Pulsatile non-Newtonian blood flow in threedimensional carotid bifurcation models: a numerical study of flow phenomena under different bifurcation

angles
K. Perktold, R.O.

Peter, M. Resch and G. Lang

Institute of Mathematics,
Received May IWI, accepted

University

Graz, Graz, Austria

ABSTRACT
Flow and stress patterns in human carotid artery brfircation models, which differ in the bifurcation angle, are analysed numerically under physiologically relevant flow conditions. The governing Navier-Stokes equations describingpulsatile. three-dimensionaljow of an incompressible non-Newtonian fluid are approximated using a pressure correction finite element method, which has been developed recently. The non-Newtonian behaviour of blood is modelled using C&ons relation, based on measured dynamic viscosity. l%e study roncentrates on jlow and slresscharacteristics in the carotid sinus. The results show that the complex flow in the sinus is affected by the angle variation. The magnitude of reversed flozv, the extension of the recirculation zone in the outer sinus region and the duration ofjlow separation during the pulse cycle as well as the resulting wall shear stress are clearly different in the small angle and in the large angle bifurcation. The haemodynamic phenomena, which are important in atherogenesis, are more pronounced in the large angle bifurcation.
Keywords: carotid artery, bifurcation models, Navier-Stokes equations, numerical analysis

INTRODUCTION
Atherosclerotic alterations occur preferentially in the carotid artery bifurcation and especially in the carotid sinus. From the fluid dynamic point of view the flow is strongly disturbed in this region. The influence of disturbed flow on atherogenesis has been studied intensively over the last twenty years. Correlations between fluid dynamics and the localized genesis and the development of atherosclerosis are well known. The findings indicate that flow separation and recirculation are important factors in the deposition of platelet thrombi and in the occurrence of early atherosclerotic lesions. To investigate the flow field and the stresses in the carotid bifurcation considerable experimental- and theoretical- research work has been performed. The fluid dynamics studies show that flow separation phenomena occur in the carotid sinus. This investigation analyses the bifurcation angle dependency of the flow characteristics in the sinus during the pulse cycle. In the study four carotid bifurcation models, which differ in the bifurcation angle, are analysed numerically. The angle between common carotid axis and internal axis was chosen from the range 1.5 to 50. The basic shape of the model, except the bifurcation
Correspondence and reprint requests to: Karl Perktold, Institute of Mathematics, Technical University Graz, Steyrergasse 30/3, A-8010 Graz, Austria 0 19!)1 Butterworth-Heinemann 011 I-5425/91/060507-O!) for BES

angle, agrees with data from the literature-. The calculations have been carried out under the physiological pulse waveform and flow division ratio internal to external carotid published by Ku et al.. The mean flow rate in the common carotid artery is 5.1 mls-. The calculations are based on the pulsatile threedimensional (SD) N avier-Stokes equations where non-Newtonian blood viscosity is assumed. The nonNewtonian behaviour of blood has been described using the Casson relation, to which experimental dynamic viscosity data are fitted. The computer simulation of the pulsatile flow field and of the wall shear stress has been carried out using a recently developed pressure correction finite element procedure. The numerical method is described only briefly here. Our study concentrates on the presentation and discussion of the numerical results for flow velocity (axial and secondary velocity, flow separation, stasis and transient flow reversals) and wall shear stress distribution during the pulse cycle.

GEOMETRICAL MODELS AND FLOW CONDITIONS

J. Biomed.

Eng. 1,091, Vol. 13. November

507

Carotid flow phenomena: K. Perktold et al.

t
D=62mm

Figure 1 Four human carotid artery bifurcation models. a, CAR 1; b, CAR 2; C, CAR 3; d, CAR 4. The essential geometric data are indicated; A, B, C, D; E, F indicate flow cross-section levels where numerical results are displayed

