Project Number: P09026 Hemodynamic Simulator II - Edge ...

skirlorangeBiotechnology

Oct 1, 2013 (4 years and 1 month ago)

130 views






Multi
-
Disciplinary Engineering Design Conference

K
ate Gleason College of Engineering

Rochester Institute of Technology

Rochester, New York 14623




Copyright © 200
8
Rochester Institute of Technology

Project Number
:
P0
9026
HEMODYNAMIC SIMULATO
R II




Alex Baxter


|



Data
Acquisition Team



Joseph Featherall

|


Lead Engineer



Mark Frisicano


|


Pump Design Team


Clarissa Gore


|


Control System Team


Liliane Pereira



|


Data Acquisition Team


Jonathan Peyton |


Pump Design Team


Gaurav Zirath


|


Team Leader




ABSTRACT


The
Hemodynamic Simulator is a modular system
that
is in intended to reproduce

the
hemodynamic
flows and pressures associated with
a
circulatory

system
. The simulator
will enable

perform
ance of

experiments and
associated with
the
fundamental
properties
associated with
typical cardiovascular
circu
latory system.
T
he system will incorporate
control

elements

that will allow generation and
measurement of arbitrary dynamic flow rates, volumes

and
pressure
s

as well as

the
evaluation of the
impact of
variations of the characteristics of system components
such as
vascular
compliance
, fluid composition and
overall topology
.


INTRODUCTION


Biomedical engineering

involves
the application
of engineering principles and techniques to
help
impr
ove the quality of liv
e
s

and overall patient
healthcare

and
consists of research and development in
areas such as bioinformatics, medical imaging, image
processing,
and
biomechanics
.


In this
project,
a Hemodynamic Simulator was
developed. The system allows for simulation of
fluid
flow


hrough a mock circulatory loop. The simulator
w
ill

be useful in research related to cardiovascular











diseases, as
it will allow

study
of
blood flow dynamics
as a function of
varying anatomical parameters such
as the heart rate

and
systolic ejection period
.


NOMENCLATURE


HR



heart rate,

measured in beats per minut
e is the
number of times a human heart contracts
, to provide
continuous supply of oxygenated blood throughout the
body.

SEP



systolic ejection
period
,

CO



cardiac output,

is the amount of blood pumped
by the heart, a ventricle in particular, in a minute.



OVERVIEW


In order to produce a functional
and aesthetically
pleasing Hemodynamic Simulator that meets the
project’s needs and requirements, the team was
divided into three distinct groups and each group was
assigned tasks specific to their role.


Proceedings of the
Multi
-
Disciplinary Engineering Design Conference




Page
2


Hemodynamic Simulator II

Project P09026

Fall ’08


Winter ‘08



The Pump Design group was responsible for
design
ing and manufacturing of all components
associated with the pumping mechanism f
or the
simulator. Some of the tasks for the group were to
identify

an actuator and servo

motor drive
, design and
manufacture

a

separate
co
mpliance

and atrial reservoir
chamber
s

and redesign

of
a
pneumatic cylinder. The
Data Acquisition group was responsible
for
identifying
suitable

sensors and transducers
to
acquire

pressures and flow
data
including any additional
signal
conditioning
circuitry
.
The

Control System group
developed
a software based

user interface and
control
system
based on LabVIEW 8.6 (National Instruments,
Austin, Texas) running under the Windows XP
operating system (Microsoft, Redmond, Washington)
on a personal computer
.




MECHANICAL CHAMBER’S

DESGIN


Previously the cardiovascular loop and pressure
generation loop were
viewed as being overly
complicated
and unable to meet the desired
performance criteria for the simulator system.
One
main design goal was to greatly simplify the system
and eliminate any un
-
needed components.
It

was
decided
to utilize
materia
l
s

that
would most closely
match the components
in

the previous system. Cast
acrylic was chosen for its
properties,
availability
and

ease of
fabrication
.


