Materials and specimen preparation
The soil used in this experimental work was a clay soil. In this work, scanning electron microscopic (SEM) and X-ray diffraction (XRD) studies were conducted to determine the composition and contents of minerals in typical soil matrix. Many irregular and rodlike quartz grains surrounded by clay minerals can be observed under SEM scanning; the most of the grain size is between 0.001 and 0.003 mm, as shown
Figure 2.56 SEM images for typical soil specimen, (a) Is the SEM picture at a magnitude of 2000 times, (b) Presents EDS spectra and charts of elemental weight percentages of point I. Accelerating voltage= 15 kV. (Cited from Wang et al. (2015).)
in Figure 2.56a-b. The detailed mineral composition was determined by XRD analysis; from the results of the analysis, it can be seen that many clay minerals (e.g., mont- morillonite (61.52%), kaolinite (26.73%), illite (6.26%), and chlorite (5.5%)) exist in the soil matrix. Lithology of rock blocks used in the experiment was white marble, whose size is less than 5 mm.
Remolded artificial SRM specimens were used to conduct the real-time ultrasonic testing. According to the Soil Specimen Preparation Standard (GB/T 50123-1999 (MWRPRC, 1999)), when the specimens were cylindrical in shape with a diameter of 50 mm and a height of 100 mm, the threshold value for the soil matrix and rock block is 2 mm and the block diameter should less 0.1 times of specimen’s height (i.e., 10 mm). The physical and mechanical properties of the soil matrix and the rock block have been described in detail in the literature of Wang et al. (2015) and Wang and Li (2015). The specimens are prepared by the compaction test (Riicknagel et al., 2013; Wang et al., 2015; Wang and Li, 2015, 2015c). And the optimal hammer count was determined based on the relationship between the density and the number of compaction, as shown in Figure 2.57. During the preparation process, some water was added into SRM with certain RBP. Using the compaction curve, the optimal water ratio was set as 9.5%. The rock blocks and the soil matrix were mixed in a certain specified proportion, then sealed in a vessel, and allowed to stand for 24 h to eliminate the moisture gradient. Then, the required amount of soil matrix (size of soil particle <2 mm) and rock blocks (size between 6 and 8 mm) was mixed and homogenized in a mixer, according to the designed RBP. Afterward, the mixtures were poured into a steel module cylinder with a diameter of 50 mm and a height of 100 mm, and the samples were prepared using the compaction method with three layers by a standard Proctor hammer. As shown in Figure 2.57a, the density of soil matrix in SRM specimens with a RBP of 20%-50% increased with increasing count. To keep the same soil density (i.e., void ratio) in SRM specimens, we draw a dotted dash line to intersect with the curves in Figure 2.57a; the value of abscissa is determined as the optimal hammer count. As shown in Figure 2.57a, the optimal hammer count for SRM specimens was
Figure 2.57 Determination of the optimal hammer counts for specimens with different RBPs. (a) Plot of density of soil in SRM against hammer count; (b) Relationship between RBP and hammer count.
determined as 14, 16, 18, 24 and 32 times, with a RBP of 20%-50%, respectively, as shown in Figure 2.57b. Each group of specimens with a RBP (i.e., 20%, 30%, 40%, and 50%) was prepared with 12 ones for the experiment.
Brief description of the testing system
The testing system is specially designed for the test, which consists of a servo-controlled testing machine, an ultrasonic detector, a pair of ultrasonic transducers (200 kHz), and a splitting clamp specially designed for this test, as shown in Figure 2.58. The axial force is measured by stress sensors. The axial force can be determined by the load controller with a precision of 0.01 kN, at each loading point. We measure the axial deformation by a micrometer gauge with a precision of 0.001 mm. An ultrasonic detector can accurately record the wave signals with good precision, which can provide a 1000 V spike with a duration of 20 ps-20 ms. During the whole experiment, the sampling length is 1024, the sampling interval is 0.1 ps, and the arrival time is 0.05 ps for each pulse. The ultrasonic pulse velocity (UPV) can be determined using the distance between the transmitter and the receiver and the recorded travel time, as follows:
where UPV (L, /) is the ultrasonic pulse velocity through SRM specimens, L is the distance between the receiver and the transmitter, and t is the travel time of the ultrasonic wave through sample.
The most common method to determine the splitting stress is the radial splitting test, also known as the “Brazilian test”, which is considered an indirect strength test method (Akazawa, 1943; Fairhurst, 1964; ISRM, 1978; Markides et ah, 2012; Jiang et ah, 2015). The principle of the specially designed clamp is similar to the Brazilian test; the purpose of using this device is to investigate the approximate tension characteristics of SRM. During the test, a cylindrical specimen is horizontally placed between the two compression cone plates (Figure 2.58). The splitting stress applied to
Figure 2.58 The real-time ultrasonic testing system for the radial splitting experiment.
SRM specimens was determined based on the elasticity theory. The splitting tensile strength a, is calculated using the following relationship:
where Q is the axial loading, N; D is the specimen diameter, m; and L is the specimen length, m. Davies and Bose (1968) demonstrated the reliability of Eq. (2.23) based on a linear elasticity theory by the finite element method (FEM). Actually, when using Eq. (2.23), it assumes that the tested material is homogenous and isotropic. From the equation, it shows the splitting tensile strength stress distribution at the center of the disk sample, which is perpendicular to the loaded diameter until the failure of the samples. Crack initiates from the center of the sample and propagates outward along the loading direction.
Macro-meso failure mechanism of SRM under splitting loading is studied in this work. First, the samples were loaded at a constant rate of 0.3 mm/min, until the failure of the samples. During this process, the splitting stress, displacement, and ultrasonic parameters (e.g., velocity, attenuation coefficient) were recorded. After the failure of specimen, we observed the failure morphology, and studied the relationship between the UPV, splitting stress, and displacement. In addition, according to the study of Wang et al. (2015, 2016), the relationship between the crack width and the splitting stress was investigated. To obtain some significant results, we used the flowchart, as
Figure 2.59 Technical flowchart for study on the ultrasonic tensile dependency tests of SRM specimens.
shown in Figure 2.59. Apart from the macroscopic descriptions, three-dimensional (3-D) laser scanner was used to scan the fracture surface of SRM to obtain the section morphology; then, we calculate the fractal dimension of the failure surface, so as to study the mesoscopic failure mechanism. The 3-D laser scanner method has great advantages over the SEM method because the SEM method can only obtain a partial morphology of fracture surfaces (Wang et al., 2016). As a result, a Win3DS-VM device was chosen to scan the surface of the specimens (Figure 2.60). The scanning distance of single picture is 300 mm x 210 mm-100 mmx80 mm, with a scanning time of <3 s. The scanning resolution is about 0.04-1.1 mm, which can satisfy the average soil particle size of the specimens. The camera resolution of the Win3DS-VM device is 1.3 million pixels.
Figure 2.60 Win3DS-VM 3-D laser scanning device.