carotid artery (D = 6.2mm), of the internal (Oi = 0.70) and of the external (DeX = 0.590) as well as the sinus shape (maximum sinus diameter D, = 1.06D) correspond to their assumptions. The angle between common carotid axis and internal carotid axis is 15 (CAR I), 25 (CAR 2), 40 (CAR 3) and 50 (CAR 4); between common carotid and external carotid it is 25 (CAR 1, 2, 3) and 50 (CAR 4). In all cases the branching plane is the plane of symmetry. The four bifurcation models, including the essential geometrical data, are shown in Figure 7. The capital letters, A, B, C, D, E, F, G, H, I indicate different cross-section locations where numerical results are presented. Cross-section A belongs to the common carotid; B, C, D, E, F, are located in the internal carotid where level B is proximal to the carotid sinus, level D agrees with the maximum sinus diameter, and level F is distal to the internal carotid. G, H, I belong to the external carotid. The flow divider is slightly rounded; the vessel wall is assumed to be rigid. Figure 2 shows a three-dimensional representation of CAR 1 where the inner, outer and side walls are defined. The calculations are carried out under unchanged non-Newtonian pulsatile flow conditions. The complex rheological behaviour of blood has been approximated using Cassons relation on the basis of

experimental viscosity data corresponding to an oscillation of 2 Hz (ref. 17). The Casson relation expresses the viscosity as a function of the shear strain Internal
carotid artery

Outer wall

Carotid

\lnner -Side

wall wall

Figure 2 Three-dimensional representation of the small angle carotid CAR I; definition of the inner, outer and side wal1

508

J. Biomed. Eng. 1991, Vol. 13, November

Carotid flow phenomena: K. Perktold et al.

Systole

Oiastole

I
10-l

I loo

I 10'

I lo2

I lo3

+a
Dynamic viscosity of human blood, hemaocrit c = 43%; the experimental data measured in oscillatory flow (2H) (ref. 17) are fitted using Cassons relation

Ol

Figure 3

0.20 Figure 4

0.40

0.60

0.80

1 .oo

rate (and of hematocrit; here 43%). The shear strain rate dependent blood viscosity is shown in Figure 3. The pulse waveform in the common carotid and in the internal carotid used in this investigation has been published by Ku et al7 Figure 4 shows the flow rate pulse waveform in the common and in the internal carotid. The mean flow rate in the common carotid is assumed to be 5.1 ml s-, and the corresponding mean flow velocity is 0;) = 16.9 cm s-. Using the common diameter D = 6.2 mm and a representative kinematic reference viscosity Y = 0.035 cm2 s- in the definition, the mean reference Reynolds number is Re = X)0 where Re = UJl/v. The assumed pulse frequency is HO strokes min-. The corresponding CY = 4.8 where number Womersley reference (Y= D/:!v%& w is angular frequency. EQUATIONS The mathematical analysis was carried out using the time-dependent non-linear Navier-Stokes equations. In this study blood is assumed to be an incompressible Casson fluid. The equation system can be written as

titp --c Pulsatile velocity waveform in the common and internal carotid artery according to Ku ef al.; tp denotes the time for one pulse cycle; mean common carotid flow 5.1 mls . -. common carotid flow; ---, internal carotid flow

-.L = () ilx,

il u

i,j=

1,2,3

where ui, i = 1, 2, 3, are velocity components in the Cartesian coordinate system Xi, i = I, 2, 3; p is pressure, p is constant fluid density and p is shear rate dependent dynamic viscosity. As shown by Perktold et al. Ix the viscosity in general Casson flow can be expressed as ,i =

I(2 I.&) (k,,(c)+ k , (c) w

data are fitted using the The experimental hematocrit-dependent paramaters k,) = 0.6125 and k I = 0.174 whereby hematocrit c = 43%. D is the second invariant of the strain rate tensor. Equation 3 describing the apparent viscosity fits blood data quite well at shear rates higher than 1 s-l, From previous carotid bifurcation flow studies it is known that the shear rates to be expected are higher than 1 s-.

