Mechanisms of mechanical heart valve cavitation: investigation using a tilting disk valve model

BACKGROUND AND AIM OF THE STUDY

The induction of mechanical heart valve (MHV) cavitation was investigated using a 27 mm Medtronic Hall (MH27) tilting disk valve.

METHODS

The MH27 valve was mounted in the mitral position of a simulating pulse flow system, and stroboscopic lighting used to visualize cavitation bubbles on the occluder inflow surface at the instant of valve closure. MHV cavitation was monitored using a digital camera with 0.04 mm/pixel resolution sufficient to render the tiny bubbles clearly visible on the computer monitor screen.

RESULTS

Cavitation on MH27 valve was classified as five types according to the time, site and shape of the cavitation bubbles. Valve cavitation occurred at the instant of occluder impact with the valve seat at closing. The impact motion was subdivided into three temporal phases: (i) squeezing flow; (ii) elastic collision; and (iii) leaflet rebound. MHV cavitation caused by vortices was found to be initiated by the squeezing jet and/or by the transvalvular leakage jets. By using a tension wave which swept across the occluder surface immediately upon elastic impact, nuclei in the vortex core were expanded to form cavitation bubbles.

CONCLUSION

Analysis of the shape and location of the cavitation bubbles permitted a better understanding of MHV cavitation mechanisms, based on the fluid dynamics of jet vortex and tension wave propagations.

Computational fluid dynamics study of a protruded-hinge bileaflet mechanical heart valve

BACKGROUND AND AIM OF THE STUDY

Following clinical experience with the Medtronic Parallel bileaflet mechanical heart valve, considerable interest has been shown in investigating fluid mechanics inside the hinge socket. Most of these studies involved hinges that are recessed into the valve housing, such as the St. Jude Medical (SJM), CarboMedics, Sorin and On-X bileaflet mechanical heart valves. The aim of this study was to investigate the flow fields of a protruded hinge under steady flow conditions, with the occluder in its fully open position. Computational fluid dynamics (CFD) simulation using the Fluent 4.4.7 commercial solver was applied in this investigation. This protruded hinge mechanism for pivoting the occluder is an in-house design from the Cardiovascular Dynamics Laboratory, Nanyang Technological University.

METHODS

The Fluent 4.4.7 code was run on a Silicon Graphic Inc. computer (4-CPUx185 MHz) in the CFD simulation. A body-fitted coordinates (BFC) grid was generated to cover the entire valvular flow domain, including the interior of the hinge and leaflet. Clearance between the leaflet and pivot housing was 50-70 microm. In the vicinity of the protruded hinge, mesh cells were small compared with hinge dimensions. A power law distribution of grid points was applied to optimize the number of cells used to cluster the entire flow field. The overall computational flow domain of the valve channel, including the floating leaflet and immersed hinge, was approximately 170,000 cells in total. Inside the hinge socket, approximately 10,000 cells were generated. A comparative model with recessed hinge that resembled the SJM valve hinge design was modeled. Due to geometric difficulties, an unstructured grid scheme was applied. Great attention was focused within the hinge pocket, in particular to the clearance between the hinge pivot and leaflet. A total of 2 million cells was generated for the whole computational flow domain.

RESULTS

Under steady flow conditions, with the leaflet fixed in an open position, the protruded hinge design yielded a pair of small vortices that formed behind the stoppers. A low-magnitude velocity was observed inside the hinge clearance. Vortices developed behind the protruded stopper. Migrating flow was noted beneath the leaflet clearance as a result of pressure difference across the leaflet. For the recessed hinge design, reverse flow dominated the inside of the hinge socket, and developed into a pair of vortices at high Reynolds number.

CONCLUSION

The protruded hinge mechanism was designed to expose the overall hinge region to the mainstream flow for a positive washing effect. Flow in this protruded hinge design is, in general, found to be three-dimensional. Initial results under steady flow conditions showed low laminar and turbulent shear stress, while the hinge clearance was well washed.

Digital particle image velocimetry investigation of the pulsating flow around a simplified 2-D model of a bileaflet heart valve

BACKGROUND AND AIM OF STUDY

Strong interactions are believed to exist between the pulsating valvular flow and the valve leaflet motions. Hinge position, indicated by d/W (d = distance between the two axes of the hinge pivots; W = width of the testing section in the middle plane), plays a critical role in MHV performance. An optimized hinge position for a bileaflet heart valve can be identified as a design criterion for better valve performance.

METHODS

A two-dimensional (2-D) digital particle image velocimetry (DPIV) system was used to map the transient flow field of a simplified 2-D model of a bileaflet heart valve with a hydraulic diameter enlarged three-fold under pressure waveforms which was expanded based on Womersley number and Euler number considerations. Six different hinge positions were investigated.

RESULTS

At extreme hinge positions (d/W <0.2 or d/W >0.3), large-scale and long-duration stagnation of flow was found in the central orifice, and instability and highly disturbed flow was noted in plots of velocity vectors.

CONCLUSION

The transient flow pattern in the vicinity of the valve was greatly affected by the hinge position of moving leaflets. An optimum d/W in the range 0.2-0.3 yielded good velocity field and opening and closing behaviors.

Bioprosthetic heart valve leaflet motion monitored by dual camera stereo photogrammetry

Dual camera stereo photogrammetry (DCSP) was applied to investigate the leaflet motion of bioprosthetic heart valves (BHVs) in a physiologic pulse flow loop (PFL). A 25-mm bovine pericardial valve was installed in the aortic valve position of the PFL, which was operated at a pulse rate of 70 beats/min and a cardiac output of 5 l/min. The systolic/diastolic aortic pressure was maintained at 120/80 mmHg to mimic the physiologic load experienced by the aortic valve. The leaflet of the test valve was marked with 80 India ink dots to form a fan-shaped matrix. From the acquired image sequences, 3-D coordinates of the marker matrix were derived and hence the surface contour, local mean and Gaussian curvatures at each opening and closing phase during one cardiac cycle were reconstructed. It is generally believed that the long-term failure rate of BHV is related to the uneven distribution of mechanical stresses occurring in the leaflet material during opening and closing. Unfortunately, a quantitative analysis of the leaflet motion under physiological conditions has not been reported. The newly developed technique permits frame-by-frame mapping of the leaflet surface, which is essential for dynamic analysis of stress-strain behavior in BHV.

Application of an unstructured grid algorithm to artificial heart valve simulations

The time varying flow pattern in the vicinity of mechanical heart valves (MHV) is fairly complex: it involves multiple passages and moving leaflets. The numeric simulation of unsteady flows in these multiple passages with moving boundaries presents a major challenge to computational fluid dynamics (CFD). Two major difficulties in the numeric simulation of MHV flows are 1) the generation of a body fitted grid within the multipassage device and 2) moving leaflets. The conventional finite difference and finite volume scheme obtained by using a structured grid have serious deficiencies in these applications. To fit the grid lines with the various angles of the moving MHV, the grid may often become too skewed for accurate numeric solution. To overcome these deficiencies, significant effort and attention should be placed on the grid generation and moving grid scheme. We present an unstructured moving grid finite volume method for heart valve simulations. The Navier-Stokes equations are discretized on a general tetrahedral mesh by using a finite volume scheme. With this scheme, the mesh can be automatically generated with any commercial software. The method is applied to a tilting disk (Medtronic Hall 29mm, Medtronic, Inc., Minneapolis, MN) heart valve, and results are compared with that of the steady flow solutions. Significant differences between steady and unsteady flow solutions are observed.

