How oblique bedding gives rockslides lateral resistance: Shanyang case study

In the mountains of Shaanxi Province, China, a catastrophic rockslide that killed 65 people in 2015 has now yielded the secret of its own destruction, and the answer lies in a slow chemical attack that quietly dismantled the mountain’s most important line of defense. A research team led by Jiayun Wang of the Xi’an Center of China Geological Survey, working with landslide expert Yueping Yin of the China Institute of Geo-Environment Monitoring, has reconstructed the failure of the Shanyang rockslide in unprecedented detail, showing that its stability hinged on a single lateral plane of karst-dissolved rock whose strength was steadily eaten away by rainwater. The study, published in Environmental Earth Sciences, combines unmanned aerial vehicle imagery, field mapping, laboratory rock testing, and three-dimensional numerical simulation to reveal how obliquely inclined bedding rockslides transition from silent creep to sudden, deadly collapse.

The Shanyang rockslide occurred on August 12, 2015, releasing approximately 1.68 million cubic meters of rock down a steep bedrock slope and causing nearly 500 million RMB in economic losses. The slope itself is geologically distinctive: it sits on the southern flank of the overturned Yaolinghe anticlinorium, where thick dolomite layers of the Sinian Dengying Formation, dipping at 43 to 60 degrees, overlie a much weaker foundation of carbonaceous mudstones and siliceous slates of the Cambrian Shuigoukou Formation. This hard-over-soft architecture, combined with an oblique angle between the rock strata and the slope face, places Shanyang squarely in the class of obliquely inclined bedding slopes that are widespread across the carbonate terrain of southwestern and northwestern China. These slopes are notoriously dangerous precisely because their movement is deceptive. Instead of sliding straight down the true dip direction of the bedding, the rock mass is forced to deflect sideways around a stable bedrock barrier, sliding along the so-called apparent dip, a direction that is far less visible to the eye and far harder to anticipate.

The key to this deflection, and the central focus of the new study, is a phenomenon the researchers call lateral resistance. When a rock mass begins to slide along its bedding planes, it eventually encounters stable rock on one flank. In carbonate terrain, that flank is often bounded by a karst dissolution structural plane, a steep fracture surface enlarged and altered by the dissolving action of percolating rainwater. At Shanyang, this plane strikes at 113 degrees and dips at a steep 76 degrees, extending roughly 360 meters in length with an average vertical height of 40 meters. As the sliding mass attempts to rotate from true dip movement into apparent dip movement, this lateral plane pushes back against the rock, providing the frictional force that keeps the entire slope in equilibrium. Yueping Yin and colleagues first recognized the importance of this mechanism in their studies of the Jiweishan rockslide in Chongqing, and the Shanyang event offered a fresh and tragic opportunity to test the concept quantitatively.

To do so, the team built a three-dimensional discrete element model of the slope using the 3DEC software, a code well suited to simulating large-strain deformation and failure of jointed rock masses. The model domain, 425 meters long, 257 meters wide, and 170 meters high, was discretized into prismatic elements with an average edge length of 30 meters, and four displacement and stress monitoring points were installed on the sliding mass. Critically, the model incorporated both the weak interlayer between the dolomite and the underlying slates, dipping at 48 degrees toward a direction of 15 degrees, and the karst dissolution structural plane on the left flank. Rock mass failure was evaluated with the Mohr-Coulomb criterion, while the sliding mass itself was assigned a ubiquitous-joint model to capture its anisotropic, layered behavior.

The strength parameters underpinning the simulation came from rigorous laboratory work. Eight cylindrical dolomite samples, 300 millimeters in both diameter and height, were cored from fresh bedrock exposed at the slide site and tested in uniaxial and triaxial compression at Xi’an University of Technology using closed-loop servo control, with confining pressures applied at five levels from 5.0 to 25.0 megapascals. The intact dolomite exhibited a cohesion of 12.85 megapascals and an internal friction angle of 48.53 degrees. Because jointing and dissolution dramatically reduce field-scale strength, the researchers then applied the Hoek-Brown empirical criterion to translate intact-rock values into rock-mass parameters. The weak carbonaceous mudstones and siliceous slates were assigned strength values from established engineering geology handbooks, and the slip zone mud was given a cohesion of 75 kilopascals and a friction angle of 24.23 degrees, with its long-term and argillized strengths scaled to 0.8 and 0.6 of natural values respectively. Most tellingly, the dissolution rate in the karst zone was measured at 44.8 percent, meaning the dissolved rock retains only 55.2 percent of intact strength, yielding a shear strength of 7.09 megapascals cohesion and a friction angle of just 26.79 degrees in the dissolution zone.

