figure 1.1. schematics of the heart showing the path of

Left atrium

Right atrium

Mitral valve

Left ventricle

Tricuspid valve

Right ventricle

From the lungs

To the body

To the lungs

From the body

From the body

Right ventricle

Left ventricle

Left atrium

Right atrium

Pulmonaryartery

Aortic valve

Pulmonary valve

Figure 1.1. Schematics of the heart showing the path of blood flow.

Figure 1.2. Left heart pumping cycle (2 - 540).

Systole Diastole

Aorta

Leftventricle

Left atrium

120

80

Pressure (mm Hg)

40

0

Left ventriclevolume

Figure 1.3. The twitch and tetanus mechanical responses of a muscle (12 -557).

Twitch

Unfused tetanus

Tetanus

0 1 2

Forc

e

Time (seconds)

Figure 1.4. Shear Induced Damage of Erythrocytes (64 - 500).

Crit

ical

She

ar S

tres

s (N

/m2 )

Lethal Damage

1000

100

10

1

0.001 0.01 0.1 1 10 100 1000 10000 100000

Exposure Time (seconds)

Figure 1.5. Shear Induced Damage of Platelets (the studies that provided the data for this graph were conducted at room temperature, nominally 24OC) (64 -500).

Lethal Damage

Crit

ical

She

ar S

tres

s (N

/m2 )

1000

100

10

1

0.001 0.01 0.1 1 10 100 1000


Figure 1.6. Relationship between viscosity and rate of shear at different hematocrits, Hn. For human red blood cell suspensions in plasma at 25OC. Hn=volume of erythrocytes multiplied by 0.96 put in (74 - 494).

0.2 0.5 1 5 10 50 100 500 1000

70

60

50

30

20

10

0

40

Visc

osity

(Cen

tipoi

se)

Shear Rate (1/second)

Plane A

Diaphragm

Figure 1.7. Schematic of a sac-type blood pump. The direction of fluid flow is indicated with arrows.

Diaphragm ring

Figure 1.8. Schematic of the Penn State University sac-type blood pump.

Pusher Plate

Figure 1.9. Schematic of the Penn State University electric blood pump.

Diaphragm

Figure 1.10. Schematic a pneumatic blood pump used for tracer residence time studies.

Figure 1.11. Schematic of a transparent sac-type blood pump, in which fluid flow was studied by tracer particle visualization.

Concentric sinus Triple sinus Modified triple sinus

Figure 1.12. Three cross-section schematics of blood pump outflow housing designs.

Figure 1.13. Two axisymmetric, valveless, skeletal muscle ventricles.

Inlet Outlet

central rigid support of adjustable length

Latissimus dorsi cone with muscle acting longitudinally

Firgure . Propsed method of harnessing skeletal muscle to form the SMV.

Latissimus tendon attached to rib

Central support attached to rib

Latissimus dorsimuscle cone

Inlet conduit joined to left atrium

Outlet conduit joined to descending aorta

Firgure . Proposed placement of SMV.

Crit

ical

She

ar S

tres

s (N

/m2 ) Lethal Damage

1000

100

10

1

0.001 0.01 0.1 1 10 100 1000 10000 100000


Figure 2.10. Combination of figures 1.3 and 1.4 showing the levels of shear stress and exposure time that are likely to damage blood elements.

Lower half of bearing clamp

Superpump piston rod

Cylindrical casing

Supporting rod

Hole tapped in the top of the supporting rod for attachment of blood pump proto-types

Linear bearing

Push rods

Acrylic disc

Force transducer

Figure 3.1. The mechanical system for harnessing a servo-controlled linear actuator to pump proto-type ventricles, as a complete solid model and with a section cut-away.

PressureTransducer

Adjustable resistance

Compliance

Reservoir

DownstreamFlowProbe

Figure 3.2. Schematic of the mock, closed, circulatory loop showing a ventricle in its uncompressed and compressed states at positions A and B, respectively.

UpstreamFlowProbe

A

B

Flat compression plate

Direction of fluid flow

Valves

Bubbling of polyurethane

Figure 3.3. A polyurethane ventricle made by covering a wax mould with polyurethane, dissolved in DMAC.

