Anda di halaman 1dari 28

Cleveland State University

LAB REPORT 1

EXPERIMENET 1
CALIBRATION OF PERISTALTICE PUMPS

Chief Engineer: ZAHRA MUBRAK


Methodologists: Sara Alkhabaz
Experimentalist: Maryam Albulushi

1
Abstract:
In this experiment the aim was to identify the difference between the actual flow rate obtained
by pump and the flow rate that is given by the computer. Each trial was repeated three times
and the average of the three trials was compared with the values recorded by the computer. In
addition, the percentage error of flow rate for each setting was to be determined. The data
obtained was then used to do the required calculations and the graphical data; in which there
were three graphs obtained for each: actual flow rate VS pump power percentage, computer -
calculated flow rate VS pump power Percentage and the last one both flow rates VS pump
power percentage. These graphs as well as the data collected from the experiments are used to
determine the error between the actual flow rates and the flow rates that the computer is
registering.

2
Introduction:
A peristaltic pump is a type of positive displacement pump used for pumping variety of fluids. It is
usually used to pump clean or aggressive fluid, since cross contamination should not occur. In the
precision pump head there is an elastic tube that contains fluid, it is compressed by the center rotor
and some rollers. The tube Closes when the rotor turns due to the compression, which causes the fluid
to be pumped inside the tube. As the roller passes, the tube opens to its original condition, which
causes the fluid to flow through the tube (DeBello).

In this experiment, there were two pumps tested to identify the relationship between the actual and
expected flow rate (left pump and right pump). this experiment is to determine the differences
between the actual flow rate (which is obtained) and the flow rate that the computer is giving.
Thus, we will be able to find a correlation between the pumps settings and the volumetric flow rates.
The experimental error of pumps can be identified in determining the flow rates. Also, one of the
objectives of this experiment is to understand the calibration technique of the peristaltic pumps to get
the value of the flow rates.

During calibrating the pump flow, it is important to be sure that the liquid fills the tubing and

reaches its free end. Then, the measuring process is started to get precise calibration. A total of

15 runs for each pump were performed under different potentiometer settings. Moreover, the

obtained data is absolutely necessary in order to perform a calibration to the pumps. In

addition, calibration of pumps will help to have a result with maximum accuracy and precision.

3
Theory:

Data can be describing graphically by a scatter plot diagram. The linear correlation coefficient is
calculated and a best fit line is determining which is also known as the linear regression line.
After plotting the graphs, the points were following a linear pattern.
a. To calculate relationship, the residuals,

Residuals = observed predicted

b. Formulate an equation for the best line fit using the slope (m) and the y-intercept,

Y= mx+b

M= (xi-x)(yi-y)
(xi-x)2

B= y- (b*x)

c.
To calculate the present error in the flow rate

1 2
= ( ) ( )2 + ( 2 )2 ( )2

Where t is the time interval, V is the volume of liquid collected, Ev is the error in volume
collected, and Et is the error in time interval.

4
Remember that V and t are measured by you during your experiment, and Ev and Et are the
inbuilt errors in your measurements of V and t.

Materials and Method:


Armfield CEX chemical reactor services unit.
Kimble case 250 ml beaker.
Kimble case 50 ml graduated cylinder model #20039 with 0.4 tolerance.
IPhone stopwatch.

The Armfield CEX chemical reactor service unit was equipped with two Tanks that were filled with
distilled water. The Amifield used a separate peristaltic pump for each tank to move the water from a
tank into either the Kimble Chase beaker or the Kimble Chase graduated cylinder. The flow rates of the
peristaltic pumps were controlled with a potentiometer that regulated flow based on a percentage of
maximum voltage output to the Armfield peristaltic pumps.
The Armfield's estimated flow rate is monitored separately using the Armfields CEX Included software
that must be run on a Windows PC. The graduated cylinder was used to collect and
measure the volume of liquid that was pumped. The beaker was used for waste collection
for later disposal and was never used in any measurements. The IPhone Stopwatch is a
default app that records up to the hundredth of a second. The stopwatch was estimated to
have a tolerance of 1 second due to operator error.

