Date of Award

Summer 8-1-2025

Degree Type

Thesis

Degree Name

Master of Science (MS)

Department

Geological Engineering

Committee Chair

Mary MacLaughlin

First Advisor

Yohanna Mejia Cruz

Second Advisor

Lorne Arnold

Abstract

Rock slope stability is a critical concern in seismically active regions, where earthquakes can trigger sudden and hazardous failures. Predicting these failures is challenging due to the complex interactions between seismic loading and geological characteristics. This study proposes a simplified analytical model to predict yield acceleration based on the forces acting upon a block just before sliding initiates. A key parameter in this equation is the friction acting on the bottom surface of the block. As rock blocks are commonly formed by the intersection of sets of subparallel fractures known as joints, and the roughness and other features of those joints can be quite variable, accurate joint characterization is essential. To characterize joint strength and behavior, single stage (SS), standard multistage (MS), and limited displacement multistage direct shear (LDMDS) tests were conducted on specimens with two surface types: cast replicates of a rough, naturally occurring granite surface (“bumpy M3”) and commercially available flat, red masonry paving blocks (“flat red pavers”). For the bumpy surface, LDMDS tests more closely replicated the shear strength parameters, friction angle and joint shear intercept (cohesion), of SS tests compared to standard MS test procedures. In contrast, the flat paver specimens subjected to the LDMDS tests displayed a lower friction angle than those subjected to the SS and MS direct shear tests, which was unexpected. This is likely due to the LDMDS specimen set having a lower friction angle, but could also be related to the friable, stick-slip behavior exhibited by the flat specimens, which made shear strength characterization difficult. This suggests that LDMDS testing is not particularly effective for flat, friable surfaces; however, further investigation is warranted. As the test results confirmed negligible cohesion acting on the flat surfaces, tilt tests were used to efficiently evaluate the range of friction angles displayed by the paver surfaces and surface degradation caused by repeated shearing of the surfaces. A total of 160 tilt tests were performed on eight pairs of flat red pavers, revealing a friction angle range from 17° to 50° and highlighting surface variability. These results were used as inputs to the analytical equation to estimate a range of yield accelerations for comparison with shake table experiments. Shake table experiments were conducted to simulate seismic loading of the flat red paver surfaces. Challenges with data collection and interpretation were overcome, and measured accelerations at the onset of block sliding were compared to the analytically predicted yield accelerations. The analytical pseudo-static equation proved effective: approximately 70% of measured yield accelerations fell within the interquartile range of those predicted using the results from the first five tilt tests on each of the eight surfaces (40 tests). To further validate the equation, two numerical models were developed in Itasca Consulting Group’s 3-Dimensional Distinct Element Code (3DEC) Version 9.0: a simple model and a more complex dynamic model. The simple model showed good agreement with analytical predictions based on tilt test results. However, the dynamic model exhibited discrepancies, particularly when vertical loads were applied or the block weight increased, suggesting that further refinement may be needed. Overall, experimental and numerical findings support the analytical equation as a practical tool for preliminary slope failure assessment under seismic loading. This integrated approach offers valuable insights into the blocky rock system under seismic loading and highlights the importance of interface characterization in stability assessments.

Download

Included in

Engineering Commons

Share

COinS