(a)

Figure 3.4. (a) Silicone rubber ventricle manufactured by rotating silicone rubber as it cures inside a mould (b) in an oven.(b)

Air bubbles in upper ventricle surface

Molten polymer injection

Figure 3.6. Section taken through an injection mould intended for manufacturing hemispherical pumping chambers.

Figure 3.6. Teflon dipping bowl with a screw top lid for storingpolyurethane dissolved in DMAC. The large width of the bowl allows artificial ventricle sized geometries to be dipped in thesolution.

(a)

Figure 3.7. (a) A polyurethane ventricle made by gluing two half ventricle forms together with a solution of DMAC and polyurethane. (b) The male resin former and the female mould it was cast from.

(b)

Figure 3.8. The smooth female mould, in two sections, made from acetal copolymer, and a wax mould cast from it.

Section of female mould for upper ventricle surface

Bolts

Wax form cast from the female mould

Threaded rod imbedded in wax form

Section of mould for ventricle base and quarter toroidal surface

Inlet and outlet holes

(a)

Figure 3.9. (a) The base of a polyurethane ventricle produced by coating the inside walls of a female mould with polyurethane solution. (b) A view of the ventricle showing the tear along the seam.

(b)

Tear along seam.

Extension piece

Thin section f the two part female mould

Figure 3.10. Extension for the smooth female mould inlet and outlet holes.

Orientation A Orientation B

Wax male form

Threaded rod with one end embedded in wax mould

Polyurethane drips formed at the inlet and outlet holes

(a)

Figure 3.11. (a) Two orientations in which dipped wax moulds are dried. (b) Supporting method for melting wax out of a polyurethane ventricle.

Polyurethane ventricle

(b)

Glass beaker

Supporting block of acetal copolymer

Melted wax.

Figure 3.12. A polyurethane ventricle made by dipping a high melting point wax mould into polyurethane solution. The holes in the base are trimmed by hand.

Plan view of ventricle base

14

15

16

171819

20

21

22

23

24

25

26

Plan view of ventricle upper surface

1

2

3

4

5

6

7

89

10

11

12 13

Figure 3.13. Thickness measurement positions made on each manufactured polyurethane ventricle. The measurements are numerically labeled, by order of measurement.

0

100

200

300

400

500

600

700

800

900

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26

Measurement Position

Thic

knes

s (m

icro

ns)

Mean-SDMeanMean+SD

Upper ventricle surface

Base

Quarter toroidal join between the base and upper ventricle surface

Figure 3.14. Mean thickness measurements of six 4 dip polyurethane ventricles.

(a)350

Figure 3.15. Box plots of the upper surface thickness measurements for the six 4 dip ventricles. The range is plotted together with the 25th and 75th percentiles. The filled circle represents the mean in (a) and the median in (b).

1 2 3 4 5 6

Ventricle number

100

150

200

250

300V

entri

cle

thic

knes

s (µ

m)

1 2 3 4 5 6

Ventricle number

100

150

200

250

300

350

Ven

tricl

e th

ickn

ess

(µm

)

(b)

B1P1 B1P2

B2P1 B2P2

Figure 3.16. The four geometrical variations of proto-type blood pump made from combinations of B1 (base with straight pipes) and B2 (base with angled pipes) with P1 (flat compression plate) and P2 (domed compression plate).

Figure 4.1. Signal from the force transducer and meter, on an oscilloscope, after delivering an impulse of short duration to the compression plate.

Outlet Pipe

Inlet Pipe

Region ofinterest

Region for assessing background counts

8502 counts

0 counts

Figure 4.2. Totalised image of counts from within an artificial ventricle. The region of interest defines the area within which counts are summed for the calculation of ejection fraction. The relationship between grey scale and gamma ray counts is linear.

12000

10000

8000

Cou

nts

6000

4000

2000

0

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6

Time (seconds)

Figure 4.3. Gamma ray counts during one cycle inside a region of interest, for a 19.6mm stroke length at 40bpm.

150

-100

-50Flow

Rat

e (m 0

0.3 0.5

100

l/sec

)

50

0.1 0.7 0.9 1.1 1.3 1.5

Time (seconds)

Net flow between the major pumping phases.