5
First, the minimum potentiometer setting that the pump would start operating was measured for both
pumps. Second, a maximum potentiometer setting was found that the change in flow rate would be
recorded by the software for both pumps. The tow pumps were run at five different potentiometer
settings between the minimum and tnaxinnun potentiometer settings for that pump. Three readings
were taken for each of the potentiometer sittings per pump. A trial would begin when the
potentiometer was set to the proper percentage and the pump was switched on. Once flow from the
pump was consistent and no air bubbles remained in the line, the graduated cylinder would be moved
under the tube to collect the distilled water at the same time the timer is started on the stopwatch.
The theoretical flow rate (DUA in mL/s) that was displayed by the Armfield software
was recorded. Once the graduated cylinder collected enough water (10mL to 40mL) to be
accurately examined, the tube would be moved into the beaker and the pump would be
turned off at the exact same time the stopwatch was paused. After volume and time were
recorded, the stopwatch was reset and the graduated cylinder was emptied into the beaker
and dried for the next trial. After three trials, the potentiometer was changed to the next
setting. This was repeated for all trials on both pumps. The actual flow rate (Q in mL/s)
was then calculated using the collected volumes and collection times.

Estimating the volume of water collated using a graduated cylinder creates parallax error.
Removing the tube from the graduated cylinder can cause varying amounts of water to dropout
due to pressure on the tube. The graduated cylinder produces an error ( tolerance) of +- 0.4 mL.
the stopwatch was stopped by hand, varying reaction time would also create error that was
estimated to be equal to 1 second of deviation. As peristaltic pumps run well even if air is
trapped in the system. this can force air into the fluid stream. The resulting trapped air in the
fluid can falsely increase measured volume resulting in error. Running the pump for a set period
before collecting results decreased this error and any error due to the pumping of air instead of
water.

6
Results and discussion:
The aim of this experiment is to determined how much error there is between the flow rates that the
computer is generating and the actual flow rates. This was done by using two Armfield pumps (Left
Pump and Right Pump). For each pump, a total of 15 trials were done: 3 trials for each of the 5
potentiometer settings.
The volume of the water flowed into the 50.0 mL graduated cylinder was measured over a time period
at the different power outputs shown in table 1, ranging from 20.0% to 90.0% of the potential power.
The average flow rate (Q) was found for each power output by dividing the corresponding volumes by
the times elapsed. Experimental and computer results show that the flow rates increase as the
potential power increases

7
Table 1: Data for left pump.

setting % V(m/l) t(s) Q(ml/s) Q(ml/min) DUA


(ml/min)
90 75 60 1.25 75 79.9
90 75 60 1.25 75 79.7
90 75 60 1.25 75 79.3
70 54 60 0.9 54 62.1
70 54 60 0.9 54 62.1
70 54 60 0.9 54 62.1
50 33 60 0.55 33 39.4
50 33 60 0.55 33 45
50 33 60 0.55 33 45
30 14 60 0.233333333 14 27.7
30 14 60 0.233333333 14 27.7
30 14 60 0.233333333 14 27.7
20 9 60 0.15 9 18.9
20 9 60 0.15 9 18.9
20 9 60 0.15 9 18.9