Pressure and flow fields in the hinge region of bileaflet mechanical heart valves

BACKGROUND AND AIM OF THE STUDY

Recent clinical thrombotic experiences with the Medtronic Parallel (MP) bileaflet heart valve have highlighted the need for new methods to assess preclinical valve hinge flow. The aim of the current study was to investigate hinge pivot flow fields in bileaflet mechanical heart valves using flow visualization in scaled x5 magnification transparent polymer models and computational fluid dynamic (CFD) analysis using CFD 2000 STORM code.

METHODS

Polymeric x5 flow models of the On-X, St. Jude Medical (SJM) and MP bileaflet heart valves were constructed using laser stereolithography to replicate the interior geometry while maintaining realistic manufacturing tolerances. Each hinge flow experiment was carried out by installing the transparent x5 model in a pulsatile flow loop, which was designed according to Womersley number similitude requirements. Motions of suspended microparticles in the valve hinge area, recorded by laser imaging techniques, were used to visualize hinge flow. Experimentally measured parameters were used as input for CFD analysis. CFD simulations were made by solving the Navier-Stokes equation using a finite volume method with the pressure-based algorithm for continuity, and a pressure-implicit with splitting of operators (PISO) algorithm for pressure-velocity coupling. Moving grid methodology was employed to simulate periodic motion of the valve leaflets. CFD hinge flow results were visualized on four parallel planes at different depths in the hinge socket. The hinge flow patterns of the three types of bileaflet heart valve design are discussed.

RESULTS

Prominent vortex formation and stagnant flow areas were noticed in the pivot region of the MP valve. Vortices persisted throughout both the forward- and reverse-flow phases. These flow structures were not observed in the hinge areas of the SJM and On-X valves.

CONCLUSIONS

Vortex formation observed in the MP valve may contribute to the high thrombogenic potential of this valve. The absence of such vortices and areas of stagnant flow in the On-X and SJM valves indicate that hinge flow conditions in these valves do not favor mechanically induced thrombogenesis or thromboembolic events.

In vitro evaluation of the long-body On-X bileaflet heart valve

BACKGROUND AND AIMS OF THE STUDY

In order to optimize the length-to-diameter ratio, a series of circular aluminum rings with flared inlet and varying ring lengths, with internal diameters corresponding to that of 19 mm replacement prosthetic heart valve orifices, were tested in a steady-flow hydraulic system. The study aim was to determine the ring length-to-internal diameter ratio that produces the best hydraulic efficiency (i.e. lowest pressure gradient) within the physiologic flow rate range.

METHODS

Each ring was tested at flow rates of 10, 15, 20, 25 and 30 l/min and length-to-diameter ratio effect on hydraulic efficiency determined experimentally. The hydraulic effect was most significant for a ratio of about 0.6, with an increase to 1.2 providing little additional benefit. Thus, a ratio of about 0.6 was considered optimum in terms of hydraulic efficiency and incorporated into the design of the On-X bileaflet mechanical heart valve (BHV) series. An in vitro hydrodynamic study of the smallest (19 mm) and largest (25 mm) clinical On-X aortic valves was performed at two independent laboratories. Standard St. Jude Medical BHVs were used as the study controls.

RESULTS

Steady-flow experiments showed that the pressure gradient in the On-X valve was about 50% less than that of the comparable size control. The pulsatile flow study demonstrated a similar pressure gradient advantage. Laser Doppler anemometer velocity profiles taken downstream of the On-X valve at the aortic root showed typical characteristics of bileaflet valves, with three velocity peaks. The peak velocity reached 1.6 m/s for the On-X and 1.75 m/s for the control valve. A recirculating vortex was seen in the sinus cavity during the ejection period. This vortex, found in most aortic valves (including bioprostheses), is believed to provide a beneficial wash-out of the valve region and assist in valve closure.

CONCLUSIONS

These two independent studies clearly demonstrated that the elongated valve body and comparably larger flow area helped to improve the valve hydrodynamic performance, which is especially beneficial in the smallest (19 mm) size valve.

Time-frequency analysis of transient pressure signals for a mechanical heart valve cavitation study

A series of transient pressure signals (TPSs) can be measured using a miniature pressure transducer mounted near the tip of the inflow side of a mechanical heart valve (MHV) occluder during closure. A relationship appears to exist between the intensity and pattern of the TPS and the cavitation potential of a MHV. To study the relationship between MHV cavitation and the TPSs, we installed an MHV in a valve testing chamber of a digitally controlled burst test loop. A charge coupled device (CCD) camera and a personal computer based image grabbing program was used to visualize cavitation bubbles appearing on or near the occluder surface. One bileaflet MHV was used as the model for this study. Cavitation bubbles were observed within 300 microsec of the leaflet/housing impact. The valve was tested at various driving pressures between 100 and 1,300 mmHg. MHV cavitation bubble intensities were qualitatively classified into three categories: 1) strong, 2) weak, and 3) none. Digital images of the MHV occluder inflow surface were recorded simultaneously with the TPSs. TPSs were studied by the time-frequency analysis method (spectrogram) and correlated to MHV cavitation potential. The intensity of the cavitation bubbles was found to be associated with burst test loop driving pressures during leaflet closure.

Dynamic characterization of a new accelerated heart valve tester

This paper presents a new accelerated prosthetic heart valve tester prototype that incorporates a camshaft and poppet valves. A three element Windkessel system is used to mimic the afterload of the human systemic circulation. The device is capable of testing eight valves simultaneously at a rate up to 1,250 cycles/min, while the flow rate, the pressure, and the valve loading can be monitored and adjusted individually. The tester was characterized and calibrated using a set of eight Carpentier-Edwards bioprostheses at a flow rate varying between 3 and 5 L/min. The experiment was carried out with the pressure difference across the closed heart valve maintained between 140 and 190 mmHg. Smooth and complete opening and closing of the valve leaflets was achieved at all cycling rates. This confirms that the velocity profiles approaching the test valves were uniform, an important factor that allows the test valves to open and close synchronously each time.

Hydrodynamics of a long-body bileaflet mechanical heart valve

A new bileaflet substitute mechanical heart valve (MHV) that incorporates an optimal length-to-annulus ratio for improved hydrodynamic efficiency was recently introduced. Efforts have been made on this new prosthesis to use the current advances in carbon materials and design technology to achieve higher net forward flow with minimal energy loss for the smaller aortic valves, while reducing regurgitant closure volume for the larger size valves by an elongated orifice. Benchtop experiments performed in a pulsatile flow loop, under simulated physiologic conditions, substantiated these improvements as claimed.