The simulations unfolded in three revealing stages. Under self-gravity alone, the rock mass crept slowly along the weak interlayer, and the maximum compressive principal stress concentrated sharply along the karst dissolution plane and at the slope toe. Stress at the monitoring point on the dissolution plane reached 1.787 megapascals, several times higher than the 0.762 and 0.462 megapascals recorded in the middle and rear of the slope, confirming that the lateral plane was bearing the brunt of the driving forces. When the researchers then simulated the progressive argillization of the weak interlayer, a process in which repeated interlayer shearing, combined with rainwater infiltration, gradually grinds the weak layer into mud, the stress concentration intensified further, showing that the slide had transitioned from creep to genuine bedding-parallel sliding. Yet the slope still held, propped up laterally by the dissolution plane and at its toe by the resisting rock mass.

The final simulation delivered the fatal verdict. As karst dissolution progressed, reducing both the cohesion and friction of the lateral plane, the stress picture inverted dramatically. The maximum principal stress on the dissolution plane, which had been the highest in the model under both gravity and argillization conditions, dropped sharply to below 0.5 megapascals, while stress in the toe rock mass spiked abruptly to over 4.5 megapascals. In other words, the primary anti-slide force migrated from the lateral plane to the slope toe, which was never designed by nature to carry that load. The toe buckled, and between simulated time steps 46,670 and 86,497 the rock mass underwent substantial displacement, culminating in the apparent dip slide-buckling failure that razed the hillside in 2015. The lateral friction force, the modeling showed, scales linearly with the cohesion and friction angle of the dolomite, so every increment of dissolution translated directly into lost resistance and increased downslope driving force.

The field evidence corroborates the simulation in striking detail. The slip surface in the middle and rear of the rockslide is smooth and devoid of shear striations, indicating it had already been penetrated and the mass detached before failure, held only temporarily by lateral resistance. The slip soils consist of unconsolidated mud produced by intense interlayer shearing, and antistep features on the slip surface record a long history of gravitational creep. In contrast, the karst dissolution plane preserves distinct sliding shear striations, the fingerprints of the final lateral shearing event. Laboratory analysis found the dolomite is 98.5 percent calcite, making it highly susceptible to dissolution, and the dense reticular calcareous cement on the underlying slate surfaces testifies to vigorous rainwater infiltration through dissolution channels. Even the precursors match: sporadic rockfall during the slow deformation phase gave way to markedly more frequent rockfall shortly before collapse, consistent with the simulated transition from creep to accelerated sliding.

The implications extend well beyond one mountainside. Because the creep phase driven by self-weight and interlayer argillization produces measurable, progressive displacement, the authors argue that GPS and GNSS monitoring networks can realistically capture the approach to failure and provide early warning for similar slopes. Perhaps more importantly, the identification of lateral resistance as the dominant stabilizing force redirects mitigation strategy: rather than treating these slopes uniformly, engineers should prioritize targeted reinforcement of karst dissolution structural planes, which both prevents the lateral strength loss that triggers buckling and optimizes the design of anchoring systems. The findings also carry a warning for a changing climate, since intensifying rainfall accelerates both the argillization of weak interlayers and the dissolution of carbonate flank planes, twin processes that together can strip a slope of its defenses.

Obliquely inclined bedding rockslides have now claimed attention at Lianziya in the Three Gorges, at Jiweishan in Wulong, and at Shanyang, and the pattern is consistent: the danger lies not in the most visible direction of movement but in the invisible strength of the rock holding the flank. By quantifying exactly how dissolution erodes that strength, the Shanyang study transforms a qualitative concept into an engineering variable that can be measured, monitored, and reinforced, offering a scientific foundation for risk prevention frameworks across the carbonate mountains of China and beyond.