The generalized Navier-Stokes system equations (l)-(3) can be solved for given boundary and initial conditions. Here at the inflow boundary the Womersley velocity profiles are prescribed. These fully developed profiles were calculated as long straight tube profiles correspondin to the velocity waveform in the common carotid 8Figure 4). At the rigid vessel wall the no slip condition is applied. At the outlet of the two branches time-dependent flow division ratio is prescribed. At the outflow boundary of the external carotid, which is assumed at a distance of eight times the external diameter from the divider, outflow velocity profiles are calculated in a first step. During this step fully developed flow is applied at the internal outlet located downstream at a distance of 3.5 times the internal diameter from the sinus end. The external outflow velocity profiles resulting from this step are used as boundary conditions in a second step where traction-free boundary conditions are assumed at the internal outlet. The second step is the actual calculation step. Studies to estimate the appropriate length of the two branches have been carried out. The numerical procedure recently developed is based on the pressure correction technique and uses the Galerkin finite element method and implicit finite differences for the time derivatives (1I).ILo.The nonlinear viscosity and the convection terms are linearized using Picard iteration. The finite element discretization employs eight-node isoparametric brick elements with tri-linear velocity approximation and constant pressure. The validation of the method in the Newtonian case has been carried out by Hilber? and by Perktold et al. where a comparison of curved tube results with results by van de Vosse et al_has been done. A comparison of non-Newtonian results is performed in Perktold and Resch. The finite element subdivision of the computational domain (symmetry is taken into account) creates 7228 elements. The total node number (three velocity components and pressure) is 34498.

NUMERICAL

RESULTS

The presentation of the results concentrates on the axial and secondary velocity field, the recirculation

.J. Biomed.

Eng. I!)!)]. Vol. 13, November

509

Carotidflow

phenomena: K. Perktold

et al.

t/to

= 0 80

Figure 5 Axial velocity profiles in the branching plane during systolic deceleration l/lp = 0.14 and during diastolic phase l/tp = 0.X. a, CAR 1; b, CAR 2; c, CAR 3; d, CAR 4

zones and the wall shear stress distribution during the pulse c cle. The comparison of the results illustrates uence of the bifurcation angle on the flow the in fly parameters. Comparison of the flow velocity

The axial velocity profiles in the branching plane of the four models at different locations and at different fractions of the pulse period t/ tp = 0.14 (systolic deceleration phase), t/ tp = 0.8 (diastolic phase) are shown in Figure 5. In the common carotid, upstream the symmetric inflow profiles of the branching, remain relatively unchanged. In the two branches, extremely skewed profiles with high velocity near the inner (divider) wall occur especially for high flow rates. At the outer wall the velocity decreases. During decelerated flow in the outer sinus region flow separation and recirculation appear. The magnitude of the reversed velocity and the extension of the reversed flow area increase with increasing bifurcation angle. During the diastolic phase low axial velocity in the outer sinus region can be observed; in the case of CAR 4 the plot demonstrates diastolic flow stagnation near the outer sinus wall. Figure 6 shows a three-dimensional representation of the axial velocity profiles in CAR 1 and CAR 4 during the systolic deceleration phase at different flow cross-sections. The location of the cross-sections is indicated in

Figure 6 Three-dimensional representation of axial velocity in the common carotid (level A) and in the internal carotid (levels B, D, E) of the small angle (CAR 1) and the large angle bifurcation (CAR 4) during systolic deceleration l/tp = 0.14. The location of the crosssections is indicated in Figure 7. a, CAR I; b, CAR 4

Figure 1. The comparison illustrates a more complex flow field in the sinus (locations B, D, E) in the case of CAR 4. Some results on secondary motion are displayed in Figure 7 and Table 1. The comparison of the secondary velocity profiles in CAR 1 and CAR 4 during systolic deceleration (t/ tp = 0.14) and during the diastolic phase (t/p = 0.8) demonstrates an increase of the secondary velocity with increasing bifurcation angle (Figure 7). The secondary motion

510

J. Biomed.

Eng. 1901, Vol. 13, November

Carortdjlow phenomena: K. Perktold et al.

t/tp = 0.14

j Outerwall r/tp = 0.80

0.80

t-ii>@@

a
Figure 7 Secondary flow velocity at the specified during systolic deceleration f/tp = 0.14 and during

flow cross-sections (A. common diastolic phase f/t/~ = 0.8. a, CAR 1; b, CAR 2

, 50cms- 1 internal carotid) in CAR I and CAR 4 carotid; B, C. D, E, E,

results from the branching and the curvature effect. At the location proximal sinus (B) the branching effect is dominant. Table 1 shows the magnitude of maximum secondary velocity at the location of maximum sinus diameter D during the pulse cycle. During systolic acceleration, at maximum flow rate and during diastolic phase the maximum secondary velocity approximately doubles from the small angle (CAR 1) to the large an le bifurcation (CAR 4). During decelerated flow t/ tp = 0.14) only minor differences occur. This behaviour can also be seen in Figure 7.
Table 1 Magnitude of maximum of maximum sinus diameter D Maximum magnitude secondary velocity at the location

