Hot-Film Anemometer Velocity Measurements of Arterial Blood Flow
in Horses
By Robert M. Nerem, John A. Rumberger, Jr., David R. Gross, Robert L. Hamlin, and Gary L. Geiger
Downloaded from http://circres.ahajournals.org/ by guest on June 14, 2017
ABSTRACT
Blood velocity measurements were carried out using a constant-temperature hot-film
anemometer system in both anesthetized and conscious horses. Catheter probes were
used to measure velocity wave forms in conscious horses and L-shaped needle probes inserted by direct vessel puncture were used to measure the profiles in regions of
the thoracic and abdominal aorta of anesthetized, open-chest horses. Both catheter and
L-shaped probes were used for coronary velocity wave-form measurements. The flow
conditions were characterized by peak Reynolds numbers of 200-10,000 and unsteadiness parameter of 2-30. These measurements indicate that in the thoracic aorta the flow
at peak systole is largely inviscid with a thin-wall boundary layer; in the abdominal aorta
the flow is more fully developed but skewed due to branching effects. Highly disturbed
flows were observed in the thoracic aorta of both anesthetized and conscious horses, but
not in the abdominal aortic region or in the coronary arteries. The results of this study indicate that the flow in the arterial system, although in many cases laminar and disturbance free, is extremely varied in character. It may be asymmetric and certainly is not
representative of fully developed, Poiseuille flow.
KEY WORDS
aortic blood
Reynolds numbers
laminar
flow
flow
• Fluid mechanical factors influence the location
of sites which show preferential development of
arterial disease. For example, both Caro et al. (1)
and Fry (2) have identified arterial wall shearing
stresses as potentially important factors in the development of atheroma. Although there is some
question about the exact role of these fluid mechanical factors (e.g., is it an effect of wall shear on
the fluid mass transport or on the properties of the
arterial wall), the possible importance of such
phenomena requires that more detailed knowledge
of the properties of arterial blood flow be
gathered.
Recently, point velocity and shear stress
measurements within the arterial system have
become possible with constant-temperature hotfilm anemometer systems. Reports on the use of
such systems in the aorta (3-9) agree on the fact
that in the ascending and upper descending portions of the aorta a clearly defined aortic wall
velocity boundary layer exists. Furthermore, for
From the Physiological Fluid Mechanics Croup, The Ohio
State University Bio-Medical Engineering Center, Columbus,
Ohio 43210.
This investigation was supported by the National Science
Foundation under Grant CK-31026 and by the Central Ohio
Heart Association.
Received July 2, 1973. Accepted for publication November
26, 1973.
Circulation Research, VoL XXXIV, February 1974
hemodynamics
disturbed flow
velocity profile
certain conditions, observations of highly disturbed
velocity wave forms have been reported and attributed to the presence of fluid mechanical turbulence in the flow (10).
These in vivo studies have been carried out
largely in animals smaller than humans, e.g., dogs
and pigs. Limitations on the size of instrumentation
have precluded the resolution of certain important
flow details in studies on these animals. Thus, there
is little, if any, quantitative experimental evidence
about the detailed nature of the flow in the aorticwall boundary layer, in the region of branching,
and downstream from such branch points. Furthermore, with animals such as dogs and pigs, the observed values of the important flow similarity parameters, e.g., Reynolds number Re = uDIv and the
unsteadiness parameter a = R(a>/j/)% (u is mean
velocity, R is vessel radius, w is fundamental frequency of the flow pulsations, D is vessel diameter,
and v is kinematic viscosity), differ from those associated with the human arterial system (11).
The limiting factor in making measurements of
this kind is instrument resolution vs. vessel size.
Large horses have aortic diameters in the range of
5 cm, brachiocephalic artery diameters near 3 cm,
intercostal and coronary artery diameters of approximately 1 cm, and iliac artery diameters in the
range of 1-2 cm. The large size of these vessels
193
194
NEREM, RUMBERQER, GROSS, HAMLIN, QEIQER
offers the opportunity of making detailed measurements in numerous vessels with a degree of instrument resolution not possible in smaller species.
Methods
Downloaded from http://circres.ahajournals.org/ by guest on June 14, 2017
Measurements of velocity wave forms in horses which
varied in weight from 136 to 410 kg were made using a
constant-temperature hot-film anemometer system with
the film held at a temperature approximately 5°C higher
than that of the blood. The system used the Disa 55DO1
anemometer and the Disa 55D10 linearizer. The principle of operation of hot-film anemometer systems and
some of the problems encountered in their use in blood
velocity measurements have been discussed in the literature (3).
Catheter probes and L-shaped probes were inserted
by direct puncture through the vessel wall. The Lshaped probes were Disa A-87 probes or probes
manufactured in our own laboratory (Fig. 1). The catheter probe was also manufactured in our laboratory (Fig.
1). The velocity probes were calibrated during each experiment using blood from the horse. At the start of each
Wire Leads
Thin Platinum Film
L-Shape Hot Film Probe
Thin Platinum Film
-Wire Leads
E
E
7 Cottieter
10
Catheter Hot Film Probe
FIGURE 1
Hot-film velocity probes used in this investigation.
experiment, a 200-ml sample of blood was taken, heparin was added to prevent coagulation, and the sample
was placed in the calibration turntable channel (12).