-150

-200

Figure 4.4 Net fluid flow rate through a ventricle. (Flow out of the ventricle is positive.)

0

20

40

60

80

100

120V

entri

cle

Vol

ume

(ml)

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6

Time (seconds)

Figure 4.5. Ventricle volume during one cycle derived from flow measurements up and down stream of a ventricle (19.6mm stroke length at 40bpm).

0

10

20

30

40

50

60

70

80

90

100

Per

cent

age

Cou

nts

2nd Pass1st Pass of tracer through ventricle

4 20 36 52 68 84 100 116 132 148 164 180 196

Time (seconds)

Figure 4.6. Clearance curve derived from gamma ray counts, within a region of interest, after the introduction of a Tc-99m bolus upstream of a ventricle (20.9mm stroke length at 30bpm).

Inlet Outlet

Region of interest

Figure 4.7. Captured image of dye inside the polyurethane ventricle at maximum ventricle volume, during a pumping cycle.

0

20

40

60

80

100

120

140

160

0 4

Ave

rage

Inte

nsity

(on

a 0-

255

grey

sca

le)

Figure 4.8. Average intensity, within a region of interest, after the introduction of an upstream dye bolus (20.9mm stroke length at 30bpm) plotted against time.

8 12 16 20 24 28 32 36 40 44 48

Time (seconds)

5

4.5

4

0

0.5

1

1.5

2

2.5

0

log(

Ave

rage

Inte

nsity

) 3.5

3

0.5 1 1.5

Dye Tracer (ml)

Figure 4.9. Logarithm of the average intensity plotted against incremental changes of, uniformly mixed, dye within the ventricle.

Per

cent

age

Dye

100

80

60

40

20

0

0 10 20 30 40 50

Time (seconds)Figure 4.10. Comparison of clearance curves produced by the pixel method , that takes account of ventricle geometry, and the average intensity method , (20.9mm stroke length at 30bpm).

Figure 4.11. Displacement of the linear actuator during one pumping cycle of the polyurethane ventricle, 18.3mm stroke length at 40bpm, at a flow rate of 1.98 litres/minute . Positive displacement values are measured when the compressionplate moves away from the ventricle, i.e. during filling. The velocity of the compression plate is derived from the displacement data. The force data is captured from a force transducer placed in series with the linear actuator piston.

-40

-20

0

20

40

60

80

0.1 0.3 0.5 0.7 0.9 1.1 1.3 1.5

Time (seconds)

Forc

e (N

ewto

ns),

Vel

ocity

(mm

/sec

),

Dis

plac

emen

t (m

m).

Force (N)Velocity (mm/sec)

Displacement (mm)

2.50

2.00

Pow

er (W

atts

)

1.50

1.00

0.50

0.000.1 0.3 0.5 0.7 0.9 1.1 1.3 1.5

Time (seconds)

Figure 4.12. Graph of power necessary to actuate the compression plate while pumping at a flow rate of 1.98 litres/minute (18.3mm stroke length at 40bpm). This graph is in phase with the plot in figure 4.11. The power curve from 0 to 0.75 seconds corresponds to compression of the ventricle during ejection. 0.75 to 1.5 seconds corresponds to ventricle filling.

Tech

netiu

m E

ject

ion

Frac

tion

(%)

Figure 4.13. Comparison of ejection fractions measured with Tc-99m and EM flow meters at different flow rates in silicone rubber (r = 0.94, gradient = 1.02) and polyurethane ventricles (r = 0.99, gradient = 0.94).

Silicone rubber ventricle

Polyurethane Ventricle

Silicone Rubber ventricle fitted linePolyurethane Ventricle Fitted Line

40

35

30

25

20

15

10

5

00 5 10 15 20 25 30 35 40

Flow Meter Ejection Fraction (%)

Rad

ionu

clid

e C

oeffi

cien

ts

-0.2

-0.18

-0.16

-0.14

-0.12

-0.1

-0.08

-0.06

-0.04

-0.02

0

-0.2

Data

Fitted line

-0.18 -0.16 -0.14 -0.12 -0.1 -0.08 -0.06 -0.04 -0.02 0

Dye Coefficients

Figure 4.14. Comparison of exponential curves, y= aebx, fitted to the Tc-99m and dye clearance curves (fitted line: r = 0.997, gradient = 0.91).