8
Table 2: Error in measurements for left pump
E(set),% E(V),ml E(t),s E(Q),ml/s E(Q),ml/min relative sig fig
error
0.1 0.4 1 0.02187400 0.14449144 0.01749920 2.45595165
8 5 6 2
0.1 0.4 1 0.02187400 0.14449144 0.01749920 2.45595165
8 5 6 2
0.1 0.4 1 0.02187400 0.14449144 0.01749920 2.45595165
8 5 6 2
0.1 0.4 1 0.01641476 0.12265579 0.01823862 2.43797789
3 7 6 7
0.1 0.4 1 0.01641476 0.12265579 0.01823862 2.43797789
3 7 6 7
0.1 0.4 1 0.01641476 0.12265579 0.01823862 2.43797789
3 7 6 7
0.1 0.4 1 0.01133455 0.09597453 0.02060828 2.38492807
9 4 9 6
0.1 0.4 1 0.01133455 0.09597453 0.02060828 2.38492807
9 4 9 6
0.1 0.4 1 0.01133455 0.09597453 0.02060828 2.38492807
9 4 9 6
0.1 0.4 1 0.00771802 0.06271629 0.03307724 2.17944064
4 2 8
0.1 0.4 1 0.00771802 0.06271629 0.03307724 2.17944064
4 2 8
0.1 0.4 1 0.00771802 0.06271629 0.03307724 2.17944064
4 2 8
0.1 0.4 1 0.00712000 0.05044248 0.04746668 2.02258107
3 7 7 9
0.1 0.4 1 0.00712000 0.05044248 0.04746668 2.02258107
3 7 7 9

9
Table 3 shows LINEST and TINV data in order to get the best fit line of x and y values that are
determined (setting %. Q). the LINEST function was used it represents Q measured from predicted in
which we get Q predicted from the best line fit. Since the R^2 value is approximately 1, we assume that
the dependency of the flow rate on the power output percentage is positive and strong.
Table 3: LINEST and TINV data for Q and % setting

m 0.963414634 -13.09756098 b

st.error of m 0.020322076 1.177979442 st.erro of b

r^2 0.994248937 2.015884389 st.erro of regression

f- test overall 2247.451524 13 degrees of freedom

regression SS 9133.170732 52.82926829 residual SS

t value 0.693829304 0.877308557 Upper m

delta m 0.010418313 0.856471931 Lower m

delta b 0.603902815 1.818943466 Upper b

DUA,pred m*setting %+b 0.611137835 Lower b

10
Table 4: Q predicted values as well the upper and lower bounds (mL/min)

setting % Q(ml/min) Qpred+ deltaQ Qpred-deltaQ Q pred


90 75 75.69607742 71.52343477 73.6097561
90 75 75.69607742 71.52343477 73.6097561
90 75 75.69607742 71.52343477 73.6097561
70 54 56.1457837 52.53714313 54.34146341
70 54 56.1457837 52.53714313 54.34146341
70 54 56.1457837 52.53714313 54.34146341
50 33 36.59548998 33.55085148 35.07317073
50 33 36.59548998 33.55085148 35.07317073
50 33 36.59548998 33.55085148 35.07317073
30 14 17.04519626 14.56455984 15.80487805
30 14 17.04519626 14.56455984 15.80487805
30 14 17.04519626 14.56455984 15.80487805
20 9 7.270049402 5.071414013 6.170731707
20 9 7.270049402 5.071414013 6.170731707
20 9 7.270049402 5.071414013 6.170731707

11
Figure 1: Experimental flow rate (Q) versus the pump power percentage for the left pump

80
y = 0.9634x - 13.098
70 R = 0.9942

60

50

40

30

20

10

0
0 20 40 60 80 100

Qpred +delta Q Q pred delta Q Q(ml/min)


We get the best fit line from the upper and lower bound. To get the upper bound we add delta
Q by Q predicted and to get the lower bound we subtract delta Q by the predicted Q. the actual
rates are found between these bounds.