Here it was maintained at a temperature of 38°C with a
thermostatically controlled water heater-circulator
system. The probe was immersed in the blood, and by
controlling the speed of the turntable channel the probe
was calibrated at various known constant velocities.
Output signals were passed through the linearizer,
which was adjusted to provide a linear output
voltage-velocity relationship. Although only a steadystate flow calibration was performed, there have been
previous evaluations of the performance of hot-film
anemometer systems under unsteady flow conditions.
The L-shaped probes manufactured in our laboratory
are physically similar to those used by Seed and Wood •
(3) and thus should have the same frequency response
characteristics. For the catheter probes, similar performance characteristics exist for forward flow, i.e., flow
coming in over the tip of the catheter; however, for reverse flow the catheter probes are rather unresponsive.
The same is true of the Disa A-87 L-shaped probe. None
of the probes used in the present series of experiments
had a direction capability.
As noted earlier, L-shaped probes were inserted by
direct vessel puncture in anesthetized, open-chest
horses. Anesthesia was induced with a bolus injection of
sodium pentobarbital administered via a catheter previously inserted in the jugular vein. More sodium pentobarbital was administered as indicated. Fluids were
administered throughout the experiment in an attempt
to inhibit the effects of circulatory shock. For thoracic
measurements, the thorax was opened via a left or right
thoracotomy with resection of four to six ribs, depending
on the exposure required. Abdominal measurements
were made via a laparotomy incision from the paralumbar fossa ventral to the midline and cranial to the
xiphoid cartilage.
After insertion, the probe was aligned as nearly as
possible on a diameter normal to the vessel wall. A
micrometer device, which was attached to the probe
after puncture of the vessel wall, allowed for graduated
changes in probe position. The internal diameter of the
vessel at each site was measured by traversing the probe
from the near wall to the far wall and adding on the
probe width. The probe was then traversed across the
vessel in approximately 1-mm steps. Velocity wave
forms were thus recorded at a series of positions across
the vessel. By keying on the electrocardiogram, the instantaneous velocity profile could then be reconstructed. It should be noted that velocity wave
forms were measured repeatedly at the center-line station to check for changes in flow conditions and for
probe fouling due to fibrin deposition. If a drift in probe
output was noted, the film was wiped gently against the
vessel wall to remove any deposition on the sensor surface. In addition, the film cold resistance was repeatedly
checked to ensure the repeatability of the velocity
Circulationflweerrfi,VoL XXXIV, Ftbnuny 1974
HORSE ARTERIAL VELOCITY MEASUREMENTS
Downloaded from http://circres.ahajournals.org/ by guest on June 14, 2017
probe measurements.
Catheter probes were used to measure coronary artery flow and aortic flow (in conscious animals). For coronary artery flow measurements, the heart was exposed
and the pericardial sac opened. The catheter was then
inserted in the coronary vessel distal to the region of interest. Prior to insertion of the catheter, the vessel was
occluded; it was released by digital manipulation proximal to the area of catheterization periodically over a
5-10-minute period. This procedure was designed to
produce some level of tolerance of the myocardium to
ischemia. Without this procedure, ligation of the vessel
and the sudden onset of ischemia caused cardiac arrest
in some animals.
Once inserted, the catheter probe could be easily
moved upstream to any position desired. To determine
the actual position of the probe at each recording station, a record of how far the probe had been moved
from the point of initial insertion was kept. The final
probe position was always at the inlet of the coronary
vessel within the sinus of Valsalva. After the horse was
killed, the probe was kept in position and the heart was
excised; visual observation of the final probe position
and of the vessel geometry allowed retracing of the
other probe positions.
Measurements of velocity wave forms with L-shaped
probes were also carried out in coronary arteries. These
measurements used the earlier described direct vessel
puncture procedure; however, due to the limited vessel
size, the only velocity measurements possible were at
the approximate midstream of the vessel.
As noted above, catheter measurements of aortic
flows in conscious horses were also conducted. The
horses were tranquilized with Acepromazine maleate (2
mg/100 lb body weight). The area of the jugular furrow
was then instilled with lidocaine HC1 2%, and the jugular vein and the carotid artery were dissected and isolated. The hot-film catheter was taped to a Pieper
pressure transducer (13); the tip of the hot-film catheter
always extended at least 0.5 inches beyond the pressure
catheter. The two catheters were then passed down the
carotid artery through the ascending aorta and the aortic valve into the left ventricle. The various positions
were verified by typical pressure wave forms. The catheter was then withdrawn and, at selected positions,
velocity wave forms were recorded. In all catheter
measurements, recordings at various positions were
repeated to guard against undetected probe fouling.
Results
The present investigation involved four series of
velocity wave-form measurements. (1) Catheter
measurements were made in conscious horses at
selected sites obtained by progressively moving the
catheter from an initial position inside the left
ventricle in a distal direction to its final position in
the carotid artery. (2) Thoracic aorta velocity
Circulation Research, VoL XXXIV, February 1974
195
profile measurements were obtained in anesthetized horses using L-shaped probes inserted by
direct vessel puncture. (3) Velocity profile
measurements were obtained using L-shaped
probes inserted into the abdominal aorta and its
branch vessels. Finally, (4) velocity wave-form
measurements of coronary flow were made using
both catheter and L-shaped probes.