Bileaflet mechanical heart valves at low cardiac output

Several earlier studies have indicated that bileaflet mechanical heart valves behave irregularly at low cardiac output and low pulse rate conditions, and that their hydrodynamic performances are generally inadequate. The authors conducted in vitro experiments in a pulsatile mock circulatory loop to compare the performance of the St. Jude Medical (SJM) valve and a long body bileaflet prosthesis recently introduced by Medical Carbon Research Institute (MCRI) (Austin, TX). The new MCRI mechanical heart valve model was designed with emphasis on improved hydrodynamic efficacy by introducing a long body with parallel leaflets and by leaflets increasing the flow area. Experimental studies were conducted on five test valves (MCRI 19 mm, MCRI 25 mm, SJM 19 mm, SJM 23 mm, and SJM 29 mm) with cardiac outputs of 2.0, 2.5, 3.0, and 3.5 L/min at a pulse rate of 40 beats/min, and 3.5, 4.0, 4.5, and 5.0 L/min at a pulse rate of 70 beats/min. Transvalvular pressure drop and closure volume were assessed by measuring the instantaneous ventricular and aortic pressures and aortic flow. The leaflet motions of the tested valves were observed by direct video recording using a charge coupled device camera while the flow measurements were being conducted. Testing under simulated physiologic ventricular and aortic pressure waveforms, the results of this study show that the MCRI bileaflets remained fully open during the entire ejection phase, even at very low cardiac output conditions (2.0 L/min). The closure volume (defined as percentage of forward flow volume) increased with decreasing cardiac output, as reported earlier by others. Comparative results also indicate that the MCRI design has nearly a two size pressure drop advantage over the SJM, with significantly smaller closure volume.

Transient pressure signals in mechanical heart valve cavitation

The purpose of this investigation was to establish a correlation between mechanical heart valve (MHV) cavitation and transient pressure (TP) signals at MHV closure. This correlation may suggest a possible method to detect in vivo MHV cavitation. In a pulsatile mock flow loop, a study was performed to measure TP and observe cavitation bubble inception at MHV closure under simulated physiologic ventricular and aortic pressures at heart rates of 70, 90, 120, and 140 beats/min with corresponding cardiac outputs of 5.0, 6.0, 7.5, and 8.5 L/min, respectively. The experimental study included two bileaflet MHV prostheses: 1) St. Jude Medical 31 mm and 2) Carbomedics 31 mm. High fidelity piezo-electric pressure transducers were used to measure TP immediately before and after the valve leaflet/housing impact. A stroboscopic lighting imaging technique was developed to capture cavitation bubbles on the MHV inflow surfaces at selected time delays ranging from 25 microseconds to 1 ms after the leaflet/housing impact. The TP traces measured 10 mm away from the valve leaflet tip showed a large pressure reduction peak at the leaflet/housing impact, and subsequent high frequency pressure oscillations (HPOs) while the cavitation bubbles were observed. The occurrence of cavitation bubbles and HPO bursts were found to be random on a beat by beat basis. However, the amplitude of the TP reduction, the intensity of the cavitation bubble (size and number), and the intensity of HPO were found to increase with the test heart rate. A correlation between the MHV cavitation bubbles and the HPO burst was positively established. Power spectrum analysis of the TP signals further showed that the frequency of the HPO (cavitation bubble collapse pressures) ranged from 100 to 450 kHz.

Numerical study of squeeze-flow in tilting disc mechanical heart valves

BACKGROUND AND AIM OF THE STUDY

The squeeze-flow that develops during valve closure is believed to cause cavitation in mitral mechanical heart valves (MHVs).

METHODS

Squeeze-flow was studied in tilting disc MHVs using two different numerical approaches. In the first decoupled analysis, experimental measurements of valve closing velocities were input into a computer model which simulated the resulting flow field. The second coupled approach involved simulation of the occluder motion and housing deformation in response to the surrounding flow.

RESULTS

Both models predicted the likelihood of cavitation during the squeeze-flow phase of valve closure. They also indicated that valve mounting compliance influences the squeeze-flow field. The coupled analysis also showed that squeeze-flow is influenced by fluid viscosity and the geometry of the contact region.

CONCLUSIONS

These parameters could therefore influence the inception of cavitation in tilting disc mitral MHVs.

Fluid dynamics of venous valve closure

In vitro experiment was performed on a stented bovine jugular vein valve (VV, 14 mm I.D. x 2 cm long) and a stentless bovine jugular vein valve conduit (10 mm I.D. x 6 cm long) in a hydraulic flow loop with a downstream oscillatory pressure source to mimic respiratory changes. Simultaneous measurements were made on the valve opening area, conduit and sinus diameter changes using a specially designed laser optic system. Visualization of flow fields both proximal and distal to the venous valve, and the valve opening area were simultaneously recorded by using two video cameras. Laser Doppler anemometer surveys were made at three cross sections: the valve inlet, the valve exist, and 2 cm downstream of the venous valve to quantity flow reflux at valve closure. The experiment confirmed that the VV is a pressure-operated rather than a flow-driven device and that little or no reflux is needed to close the valve completely. The experiment further demonstrated that the VV sinus expands rapidly against back pressure, a critical character to consider in venous prosthesis design.

Transient pressure at closing of a monoleaflet mechanical heart valve prosthesis: mounting compliance effect

An in vitro experimental study was performed to investigate the mounting compliance effect on the occluder closing dynamics and the transient pressure at the closing of the mitral Medtronic Hall (MH) mechanical heart valve (MHV). The closing velocity and the transient pressure were simultaneously measured at heart rates of 70, 90, 120, and 140 beats/minute with cardiac outputs of 5.0, 6.0, 7.5, and 8.5 liters/minute, respectively. The experiment was conducted under simulated physiologic ventricular and aortic pressures in a pulsatile mock flow loop. The characteristics of the transient pressure were investigated by detailed mapping of the transient pressure field in the atrial chamber using high frequency pressure transducers. Simultaneous measurements of the occluder closing velocity and the transient pressure around the seat stop of the MH showed that the transient pressure generated on the inflow side dropped below the vapor pressure of liquid during the occluder's sudden deceleration at closing. The amplitude of the transient pressure reduction (TR) was proportional to the occluder approaching velocity. The development of the transient pressure in the rigid and flexible mountings were significantly different. In the rigid mounting (RM), the pressure was reduced below the liquid's vapor pressure and maintained below -350 mmHg for approximately 180 microseconds. Strong signals of high frequency pressure oscillations (HPO) were recorded in the transient pressure traces. The timing of the HPO was found to be consistent with that of the cavitation bubble collapse as observed by others. In the flexible mounting (FM), TR also occurred, but recovered quickly and was followed immediately by a positive pressure spike. Relatively weak HPO appeared in the transient pressure trace. The mapping of the transient pressure field showed that both the transient pressure reduction (on the major orifice side) or rise (on the minor orifice side) as well as the HPO were locally generated near the valve occluder surface. The transient pressure attenuated with distance away from the occluder surface. The HPO were detectable as far as 40 mm away from the occluder surface. The rigid mounting pressure signals showed characteristically two occurrences of high frequency pressure oscillations. The HPO with smaller amplitude occurred first after the initiation of the TR, followed by a burst of strong HPO at about 450 microseconds. It is believed that they were the result of the collapse of cavitation bubbles. The strong HPO did not appear in the flexible mounting signals. The study indicated that the mounting compliance played a significant role in the MHV cavitation inception and the subsequent bubble growth. It also suggested the possibility of detecting the cavitation by using a high frequency pressure transducer positioned in the atrial chamber.

