See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/301354150

Simulations of Artificial Biological Heart Valves with ANSYS

http://volitans.wix.com/essschallenge

Technical Report · April 2016

DOI: 10.13140/RG.2.1.3146.7925

CITATION READS

1 1,368

5 authors, including:

Laura Tameirão Sampaio Rodrigues Lucas Coelho e Silva

Federal University of Minas Gerais Instituto Tecnologico de Aeronautica
4 PUBLICATIONS   3 CITATIONS    2 PUBLICATIONS   1 CITATION   

SEE PROFILE SEE PROFILE

Marcelo Greco
Federal University of Minas Gerais
114 PUBLICATIONS   423 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Dynamic Instability in Shallow Arches under Transversal Forces and Plane Frames with Semirigid Connections View project

Dynamic Instability in Shallow Arches under Transversal Forces and Plane Frames with Semirigid Connections View project

All content following this page was uploaded by Marcelo Greco on 18 April 2016.

The user has requested enhancement of the downloaded file.

 FEDERAL UNIVERSITY OF MINAS GERAIS
ESSS COMPUTATIONAL MODELING CHALLENGE
CONVERGENCE 2016

Volitans Team
Laura Tameirão Sampaio Rodrigues
Lucas Coelho e Silva
Lucas Costa Machado

Simulations of Artificial Biological Heart Valves with ANSYS

Belo Horizonte, Brazil

2016
 2

1. Introduction
In Brazil, the annual heart valve transplant numbers are significant and growing.
30% of cardiac surgical procedures in the country are caused by valvular sequelae,
therefore valve replacement surgeries represent a considerable portion of the Unified
Health System (SUS) budget, due to high costs of the valves.
In these surgeries, the valve that presents dysfunction is replaced by a mechanical
or biological prosthesis. The mechanical valve has the advantage of its high durability.
However, this type of valve alters the system in a way which leads to the increased risk
of thromboembolic complications, ergo, patients with mechanical prosthesis require
lifelong anticoagulation medication. Biological valves, in turn, are made of bovine
pericardium or porcine, properly treated and reconstituted, reducing the risk of immune
rejection. However, the biological valve has increased wear over time.
In order to predict the behavior of biological valves, specialized companies
conduct several laboratorial tests, which are extremely important for valve validation and
insurance of the manufacturing process quality. However, these are high-cost tests.
Aiming reduction of the valve cost and hence the surgical procedure cost,
simulations in ANSYS reproducing tests in the laboratory have been proposed, forming
a more viable alternative to validate the development process of the valves.

2. Objective
The objective of this project is to assess the possibility of implementation of
simulations in ANSYS in the development process of biological artificial heart valves.
With these simulations, reduction of costs and engineering time is achievable, apart from
being possible to use the results to increase the product’s efficiency.

3. Methodology
In order to maintain a high-fidelity mathematical model of reality, a technical visit
to Labcor was carried out to follow tests performed during the manufacturing process of
the valves, allowing for the definition of the most suitable simulations settings. Labcor is
a company which manufactures biological heart valves, settled in Belo Horizonte, Brazil.
The valve geometry was developed based on CAD drawings provided by the
company.
The problem stated is a transient situation, with a high coupling level between the
fluid flow and the structural deformation, thus the numerical problem must also be highly-
coupled. Therefore, a simulation coupling Fluent with Transient Structural has been
proposed.
Given that a fluid-structure interaction problem faces more complicated
convergence and solution instabilities, preliminary simulations were performed in order
 3

to detect possible modeling errors, ensuring the most appropriate model would be used at
the final simulation.
Therefore, the problem was divided on three phases:
 Fluid Dynamics Simulation of the test section
 Static Structural Simulation
 Fluid-Structure Interaction Simulation with a One-Way Coupling

4. Development
4.1 Fluid Dynamics Simulation
Geometry and mesh
A CAD model of the test section was made for the fluid dynamics simulation.

Figure 4.1.1 – CAD model of the test section

For the mesh design, the statistic elements of the mesh such as Element Quality,
Aspect Ratio and Skewness were assessed initially. The mesh was refined based on the
simulation results, checking how the maximum pressure on the leaflets’ walls changed
with mesh quality increase.
 4

Figure 4.1.2 – Fluid domain mesh

Figure 4.1.3 – Fluid domain mesh

Solver Configuration
Fluent was used for the fluid dynamics simulation. The solver was configured as
transient, and the working fluid, viscous and laminar.
 5

Working Fluid
A model for the human blood was used to define the working fluid. The density used was
of 1060 kg/m³, and for viscosity the Carreau model was adopted.

The following coefficients were used:

Table 4.1.1 – Blood viscosity modeling
Coefficient Value
Relaxation time 𝜆 (s) 3.313
Power index 𝑛 0.3568
Viscosity at zero shear rate 𝜇0 (kg/m-s) 0.056
Viscosity at infinite shear rate 𝜇𝑖𝑛𝑓 (kg/m-s) 0.0035

Boundary Conditions
The outlet was set as pressure-outlet, with a static gauge pressure of 14 kPa,
corresponding to the average pressure of the systolic and diastolic phases in a healthy
human. The inlet was defined as a velocity-inlet, and the time-varying periodic velocity
profile was set via UDF. The velocity pulse modeled the human pulsatile blood flow and
was consistent with the pulse applied by the testing device.