Additional information on the secondary motion is gained from the mean axial vorticity2,, which is defined as [,= l/A$siidY where A = nd2/8 is the area surrounded by the path S along the tube diameter d in the symmetry plane and the semicircle. Figure 8 shows the mean axial vorticity in the internal

of secondary location D

velocity

in cm s

t/t/l o.oF, 0.10 0.14 0.2kI 0.27 0.36 0.46 0.80

CAR 2.7 !?.!I 16.1 6.2 5.1 4.2 3.5 2.4

CAR 2 3.6 12.8 17.3 6.3 7.2 5..5 4.8 3.3

CAR 3 5.0 16.6 17.0 7.3 7.5 6.8 6.2 4.4

CAR 4 5.9 18.3 15.7 7.1 7.5 7.4 7.0 5.1

Figure 8 Mean axial vorticity in the internal carotid from the location proximal sinus (B) to the distal internal level F at l/lp = 0.14 (top) and at t/tp = 0.8 (bottom) for CAR l-4 (a-d, respt)

j. Biomed.

Eng. 19!)1, Vol. l.i, November

511

Carotidflow

phenomena: K. Perktold et al.

carotid from location B (proximal sinus) to location F t/ tp = (distal internal) d uring systolic deceleration 0.14 and during the diastolic phase t/tp = 0.8. At the proximal sinus the branching effect is dominant, and high axial vorticity occurs in all models considered. Along the sinus the vorticity decreases and reaches a minimum at location E. This minimum is caused by the cross-section reduction at the downstream end of the sinus together with the high axial velocity in the region of the inner sinus wall. The resulting effect is contrary to the secondary motion effect from the curvature and the branching. For increasing bifurcation angle an increase of the axial vorticity is apparent. Comparison sinus of reversed flow zones in the carotid

The analysis of the bifurcation angle-dependent location and extension of flow separation zones, and of the duration of reversal flow during the pulse cycle are crucial points in the present study. In these zones, relatively low flow velocity and velocity gradients and consequently low wall shear stresses occur. These
0.32 0.1 0.28 0.26 0.24

facts are of essential importance in the deposition of platelet thrombi and in the occurrence of early atherosclerotic lesions. The zones of reversed axial flow in the plane of symmetry of the models CAR 1 and CAR 4 during the pulse cycle are plotted in Figure9. The lines in the flow field are zero axial velocity contour lines corresponding to different pulse phases. Major flow separation and recirculation occur during systolic deceleration. The plots show that flow separation at the outer sinus wall first appears at the location proximal sinus (level B) during late systolic acceleration. In the case of CAR 1 the beginning of the flow separation is detected at t/tp = 0.1; with an increasing angle, separation can be seen somewhat earlier. In CAR 4 the beginning is at t/tp = 0.06. During the systolic phase the zone of reversed axial velocity grows and reaches its maximum extension in the second half of the systolic deceleration phase, approximately at t/ tp = 0.16, 0.18. Subsequently the recirculation zone decreases and disappears approximately at t/ tp = 0.28 in CAR 1 and t/ tp = 0.26 in CAR 4. The location of the disappearance is shifted downstream up to the maximum sinus diameter level. In all cases at the proximal sinus location, minor separation reoccurs during the end of the systolic phase. In the figure it can be seen that there are only minor differences in the duration of the systolic flow separation, but in CAR 4 additional separation is

0.24 \

0.46-l

.O 3

Figure 9 Zones of reversed flow at the branching plane during the pulse cycle at discrete pulse phases (A(Ulp) = 0.02). The lines are zero axial velocity levels. a, CAR 1; b, CAR 4

cross-sections (A, common carotid, B, C, D, E, F, internal carotid) in Carotid 1 and Carotid 4 during systolic deceleration t/tp = 0.14. a, CARl;b,CAR4

b Figure 10 Zones of reversed flow (hatched areas) at specified flow

512

J. Biomed. Eng. 1991, Vol. 13, November

Carotidjlow phenomena: K. Perktold et al.

