NIOSH researchers have identified a pattern of fracture zone development that suggests an explanation for fracture formation around rectangular openings in underground mines. This pattern is characterized by shearing and dilation that result in either faults or in the repeated formation and propagation of en echelon fractures from sites of tension. Two computer modeling codes, Fast Lagrangian Analysis of Continua (FLAC) and Particle Flow Code (PFC), were used to model different aspects of this pattern. Use of very small elements with FLAC enabled the identification of sites of initial tension near the skin of square-cornered rectangular openings, while PFC allowed the initiation and progressive development of fractures from these sites to be followed as the fractures evolved.
Such studies can lead to a greater understanding of how roof support can be better selected and installed for specific conditions in underground mines prone to roof falls and rock bursts. These studies may also lead to modifications of corner and opening shapes that could be incorporated into mine designs to produce more stable mine openings and reduce the risks of rock falls and rock bursts.
Mining-induced fractures are typically either hidden from view or become obscured when fractured rock around mine openings collapses or is ejected in a rock burst. Consequently, knowledge of the distribution, geometry, and extent of mining-induced fractures has been limited. However, fracture patterns that suggest mechanisms of fracture formation are occasionally seen where new crosscuts intersect old ones. The old fractures are exposed in the walls of the new opening, but well-described examples are rare in the literature.
Most investigators agree that extension fractures form parallel to the direction of maximum principal stress. Recently favored explanations for how decimeter and longer (macroscale) fractures develop have mainly involved the proliferation, interaction, and coalescence of smaller fractures.
Fairhurst and Cook  proposed that stress-induced microcracks first form an incipient cleavage parallel to the surface of a mine opening. These microcracks then grow into larger fractures as a result of the applied stress. These fractures shorten with depth, and their ends define a shape concave toward the opening that represents the limit of breakage in the case of a rock burst, roof collapse, or pillar hour-glassing. However, in a field example where thin layers were displayed at the margin of a rock burst breakout, White  concluded that the closely spaced fractures had not extended across the entire volume of the ejected rock, but were present only near the periphery of the breakout. Other examples that support this scenario are described in White et al.
Stacey  proposed that failure about mine openings will occur if extension strain reached a certain value, which he considered a material property. He noted that extension strain is highest near the corners of rectangular openings and suggested that failure begins at these locations. He proposed that the extent of failure is determined by how much rock around the opening exceeds the requisite extension strain. For rectangular openings, critical extension strain is deepest at the midpoint between corners and its limit duplicates the concave shape identified by Fairhurst and Cook  and commonly seen after roof falls and rock bursts. However, Stacey did not differentiate extension resulting from tension from Poissonresponse extension.