Professional Documents
Culture Documents
*DHI-NTU Water & Environment Centre and Education Hub, Nanyang Technological University; †School of Mechanical
and Aerospace Engineering, Nanyang Technological University; and ‡Maritime Research Centre, Nanyang Technological
University, Singapore
Abstract: A magnetically suspended axial flow blood transparent model, the blood analog fluid with the refrac-
pump is studied experimentally in this article. The pump tive index close to that of acrylic was used to avoid
casing enclosed a three-blade straightener, a two-blade refraction. According to the CFD results, the axial flow
impeller shrouded by a permanent magnet-embedded cyl- blood pump could generate adequate pressure head at the
inder, and a three-blade diffuser. The internal flow fields rotating speed of 9500 rpm and flow rate of 5 L/min, and
were simulated earlier using computational fluid dyna- the same flow condition was applied during the PIV
mics (CFD), and the pump characteristic curves were measurement. Through the comparisons, it was found that
determined.The simulation results showed that the internal the experimental results were close to those obtained by
flow field was basically streamlined, except the diffuser CFD and had thus validated the CFD model, which could
region. Particle image velocimetry (PIV) measurement of complement the limitation of the measurement in assessing
the 1:1 pump model was conducted to validate the CFD the more detailed flow fields of the axial flow pump. Key
result. In order to ensure the optical access, an acrylic pro- Words: Computational fluid dynamics—Axial flow blood
totype was fabricated with the impeller driven by a servo- pump—Particle image velocimetry.
motor instead, as the magnet is opaque. In addition to the
Each year, approximately 30 000 patients at risk of 6000~12 000 rpm (8–11). The hemolysis induced by a
dying from severe heart failure are awaiting heart rotary blood pump is the main concern besides the
transplant, but only around 3500 cases are performed hydraulic performance (12). It is found that the
(1) mainly constrained by the limited heart donors hemolysis is significantly related to the flow behavior,
available. Ventricular assist devices (VADs) have and the scalar shear stress and exposure time are two
been developed for decades as an alternative main parameters used to estimate the hemolysis (13).
therapeutic modality and can be produced in the Therefore, the rotary blood pump poses a challenge,
quantity needed. VADs are used for bridge-to- especially for axial flow blood pump with higher
transplantation, bridge-to-recovery, or even long- rotating speed.
term cardiac support (2,3). The continuous-flow Computational fluid dynamics (CFD) has been
VADs usually include a rotating impeller to generate widely used as a powerful tool to investigate many
adequate flow rate and pressure head, and so are also blood-contacting medical devices including blood
named as rotary blood pumps. Typically, the rotating pumps (14,15). With the aid of CFD, the design phase
speed of a centrifugal blood pump is 2000~4000 can be shortened significantly from several years to
rpm (4–7), and that of an axial blood pump is several months (16). It can also support and explain
experimental results. Compared with the traditional
design, CFD can predict the detailed flow fields,
doi:10.1111/j.1525-1594.2011.01362.x pump characteristics curves, and other parameters
Received November 2010; revised July 2011. such as hemolysis (12,17). However, the numerical
Address correspondence and reprint requests to Mr. Boyang Su, results still need validating using the nonintrusive
School of Mechanical and Aerospace Engineering, Nanyang Tech-
nological University, Nanyang Drive, Singapore 639798, Singapore.
measurements, although they are usually difficult and
E-mail: bysu@ntu.edu.sg limited by the optical access. For the whole flow field,
aor_1362 359..367
360 B. SU ET AL.
rotor–stator interaction with 300 time steps in one The test model is partially fabricated from acrylic for
period (33). Velocity inlet and pressure outlet were the illumination of the laser sheet and image pairs
applied at the pump inlet and outlet, respectively. The capturing as demonstrated in Fig. 4b. In order to
static pressure head across the numerical model was minimize the vibration of the operating impeller, a
used as the macro parameter to monitor the conver- bearing holder, which is a hollow step cylinder, is
gence, and the iteration was stopped when the con- designed as shown in Fig. 4c. Its front end is designed
vergence criterion of 10-4 was satisfied (34). to place the hollow diffuser in position with the diam-
eter of the diffuser hub equal to that of the bearing
Experimental method holder at Section B. Note that the influences of the
Figure 4a shows the schematic drawing of the con- vibrations of the inner components such as impeller
structed experimental test rig. The pump model con- and diffuser are not considered in this study. The
structed is the exact size of the prototype in order to straightener, impeller, and diffuser are all made of
avoid any scaling problem that alters the results. acrylic as well as the impeller shroud that is glued
When the pump operates, the flow rate is controlled together with the impeller. The impeller is driven by
by the throttle valve located downstream of the flow- a steel shaft, which is inserted through the bearings
meter, and the pressure head across the pump model and bearing holder. These components are enclosed
and the flow rate is obtained from the pressure trans- in the chamber, which is capped by two components
ducer and the flowmeter, respectively. connected to the test rig.