Asynchronous closure and leaflet impact velocity of bileaflet mechanical heart valves

The effect of gravitational field on the asynchronous closure of four different types of bileaflet heart valves (BHV) were investigated in an in vitro mock circulation system. The experimental study involved the 29 mm St. Jude Medical Standard, 29 mm CarboMedics, 29mm Edwards-Duromedics and 29 mm Edwards-Tekna BHVs. The valves were tested in the mitral position on an inclined 45 degrees plane of the pulsatile mock flow loop. The test valves were orientated with their pivotal axis horizontal so that the gravitational vectors on the two leaflets were clearly disparate. Using a specially designed laser optical system, the time difference at which the closing leaflets made their first contacts with the valve housing was measured. The closing velocity of each leaflet was measured separately by the laser sweeping technique (LST). The experiments were conducted under physiologic ventricular and aortic pressures at the heart rates of 70, 90, and 120 beats/minute with the corresponding flow rates of 5.0, 6.0, and 7.5 liters/minute. Asynchronous closing motions were registered in all four tested BHVs. The leaflet closing motions were random in nature, but indicated clear dependency on the orientations with respect to the gravitational field. The lower leaflet which makes a generally level swing to close, was assisted initially by the gravity and was always found to close earlier than the upper leaflet, which swing upward to close. The initial closure of the upper leaflet was against the gravity which delayed the leaflet motion. Depending on the BHV designs, the time delays between the two leaflets were found to vary from beat to beat but to follow certain probability distributions. In general, the leaflet/housing contact time between the two valve leaflets exhibited the clear trend of decreasing delay time at closure with the increase of the heart rate. For each of the valves tested, the average impact velocity of the first closing leaflet was found always smaller than that of the second closing leaflet at all three heart rates tested. The impact velocities of both the BHV leaflets were found to increase with the heart rate. The difference in the closing velocities between the two leaflets decreases with the increase of the heart rate and is generally proportional to the impact time delay between the two leaflets.

Ventricular pressure slope and bileaflet mechanical heart valve closure

The maximum left ventricular pressure slope (dP/dt) value has been used by several investigators as the criterion for studying mitral valve closure. In this article, the relationship between the ventricular pressure slope (dP/dt) and the leaflet closing behavior of bileaflet mechanical heart valves (BMV) is investigated. Two current BMVs, the St. Jude Medical 29 mm and CarboMedics 29 mm, installed in the mitral position of a mock circulatory pulsatile flow loop were used as the study model. Under simulated physiologic pressures and flow conditions, the experiment was conducted at 70, 90, and 120 beats/min with corresponding flow rates of 5.0, 6.0, and 7.5 liters/min, respectively. A laser sweeping technique was used to monitor the leaflet closing motion within the last 3 degrees excursion at valve closure. A modified dual beam laser sweeping technique system was used to register the difference of leaflet/housing impact time between the two BMV closing leaflets in asynchronous closure. Common BMV asynchronous closures were found in both BMVs at all three heart rates tested. The second closing leaflet was found to always close at higher velocity than the first. Simultaneous measurements of the ventricular pressure (Pv) and the leaflet closing time showed that Pv exhibited three stage characteristics. In the first stage, Pv gradually increased as the ventricle was filled. A sudden rise of Pv occurred immediately after closing of the first leaflet. The maximum dp/dt occurred in the third stage after closure of both BMV leaflets. The BMV closing behavior and the corresponding Pv pattern were found to depend strongly upon valve type and heart rate. The time averaged ventricular pressure slope (dp/dt) values at 70, 90, and 120 beats/min were about 40, 70, and 150 mmHg/sec for the St. Jude Medical valve and 40, 105, and 205 for the CarboMedics valve during the first closing stage. The maximum dp/dt values were 2670, 4350, and 5000 mmHg/sec for the St. Jude Medical valve and 1210, 2530, and 3210 mmHg/sec for the CarboMedics valve at the three heart rates tested, respectively. The study showed that the left ventricular pressure patterns (dP/dt) at valve closure were the result of valve operation under given driving conditions. The dp/dtmax cannot be used as the criterion for studying BMV closure.

The closing velocity of mechanical heart valve leaflets

The closing motion of the occluder leaflets in bileaflet type mechanical heart valves (MHV) was monitored with a laser sweeping technique. The angular displacements of the leaflets were registered with precision of 0.2 microsecond steps. Experimental measurements were made using five 29 mm Edwards-Duromedics including three original specification (EDOS) and two modified specification (EDMS), and two 29 mm St Jude Medical MHVs. The testing valve was installed in the mitral position of a physiologic pulsatile mock circulatory flow loop using water-glycerine solution as the testing fluid. Each valve was tested by: (1) direct mounting the valve on metal washers, and (2) mounting the valve with its sewing ring. Experiments were carried out at pulse rates of 70, 90, and 120 beats min-1, with the corresponding cardiac output of 5, 6, and 7.5 litres min-1, and maximum left ventricular pressure gradients (dp/dt) of 1,800, 3,000 and 5,600 mm Hg s-1, respectively. The maximum leaflet closing velocity of each of the tested valve types are presented. The difference in leaflet closing movements between the direct rigid mounting and the sewing ring mounting are discussed. The details of the laser sweeping technique are presented.

A squeeze flow phenomenon at the closing of a bileaflet mechanical heart valve prosthesis

In vivo cavitation in cardiovascular flow fields may occur under very unusual circumstances as a localized transient phenomenon which are confined to very small regions in the vicinity of the valve body or leaflet surface. The violent collapse of cavitation bubbles induces local erosion that may lead to structural damage. The fluid mechanical factors that may cause in vivo cavitation inception in mechanical heart valve (MHV) prostheses are investigated. It is established that the closing velocity of the leaflet holds the key to MHV cavitation. During the final phase of valve closing, the fluid mass in the gap space between the closing occluder and the valve's body is squeezed into motion by the rapidly approaching boundaries. The flow pattern created by this motion (termed 'squeeze flow'), is found to be related to the valve geometry, and the impact velocity of the closing leaflet. Given the closing velocity of the leaflet and the geometry of the MHV, computational flow dynamics (CFD) are made to determine the velocity distributions in the gap flow field of a bileaflet MHV in the mitral position. A two dimensional, time dependent model of the gap space show that flow velocity in the gap space can reach values as high as 30 ms-1 in regions near the edge of the inflow surface of the Edwards Duromedics (ED) MHV leaflet. This high speed stream ejected from the gap channel can create the conditions that characterize cavitation. The location of the isolated high speed region corresponds to the surface erosion that was observed in a number of damaged ED-MHV explants.

On the transport of a suspended particle in flow through the entrance region of a tube

The motion of a single, spherical particle, released at different radial positions at the inlet of the entrance region of a straight circular laminar flow tube (Re = 260), was studied theoretically. Radial migration of the particle, either toward the tube center or toward the tube wall, was predicted. Based on the hypothesis that the particle experienced a lift force which was produced by the vorticity in the boundary layer and a velocity difference between the center of the suspended particle and the fluid medium, an inertia-vorticity fluid dynamic model was formulated to analyze the particle radial motions. Computational flow dynamics (CFD) solutions obtained from a 9.8 mm diameter tube model included the resulting particle loci for three particle radii (a = 0.1 cm, 0.085 cm, 0.050 cm), with the particle entry at various radial positions. The computation also covered a range of different particle entry speeds. The results showed that the particle migrates toward the tube center if it lags behind the medium in the core region; otherwise, it migrates toward the tube wall. Additional flow experiments were conducted in a circular (2R = 10.2 mm), 300 mm long straight tube. A small polystyrene sphere (2a = 1.72 mm, density rho p = 1.014 g.cm-3) was released at the inlet (X = 0, eta/R = 0.48) with two dimesionless release velocities (omega p = 0, and omega p > 1.0). The recorded particle traces agree well with the computational model.

