ME2 Fluid Mechanics Lab

Michael Pan

CID: 01853189

Department of Mechanical Engineering

Imperial College London

21 Mar 24

Word Count: 1987

Abstract
In this investigation, 5 activities were carried out to show evidence for the existence and
characterise boundary layers in a fluid. A boundary layer for water flow was observed in our
experimental set up, and boundary layer thickness 𝛿 was found to be 2.0 mm ± 0.5 mm.
Boundary layer of air in the wind tunnel was found to have thickness of 15 mm, suggesting
an inverse relationship between 𝛿 and 𝑣. Boundary layers were also found to stay attached
at any negative pressure gradient and small positive pressure gradients but detaches at
large positive pressure gradients. The direction of gravity acting on the boundary layer did
not affect any results, leading to the assumption of no body forces still holding for boundary
layers.

ii
Table of Contents
1.

Nomenclature and Abbreviations .......................................................................................................... iv

1. Introduction .................................................................................................................................... 1
2. Materials and Methods ................................................................................................................... 2
2.1 Experimental set-up .............................................................................................................. 2
2.2 Experimental Procedure ........................................................................................................ 3
2.2.1 Activity 1 ............................................................................................................................ 3
2.2.2 Activity 2 ............................................................................................................................ 3
2.2.3 Activity 3 ............................................................................................................................ 4
2.2.4 Activity 4 ............................................................................................................................ 4
2.2.5 Wind Tunnel Activity ......................................................................................................... 4
2.3 Data Processing ..................................................................................................................... 4
3. Results ............................................................................................................................................. 5
3.1 Activity 1 ................................................................................................................................ 5
3.2 Activity 2 ................................................................................................................................ 5
3.3 Activity 3 ................................................................................................................................ 7
3.4 Activity 4 ................................................................................................................................ 8
3.5 Wind Tunnel Activity ............................................................................................................. 9
4. Discussion...................................................................................................................................... 10
4.1 Boundary Layer Thickness ................................................................................................... 10
4.2 Boundary Layer Detachment ............................................................................................... 11
4.3 No Body Forces Assumption ................................................................................................ 11
4.4 Limitations ........................................................................................................................... 12
5. Conclusion ..................................................................................................................................... 13
References ............................................................................................................................................ 14
Appendix ............................................................................................................................................... 15
Appendix A – Results of Activity 1 .................................................................................................... 15

iii
Nomenclature and Abbreviations
𝐴 Working section cross sectional area
ℎ Manometer height of pitot tube
𝑔 Acceleration of free fall
𝐿 Characteristic length
𝑅𝑒 Reynolds Number
𝑡 Time taken to fill measuring cylinder
𝑈∞ Free stream velocity
𝑣 Dynamic viscosity
𝑉 Volume of water collected

𝛿 Boundary layer thickness

𝜌𝑎 Density of air
𝜌𝑤 Density of water

iv
1. Introduction
This report seeks to characterise the boundary layer using flow visualisation in a working
section with laminar flow. Four experiments were conducted to prove the existence of
boundary layers, find the thickness of the found boundary layer, and find when laminar
boundary layers separate from the boundary surface.

In steady fluid flow at high Reynolds numbers, an inviscid model can be considered for flow
that is away from any walls. However, due to the no slip condition where fluid velocity must
be 0 at any stationary surfaces, the inviscid assumption no longer holds at fluid-surface
boundaries and viscous properties must be considered. The viscosity of the fluid leads to a
region of reduced flow velocity from the viscous transfer of momentum, known as the
boundary layer. The distance from the surface to the point where the flow velocity is 99% of
the free stream velocity 𝑈∞ is the boundary layer thickness 𝛿. The development of this
boundary layer for a fluid with free stream velocity 𝑈∞ is shown below in Figure 1.

Figure 1: Development of boundary layer across a flat plate (Giusti et al., 2023)

As the existence of the boundary layer is a viscous effect, laminar flow would most
effectively show its characteristics. Laminar flow is a type of highly ordered, often steady
flow where viscous forces are dominant. Fluid flow is generally considered laminar when the
Reynolds number 𝑅𝑒 is below 5 × 105 . Reynolds number can be calculated using Equation 1
below, where 𝐿 is the characteristic length and 𝑣 is the kinematic viscosity of the fluid.