In the first series, aortic flow measurements
were carried out in conscious horses. These experiments are summarized in Table 1. Although the
catheter was not rigidly positioned in the vessel
cross section because of the flatness of the profile
(see Figs. 3 and 4), it was felt that the recorded
measurements were representative. Included in
Table 1 (as well as in subsequent tables) are both
the peak center-line velocity (u) and the mean
center-line velocity (u). It is possible to obtain u
directly from the measurement. However, u is an
average calculated from the center-line velocity
wave form. The probes used were not direction
sensitive, and, although a correction for reverse
flow was made for those positions at which it was
felt to be appropriate, the values of 0 must be considered to be only approximate.
Peak aortic center-line flow velocities of nearly
100 cm/sec were obtained in this series of measurements. The corresponding peak Reynolds number
Re was on the order of 10,000, and the range of
values for the unsteadiness parameter a was 10 to
30. The estimated ratio of peak velocity to mean
velocity at the center line ranged from 3.5 to 7. The
observed degree of aortic flow disturbance is noted
in Table 1. In this table, the flow has been characterized as undisturbed (U), disturbed (D), or highly
disturbed (HD) following the definitions in the
paper by Nerem and Seed (10). As may be seen,
disturbed flows were frequently encountered.
Figure 2 shows disturbed flow in a series of
measured velocity wave forms for three positions:
the left ventricle, just distal to the aortic valve, and
the carotid artery. Also shown are the measured
pressure wave forms. It should be emphasized that
these pressure measurements were uncalibrated
and that the crude wave form obtained was only
used to determine the catheter location. However,
the mean arterial blood pressure was estimated to
be between 85 and 105 torr. It should also be noted
that the tip of the hot-film catheter was located 2
inches proximal to the tip of the pressure
transducer.
The measured velocity in the left ventricle (Fig.
2) must be considered qualitative, since the orien-
196
NEREM, RUMBERGER, GROSS, HAMLIN, GEIGER
TABLE 1
Aortic Flow Measurements in Conscious Horses
Horse no.
Position
(min-1)
2R
(cm)
1
Left ventricle
Ascending aorta just distal to aortic valve
Ascending aorta 2 inches distal to
aortic valve
Arch of aorta 6 inches distal to
aortic valve near carotid artery inlet
Carotid artery 9 inches distal to
aortic valve
Carotid artery 11 inches distal to
aortic valve
Left ventricle
At aortic valve
Ascending aorta
Arch of aorta near inlet to carotid
artery
Carotid artery near incision
Left ventricle
At aortic valve
6 inches distal to aortic valve near
inlet to carotid artery
Carotid artery
76
78
3.8
83
HR
2
Downloaded from http://circres.ahajournals.org/ by guest on June 14, 2017
3
0
u/u
a
Re
87
3.8
29.0
9460
HD
HD
3.6
62
3.4
28.4
6380
D
85
3.0
33
4.0
23.9
2830
D
93
2.1
16
3.7
17.5
960
U
79
1.8
2.8
13.8
437
U
67
33
41
40
3.0
2.6
2.5
6.6
5.9
5.6
14.8
14.4
13.7
7130
3490
2000
3.4
7.8
200
7.2
6.0
18.6
14.4
8700
1388
6.3
9.06
340
(cm/sec)
48
1.4
8.5
50
93
53
28
42
40
38
38
3.5
2.7
5.0
70
87
18
38
1.7
7
Character of flow
HD
HD
D
D
U
HD
HD
U
u
HR = heart rate, R = radius, u= peak center-line velocity, u =• mean center-line velocity, a = unsteadiness parameter, Re •» peak
Reynolds number, HD -highly disturbed flow, D -disturbed flow, and U -undisturbed flow.
Catheter within Ventricle
15-
o-
Catheter I" Distal to Aortic Valve
Catheter in Carotid Artery
FIGURE 2
Hot-film catheter velocity probe recordings in a conscious horse
(no. 1) as the catheter is withdrawn from the left ventricle out
into tlie carotid artery. P — pressure wave form, V - velocity
wave form, T = time marks.
tation of the probe to the flow and thus the calibration are unknown. However, the presence and the
associated intensity of the high-frequency fluctuations are striking. These fluctuations are associated
with the filling process of the left ventricle. At a
position 1 inch distal to the aortic valve, the flow in
the aorta is still highly disturbed; however, in the
carotid artery the flow appears to be smooth and
undisturbed, and the peak velocity is considerably
lower, being on the order of 20 cm/sec.
The experiments in which thoracic aorta
velocity measurements were obtained using Lshaped probes are summarized in Table 2. As noted
above, these measurements were performed in
open-chest, anesthetized horses in which the probe
was inserted by direct vessel puncture. Both the
Disa A-87 and the probes manufactured in our
laboratory were used. The location of each of the
measurements is indicated in Table 2. Because-of
access limitations, the measurements were carried
out in a plane extending through the center of the
vessel and perpendicular to the plane of curvature
of the aorta. In this series of measurements, peak
aortic flow velocities of 90 cm/sec were seen with a
Circulation Research, Vol. XXXIV, February 1974
197
HORSE ARTERIAL VELOCITY MEASUREMENTS
TABLE 2
Flow Measurements in the Thoracic Aorta of Anesthetized Horses
Horse no.
