Computational Fluid Dynamics Analysis to Establish the Design Process of a Centrifugal Blood Pump: Second Report

To establish an efficient design process for centrifugal blood pumps, the results of computational fluid dynamics (CFD) analysis were compared to the results of flow visualization tests and hemolysis tests, using the Nikkiso centrifugal blood pump. CFD analysis revealed that the radial gap greatly affected the shear stress in the outlet diffuser. The hemolysis study also indicated a similar tendency. To see the flow behind the impeller, we conducted a comparative study between models with and without washout holes using the CFD technique. CFD analysis indicated that flow and pressure distributions behind the impeller were different between both models, and a particle was observed to remain longer behind the impeller in the model without washout holes. In the future, CFD analysis could be a useful tool for developing blood pumps in comparison to flow visualization tests and hemolysis tests.     

Development of Design Methods of a Centrifugal Blood Pump with In Vitro Tests, Flow Visualization, and Computational Fluid Dynamics: Results inHemolysis Tests

Abstract There are few established engineering guidelines aimed at reducing hemolysis for the design of centrifugal blood pumps. In this study, a fluid dynamic approach was applied to investigate hemolysis in centrifugal pumps. Three different strategies were integrated to examine the relationship between hemolysis and flow patterns. Hemolytic performances were evaluated in in vitro tests and compared with the flow patterns analyzed by flow visualization and computational fluid dynamic (CFD). Then our group tried to establish engineering guidelines to reduce hemolysis in the development of centrifugal blood pumps. The commercially available Nikkiso centrifugal blood pump (HPM-15) was used as a standard, and the dimensions of 2 types of gaps between the impeller and the casing, the axial and the radial gap, were varied. Four impellers with different vane outlet angles were also prepared and tested. Representative results of the hemolysis tests were as follows: The axial gaps of 0.5, 1.0, and 1.5 mm resulted in normalized index of hemolysis (NIH) values of 0.0028, 0.0013 and 0.0008 g/100 L, respectively. The radial gaps of 0.5 and 1.5 mm resulted in NIH values of 0.0012 and 0.0008 g/100 L, respectively. The backward type vane and the standard one resulted in NIH values of 0.0013 and 0.0002 g/100 L, respectively. These results revealed that small gaps led to more hemolysis and that the backward type vane caused more hemolysis. Therefore, the design parameters of centrifugal blood pumps could affect their hemolytic performances. In flow visualization tests, vortices around the impeller outer tip and tongue region were observed, and their patterns varied with the dimensions of the gaps. CFD analysis also predicted high shear stress consistent with the results of the hemolysis tests. Further investigation of the regional flow patterns is needed to discuss the cause of the hemolysis in centrifugal blood pumps.     

Development of Design Methods for a Centrifugal Blood Pump with a Fluid Dynamic Approach: Results in Hemolysis Tests

The purpose of this study was to examine the relationship between local flow conditions and the hemolysis level by integrating hemolysis tests, flow visualization, and computational fluid dynamics to establish practical design criteria for centrifugal blood pumps with lower levels of hemolysis. The Nikkiso centrifugal blood pump was used as a standard model, and pumps with different values of 3 geometrical parameters were tested. The studied parameters were the radial gap between the outer edge of the impeller vane and the casing wall, the position of the outlet port, and the discharge angle of the impeller vane. The effect of a narrow radial gap on hemolysis was consistent with no evidence that the outlet port position or the vane discharge angle affected blood trauma in so far as the Nikkiso centrifugal blood pump was concerned. The radial gap should be considered as a design parameter of a centrifugal blood pump to reduce blood trauma.     

Computational modeling of the Food and Drug Administration’s benchmark centrifugal blood pump

