SIMULATION AND ANALYSIS OF FLOW IN DIFFERENT GEOMETRICAL ARRAYS OF A VASCULAR GRAFT

Garca J.*, Gonzlez *, Briceo J.C. * * Group of Biomedical Engineering, Mechanical Engineering Department, University of Los Andes

Abstract
Blood ow dynamics has a fundamental role in the success of vascular grafts. In particular, shear stresses and secondary ows are involved in platelet activation and aggregation. In order to investigate these three dimensional ow characteristics, a program was developed and applied to the results of computational simulations. The program evaluated shear stress and residence time values to get a level of platelet activation. The used geometries were a simplication of real geometries obtained by injection - corrosion method and angiographic images. The partial results show that shear stresses reached abnormally high values in the anastomosis region, and presence of secondary ows in the post - anastomosis region. Key words: Blood ow dynamics, wall shear stress, graft, platelet activation, computational uid dynamics.

Introduction
Cardiovascular diseases are the principal cause of death in the world. In 2005, 17.5 million people died because of cardiovascular disease, being coronary cardiopaties and cerebrovascular events the principal causes [6]. The group of biomedical engineering (GIB) has developed a regenerative vascular graft of porcine small intestine submucosa (SIS). The invivo experiments in animal veins showed a high rate of occlusion, therefore the study of ow is relevant to improve the design and implant protocol of the graft. Several groups have proposed computational evaluations of ow characteristics in cardiovascular devices and pathologies. These studies showed that shear stress along the particle trace was involved in blood damage, creating a blood damage index for hemolysis analysis. This index was based on shear stress and exposure time without checking platelet activation [1]. Other studies analyzed the ow in axisymmetric stenosis, and created a platelet activa1

tion index, based on shear stress and exposure time [2]. However, there is a shear stress threshold for platelet activation independent of the exposure time to each shear stress. Furthermore, most studies were interested in designing vascular grafts to improve the anastomosis in bypass surgery, not in a vessel replacement (Termino - terminal anastomosis) [3, 4]. As a result, in this work, ow through a simplied geometry of real collagen graft in jugular vein of leporids was analyzed. In addition, animal experiments have shown persistent occlusion of grafts. The occlusion taking into account ow variables only and the possible relationship with physiological variables will be explained. The real geometry is obtained in the rst minute after surgery and with help of angiographic images, a group of simplied geometries was built with dierent rate between graft and vessel diameter. Computational simulation was based on the nite volume method. Moreover a particle distribution at inlet of the model to be trajectorized was created. The results were exported to MATLAB for further analysis of platelet activation.

SIS Graft
GIB investigates invivo behavior of collagen regenerative grafts tested in jugular veins of leporids. The average diameter of the vein is 3mm and there were discussions about the ideal graft diameter to improve ow conditions at the anastomosis. These group completed works about the mechanical properties of the graft in dierent stages of regeneration [8, 16]. Regarding the ow there are no precedent works. The moment to analyze the ow is the rst postimplantation minute. It is in this moment when phenomena like inammation and vasoconstriction occur generating particular ow conditions. These conditions include high shear stress and secondary ows, as potential causes of thrombogenicity and eventual graft occlusion.

Objectives
General objective
To simulate and analyze ow in dierent geometries of the graft, varying the relation of graft-vessel diameter and the anastomosis diameter.

Specic objectives
To acquire and create geometries. To simulate ow. To analyze ow variables: shear stress, residence times and platelet activation.

Newtonian uid with viscosity = 3.5cP was considered. These conditions put the problem in laminar regimen with Re = 90. Streak lines were used to determinate shear stress and exposure time, a Lagrangian tracking approach was employed with displacement of each streak line being computed using forward Euler integration of the velocity over a time interval. The streak lines began at the geometry inlet in user dened position (Figure 1) and not all ended at the outlet, some streak lines entered in vortexes. Hydrodynamic conditions kept the platelets away from the vessel wall. The starting point for platelets at the inlet had a minimum distance from the vessel wall of 0.1mm taking into account hydrodynamic behavior.