In order to reduce the possibility of hemolysis, The PIV system used in the present study was the
magnetic suspension technique is widely adopted in Dantec FlowMap PIV system (Dantec Dynamics
rotary blood pumps, including the present study. A/S, Skovlunde, Denmark). The Gemini PIV 200-15
However, it is impossible to perform PIV measure- (New Wave Research, Fremont, CA, USA), a double-
ments due to the opaque magnets. Therefore, the cavity pulsed neodymium yttrium aluminum garnet
impeller is driven by a servomotor through a shaft. (Nd:YAG) laser system, was used as the illumination
(a)
(b)
(c)
source. The two pulsing beams from a double-cavity volume (35). Its refractive index was 1.488 to match
Q-switched Nd:YAG laser are combined and formed with that of the acrylic model with a density of
into pulsing light sheets using optics. Q-switched 1155 kg/m3 and a dynamic viscosity of 3.8 mPa·s. The
pulse can effectively freeze the motion of particle, kinetic viscosities of the working fluid and blood
and two cavities ensure the widest possible ranges of were 3.29 and 3.33 mm2/s, respectively, so as to main-
time between pulses. The CCD camera used in tain Reynolds number similarity. The solution was
the FlowMap PIV system is HiSense MkII seeded with neutral buoyant hollow glass spheres
camera (Hamamatsu C8484-05 digital CCD chip with the mean diameter of 10 mm. In the present PIV
(Hamamatsu City, Shizuoka, Japan); resolution measurement, 100 image pairs were captured at each
1280 ¥ 1024) equipped with Nikon AF Micro-Nikkor measured area. The measurement area was 10.76 ¥
60 mm f/2.8D (Tokyo, Japan). FlowManager software 8.26 mm2 with the scale factor, S, of 1.25. The record-
(Dantec Dynamics A/S) controlled the synchroniza- ing aperture was f/5.6~8, and thus, the camera image
tion sequence to acquire data, and stored the record- of the seeding particle had a diameter 2~3 pixels
ing data. The “once-per-revolution” signal from the under the wavelength of 532 nm. The interrogation
shaft encoder was sent to the processor, which then area of 32 ¥ 32 pixels was used with 50% overlap. The
fires the laser and acquires the images from the PIV pulse duration of this PIV is 10 ns, so the particle
camera as shown in Fig. 3a, so images could be taken movement during this time duration was negligible.
at a certain angular position of the impeller. More- In the straightener, the pulse separation varied
over, the timing signals are used to control the 10~25 ms at different measurement areas according to
camera to make sure that the CCD sensor is only the local flow velocity. In order to minimize the
exposed to the light from the illumination system, as effects of the out-of-plane velocity component, the
the ambient light will reduce the signal : noise ratio. thickness of the laser sheet was 0.5 mm, and the time
In order to measure the flow field at a different between light pulses was 10 ms so as to satisfy the
angular position, of the impeller a delay was equation below (36):
set between the external signal and the image
acquisition. Two successive images are then trans- vz ⋅ Δt
≤ 0.25
ferred to the FlowMap processor and stored finally. Zl
As a result, a pair of seeding particle images with
time delay was captured at a certain impeller angular where vz is the out-of-plane velocity, Dt is the time
position, and the velocity vector map was derived between pulses and Zl is the thickness of the laser
from the movement of particles during the time sheet. Postprocessing validation methods including
delay. peak height validation and velocity range validation
In order to avoid refraction, the working fluid were applied on the raw vector. Finally, the mean
involved mixing 40% of glycerin and 60% of aqueous phase-averaged velocity profiles over 100 recordings
ammonium thiocyanate (NH4SCN) solution by were obtained in each region.
(a) Straightener
(b) Impeller
(c) Diffuser
at the regions close to blades, hub, and inner casing laser thickness and the time between pulses as men-
surfaces due to a number of reasons. First, the tioned earlier in the experimental method section.
trapped seeding particles on surfaces could induce The L2 relative error norm of the differences in axial
relatively strong reflection, resulting in error in PIV velocity between CFD and PIV with respect to the
measurements. Second, the fluid close to the inter- PIV in the straightener, impeller, and diffuser is 11,
faces usually has higher velocity gradient in radial 19, and 27%, respectively. Note that the L2 relative
direction, and it could cause error due to the difficulty error norm is defined as
in PIV measurements. In order to calculate the esti-
N
mated velocity accurately, the seeding particles
within an interrogation area, where the camera image ∑ ( fi E − fi N )2
i
is usually divided into 32 pixels by 32 pixels in size, N
33. Chua LP, Song GL, Zhou TM. Numerical analysis of the inner 35. Doyle MG, Vergniaud JB, Tavoularis S, Bourgault Y.
flow field of a biocentrifugal blood pump. Artif Organs Numerical simulations of blood flow in artificial and natural
2006;30:467–77. hearts with fluid–structure interaction. Artif Organs 2008;
34. Throckmorton AL, Untaroiu A, Allaire PE, et al. Computa- 32:870–9.
tional analysis of an axial flow pediatric ventricular assist 36. Dantec-Dynamics. Flowmanager software and introduction to
device. Artif Organs 2004;28:881–91. PIV instrumentation (software user’s guide). 2002.