In order to simulate hemodynamics within centrifugal blood pumps and to predict pump hemolysis, CFD simulations must be thoroughly validated against experimental data. They must also account for and accurately model the specific working fluid in the pump, whether that is a blood-analog solution to match an experimental PIV study or animal blood in a hemolysis experiment. Therefore, the Food and Drug Administration (FDA) benchmark centrifugal blood pump and its database of experimental PIV and hemolysis data were used to thoroughly validate CFD simulations of the same blood pump. A Newtonian blood model was first used to compare to the PIV data with a blood analog fluid while hemolysis data were compared using a power-law hemolysis model fit to porcine blood data. A viscoelastic blood model was then incorporated into the CFD solver to investigate the importance of modeling blood’s viscoelasticity in centrifugal pumps. The established computational framework, including a dynamic rotating mesh, animal blood-specific fluid properties and hemolysis modeling, and a k-ω SST turbulence model, was shown to more accurately predict pump pressure heads, velocity fields, and hemolysis compared to previously published CFD studies of the FDA centrifugal pump. The CFD simulations were able to match the FDA pressure and hemolysis data for multiple pump operating conditions, with the CFD results being within the standard deviations of the experimental results. While CFD radial velocity profiles between the impeller blades also compared well to the PIV velocity results, more work is still needed to address the large variability among both experimental and computational predictions of velocity in the diffuser outlet jet. Small differences were observed between the Newtonian and viscoelastic blood models in pressure head and hemolysis at the higher flow rate cases (FDA Conditions 4 and 5) but were more significant at lower flow rate and pump impeller speeds (FDA Condition 1). These results suggest that the importance of accounting for blood’s viscoelasticity may be dependent on the specific blood pump operating conditions. This detailed computational framework with improved modeling techniques and an extensive validation procedure will be used in future CFD studies of centrifugal blood pumps to aid in device design and predictions of their biological responses.     

Prediction of leakage flow in a shrouded centrifugal blood pump

This article proposes a phenomenological model to predict the leakage flow in the clearance gap of shrouded centrifugal blood pumps. A good washout in the gap clearance between the rotating impeller surfaces and volute casing is essential to avoid thrombosis. However, excessive leakage flow will result in higher fluid shear stress that may lead to hemolysis. Computational fluid dynamics (CFD) analysis was performed to investigate the leakage flow in a miniaturized shrouded centrifugal blood pump operating at a speed of 2000 rpm. Based on an analytical model derived earlier, a phenomenological model is proposed to predict the leakage flow. The leakage flow rate is found to be proportional to h(α) , where h is the gap size and the exponent α ranges from 2.955 to 3.15 for corresponding gap sizes of 0.2-0.5 mm. In addition, it is observed that α is a linear function of the gap size h. The exponent α compensates for the variation of pressure difference along the circumferential direction as well as inertia effects that are dominant for larger gap clearances. The proposed model displays good agreement with computational results. The CFD analysis also showed that for larger gap sizes, the total leakage flow rate is of the same order of magnitude as the operating flow rate, thus suggesting low volumetric efficiency.     

Computational Prediction of Hemolysis in a Centrifugal Ventricular Assist Device

Abstract: This paper describes the use of computational fluid dynamics (CFD) to predict numerically the hemolysis in centrifugal pumps. A numerical hydrodynamical model, based on the full Navier-Stokes equation, was used to obtain the flow in a vaneless centrifugal pump (of corotating disks type). After proper postprocessing, critical zones in the channel were identified by means of two-dimensional color-coded maps of %Hb release. Simulation of different conditions revealed that flow behavior at the entrance region of the channel is the main cause of blood trauma in such devices. A useful feature resulting from the CFD simulation is the visualization of critical flow zones that are impossible to determine experimentally with in vitro hemolysis tests.     

Hemolytic evaluation using polyurethane microcapsule suspensions in circulatory support devices: normalized index of hemolysis comparisons of commercial centrifugal blood pumps

Abstract:  We have been developing some types of microcapsule suspensions with polyurethane membranes to evaluate the absolute hemolytic characteristics of the centrifugal blood pumps used in circulatory support devices such as artificial hearts. In order to facilitate/realize hemolysis testing on centrifugal blood pumps that have hemolysis levels as low as those of commercial centrifugal blood pumps, we eliminated capsules with diameters less than 72.2 µm, amounting to 15.4% of all capsules in the conventional suspension (crude suspension [CS]), and adjusted the capsule volume ratio to correspond to a hematocrit of 40%. In this way we succeeded in enhancing the sensitivity of the suspension to microcapsule destruction 61 fold. We used this new suspension (fine suspension [FS]) to perform hemolysis tests on four types of commercial pump with mock circulation systems. Under conditions of 500 mm Hg and 11.2 L/min, we successfully determined the hemolytic characteristics (normalized index of hemolysis [NIH]) of some of the centrifugal blood pumps; the results showed some correlation with those of hemolysis tests on bovine blood and suggest that microcapsule suspensions with polyurethane membranes are useful as standard test solutions for the absolute evaluation of centrifugal blood pumps.     