The new components t
hat were

designed are
an
aortic compliance chamber and
an
atrial reservoir.
The aortic compliance chamber was sized to
incorporate a

hydraulic column as seen in the
buffering chamber in the pressure generation
loop
.
The closest stock size of acrylic tubing was four inch
outer diameter (
OD
)

and 3.75 inch
inner diameter
(
ID
)
. To reduce the risk of a fountain like effect
at
the
in
let of the compliance chamber
,

a taper was placed
into the riser inlet from the ventricular chamber. Also
the chamber was sized to have at least a six inch tall
water column and
a four inch
(
max
imum)

air buffer.
The outlet

incorporates

one inch NPT

thread, which
was also chosen
to be the common fitting

thread

in the
overall system flow
loop. The outlet was placed as
low as possible in the chamber to keep pressure
changes from the inlet
due to its
elevation to a
minimum as preliminary calculations pointed that
elevation change was the main factor in the head loss
seen thru the
system
.



Determing the

the diameter for the atrial
reservoir
took into account the fact

that the

reservoir’s
air pocket would be pressurized to a much lower
pressure than in the aortic chamber.

As such, it was
designed
to have a surface area at least twice the size of
the aortic chamber

resulting in a cylinder with a
six
inch OD by 5.75 inch ID was chosen for the atrial
chamber. The inlet and outlet
also incorporate
one inch
NPT
threads
and were placed five inches from the
bottom of the chamber to allow for the incoming flow
to disperse energy into the body of water evenly as the
water level in the chamber was aimed at being ten to
twelve inches
above the base of the chamber.
.



Cr
itical

aspect
s

of the

design of the chambers
were

ease of manufacture and
ease of assembly

of the
overall

system
. Common
hose
fasteners were chosen
and a common top design

for both the aortic
compliance chamber an
d atrial reservoir

was

implemented

to ensure proper sealing while allowing
easy emptying and filling of the chambers.
The
chamber and reservoir
tops are easily modified


they
are essentially
disk
s

into which

any fitting style can
be
incorporated
. The top disk bolts down to a ring that
is turned on a lathe and then welded to the body
.

T
his
ring

hold
s

the nuts for the bolts to allow for one hand
tightening. The top ring also has a groove cut
into it to
allow for the placement of an o
-
ring to seal between
the top ring and top.
.



Figure
1
:
Display of taper in the inlet
from the
Ventricular chamber.


Proceedings of
the
KGCOE Multi
-
Disciplinary Engineering Design Conference



Page
3



Copyright © 200
8
Rochester Institute of Technology




Figure
2
: Cutaway of top ring and top lid
fastener
interface
.



Manufacturing:


Acrylic Manufacturing Study


During the design process, acrylic fabrication
was selected as the most effective process for
manufacturing the compliance chambers. Acrylic
material is easily allocated, inexpensive, easily
machined, and can be welded using solvent bonding
techniques. The
se material properties would allow the
team to use preformed tubing and sheet stock to
construct the final geometry of the compliance
chambers. This process cut down drastically on
material costs, machining time, and waste material.



Figure
3
: Acrylic based connector

Although the acrylic is soft and easily
machined, the builders needed to experiment with
different cutting and fabrication methods to optimize
the process and avoid some the negative idiosyncrasies
of clear plastic machining. Some of these
in
clude:
”c
razing


or miniscule cracking during welding,
overheating during machining, and solvent selection
and surface preparation.


To avoid overheating, surface melting and
creating thermally induced residual stress, all cutting
tools need to be razor
sharp. The use of dull or nicked
cutting tools will increase the risk of crazing especially
in thin cross sections and when removing large
amounts of material. If crazing does not occur
immediately, applying solvent to the surface for
welding after cuttin
g can weaken the material enough
to allow crazing to propagate. T
able 1 lists
the cutting
speeds and feeds used as guidelines for machining
operations.





Machining Acrylic ( w/ HSS)



sfm

Chip load (in)

Milling

315

0.002

Turning

600

0.01

Boring

150
-
200

0.0015


Table 1:

SFM values for various machining types and
their appropriate chip loads


In addition to these speeds
,

consideration should
be taken to reduce chip load and therefore the cutting
force near the edge of the part to prevent chipping.
Lubricant is not necessary except during high friction
operations such as
tapping
.


During the welding process a number of

different
fitment

tolerances, solvent adhesives, and
surface preparations were tested. Thinner solvents
required extensive surface preparation, and extremely
tight
tolerances
, while medium bodied adhesives
provide excellent bonding, sealing
and product clarity
with minimal surface preparation time. It is also noted
as critical that the solvent bond be held in place for
curing with light pressure so as not to squeeze the
solvent out of the joint creating a “dry” bond.
Table

2

provides the developed solvent welding guidelines.