detected during the diastolic phase from t/p = 0.46 to the end of the pulse cycle. A further difference can be observed in the location of the separation zones. In model CAR 1 flow separation occurs already at the terminal end of the common carotid, while in model CAR 4 the separation zone develops at the beginning of the internal carotid artery. At the outer wall of the external carotid in CAR 4 flow separation is also observed where the maximum extension appears again at the systolic deceleration phase t/ t$t = 0.16. Figure 70 illustrates the cross-sectional extension of flow separation in the sinus of CAR 1 and CAR 4 at systolic deceleration t/ tp = 0.14. The areas of reversed axial velocity are hatched. The plots show the different shape of the cross-sectional separation areas. With increasing bifurcation angle the shape of the zero axial velocity contour lines indicates the development of so-called C-shaped axial velocity distribution, where the area of reversed flow is relatively concentrated in the vessel centre. This Cshaped axial velocity distribution occurs because of the high axial velocity near the inner sinus wall and due to the movement of the high velocity particles towards the outer wall by circumferential secondary motion. Comparison of wall shear stress

B, D). The maximum negative shear stress is 2.6 Nm-2 and a pears at the location of maximum sinus diameter ?D) in CAR 4. Negative wall shear stress indicates reversed flow at the specified wall locations B, D. From the shear stress distribution it can be detected that the duration of systolic flow separation at the outer wall locations B and D of CAR 1 and CAR 4 are only slightly different. However, in CAR 4 the flow separation begins somewhat earlier

Y
E z :

al
a

The wall shear stress at different locations at the inner and the outer wall (A, B, D, F) in the cases of CAR 1 and CAR 4, as well as at the side wall (B,D) in CAR 1, 2, 3, 4 during the pulse cycle is displayed in Figures 7 l- 73. The sign corresponds to the tangential velocity direction near the wall. Figure 77 shows the difference in the shear stress distribution at the inner and at the outer wall. Shear stress is significantly higher at the inner wall; the systolic maximum is 14 N rnd2 and can be seen near the flow divider tip (level B) in CAR 1. While the shear stress at the inner wall of the sinus is maximum, negative shear stress occurs at the outer wall (location,

_ 7 E
Z 2

t/tp

1 .o

Figure 12 Wall shear stress magmtude at the side wall of the carotid smus (levels R and D) during the pulse cycle. n, CAR I; 13, CAK 2; +. CAR 3; A. CAR 4. a, B (side); b, D (side)

CA

t/tp

-0

ok-0 t// tp
J. Riomed. Emq. 1991, Vol. 13, November 513

Figure 11 Shear stress at the outer and at the inner wall in the small angle bifurcation Carotid 1 and in the large angle bifurcation Carotid 4; a, level A: common carotid; b, level B: proximal carotid sinus; C, level D: maximum carotid sinus diameter; d, level F: distal internal carotid (level locations shown in F&W 7). III, CAR I; A, CAR 4 (open and solid symbols represent inner and outer wall respectively)

Carotidjow phenomena: K. Perktold et al.

in the pulse cycle. The diastolic outer sinus wall shear stress is low and nearly constant. In contrast to the behaviour of the other three models used, in CAR 4 negative diastolic wall shear stress occurs at the location proximal sinus (B). This implies flow separation during the diastolic phase. The wall shear stress results indicating the duration of flow separation agree with the results shown in Figure 9. The magnitude of the shear stress vector at the side wall locations B and D is displayed in Figure 72. During the entire pulse cycle the highest side-wall shear stress occurs in the case of CAR 4. At location B the systolic peak shear stress is about four times higher, and during the diastolic phase about twice as high, when compared with CAR 1. Figure 73 illustrates the angle of the shear stress vector at the side wall (locations B, D) with respect to the internal axis. The maximum angle is about 9.5 and occurs in CAR 1 during the systolic peak flow. During the diastolic phase the maximum angle of the side wall shear stress vector appears in CAR 4.