Probe
4
Disa A-87
5
Disa A-87
6'
Catheter
L-shaped
Downloaded from http://circres.ahajournals.org/ by guest on June 14, 2017
L-shaped
L-shaped
Position
Descending aorta 7 inches
distal to aortic arch
Descending aorta 4 inches
distal to aortic arch
Descending aorta 5 inches
distal to aortic arch
Descending aorta 11 inches distal to aortic
arch
Descending aorta 15 inches distal to aortic
arch
Descending thoracic aorta
4 inches distal to aortic
arch
Descending thoracic aorta
5 inches distal to aortic
arch
Descending thoracic aorta
6 inches distal to aortic
arch
HR
(mlrr 1 )
2R
(cm)
u
(cm/sec)
u/u
96
3.05
22
AA
70
2.69
33
45
2.41
51
Re
Character of flow
25.8
1920
D
5.3
19.4
2536
D
22
4.0
13.9
1514
D
2.08
70
3.0
12.8
4160
D
59
1.40
96
2.8
3840
D
45
2.5
33
3.8
14.5
2357
80
2.6
23
2.4
20.1
1709
51
2.5
33
3.5
15.44
2357
9.29
D
See Table 1 for abbreviations.
"Horse was in shock.
corresponding Reynolds number of 4,000 based on
diameter. The range of values of the unsteadiness
parameter was from 10 to 25. The estimated ratio
of peak velocity to mean velocity ranged from 2.5
to 5.5. This finding is reasonably consistent with the
measurements in Table 1 and with measurements
in dogs (14).
In selected horses complete velocity profiles
were obtained as a function of temporal position in
the cardiac cycle. Representative measurements of
such a series of velocity profiles in the thoracic aorta are shown in Figures 3 and 4. These profiles are
based on averaging over ten cardiac cycles at each
station across the vessel lumen. As the probe is
moved sequentially across the vessel, the time of a
measured velocity during the cardiac cycle is obtained by keying on the peak of the R wave in the
electrocardiogram. The finite width of the probe
prevented any measurements in the immediate
region of the near wall. As a result, it was not always possible to obtain measurements in the nearwall boundary layer. In Figure 3 only measurements in the far-wall boundary layer are included.
Thus, in this figure the apparent asymmetry is due
to the absence of any near-wall boundary layer
measurements. In Figure 4, portions of both the
Circulation Research. Vol. XXXIV, February 1974
near- and far-wall boundary layers are evident, and
no noticeable skewing is present. These measurements were made distal to the aortic arch and not
cm /sec
Time
A
Time B
30
30JL»
20
16
0.8
16
0.8
0.8
O
0.8
16
Distance from <£_ cm
cm/sec
Time
C
30
2 0
I0
Time
16
0.8
0
0.8
16
Distance from ^ cm
FIGURE 3
Thoracic aorta velocity profile in an anesthetized horse (no. 5)
at various times during the cardiac cycle as indicated. Measurements were performed in the plane orthogonal to that of aortic
curvature.
198
NEREM, RUMBERGER, GROSS, HAMLIN, GEIGER
Time
B
cm/sec
29
cm/sec
cm /sec
[50
Time
B
• * •
i
ioo
K X > .
• * •••«
50
1
1.0
0.6
0.2
0.2
0.6
1.0
cm/sec
Time
C
30
1.0
0.6
0.2
02
Time
cm/sec
a
4°
0.6
0.2
10
0
0.2
06
10
10
0.6 "02
1
1
0.2
0.4
02
06
Time
Too
•
J L
50
0.2
0
Time
0
•
02
0.2
cm /sec
20
0.6
04
Time
c •
Time
10
0.4
•
To
Distance from £ cm
•
04
1
1
02
I
0
Downloaded from http://circres.ahajournals.org/ by guest on June 14, 2017
in the plane of curvature. Thus the absence of any
skewing of the profile was not unexpected.
The third series of measurements was carried out
in the region of the abdominal aorta using Lshaped velocity probes inserted by direct vessel
puncture. These measurements are summarized in
Table 3. Velocity profiles were obtained in the abdominal aorta proximal to the mesenteric artery,
between the mesenteric and renal arteries, and distal to the point at which the renal artery branches
off. Profiles were also measured in the terminal aorta and the internal "and external iliac arteries. In the
horses used, the distal aorta usually gave off paired
external iliac arteries which bifurcated within 1-3
cm into the internal iliac arteries. In addition,
limited velocity wave-form measurements were
obtained in the mesenteric and renal arteries themselves.
Profile measurements in the abdominal aorta
were also obtained by sequentially moving the
velocity probe across the vessel, keying on the
electrocardiogram to reconstruct the velocity
profile. Velocity profiles measured in this manner
are shown in Figures 5-8. It should be emphasized
that many more profiles were measured and that
Figures 5-8 have only been selected as representative examples. Included in each figure is a centerline velocity wave form; the time corresponding to
the associated velocity profiles is indicated on each
wave form. As is obvious, these profiles are
different in character with respect to each other
and to the thoracic aorta profiles of Figures 3 and 4.