The closing behavior of Medtronic Hall mechanical heart valves

The rapid deceleration of mechanical heart valve leaflets or discs at closure, and their rebound after impact between the leaflets or discs and housing may produce conditions that favor cavitation inception. Precision measurements of the disc closing velocity using a laser sweeping technique (LST) were made on two Medtronic Hall (Med Hall) mitral valve models, the 29 mm Standard and D-16 Med Hall valves. The experiment was carried out in a pulsatile mock flow loop (PFL) by installing the tested valve in the mitral position of the PFL. Tests were conducted under physiologic pressures at heart rates of 70, 90, and 120 beats/min with flow rates of 5, 6, and 7.5 l/min, respectively. For each valve tested, the experimental series was carried out to include both a flexible mounting and a rigid mounting mechanism. The time history of the disc closing velocity during the last 3 degrees before final closure was measured. Results show that the disc approached the valve housing at a near constant velocity of approximately 1.3, 1.5, and 2.0 m/sec for the Med Hall Standard valve, and 1.9, 2.5, and 2.6 m/sec for the Med Hall D-16 valve at 70, 90, and 120 beats/min, respectively. After impact, the valve of disc rebounded with velocities that depended on the mounting modalities and heart rates. In the rigid mounting, the disc rebound velocities of the Med Hall Standard and D-16 valves were about 60%-70% and 70%-80% of their respective approaching velocities. For the flexible mounting, the rebound velocities were 15%-25% and 20%-30% of their respective approaching velocities. The results demonstrate that a slight modification in the seat stop geometry of Med Hall valves can significantly affect disc closure behavior.

An experimental-computational analysis of MHV cavitation: effects of leaflet squeezing and rebound

A combined experimental-computational study was performed to investigate the flow mechanics which could cause cavitation during the squeezing and rebounding phases of valve closure in the 29 mm mitral bileaflet Edwards-Duromedics (ED) mechanical heart valve (MHV). Leaflet closing motion was measured in vitro, and input into a computational fluid mechanics software package, CFD-ACE, to compute flow velocities and pressures in the small gap space between the occluder tip and valve housing. The possibility of cavitation inception was predicted when fluid pressures dropped below the saturated vapor pressure for blood plasma. The computational analysis indicated that cavitation is more likely to be induced during valve rebound rather than the squeezing phase of valve closure in the 29 mm ED-MHV. Also, there is a higher probability of cavitation at lower values of the gap width at the point of impact between the leaflet tip and housing. These predictions of cavitation inception are not likely to be significantly influenced by the water-hammer pressure gradient that develops during valve closure.

Occluder closing behavior: a key factor in mechanical heart valve cavitation

A laser sweeping technique developed in this laboratory was found to be capable of monitoring the leaflet closing motion with microsecond precision. The leaflet closing velocity was measured inside the last three degrees before impact. Mechanical heart valve (MHV) leaflets were observed to close with a three-phase motion; the approaching phase, the decelerating phase, and the rebound phase, all of which take place within one to two milliseconds. The leaflet closing behavior depends mainly on the leaflet design and the hinge mechanism. Bileaflet and monoleaflet types of mechanical heart valves were tested in the mitral position in a physiologic mock circulatory flow loop, which incorporated a computer-controlled magnetic drive and an adjustable afterload system. The test loop was tuned to produce physiologic ventricular and aortic pressure wave forms at 70-120 beats/min, with the maximum ventricular dp/dt varying between 1500-5600 mmHg/sec. The experiments were conducted by controlling the cardiac output at a constant level between 2.0-9.0 liters/min. The measured time-displacement curve of each tested MHV leaflet and its geometry were taken as the input for computation of the squeeze flow field in the narrow gap space between the approaching leaflet and the valve housing. The results indicated rapid build-up of both the pressure and velocity in the gap field within microsecs before the impact. The pressure build-up in the gap space is apparently responsible for the leaflet deceleration before the impact. When the concurrent water hammer pressure reduction at closure was combined with the high energy squeeze jet ejected from the gap space, there were strong indications of the environment which favors micro cavitation inceptions in certain types of MHV.

Closing behavior of a new bileaflet mechanical heart valve

Recently, the in vivo cavitation potential has become a primary concern among manufacturers of new mechanical heart valves (MHV). An experimental/computational program was designed to investigate each of the flow parameters involved. It was established that the closing velocity of the leaflet holds the key to MHV cavitation. One of the novel concepts of the new bileaflet mechanical heart valve (1205-MHV) was its ability to operate with a relatively small angular excursion that led to a much smaller closure velocity at impact (as compared with control valves). This is believed to significantly reduce the cavitation potential. The 1205-MHV is characterized by a longer valve body, with the hinges protruding further upstream. The unique design allows the valve the freedom to open as much as 90 degrees. The closure velocities are reduced by a smaller leaflet excursion (50 degrees), combined with a floating hinge that allows absorption of part of the impact energy at closure. The impact velocities of the 1205-MHV leaflets at closure were measured by a laser sweeping technique that monitored the leaflet closing motion with a precision of 5 microseconds within the last 3 degrees before impact. The 27 mm 1205-MHV (the largest size) was tested in the program by mounting the valve in the mitral position of a physiologic mock circulatory loop. The valve was tested at 70, 90, and 120 bpm. The results were compared with those of a St. Jude Medical 29 mm MHV. The closure velocities measured with the 1205-MHV were significantly lower than those measured with the control valve.(ABSTRACT TRUNCATED AT 250 WORDS)

In vitro performance of venous valve prostheses. An experimental model study

The performance of fresh and glutaraldehyde fixed bovine jugular vein valves (10 mm diameter) was investigated in an experimental flow loop that provides adjustable flow rates and a downstream oscillatory pressure. Three different venous valve (VV) conduit geometries (curved [C], straight [S], and tapered [T]), were tested. The flow loop consisted of two independently adjustable components, with the mean flow generated by adjusting the elevation difference between the head tank and outflow chamber. An adjustable sinusoidal pressure pulse was superimposed on the downstream of a VV to mimic the respiratory effects. Flow visualizations were made using 100 microns mica chip tracers in the laser illuminated flow fields. To assess VV performance under various flow conditions, the closure opening (CO), partial opening (PO), and leaflet fluttering (LF) were evaluated. At a given pulse pressure, the three conduits required different flow rates to reach CO mode. At 12 cm H2O pulse pressure, the fresh valve in C-conduit exhibited stable CO operation at a flow rate of 1.01 ml/sec. That in S and T conduits required 1.67 ml/sec and 2.25 ml/sec, respectively. At higher flow rates, PO and LF performances were observed in all three conduits. Different threshold values of pulse pressure were needed to reestablish the CO operation mode for the C, S, and T conduits, individually. These observations provide some insight into the role of conduit geometry and sinus configuration in the function of VV.

Fluid mechanics of left ventricular assist system outflow housings

Using the Novacor (Baxter Novacor, Oakland, CA) Left Ventricular Assist System (LVAS) as a test bed, phasic flow patterns were analyzed for three outflow valve housing designs: 1) a triple sinus; 2) an axisymmetric concentric sinus (CS); and 3) a modified triple sinus (MS). The 21 mm Carpentier-Edwards trileaflet pericardial heart valve prosthesis was used for all experiments done on the three housing designs. The LVAS was actuated by a laboratory model of the Novacor LVAS control console, and it was connected to a mock flow loop with an adjustable afterload system to provide physiologic pressures and flows (Pao, 120/80 mmHg; pump output [PO], 2-6 L/min). Laser illuminated flow visualization techniques were used to investigate the phasic flow patterns of the housings, and the visualization derived velocity was verified by laser Doppler velocimetry at several selected points in the field. Formation of vortices behind the leaflets during the LVAS ejection phase was observed in each of the housing designs. They were well organized, and they circulated with the greatest strength in MS. These vortices tended to lie in a plane parallel to the main flow axis, with the rotational velocity increasing with the stroke volume of the LVAS. In the CS and the MS housings, a circumferential flow that provided good washing of this region was observed behind the stents.

