Sandstone specimen preparation

The bedding sandstone used in this experiment is wood grain sandstone gathered from a mining engineering in Yunnan, with small and dense mineral particles. The rock is overall pale yellow in the dry state, and the bedding structure is distinguishable by color with the pore strips distributed along the bedding direction. Structurally, the rock belongs to transversely isotropic material. Appraised by the Xi’an Mineral Resources Supervision and Testing Center of the Ministry of Land and Resources, the mineral composition of the rock specimen is composed of 48.6% quartz, 28.4% potassium feldspar, 11.1% illite, 6% dolomite, 2.9% talc, 2.5% calcite and 0.5% diopside. The specimen exhibited a typical crumb structure containing mineral particles, fillers and partial pores.

In order to comprehensively and effectively study the development of the mechanical properties, the damage response and failure mode of the bedding rock under F–T cycles, the specimen was processed into two directions ((alpha { = }0^circ) and (alpha { = 9}0^circ)) according to the original bedding direction of wood grain sandstone. Wherein, the specimen vertical to the loading direction ((alpha { = }0^circ)) was named V-specimens (Fig. 1); The specimen parallel to the loading direction ((alpha { = 9}0^circ)) was named P-specimens (Fig. 2).

Figure 1

Vertical bedding specimens.

Figure 2

Parallel bedding specimens.

In addition, according to the previous research results, the wood grain sandstone was processed into two kinds of test specimens with different sizes to study the static and dynamic mechanical properties of the bedding sandstone damaged by F–T cycles systematically:

  1. 1.

    Static compression specimens: According to «GB/T 50266-2013 Engineering Rock Mass Test Method Standard»22, the root was divided and polished into a cylindrical standard test piece with a height to diameter ratio of 2:1 (Fig. 3). After processing, the test piece was 100 mm high and 50 mm in diameter; the parallel error of the two end faces was less than 0.05 mm; the vertical error between the end face of the test piece and the axis of the cylinder was less than 0.25°.

  2. 2.

    Dynamic compression specimens: Following the methods and dynamic experimental research results recommended by the International Society of Rock Mechanics (ISRM)23, the wood grain sandstone was processed into a standard cylinder of Φ96 mm × 48 mm (Fig. 4). The parallel error of both ends was controlled within 0.5 mm, and the surface flatness was controlled within 0.2 mm.

Figure 3

Static compression test specimens: (a) vertical bedding; (b) parallel bedding.

Figure 4

Dynamic compression test specimens: (a) vertical bedding; (b) parallel bedding.

F–T treatment of specimen

Firstly, the appearance, volume, quality and ultrasonic signals of the test pieces were screened and measured. The test pieces with large discrete physical indexes were excluded from the following analyses. Subsequently, the screened specimen was placed in a 101-2ASB type electric hot air blower for blast drying at a constant temperature of 107 ± 1 °C for 24 h. Then the specimen was naturally cooled and weighed. After weighing, the drying process was repeated at the same temperature and time duration, and the mass was weighed again after cooling. We repeated the drying–cooling–weighing processes until the mass difference of two measurements was less than 0.1%. Natural cooling was carried out in the drying cabinet. The blast was maintained, and the temperature was turned off to prevent the specimen from absorbing water. Then the dried specimen was put into boiling water to get saturated. The temperature was set to be 100 °C. In the boiling process, the liquid level was ensured to be higher than the specimen constantly. Then, the specimens were kept boiling for 6 h. After the specimen was naturally cooled in the water, it was taken out and put it into the water tank for storage. Finally, F–T test was carried out on the saturated specimen in JCD-40J automatic building material F–T cycle test machine with one F–T cycle procedure set as follows: 4-h freezing in air after the test chamber temperature reaching − 20 °C, followed by a 4-h thawing in water at 20 °C (Fig. 5). During the F–T process, no water existed in the cavity during freezing, and warm water (+ 20 °C) was later injected during melting. The water level was always higher than the specimen, and the temperature change followed almost the same path during each F–T cycle respectively.

Figure 5

Temperature–time schematic diagrams of F–T cycles.

And in order to control the systematic uncertainties, 3 mechanical specimens and 2 mesoscopic observation specimens were tested under each experimental condition. For the static pressure test, the number of F–T cycles of the two types of bedding specimens was set to 0, 10, 20, 30, and 40 times.

Experiment methods

Static compression test

As the most conventional mechanical test, the static compression test can provide strength and deformation parameters. These data can provide fundamental reference for the mechanical properties of rock under various loading environments and loading conditions. Therefore, the static compression tests of the two bedding sandstones were carried out in the HYY type electro-hydraulic servo pressure tester at a loading mode of 20 kN/min. The HYY type electro-hydraulic servo pressure test system consists of a pressure table, a hydraulic pump, and a microcomputer control system. The loading limit of the press was 2,000 kN, and the complete stress–strain curve of the specimen was obtained here directly.