𝑈∞ 𝐿 (1)
𝑅𝑒 =
𝑣

1
2. Materials and Methods
2.1 Experimental set-up

Pictures of fluid flow across a working section were taken in this investigation. The
experimental set up (Figure 2) consisted of a central reservoir and pump that supplied water
to a working section. Flow rate could be adjusted by rotating the upstream valve. The
working section (Figure 3) contained camber that provided sections of different pressure
gradients, and connections for ink to be inserted into the flow via syringes to visualise
streamlines within the flow. The upstream syringe port injected ink into a honeycomb that
followed the general fluid flow. The downstream syringe port injected ink through a hole in
section B (Figure 3) that allowed flow visualisation closer to the surface. The working section
had characteristic length 𝐿 of 40 mm.

Figure 2: Assembled experimental set-up (Johnson, 2024)

Figure 3: Working section (Johnson, 2024)

2
Measuring instruments used are listed in the table below, along with their ranges and
uncertainties.

Table 1: Table of measuring instruments

Instrument Quantity measured Range Uncertainty

Meter Rule Boundary layer 0–1m ± 0.5 mm

thickness, 𝛿

Measuring Volume, 𝑉 0 – 500 ml ± 1 ml

cylinder

Stop-watch Times, 𝑡 Sufficient ± 0.3 s

Pitot tube Stagnation pressure, ℎ Sufficient Unknown

2.2 Experimental Procedure

The experiment was split into 5 different activities that were conducted one after another.

2.2.1 Activity 1
The experimental set up was first assembled, making sure that no leaks were present, and
bubbles were not present within the working section. The flow rate of the system was
measured by collecting a volume 𝑉 of water within a measured time 𝑡 from the downstream
tube outlet while the upstream valve was fully open. Two measurements were taken, and
the procedure was repeated after the valve was closed slightly. When the valve was fully
open, water was collected over 5 s, and 10 s for when the valve was slightly closed. All
further procedures were then done with the valve fully open.

2.2.2 Activity 2
To characterise the general flow, ink was injected at constant intervals into the upstream
syringe port to compare streamlines at different times. Thick layers of ink were also injected
to compare streamlines at different heights. The process was videoed to show results.

3
2.2.3 Activity 3
To observe the boundary layer, ink was injected through the downstream syringe port. The
syringe was squeezed just hard enough produce small ink blobs. Ink was also injected at the
upstream port at the same time to compare fluid velocities.

2.2.4 Activity 4
The procedure for activity 3 was then repeated with the working section rotated radially by
90o to observe if the no body forces assumption applies to boundary layers.

2.2.5 Wind Tunnel Activity

A wind and tunnel and pitot tube were used to construct a velocity profile of the boundary
layer of air flowing against a flat plate. The pitot tube was placed at 𝑥 = 0.86 𝑚 and
𝑦 = 1 𝑚𝑚. 20 pressure readings were taken at 1 mm intervals along the y-axis.

2.3 Data Processing

Equation 2 was used to calculate fluid velocity, where 𝑉 is volume of water collected, 𝑈 is
fluid velocity within the working section, 𝑡 is time taken, and 𝐴 is cross sectional area of the
working section.

𝑉 (2)
𝑈=
𝐴𝑡

𝐴 = 1.918 × 10−3 𝑚2 (3)

Reynolds number was then calculated with Equation 1, using fluid velocity from Equation 2.

To find the velocity of air from the pitot tube pressure reading, the following equation was
used,

(4)
ℎ 𝜌𝑤
𝑈 = √2𝑔 − ( )
1000 𝜌𝑎

where ℎ is the manometer heigh in units [mm H2O], 𝜌𝑤 is the density of water, 𝜌𝑎 is the
density of air (Engineering Toolbox, 2019), and 𝑔 is the acceleration of free fall.

4
3. Results
All videos and images from this experiment can be found at this link

3.1 Activity 1

The results from activity 1 show laminar flow across all valve positions and can be found in
Appendix A.

3.2 Activity 2

Figure 4 shows an image when ink was injected at constant intervals.

Figure 4: Image of ink injected upstream

Video of ink injected at constant intervals can be found here, or at link 1:

https://drive.google.com/file/d/1i6NacDmev4iw0ucE-
qeDZ6AgJvIl8Bw0/view?usp=drive_link

5
Figure 5 shows an image when a thick layer of ink was injected.

Figure 5: Image of thick layer of injected ink

The video can be found here, or at link 2:

https://drive.google.com/file/d/1vbE-
wHETFrh6M0LrhW9Pvpya1FeZM18F/view?usp=sharing

The sign of the pressure gradient and its relative magnitude at different sections of the
working section is presented in table 2. The relative magnitude is a scale from 1 to 3 with 1
being the greatest.