I
0.4
04
0.2
0
0.2
04
Distance from <f_ cm
FIGURE 4
Thoracic aorta velocity profile in an anesthetized horse (no. 7)
at various times during the cardiac cycle as indicated. Measurements were performed in the plane orthogonal to that of aortic
curvature.
I
02
FIGURES
Abdominal aorta velocity profile in an anesthetized horse (no.
14) at various times during the cardiac cycle as indicated.
Measurements were performed proximal to the branching of the
mesenteric artery in the plane orthogonal to the plane of
branching.
These differences will be discussed in the next section; however, it is obvious that the flow is in many
cases complex and certainly not indicative of fully
developed Poiseuille flow. From the center-line
velocity wave form, the peak Reynolds number Re
(based on the peak center-line velocity u), the ratio
of peak center-line velocity to mean center-line
Time
B
•
20
Cross - Sectional View
Abdominal Aorta
T)4
0
60
20
04
0
0.4
cm/sec
04
cm sec
Time
D
Time
C
60
6
g
4 0
•
t
Time
•
20
20
1
» •
• •
i
1
Distance from C cm
FIGURE 8
Abdominal aorta velocity profile in an anesthetized horse (no.
10) at various times during the cardiac cycle as indicated.
Measurements were performed distal to the branching of the
renal artery and proximal to the iliac bifurcation in the plane
indicated.
Circulation Research, Vol. XXXIV, February 1974
199
HORSE ARTERIAL VELOCITY MEASUREMENTS
TABLES
Flow Measurements in the Abdominal Aorta and Its Branching Vessels in Anestlietized Horses
Position
Horse no.
10
11
12°
13
Downloaded from http://circres.ahajournals.org/ by guest on June 14, 2017
14
HR
(min-i)
Abdominal aorta proximal to iliac 92
trifurcation
Internal iliac distal toflowdivider 47
Abdominal aorta 1 inch distal to 63
renal artery
Abdominal aorta proximal to 68
renal inlet
External iliac distal to flow 115
divider
Terminal aorta distal to flow 115
divider
Internal iliac distal toflowdivider 41
Abdominal aorta 2 inches distal
40
to renal inlet
Abdominal aorta 1 inch proximal 40
to renal artery inlet
Renal artery
77
External iliac distal to flow 72
divider
Abdominal aorta distal to renal 81
2R
(cm)
u
(cm/sec)
u/u
1.25
63
1.8
0.78
1.30
53
15.4
7.1
4.5
1.20
23
4.5
1.7
16
3.0
15.8
4.2
16.1
1.74
1.6
0.79
1.42
9
23
1.54
26
0.3
0.78
120
10.8
0.91
84.5
87
Re
Character of flow
2250
U
4.62
8.92
1181
572
U
8.55
788
U
111
U
10.4
79.5
U
U
4.37
7.76
202
925
8.42
1140
1.25
4.9
2.27
5.7
1028
240
30
5.0
7.08
780
0.4
0.8
188
118
3.1
4.2
3.17
6.45
2148
2697
54.5
0.4
29
4.7
2.55
331
u
u
u
53.5
0.53
14
3.0
3.35
212
u
51
0.79
3
4.0
4.84
50
1.4
15.2
3.9
8.55
1.4
2.0
U
u
u
u
u
u
artery
15
Mesenteric artery just off aorta
Abdominal aorta proximal to
mesenteric artery
External iliac distal to flow
divider
Terminal aorta distal to flow
divider
Abdominal aorta distal to renal
artery
Abdominal aorta proximal to
renal artery
67.2
608
u
u
See Table 1 for abbreviations.
'Animal was anoxic throughout experiment.
velocity u/u, and the unsteadiness parameter a can
be calculated. This information is included in Table
3; however, compared with those for the thoracic
aorta, the peak Reynolds numbers are lower, the
unsteadiness parameter values are smaller, and the
value of u/u is in some cases considerably less.
The final series of measurements was carried out
in the right and left coronary arteries with both the
catheter and the L-shaped probe. These measurements are summarized in Table 4, and typical wave
forms are shown in Figure 9. The apparent
difference in wave forms may be partially explained by the difference in heart rate and the condition of the preparation at the time of measurement. Maximum peak velocities of 50-60 cm/sec
were measured. The corresponding maximum
Circulation Research, VoL XXXIV, February 1974
value of the peak Reynolds number was approximately 1,500, and the range of values for the
unsteadiness parameter was 1.5 to 4.0. The estimated ratio of peak velocity to mean velocity
ranged from 1.5 to 3.
In Figure 9a a double wave form per cardiac cycle is apparent; it is believed that both wave forms
correspond to forward flow (the catheter probe is
not direction sensitive). The first hump is associated
with systole and the second with diastole. Assuming
that both humps correspond to forward flow, then a
mean velocity can be calculated. The resulting
values are indicated in Table 4. Of course, these
measurements must only be considered as representative, since the exact position of the probe in
the vessel cross section was not known.