12
The same methods that were used to calculate the best fit line data for Q are used to calculate
the best fit line for DAU.
Table 5: LINEST data and TINV data for flow rate DUA and % setting

m 0.866890244 1.21504065 b
st.error of m 0.015015672 0.870391048 st.erro of b
r^2 0.996114794 1.489506237 st.erro of regression
f- test overall 3333.025812 13 dgrees of freedom
regression SS 7394.747159 28.8421748 residual SS
t value 0.693829304 0.877308557 upper m
delta m 0.010418313 0.856471931 lower m
delta b 0.603902815 1.818943466 upper b
DUApred m*setting %+b 0.611137835 lower b

13
Table 6: DAU predicted values and the upper and lower bounds in ml/min

setting % DUA DAU+delta DAU-delta DUA pred


(ml/min) DUA DAU
90 79.9 80.77671362 77.69361159 79.2351626
90 79.7 80.77671362 77.69361159 79.2351626
90 79.3 80.77671362 77.69361159 79.2351626
70 62.1 63.23054247 60.56417297 61.89735772
70 62.1 63.23054247 60.56417297 61.89735772
70 62.1 63.23054247 60.56417297 61.89735772
50 39.4 45.68437133 43.43473436 44.55955285
50 45 45.68437133 43.43473436 44.55955285
50 45 45.68437133 43.43473436 44.55955285
30 27.7 28.13820018 26.30529575 27.22174797
30 27.7 28.13820018 26.30529575 27.22174797
30 27.7 28.13820018 26.30529575 27.22174797
20 18.9 19.36511461 17.74057645 18.55284553
20 18.9 19.36511461 17.74057645 18.55284553
20 18.9 19.36511461 17.74057645 18.55284553

14
Figure 2: Instrumental flow rate (DUA) versus the pump power percentage for the left pump

DAU VS PUMP POWER


90

y = 0.8669x + 1.215
80
R = 0.9961

70

60

50

40

30

20

10

0
0 10 20 30 40 50 60 70 80 90 100

DUA (ml/min) DAU+delta DUA DAU-delta DAU Linear (DUA (ml/min))

in which the x and y values are labeled as (DAU, Pump power)

15
Table 7:

setting DUA Q(ml/min)


% (ml/min)
90 79.9 75
90 79.7 75
90 79.3 75
70 62.1 54
70 62.1 54
70 62.1 54
50 39.4 33
50 45 33
50 45 33
30 27.7 14
30 27.7 14
30 27.7 14
20 18.9 9
20 18.9 9
20 18.9 9

16
Figure 3: Experimental and Instrumental flow rate (Q, DUA) versus the pump power percentage for
the left pump

DAU vs. Q

90
80
70
60
50
40
30
20
10
0
0 10 20 30 40 50 60 70 80 90 100

DUA (ml/min) Q(ml/min) Linear (DUA (ml/min)) Linear (Q(ml/min))

In which the x and y axis are labeled as (DAU, Q)

The graph DAU vs. Q produced a steeper slope than what the computer measured which tells
that the flow rates could be greater than what has been generated.

17
The same procedure we used to collect data for the right pump.
The volume of the water then flowed into the 50.0 Ml graduated cylinder was measured over a
time period at the different power outputs shown in Table 6, ranging from 90.0% to 20% of the
potential power. The average flow rate (Q) was found for each power output by dividing the
corresponding volume by the time elapsed. Experimental and computer results show that the
flow rates increase as the potential power increases
Table 8: data collected during the experimental for the right pimp
%setting V, mL t, s Q, mL/s Q, DUA, mL/min
mL/min
90 76 60 1.266667 76 78.3
90 76 60 1.266667 76 78.3
90 76 60 1.266667 76 78.3
70 55 60 0.916667 55 61.1
70 55 60 0.916667 55 61.1
70 55 60 0.916667 55 61.1
50 35 60 0.583333 35 43.9
50 35 60 0.583333 35 43.9
50 35 60 0.583333 35 43.9
30 19 60 0.316667 19 26.7
30 19 60 0.316667 19 26.7
30 19 60 0.316667 19 26.7
20 9 60 0.15 9 18.2
20 9 60 0.15 9 18.2
20 9 60 0.15 9 18.2

