Tutorial 5: Kinematics of Fluid Motion (I) and

The Bernoulli Equation (II) [Solutions]

MECH 2210 - Fluid Mechanics (L2)
Spring 23-24

Justin Sun
kijesun@connect.ust.hk

11th March 2024

1 Concept Recap
1.1 Multiple Choice Questions
(Q1): Given fluid description ρ, T, p, V, a with particle A, where V = V(x, y, z, t), which of the
following is the Eulerian description of fluid flow?

(a) V = V(x0 , y0 , z0 , t) where x0 , y0 , z0 vary with t.

(b) VA = VA (t) for a particle A where t varies.
(c) V = V(x0 , y0 , z0 , t) where x0 , y0 , z0 are fixed.
(d) V = V(x0 , y0 , z0 , t) where x0 , y0 , z0 , t are fixed.

In the Eulerian description, fluid motion is described by fluid properties at different

locations and times, i.e., we observe the difference of fluid properties at a fixed position.

(Q2): Given fluid description ρ, T, p, V, a with particle A, where V = V(x, y, z, t), which of the
following is the Lagrangian description of fluid flow?

(a) VA = VA (t) for a particle A where t varies.

(b) VA = VA (t) for a particle A where t is fixed.
(c) V = V(x0 , y0 , z0 , t) where x0 , y0 , z0 are fixed and t varies.
(d) V = V(x0 , y0 , z0 , t) where x0 , y0 , z0 , t varies.

In the Lagrangian description, fluid motion is described by properties of fluid particles at

different times, i.e., we observe the fluid properties of a fluid particle no matter the position
across time.

1
(Q3): Which of the extensive properties cannot have an equivalent intensive form?

(a) Kinetic Energy

(b) Momentum
(c) Temperature
(d) Volume

To have an intensive form of a property, the property itself must be mass-dependent.

If we recall the equations of kinetic energy, momentum, and volume, we realise that these
quantities all depend on mass, while temperature is defined by the speed of the particles
inside the volume which we want to measure the temperature. Also, note that density does
not have an intensive form despite the equation involving mass since volume is also
mass-dependent, therefore these mass-dependencies cancel out and therefore density is mass-
independent. We may also think about this logically, since the density does not change with
additional mass.

(Q4): Given the following expression of acceleration a in streamline coordinates (s, n):

   
∂V ∂V ds ∂V dn ∂s ∂s ds ∂s dn
a= + + s+V + +
∂t ∂s dt ∂n dt ∂t ∂s dt ∂n dt

Assuming steady flows and along streamlines, which of the following is the most simplified,
commonly seen, and correct form?
   
∂V ds ∂V dn ∂s ds ∂s dn
(a) a = + s+V +
∂s dt ∂n dt ∂s dt ∂n dt
∂V V2
(b) a = V s+ n
∂s R
∂V ∂s
(c) a = V s+V2
∂s ∂s
   
∂V ds ∂s ds
(d) a = s+V
∂s dt ∂s dt

2
 Given steady flow, fluidic properties at a particular position do not change over time, there-
∂V ∂s
fore the terms = = 0. Since we are considering flow along streamlines, we have
∂t ∂t
ds
= V (velocity is always tangent to streamline). We also have that n is constant, there-
dt
dn
fore = 0. This leaves us with:
dt
∂V ∂s
a=V s+V2
∂s ∂s
∂s ∂s s(t + ∆t) − s(t) |s|δθn n
Lastly, notice that = lim = lim = = :
∂s δs→0 ∂s δs,δt→0 s(t + ∆t) − s(t) Rδθ R

which leads to our final answer of (b).

Note: The topics thereafter involve a similar step to the above - simplifying complex differ-
ential equations (the Navier-Stokes Equations more precisely) based on assumptions. It
is important to understand what the terms mean so that we may simplify them correctly.
∂s
(The step for simplifying is not so important here).
∂s
(Q5): What is the difference between kinematics and kinetics?

(a) Kinematics is a description of motion whereas kinetics describes the reasons why we
have particular changes in motions.
(b) Kinematics is a description of motion whereas kinetics describes the kinetic energy
fluctuations of a particular system.
(c) Kinematics is a description of motion from a spatial and temporal perspective without
reference to forces, whereas kinetics is concerned about the forces acting on a system
that causes changes in motion.
(d) There are no discernible differences.