200
NEREM, RUMBERGER, GROSS, HAMLIN, GEIGER
cm/sec
cm /sec
60
60
Time
B
O
a>
E
o
in
03
0.2
01
0
OJ
02
03
03
02
0.1
OJ
0.2
03
cm/sec
S|0
Time
C
Time
..
03
t
t
02
0.1
01
0.2
03
Distance from <t , cm
Downloaded from http://circres.ahajournals.org/ by guest on June 14, 2017
FIGURE 7
External iliac artery velocity profile in an anesthetized horse
(no. 11) at various times during the cardiac cycle as indicated.
Measurements were performed in the plane orthogonal to the
plane of branching.
Discussion
From the measurements performed in this series
of in vivo experiments in horses, a more complete
picture of the general nature of the blood flow in
the aorta and the larger arteries is available. This
picture obviously is partially due to the results of
previous investigations. However, as noted in the
introduction, the large size of the vessels of a horse
cm/sec
6
cm/sec
6
u
<o
in
E
o
OJ
IO 1
^Velocity
AJ^A^^
EKG
(b)
FIGURE 9
Recordings of coronary artery velocity wave forms in
anesthetized horses, a: Catheter probe in right main coronary
artery, b: L-shaped probe in left anterior descending coronary
artery.
Time
Time
A
B
4
•
2
•
4
0.8
04
0
0.4
0.8
3.8
0.4
04
08
cm/sec
6|
Time
C
Time
08
0.4
0
0.4
08
Distance from £ , cm
FIGURE 8
Terminal aorta velocity profile in an anesthetized horse (no. 12)
at various times during the cardiac cycle as indicated. Measurements were performed in the plane orthogonal to the plane of
branching.
provided an access and a resolution not afforded by
the use of smaller animals, and thus the present
measurements provide some unique results which
in some cases are a verification of what before
could only be suspected.
Starting with the left ventricle and itsfillingprocess, it is clear from Figure 2 that the flow in this
chamber has a reasonably high velocity and considerable high-frequency content. As noted in the
previous section, the orientation of the catheter
probe and the flow within the ventricle is not
known in these measurements in conscious horses.
Since differences in probe orientation can cause a
change in the calibration characteristics of as much
as a factor of two, the peak velocities measured
may range from a low of 30 cm/sec to the indicated
value in Figure 2 of approximately 60 cm/sec.
The high-frequency content is of interest,
Circulation Research, Vol. XXXIV, February 1974
201
HORSE ARTERIAL VELOCITY MEASUREMENTS
TABLE 4
Coronary Flow Measurements in Anesthetized Horses
Horse no.
16
17
Probe
Catheter
Catheter
Downloaded from http://circres.ahajournals.org/ by guest on June 14, 2017
12
L-shaped
13
L-shaped
15
L-shaped
Position
HR
2R
u
(min-1)
(cm)
(cm/sec)
u/u
a
Re
0.70
53
1.4
4.36
1060
U
0.80
60
1.8
4.89
1371
U
0.90
52
2.6
6.36
1337
u
0.90
65
2.0
6.91
1671
u
0.70
0.70
0.70
17
11
64
3.1
5.48
4.76
7.58
340
220
u
u
u
0.6
32
2.4
5.6
548
u
0.5
63
3.0
2.80
900
u
0.4
45
1.6
2.55
512
u
Descending branch of 52
right main coronary artery
Upstream of descending 50
branch of right main
coronary artery
0.5 inches within right 67
main coronary artery
downstream of sinus of
Valsalva
3 inches distal to sinus in 79
right main coronary artery
8 inches distal to sinus in 82
descending branch of 62
right main coronary ar- 157
tery near incision"
Left anterior descending 117
coronary artery
Left anterior descending 42
coronary artery
Left anterior descending 54.5
coronary artery
1280
Character of flow
See Table 1 for abbreviations.
"Drugs administered to produce three different flow conditions.
because these disturbances are undoubtedly convected into the aorta itself and are thus important
to an understanding of the nature of and the conditions necessary for the presence of highly disturbed
aortic flows. Observations of such flows have been
reported by several investigators (4-6) and have
been considered in detail (10). They have also been
observed in the present study, as is illustrated in
Figure 2 by the velocity wave-form measurement 1
inch distal to the aortic valve.
In addition to this possible presence of high-frequency disturbances, the most striking feature of
the flow in the thoracic aorta is the flatness of the
profile in the center region of the vessel. Such a
profile is illustrated for the aortic arch in Figures 3
and 4, where the flow can be seen to be characterized by an inviscid core and a thin-wall boundary layer region to which viscous effects are by in
large confined. Based on steady-state pipe flow
data and for the Reynolds numbers characterizing
the mean aortic flow (15), the entrance length, i.e.,
the distance required for a fully developed viscous
flow to be attained, would be approximately 30-40
tube diameters. The thoracic aorta measurements
Circulation Retearch, VoL XXXIV. February 1974
of Figures 3 and 4 were performed within 15 cm of
the aortic valve (3-5 tube diameters), and thus the
presence of an inviscid core and a thin-wall boundary layer was not surprising. The nature of the
boundary layer, of course, must be a combination of
the properties due to a steady-state mean flow and
to the unsteady, pulsatile flow. In terms of unsteady
effects, for pulsatile flow in an infinite cylindrical
tube, the viscous effects in the limit of a becoming
large are confined to a thin-wall boundary layer
(16). Thus, this effect also would suggest that the
velocity profile should appear as it is shown in
Figures 3 and 4. This finding also has been suggested by the measurements in dogs (4-6).