Parametric Study of Blade Tip Clearance, Flow Rate, and Impeller Speed on Blood Damage in Rotary Blood Pump

Phenomenological studies on mechanical hemolysis in rotary blood pumps have provided empirical relationships that predict hemoglobin release as an exponential function of shear rate and time. However, these relations are not universally valid in all flow circumstances, particularly in small gap clearances. The experiments in this study were conducted at multiple operating points based on flow rate, impeller speed, and tip gap clearance. Fresh bovine red blood cells were resuspended in phosphate-buffered saline at about 30% hematocrit, and circulated for 30 min in a centrifugal blood pump with a variable tip gap, designed specifically for these studies. Blood damage indices were found to increase with increased impeller speed or decreased flow rate. The hemolysis index for 50-µm tip gap was found to be less than 200-µm gap, despite increased shear rate. This is explained by a cell screening effect that prevents cells from entering the smaller gap. It is suggested that these parameters should be reflected in the hemolysis model not only for the design, but for the practical use of rotary blood pumps, and that further investigation is needed to explore other possible factors contributing to hemolysis.     

The Effect of the Impeller-Driver Magnetic Coupling Distance on Hemolysis in a Compact Centrifugal Pump

Abstract: Blood trauma is one of the important performance parameters of centrifugal pumps. To investigate the blood trauma induced by these pumps, in vitro hemolysis tests have become an important procedure and are increasingly used for pump development and comparisons. The Baylor compact eccentric inlet port (C1E) centrifugal blood pump was developed as a long-term centrifugal ventricular assist device (VAD) as well as a cardiopulmonary bypass pump (CPB). The Baylor C1E pump incorporates a seal-less design with a blood stagnation-free structure. This pump can provide flows of 5 L/min against 350 mm Hg of total pressure head at 2,600 revolutions per minute. The pump impeller is magnetically coupled to the driver magnet in a seal-less manner. The latest hemolysis study revealed that hemolysis may be affected by the gap distance between the driver and the impeller magnet. The purpose of this study was to verify the effect of the magnetic coupling distance on the normalized index of hemolysis (NIH) with the C1E model and to obtain an optimal gap distance. The NIH value was clearly decreased by alteration of the magnetic coupling distance from 7.7 to 9.7 mm in CPB and left ventricular assist device (LVAD) conditions. The NIH, when using the pump as an LVAD condition, was reduced to a level of 0.0056 from 0.095 when the magnetic coupling distance was extended. The same results were also obtained when the pumps were used in a CPB condition. The magnetic coupling distance is an important factor for the C1E model in terms of hemolysis. Different coupling forces effect the bearings and impeller stability. These results suggest that an optimal driving condition with a proper magnetic coupling and an optimal force between the impeller and driver is necessary to develop an atraumatic centrifugal pump.     

Flow visualization analysis for evaluation of shear and recirculation in a new closed-type, monopivot centrifugal blood pump

Flow visualization has great potential in analyzing flow patterns of centrifugal blood pumps to locate possible hemolysis and thrombus formation sites. This study focused on the said phenomena thought to correlate with areas of high shear velocity and stagnation and analyzed a new closed-type centrifugal blood pump. As a result of analyzing the flow of inlet and front gap of the impeller, flow in the front gap was approximately 30% of the external flow. Visualization in the back gap showed sufficient exchange also. Analysis in the volute area and around the washout holes revealed high shear locations and quantified the highest shear velocity. Maximum shear on the volute wall was found to be 9,000-19,000 s-1 and was located in the 0.2-mm vicinity of the wall. Based on these results, previous hemolysis tests, and small pump size, one concludes that the analyzed closed-type centrifugal pump has a relatively smooth flow suitable for a totally implantable artificial heart.     

An estimation method of hemolysis within an axial flow blood pump by computational fluid dynamics analysis

Evaluation of hemolysis within a blood pump on a computer is useful for developing rotary blood pumps. The flow fields in the axial flow blood pump were analyzed using computational fluid dynamics (CFD). A blood damage index was calculated based on the changes in shear stress with time along 937 streamlines. Hemolysis of the pumps was measured using bovine blood. A good correlation between the computed and measured hemolysis results was observed. CFD analysis is useful for estimating hemolysis of rotary blood pumps on a computer.     

A review of leakage flow in centrifugal blood pumps

This article presents a new approach in determining the functional relationship between the leakage flow in a centrifugal blood pump and various parameters that affect it. While high leakage flow in a blood pump is essential for good washout and can help prevent thrombus formation, excessive leakage flow will result in higher fluid shear stress that may lead to hemolysis. Dimensional analysis is employed to provide a functional relationship between leakage flow rate and other important parameters governing the operation of a centrifugal blood pump. Results showed that pump performance with a smaller gap clearance is clearly superior compared to those of two other similar pumps with larger gap clearances. It was also observed that the nondimensional leakage flow rate varies almost linearly with dimensionless pump head. It also decreases with increasing volume flow rate. A smaller gap clearance will also increase the flow resistance and hence, decrease the nondimensional leakage flow rate. Increasing surface roughness, length of the gap clearance passage, or loss coefficient of the gap geometry will increase losses and hence, decrease the leakage flow rate. It is also interesting to note that for a given pump and gap clearance geometry, the nondimensional leakage flow rate is almost independent of the Reynolds number when specific speed is constant.     