Table 2: Sign and relative magnitude of pressure gradient at different sections

Section F E D C B A

Sign 0 Positive Positive 0 Negative 0

Relative n.a. 3 2 n.a. 1 n.a.

magnitude

6
3.3 Activity 3

Figure 6 shows an image of the blobs of ink injected from the downstream port.

Figure 6: Image of ink blobs from downstream port

Figure 7 shows streamlines from the free stream together with ink injected from
downstream.

Figure 7: Image of upstream and downstream ink

The video can be found here, or at link 3:

https://drive.google.com/file/d/1g5bVSZQyOfJIeWuo78XPeI3EAYd8GY82/view?usp=sharing

Table 3 shows if the boundary layer remains attached in each section of the working section.

7
 Table 3: Attachment of boundary layer at different sections

Section F E D C B A

Attached / n.a. Separated Attached Attached Attached n.a.

Separated

Pressure n.a. Positive Positive 0 Negative n.a.

Gradient

3.4 Activity 4
Figure 8 shows an image of ink injected after the set up was rotated.

Figure 8: Image of ink injected after set up rotation

The video can be found here, or at link 4:

https://drive.google.com/file/d/12WaJGO0MUkT0qKvQYoknNsaZb5pr8TbF/view?usp=sharing

Table 4 shows if the boundary layer is attached at each section.

8
 Table 4: Boundary layer attachment at different sections

Section F E D C B A

Attached / Separated Separated Attached Attached Attached n.a.

Separated

Pressure 0 Positive Positive 0 Negative n.a.

Gradient

3.5 Wind Tunnel Activity

A velocity profile was obtained from the wind tunnel data and is shown in figure 9.
25

20
Distnace from sruface / mm

15

10

0
0 2 4 6 8 10 12 14 16 18
Fluid Velocity / m s-2

Figure 9: Velocity profile of wind tunnel boundary layer

9
4. Discussion
Looking at the video from link 1, two consecutive lines of ink both took 3 seconds to cross
the working section and followed the same path. This suggests that the flow across the
working section is steady and laminar. In link 2, the thick layer of ink injection shows a non-
uniform velocity profile both axially and radially. The ink stream accelerates and decelerates
along the axis of the working section due to the pressure gradients exerted by the shape of
the camber. There is also axial variation in the velocity profile, likely due to fluid further
away from the surface being less affected by the shape of the surface. This is due to the
viscous transfer of momentum across the axial direction of the fluid.

Positive pressure gradients are associated with a deceleration in fluid velocity while
negative pressure gradients are associated with an acceleration in fluid velocity
(Giusti et al., 2023), which can be observed in the video. The pressure gradients in Table 1
agree with the shape of the camber in the working section. In section B, the cross-sectional
area decreases by a large amount, thus resulting in a negative pressure gradient and
acceleration of the fluid flow. In sections E and D, the cross-sectional area increases,
resulting in positive pressure gradients and deceleration of the fluid flow. Section E has a
steeper angle than section D and thus has a larger magnitude of pressure gradient. Section B
has the largest decrease in cross sectional area over a short horizontal distance and thus has
the largest pressure gradient magnitude. Sections A, C and F have a constant cross-sectional
area and thus have no pressure gradient across them and a constant velocity profile.

4.1 Boundary Layer Thickness

In Figure 6, ink can be seen to follow streamlines at the camber surface. However, this
stream of ink cannot be assumed to represent the entire boundary layer, or even prove its
existence. Towards the end of the video in link 3, the stream of ink at the surface of section
C has a lower velocity than the stream of ink much higher above it. This suggests the
presence of a boundary layer, as the velocity profile at section C should be uniform due to
the lack of a pressure gradient. The velocity profile in Figure 9 further proves the existence

10
of a boundary layer in air, as the gradual increase form 0 velocity to a uniform 𝑈∞ can be
seen. From Figure 9, the thickness of the air boundary layer was found to be 15 mm.

