Loading...
Analysis of rockmass mechanical behavior using rockmass analog numerical models and laboratory specimens
West, Isabella G.
West, Isabella G.
Citations
Altmetric:
Advisor
Editor
Date
Date Issued
2023
Date Submitted
Collections
Research Projects
Organizational Units
Journal Issue
Embargo Expires
Abstract
The ability to characterize and predict the mechanical behavior of rock has important applications for the design of civil and mining engineering structures, as well as applications for industries like geothermal, oil and gas, etc. Rock can be characterized mechanically by attributes such as strength, stiffness, and brittleness. Such material attributes can be estimated through laboratory testing on representative intact rock specimens. However, since field-scale rock formations (i.e. rockmasses) are composed of sections of intact rock separated by a network of preexisting discontinuities (e.g. joints, faults), their mechanical behavior is difficult to quantify in the lab due to the required scale. The representative elementary volume (REV) of a rockmass is typically on the scale of meters, whereas standard laboratory equipment accommodates rock specimens on the decimeter scale. The overarching goal of this research is to investigate rockmass mechanical behavior using two different methods, given the aforementioned challenges: (1) small-scale rockmass laboratory analog specimens and (2) numerical models. Specifically, analog laboratory specimens have been used to investigate the effect of intact rock brittleness on rockmass strength and the capability of Bonded Block Models to predict the behavior of jointed rock has been evaluated.
The first method for studying rockmass behavior is small-scale rockmass analog laboratory specimens. These analog specimens have been used to study trends in rockmass behavior given various joint and joint set properties. However, little research has been conducted on the influence of intact rock attributes (i.e. strength, stiffness, brittleness) on the effect that preexisting discontinuities in rockmass have on its mechanical behavior. This thesis considers the influence of intact rock brittleness, which is investigated using small-scale rockmass analog laboratory specimens of Carrara marble, a relatively ductile rock. The analog specimens used contain smooth saw-cut joints in two distinct joint sets. The results of this research are compared to a previous study that tested similar rockmass analog laboratory specimens of Blanco Mera granite, which is a much more brittle rock (despite having a similar unconfined compressive strength). Notably, the drops in peak strength with increasing degrees of jointing are smaller for Carrara Marble than for Blanco Mera granite. At the upper end of the range of confinement levels tested, the strength of Carrara marble was virtually unaffected by the presence of the preexisting joints.
An alternative method for studying rockmass mechanical behavior is numerical modeling. Specifically, Bonded Block Modeling (BBM), a subset of Discrete Element Method (DEM) modeling, has become increasingly used to simulate intact rock damage, as it is capable of replicating grain-scale damage in rock. BBM uses continuous polygonal blocks to represent intact mineral grains in rock, which are able to detach and move freely relative to each other. Input parameters are required that govern the stiffness, strength, and dilatancy of the blocks and their contacts. Part of this thesis focuses on exploring the BBM sensitivity of the emergent mechanical behavior of the BBMs to input parameter heterogeneity, which has not been thoroughly documented previously in the literature.
BBM has been used previously to create Synthetic Rockmass Models (SRM). The SRM approach is based on the premise that if a numerical model replicates the micromechanical damage mechanisms (i.e. inelastic yield of blocks and the formation of cracks along block boundaries) occurring at a small scale in real rock, in addition to matching emergent material attributes (i.e. strength, stiffness, dilatancy, brittleness), then the model will act mechanistically realistic under other conditions and at other scales. Although SRM is used in practice to achieve reasonable representations of field-scale ground behavior, the predictive capabilities of such models under controlled conditions have not yet been thoroughly verified or peer-reviewed by the scientific community. The availability of the jointed rockmass analog laboratory specimens used in this thesis provides a unique opportunity to evaluate the predictive capabilities of BBM for simulating rockmass mechanical behavior and validate the concept of SRM in general. A forward prediction is evaluated by first calibrating intact BBMs of both Blanco Mera granite and Carrara marble to intact rock laboratory data. Next, preexisting smooth joints that match those from the jointed rockmass analog laboratory specimens are added to the intact BBMs. The behavior of these jointed models is then compared to the behavior of the jointed laboratory specimens. In the case of the brittle granite, its BBM is able to predict the behavior of jointed rockmass analog specimens without further calibration. The validation of such a predictive model for brittle rock allows for further research regarding joint set properties and their effect on rockmass mechanical behavior. However, in the case of the more ductile marble, realistic predictions of the jointed specimen behavior are not achieved due to the inability of the BBM to realistically replicate more ductile damage mechanisms in rock.
The BBM of Blanco Mera granite was then used in a joint property sensitivity analysis to investigate the effect of joint roughness/alteration, orientation, persistence, and the number of joints on the strength of rockmass analog laboratory specimens. Joint friction angle was found to have a small effect on the model’s strength and only certain ranges in joint orientation notably affected the strength of the model. Both the number of joints and joint persistence were found to have large effects on the model’s strength.
Associated Publications
Rights
Copyright of the original work is retained by the author.