Method
Acquisition and creation of geometries
A real geometry was acquired by using injection corrosion technique. This technique consisted of injecting a polymer in the vessel, sealing it and waiting for the hardening. Then the piece was immersed in a corrosive solution to acquire the internal geometry. Based on this geometry and angiographic images, a group of geometries was created in Solid Edge (Synchronous Technology 2 - SIEMENS CAD software) with dierent rate of vessel - graft diameters and dierent diameter of anastomosis, the vessel diameter was constant (3mm) and the rate varies from 1 to 1.5, the last one was equivalent to the actual diameter of graft in the animal model.

Figure 1: Starting points for platelets at inlet

Flow analysis
The results of streak lines were exported to MATLAB (R), and the stress tensor (ST ) at a given point in space was constructed as follows: xx xy xz ST = yx yy yz (1) zx zy zz v2 v1 xy = yx = + (2) x1 x2 v1 v3 xz = zx = + (3) x1 x3 v3 v2 yz = zy = + (4) x2 x3 2

Flow simulation
The simulation was performed in the commercial software Fluent (Versin 12.0, ANSYS). The geometries were meshed with tetrahedral elements changing the growth rate. Pressure convergence was checked increasing the number of calculated elements and calculating the head loss pressure in each simulation. Boundary conditions were pressure outlet at the end of the geometry (P = 666P a) and a paraboloid velocity prole at inlet with average velocity (V = 0.1 m ) for the worst high velocity s case. The convergence criterion for simulation was 106 for residuals values (Continuity and velocities).

zz

2 v3 = p .V + 2 3 x3

Results and analysis

(5) (6) (7)

Acquisition and creation of geometries

The principal characteristic of the geometries is the inuence of the sutures that creates a parachute shape in the anastomosis region (Figure 2). The geometries created have six dierents relation of diDG ameters graft - vessel DV for 1.0 to 1.5, and three dierent relation of diameters anastomosis - vessel DA DV (Figures 3 and 4).

2 v2 yy = p .V + 2 3 x2 2 v1 xx = p .V + 2 3 x1

where is the normal stress and is the shear stress. Once the shear stresses were determined, the Eigen values were found for every point of all trajectories, providing the 3D principal stresses, and the maximum shear stress: max = 3D = (max min ) 2 (8)

The obtained data was processed to calculate a level of platelet activation (LPA) [2]: LP A = ti (9)

where is the maximum shear stress calculated by equation 8, and ti is the exposure time exported from the simulation. The program creates a group of symbolic platelets simulating the whole blood volume, a number corresponding to trajectories is randomly selected, and randomly placed in each trajectory, for symbolic platelets, the program calculates the LPA and nds the maximum shear stress. A platelet was activated if the LPA is higher than 1.7P a s or if max is higher than 10.5P a [2]. These values were considered for young platelets, the capability to become activated decreases with the age of the platelet [5]. An age factor is randomly given to each platelet, the older platelet is activated under LP A = 10, 5P a s and max = 21P a or higher values. The information was saved and the platelets returned to the blood volume, which was randomly mixed. A degeneration factor was included for platelets, randomly replacing some of them in each cycle for new platelets and replacing the older ones. This cycle was repeated ten times to calculate the percentage of activated platelets. The complete process was repeated ten times to get an average to eliminate the random eect. Results were analyzed and related to possible physiological factors that could accelerate the activation and occlusion process. 3

(a)

(b)

Figure 2: Suture inuence in injection - corrosion model (a), global shape in angiography image (b)

ary condition of velocity at inlet is shown in gure 6.

Figure 3: Simplied geometries, three dierent anasDA DA tomosis diameters a) DV = 0.5, b) DV = 0.66, c) DA DV = 0.86

Figure 5: Problem convergence graphic

Figure 4: Simplied geometries, group with DG 0.66 and dierent relation of diameter DV

DA DV

= Figure 6: Paraboloid velocity inlet contour

Flow simulations
The convergence analysis starts with a calculation of 50.000 tetrahedral elements and the number of elements was duplicated until 1.6 106 . The simulation results showed that the adopted model with 0.8 106 elements and a rened model with 1.6 106 element has less than 1% dierence between head loss pressure (Figure 5). The mesh was controlled by varying the growth rate. The paraboloid bound4

