and the influence of the prestress on the impact resistance was also obtained. In the test, the impact height of the hammer was 0.35 m. According to the law of energy conservation, the energy was converted to 4630.5 kJ [19]. The height was fixed and initial prestress of 10, 20, 30 and 40 kN was applied, and then the peak strain in the strain-time curve of ECSS under four working conditions was selected for comparative analysis. The specific failure mode in the test is shown in Figure 4. When the prestress of 10 kN was applied, most wires of ECSS broke, accompanied by huge sound and instantaneous spark. The test phenomenon in this group was obvious, but there were still unbroken monofilaments. When the prestress of 20 and 30 kN was applied, the steel strand only had bending and no obvious damage; cable bodies both produced a phenomenon of spring with the action of the hammer under the two working conditions, but the bending deflection for prestress of 20 kN was  larger than that for prestress of 30 kN. When prestress of 40 kN was applied, ECSS showed obvious bending changes and a few wires broke. Almost all the wires of ECSS with the prestress of 10 kN were broken, so the strain-time curve increased sharply and disappeared rapidly. However, under the prestress of 20 and 30 kN, the strain-time curve of ECSS increased sharply at first, then fluctuated continuously and finally tended to be flat. This indicates that the rigidity, toughness and bearing capacity of steel strand can be improved by applying a certain amount of prestress to ECSS. From the strain-time curve (Figure 5), the peak strain of the ECSS with an initial prestress of 20 kN under the impact action was larger than that of an initial prestress of 40 kN, which indicated that the increase of prestress was conducive to reducing the overall deformation of the ECSS. The coincidence degree of the straintime curve of ECSS with the prestress of 30 kN between the anchor end and the impact point was high. Combined with the failure phenomenon, it can be concluded that initial prestress was helpful to improve its impact resistance when it was less than 30 kN (0.11f u ); combined with the strain-time curve of ECSS with the prestress of 40 kN (0.15f u ), it showed that exceeding this range will accelerate the failure prcess.
In the impact test, ECSS was in an elastic-plastic state for the wires that had not fractured yet; the deformation of the specimen is shown in Figure 6. Assuming  that the strain is uniformly distributed in the elasticplastic stage, the overall elongation of the specimen can be obtained, as shown in equation (1) [20]. The distance of the steel strand between the two anchoring ends is 1.6 m, so the elongation of ECSS can be obtained using equation (1). Based on the peak strain measured in the test, the maximum deflection of ECSS (in Figure 6) can be obtained using equation (2) [20]. When the impact height was greater than 0.35 m (4630.5 kJ), the impact load was close to the failure strength of ECSS, and the deflection of ECSS increases with the increase of prestress as shown in Figure 7. Combined with the strain-time curve in Figure 5, it can be determined that appropriate application of initial pre-tension can effectively increase the stiffness of ECSS, but excessive application is detrimental to its load-bearing performance.

Effect of energy on the impact resistance performance
In this group of tests, by changing the height of drop hammer, the stress characteristics and failure model of ECSS under the same initial prestress were studied, and the influence of impact energy on the load-carrying capacity of ECSS was determined. Based on the energy conservation principle, the specific values of impact energy under different heights of the hammer are shown in Table 3.
The initial pretension of 30 kN was fixed based on the test phenomena and data in Section 3.1.
With the increase of impact height, the damage phenomenon of ECSS becomes more and more serious, as shown in Figure 8. When the impact height was 0.15 m, only bending occurred. When the impact height was 0.25 m, only single wire broke and other wires bent. When the impact height was 0.35 m, most of the wires broke and the damage phenomenon was obvious. When the impact height was 0.45 m, there was a violent sound and all wires fractured and generated a flash of spark. During the test, it was found that different impact energy has a significant effect on the strain-time curve of ECSS. When the impact energy was small, the strain-time curve presents an approximate wave-peak shape; when the impact energy was high, the strain changed suddenly with time, and the impact energy had a greater impact on the strain at the impact point of ECSS than its anchor end, as shown in Figure 9.

Effect of defects on the impact resistance performance
Different kinds of defects are likely to occur in the actual production, transportation and installation of ECSS, which will lead to reduction of its stiffness and the occurrence of stress concentration during use [21,22]. Therefore, it is very important to study the impact performance of ECSS with initial defects. In this group, the impact test was carried out after the gap was imposed on the anchorage end and near the impact position. As mentioned above, the impact  failure characteristics of ECSS were studied when the initial pretension was fixed at 30 kN and the impact heights were 0.15 and 0.45 m. When the impact height was 0.45 m, necked phenomenon occurred in the wires at the impact position of ECSS, shear damage occurred at the part of the wires near the anchoring end and the epoxy resin coating was torn off near the fracture, as shown in Figure 10. However, when the height was 0.15 m, the whole strand bent but did not fracture. The peak strains of ECSS without initial defect and with initial defect at the anchor end and near the impact position under different impact heights are shown in Table 4. It can be seen that the peak strain of ECSS without initial defects was increasing with the increase of impact height, and its damage showed a certain ductility at the anchorage end or near the impact point. However, for ECSS with initial defect, the peak strain increased first and then decreased with the increase of impact height, and its failure at the defect location showed brittle shear form. In addition, irrespective of whether at the anchor end or near the impact point, for ECSS with and without initial defects, the peak strains of the two were close when the impact height was small, while the strain values of the two displayed opposite tendency with the increase of the impact height especially near the impact point, as shown in Figure 11. On the whole, the impact resistance of ECSS with initial defects decreased with the increase of impact height.