B1P1

B1P2

B2P1

B2P2

0

0.5

1

1.5

2

2.5

3

0

Flow

Rat

e (li

tres/

min

ute)

5 10 15 20 25 30 35

Stroke length of linear actuator (mm)

Figure 5.1. Calibration of the four designs for flow rate against superpump stroke length. The r values for the least squares linear fitted lines are all above 0.999.

-0.2

0

0.2

0.4

0.6

0.8

1

1.2

Vol

ume

of D

ye (m

l) = ± 1 standard deviation.

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29

Time (seconds)

Figure 5.2. The mean and standard deviation of ten clearance curves, for the same pumping chamber, at the same flow and pumping rate.

Figure 5.3. Process for converting a video frame of dye, inside a pumping chamber, to a colour map of dye concentration along each line of site through the pumping chamber.

Path length corrected image (grey scale corresponds, directly, to average concentration of dye along each line of site).Raw video image.

Background subtracted from raw video image. Colour scale applied to path length corrected image.

Figure 5.4. Path lengths as a grey scale image, with a linear scale from white (grey value of 255) to black (grey value of 0) corresponding to a scale of 0mm to 56mm path length.

Figure 5.5. A typical calibration image after the background effects have been subtracted. 0.9 ml of dye, uniformly concentrated, present in a ventricle.

0.6

0.5

Figure 5.6. Resolution curve showing the relationship between the amount of dye along the path length at a pixel (K) and the measured grey scale intensity of the pixel (Lp).

0.4

0.2

0.3

K(m

l/cm

2 )

0.1

05 55 105 155 205 255

Lp (grey scale)

Raw video image.

Figure 5.7. Enhancement of a raw video image. The inflow jet is visualized with dye.

Enhanced image.

-200

-150

-100

-50

0

0.5 1 1.5 2

Time (seconds)

Flow

Rat

e (m

l/sec

ond)

B1P1

B1P2

B2P1

B2P2

200

150

100

50

-250

Figure 5.8. Resultant flow curves for the four designs at a flow rate of 2.1 litres/minute. Outflow is posivitve and inflow negative.

200

0.5 1 1.5 2

Time (seconds)

Resultant Flow Rate (ml/second)

Force (Newtons)

B

150

100

50

0

-50

-100

-150

-200

A-250

Figure 5.9. Comparison of force and flow curves for B2P2, flow rate of 2.0 litres/minute at 30bpm.

3

B1P1

B1P2

B2P1

B2P2

2.5

2

Pow

er (W

atts

)

1.5

1

0.5

0

0 0.5 1 1.5 2

Time (seconds)

Figure 5.10. Power curves for the four designs at a flow rate of 2.0 litres/minute, pumping at 30bpm. The curves from 0 to 1 seconds represent power expended in pumping fluid by compressing an artificial ventricle, while from 1 to 2 seconds represents power imparted to the compression plate by in-flowing fluid and back pressure.

10

20

30

40

50

Per

cent

age

Dye B1P1

B1P2

B2P1

B2P2

100%

100

90

80

70

60

00 10 20 30 40 50 60

Time(seconds)

Figure 5.11. Clearance curves for the 4 geometries at a flow rate of 1.0 litres/minute and a 30 bpm pumping rate. The characteristic quick filling of the pumping chambers is evident, followed by a decreasing exponential emptying.

Per

cent

age

Dye

B1P1

B1P2

B2P1

B2P2

100%

100

10

1

0.10 5 10 15 20 25 30

Time (seconds)

Figure 5.12. Clearance curves for the 4 geometries at a flow rate of 1.0 litres/minute and a 30 bpm pumping rate.

Per

cent

age

Dye

B1P1B1P2B2P1B2P2100%

100

10

1

0.10 2 4 6 8 10 12 14 16 18

Time (seconds)

Figure 5.13. Clearance curves for the 4 geometries at a flow rate of 1.55 litres/minute and a 30 bpm pumping rate.