On transport of suspended particulates in tube flow

Blood cells suspended in shear flows exhibit much larger dispersive motions than those predicted by the Stokes-Einstein formula for Brownian diffusion. The lateral migration and the erratic motions of the 8 microns red blood cells (RBC) is thought to be analogous to a diffusive process. It is shown that the often cited convective-diffusion theory may not be an adequate model for describing the transverse migration of suspended cells in blood flow. A comprehensive review of both the classical theory and of contemporary work in particle transport is presented, with particular emphasis on low Reynolds number tube flows. The mechanisms of Taylor dispersion, the effects of Brownian perturbations on translational and rotational motions of the suspended particles in shear fields, and the influence of integratable and chaotic advections, are individually examined. The classical experiment by Segre and Silberberg (1962) lead us to believe that particle hydrodynamics may play an important role in transverse migrations. In this light, we have further examined the hydrodynamic aspects of the so-called "tubular pinch" effect, the lateral migration of rigid spheres. We have also discussed the transverse motions of liquid drops, and the reversibility of the organization of suspensions in transport. The convective accelerations in the entrance region of a tube can produce relative velocities between fluid medium and various type of particulates if there is a difference in density. The deformable RBC, an "active-type" particle, can provide feedback to the flow from both mass and momentum considerations; the more rigid platelet, a "passive-type" particle, will experience a much smaller relative velocity as compared to the RBC. We may expect that particles of different densities are transported to different equilibrium annular positions before entering the fully developed flow region. The erratic, lateral movement of suspended particulates in steady laminar tube flow can be described by the usual Lagrangian coordinates.

Venturi pressure cannot cause cavitation in mechanical heart valve prostheses

Surface pitting of certain mechanical heart valve (MHV) explants has prompted investigation into possible causes of cavitation during MHV operation. Leaflets of a 29 mm MHV were glued shut with B-datum (BD) gaps fixed at 0.0089, 0.0174, and 0.0219 cm. Each BD gap setting was tested in a steady flow chamber, with leakage flow established at transvalvular pressures (delta P) of 20 to 200 mmHg. Laser Doppler velocimeter (LDV) velocity measurements were recorded 220 microns distal to the BD, along with leakage flowrates. Maximum LDV velocities were compared with those calculated using the mass conservation equation. At identical P, the LDV flow velocities for the three BD settings were found to be approximately equal. This indicates a geometric independence of the leakage flow velocity. At atmospheric pressure, the local velocity necessary to cavitate blood as a liquid is approximately 13 m/sec. These results demonstrate that the leakage velocity is insufficient to cause cavitation. A simplified theoretical model is proposed to illustrate the necessary delta P to produce Venturi related cavitation.

Haemodynamics of angioaccess venous anastomoses

A large percentage of arteriovenous haemodialysis angioaccess loop grafts (AVLG) fail within the first year after surgery, the occlusive lesions being found predominantly at the venous anastomosis site. This paper presents a detailed flow dynamic study of the AVLG system using three elastic, transparent bench-top flow models, which were based on the geometry of silicone rubber casts obtained at different times from a chronic animal model. Each model thus represented a different stage of the lesion development. Flow visualization and laser Doppler anemometer surveys of the flow field confirmed that the hydrodynamic factors favour lesion development near the stagnation point opposite the anastomotic toe, where the momentum of the impinging jet stream, combined with the oscillating wall shear stress generated in the vicinity of the stagnation point, acts in both directions. The accumulation of tracer particles in the region of flow separation is believed to be a combined contribution from the hydraulic forces and the inward motion of the vessel wall. As these hydrodynamic factors are enhanced upon further development of the occlusive lesion, a vicious cycle may be formed.

Laser assessment of leaflet closing motion in prosthetic heart valves

The dynamics of leaflet motion in heart valve prostheses (HVP), and in particular the closing velocity, is believed to be related to the valve sound and possibly to the phenomenon of valve cavitation. This paper describes a non-intrusive laser sweeping technique enabling the study of leaflet motion. The principle of measurement and the equipment involved are presented, together with the results of two commercially available, 29 mm bileaflet mitral valves, a St. Jude Medical, and an Edwards Duromedic valve. Experiments were carried out in a pulsatile mock flow testing loop designed to mimic physiological pressure waveforms and ventricular contraction. Measurements of heart rate were made in the range 70-120 beats min-1, with a ventricular pressure slope range of 1800-5600 mm Hgs-1 and a cardiac output range of 5.0-7.5 litres min-1. Motion analysis of the measured data focuses on the velocity of the leaflet immediately before closure.

Estimation of the rotational undamped natural frequency of bileaflet cardiac valve prostheses

The angular momentum balance is solved numerically for a size 29 mm CarboMedics prosthetic heart valve. The lift force is estimated from potential flow theory, while the drag force is estimated from the lift force and a blunt body empiricism. Buoyancy and gravitational effects are calculated based on the assumption of homogeneous leaflets. Other assumptions include uniform flow, negligible friction at the pivot axis, negligible viscous damping and fluid inertance, and a symmetry flow condition. Oscillations are predicted in the opening dynamics in the range of 2-25 Hz, for flow rates through one-half of the orifice in the range of 0.1-10.0 l/min. The frequency of these oscillations is dependent upon the orientation of the leaflet in relation to the gravitational field and the magnitude of the flow rate. In vivo and in vitro measurements by other investigators demonstrate similar effects of gravity and flow rate, with flutter frequencies of the order of 10-100 Hz. Excitation frequencies, based on vortex shedding, are estimated to be of the order 2-200 Hz, for the range of flow rates of the theoretical model. These results suggest that the natural frequency of this rotational second order system may, in theory, be a contributing factor to the flutter observed in bileaflet cardiac valve prostheses. The clinical significance of this finding is yet to be established.

The closing velocity of Baxter Duromedic heart valve prostheses

The closing velocity of a mechanical heart valve (MHV) leaflet has been conventionally related to the valve sound, and possibly to recently observed MHV cavitation phenomenon. Presently, the MHV leaflet terminal closing velocity has been indirectly assessed either by listening to the valve sound and/or averaging over the closing period. Leaflet motion during the closing phase is not uniform. Using a laser sweeping technique developed in this laboratory, an experimental study was carried out to analyze the closing motion of Baxter Duromedic (DM) 29 mm mitral MHVs. The results were compared with that of a St. Jude Mechanical (SJ) 29 mm mitral MHV, using the same technique. The in vitro experiment was carried out by mounting the testing MHV at the mitral position of a mock circulatory testing facility, with heart rates ranging from 70-120 beats/min, ventricular pressure slopes (dp/dt) from 1800-5600 mmHg/sec, and cardiac outputs from 5.0-7.5 liters/min. This paper introduces the laser sweeping technique developed for precision monitoring of MHV leaflet motion, and presents the detailed leaflet motions within the last 3 degrees before final closure. The experimental results showed that the final closing velocity of DM, which is known to generate a louder valve sound and has a shorter closing period than that of SJ, actually has approximately the same closing velocity within the range of the experiment. Theoretical analysis further confirms that a short closing period may not lead to a higher leaflet final closing velocity.