18
The errors in equipment and experimental data collected are shown in Table 9. E(set) is the
given error in the potentiometer reading of the right pump. E(t) is the error in the time
recorded by the iPhone stopwatch. The error in the flow rate E(Q) is a function of E(t) and E(V).
the relative error represents the range from which the measured flow rates deviate from the
actual flow rate. Calculation for the number of significant figures are from the equation:
E = 0.5 * 10^(1-NS)
Where E is the relative error and NS is the number of significant figures. Algebraic rearranging
and taking the logarithm of both sides give NS as a function of E:
NS = 1 log(E/0.5)
Table 9: the error in the measurements for the right pump
E(set), E(V), E(t), s E(Q), E(Q), relative sig fig
% ml ml/s ml/min error
0.1 0.4 1 0.022139 0.145449 0.017478 2.45648
0.1 0.4 1 0.022139 0.145449 0.017478 2.45648
0.1 0.4 1 0.022139 0.145449 0.017478 2.45648
0.1 0.4 1 0.016669 0.123783 0.018184 2.439272
0.1 0.4 1 0.016669 0.123783 0.018184 2.439272
0.1 0.4 1 0.016669 0.123783 0.018184 2.439272
0.1 0.4 1 0.011788 0.098826 0.020209 2.393432
0.1 0.4 1 0.011788 0.098826 0.020209 2.393432
0.1 0.4 1 0.011788 0.098826 0.020209 2.393432
0.1 0.4 1 0.008503 0.072954 0.026851 2.270005
0.1 0.4 1 0.008503 0.072954 0.026851 2.270005
0.1 0.4 1 0.008503 0.072954 0.026851 2.270005
0.1 0.4 1 0.00712 0.050442 0.047467 2.022581
0.1 0.4 1 0.00712 0.050442 0.047467 2.022581

19
The data in Table 10 shows the LINEST data and TINV data. To formulate a best fit line, the x and y
values determined (setting%, Q), the method of least squares was used. In order to do this, the LINEST
function from Microsoft Excel was used. It is a representation of the derivation of Q-measured from Q-
predicted, where Q-predicted is from the line of best fit. The dependence of the flow rate on the
power output percentage is a strong, positive, linear correlation since the R-squared value is nearly 1.
Table 10: LINES data and TINV data for flow rate Q and % setting:

m 0.945732 -10.378 b
st.error of m 0.01294 0.750097 st.erro of b
r^2 0.997572 1.283647 st.erro of regression
f- test overall 5341.215 13 dgrees of freedom
regression SS 8800.979 21.42073 residual SS
t value 0.693829 0.95471 upper m
delta m 0.008978 0.936753 lower m
delta b 0.520439 -9.85761 upper b
Q pred m * %setting + b -10.8985 lower b

The line of best fit is then enclosed by the upper and lower bounds. The upper bound is found
by adding delta Q to Q-predicted. The lower bound is found by subtracting delta Q from Q-
predicted. The actual flow rates should be in between these bounds. These values are shown in
Table 11.

20
Table 11: Q predicted values as well the upper and lower bounds (mL/min):

Qpred Qpred- Q pred


+delta y delta y
76.0663 73.40931 74.7378
76.0663 73.40931 85.86595
76.0663 73.40931 86.3995
56.9721 54.67424 79.20122
56.9721 54.67424 87.62195
56.9721 54.67424 66.20122
37.8779 35.93918 47.28659
37.8779 35.93918 47.28659
37.8779 35.93918 47.28659
18.78369 17.20411 28.37195
18.78369 17.20411 28.37195
18.78369 17.20411 28.37195
9.236593 7.836577 18.91463
9.236593 7.836577 18.91463
9.236593 7.836577 18.91463

21
As shown in figure 4, the experimental flow rate (Q) is graphed versus the pump power
percentage.

Figure 4: Experimental flow rate (Q) versus the pump power percentage for the right pump:
in which the x and y axis are labeled as (Q, % setting)

80
y = 0.9457x - 10.378
70 R = 0.9976

60

50

40

30

20

10

0
0 20 40 60 80 100

As shown in table 12, the same methods that are used to calculate the best fit line data
for Q were used to calculate the best fit line for DUA (the flow rate determined by the
computer). The upper and lower bounds are shown in Table 13.

