Dissertation – John Shamu

Shamu, J. (2021) Rock grouting design: Rheological aspects and radial flow visualizations with ultrasound. PhD thesis, TRITA-ABE-DLT-2140: Stockholm: KTH Royal Institute of Technology

urn:nbn:se:kth:diva-304246

Abstract: The rheological properties of cement-based grouts play a crucial role in determining the final spread in grouted rock formations. In rheological terms, cement grouts are known to be complex time-dependent yield stress fluids, but their steady flow behavior is often described by the simple Bingham constitutive law. The Bingham parameters obtained from the linear curve fitting to flow curve data are then used in grout propagation calculations during the design phase, e.g., for rock fracture grouting in tunnel construction. Since cement grouts are time-dependent and thixotropic suspensions, the interpretation of their flow curves during conventional rotational rheometry is often complicated by the presence of wall slip, thixotropy, flow localization, and sedimentation, particularly at low shear rates. A systematic approach was carried out as part of the research work to study these effects within the constraints of the concentric cylinder geometry (Couette) and for different cement grout concentrations. Of particular interest were the influence of geometry and flow sweep measurement interval on flow curves, including the characteristic unstable flow branch that appears at applied shear rates that are below the critical shear rate. The unstable flow branch observed below the critical shear rate has been described as a characteristic feature in the flow curves of thixotropic suspensions, e.g., cement grouts and laponite. From a practical standpoint, these crucial shear rate aspects, including wall slip, have not been considered during grouting design calculations while using the common Bingham model. Thus, the research also considered these shear rate aspects as part of grouting design by incorporating them into the design approach within the existing framework of the Real Time Grouting Control (RTGC) method. Another interesting part of the research work presented in this thesis relates to studies on the radial flow of yield stress fluids. The radial flow between parallel plates is an idealized fundamental flow configuration that is often used to understand grout spread estimation in rock fractures. In part, the entire radial flow work was motivated by the ongoing discussions in the literature regarding the different analytical solutions for radial flow. Moreover, compared to other flow configurations, e.g., pipes and channels, only a limited amount of work has presented analytical solutions, numerical models, and especially experimental work for radial flow. Thus, during the doctoral work, a radial flow experimental device was designed, manufactured, and subsequently used to acquire Carbopol YSF radial flow velocity profiles for the first time. The velocity profile measurements were carried out using the pulsed Ultrasound Velocity Profiling (UVP) technique. The velocity profiles from the initial radial flow study showed that significant wall slip was present. An analytical solution with a Navier slip term was used to describe the velocity profiles, resulting in a good agreement in the velocity profile magnitude. However, the plug-flow region extent was smaller in the analytical solution. Subsequent studies on radial flow sought to address the wall slip issue using different wall slip reduction procedures (chemical treatment and sandblasting). Both treatments showed substantial wall slip reduction; however, some wall slip effects persisted, especially for the thicker Carbopol gels. In addition, a plug-point estimation algorithm using Tikhonov regularization was developed to calculate the yield points from the non-smooth velocity profiles accurately. A final modification was the addition of frame reinforcement to maintain the required constant aperture better. The tests carried out in the final radial flow study followed the test scheme from the previous studies, but with two concentrations of Carbopol. The aim was to compare the measured velocity profiles with two radial flow analytical solutions based on different assumptions that have recently been discussed in the literature. The measurement results showed a good agreement in the velocity profile shape, but with some velocity magnitude discrepancies, particularly in the central part of the velocity profiles. Such discrepancies could result from remaining wall slip together with other higherorder flow effects, e.g., nonlinear flow due to inertial effects, that are not accounted for by the analytical solutions. Nevertheless, within the context of grouting practice, such magnitudes of differences could be considered reasonable for scoping calculations during grouting design and execution. Future studies related to radial flow can improve the current understanding by conducting similar tests, but with improved experimental setups, e.g., better wall slip reduction, larger aspect ratios, and more detailed spatial resolution for smaller flow apertures. Additionally, the understanding of the rheological behavior of cement grouts would be improved from a practical standpoint, i.e., grouting design and execution, if the wall slip phenomenon is studied in more detail and considered an inseparable feature of yield stress fluid flow.