Philip Griffin

The Role of Shear Stress in Arteriovenous Fistula Maturation and Failure: A Systematic Review

Introduction Non-maturation and post-maturation venous stenosis are the primary causes of failure within arteriovenous fistulae (AVFs). Although the exact mechanisms triggering failure remain unclear, abnormal hemodynamic profiles are thought to mediate vascular remodelling and can adversely impact on fistula patency. Aim The review aims to clarify the role of shear stress on outward remodelling during maturation and evaluate the evidence supporting theories related to the localisation and development of intimal hyperplasia within AVFs. Methods A systematic review of studies comparing remodelling data with hemodynamic data obtained from computational fluid dynamics of AVFs during and after maturation was conducted. Results Outward remodelling occurred to reduce or normalise the level of shear stress over time in fistulae with a large radius of curvature (curved) whereas shear stress was found to augment over time in fistulae with a small radius of curvature (straight) coinciding with minimal to no increases in lumen area. Although this review highlighted that there is a growing body of evidence suggesting low and oscillating shear stress may stimulate the initiation and development of intimal medial thickening within AVFs. Further lines of evidence are needed to support the disturbed flow theory and outward remodelling findings before surgical configurations and treatment strategies are optimised to conform to them. This review highlighted that variation between the time of analysis, classification of IH, resolution of simulations, data processing techniques and omission of various shear stress metrics prevented forming pooling of data amongst studies. Conclusion Standardised measurements and data processing techniques are needed to comprehensively evaluate the relationship between shear stress and intimal medial thickening. Advances in image acquisition and flow quantifications coupled with the increasing prevalence of longitudinal studies commencing from fistula creation offer viable techniques and strategies to robustly evaluate the relationship between shear stress and remodelling during maturation and thereafter.

The Effect of Reynolds Number, Compressibility and Free Stream Turbulence on Profile Entropy Generation Rate

Detailed aerodynamic data from the suction surface boundary layer on a turbine blade arranged in a linear subsonic cascade was acquired under high free stream turbulence conditions (∼ 5.2%) generated using a perforated plate placed upstream of the cascade. In addition, data was also obtained from a transonic turbine cascade utilizing the same blade profile but of much smaller chord at free stream turbulence levels of 3.5%. Velocity profiles from the laminar, transitional and turbulent boundary layers were measured at various locations along the airfoil suction surface for the incompressible regime at ReC of 76,000. For the compressible test cases, boundary layer velocity profiles were measured at two locations towards the aft section of the blade at ReC of 163,000 and MEx of 0.37 respectively. For both cases the boundary layer velocity profiles were acquired by traversing a single normal hot wire probe normal to the blade surface. In addition the extent of the transition region over the blade surface was determined for both compressible and incompressible regimes by the use of an array of heated thin film sensors over a range of Reynolds and exit Mach numbers. It was observed that an earlier transition ensued at high free stream turbulence conditions in comparison to a previous investigation at comparable ReC and lower turbulence level (0.8% Tu). In addition comparisons were made to existing incompressible data at ReC = 185,000 and 0.8% free stream turbulence intensity. One of the primary observations resulting from an earlier transition was a thicker turbulent boundary layer, but in addition it was also noted that shear strain rates in the laminar boundary layer were significantly higher than those obtained at the 0.8% turbulence intensity. Further analyses also elucidated the presence of fluctuating components of velocity in the laminar boundary layer and were attributed to the effects of the free stream turbulence. This leads to the notion of a hybrid boundary layer, possessing both laminar and turbulent characteristics. These findings have implications regarding the profile loss of the blade, that is the loss generated in blade boundary layers and wakes normally associated with phenomena such as viscous shear, Reynolds stress production, shock wave formation and heat transfer across temperature differences and can be quantified in terms of the amount of entropy generated. For the purposes of this study entropy creation is solely restricted to that arising due to fluid dynamic phenomena, thereby assuming an adiabatic and quasi-isothermal flow. The entropy generation rate per unit volume is obtained directly from the boundary layer velocity profile; further integration gives rise to the entropy generation rate over the boundary layer at a point or over the entire suction surface length. Even though the number of quantitative measurement points on the transonic cascade was limited due to the very thin boundary layer present, no effects attributable to compressibility were observed on the entropy generation rate at the Mach number in question. Increased free stream turbulence had a greater effect on the generated entropy due to increased viscous shear in the laminar boundary layer and increased Reynolds stress production. In contrast, free stream turbulence did not have any significant effect on the turbulent boundary layer in the context of this study, as it was observed that the amount of entropy generated in the turbulent boundary layer was approximately equivalent for both turbulence levels at comparable Reynolds number.Copyright

Numerical and experimental investigation of the mean and turbulent characteristics of a wing-tip vortex in the near field:

The near-field (up to three chord lengths) development of a wing-tip vortex is investigated both numerically and experimentally. The research was conducted in a medium speed wind tunnel on a NACA 0012 square tip half-wing at a Reynolds number of 3.2 × 105. A full Reynolds stress turbulence model with a hybrid unstructured grid was used to compute the wing-tip vortex in the near field while an x-wire anemometer and five-hole probe recorded the experimental results. The mean flow of the computed vortex was in good agreement with experiment as the circulation parameter was within 6% of the experimental value at x/c = 0 for α = 10° and the crossflow velocity magnitude was within 1% of the experimental value at x/c = 1 for α = 5°. The trajectory of the computed vortex was also in good agreement as it had moved inboard by the same amount (10% chord) as the experimental vortex at the last measurement location. The axial velocity excess is under predicted for α = 10°, whereas the velocity deficit is in relatively good agreement for α = 5°. The computed Reynolds shear stress component 〈u′v′〉 is in good agreement with experiment at x/c = 0 for α = 5°, but is greatly under predicted further downstream and at all locations for α = 10°. It is thought that a lack of local grid refinement in the vortex core and deficiencies in the Reynolds stress turbulence model may have led to errors in the mean flow and turbulence results respectively.

