Sie sind auf Seite 1von 6

Siddharth

Pai 1001545964

PRA 0104

Microfluidics - An Observation of the Flow of a Fluid through Four Different channels


Siddharth Pai 1001545964

Introduction:
The behaviour of fluids as it flows through certain paths is a very important concept to understand as
they can be used in practical applications such as understanding blood flow patterns, modern DNA
sequencing, and all biological microchip applications [1]. Given the nature of these application, it can
be seen that the behaviour of the flow of fluids need to be examined on a much smaller scale where
properties like surface tension, pressure and speed of the fluid start dominating and behave
differently [1]. Past literature on this mater reveals the work of great scientists who formulated
equations and mathematical models like the Bernoullis equation and Poiseuilles Law [2], which help
understand the vector profile of a fluid in a given channel. In this experiment, flow of a fluid is
examined through four different channels and their behaviors are later discussed.

Experimental Procedure and results:
The procedure detailed in the laboratory manual [1] was followed for the most part of this
experiment. The discrepancies/ additional information regarding the procedures performed during
the experiment are listed below:
The software used for capturing the pictures was: Infinity Capture;
A 10x (overall 100x) magnification was used under the microscope instead of the
40x (overall 400x) as mentioned on the lab document;
Channels 1, 3, and 4 were observed under 200ms of exposure time;
Channel 2 was observed under 150ms of exposure time;
The hemocytometer was placed slightly off centre such that the middle section with
the narrowest line separation was closer towards the bottom left corner.

Measuring Scale:
A hemacytometer (shown in figure 1) was placed under a microscope first with 4x magnification and
second with 10x magnification. A picture was taken after placing the hemocytometer appropriately
in the viewing frame and an on-screen pixel ruler was used to calculate a magnification scale, M
[units: m/pix], found to be M = 1.06E-6 m/pix. Using this magnification number, the velocity in m/s
was found using the following equation: = ()()/ ( 1) where is the exposure time.
For each channel and streaks of approximately the same brightness and at the same height (h) from a
defined origin were used to calculate the velocity. The different velocities at specific heights were
averaged.

Channel Imperfections: A picture of the straight channel was taken (fig. 2) under 4x magnification
and the subtle imperfections on the channel edges can be seen. These imperfections include a slight
bend or asymmetric pattern in the edge of the channel, which affect the flow of the fluid. These
imperfections were also consistently seen in all the other channels and it is prudent to keep this in
consideration while evaluating the results from the experiment.

Straight channel: It was found that the velocity profile of the fluid flowing through the straight
channel, as a function of the height, h (measured from one edge of the channel) was somewhat
parabolic (refer to figure 2). Additionally, the observed flow was laminar flow throughout the
channel.

Siddharth Pai 1001545964

PRA 0104

Diffuser: Laminar flow was observed throughout the channel, with the maximum velocity flowing
through the centre of the cross section of the channel. Additionally, the velocity of the fluid decreased
as it flowed from a smaller to a wider width of the cross section.

Step: It was found that the velocity of the fluid was maximum at the point slightly closer to the inner
side of the curve from the centre. Some turbulent flow was observed on the outer edge of the channel
and the velocity of the fluid slowed during the turn and stayed constant before and after the turn.

Bent: The velocity of the fluid was greater towards the outer edge of the turn and slowed towards the
inner edge. Laminar flow was observed throughout the channel unlike in the step channel.
Error Analysis (Sample Calculation):
I) Propagated error in the calculation of magnification factor, M:

1
1

0.05 3
= ;
= =
= 0.02;
= ! =
= 2.26 8

47

47!
() = 5 6, () = 4

where is the length in m and p the equivalent measure in pixels.

)! + (
)! = (1.06 7)! + (9.05 8)! = 1.4 7


II) Propagated error (sample calculation) in length of streak, :

= !"#$% ;
= = 1.06 6;
= !"#$% = 69
!"#$%

!"#$% = 4,
= 1.4 7
=

!"#$%

!"#$% )! + (

)! =

(4.24 6)! + (9.66 6)! = 1.05 5


III) Propagated error (sample calculation) in velocity, v (m/s):

7.3 5
= ; = = 5; = ! =
= 1.8 3

0.2!
() = 1.05 5, () = 0.05
=

!"

!"