To find the edge of the boundary layer and hence the thickness 𝛿, the blobs of ink at the
boundary were observed closely. In Figure 6, the leading point of each stream of ink can be
observed to bunch up into a blob as the stream progresses down section C. This is likely due
to the leading point of the stream of ink approaching 𝑈∞ of the flow as the boundary layer
develops. As the axial component of velocity does not increase significantly after that point,
any small radial component of velocity can cause the stream of ink to separate and shift
radially, forming the blob. Therefore, the boundary layer thickness can be found from taking
the distance from the surface to the bottom of a blob. From Figure 6, 𝛿 was found to be
around 2.0 mm ± 0.5 mm. This suggests that 𝛿 decreases with increasing density and
kinematic viscosity.

However, this is a very crude measurement as there is not a distinct point to measure the
blob from. The ruler used to measure 𝛿 also only has a precision of 0.5 mm, which is very
large compared to the value of 𝛿 itself.

4.2 Boundary Layer Detachment

Flow reversal can be observed in sections E and F in Figure 8, suggesting that the boundary
layer has separated (Giusti et al., 2023). Boundary layer attachment and separation is listed
against pressure gradient in table 3. As the boundary layer remains attached at section B
despite having the largest magnitude of pressure gradient, boundary layers will likely remain
attached at any negative pressure gradient. However, the boundary layer detaching at
section E and not E suggests that boundary layers stay attached at smaller magnitudes of
positive gradients but detach after the positive pressure gradient reaches a critical
magnitude.

4.3 No Body Forces Assumption

The results from Figure 8 and link 4 are identical to those from Figure 6 and link 3. Rotating
the working section should have resulted in gravity acting in a different direction relative to

11
the axes of the working section. This suggests that the assumption of no body forces holds
for boundary layers as it is not affected by gravity.

4.4 Limitations

Results from this investigation have largely been qualitative descriptions of observations
due to difficulty and lack of fluid velocity measurements, making conclusions less precise.
Measurement of boundary layer thickness is also limited by the precision of the ruler and
the imprecise boundary. One major limitation is the use of ink to visualise flow and
boundary layer, as it is not truly representative of the entire boundary layer. Ink also has
different properties than water that may affect how accurate the visualisation is.

12
5. Conclusion
Through this investigation, the existence of boundary layers in water and air was proven and
observed. Boundary layer thickness in section C of the working section was found to be
2 mm ± 0.5 mm, while boundary layer of air in the wind tunnel was found to have thickness
of 15 mm. This suggests an inverse relationship between 𝛿 and 𝑣. Boundary layers were also
found to stay attached at any negative pressure gradient and small positive pressure
gradients, but detaches at large positive pressure gradients. The direction of gravity acting
on the boundary layer did not affect any results, leading to the assumption of no body
forces still holding for boundary layers.

In future experiments, more quantitative data can be recorded by measuring the fluid
velocity of water at different radial distances from the camber surface, similar to the wind
tunnel activity.

13
References
Engineering ToolBox (2019). Water - Dynamic and Kinematic Viscosity. [online]
Engineeringtoolbox.com. Available at: https://www.engineeringtoolbox.com/water-
dynamic-kinematic-viscosity-d_596.html.

Giusti, A., Johnson, P.B. and Touber, E. (2023). Imperial College Fluid Mechanics 2. 3.1.2 ed.

Johnson, P.B. (2024). ME2 Fluid Mechanics Lab Handout 2024.

14
 Appendix
Appendix A – Results of Activity 1

Table 5 below shows the volume of water collected and the resulting Reynolds numbers
when the valve is fully and partially open.

Table 5: Results from activity 1

Valve Position Fully Open Partially open

Recording 1 2 3 4
number

Time taken / s 5.40 ± 0.3 5.37 ± 0.3 10.30 ± 0.3 10.29 ± 0.3

Volume 370 ± 1 380 ± 1 165 ± 1 160 ± 1

collected / ml

Fluid velocity (3.55 (3.69 (8.35 (8.10

/ m s-1 ± 0.21) × 10−2 ± 0.22) × 10−2 ± 0.29) × 10−4 ± 0.29) × 10−4

Reynolds 1423 ± 83 1469 ± 86 333 ± 12 323 ± 11

number

Flow regime Laminar Laminar Laminar Laminar

Figures 10 to 12 show the pictures taken at each data reading. Figure 4 and link 1 shows
reading 4.

Figure 10: Image from reading 1

15
 Figure 12: Image from reading 2

Figure 11: Image from reading 3

As Reynolds numbers are below 5 × 105 even when the valve is fully open, the flow regime
is laminar across all valve positions. Further procedures were therefore done with the valve
wide open for a constant, repeatable flow rate.

16