As noted in the previous section, the measurements in Figures 3 and 4 were not performed in the
plane of aortic arch curvature, and thus no skewing
due to a curvature effect, such as found by Seed
and Wood (6), was anticipated. It should also be
noted that, although the peak velocities in Figures
3 and 4 may appear low, this fact is undoubtedly
due to the condition of the horse, e.g., the effects of
anesthesia and trauma. As may be seen in Table 2,
varying peak center-line velocities were recorded.
202
Downloaded from http://circres.ahajournals.org/ by guest on June 14, 2017
By comparing with Table 1, it can be seen that the
higher values correspond to the conditions in conscious horses. It is felt, however, that, in terms of
the profiles shown in Figures 4 and-5 and in subsequent figures, the important thing is the qualitative
nature of the velocity profiles and not the exact
quantitative value of the velocity.
A similar flat profile to that in the thoracic aorta
region is seen in Figure 5 for theflowjust proximal
to the point at which the mesenteric artery
branches off of the abdominal aorta. The high
velocity at this point is in marked contrast with that
observed in the aorta of the same horse distal to the
point at which the renal artery branches off. In this
particular animal (horse 13), the mesenteric and
renal arteries branched almost immediately adjacent to one another; the large diversion of blood
into these branches thus resulted in low velocities
at distal points in the abdominal aorta and also in
the iliac arteries. Although the velocity profile in
the mesenteric artery could not be resolved
because of the small vessel size, an approximate
midstream velocity wave form was recorded; this
wave form is compared in Figure 10 with centerline wave forms for the aorta, both proximal to the
mesenteric artery and distal to the renal artery,
and for the external iliac artery. As is evident, the
higher velocities are observed proximal to the
point at which the mesenteric artery branches off
and in the mesenteric artery itself.
It has already been noted that distal to the point
at which the renal artery branches off the
velocities are sharply reduced from those further
upstream in the aorta. However, of more interest is
the fact that the profiles in this region are no longer
necessarily flat but demonstrate a more fully developed viscous flow character. In addition, there is a
skewing of the profile which is illustrated in Figure
6 and is believed to be associated with the effects
of the branching off of the mesenteric and renal arteries. The profile shown in Figure 6 was obtained
in a plane such that the near wall was closer to the
side from which the renal artery branched oft than
it was to the far wall side, and the skewing is
believed to be associated with the development of
a new boundary layer distal to the renal artery
branch point. Thus, though the velocities are
relatively low in this region, there appear to be
marked profile characteristics associated with
branching effects.
Within 10-20 cm (depending on the size of the
horse) of the point at which the renal artery
branches off, the abdominal aorta then branches
NEREM, RUMBERGER, GROSS, HAMLIN, GEIGER
120
Abdominal Aorta
Proximal to Mesenteric
Artery
o
190
Mesenteric Artery
30
o
o
Abdominal Aorta
Distal to Renal
Artery
io
External Iliac Artery
o
Time
FIGURE 10
Recordings of center-line velocity wave forms at various positions along the abdominal aorta and in the branching vessels for
an anesthetized horse (no. 14).
into the iliac arteries. As illustrated in Figures 7 and
8, the profiles in the external iliac artery and in the
terminal aorta just prior to the branching off of the
internal iliac arteries are characterized by the same
low velocities seen in the abdominal aorta just
proximal and by profiles indicative of a somewhat
fully developed viscous flow. The terminal aorta
profile is of particular interest because the velocity
on the center line is much lower than that in the
region further out near either wall. Thisfindingwas
observed in several horses, but it was most noticeable in horse 11 in which, because of the high heart
rate, the velocities and thus the Reynolds numbers
were reduced. Comparing the terminal aorta
profile measured in horse 12 with that measured in
horse 11 (Fig. 8), it appeared that in the former the
dip in the profile in the center was not as marked.
Horse 11 also had higher Reynolds numbers, and
the suggestion of a decrease in the center-line dip
with increasing Reynolds number and the general
shape of these profiles is in agreement with previously reported measurements in branching air
flows (17).
Circulation Research, Vol. XXXIV, February 1974
203
HORSE ARTERIAL VELOCITY MEASUREMENTS
Downloaded from http://circres.ahajournals.org/ by guest on June 14, 2017
It was noted previously that disturbed flows
were observed in a number of experiments. These
observations were in the thoracic aorta of both
conscious and anesthetized horses. High-frequency
disturbances were not observed in the coronary arteries or the abdominal aorta and its branch vessels.
However, no measurements in conscious horses
were performed in these vessels, and thus no
definite conclusions can be stated. Low-frequency
disturbances on the order of 5-10 Hz were observed in the abdominal aorta distal to the renal artery branch point in certain horses. This flow appears to be laminar with a low-frequency oscillation which may be the result of complexities associated with the bifurcation at the renal artery branch
point, e.g., a shed vortex, and thus not indicative of
fluid mechanical turbulence.