Graft compliance and anastomotic flow patterns

The oscillatory flow patterns at the venous anastomosis of a hemodialysis angioaccess loop graft system were studied using two new compliant vascular prostheses: a longitudinally compliant polytetrafluoroethylene-composite (Baxter Ultraflex PTFE-Plus) graft (BA) and a radially compliant ultrafine polyester fiber (TORAY-UFPF) graft (TR). A non-compliant Gore-Tex polytetrafluoroethylene graft was used as the control. The experimental grafts were 8 mm inside diameter x 25 cm long. Flow experiments were done in a transparent, elastic bench-top flow model; fabrication was based on silicone rubber casts obtained from femoral-to-femoral arteriovenous loop grafts surgically implanted in dogs. The loop graft constructed in the dog model was made to mimic the branchial-to-cephalic angioaccess loop graft commonly used in hemodialysis patients. The flow model was connected to a pulse generator, an adjustable arterial afterload, and a venous afterload. Under identical input conditions, the pressure and flow waveforms were monitored simultaneously at the proximal and distal ends of both the arterial and venous anastomoses. For each graft studied, the anastomotic flow field was visualized using laser illuminated hydrogen bubbles as tracers. At pulse rates of 60 and 90 beats/min, graft flow rates were 2.2 and 2.5 L/min, respectively. Among the grafts studied, measurable differences in pressure and flow wave attenuation and their respective phase lags resulted in characteristically dissimilar flow patterns at the venous anastomosis. Growth of the separation zone at the toe of the anastomosis, and the pattern of retrograde flow in the distal vein are visibly different in all three grafts.

Vortex shedding in bileaflet heart valve prostheses

A dynamic study of two geometrically similar bileaflet heart valve prostheses (HVP) was performed using a physiologic mock circulatory flow loop. The HVPs studied were the 25 mm St. Jude Medical (SJM) and the 25 mm Carbomedics (CMI) in the aortic position and the 27 mm SJM and 27 mm CMI in the mitral position. All data were collected at a heart rate of 70 beats/min and a cardiac output of 5.0 L/min. Flow visualization was conducted in the transparent flow chambers of the pulsatile mock circulatory flow loop using a 15 mW He-Ne laser light source. A cylindrical lens and optics system converted the incident laser beam into a thin parallel light plane, and 420 microns tracer particles were suspended in the testing fluid to illuminate the flow field at selected planes. Frame-by-frame analysis of the 16 mm high-speed cine provides detailed phasic flow patterns in the vicinity of the HVP. A series of still photographs of flow patterns, taken at approximately 22.5 degrees phase intervals, are sequentially presented for each HVP. In the aortic position, a Karman-like vortex pattern appears downstream of the SJM at the end of the ejection phase. The CMI exhibits a rather symmetrical ejection flow pattern that turns into random motion immediately after the onset of ejection. In the mitral position, the SJM again exhibits a strong core flow during ventricular filling, whereas the CMI produces a more diffuse pattern during the same period. A pair of vortices shed from both the SJM and CMI are clearly visible toward the end of the ventricular filling phase. The vortex mechanisms are discussed in light of leaflet boundary layer formation.

Bursting strength in CO2 laser-assisted microvascular anastomoses

Iliac artery end-to-end anastomoses were performed in 42 Sprague-Dawley rats, divided into seven groups, to determine the welding effects of CO2 laser radiation in microvascular anastomoses. Conventional suture techniques were performed on right iliac arteries, and left iliac arteries were anastomosed with a laser-assisted technique. Bursting strength and diameters of the anastomotic sites were measured at different intervals (from one day to five weeks) post surgery. The anastomotic patency rate was 100 percent in both groups, and the aneurysm rate was only 2 percent in the laser group. Bursting strength was low at one and three days post surgery in both groups; then, it increased gradually until both groups could withstand higher than physiologic pressures. Results of high patency rates, low aneurysm formation, and the ability to withstand pressures higher than physiologic, suggest that the laser-assisted anastomotic technique can play an important role in microvascular surgery.

Flow phenomena in compliant and noncompliant arteriovenous grafts

Oscillatory flows at venous anastomoses of two end-to-side angioaccess grafts were studied using: 1) a noncompliant Gore-Tex polytetrafluoroethylene (PTFE) graft and 2) a longitudinal compliant Ultraflex graft. The study was done in a transparent flow model fabricated from a femoral-to-femoral arteriovenous loop graft (AVLG) surgically implanted in dogs. The 6 mm x 25 cm loop graft reproduces that of the brachial artery-cephalic vein hemodialysis angioaccess loop graft commonly used in patients. At a pulse rate of 70 beats/min and blood flow of 0.7 L/min, flow visualization was done using a 15 mW He-Ne laser and a cylindrical lens system which converts the incident laser beam into a thin parallel light plane. Hydrogen bubble technique was used to illustrate the flow field at selected planes. Phasic pressure and flow oscillations were observed in the hypertensive host vein. The pressure wave arrived at the venous anastomosis with higher amplitude and shorter time of travel in the Gore-Tex graft than in the Ultraflex graft. The differences in wave attenuations and wave transmission time between grafts resulted in apparent changes of flow pattern in both the proximal (heart side) and distal (foot side) veins. The negative lags between the flow peaks in the distal vein and proximal artery are explained by the phasic pressure gradients measured at the anastomoses. Spectral analysis was done on the venous thrills, which are characteristic of almost all clinical AVLG implants. The power spectra taken at 60, 70, 90, and 120 beats/min clearly demonstrated a concentration of fluctuation energy at about 100 Hz in the Gore-Tex AVLG.

Segmental volume distensibility of the canine thoracic aorta in vivo

Segments of the canine ascending aorta, upper descending thoracic aorta, and middle descending thoracic aorta were instrumented with ultrasonic dimension gauges and a cathetertip manometer simultaneously to measure changes in segment diameter, length, and intravascular pressure. Volume distensibility (EV) was calculated as the sum of circumferential extensibility (EC), longitudinal extensibility (EL), and high order extensibilities (EK) for each segment. The EC and EL were linear expressions that represented percentage volume changes per mmHg pulse pressure due to circumferential and longitudinal dimensional changes. The high order extensibilities (second and third order) accounted for the percentage volume changes per mmHg pulse pressure due to the interactions between circumferential and longitudinal dimensional changes. Mean(SEM) EV values from six dogs were 1.62(0.31), 0.84(0.08), and 0.62(0.08)% delta V/mmHg delta P for the ascending aorta, upper descending thoracic aorta, and middle descending thoracic aorta segments respectively. The EV, EL, and EK of the ascending aorta segment were significantly greater than those of the upper descending thoracic aorta and middle descending thoracic aorta segments, whereas EC was significantly less in the ascending aorta than in both the upper descending thoracic aorta and middle descending thoracic aorta segments. It is concluded that there are regional differences in aortic distensibility and its components in vivo. Longitudinal wall motion is an important determinant of these aortic mechanical properties.

Dynamic analysis of flutter in disk type mechanical heart valve prostheses

Parametric study of the low frequency oscillations occasionally observed in certain types of disc type prosthetic heart valves (PHV) are carried out using a finite element technique. The analysis is performed to determine the frequencies of the dynamic fluttering with the help of the 'ANSYS' computer program. The results show that the frequencies of the dynamic fluttering for both the circular occluders and the semi-circular occluders are at least two orders of magnitude higher than that observed in vivo. It is thus concluded that the clinically observed leaflet oscillations should not be a dynamic flutter phenomenon. Rather, the vortex shedding has been assumed to be the cause of these oscillations.

