The burst experiment is a common method that is mostly used in the petroleum industry to estimate the fracture toughness (KIC) of rocks under confinement in the laboratory. Unlike other methods, it tests rock samples subjected to confining stress, which represents the actual subsurface condition of a stressed wellbore with bi-wing fractures, and therefore it is considered advantageous to other methods. Fracture propagation in this burst experiment is only indicated by a sudden burst of the jacketed sample, where the burst pressure is measured. The fracture toughness of the sample can then be calculated after taking this recorded pressure as the critical pressure required to propagate the axially pre-notched wing fractures. However, the fracture growth may exist before the sample bursts, which might be a possible weakness of this technique. The cause is the existence of stable crack growth before the unstable “burst”. Due to the challenges in the detection of this stable crack growth, there is significant uncertainty in the crack length used in the analysis of the results.
The purpose of this thesis is to estimate fracture toughness accurately by comparing the stable and unstable behavior types from numerical modeling predictions with the results of a series of modified burst experiments using acoustic emission (AE) detection for certain configurations. The results indicate that stable growth will occur before sample rupture for certain sample geometry and loading combinations. The difference between stable and unstable growth cases is detected from acoustic emission monitoring records. Moreover, modifications are proposed for more accurate fracture toughness calculations after investigating the behavior of the estimated fracture toughness results. More consistent estimations of fracture toughness are obtained from experiments with unstable configurations. The results of experiments with unstable behavior indicate a positive correlation between confining pressure and fracture toughness estimations, which complies with many observations reported in the literature. Lastly, fixed confinement experiments are performed for a legitimate calculation of KIC with a range of geometry and loading conditions based on the global stability criterion. The effect of fixed confining pressure on the fracture toughness estimations has been shown, offering guidance for improved design of this frequently used test method.