The ow phenomenon observed in each geometry has particular elements to evaluate. Therefore, it is important to reduce the shear stresses and to avoid the secondary ows. The general results present an acceleration at the anastomosis and secondary ows after the anastomosis, consequences of the reduction and expansion of diameter. The exception is the small relation of graft - vessel diameter in the bigger anastomosis, D = 2.5mm. In these cases, the geometry has low perturbation of ow, but in the others

geometries the phenomenon change completely. The size of the secondary ows increases with the relation of diameters and the inverse of the anastomosis diameter. These secondary ows have a low rotational speed when the relation of diameters is large. Consequently, low speeds could promote aggregation of platelets when groups of activated platelets fall in these zones. Figures 7 - 12 show the pathlines for the smaller and the bigger relation of each anastomosis diameter.

Figure 9: Pathlines

Dgraf t Dvessel

= 1 and D = 2mm

Figure 7: Pathlines

Dgraf t Dvessel

= 1 and D = 2,5mm

Figure 10: Pathlines

Dgraf t Dvessel

= 1.5 and D = 2mm

Figure 8: Pathlines

Dgraf t Dvessel

= 1.5 and D = 2,5mm

Figure 11: Pathlines

Dgraf t Dvessel

= 1 and D = 1,5mm

was found. In anastomosis with a rate smaller than DA DV = 0.66, the number of activated platelets by cycle increase with the relation of diameters, and when the anastomosis diameter decreases, this behavior is more evident (Figure 14).

Figure 12: Pathlines

Dgraf t Dvessel

= 1.5 and D = 1,5mm

Flow analysis
Particle trajectories were exported from the computational calculation with values of velocity and residence time. This information was organized to work particle by particle and the shear stress is calculated using the equation 8. An example of trajectories is shown in gure 13 where all trajectories start at inlet and almost everyone escape for outlet, some of them fall in secondary ow zones, and recirculate until a threshold of 10000 steps of trajectory calculation.
(a)

(b)

Figure 13: Example of streak lines

For each geometry LPA and maximum shear stress were calculated and the percentage of activated platelets was found for each cycle. The results of activated platelets do not show a general dependency with relation of diameters. However, evaluating each anastomosis diameter, a weak relation between activated platelets and relation of diameters 6

(c)

Figure 14: Graphic of activated platelets vs numDA ber of cycles, anastomosis diameter a) DV = 0.5, DA DA b) DV = 0.66, c) DV = 0.86, legend at right side DG indicates DV

Histograms were constructed for LPA and max in each geometry and cycle (Figures 15 and 16). In DA DV = 0.86, there is no activation by max , just by LPA. In the other two anastomosis diameters the principal activator is shear stress. The evolution in time of LPA in platelets circulating through the graft (Figure 15) has a particular behavior. In general, the LPA values are higher at DG high DV , but the relation of the evolution and values with the anastomosis diameter does not have a

increasing or decreasing tendence. The LPA is the product of two variables, shear stress and residence time, which are dependent on each other, because both depend on velocity. If shear stress is high, the velocity also has to be high, thus, the residence time is going to be shorter. This perspective explains the DA evolution of LPA. In DV = 0.86 the velocity does not have an important increase at the anastomosis region, thus, the residence time is longer and the shear stress values are higher than normal by geometric conditions.

(a)

DA DV

= 0.86,

DG DV

=1

(b)

DA DV

= 0.66,

DG DV

=1

(c)

DA DV

= 0.5,

DG DV

=1

(d)

DA DV

= 0.86,

DG DV

= 1.5

(e)

DA DV

= 0.66,

DG DV

= 1.5

(f)

DA DV

= 0.5,

DG DV

= 1.5

Figure 15: Histograms LAP, legend at right side indicates number of cycles

DA In DV = 0.66, velocity at anastomosis is higher than velocity in last geometry and the residence time is shorter. On the other hand, the secondary ow zones are larger, so there are larger regions of high shear stress between main accelerated ow and secondary ows. DA In DV = 0.5, velocity at anastomosis are higher than the other geometry velocities, and the residence time shorter. In addition, the secondary ow zones increase their size again.