Per

cent

age

Dye

B1P1

B1P2

B2P1

B2P2

100%

100

10

1

0.10 2 4 6 8 10 12 14

Time (seconds)

Figure 5.14. Clearance curves, on a log scale, for the 4 geometries at a flow rate of 2.1 litres/minute and a 30 bpm pumping rate.

Figure 5.15. Clearance curves for the two P1 geometries at a flow rate of 2.9 litres/minute and a 30 bpm pumping rate.

B1P1

B2P1

100%

Per

cent

age

Dye

100

10

1

0.1

0 2 4 6 8 10

Time (seconds)

0.037 0 0.046 0

B1P1 B1P2

0.03 0 0.03 0

B2P1 B2P2

Figure 5.16a. Linear scales of dye concentration (ml of dye / cm3) against colour for the quantitative images for each pumping chamber geometry.

Figure 5.16b. Images of dye concentration, at the instant of maximum pumping chamber volume, over consecutive cycles..

B1P1 B1P2

B2P1 B2P2Cycle 1

B1P1 B1P2

B2P1 B2P2

Cycle 2

B1P1 B1P2

B2P1 B2P2

Cycle 3

B1P1 B1P2

B2P1 B2P2

Cycle 4

B1P1 B1P2

B2P1 B2P2

T0

B1P1 B1P2

B2P1 B2P2

T1

B1P1 B1P2

B2P1 B2P2

T2

B1P1 B1P2

B2P1 B2P2

T3

Figure 5.17. Consecutive video images (40 msec exposure) of a dye bolus entering each pumping chamber geometry.

Figure 5.18a. Conveyor belt (Mode a) and 100% mixing (Mode b) clearance curves at 50% ejection fraction.

010

2030

405060

7080

90100

Per

cent

age

Dye

Mode (a)Mode (b)

0 1 2 3 4 5 6 7

Cycle

15.7%

26.1%3

6.1%

48.3%

56.2%

65.4%

Figure 5.18b. Schematic of the artificial ventricle showing the regions within which apparent ejection fraction is calculated. The %standard deviation, from the mean, of each region’s apparent ejection fraction is shown, below the region number (n=5).

Figure 5.18c. Apparent ejection fractions for each design in six regions, at 2.1 litres/minute. The regions are defined and labeled in the schematic of figure 13b.

0.6

0.65

0.7

0.75

0.8

0.85

0.9

0.95

1

Region

1

Region

2

Region

3

Region

4

Region

5

Region

6

App

aren

t Eje

ctio

n Fr

actio

ns

B2P2

B2P1

B1P2

B1P1

Viewing tank

Blood pump

Figure 6.1. Three dimensional image of a proto-type blood pump suspended inside a viewing tank. Laser light is incident perpendicular to the orientation of the camera.

(a) Inlet Outlet

(b) Inlet Outlet

Figure 6.2. Fluid flow inside B2P1 recorded with single shot camera on 400ASA film at an exposure of 50msec and widest aperture (f=4.5), laser at maximum power output (5W). (a) flow during filling and (b) during ejection.

Figure 6.3. Fluid flow inside B2P1 visualized with polystyrene (above) and fluorescent (below) tracer particles, at a flow rate of 2.1 l/min at 30bpm. The fluorescent particles were made in the laboratory. The image grey scales have been reversed, so particle streaks are black rather than white.

S1S2 S3

Inlet

Outlet

Figure 6.4. The schematics of two views of the ventricles show the positions of the sections (S1, 2.. 8) as viewed in figures 6.6, 6.7, 6.8 and 6.9. (a) The 3 two-dimensional sections in which fluid flow is studied parallel to the long axis of the ventricle, selected with laser light sheets S1, S2 and S3. (b) The 5 two-dimensional sections in which fluid flow is studied, selected with laser light sheets S4, S5, S6, S7 and S8.