3
2 Exercises
2.1 Question 1
Determine the acceleration field for a three-dimensional flow with velocity components:
• u = −x
• v = 4x2 y 2
• w =x−y

2.1.1 Solution
Our complete acceleration term is given as follows:
DV ∂V ∂V ∂V ∂V
a= = +u +v +w
Dt ∂t ∂x ∂y ∂z
Note that the velocity components given are not dependent on time, i.e., steady flow, and therefore
∂V
we have = 0. We can also decompose the acceleration terms such that it is dependent on the
∂t
corresponding velocity terms in the same direction, i.e.:
∂u ∂u ∂u
ax = u +v +w
∂x ∂y ∂z
∂v ∂v ∂v
ay = u +v +w
∂x ∂y ∂z
∂w ∂w ∂w
az = u +v +w
∂x ∂y ∂z
such that a = ax i + ay j + az k. Therefore we compute the acceleration components as follows, first
in the x-direction:
∂u ∂u ∂u
ax = u +v +w
∂x ∂y ∂z
2 2
= (−x)(−1) + (4x y )(0) + (x − y)(0)
ax = x
then in the y-direction:
∂v ∂v ∂v
ay = u +v +w
∂x ∂y ∂z
= (−x)(8xy ) + (4x2 y 2 )(8x2 y) + (x − y)(0)
2

ay = 8x2 y 2 (4x2 y − 1)
and finally in the z-direction:
∂w ∂w ∂w
az = u +v +w
∂x ∂y ∂z
2 2
= (−x)(1) + (4x y )(−1) + (x − y)(0)
az = −4x2 y 2 − x
Therefore, we finally have the following:
a = ax i + ay j + az k = xi + 8x2 y 2 (4x2 y − 1)j + (−4x2 y 2 − x)k

4
2.2 Question 2
Air flows from a pipe into the region between two parallel circular disks as shown below. The
V0 R
flow velocity in the gap between the disks is closely approximated by V = , where R is the
r
radius of the disk, r is the radial coordinate, and V0 is the fluid velocity at the edge of the disk.
Determine the acceleration for r = 1, 2, or 3 cm if V0 = 5cm/s and R = 3cm.

2.2.1 Solution
We can change our view of the problem to the diagram below:

where we observe that we can better model the problem using streamline coordinates since the
fluid is only moving in the radial direction. Therefore, in streamline coordinates, our acceleration
is as follows:

a = an n + as s
= as s
∂V
=V s
∂s
∂V
a=V s
∂r

V0 R
Since we are given that V = , we can substitute it into the above equation:
r
∂V
a=V s
∂r
5
  
V0 R V0 R
= − 2 s
r r
2 2
V R
a=− 03 s
r
By the above equation, we can compute the desired acceleration along the streamline for (i)
r = 1cm, V0 = 5cm/s, R = 3cm:

V02 R2
as = − 3
r
(5)2 (3)2
=−
13
as, r=1cm = −225cm/s2

For (ii) r = 2cm, V0 = 5cm/s, R = 3cm:

V02 R2
as = −
r3
(5)2 (3)2
=−
23
as, r=2cm = −28.125cm/s2

And for (ii) r = 3cm, V0 = 5cm/s, R = 3cm:

V02 R2
as = −
r3
(5)2 (3)2
=−
33
25
as, r=3cm = − cm/s2
3

6
2.3 Question 3
As a valve is opened, water flows through the diffuser as shown below at an increasing flow rate
so that the velocity along the center line is given by:
 x
V = uî = V0 (1 − e−ct ) 1 − î
ℓ
where u0 , c and ℓ are constants. Determine the acceleration as a function of x and t. If V0 = 3m/s
and ℓ = 1.5m, what value of c where c ̸= 0 is needed to make the acceleration zero for any x at
t = 1s. Explain how the acceleration can be zero if the flow rate is increasing with time.

2.3.1 Solution
Our complete acceleration term is given as follows:
DV ∂V ∂V ∂V ∂V
a= = +u +v +w
Dt ∂t ∂x ∂y ∂z

Given that V = uî where u = u(x, t), we have v = w = 0, such that we can simplify the above
into the following:
 
∂u ∂u
a= +u î
∂t ∂x

Substituting u = V0 (1 − e−ct ) 1 − xℓ , we can compute the following acceleration term:

∂u ∂u
ax = +u
∂t ∂x   
 x  −ct h −ct
 x i −ct 1
= V0 1 − ce + V0 (1 − e ) 1 − V0 (1 − e ) −
ℓ ℓ ℓ
 
 x   x  1
= V0 1 − ce−ct + V02 (1 − e−ct )2 1 − −
ℓ ℓ ℓ
 
 x  V0
ax = 1 − ce−ct − (1 − e−ct )2
ℓ ℓ

7
For acceleration to be zero for any x, i.e., a = ax = 0 at t = 1s, and given V0 = 3m/s and ℓ = 1.5,
we must have the following:
 
 x −ct V0 −ct 2
ax = 1 − ce − (1 − e ) = 0
ℓ ℓ
V0
ce−ct − (1 − e−ct )2 = 0
ℓ
3
ce−c(1) − (1 − e−c(1) )2 = 0
1.5
ce−c − 2(1 − e−c )2 = 0

The above is a rather complex differential equation, so we just use a calculator (or graphical
method) to solve it (the precise mathematics to solve this equation is not the objective of this
course, and we will never give you something as complex during the exam), where we get:

c = 0.4901 s−1

Now, the acceleration (material derivative) can be zero despite the flow rate increasing with time
is due to the fact that the acceleration can be decomposed into 2 parts, (i) the local derivative,
∂V ∂V ∂V ∂V
, and (ii) the convective derivative, (V ·∇)V = u +v +w , where the summation
∂t ∂x ∂y ∂z
of both terms is the acceleration. The flow rate increase only gives us that the local derivate is
positive, and therefore in order to have acceleration a = 0, we simply require the convective
derivative term to be negative.