CONCLUSION Detailed analysis of the flow dynamic characteristics were the basis for the investigation of the role of haemodynamic phenomena in atherogenesis. In our study four human carotid bifurcation models, which differ in the bifurcation angle, were analysed numerically under physiologically relevant pulsatile flow conditions. In the investigation the axial and secondary velocity, the extension and location of the flow recirculation zones, and the duration of flow separation in the carotid sinus, and the shear stress distribution at specified locations at the inner, outer and side wall are calculated. In a previous investigation concerning the geometric factor in atherogenesis flow studies have been carried out for carotid bifurcation models with different sinus shape whereby idealized constant flow division ratio was assumed. The parameter study showed that the fluid dynamic variables in the sinus are significantly affected by the angle between common carotid axis and internal carotid axis. With increasing bifurcation angle from CAR 1 to CAR 4 increased reversed flow and enhanced flow recirculation in the outer sinus region occur. In the case of the large angle bifurcation CAR 4 at location proximal sinus flow separation accompanied with low negative wall shear stress was found over the major part of the pulse cycle, while in the small angle bifurcation CAR 1 flow separation occurs only during the systolic phase. In CAR 1 flow separation was found at the terminal end of the common carotid; in CAR 4 separation develops at the beginning of the internal carotid. From the clinical point of view it is known that large angle bifurcations have unfavourable characteristics in the development of atherosclerosis?. In the literature it is documented that flow separation and recirculation and the resulting low wall shear stress development of favour the atherosclerotic lesions 4,7.27-2. The present study supports the haemodynamic theory of atherogenesis, that the specific occurrence of this disease may reflect the individual variations in the vascular geometry. One aspect, which confirms the crucial importance of flow separation and recirculation and of low wall shear stresses in atherogenesis, is described by DeSyo. It is reported that in higher level carotid bifurcations (larger angles) atherosclerotic plaques are typically located a short distance downstream from the starting point of the internal carotid artery, while in lower level bifurcations (smaller angles) typical atherosclerotic processes can be observed at the terminal end of the common carotid artery and at the very beginning of the internal carotid artery. This bifurcation angle-dependent location of deposits and stenosis, reflects the variation in the position of the zones of flow separation as analysed in this study.

60 B E 8 Y c

30

01

01

1.0 t/tp

ACKNOWLEDGEMENT This investigation is supported by the Austrian Research Fund (Fonds zur. Forderung der wissenschaftlichen Forschung in Osterreich), Projekt-Nr. P 7726 PHY, Vienna, Austria.

Figure 13 Variation of the direction of the wall shear stress vector at the carotid sinus side wall (levels B and D) during the pulse cycle; the direction 0 = 0 agrees with the direction of the internal axis. 0, CAR 1; 0, CAR 2; 0, CAR 3; L?, CAR 4. a, B (side); b, D (side)

514

J. Biomed. Eng. 1991, Vol. 13, November

Carotidflow phenomena: K. Perktold et al.

REFERENCES
BK, Mabon RF, Giddens DP. Steady flow in a model of the human carotid bifurcation. Part I - Flow visualization. JBiomechanics 1982; 15: 34!)-62. L. Bharadvaj BK, Mabon RF, Giddens DP. Steady flow in a model of the human carotid bifurcation. Part II - Laser Doppler anemometer measurements. J Biomechanics 19X2; 15: 363-78. 3. Ku DN, Giddens DP. Pulsatile flow in a model carotid bifurcation. Arteriosclerosis 1983; 3: 3 I-9. .4. Zarins CK, Giddens DP, Bharadvaj BK. Sottiura VS, Mabon RF, Glagov S. Carotid bifurcation atherosclerosis. Circ Res 1983; 53: 502-14. 5. Motomaiya M, Karino T. Flow patterns in the human carotid artery bifurcation. Stroke 19X-l; 15: 50-X. 6. Reneman RS, van Merode T, Hide P, Hoeks APG. Flow velocity patterns in and distensibility of the carotid artery bulb in subjects of various ages. Circulation l!lH5; 71: .500!I. 7. Ku DN. Giddens DP, Zarins CK, Glagov S. Pulsatile flow and atherosclerosis in the human carotid bifurcation. Arteriosclerosis 198.5;5: 293-302. 8. Ku DN, Giddens DP. Laser Doppler anemometer measurements of pulsatile flow in a model carotid bifurcation. J Biomerhanics l!)X7; 20: 407-2 1. !I. Perktold K, Hilbert D. Numerical simulation of pulsatile flow in a carotid bifurcation model. J Biomed Eng l!lX6; 8: I!&!). 10. Rindt CCM. Analysis of the three-dimensionaljowfield in the carotid artery bifurcation. IPhD Thesis]. Eindhoven University of Technoiogv 19X!). I I. Rindt CCM, van.Steenhoven AA, Janssen JD, Reneman KS, Segal A. A numerical analysis of steady flow in a 3Dmodel of the carotid artery bifurcation. J Biomechanirs 1lOO. .\ , 23.. 461-73. 12. Keuderink PJ, Willems PJB, Schreurs PJG and van Steenhoven AA. Fluid flow through distensible models of the carotid artery bifurcation. Proc 2nd Int Symp Biofluid Mechanics and Biorheology, Munich l!jH!l; 455-62. 1:-i. Van de Vosse FN, van Steenhoven AA, Janssen JD and Reneman KS. A two-dimensional numerical analysis of unsteady flow in the carotid artery bifurcation. Biorheology l!l!)O; 27: lti%H!). 14. Perktold K, Resch M. Numerical flow studies in human carotid bifurcations: Basic discussion of the geometric factor in atherogenesis. J Biomed Eng l!l!lO; 12: 11 l-23. 15. Perktold K, Kesch M and Peter R. Three-dimensional numerical analysis of pulsatile flow and wall shear stress in the carotid artery bifurcation. J Biomech I!)91 ; 24: Xl!)-LO.
I. Bharadvaj