Solvent Welding Acrylic

Su
r
face Prep.

400 grit sandpaper

Surface Cleaning


Denatured Alcohol

Tolerancing

Slip fit .003

-
.007


clearance

Solvent Adhesive

Weld
-
On #1802

Applicator

Hypo. 65 needle

Work time

15 sec.

Proceedings of the
Multi
-
Disciplinary Engineering Design Conference




Page
4


Hemodynamic Simulator II

Project P09026

Fall ’08


Winter ‘08



Supplier

www.rplastics.com

Cure time

24 h
rs
. (1 hr. till handlable)


Table 2:
Developed solvent bonding guidelines


ACTUATOR & SERVO SEL
ECTION


One major area for mechanical design was the
actuator and pump that
were
to be used in the system.
The first step in selecting the actuator was to calculate
the needed

travel

force the actuator
provi
de
.
Based on
fluid calculations
,

three
feasible
scenarios
were
determined and travel and force values were
derived
.
I
t was assumed the buffer chamber would be entirely
filled with air. This would be the worst case scenario
for the system because the system would have the
most air to compress in this scenario. If the buffering
chamber were completely filled with
air
,

the travel of
the actuator needed would be 200mm in order to
create an assumed pressure of 200mm
Hg
. The next
calculation assumed the buffering chamber would be
half
-
filled
with water. With this assumption the travel
needed to create 2
00mm Hg of assumed pressure
would be
125mm
.
The third scenario

assumed the
chamber would be filled entirely with fluid, this means
there would be minimal air to compress. This yielded
a travel of 88mm to gain the 200mm Hg of assumed
press
ure
.


The next property that was necessary to calculate
was how much force was required to compress the air
cylinder. Using a spring gauge we were able to
measure the resistance force from the seals in the air
cylinder
.
Based on
10 trials
,

the average force was
found to be
6.5

lbs. Next it was necessary to calculate
the force required to compress the air in the cylinder.
Using the pressure and the piston area along with the
assumed pressures it was shown that 11.5lbs were
needed to compre
ss the
air
. This means the total force
required is about 18lbs. The last factor was the
velocity of the actuator. Given the heart rate we
wanted to achieve
we

determined
that

the required
would

be a factor of four larger than the travel that

was calculated. THK
was

offering the most
reasonable prices for actuators and had one product
line that would meet the required
specifications
. It
was determined the system would use the VLA
-
ST
-
60
-
12
-
0250 actuator;
which has

a travel of 250
mm.
The longer travel was selected so there was extra
travel for expansion of the project as needed. The
actuator has a maximum speed of 1000mm/s which
yielded a factor of safety of 2, this was the only
actuator found that could meet our specifications

in the
velocity category.
The

actuator was rated for 45lbs of
force which is much larger than the required 18lbs
giving us a factor of safety of 2.5.


Figure
4
:
THK Linear Actuator


Through research it was determined that a
Yaskawa
motor

to power our linear actuator would be
the best
option
. This specific motor comes with a PCI
mechatrolink

car that was specifically designed to
accurately control a linear actuator using
Labview
VI’s

created by Yaskawa. The motor was a 100W
motor
with a torque output
of 1.15 N/m
.
The

motor
was then coupled to the actuator using a coupling
supplied

from THK that fit both the motor and
actuator shafts.



Figure
5
: Yaskawa Servo Motor



Figure
6
:
Yaskwawa Servo Controller



Once

the actuator and motor were selected
,
a
suitable

mounting
design
for this part of the system

was developed
. The
main
goal of the mounting
included

withstand
ing

the continuous operation of the
system

and the associated vibrations that were
produced
.
The

mounting was
also
designed
or ease of

manufacture and
serviceability of the connected parts.
An important

design consideration was
low cost.
T
o
dampen the shaking

of the overall system due
the
actuator motion
,

rubber feet were included on the
Proceedings of
the
KGCOE Multi
-
Disciplinary Engineering Design Conference



Page
5



Copyright © 200
8
Rochester Institute of Technology


bottom of the mounting plate. Urethane bushings
were made from urethane stock and press fit into
counter
-
bores, this would allow for any misalignment
in the mounting system to be dispersed into the
urethane
.