22
Table 12: LINEST data and TINV data for flow rate DUA and % setting:
m 0.859024 0.970732 b
st.error of m 0.000338 0.019606 st.erro of b
r^2 0.999998 0.033552 st.erro of regression
f- test overall 6450332 13 dgrees of freedom
regression SS 7261.161 0.014634 residual SS
t value 0.693829 0.859259 Upper m
delta m 0.000235 0.85879 Lower m
delta b 0.013603 0.984335 Upper b
DUA pred m* %setting + b 0.957129 Lower b

23
Table 12: DUA predicted values as well the upper and lower bounds (mL/min):

DAU+delta DAU- DUA


DUA delta pred
DAU
78.31765 78.2482 78.28293
78.31765 78.2482 78.28293
78.31765 78.2482 78.28293
61.13247 61.07241 61.10244
61.13247 61.07241 61.10244
61.13247 61.07241 61.10244
43.94729 43.89661 43.92195
43.94729 43.89661 43.92195
43.94729 43.89661 43.92195
26.76211 26.72082 26.74146
26.76211 26.72082 26.74146
26.76211 26.72082 26.74146
18.16952 18.13292 18.15122
18.16952 18.13292 18.15122
18.16952 18.13292 18.15122

Figure 5 is a scatter plot of computer-generated flow rate (DUA) versus the potentiometer
setting (%). The linear regression line is bounded by the upper and lower limits of DUA. This
graph shows a strong linear correlation between the flow rates and the pump potentiometer
setting.

24
Figure 5: Instrumental flow rate (DUA) versus the pump power percentage for the right
pump:
In which x and y axis are labeled as: (DAU, % setting)

90

80 y = 0.859x + 0.9707
R = 1

70

60

50

40

30

20

10

0
0 10 20 30 40 50 60 70 80 90 100

DUA, mL/min DAU+delta DUA DAU-delta DUA

25
Figure 6 shows the experimental (Q) and computer-generated (DUA) flow rates as a function of
pump power percentage. In which the y axis is Q and x axis is DAU. The experimental results
produced a steeper slope than that of the computer measurements which suggests that the
flow rates may have been greater than what was generated by the computer.

Figure 6: Experimental and instrumental flow rates (Q, DUA) versus the pump power
percentage for the right pump:

Q vs. DAU
90
80
70
60
50
40
30
20
10
0
0 20 40 60 80 100

Series1 Series2 DUA(mL/min) Q(mL/min)

26
Conclusion

From this experiment we got two linear equations for each pump, which are the actual flow
rate, Q, and the theoretical flow rate, DAU. Both equations can be used to calibrate the pumps
in order for the theoretical flow rate DAU shown using the Armfield software is more accurate.
Both equations were concluded from the best-fit linear equation for the values of DAU and at
multiple and different potentiometer settings. Three readings were taken at each setting in
order to reduce error. To find the equations the regression analysis method of least squares
was used using LINEST function on Excel. The results of our experiment shows a strong
relationship with a little difference from the predicted flow rate equation at higher and lower
setting reading. This results because the pump is designed to operate very accurately at an
average flow rate.
In order to increase precision and accuracy, we run the each trial at the same setting a few
times to get the same amount of errors and be sure of it, therefore the more we run the
experiment at the same setting the less error we are going to see and it would increase the
accuracy as well.

27
Reference:

http://armfieldonline.com/en/news/view/chemical-reaction

http://www.kimblechase.com/advancedwebpage.aspx?cg=206&cd=4&skutype=202&skufld=sk
u&dm=1250&webid=522

28

Anda mungkin juga menyukai