In general there are three important variables asociated with LPA, two of which are directly asociated: residence time and shear stress. The otherone is from the observation of ow: secondary ow zones size. The results suggest that the relationship of the LPA with the anastomosis diameter has a maximum DA DA of between DV = 0.5 and DV = 0.86, because in the DA histograms the rate DV = 0.66 always has higher values, where the three variables generate a higher total.

(a)

DA DV

= 0.86,

DG DV

=1

(b)

DA DV

= 0.66,

DG DV

=1

(c)

DA DV

= 0.5,

DG DV

=1

(d)

DA DV

= 0.86,

DG DV

= 1.5

(e)

DA DV

= 0.66,

DG DV

= 1.5

(f)

DA DV

= 0.5,

DG DV

= 1.5

Figure 16: Histograms maximum shear stress, legend at right side indicates number of cycles

Histograms of shear stress have a relationship with the rate of diameters and the anastomosis diameter. If the reate of the diameters increases, the histograms of shear stress register higher values and more population in high values. This happens because the size of secondary ow zones increase and the acceleration of ow at graft outlet is higher (Figure 16). The shear stress increases with the inverse of the anastomosis diameter because the velocity at anastomosis increases and these high velocity values are closer to the wall.

Moreover, collagen, the main graft material, is recognized by platelets, for example in the presence of injury, and this gets them activated starting a coagulation process. If enough platelets become aggregated, the coating starts and the possibility of adhesion to the graft wall increases. Physiological factors help ow variables to occlude the graft, but the largest number of platelets and coagulation factors are present 7 to 10 days after injury. This period of time is important because the occlusion could occur during those conditions.

Discussion
LPA and shear stress were the two activating factors taken into account, LPA has low values at relations DA DA DV = 0.86 DV = 0.5, and shear stress has low valDA ues at DV = 0.86. In general, these two factors have minimum values at small relation of diameters. The DG analysis suggests a relation of diameter DV = 1 for the size of the secondary ow zones and an anastomosis diameter closer to vessel diameter. When a platelet is activated its morphology changes to get more chances of aggregation with other platelets and adhesion to the vascular wall. 8

Conclusion
In this simplied model, the analysis of ow variables and the activation of platelets suggest an important role of geometry, especially the anastomosis diameter, in the phenomenon of occlusion in collagen vascular grafts. This analysis could be related to physiological coagulation factors and increase the probability of occlusion. An improvement in implantation method is recommended, trying to reduce the inuence of sutures, avoiding the reduction of diameter at the anastomosis. Another important moment of ow evaluation is the period between 7 to 10 days

after surgery, when an inammatory response happens and the number of platelets increase to a maximum. Possible future work can include a platelet aggregation model to make a more accurate prediction of graft occlusion.

References
[1] Chua, Leok Poh. Su, Boyang. Lim, Tau Meng. Zhou, Tongming. Numerical Simulation of an Axial Blood Pump. Articial Organs. Volume 31. Issue 7. Blackwell Publishing Inc. 15251594. [2] Bluestein, D.; Niu, L.; Schophoerster, T.; Dewanjee, M. K.. Fluid mechanics of arterial stenosis: Relationship to the development of mural thrombus. Annals of Biomed. Eng, 25:344-356, 1997. [3] Qiao, Aike, Liu, Youjun, Guo, Zhihong. Wall shear stresses in small and large two-way bypass grafts, Medical Engineering & Physics, Volume 28, Issue 3, April 2006, Pages 251-258, ISSN 1350-4533, DOI: 10.1016/j.medengphy.2005.05.004. [4] Lei, Ming. Archie, Joseph P.. Kleinstreuer, Clement. Computational design of a bypass graft that minimizes wall shear stress gradients in the region of the distal anastomosis, Journal of Vascular Surgery, Volume 25, Issue 4, April 1997, Pages 637-646, ISSN 0741-5214, DOI: 10.1016/S0741-5214(97)70289-1. [5] Fulton, John F.. A Textbook of Physiology. W. B. Saunders Company. Seventeenth edition. Philadelphia, 1955. Pages 524 - 526. [6] Organizacin Mundial de la Salud (OMS). Enfermedades cardiovasculares, datos principales. Consultado el 6 de noviembre de 2009. Disponible en <www.whi.int> [7] Yalcin, Bulent. Komesli, Gokhan H. Ozgok, Yazar. Ozan, Hasan. Vascular anatomy of normal and undescended testes: Surgical assessment of anastomotic channels between testicular and deferential arteries, Urology, Volume 66, Issue 4, October 2005, Pages 854-857. 9