(a)

S1 S2S3

S4 S5 S6 S7 S8

(b)

Inlet Outlet

S4 S5 S6 S7 S8

Inlet Outlet

streamline

Figure 6.5. Key to flow pattern symbols used for figures 6.6, 6.7, 6.8 and 6.9.

small vortices (eddies)

fluid tracer particles either stationary or passing through the plane of the light sheet

slow vortex

streamlines representing slower velocity particles

fluid tracer particles either virtually stationary or passing through the plane of the light sheet, but exhibiting some slow movement.

small slow vortices

Figure 6.6. Fluid flow in B1P1 (a) at frame 10, (b) at frame 16, (c) at frame 24 and (d) at frame 40.

S1

S2

S3

S4

S5

S6

S7

S8

Inlet Outlet

S1

S2

S3

S4

S5

S6

S7

S8

Inlet Outlet

Inlet Outlet

Figure 6.6a.

Figure 6.6b.

S1

S2

S3

S4

S5

S6

S7

S8

Figure 6.6c.

S1

S2

S3

S4

S5

S6

S7

S8

Figure 6.6d.


Figure 6.7a.

Inlet Outlet

Inlet Outlet

S1

S2

S3

S4

S5

S6

S7

S8

Inlet Outlet

Figure 6.7b.

S1

S2

S3

S4

S5

S6

S7

S8

S1

S2

S3

S4

S5

S6

S7

S8

Figure 6.7c.

S1

S2

S3

S4

S5

S6

S7

S8

Figure 6.7d.

S1

S2

S3

S4

S5

S6

S7

S8


Inlet Outlet

S1

S2

S3

S4

S5

S6

S7

S8

Inlet Outlet

Inlet Outlet

Figure 6.8a.

Figure 6.8b.

S1

S2

S3

S4

S5

S6

S7

S8

Figure 6.8c.

S1

S2

S3

S4

S5

S6

S7

S8

Figure 6.8d.

Figure 6.9b.

S1

S2

S3

S4

S5

S6

S7

S8

Inlet Outlet

Inlet Outlet

Figure 6.9a.

S1

S2

S3

S4

S5

S6

S7

S8


Inlet Outlet

S1

S2

S3

S4

S5

S6

S7

S8

Figure 6.9c.

S1

S2

S3

S4

S5

S6

S7

S8

Figure 6.9d.

CCD camera

Compliance chamber

Reservoir

Pump

Direction of flow

Line generator

Viewing tank

Light sheet

Figure 7.1. Rendered solid model of the experimental flow circuit for producing developed laminar flow in a cylindrical pipe, and capturing illuminated fluid tracer particles on video.

(b)

(a)

Figure 7.2. The steady, developed, velocity profile in a pipe calculated by fitting a quadratic function to velocity measurements made by measuring (a) the distance between dots and (b) the length of streaks.

0

Vel

ocity

(mm

/sec

)

60

20

40

-20 -10 0 10 20

Radial position (mm)

-20

0

20

-20

Vel

ocity

(mm

/sec

)

60

40

-10 0 10 20

Radial position (mm)

Data points.Quadratic fitted to data points.Velocity profile calculated from equation 7.1.

Area 2Area 1

Area 3

Figure 7.3. Images of particles in design B1P1 in section S1. The light sheet has been pulsed to show particle paths as dots. The black rectangles are the areas in which fluid shear stresses have beenmeasured more accurately by zooming the camera in on them.

Area 1 Area 2

Area 3 Area 4


Area 2Area 1

Area 3 Area 4

Area 6Area 5


Area 2Area 1

Area 4Area 3

Area 5


Area 1 Area 1

Area 1


Area 2Area 1

Area 3 Area 4


Area 1 Area 2

Area 4Area 3


Uniform inlet velocity profile 40mm

upper solid boundary

Figure 7.10. Schematic of the numerical fluid model of a submerged jet impacting a plate.

30 mmExit

lower solid boundary

50mm

Axis of rotational symmetry

Figure 8.1. A potentially improved inlet/outlet pipe configuration for the proposed skeletal muscle powered VAD. The light grey outlet pipe is behind the inlet pipe.

Figure 8.2. Curved tube type pumping chamber.

Figure 8.3. A pumping chamber based upon the shape of the natural ventricle in a human heart.

figure 1.1. schematics of the heart showing the path of

Documents