Biomedical Engineering Reference
In-Depth Information
FIGURE 8-54
Structure of an
axial-flow pump.
[Adapted from
(Mitamura,
Nakamura et al.,
2001).]
FIGURE 8-55
Turbulent kinetic
energy in an axial
pump. (a) Four vane
impeller. (b) Six vane
impeller. (Mitamura,
Nakamura et al.,
2001), copyright
Informa Healthcare,
reproduced with
permission.
kinetic energy along streamlines through the pumps for a 100 mmHg head and a flow
rate of 5 L/min. Analysis was conducted for pumps with four and six vanes, as shown in
Figure 8-54.
From the turbulent kinetic energy results shown in Figure 8-55, the Reynolds's shear
stress was then calculated, which allowed hemolysis to be estimated. To achieve this, a
similar analysis to that discussed in the section on fluid dynamics of pulsatile devices was
performed.
The predicted hemolysis based on the integrated shear stress along 30 stream lines
flowing through each of the pumps is compared with measurements made using animal
blood. A correlation coefficient of 0.87 was obtained, indicating that the modeling process
is reasonably accurate.
Because the hemolysis level is a function of both the Reynolds's shear stress and the
exposure time of the red blood cell membranes to these forces, it is possible to achieve
the required low level even from a fast-operating axial pump if the exposure time is short
enough.
Care should be taken when using Wurtzinger's formula, equation (8.5), because it has
been found to be incorrect for high shear but short episodes (
5 ms) or in
low shear but long episodes (
τ<
5Pa
t
>
2 s). In these cases, the hemoglobin release is
smaller than predicted by two or three orders of magnitude.
It is clear from this brief introduction that off-the-shelf designs are not suitable for use
as blood pumps and that comprehensive analysis using CFD is essential if a biocompatible
device is required.
τ>
100 Pa
t
<
