Search

• Positive and negative oscillations of K II under compression are the reason for zigzag crack propagation paths modeled by the traditional fracture criterion. • GMTS criterion considering T -stress can effectively eliminate K II oscillations but requires complicated parameter calculations. • SLMS criterion proposed in this study can conveniently and efficiently model the propagation of both tensile and shear cracks. • The effects of flaw inclination, friction coefficient, loading level, and flaw surface contact status on the crack propagation paths were studied. The traditional fracture criterion performs well when calculating crack propagation under tensile conditions, but a zigzag crack propagation path was obtained under compression, which is inconsistent with experimental results. This study indicates that the positive and negative oscillations of the mode-II stress intensity factor (K II) under compressive loading during crack propagation is the reason for the zigzag trajectory of crack propagation. Such oscillation of K II can be effectively eliminated when the non-singular stress (T -stress) at the flaw tips is considered, and thus, the path of crack propagation can also be smoothed. However, due to the complicated calculations of the stress intensity factors and T -stress at each step of crack propagation, it is not suitable for simulation in fast, complex environments and multi-flaws. Also, the lack of a shear fracture criterion in traditional fracture methods leads to difficulty in studying shear crack propagation in rock. Therefore, a strength-based localized maximum stress (SLMS) criterion was proposed in this study to model both tensile and shear crack propagations in rock more efficiently, conveniently, and accurately. Then, the crack propagation processes in both plate and Brazilian disc specimens with a single flaw under uniaxial and biaxial compression were modeled to investigate the effects of the flaw inclination angle, friction coefficient, and loading level on the crack propagation path. The influence of the contact friction between the flaw surfaces on the contact status and the crack propagation path during the crack propagation process was also analyzed. Also, the crack propagation in a plate with a single flaw under biaxial tension was modeled, which shows bifurcation crack propagation when the lateral tension is several times larger than the axial tension. All the modeled results indicated that the proposed SLMS criterion can get better results when modeling crack propagations for open or closed flaws under tension or compression conditions. [ABSTRACT FROM AUTHOR]

• The SIFs critical values of sandstone were tested by SCB specimen. • Relationships between water content and SIFs were studied. • SIFc of sandstone subjected to loading during soaking is lower than that without loading. • Water soaking duration and loading-soaking sequence of rock significantly affect the SIFs. Fracturing behavior of rock is significantly affected by water. To study the influence of water on the critical values of stress intensity factors (the SIFc, including K If and K IIf) corresponding to the onset of fracture of rocks, edge-cracked semicircular bend (SCB) testing was performed on sandstone specimens (collected at a site in Linyi City, Shandong Province, China) with different water contents. First, water absorption was tested to study the evolution of the water content in the specimens. Then, the SIFc of SCB specimens after different soaking durations in water (yielding conditions ranging from dry to saturation) was tested under three-point loading. The results showed a linear decrease in SIFc with increasing water content for all the specimens with different notch angles; however, with increasing notch angle, the K If exponentially decreased, while the K IIf increased. To study the effect of sustained load on SIFs of rock in water, a new experiment was designed on the specimen subjected to loading while soaking for a certain duration, then the SIFc was tested. Comprehensive predictive empirical relationships are established between SIFc and water content/soaking duration for the sandstone. A novel observation is for two specimens with the same water soaking duration, the SIFc of a specimen with loading during soaking is lower than that when it was tested after soaking, which is attributed to the fact that water is more likely to penetrate into the rock, because the bending load opens pores and cracks then leads to an increase in water content under such conditions. In other words, the water content of sandstone specimens used in this study is the key factor affecting its SIFc. [ABSTRACT FROM AUTHOR]