When the static compression test was carried out based on the hydraulic servo tester, after placing the sample on the pressure-bearing table, there was a gap between the sample and the upper platen. Therefore, in the initial stage of the static compression test, that is, before the pressure plate and the end face of the sample are closely adhered, the stress–strain curve had obvious fluctuation phenomenon. It couldn’t reflect the true strength deformation property of the material. In order to eliminate the turbulent data generated by the adjustment of the contact between the compressor and the test piece at the beginning of the test, this article referred to the treatment of Nie et al.24 and Lu et al.25. When the stress–strain curve was processed, the data with the stress greater than 0.5 MPa in the original data was selected as the effective data, and the strain value of the effective data was subtracted from the disorder value to obtain the final stress–strain curve. After processing, there was no large fluctuation phenomenon, which could reflect the strength and deformation performance of the specimen (Fig. 6).

Figure 6

Stress–strain curves of disorder data: (a) before treatment; (b) after treatment.

Dynamic compression test

To explore the influence of impact velocity, the direction of the bedding structure and the number of F–T cycles on the mechanical properties of the sandstone, the split Hopkinson pressure bar (SHPB) was used for dynamic compression test. The impact test plan is formulated as shown in Table 1.

Table 1 Impact test program of the bedding sandstone under F–T cycles.

The split Hopkinson pressure bar mainly includes an energy storage module, an impact module, and a data acquisition module. The energy storage modules mainly include air compressors, high-pressure tanks (atmospheric packages) and connecting pipes; the impact modules mainly include launching devices, transmission rods, equipment brackets, energy absorption components and consoles (Fig. 7); the data acquisition modules mainly include strain detectors and Bullet velocity measuring instrument. The transmission rod consists of a striking rod (bullet), an incident rod, a transmission rod, and an energy absorbing rod. They are all made of 48CrMoA high-strength alloy steel, with a diameter of 100 mm, an elastic modulus of 210 GPa, a Poisson’s ratio of 0.25 to 0.3, and a density of 7.85 g/cm3. The strain gauge is attached to the appropriate position of the incident rod and the transmitting rod, which are connected to the data acquisition system. Through dynamic strain tester, the incident, reflection and transmission waveforms in the rod can be collected.

Figure 7

Schematic diagram of Φ100 mm SHPB test system.

When conducting an impact test, first of all, a high pressure was formed in the high-pressure tank by an air compressor. Then, the console air valve was opened and the preset high-pressure potential energy was instantly converted into the kinetic energy of the striking rod. Then, the striking rod impacted the incident rod at a certain speed, thereby exciting the impact pulse, and then compressing the specimen. The data was collected through a strain gauge attached to the rod and an external strain detector.

To ensure dynamic equilibrium and deform uniformity, the annealed copper sheet, 1.0 mm in thick and 10–25 mm in diameter, was used as the wave shaper (Fig. 8). It has shown that the wave shaper can effectively reduce the high-frequency oscillations in the stress pulse. At the same time, it is beneficial to obtain a smooth sinusoid-like incident pulse waveform, extend the rise time of the pulse, and effectively improve the uniformity of the internal stress of the sample. In the impact test, it was confirmed that the sum of the incident wave and reflected wave agreed well with the transmitted wave, reflecting the stress balance of specimens during impact process (Fig. 9).

Figure 8

Copper sheet shaper before and after impact.

Figure 9

Stress waves in the SHPB impact compression tests.

Microscopic test

The EM-30 scanning electron microscope developed by COXEM was carried out to analyze the bedding sandstone specimens under the F–T cycle conditions. Since the rock was a non-metallic material, it was highly insulated from charge. Therefore, in the process of electron beam bombardment scanning of the bedding sandstone specimen, it was easy to form an electron accumulation band on the surface of the specimen, which formed a negative charge region resulting in local discharge phenomenon. This phenomenon would adversely affect the normal scanning of the incident electron beam, indirectly reducing the quality and authenticity of the scanned image. In order to improve the accuracy of electron microscopy, the ETD-800 ion sputter was used to coat gold on the surface of the specimen before electron microscopy forming a conductive loop, avoiding the electron accumulation and negative charge regions.

Microcracks and surface micro-structure of red-sandstone specimens before accelerated weathering were clear from SEM images as shown in Fig. 10.

Figure 10

SEM image of wood grain sandstone free from F–T cycles.

Source link

Leave a Reply

Your email address will not be published. Required fields are marked *