16. Perktold K, Resch M and Florian H. Pulsatile nonNewtonian flow characteristics in a three-dimensional human carotid bifurcation model. ASME J Biomech Eng l!)!U; 114. 17. Liepsch D, Moravec S. Pulsatile flow of non-Newtonian fluid in distensible models of human arteries. Biorheologr 19x4; 21: .57 l-86. IX. Perktold K, Peter R and Resch M. Pulsatile nonNewtonian blood flow simulation through a bifurcation with an aneurysm. Biorheology I!%!); 26: 101 l-30. l!). Perktold K. On numerical simulation of threedimensional physiological flow problems. Ber Math Stat Sektion, ForschungsgesJoanneurn, Graz, 1!)87; 280. 20. Hilbert D. An efficient Navier-Stokes solver and its application to fluid flow in elastic tubes. In: Greenspan D, Rozsa P, eds. Numerical Methods, Coil. Sot. J Bolyai 50, Amsterdam: North Holland, l!*X7; -4X1-~11. 2 1. Hilbert D. Ein Finite Elementr Aufspaltungsve~ahren our numerischen LDsung der Navier-Stokes Gleichungen und seine Anwendung auf die Stramung in Rohren mit elastischen Wanden. (Thesis]. Technical University Graz, Austria, 1087. 22. Perktold K, Nerem RM, Peter RO. A numerical calculation of flow in a curved tube model of the left main coronary artery. J Biomechanics l!l!l I ; 24: 175-X9. Xi. Van de Voss FN, van Steenhoven AA, Segal A,Janssen JD. A finite element analysis of the steady laminar entrance flow in a !I() curved tube. Int J Num Meth Fluia!s l!lXl; 9: L7.5-x7. 2-C. Perktold K, Resch M. Numerische Losung der NavierStokes Gleichungen fur verallgemeinerte Newtonsche Fluide. ZAMX4 - Z angew Math Mech 1!)91; 71: T-llti-18. 2.5. Olson DE, Snyder B. The upstream scale of flow development in curved circular pipes. ./ Fluid Mech 1985; 150: 1X-58. 26. DeSyo D. Radiogammetric analysis of carotid bifurcation: Hemodynamic-atherogenetic repercussions on surgical practice. Proc. 2nd Int. Symp. Biofluid Mechanics and Biorheology, Munich l!l#); 71-2. 27. Caro CG, Fitz-Gerald JM, Schroter KC. Atheroma and arterial wall shear: Observation, correlation and proposal of a shear dependent mass transfer mechanism of atherogenesis. Pro< Roy Sttc Land 1117 1; 177: IO!)-59 28. Friedman MH, Deters OJ, Mark FF, Bargeson CB, Hutchins GM. Arterial geometry affects hemodynamics a potential risk factor for atherosclerosis. Atherosclerosis 1083; 46: 22.5-31. Y!). Mark FF, Bargeson CB, Deters OJ and Friedman MH. Variations in geometry and shear distribution in casts of human aortic bifurcations. J Biomech 1989; 22: 577x2.

J. Eiomed.

Eng. 1991. Vol. 13. November

515