8
2.4 Question 4
The fluid velocity along the x axis shown below changes from 6m/s at point A to 18m/s at
point B. It is also known that the velocity is a linear function of distance along the streamline.
Determine the acceleration at points A, B, and C. Assume steady flow.

2.4.1 Solution
Our complete acceleration term is given as follows:
DV ∂V ∂V ∂V ∂V
a= = +u +v +w
Dt ∂t ∂x ∂y ∂z
Given that we are travelling along a streamline that is along the x-axis, we have v = w = 0 and
assuming steady flow, i.e., ∂u
∂t
= 0, we can simplify the above into the following:
∂u
a=u
∂x
Furthermore, given that the velocity is a linear function, we can model the velocity component u
as u = c1 x+c2 . To solve for c1 and c2 , note that at point A, where x = 0, we have VA = u = 6m/s,
therefore:

V A = u = c1 x + c2
6 = c1 (0) + c2
c2 = 6

and at point B, where x = 0.1m we have VB = u = 18m/s, therefore:

VB = u = c1 x + c2
18 = c1 (0.1) + (6)
c1 = 120

Therefore, we have the velocity function in terms of x as follows:

u = 120x + 6

where we can find the acceleration as follows:

∂u
ax = u
∂x
9
 = (120x + 6)120
ax = 1202 x + 720

We can then compute the acceleration at point A where x = 0:

ax, pointA = 1202 x + 720

= 1202 (0) + 720
ax, pointA = 720 m/s2

then the acceleration at point B where x = 0.1:

ax, pointB = 1202 x + 720

= 1202 (0.1) + 720
ax, pointB = 2160 m/s2

and finally the acceleration at point C where x = 0.05:

ax, pointB = 1202 x + 720

= 1202 (0.05) + 720
ax, pointB = 1440 m/s2

10
3 Additional Problems - Bernoulli Equations
3.1 Question 5
When the drain plug is pulled, water flows from a hole in the bottom of a large, open cylindrical
tank. Show that if viscous effects are negligible and if the flow is assumed to be quasi-steady,
then it takes approximately 3.41 times longer to empty the entire tank than it does to empty the
first half of the tank. Explain why this is so.

3.1.1 Solution
We first check our 4 assumptions of the Bernoulli Equation:

1) Inviscid Flow: Given that viscous effects are negligible.

2) Steady Flow: Given that it is a quasi-steady flow.

3) Incompressible Flow: Yes, sufficiently low speed.

4) Along a streamline: Yes, depending on the points we choose.

At a particular time t, the free surface height (or depth of water) is h. Let us apply the Bernoulli
Equation from that free surface to the drain (hole). Given the free surface assumption, we have
V1 = 0, p1 = p2 = 0, z1 = h, and z2 = 0. We have the following:
1 1
p1 + ρV12 + γz1 = p2 + ρV22 + γz2
2 2
1 2
ρgh = ρV2
2
p
V2 = 2gh

Suppose we consider a differential (small) time interval dt, where the differential volume drained
dV– within the time interval dt can be expressed as follows:
1
dV– = − πd2 V2 dt
  4
1 1
πD2 dh = − πd2 V2 dt
4 4

11
 d2 p
dh = − 2 2gh dt
r D
D2 1 − 1
Z Z
− 2 h 2 dh = 1 dt
d 2g
r
D2 1 1
−2 2 h2 + C = t
d 2g
s
D2 2h
t=− 2 +C
d g

To find the integration constant C, we use the boundary condition where t = 0, h = H, therefore
we get the following:
s
D2 2h
t=− 2 +C
d g
s
D2 2H
0=− 2 +C
d g
s
D2 2H
C= 2
d g

Therefore, we have:
s
D2 2h
t=− 2 +C
d g
s s
D2 2h D2 2H
=− 2 + 2
d g d g
D 2 2 √ √ 
r
t= 2 H− h
d g
H
Now we first find the time to empty half the tank, i.e., when h = 2
, as follows:
r !
D2 2 √
r
H
t1 = 2 H−
d g 2
√ !
D2 2 √
r
2
t1 = 2 H 1−
d g 2

We then find the time to empty the entire tank, i.e., when h = 0, as follows:

D 2 2 √
r 
t2 = 2 H −0
d g
2√
2
r
D
t2 = 2 H
d g

12
 t2
The ratio exactly represents how many times it takes to empty the entire tank than it does
t1
the first half of the tank, therefore we have the following:
q √
D2 2
t2 d2 g
H
= 2q √  √ 
t1 D 2
H 1− 2
d2 g 2
1
= √ 
2
1− 2

t2
= 3.414213562
t1

13