As is apparent from the data in Tables 1-4, the
flow conditions corresponding to the present
measurements cover a wide range. Considering the
two basic fluid mechanical parameters —peak
Reynolds number Re and unsteadiness parameter
a —the range of conditions encountered in the
present experiments includes peak Reynolds numbers ranging from 200 to 10,000 and unsteadiness
parameter values ranging from 2 to 30. Associated
with these widely different conditions were
markedly different velocity wave forms, velocity
profiles, and flow disturbance characteristics. It is
obvious from these results that the flow in the arterial system, although in many cases laminar and
disturbance free, is extremely complex in
character. Further studies will provide additional
insight into the details of these fluid mechanical
characteristics, and it appears that the horse,
because of the size of its vessels, offers itself as an
excellent experimental animal for such studies.
2.
3.
4.
Atheroma and arterial wall shear: Observation, correlation and proposal of a shear dependent mass transfer
Ctrcuiatim RtttcnK VoL XXXIV, February 1974
SCHULTZ, D.L., TUNSTALL-PEDOE, D.S., LEE, G.DEJ., GUNNING. A.J., AND BELLHOUSE, B.J.: Velocity distribution and
transition in the arterial system. In Circulatory and
Respiratory Mass Transport CIBA Foundation Symposium, edited by G.E.W. Wolstenholme and J. Knight.
London, Churchill, 1969, pp 172-199.
5.
LINC, S.C., ATABEK. H.B., FRY. D.L., PATEL, D.J., AND JANICKI.
J.S.: Application of heated film velocity and shear
probes to hemodynamic studies. Circ Res 23:789-801,
1968.
6.
SEED, W.A., AND WOOD. N.B.: Velocity patterns in the aorta.
7.
SEED. W.A., AND WOOD, N.B.: Use of a hot-film probe for
Cardiovasc Res 5:319-330, 1971.
cardiovascular studies. J Sci Instr 3:377-384, 1970.
8. TUNSTALL-PEDOE, D.S.: Velocity distribution of blood flow
in major arteries of animal and man. Ph.D. Thesis,
University of Oxford, 1970.
9.
REUBEN, S.R., SWADUNG, J.P., AND LEE, G.DEJ.: Velocity
profiles in the main pulmonary artery of dogs and man
measured with a thin film resistance anemometer. Circ
Res 27:995-1001, 1970.
10.
NEREM.R.M., AND SEED. W.A.: In vivo study of the nature of
11.
aortic flow disturbances. Cardiovasc Res 6:1-14, 1972.
MCDONALD, D.A.: Blood Flow in Arteries. London, Arnold,
1960, pp 53-90.
12.
SEED, W.A., AND WOOD. N.B.: Apparatus for calibrating
13.
14.
15.
16.
References
CARO, C.G., FITZ-GERALD, J.M., AND SCHROTER, R.C..
mechanism for atherogenesis. Proc R Soc Lond [Biol]
117:109-159,1971.
FRY, D.L.: Acute vascular endothelial changes associated
with increased blood velocity gradients. Circ Res
22:165-197, 1968.
SEED. W.A., AND WOOD. N.B.: Development and evaluation
of a hot-film velocity probe for cardiovascular studies.
Cardiovasc Res 4:253-263, 1970.
17.
velocity probes in liquids. J Sci Instr 2:896-898, 1969.
PlEPER, H.: Catheter-tip manometer for measuring blood
pressure during G-changes. J AppI Physiol 22:352-353,
1967.
FALSETTI, H.L., KISER. K.M., FRANCIS. G.P., AND BEL-MORE.
E.R.: Sequential velocity development in the ascending
and descending aorta of the dog. Circ Res 21:328-338,
1972.
GOLDSTEIN, S.: Modern Developments in Fluid Dynamics,
vol. 1. Oxford, Clarendon, 1938, p 301.
NEHEM. R.M.: Fluid mechanical aspects of blood flow. In
Proceedings of the 8th International Symposium on
Space Technology and Science, Tokyo, Japan, 1969, pp
1201-1216.
SCHROTEB, R.C., AND SUDLOW, M.F.: Flow patterns in
models of the human bronchial airways. Respir Physiol
7:341-355, 1969.
Hot-Film Anemometer Velocity Measurements of Arterial Blood Flow in Horses
ROBERT M. NEREM, JOHN A. RUMBERGER, Jr., DAVID R. GROSS, ROBERT L. HAMLIN
and GARY L. GEIGER
Downloaded from http://circres.ahajournals.org/ by guest on June 14, 2017
Circ Res. 1974;34:193-203
doi: 10.1161/01.RES.34.2.193
Circulation Research is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1974 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7330. Online ISSN: 1524-4571
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://circres.ahajournals.org/content/34/2/193
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published in
Circulation Research can be obtained via RightsLink, a service of the Copyright Clearance Center, not the
Editorial Office. Once the online version of the published article for which permission is being requested is
located, click Request Permissions in the middle column of the Web page under Services. Further information
about this process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Circulation Research is online at:
http://circres.ahajournals.org//subscriptions/