Dynamic tensile behavior of steel strands at different strain rates

Abstract

In engineering structures, steel strands can be subjected to dynamic loads associated with earthquakes, vehicle impacts, explosions and other special circumstances occurring in addition to static loads and resulting in higher strain rates. Therefore, it is of great significance to study the mechanical performance of steel strands subjected to dynamic loads at high strain rates. In this study, first an electrohydraulic high-speed impact test system, an electrohydraulic servo shear-stress test system, and a universal testing machine were used to conduct the dynamic tensile tests on single-wire, single-bundle, and four-bundle steel strands at ten different strain rates (2.73 × 10⁻³–10.714 s⁻¹), and the influences of the strain rate and strand diameter on the mechanical behavior of steel strands were analyzed based on the experimental data. At the same time, the wire breakage, failure modes, and fracture forms were analyzed, and the Ramberg-Osgood and Johnson-Cook constitutive models were revised to obtain a constitutive relationship which, along with the related material parameters, better describes the stress–strain behavior of the tensile steel strands subjected to dynamic tensile loads. The results show that the steel strands are clearly sensitive to strain rate, which is higher in quasi-static states and at low strain rates. At higher strain rates, the sensitivity to strain rate decreases, and the sensitivity of yield strength is higher than the ultimate strength. The strand diameter has certain influence on the steel strands performance. The ultimate strains of the single-wire and four-bundle steel strands are all smaller than that of the single-bundle steel strands. Twisting of the single-bundle steel strands weakened the wires and resulted in premature wire breakage. The failure modes of steel strands subjected to different tensile strain rates shifted from ductile to brittle. The failure modes were mainly of three types, namely, necking-milling, splitting-milling, and splitting. The Ramberg-Osgood and the Johnson-Cook constitutive models with the Cowper-Symonds strain rate parameter correction can more accurately reflect the variation in the strain hardening characteristics of steel strands with strain rate.

Introduction

Steel strands are made of hot-rolled wires, and are formed from individual high-strength steel rods through several cold-drawing operations. Subsequently, several high-strength steel wires are twisted around a long, straight central steel wire. Because steel strands exhibit high strength, low relaxation, and good mechanical properties, they are widely used in all types of civil engineering structures, as bridge cables, rock and soil anchorages, prestressed concrete strands, and spatial-grid cables.

In the past, many studies have investigated steel strands subjected to quasi-static tension, which also involved the influence of temperature [1], [2], fatigue [3], [4], corrosion [5], [6], stress relaxation [7], [8], equivalent breakage force [9]. In practical engineering applications, in addition to static loads steel strands are inevitably subjected to dynamic loads caused by earthquakes, winds, vehicles, explosions, etc. [10], [11], [12], [13], [14], resulting in a high strain rate of the steel strands. For example, existing studies have shown that the maximum strain rate in structures during earthquakes reaches about 0.1 s⁻¹, while the strain rate of steel materials under impact and explosion load is generally greater than 1 s⁻¹ [15], [16], [17]. Since steel strands are mainly used in structural engineering and their mechanical properties under seismic load are often of importance, this paper focuses on and studies the strain rate range that may be encountered under seismic actions, thus the experimental strain rate range in this paper is 2.73 × 10⁻³−10.714 s⁻¹.

At present, the static assumption is often used in engineering design to account for the dynamic effects, which clearly is inconsistent with the reality. Therefore, it is necessary to study the dynamic tensile performance of steel strands.

In recent years, as the scope of applications of steel strands has wideend, their dynamic mechanical performance has received increased attention. The research is, however, still in its infancy, and the mechanical performance of steel strands at diffrent strain rates is rarely reported. However, studies on the dynamic mechanical performance of different steel materials and similar alloys have been carried out [18]. Although differences may exist in specimen structure and material properties, it is useful to review the dynamic mechanical performance of other steels and alloys. For example, Yang et al. [19] conducted quasi-static and dynamic tensile tests at different strain rates on the Q550 steel. The results showed that the strain rate of the Q550 steel increased with the increase of strain rate, but its sensitivity to the strain rate was lower than that of the ordinary low-carbon steel. Xu et al. [20] studied the dynamic mechanical performance of the TWIP steel, and pointed out that its strength was negatively correlated with the strain rate at low strains, but positively correlated at high strains. Singh et al. [21] showed that the MP800HY high-strength steel was not sensitive to strain rate. Yu et al. [22] found through tests that the Q345 steel was a strain-rate sensitive material, but its strength did not change much at ultra-high strain rates. Huang [23], Lin [16], and Li et al. [17] conducted dynamic tensile tests on HPB235, HRB335, and HRB400 reinforcing bars, and found that the yield and ultimate strength of these three steels all increased with the increase in strain rate. The increase in the yield strength was greater than that in the ultimate strength, and the steel with lower static-load yield stress was more sensitive to the strain rate.

The constitutive model of steel refers to the mathematical expression that can describe the relationship between the real stress and strain observed experimentally. Currently, models that are commonly used to describe the stress–strain relationship of steel include the L-H model [24], the Ramberg-Osgood model [25], the Johnson-Cook model [26], and the H/V-R model [27]. In view of the mutual coupling between the strain rate effect and strain hardening effect of steel subjected to dynamic loads, many researchers suggested modifying the constitutive models by assuming the steel strain hardening expression as a function of both the strain itself and the strain rate. For example, Chen et al. [28] introduced the Cowper-Symonds model [29], which considers the effect of metal strain rate, into the H/V-R model to obtain a modified model that can better describe the change of strain hardening characteristics of the Q420 steel with the change of strain rate. Zeng et al. [30] conducted dynamic tensile tests on the HRB500E steel, modified the Cowper-Symonds and Malvar models [31] that could predict the dynamic yield stress according to the test results, and tested the Johnson-Cook model and the constitutive model that took the strain rate effect into consideration. Yang et al. [19] introduced the Cowper-Symonds model into the Johnson-Cook model to obtain a modified model that can better describe the variation of strain hardening characteristics of the Q550 steel with the change in strain rate.

This paper studies seven-wire steel strands consisting of one central wire and six side wires. The strength grade of steel is 1860 MPa. The steel belongs in a strand steel specification widely used in China at present. Therefore, studying its dynamic mechanical performance at different strain rates is of great significance for actual engineering applications. First, an electrohydraulic high-speed impact test system, an electrohydraulic servo shear-stress test system, and a universal testing machine were used to conduct the dynamic tensile tests of single-wire, single-bundle, and four-bundle steel strands at ten different strain rates (2.73 × 10⁻³–10.714 s⁻¹), and the influence of strain rate and strand diameter on the mechanical behavior of the steel strands were analyzed based on the experimental data. At the same time, the wire breakage, failure modes, and fracture forms were analyzed, and the Ramberg-Osgood and Johnson-Cook constitutive models were revised to obtain a constitutive model, which, along with the related material parameters, can better describe the stress–strain relationship of the tensile steel strands subjected to dynamic tensile loads.