Development of the Small Caliber Centrifugal Blood Pump

Regarding the development of a centrifugal blood pump to be connected directly with small diameter tubings for pediatric use while minimizing hemolysis, we have studied the inlet port side configurations of a pump using both a hemolysis test and computational fluid dynamics (CFD) analysis. We have conducted a hemolysis test on 2 models. The tapered shape inlet has proven to be lower in the index of hemolysis (IH) than the straight shape. CFD analyses utilizing expanded flow paths indicated that the flow velocity decreased as the fluid path became larger within the tapered nozzle. When entering the pump chamber, the flow rushed in at a greater velocity through the straight nozzle due to its small diameter. The straight shape showed an abrupt change in pressure around the entrance of the pump chamber while the tapered shape did not. The flow inlet angle of the straight model was observed to be larger than that of the tapered model because of its smaller turning radius.     

Effect of Surface Roughness on Hemolysis in a Pivot Bearing Supported Gyro Centrifugal Pump (C1E3)

Abstract: The blood contacting surface quality is an important pump parameter for blood compatibility and cell damage. This study investigates the surface roughness and the effect it has on hemolysis in a centrifugal blood pump. In vitro hemolysis tests were performed with a pivot bearing supported Gyro centrifugal pump (C1E3) simulating cardiopulmonary bypass (CPB; 5 L/min, 350 mm Hg) and left ventricular assist device (LVAD; 5 L/min, 100 mm Hg) conditions. To produce 4 different grades of surface roughness, the impellers and housings were subjected to vapor polishing, sand papering, fine sand blasting, or coarse sand blasting. Seven pumps were assembled with different impeller and housing surfaces. These surfaces were then examined by a surface profile instrument and a scanning electron microscope. The results of this study are as follows. First, the effect of surface roughness on hemolysis was significantly greater in the CPB condition than in the LVAD condition. Second, surface roughness, regardless of whether it is the impeller or pump housing, had little effect on hemolysis in the LVAD condition. Third, in the CPB condition, the surface roughness of the pump housing has a greater effect on hemolysis than does that of the impeller. From a hemolytic point of view, an extremely smooth pump housing is required for use of an impeller type centrifugal pump as a CPB device. In contrast, it is conceivable that a smooth surface is not always essential for an impeller type centrifugal pump that is used as an LVAD.     

Study of Secondary Flow in Centrifugal Blood Pumps Using a Flow Visualization Method with a High-Speed Video Camera

Abstract Four pump models with different vane configurations were evaluated with flow visualization techniques using a high-speed video camera. These models also were evaluated through in vivo hemolysis tests using bovine blood. The impeller having the greatest fluid velocity relative to the impeller, the largest velocity variance, and the most irregular local flow patterns in the flow passage caused the most hemolysis. Even if the pumps were operated at almost the same speed (rpm) at the same output, the impeller showing more irregular flow patterns had a statistically greater rate of hemolysis. This fact confirms that the existence of local irregular flow patterns in a centrifugal blood pump deteriorates its hemolytic performance. Thus, to optimize the design of the pump, it is very important to examine the secondary flow patterns in the centrifugal blood pump in detail using flow visualization with a high-speed video camera.     

Flow Visualization as a Complementary Tool to Hemolysis Testing in the Development of Centrifugal Blood Pumps

Abstract With a 250% scaled-up pump model, high speed video camera, and argon ion laser light sheet, flow patterns related to hemolysis were visualized and analyzed with 4 frame particle tracking software. Different flow patterns and shear distributions were clarified by flow visualization for pumps modified to have different hemolysis levels. A combination of in vitro hemolysis tests, flow visualization, and CFD analysis suggested a close relationship between hemolysis and high shear caused by small impeller/casing gaps. Because arbitrary cross sections can be illuminated by laser light sheet, flow visualization is a useful tool in finding locations related to hemolysis in the design process of rotary blood pumps.     