( !" )! + ( !" )! = (5.25 5)! + (9.1 5)! = 1 4 /


IV) Statistical error in velocity due to averaging (sample calculation for channel 1):
Standard deviation = 0.000142702. Applying Central Limit Theorem, a better error approximate is
got by = 0.000142702/ 21 = 3 5/. Since 3E-5 < 1E-4, the propagated error was
chosen to be the error in the velocity calculated.

Additional sources of error in the lab include the error introduced through the hemocytometer since
the lines of this instrument have a width of ~2px, meaning an error of 2E-6m. Moreover, in the
bent channel, when the streak of the beads are a curve, the arc is an approximation using the
trigonometry, since the pixel ruler could only measure distances in straight lines.

Siddharth Pai 1001545964

PRA 0104

V) A slightly different approach was also taken in order to deduce the possible error associated with
the velocity profile of the straight channel. It is well known and widely accepted to apply Poiseuilles
Law with laminar flow in a given channel. As such, the theoretically calculated velocities were plotted
on the same graph (refer to graph 1) as the observed velocities and the deviation between the two
were treated as the error and were calculated through the following equation:
Error = (!"#$%&$' !!!"#!$%&'( )! ( 1) ( ) , where !!!"#!$%&'( is
!
calculated using Poiseuilles Law, which states that !!!"#!$%&'( = ! (1 ! ). Note that ! is the
maximum velocity in the centre of the cross section, r is the radius away from the centre of the cross
section and lastly R is the radius of the entire cross section. It was found that the standard deviation
of the data was about 3.6E-5 m/s, which resulted in an error of about 9.6%.

Discussion:
Straight: The velocity of the fluid was expected to be the greatest at the centre of the channel due to
the symmetry of the shape of the channel and the no-slip boundary condition. This hypothesis was
confirmed through the velocity profile that was created with the data collected through the lab. The
velocity profile showed a parabolic relationship along the channel (graph 1), which agrees with
Poiseuilles Law of laminar flow [2]. Laminar flow was observed throughout the channel since straight
streams of beads could be seen across the section of the channel and no signs of turbulence
(including stagnation, vortex clusters, etc.) were noticed. The velocity of the fluid could be
manipulated by varying the height at which the syringe was placed. A higher height of the syringe
would provide the fluid a higher gravitational potential to begin with and this additional energy
would result in a higher flow velocity at the channel. Moreover, the velocity of the fluid could be
increased by pressing on the gravity head or by increasing the amount of fluid entering the channel.
The inverse of the latter is also true, which could be implemented to reduce the velocity. The above
hypotheses were also verified from the data collected through the lab.

Diffuser: The velocity of the fluid and the cross-sectional area of the channel were assumed to be
inversely related: (A)(v) = C. Given this relationship, as the fluid flowed from a region of the channel
with a smaller width to a larger width (smaller cross sectional area to a larger cross sectional area)
the velocity of fluid decreased, thus validating the initial assumption. The latter was found by
examining the difference in width sizes of the beads as it flowed through the region with a smaller to
a larger width. Additionally, it was assumed that the depth of the channel was constant throughout
the region of examination, thus A x, where x is the distance from the mouth of the nozzle. Since Av
= C, xv = D, and this would suggest an inverse (reciprocal) relationship between x and v, which was
the case from the acquired data, as seen in graph 2. It was also observed that in the velocity profile
along the vertical direction of this channel resembled that of the straight channel a quadratic
relationship, thus was not discussed in much detail. Varying the height of the syringe varied the
velocity of the fluid however it did not change the flow pattern and characteristics or introduce any
unexpected behaviour. Lastly, just like in the previous case, the maximum velocity of the fluid was
found to be in the centre of the cross section of the channel.

Step: The velocity profile of this channel suggested that the maximum velocity occurred slightly
closer to the inner side of the curve (near the abrupt turn) and slowed down further away from this
point. This was an asymmetric velocity profile unlike the ones observed earlier. The velocity before
and after the turn however was the same. Additionally, at the very outer edge of the turn (refer to
figure 4) it was observed that some beads tended to not flow with the rest of the fluid but rather

Siddharth Pai 1001545964

PRA 0104

form a circular flow (eddy current) counter clockwise (opposite to the direction of flow). This
suggested turbulent flow in the channel at that point. Manipulating the velocity of the fluid by
varying the height of the syringe did not change the flow pattern and the observations mentioned
above stayed the same.