Effects of Freestream Turbulence on the Characteristics in the Boundary Layer Near the Transition Onset Location

Experimental data is presented for a flat plate test facility with augmented levels of freestream turbulence (FST). The turbulence decay downstream of two turbulence generating grids, in addition to the integral length scales, is provided and good comparison with established correlations is presented. Boundary layer measurements using a single normal hotwire probe were obtained at FST intensities of 7%, 6%, 5.5%, 1.55%, and 1.45%, and the results presented include the fifth and 95th percentile of the velocity fluctuations and the root mean squared (RMS) velocity profiles near the transition onset region. The transition onset Reynolds number for each of the turbulence levels considered is consistent with theoretical findings. In all cases analyzed, the maximum fifth and 95th percentile far exceeded the maximum RMS values, with the location of the maximum 95th percentile closer to the wall compared to the maximum fifth percentile. Using probability density function (PDF) analysis, it is demonstrated that there is a dominating positive velocity fluctuation in the near-wall region and a dominating negative velocity fluctuation further out in the boundary layer and that the fluctuations in the boundary layer are greater compared to the freestream. The effect of the FST on the boundary layer is discussed with comparison to the Blasius solution and the influence of the fluctuations on the deviation from the Blasius profile is presented and discussed. Through investigation of the energy spectrum of the fluctuating velocity component within the boundary layer, it is shown that there is a higher energy content at lower frequency in the boundary layer when compared to that of the freestream.

Quantifying Turbulent Wall Shear Stress in an Arteriovenous Graft Using Large Eddy Simulation

Hemodialysis patients require a vascular access capable of accommodating the high blood flow rates required for effective dialysis treatment. The arteriovenous graft is one such access. However, this access type suffers from reduced one year primary & secondary patency rates of 59–90% and 50–82% respectively [1]. The main contributor to the failure of this access is stenosis via the development of intimal hyperplasia (IH) that predominately occurs at the venous anastomosis. It is hypothesized that the resulting transitional to turbulent flow regime within the venous anastomosis contributes to the development of IH. The aim of this study is to investigate the influence of this transitional to turbulent behavior on wall shear stress within the venous anastomosis via the use of large eddy simulation.Copyright

Experimental/Numerical Investigation of a Wingtip Vortex in the Near-Field

turbulence intensity of 9% just after the trailing edge. Maximum turbulence intensity after x/c = 0 is observed in the vortex core which decays with downstream distance. Five-hole probe mean velocity measurements revealed jet like and wake like axial velocity profiles depending on wing angle of attack, with values of 1.33U∞ and 0.83U∞ being measured for 10° and 5°. Axial vorticity is also observed to increase gradually with downstream distance. A full Reynolds stress turbulence model with a second order upwind differencing scheme is used to compute the vortex in the near-field. Vortex structure and trajectory correlate well with experiment. It is thought that the dissipative nature of the second order convection scheme led to discrepancies between numerical and experimental results.

Capillary Driven Fluid Flow in Medical Devices

Microscale fluid dynamics has played a significant role in the development of many applications in the medical diagnostic sector and in recent years many devices have implemented advances in this field. Capillary driven assays are commonly used in diagnostic areas, such as, cardiac risk, fertility, drug abuse and infectious diseases1. Typically a platform is used where bodily fluids or samples are taken, filtered and by means of small microchannels, transported and mixed with a variety of antibodies2. In order to perform correctly, these antibodies need to bind to the proteins in the fluid. It is therefore essential that the exposure of the proteins to the antibodies is maximized. To achieve this, capillary dimensions can be altered to obtain the required flow rates and exposure times3. This paper focuses on controlling these parameters.In this paper, a study was conducted in which the flows in four straight rectangular microchannels of varying cross sectional areas were assessed. The four microchannels were fabricated from an epoxy material. The microchannel widths varied from 100μm to 1000μm with each channel having a dept of 200μm. The four microchannels had aspect ratios of 0.5, 1, 2 and 10. The microchannels were sealed using a heat sealing hydrophilic tape. Fluid velocity rates were measured experimentally using an X-Stream XS-4 high speed camera at 500 frames per second. Preliminary contact angle results between water and the epoxy material gave a contact angle of 81.5 degrees +/− 6 degrees. Computational models of the four microchannels were performed using a Volume Of Fluid (VOF) model in Fluent 6.2.16, a commercially available CFD code. The computational models had four boundary types: pressure inlet, pressure outlet, epoxy wall and hydrophilic tape wall. The inlet boundary has an initial pressure applied to it, capillary pressure between the water and air interface in the microchannel. It was found that as the dimensions of the microchannels increased, the governing equations were less accurate in predicting the experimental fluid velocity in the microchannels.Copyright