CFD Simulation of Blood Flow Through Artificial Heart Valves
CFD Simulation of Blood Flow Through Artificial Heart Valves
M. Bajgrowicz 1
Summary: Artificial valve replacement is a common procedure in case of heart valve disorders treatment. The main task of this prosthesis is to generate blood flow similar to flow
generated by natural, human valve. Velocity distribution depends mainly on the geometry
of the valve. That is why it should be analysed and modified in case to obtain the best results. The objective of this study was to numerically investigate how geometry of valve (the
maximal opening angle) influences velocity distribution behind a valve in terms of potential
negative effect on blood components. For this purpose seven 2D models of artificial heart
valves (AHV) with different maximal opening angle were created in Gambit. The blood flow
was simulated in Ansys Fluent. This paper demonstrates the effectiveness of CFD method
in artificial heart valves design process. Using Ansys Fluent, it is possible to simulate flow
thorough AHV and analyze blood distribution in a region close to the valve, where physical
measurements are very difficult to conduct.
Keywords: CFD, artificial heart valve, blood flow simulation
1 Introduction
The most common valve defects are: valve stenosis and mitral regurgitation. In case of the above-mentioned
disorders, the most common procedure is an artificial valve replacement. Thanks to the development of
biomedical engineering for 50 years, there is the possibility to perform such operation [1, 2, 3]. Although
many companies are working on improving geometry of AHV, there is still no perfect prosthesis which could
replace the natural valve. The main problem faced by developers is to ensure adequate blood flow, as close
as possible to the natural. Researches show that AHV produce higher velocity gradients than natural heart
valves. That, create undesirable shear stress acting on the surfaces of blood cells. This can lead to changes in
the membranes of these cells. Changes in membrane lipid fluidity of red cells may decrease their deformation
abilities, and increase the platelet aggregation, what consequently influence the rheological properties of the
blood and increase probability of thrombus formation [4]. That is why it is necessary to analyze how velocity
is distributed in area close to the valve, and to modify it by improving geometry of valve, in case to achieve
the best velocity distribution [5, 7, 7, 9]. Formed flow should be uniformly distributed, to prevent creating
zones in which the liquid is stationary and zones where it moves at a high speed. Both of these zones are
harmful to our circulatory system. Congestive zone conduct to platelet aggregation and thrombus formation,
while in the zone of strong turbulence, deformation and destruction of blood cells is more likely to appear
[10].
2 Material and methods
2.1 Geometric model - Gambit
Two different kinds of artificial heart valves were designed: tilting disc valve and bileaflet valve. Main valves
dimensions: external suture ring diameter: 28 mm, flange diameter: 19 mm, disc thickness: 0.5 mm. Cross
sections and all dimensions are shown in figure 1.
1
Dipl. Ing. Magdalena Bajgrowicz, Institute of Material Science, Wroclaw University of Technology, wybrzeze Stanislawa
Wyspianskiego 27, 50-370 Wroclaw, email: [email protected]
XIIIth Youth Symposium on Experimental Solid Mechanics, June 29th July 2nd , Decn, Czech Republic
Figure 1: Cross sections and dimensions [mm] of designed valves. A - titling disc valve, B - bileaflet valve.
Four 2-D models of titling disc valve, and 3 models of bileaflet valve with different maximal opening
angles were created in Gambit software. Disc positions for tilting disc valve were: = 60 , 70 , 80 , 90 .
For bileaflet valve: = 70 , 80 , 90 ,
2.2 Simulation - Fluent
Simulations of blood flow through prepared models were proceed by using Ansys Fluent software. It was
steady, turbulent and Newtonian flow. K-epsilon model was used. For this study, the elasticity of vessel walls
was neglected, valve and aortic walls were assumed rigid. Models represent situation when disks are fully
open.
FLOW PARAMETERS Simulations were performed for the intensive flow of blood, when volumetric
flow rate Q = 25 l/min, that corresponds to the peak physiological flow through the human aorta. Velocity
of blood flow in simulations was 1.6 m/s. The Reynolds number was 12057. Calculations are performed for
the continuous flow with constant velocity at the inlet, and for a constant position of the valve disk.
BLOOD MODEL Blood was modeled as isotropic homogeneous, Newtonian liquid with the following
parameters: density = 1055 kg/m3 , dynamic viscosity: = 0.0035 kg/m s.
3 Results
The analyzed parameter was velocity distribution. Figure 2 shows velocity fields generated by seven different
valves simulated in Ansys Fluent.
4 Discussion
Velocity profile generated by tilting disc valve is unsymmetrical. The major steam is skewed towards one side
of the aorta in the direction in which disc opens. Width and speed of this stream is directly connected with
opening angle of the valve. The bigger opening angle, the wider is the stream and the lower is the speed. The
fact that the whole flow and pressure is focused in one side of artery may have negative influence on aortas
walls. It may lead to unnecessary aorta fatigue. Another characteristic feature of tilting valves connected
with unsymmetrical flow is fact that two main zones are created. First zone covers area of main stream, it
is a zone where blood flows with high speed. The second one is a zone where blood flows slower. It can
be seen very clearly, especially when maximal opening angle is small. In case of valve with opening angle
60 , speed amplitude between zones is very high, which is something that must be avoided. By increasing
the opening angle, velocity amplitude between both zones can be decreased and that improve the flow of the
blood.
Bileaflet valves, unlike tilting disc valves generate symmetrical blood flow. Two main streams on both
sides of aorta, and one smaller in a centre of artery can be distinguished. It can be assumed that this flow
8
XIIIth Youth Symposium on Experimental Solid Mechanics, June 29th July 2nd , Decn, Czech Republic
[4] I. Zakowska,
Ocena zmian wasciwosci bon plazmatycznych krwinek czerwonych u pacjentow ze sztuczna zastawka serca.
9
XIIIth Youth Symposium on Experimental Solid Mechanics, June 29th July 2nd , Decn, Czech Republic
[5] S. Petkova, A. Hossain, J. Naser, E. Palombo, CFD modeling of blood flow in portal vein hypertension
with and without thrombosis, 2003
[6] D. Hazer, R. Unterhinninghofen, M. Kostrzewa, H. U. Kauczor, R. Dillmann, G. M. Richter A workflow
for Computational Fluid Dynamics Simulations using Patient Specific Aortic Model
[7] S. Aluri, K. B. Chandran, Numerical simulation of mechanical heart valve closure, Uniwersity of Iowa
[8] A. Torcu, A. Lungu, S. Vlad, Closure Mechanics of Mechanical Heart Valve Prostheses Experimental
and Computational Analysis, Technical University of Cluj-Napoca, Romania
[9] C. C. Boter, T. Vasile, S. Sfrangeou, CFD Simulation of the Portal Vein Blood Flow, Technical University of Cluj-Napaca, Romania
[10] M. S. Sacks, W. D. Merryman, D. E. Schmidt, On the biomechanics of heart valve function, Journal of
biomechanics 42 (2009), 1804-1824
10