Bent: The velocity profile of this channel was not determined due to the complexity associated with
the shape of the channel and definition of the axes required. However certain streaks of the beads
were used at different distances away from the inner edge of the channel and it was found that the
velocity of the fluid was the greatest, furthest away from the inner edge and slowed down as one
came closer towards the inner end of the curve. Laminar flow was observed throughout the channel
since straight streams of beads could be seen across the section of the channel and no signs of
turbulence (including stagnation, vortex clusters, etc.) were noticed. This suggested that the flow of
the fluid does in fact depend on whether or not the turn is a sharp turn as supposed to a smoother,
bent turn. A sharp turn introduces turbulence in the flow of the fluid whereas a smoother curve
keeps the fluid in a laminar flow throughout the channel.

Conclusions:
It was found that for the straight and diffuse channels, the velocity profile revealed that the velocity
of the fluid was maximum at the centre of the cross section of the channel, which was in agreement
with Poiseuilles Law (max velocity ~5.8E-4m/s in straight channel). However for the sharp turn
channel, it was found that the velocity was maximum at a point slightly closer to the inner edge of the
channel during the turn (max velocity ~2.93E-04, at a h of 51px). Additionally, the bent channel had
the maximum velocity on the outer edge of the channel (~6.57E-04). Laminar flow was observed for
all the channels except in the step channel, where some eddy currents were observed on the outer
edge of the turn (refer to fig. 5). Lastly, an inverse relationship was found between the cross sectional
area and the velocity of the fluid (refer to graph 2). Manipulating the velocity by varying the syringe
height, changing the amount of fluid entering the channels, and using the gravity head did not change
the behaviour of the fluid flow through the channels.

These observations regarding the flow of fluids on a small scale basis help to understand the
way fluids work and can be used to simulate their behaviour when used in real world applications,
which is of great value. With certain boundary parameters in a given project, desired results can be
achieved by manipulating variable parameters for example achieving a certain fluid velocity by
varying the cross-sectional area.






References and Bibliography:
[1] A. Shukalyuk, B. Keith, E. Chung, T. Dell, E. Brisson and D. Tremelli, "Introduction to Microfluidics,"
University of Toronto, Toronto, 2006.
[2] Fluid Velocity Profile. (2015). Available: http://hyperphysics.phy-astr.gsu.edu/hbase/pfric.html

Siddharth Pai 1001545964

PRA 0104

Tables and Figures:

Figure 1: Hemacytometer under 4x Figure 2: Imperfections in Straight Channel


Figure 4: Diffuser Channel

Figure 5: Step Channel

Figure 6: Bent Channel

Figure 3: Straight Channel

Graph 1 Velocity profile with a quadratic regression curve for channel 1 straight

Siddharth Pai 1001545964

PRA 0104

Graph 2 Velocity profile with an inverse relationship for channel 2 diffuser channel

Graph 3- Velocity Profile for channel 3 step channel


Trial
1
2
3
4
5
6
7

Data from Channel 3 Step (Sharp turn)


h in px
l in px
l in m
time
h in m
11
15
1.60E-05
0.2
1.17E-08
33
41
4.36E-05
0.2
3.51E-08
51
55
5.85E-05
0.2
5.42E-08
83
23
2.45E-05
0.2
8.83E-08
109
25
2.66E-05
0.2
1.16E-07
124
9
9.57E-06
0.2
1.32E-07
146
2
2.13E-06
0.2
1.55E-07
Table 1 Data from channel 3 Step (Sharp turn)

speed
7.98E-05
2.18E-04
2.93E-04
1.22E-04
1.33E-04
4.79E-05
1.06E-05

Trial start x
1
345
2
303
3
306
4
406
5
286

Data from Channel 4 - Bent


start y end x end y l in px
l in m
time (s) speed (m/s)
308
454
326 110.48 1.18E-04
0.200
5.88E-04
356
426
368 123.58 1.31E-04
0.200
6.57E-04
354
341
353
35.01 3.72E-05
0.200
1.86E-04
321
486
340
82.23 8.75E-05
0.200
4.37E-04
251
352
254
66.07 7.03E-05
0.200
3.51E-04
Table 2 Data from channel 4 - Bent

Das könnte Ihnen auch gefallen