Comparison and analysis of impact resistance of OSS and ECSS
This group studied the impact resistance of OSS with smooth surface. The experimental scheme was to keep  the initial pretension of 30 kN unchanged and only change the impact height. The test phenomenon is shown in Figure 12. Similar to ECSS, larger bending of the strand occurred only when the impact height was lower, and the fracture of the wires increased gradually with the increase of impact height. When the impact height reached 0.45 m, all strands broke with an instantaneous spark phenomenon near the impact point. Figure 13 shows the comparison of the results of OSS and ECSS under the same test conditions. With the increase of impact height, in the elastic-plastic stage without fracture, the peak strain near the impact point and anchorage end of OSS increased by 41.6 and 54.8%, respectively, while that of ECSS was 28 and 21.3%. Similarly, the maximum deflection of the two types of steel strands can be obtained by equation (2) when the peak strain was known. The maximum deflections of OSS and ECSS at the impact point were 11.8 and 11.6 cm, respectively, while those at the anchor end were 11.1 and 10.8 cm, respectively. This shows that ECSS has better stability within the range of its bearing capacity, that is, it shows better absorption characteristics for the increased impact energy. The molecular structure of epoxy resin is dense and shows strong cohesion.
Therefore, comparing with OSS, this type of ECSS has better mechanical properties for impact resistance.

Secondary impact performance of ECSS
The impact performance of ECSS without fracture from the first impact was carried out to study its deformation resistance capacity under the secondary impact energies. The strain-time curves of non-prestressed and prestressed ECSS under the secondary impact are shown in Figure 14.
The secondary impact strain of ECSS which had applied prestress at first impact was higher than that of the first impact, indicating that the secondary impact resistance of the ECSS was reduced. The strain variation near the impact point of ECSS was more obvious than that near the anchor end. The strain for the secondary impact near the anchor end increased by 13% compared with the first impact, and the strain for the secondary impact near the impact point also increased by 10%. The height of hammer when fracture of ECSS took place was decreased from 0.45 to 0.25 m, which also indicated that the secondary impact resistance decreased significantly when it was subjected to secondary impact. By comparing the fracture phenomena of the wires with the two impacts, it was found that there were unbroken wires and smooth wires with the first impact, and the epoxy coating on the steel strand did not fall off obviously. With the second impact, the fracture notch of the wires presented at 45°and the epoxy coating was torn and detached. It showed that the defect or damage of the wires was amplified after the secondary impact, so the fracture section of the wires under the impact energy was irregular, but the impact energy could still be dissipated by its deformation.
Under the secondary impact, the bearing capacity of ECSS was lower than that after the first impact, and the stress value was also larger when it was not broken. The ductility of the wires at the secondary impact was lower than that of the wires at the first impact. By observing the section where the wires were damaged and broken, it was found that there were three types of failure mode of ECSS.
(1) Shear failure. It mainly occurs near the anchor end. The section of each strand was basically flat and smooth. The connection between the strand and anchor was broken due to excessive deflection after multiple impacts, as shown in Figure 15(a). (2) Shrinkage fracture. The stress at the impact position was transmitted along the longitudinal direction of ECSS, which led to excessive deflection and contraction when the stress rebounded due to impact, as shown in Figure 15(b). (3) Matrix destruction. It mainly occurred at the upper part of ECSS in the twisting direction which was completely damaged after being impacted and its section displayed an irregular wear shape in the cross section. The notch was about 45°c racked due to excessive tensile stress, and the fracture phenomenon at the impact point was more serious, as shown in Figure 15(c).
When ECSS is impacted, it will produce bending deformation. At that time, the wires at the impact position will undergo plastic deformation. Static tensile test was conducted on the residual tensile strength of the middle and outer wires of ECSS after the secondary impact. The test results are shown in Table 5 ( Figure 16).
The results of the analysis show that the fracture of ECSS occurs at the impact position after the secondary impact. Compared with the unimpacted ECSS, the tensilebearing capacity of ECSS after the secondary impact still retains a high residual strength, up to more than 85%. The residual strength of the outer wires of ECSS was lower than that of the central strand after the secondary impact. With the increase of impact energy, the residual strength of the outer wire decreased more obviously than that of the central wire.

Conclusion
In this article, the impact resistance of ECSS was studied through experiments, and the effects of impact energy, initial pretension, initial defects and epoxy coating on the  Figure 16: Residual-bearing capacity of ECSS after the secondary impact.
transverse bearing performance of steel strand were analyzed. The variation of strain-time curve and residualbearing capacity of ECSS after secondary impact were revealed. The main conclusions were as follows: 1) Compared with the performance of the non-prestressed ECSS, the initial pretension reduces the strain at both the anchorage end and the impact position and also reduces the impact deflection of ECSS. When the initial prestress is less than 0.11f u , it helps to improve its impact resistance capacity; exceeding this value will accelerate its failure process.
2) The effect of impact energy on the strain at the impact point of ECSS is significantly greater than that near the anchor end. With the increase of impact energy, the impact resistance performance of ECSS with initial defects decreases significantly. 3) Compared with OSS, ECSS has better stability which shows better absorption characteristics for the increased impact energy, and the impact-bearing capacity of ECSS is more obvious when the impact energy is higher. 4) The strain of ECSS under the secondary impact is larger than that without prestress. After the secondary impact, ECSS still has strong bearing capacity, the impact of the outer wire is higher than the central wire and its residual tensile-bearing capacity can still reach 0.85f u .