CONTR
OL SYSTEM


The Controls Team’s specific goals for this
project included the development of a system that
provided the user with accurate and precise control
over the actuator parameters. This was achieved by
integrating
LabVIEW Vis provided with the
Ya
skawa
Mecatrolink II
control electronics.
.


The speed and position of the servo motor can be
controlled both by software settings and by altering
the device parameters. Changing the device
parameters is done in the setup util
ity that is obtained
as part of the Mechatrolink software. The acceleration
of the linear actuator is controlled by the 1
st

and 2
nd

linear acceleration constants and the switching speed.
Each “move” of the motor progresses through the
sequence
seen in Figu
re
6

which

defines to
pump
motion.

The acceleration/deceleration constant
and switching speed are used to adjust the amount of
the displacements during flexion. The goal of the
program is to enable the user to control the motion of
the actuator without having to change the setup of the
mot
or. A LabVIEW© (National Instruments, TX)
interface was designed to be user
-
friendly so that the
user can easily define the stroke volume and heart rate
without needed to alter the parameters in the setup
utility. In order to achieve this goal the control
team



SIGNAL CONDITIONING


The pressure transducers used to measure the
Atrial
, aortic and venous pressures are disposable,
medical grade pressure transducers donated to the
team by Dr. Schwarz.
The pressure transducers are not
normally
commercial
ly

av
ailable for research or
industrial
use
.


These
transducers utilize strain gauge mechanism
to convert mechanical stress into an electrical signal.
According to authors
Webster and Pallas
-

Arney

in

Sensors and Signal Conditioning
, strain gages are
bas
ed on the variation of resistance of a conductor or
semiconductor when subjected to mechanical stress.
The electrical resistance of a wire with length, l and
cross sectional A, and resistivity
ρ

can be defined as:










(1)


and for small variations, the resistance for the metallic
wire ca
n

be expressed as:








(



)





(



)

(2)


where Ro is the resistance when there is no applied
stress, G is the gage fact
or, which is a constant for any
specific metal.


A Wheatstone
bridge

is known be an effective
method for measuring small resistances. This
technique

was first
proposed by S.H. Christie in 1833
and then reported by Sir Charles Wheatstone to the
Royal Society (London) in
1853
. The bridge circuit, as
shown in
Figure
7
, is

based on a feedback, in order to
adjust the value of the standard until the current
thro
ugh the current meter indicates zero

[1]
.



Figure
8
: Wheatstone bridge configuration

http://zone.ni.com/cms/images/devzone/tut/a/83a1fe69
766.gif


When the bridge circuit is in balanced condition,
resistance R
3

is:














(3)


In the equation above, R
3

is directly proportional to
corresponding changes in R
2

in order to balance the
circuit. This condition is ach
ieved independent to the
supply voltage or the current, and any possible
variations.



The sensitivity of the pressure transducers used
for the system is stated

as a
voltage
:












(4)


From the specification above, the change in output
voltage generated by the balanced bridge circuit is



for a change in one mmHg of pressure and



for increase for every volt provided to the circuit
in form of excitation.



Figure
7
: Pump Motion

Proceedings of the
Multi
-
Disciplinary Engineering Design Conference




Page
6


Hemodynamic Simulator II

Project P09026

Fall ’08


Winter ‘08



The system was designed for pressures 0


140
mmHg,
and
therefore
according to the sensitivity of
the sensors, the output voltage would range from 0


5
mV, given
a chosen
excitation voltage
of 5V. Any
variations in the output voltage will be very small

to
be captured by the NI USB 6008 data acquisition
board
.
Hence, signal conditioning must be designed in
order to amplify the signal to amplitudes that are
capture
able by the
DAQ
. Also, the signal conditioning
must eliminate any noise propagating t
hrough to the
output, before it detected and transmitted to the
computer.
The signal conditioning
for the pressure
transducers was designed for 0


200mmHg

as
indicated in
equations






Signal Conditioning Gain Computations













































For our purposes, a two stage signal conditioning is
chosen.
The

first stage is the
DC gain stage and the
second stage is identified as the low pass filter stage.
Both the stages were built and te
sted on a breadboard.
Due

to wide range of tolerances on the components
used, the
physical gain was calculated to be 475V/V
and 487V/
V
.


