Dynamic fracture propagation in steel fiber-reinforced concrete

Resumen

It is well known that concrete is one of the most widely used materials in modern construction. However, concrete structures are not only subjected to static loads, but may also be subjected to dynamic loads such as seismic, blast and impact, which can cause damage and the collapse of building structures, that results in the loss of property and lives. Steel fiber-reinforced concrete (SFRC) is concrete with steel fibers added and randomly distributed throughout the matrix, which forms a multi-phase non-homogeneous composite material. As a result of the crack resistant and reinforcing effects of the steel fibers, SFRC has been increasingly used in a wide range of engineering applications. Moreover, with the development of digital imaging technology, the use of digital image correlation analysis in the field of civil engineering provides a new approach to study the dynamic fracture of structures. Up to present, many researchers have conducted extensive research on the process of crack propagation in concrete under static loading conditions. However, the investigations on dynamic fracture behavior of concrete and SFRC are still scarce. Thus, the thesis studies the complete process of crack propagation in mode I, and mixed mode of a plain concrete and SFRCs, under various loading rates by adopting a servo-hydraulic testing machine and a drop-weight impact device combined with the digital image correlation (DIC) method. Moreover, size effect on the compressive behavior of SFRC, and tensile tests of high-strength bolted joint by using DIC are studied as well. The primary findings and conclusions are as follows: In the first place, regarding the study on the dynamic mixed-mode crack propagation in self compacting steel fiber-reinforced concrete, DIC is employed to measure the propagation speed of the observed multiple cracks under different loading rates. Specifically, beams with three different steel fiber contents, 0%, 0.4% and 0.8% in volume fraction, were casted for tests under a three-point bending configuration. A notch up to half of the beam height was made at 1/4 of the span to the central section. Tests were carried out at low loading rates (2.2 µm/s and 22 mm/s) by using a servo-hydraulic machine, under impact loading conditions (1.77 m/s and 3.55 m/s) by using a drop-weight impact device. Moreover, a high-speed camera was used to record the process of (multiple) crack propagation. Besides the shear crack started from the notch tip, the flexural cracks initiated from the central bottom surface were also formed in all beams except those with a fiber content of 60 kg/m 3 (0.8% in volume ratio) impacted at 1.77 m/s. Such high speeds of multiple-crack propagation in concrete, more than 50% of the Rayleigh wave speed, have been rarely measured in the lab. In addition, the evolution of simultaneous crack formation was recorded; the influence of fiber content and loading rate is explored. Furthermore, all three types of concrete were previously characterized through independent tests, the flexural strength, the toughness indices as well as the residual strength factors are reported. In the second place, this study is to measure the crack propagation speed in another three types of self-compacting concrete reinforced with steel fibers loaded under four different loading rates by using the strain gauge technique. Central-notched prismatic beams with two types of fibers (13 mm and 30 mm in length), three fiber volume ratios, 0.51%, 0.77% and 1.23%, were fabricated. Four strain gages were glued on one side of the specimen notch to measure the crack propagation velocity, a fifth one at the notch tip to estimate the strain rates upon the initiation of a cohesive crack and the stress-free crack. The servo-hydraulic testing machine and the drop-weight impact device were employed to conduct three-point bending tests at four loading-point displacement rates. Numerical simulations based on cohesive theories of fracture and preliminary results based on the technique of Digital Image Correlation are also presented to complement those obtained from the strain gages. In addition, the toughness indices are calculated under all four loading rates. Strain hardening (softening) behavior accounting from the initiation of the first crack is observed for all three types of concrete at low (high) loading rates. Significant enhancement in the energy absorption capacity is observed with increased fiber content. Next, the objective of this study is to explore the size effect observed in compressive tests carried out over cubes with three different edge lengths (150 mm, 80 mm and 40 mm, respectively). A servo-hydraulic machine was employed for tests performed at a quasi-static loading rate. Moreover, a series of images were recorded by using a digital single-lens reflex camera for posterior DIC analysis. The results show that the failure strains of the specimens manifest a remarkable increase with a decrease in cube size. At peak load, for a small-sized cube, the number of cracks is less, the strain distribution is more concentrated, and the maximum strain is larger. Finally, the European standard EN 1090-2:2018 establishes the requirements for the execution of steel structures, including the testing method to measure the delayed slip at constant load in bolted joints. One specification sets that the relative displacement between the joined plates from the first 5 minutes under load until 3 hours must be lesser than 2 µm in total. In this work, it is highlighted that this tolerance is too small for obtaining reliable measurements in laboratory with the extensometers commonly available. These instruments usually present a resolution of 1 µm, and a worse precision, which could be 2 µm or more and, thereby, equal or higher than the total displacement 2 µm. Alternatively, using DIC would overcome such shortcomings, the precision could reach 0.39 µm in our case.