Fluid mechanics of cavitation in orbital atherectomy.

Abstract

Orbital atherectomy is a means of removing plaque from a stenosed artery which uses a grinding head attached to a catheter to break up the accumulated plaque. The grinding head, referred to as crown or burr, is rotated at very high rotational velocity during the operation, from 80, 000rpm to 200, 000rpm. Because of this high rotational velocity, there are many concerns regarding the possible harm that can be done to the patient. The original motivation for this thesis was to explore and quantify the fluid mechanics of rotational atherectomy. However, as the topic developed, it broaden appreciably to encompass the exploration of the efficacy of turbulence modeling, bubble nucleation and growth, two-phase flow in rotational systems, and means for determining difficult to measure fluid flow characteristics. The work that finally emerged was a synergistic blending of experimentation and numerical simulation. In some instances, the simulation guided the experimental work, while in others the experiments served to guide and validate the simulation models. Since rotational fluid mechanics underlay the entire enterprise, an imperative initial task was to deeply explore the applicability of various turbulence models to such flows. In that regard, the situation of single-phase flow in an annular space bounded by a rotating inner cylinder and a stationary outer cylinder was used as a standard. That physical situation had been the subject of an in-depth experiment-based doctoral thesis noteworthy for its care and attention to detail. The examination of the efficacy of turbulence modeling encompassed two different categories of models. One of these categories dealt with models based on isotropic turbulence. Models in this category result in a turbulent viscosity that is the same for all three possible directions of fluid flow. However, in consideration of the ultimate fluid flows to be considered here, which include superimposed axial and rotational motions, isotropic turbulence cannot be expected to prevail. In this light, consideration was also extended to models in which the isotropic turbulence assumption was dropped in favor of different turbulence intensities in each of the possible directions of fluid flow. To implement the use of non- isotropic turbulence models, certain existing formulations were extended to levels rarely encountered in the published literature. The outcome of the extensive study of turbulence models led to a logic-based selection of the optimum one for the fluid mechanic investigation that is central to this work. Another issue dealt with in preparation for the study of the rotational atherectomy device is the nucleation and growth of bubbles. The need for this focus was the concern, often suggested by certain medical practitioners, that the high-rotational velocities of the device would give rise to locally low pressures in the flowing medium (blood and additives). The existence of pressures below the vapor pressure of the medium would give rise to cavitation bubbles. The bursting of such bubbles is known to create a high-velocity jet which, if impinged on an artery wall, would cause necrosis. Bubbles may be created by a number of different physical processes other than cavitation. In particular, the presence or absence of nucleation sites is a major factor in the creation of bubbles. To gain a thoroughgoing understanding of the entire process of bubble creation and collapse, a theoretical development was pursued. That development was guided by experimental results present in the literature. The model that was created for the numerical simulation yielded results that were consonant with the experimental data. The possible presence of bubbles in a liquid flow creates a fluid regime termed twophase flow. To adhere to the rotational fluid theme, experiments and corresponding modeling was performed for an impeller-driven flow in a contained fluid environment. This physical situation is closely aligned with rotational atherectomy. The investigated situation was designed to enable an initial configuration in which the liquid interfaced with a gas at a horizontal free surface to metamorphize into a curved free-surface interface. In particular, a method of dealing with two-phase flows was evaluated and then successfully implemented. The main focus of the work was a synergistic fluid-mechanic analysis of the rotating atherectomy device positioned in two independent environments: (a) a transparent horizontal tube whose diameter was chosen to model that of the superficial femoral artery and (b) a large open-topped transparent container. The atherectomy device consisted, in essence, of a shaft on which is mounted an enlarged section called the crown. The crown is coated with an abrasive material whose function is to grind hardened plaque and thereby rejuvenate the arterial function. The tube-based experimentation provided both observational and quantitative data. With respect to former, flow visualizations implemented by means of a tracer medium did not reveal the presence of bubbles. With regard to this finding, it is relevant to convey the caveat that inherent optical constraints provided a bound on the smallest observable bubbles. The extracted quantitative information included velocity magnitudes which were compared with those of the numerical simulations and virtual congruence was found to occur. The injected tracer medium also enabled the observation of patterns of fluid flow. These patterns were found to be in close accord with those predicted by the simulations. An additional product of the experimentation was the opportunity provided to investigate situations which were beyond those that could be modeled numerically. These situations included the case in which the crown was positioned eccentrically and in which the shaft was flexible rather than rigid. These two realities brought in laboratory experimentation into close accord with the operational experience. From the experimentation in the open-topped container, both observational and quantitative results were also extracted. Again, these findings strongly supported those from the numerical simulations. Overall, the four interrelated parts of this thesis provided ample opportunity to delve deeply into highly complex fluid-mechanic phenomena. The logic-based selection of turbulence models represents the most complete study of this category compared with the less thoroughgoing comparable studies in the literature. The bubble nucleation and growth models implemented here were strongly supported by experimental data. With regard to the two-phase fluid-mechanic investigation, the overall satisfactory agreement of the numerical predictions with the experimental data provide license for the use of the simulation model for related problems involving the separation of particles immersed in a liquid medium. Finally, all the fluid-mechanic issues related to rotating atherectomy were fully resolved.