Stage 1: DC Gain Stage


This stage is implemented with

an INA128
instrumentation
amplifier
.
This stage
provides a buffer
for the input circuitry and more importantly reduces
the common mode noise to great extent.

This stage
was designed for a gain of 50V/V.





(







)


























(

)





(







)














Stage
2: Low Pass Filter w/ small gain


The output from the first stage is treated as an input to
the second stage of the signal conditioning circuitry.
Since the first stage is providing 50V/V of the overall
500V/V gain, o
nly 10V/V is needed out of this stage.
The low filter is implemented with a negative
feedback, based non
-
inverting op
-
amp, as shown in
Figure 2.




Figure
9
: Low Pass Filter Design

w
ith
a single supply





(







)
































(

)










(







)



























The pressures associated with human cardiovascular
system are known to be around 20Hz. And theref
ore
the low pass filtering is designed with a roll frequency
of 50Hz.












































(

)















































Proceedings of
the
KGCOE Multi
-
Disciplinary Engineering Design Conference



Page
7



Copyright © 200
8
Rochester Institute of Technology


C
1

is arbitrarily selected to be 0.33
μ
F.


































The schematic of the overall signal conditioning
circuitry

is shown in Figure
4
.


Development Procedures




Software Development

1.

Actuator control

2.

Flow and Pressure Measurements


TESTING THE
FINAL
DESIGN


HIGHLIGHTS OF FINAL
DESIGN


The current Hemodynamic Simulator is intended
to be a mechanically and
elect
r
ically robust system,
which i
s capable for generating controlled motion of
the actuator, which further provides a controlled fluid
flow through the circulatory
loop
.


Further enhancements made to the simulator
allows for easy transportability fro
m one classroom to
another. Easy filling and draining procedures have
been implanted. The new Mechatrolink II PCI board
provides
not only provides
a direct electronic link
from the servo to
computer but also to LabVIEW,
which allows for accurate and real t
ime control of the
system.

FUTURE WORK


In future senior design projects,
the primary focus
s
hould be to enhance the simulator’s controllability.
Refinement of flow compliance and resistance
mechanisms as well as final system integration, testing
and
validation

are other suggested areas of future
enhancements.


Currently, the control system can simulate for the
system for one motion profile and arbitrary motion
input is not possible. Pressure and flow readings are
accessible to the user graphically an
d numerically.
With the numerical data, a variety of manipulations
can be carried out to monitor different hemodynamic
activities in real time.
team
couldn’t finish
development of
.
It would be
very useful if

develop
a
theoretical model that would represent the dynamics of
the overall system were
develo
ped
.



Lastly,
future groups may also reduce the size of
the overall system significantly by replacing
a regular
sized PC with a

smaller single board computer, and
furthermore replace the LCD monitor with a touch
screen, which would eliminate the need f
or a dedicated
keyboard and
mouse
.


CONCLUSIONS


The final design of Hemodynamic Simulator
successfully mimics the activity of left ventricular of a
human
heart
. Although, the not all parameters are fully
controllable, the simulator has gone through
multiple

folds reduction in size, making the system portable.
Reduction in number of hose clamps in the circulatory
loop is another major customer spec that the final
design complies with.


The final produc
t ser
ves as a valuable teaching
tool, more importantly a research tool to dynamics of
blood flow. The simulator may also be used to test
operation
s of heart valve, LVAD and other assist
devices. The project may also be used to
showcases
the newly developin
g bioengineering discipline at
Rochester Institute of Technology.

ACKNOWLEDGMENTS


The team would like to express its sincerest
gratitude to those who have made invaluable
contributions to this project.
Many thanks to the
advisor
s
,
Dr. Daniel Phillips
and Dr. Karl Schwarz

for
t
he
i
r guidance and support. Additionally,
the team
would like to
express thanks

to
consultants
,

Mr. John
Wellin, Dr. Steven Day, Dr. Mark Kempski, Dr.
Jeffrey Kozak
,
who provided prompt and
important
assistance, when necessary
.



REFERENCES


[1]
Pall, Ram, and John G. Webster.
Sensors and
Signal Conditioning
. New York: Wiley
-
Interscience,
2000.