Phasic flow patterns at a hemodialysis venous anastomosis

A phase-by-phase analysis of local flow patterns at the venous anastomosis of an arteriovenous hemodialysis angioaccess loop graft (AVLG) was made. The study was carried out in an elastic, transparent Silastic in vitro flow model, which duplicates the detail geometry of the AVLG obtained from an animal model (30+ kg dogs with 12 weeks bilateral femoral AVLG implantation). The flow model was installed in a mock pulsatile flow loop system designed to simulate physiological conditions. Flow visualization was made in laser-illuminated flow fields using a high-speed cine camera. Analysis of the high-speed cine indicates there is a distinct separation region downstream of the anastomotic toe in the median plane and a stagnation region that oscillates along the opposite wall. During inward motion of the vessel wall, accumulation of particles in the separation region and the nearby stagnation region is observed. A large swirl appears in the distal vein during end-systolic period. A double-helical flow pattern occurs further down in the distal vein. Retrograde flow in the distal vein occurs in an "oscillating" manner following each cardiac cycle.

Flow profiles and wall shear stress distribution at a hemodialysis venous anastomosis: preliminary study

The phasic velocity field in the vicinity of the venous anastomosis in a hemodialysis angioaccess arteriovenous fistula loop graft (AVLG) is investigated employing a laser Doppler anemometer (LDA) system. Detailed LDA velocity profiles are obtained by sectional survey performed in a transparent, elastic flow model which was fabricated to represent the geometry of the AVLG system under physiological pressure and flow waveforms. The geometry of the flow model was based on a silicone rubber cast obtained from an experimental dog model. In the present study, detailed distribution of velocity profiles is obtained. The distribution of wall shear stress in the model is computed from the slope of the local velocity profiles near the wall. The relationship between the results obtained by flow visualization and the LDA measurement is discussed.

Developing oscillatory flow in a circular pipe: a new solution

The problem of oscillatory flow in a circular pipe was analyzed by Atabek and associates more than two decades ago. Their formulas for velocity and pressure distributions in developing pipe flows under oscillatory conditions have been often cited. However, the application of these formulas for flow field computations requires a rather complex procedure involving plotting of a set of curves and predetermination of the phase angles. This paper presents a method using the imaginary argument of the Bessel function to solve the Navier-Stokes equations. A different set of solution formulas are obtained. A comparison of the formulas obtained in this paper with those of Atabek shows that the former is considerably simpler and more convenient to use in flow computations. Numerical results computed using this paper's formulas are consistent with Atabek's and with the experimental measurements.

Wall shear stress distribution in a model human aortic arch: assessment by an electrochemical technique

Wall shear stress (WSS) distribution in a human aortic arch model is studied using 130 cathode electrodes flush-mounted on the model walls. Flow visualizations are made in a transparent geometry model to identify the regions of fluid mechanical interests, e.g. regions of flow separation, eddy formation and flow stagnancy. The 130 electrodes are strategically positioned in the arch based on information obtained from the flow visualizations. The measured data indicate that the aortic arch may be categorized into eight regions: three along the inner wall of the arch (A,B,C); and five near the outer wall (D,E,F,G,H). (1) The regions of low WSS are distributed along the inner wall of the ascending aorta A; the inner wall of the descending aorta C; and the upstream inner wall of the innominate and the common carotid branchings F. (2) The high WSS regions are distributed along the outer wall of the arch E; and the inner wall in the arch opposite to the left subclavian branching B. (3) In certain regions, high and low WSS may be found next to each other (e.g. G and H) without a definable boundary in between; and (4) as the Reynolds number increases, the areas of low WSS decrease, while the high WSS areas increase with no obvious change in magnitude of the stress along the inner wall of the arch. At the branchings, the WSS distribution is not affected by the Reynolds number within the range of observations. The measured WSS distribution is compared with Rodkiewicz's map of early atherosclerotic lesions in the aortic arch of cholesterol fed rabbits.

Human red blood cell hemolysis in a turbulent shear flow: contribution of Reynolds shear stresses

Various previous models used in studying red blood cell (RBC) hemolysis in turbulent shear flows are reviewed from a fluid dynamic point of view. The effect of turbulent shear stress (Reynolds shear stress, tau R) on RBC hemolysis is investigated utilizing a submerged axisymmetric jet flow field. A detailed survey of the flow field is made with a laser Doppler anemometer system to obtain contour maps of the mean velocity distributions, relative turbulence intensities, and tau R distributions in the field prior to conducting the experiment of sampling and analyzing the cells free-hemoglobin in the field. A new two-point sampling technique, developed in this laboratory, allows collections of RBC samples from selected locations in the flow field so that a relationship between the local shear stress level and the cell damage may be established. The threshold level of tau R responsible for incipient hemolysis is found to be approximately 400 Newtons per square meter (N/m2), below which a sublethal region of zero hemolysis is observed.

Influence of red blood cell concentrations on the measurement of turbulence using hot-film anemometer

Measurement of local velocity fluctuations was made with an L-shaped conical hot-film probe in a submerged circular jet. The experiment was carried out in solutions of washed human red blood cells (RBC) in a phosphate buffer solution (PBS), at hematocrit concentrations (Ht percent) of 10, 19, 29, and 38 percent. The viscosity of the testing solutions was kept at 3.2 c.p. by adding proper amount of dextran. The experiment was conducted at Reynolds numbers (NR) 674, 963, 1255 and 1410, based on the jet exit velocity and exit diameter. Statistical analyses were performed on the recorded instantaneous velocity signals to obtain the root-mean-square (rms) values, the probability density functions (PDF) and the power spectral density functions (PSDF) of the signals. Within the range tested, we noticed an incidental rise in rms values at 19 to 29 Ht percent for NR = 963 similar to those reported earlier in the literature. Further analyses using PDF and PSDF, however, showed neither a trend nor any physical significance of this rise. Based on the analyses of both the PDF and the PSDF, we believe that the incidental rise in rms value can be partially attributed to the high spikes registered by the probe in a high RBC concentrations fluid flow. The bombardment of RBC on the probe thermal boundary layer may cause a characteristic change in the probe response to certain flow phenomenon, at least within the Reynolds number range used in this study.(ABSTRACT TRUNCATED AT 250 WORDS)

A model investigation of the velocity and pressure spectra in vascular murmurs

A physical model consisting of an axisymmetrical jet in a rigid plexiglass pipe was used to study the flow and pressure fluctuations downstream from an aortic stenosis. The fluctuating velocity components, u and v, at several locations in the steady liquid jet were measured using a laser Doppler anemometer system. Simultaneous wall pressure fluctuations were monitored by an array of nine miniature pressure transducers wall mounted in the axial direction. This paper presents the detailed measurements of mean velocity profiles, turbulent intensity distributions and RMS pressure fluctuations. The energy spectra obtained for the pressure fluctuations and the u and v velocity components are compared. Contrary to earlier works, we found that the differences between peak frequencies of the pressure spectra and the characteristic frequencies of the velocity spectra vary with positions downstream from the nozzle. These differences are discussed in light of pseudosound generation by the eddy structures in the stenotic flow field.