Verification
Mass conservation and inlet boundary conditions
In order to inspect the fluid dynamics simulation, the mass conservation and inlet
boundary conditions were checked. The inlet boundary conditions were checked in order
to ensure the UDF behavior was the expected one.

Mesh and time step refinement

Simulations were reconducted with a finer mesh and smaller time step in order to
investigate the differences caused by numerical discretization.

4.2. Static Structural Simulation

In order to visualize the deforming beahavior of the valve in a simpler and quicker
way, a static structural simulation was performed.
The material used was set as Multilinear Isotropic Hardening and its properties
based on data provided by Labcor.
 6

Contact and Boundary Conditions

Frictionless was used as the contact conditions between the leaflets surfaces. Fixed
support was defined on the surfaces that are fixed on the lab testing. A pressure of 14 kPa
was applied on the internal leaflets surfaces.

Figure 4.2.1 – Fixed Support on lateral faces

Figure 4.2.2 – Contact condition: faces of application

 7

Mesh
A mesh independency study was performed, and the parameter used for
comparison was that of maximum displacement.

Figure 4.2.3 – Mesh independency study

Figure 4.2.4 – Structural mesh of the leaflets

 8

4.3 Fluid Structure Interaction with a One-Way Coupling

Based on the fluid dynamics and static structural simulations, a transient fluid-
structure interaction simulation was performed. The simulation was set on ANSYS
Workbench using Transient Structural and Fluent, coupled via the System Coupling
module.

Figure 4.3.1 – FSI simulation on ANSYS Workbench

Geometry and Mesh

The geometry and mesh used for this simulation were based on the results
achieved on the two former phases.

Transient Structural
The fluid-structure interface region was defined on Transient Structural:

Figure 4.3.2 – Fluid-structure interface region

Once again, the lateral faces of the leaflets were defined as fixed support.
 9

The total simulation time was 1 s, corresponding to two periods of the modeled
velocity profile, with a 5 ms time-step.

Fluid Dynamics
In addition to the Fluent settings presented on the fluid dynamics simulation, the
Dynamic Mesh resource was also configured, in order to assess the mesh movement
during fluid-structure interaction.

One-Way Coupling
The System Coupling module in ANSYS Workbench was used to couple
Transient Structural with Fluent, setting the data transfer between the interface regions.

5. Results

Figure 5.1 – Deformation of the leaflets (Static Structural Simulation)

 10

Figure 5.2 – Deformation of the leaflets (Transient Structural). Time = 0.6 s.

Figure 5.3 – Comparison between testing by Labcor and simulation performed

By comparing the figures above, it can be seen that the maximum deformation of
the valve in the experimental test section and the maximum deformation of the valve
model in the simulation are qualitatively very close. However, it’s necessary perform
experimental measurements of the valve deformation to get a quantitative comparison.
Furthermore, the results also resemble qualitatively those obtained by Haj-Ali et al. [8].
 11

6. Conclusion
The project carried out for the 6th Computational Modeling ESSS Challenge had
the objective to evaluate the possibility of implementing simulations in ANSYS in the
development process of biological artificial heart valves.
Based on the simulations, it is observed that, qualitatively, the results were
satisfactory. However, it is still necessary to perform experimental testing in order to
measure the valve deformation quantitatively and compare it with the results obtained in
the simulations, so that the simulations could be implemented in development of the
valves.

7. Future Directions
Experimental validation with quantitative measurements of the study conducted
is suggested. It is also suggested a fatigue study of the valves using simulation aiming
reducing costs with the preliminary tests, since the fatigue tests are extremely time-
consuming and expensive. Furthermore, it is proposed a simulation of the valve coupled
with an aorta model (Figure 7.1). Thus, the influence on the hemodynamics of the valve
with the aorta can be examined in order to develop more efficient valves.

Figure 7.1 – Future work: valve and aortal model coupling

 12

8. Bibliography
[1] Wattona, P.N.; et al.; Dynamic modelling of prosthetic chorded mitral valves using
the immersed boundary method; 2006.
[2] Nguyen, V.T.; et al.; Experimentally validated hemodynamics simulations of
mechanical heart valves in three dimensions; 2011.
[3] Alia, R.H.; et al.; Structural simulations of prosthetic tri-leaflet aortic heart valves;
2008.
[4] Tu, J.; et al.; Computational Hemodynamics – Theory, Modelling and Applications;
2015.
[5] Teixeira, G.M.; Kessler, M.P.; Maliska, C.R.; Fontes de erro: identificar, quantificar
e reduzir a intervalos aceitáveis.
[6] Fernandes, A.M.S. ; et al.; Impacto do perfil socioeconômico na escolha da prótese
valvar em cirurgia cardíaca; 2012.
[7] Haj-Ali, R; et al.; Strucutral simulations of prosthetic tri-leaflet aortic heart valves;
2008.
[8] de Hart, J; Fluid-Structure Interaction on the Aortic Heart Valve, a three-dimensional
computational analysis; 2002.

View publication stats