[8] Snchez, Diana M.. Beltrn, Ricardo. Silva, Ana C.. Quijano, Lina M.. Solano, Omar. Vega, Francisco. Mugnier, Jaqueline. Pieros, Diego. Espinel, Camilo. Moreno, Julian M.. Briceo, Juan C.. Evolution of the remodeling of the arterial wall in a growing animal model using small intestine submucosa vascular grafts. Universidad de los Andes. Facultad de Ingeniera. Grupo de Ingeniera Biomdica. 2008. [9] Meyer, Eric P.. Beer, Gertrude M.. Lang,Axel. Manestar, Mirjana. Krucker, Thomas. Meier, Sonja. Mihic-Probst, Daniela. Groscurth, Peter. Polyurethane elastomer: A new material for the visualization of cadaveric blood vessels. Clinical Anatomy. Volumen 20. Issue 4. 2007. Pages 448-454. [10] Garca Rodrguez, Javier Francisco. Protocolo para la utilizacin de la mquina de PIV en modelos cardiovasculares. Bogot, 2009. Tesis (Pregrado en Ingeniera mecnica). Universidad de Los Andes. Facultad de ingeniera. Departamento de ingeniera mecnica. [11] Garca Rodrguez, Javier Francisco. Soluciones para el uso en un simulador cardiovascular. Bogot, 2008. Proyecto inteermedio (Ingeniera mecnica). Universidad de Los Andes. Facultad de ingeniera. Departamento de ingeniera mecnica. [12] Brunette, J.. Mongrain, R.. Laurier, J.. Galaz, R.. Tardif, J. C.. 3D ow study in a mildly stenotic coronary artery phantom using a whole volume PIV method, Medical Engineering & Physics, Volumen 30, Issue 9, Noviembre 2008, Pages 1193-1200, ISSN 1350-4533, DOI: 10.1016/j.medengphy.2008.02.012. [13] Fasel, Jean H.D.. Portal Venous Territories Within the Human Liver: An Anatomical Reappraisal. The Anatomical Record: Advances in Integrative Anatomy and Evolutionary Biology. Volumen 261. Issue 6. 2008. Pages 636-642. [14] Maldonado Martinez, Natalia. Desarrollo y validacin de modelos computacionales e in vitro para el estudio de la conexin total cavopulmonar. Bogot. 2007. Tesis (Maestra en Ingeniera mecnica). Universidad de Los Andes.

Facultad de ingeniera. Departamento de ingeniera mecnica. [15] Manual Merck de informacin mdica para el hogar, Seccin 3: Enfermedades cardiovasculares, Capitulo 26: Aterosclerosis. Consultado el 28 de febrero de 2010. Disponible en <www.msd.es/publicaciones/mmerck_hogar>. [16] Martinez Movilla, Rosalba Rebeca. Estudio de la deexin y los esfuerzos constantes en la interfase entre la arteria e injerto vascular. Bogot. 2007. Tesis (Pregrado en Ingeniera mecnica). Universidad de Los Andes. Facultad de ingeniera. Departamento de ingeniera mecnica. [17] Medline Plus Encilopedia Medica, Aneurisma. Consultado el 28 de febrero de 2010. Disponible

en <www.nlm.nih.gov/medlineplus>. [18] Rael, M., Willert, c. E.,Wereley, S. T., &Kompenhans, J. Particle Image Velocimetry: A practical guide. Segunda edicin. Nueva York. Springer Berlin Heidelberg. 2007. [19] Budwig, R. (1994). Refractive index matching methods for liquid ow investigations. Experiments in uids, 17, (5), 350 355. Recuperado el 21 de febrero de 2008, de la base de datos Springerlink. [20] Snchez Palencia, Diana Marcela. PIV Fontan experiments, Protocol and results. Atlanta, 2007. Cardiovascular Fluid Mechanics Laboratory. Georgia Institute of technology. Biomedical engineering department.

10