Hemolytic Effect of Surface Roughness of an Impeller in a Centrifugal Blood Pump

Abstract: The present study investigates how the surface roughness of an impeller affects hemolysis in the pivot bearing supported Gyro C1E3 pump. This study focuses on particular areas of the impeller surface in the impeller type centrifugal pump. Seven Gyro C1E3 pumps were prepared with smooth surface housings and different impeller parts with different surface roughnesses. The vanes, top side, and backside of the impeller were independently subjected to vapor polishing, fine sand blasting, or coarse sand blasting to produce three different grades of surface roughness. These surfaces were then examined by a surface profile instrument. Using these pumps with different impellers, in vitro hemolysis tests were performed simulating cardiopulmonary bypass (5 Wmin, 350 mm Hg). The findings of this study conclusively proved that surface roughness of the back side of the impeller has the greatest effect on hemolysis, followed by the top side and then the vanes. The following are reasons for these findings. First, the shear rate may be greater on the back side than on the top side because of the smaller gap between the back and the housing and the greater relative speed against the impeller. Second, the fluid beneath the impeller may have a longer exposure time because there is little chance for the fluid to mix beneath the impeller. Third, the shear rate may be greater on the top side of the impeller than on the vanes because a vortex formation occurs behind the vanes.     

Measurement of the rotor motion and corresponding hemolysis of a centrifugal blood pump with a magnetic and hydrodynamic hybrid bearing

This study proposed a centrifugal blood pump with a novel magnetic and hydrodynamic hybrid passive bearing, which consisted of a plain journal bearing for radial stability and a permanent magnetic bearing for axial and tilting stability. We measured the radial motion of the bearing and performed hemolysis tests for the different radial clearance sizes. In the results, it appeared that the radial motion had two modes: the stable center mode, in which the radius of the radial motion rapidly converged to less than 20 microm; and the unstable circle mode, in which the rotor suspension linearly increased with the rotation speed. It also appeared that the pumps with the radial clearance of 80 microm caused more hemolysis than with the smaller clearance sizes in the circle mode. The circle mode was avoidable by the higher rotation or the asymmetric pump structure, but the mechanism of hemolysis in this mode was still unclear.     

Improvement of Hemocompatibility in Centrifugal Blood Pump With Hydrodynamic Bearings and Semi-open Impeller: In Vitro Evaluation

We have developed a noncontact-type centrifugal blood pump with hydrodynamic bearings and a semi-open impeller for mechanical circulatory assist. The impeller is levitated by an original spiral-groove thrust bearing and a herringbone-groove journal bearing, without any additional displacement-sensing module or additional complex control circuits. The pump was improved by optimizing the groove direction of the spiral-groove thrust bearing and the pull-up magnetic force between the rotor magnet and the stator coil against the impeller. To evaluate hemocompatibility, we conducted a levitation performance test and in vitro hemocompatibility tests by means of a mock-up circulation loop. In the hemolysis test, the normalized index of hemolysis was reduced from 0.721 to 0.0335 g/100 L corresponding to an expansion of the bearing gap from 1.1 to 56.1 µm. In the in vitro antithrombogenic test, blood pumps with a wide thrust bearing gap were effective in preventing thrombus formation. Through in vitro evaluation tests, we confirmed that hemocompatibility was improved by balancing the hydrodynamic fluid dynamics and magnetic forces.     

Flow Visualization Study to Improve Hemocompatibility of a Centrifugal Blood Pump

A correlation study was conducted among quantitative flow visualization analysis, computational fluid dynamic analysis, and hemolysis tests regarding the flow in a centrifugal blood pump to prevent hemolysis. Particular attention was paid to the effect of the impeller/casing gap widths on the flow in the volute and in the outlet. Flow vector maps were obtained for 250% scaled-up models with various geometries, using an argon ion laser light sheet, a high speed video camera, and particle tracking velocimetry. In terms of the results, in the small radial gap model, high shear occurred near the inside wall of the outlet and stagnation near the outside wall of the outlet whereas the standard model maintained smooth flow and low shear. The small radial gap model showed a lower head and greater hemolysis than the standard model. This head decrease could be partly restored by relocating the outlet position; however, the hemolysis level hardly decreased. From these results, it was found that the small radial gap itself is important. It was also confirmed by detailed flow visualization and simple laminar shear analysis near the wall that the small radial gap caused a wider high shear layer (110-120 microm) than the standard model (approximately 80 microm). In the small radial gap model, the high shear layer in the outlet (approximately 50 microm) is much narrower than that in the volute. Flow visualization together with the aid of computational fluid dynamic analysis would be useful to eliminate the causes of hemolysis.