Momentum Balances: Qin Q Out Q Q
Momentum Balances: Qin Q Out Q Q
Momentum Balances: Qin Q Out Q Q
7. Momentum balances
Partly based on Chapter 7 of the De Nevers textbook (sections 7.1-7.3).
Introduction
Next to mass and energy, momentum is an important quantity in fluid mechanics. Momentum
is mass times velocity. Since velocity is a vector quantity, momentum also is a vector
quantity. A momentum balance therefore is a vector equation. However, in many cases in this
course we will only consider momentum is one specific (and specified) direction, and (thus)
one component of the momentum balance.
Recall the general form of a balance for a scalar quantity q in a well-specified volume with
inflow and outflow as discussed in LN03:
dq
= φq ,in − φq ,out + Bq − Dq
dt
Generalizing this to the vector quantity momentum p (note the notation: vectors are bold
face):
dp
= φp,in − φp,out + Bp − Dp (1)
dt
The rest of these Lecture Notes (and Chapter 7 of the De Nevers text book) focuses on
examples of applications of Eq. 2. Most of these examples consider a steady state:
(note that since Eq. 3 is a vector equation, its right hand side has a zero vector 0).
Examples
Jet on a surface
Figure 1
We estimate the forces that surfaces experience when hit by a water jet, see Figure 1. Here the
vector character of the momentum balance becomes apparent. What is the force the jet exerts
on the surface? We make some simplifying assumptions here. The major one is that the liquid
flows out sideways (i.e. in the radial (r) direction) on the surface. In the real situation the jet
would break up in drops that move in a spectrum of directions. Setting up a steady-state
momentum balance in the x-direction should give us the force on the surface that (given the
symmetry of the problem) we expect to be in x-direction. The general, steady state form is
φm,in v in − φm ,out v out + ∑ F = 0 . In x-direction this gets φm,in vx ,in − φm,out vx,out + Fx = 0 (only one
force F with x-component Fx). The essential step is to realize that vx ,out = 0 ; the outflow has
no velocity (and thus momentum) in the x-direction, only in the radial direction. Then the
expression for the force is Fx = −φm ,in vx ,in .
The mass flow is φm ,in = 20 kg/s, the nozzle that produces the jet has a diameter d=35
4φm
mm so that the velocity with which the water exits the nozzle is =20.8 m/s. If we
ρπ d 2
assume the vx ,in to be the same velocity with which the water exits the nozzle (20.8 m/s), then
Fx = −416 N . What is the meaning of the minus sign? It means that to keep the surface at its
place we have to exert a force of 416 N in the negative x-direction. In other words, the jet
exerts a force on the surface of 416 N in the positive x-direction. If we would not hold the
plate, it (obviously) would be blown away in the positive x-direction.
The force would be higher if the water would (partly) bounce back from the surface, as
schematically shown in Figure 2; think through for yourself why this is.
Figure 2
Pipe bend
Consider a horizontally placed pipe bend, with an angle θ as defined in Figure 3. The pipe has
a constant diameter D. If the liquid flowing through the bend has a constant density ρ the
average velocity magnitude of the liquid flowing through the pipe is the same in every cross
section of the bend. The direction of the velocity, however, changes and we expect that we
need forces (in x and y direction) to keep the bend in place. Assume a steady mass flux φm .
4φm
The velocity magnitude in the pipe we call U; U = .
π D2 ρ
Figure 3
The steady state momentum balance over the system within the dashed rectangle in x-
direction reads: φmU − φmU cos θ + Fx = 0 . In y-direction it reads −φmU sin θ + Fy = 0 (the
minus sign in the latter equation is there since only the outflow contains positive y-
momentum). The forces to hold the pipe bend in place therefore are
Fx = φmU ( cos θ − 1) , Fy = φmU sin θ . The force in x-direction is negative for any θ.
Calculation of such forces plays a major role when designing pipeline systems. Suppose
we put a bend of 90o in the oil pipeline from Lecture Notes 06b. It transported 0.8 m3 oil per
4φ
second through a pipe with diameter 40 cm. The velocity in the pipe was U = V2 = 6.4 m/s .
πD
The density of the oil was ρ=920 kg/m so that φm =736 kg/s. The forces then are
3
U-bend
Figure 3b
In many cases forces due to pressure (and pressure differences) are part of momentum
balances. Consider the following example: A steady-state, turbulent oil flow is sent through
the horizontally placed 180o pipe bend shown in the figure. The cross sectional area at the
inlet (1) is A1=1 cm2, at the exit (2) it is A2=0.5 cm2. The oil has constant density ρ=800
kg/m3. The volumetric flow rate is φV=0.4 litre/s; the gauge pressure at cross section (1) is
p1=0.25 bar. (A) Determine the oil velocity at cross sections (1) and (2). (B) If we assume
frictionless flow, what is the gauge pressure at cross section (2)? (C) Determine the force in x-
direction to keep the pipe bend in place.
φV φV
(A) Mass balance: v1 = = 4 m/s, v2 = = 8 m/s.
A1 A2
Figure 4
Water rockets are relatively easy to make launching devices. They usually consist of a (upside
down) soda bottle, filled with water and pumped to a certain pressure with e.g. a bicycle
pump. Once a hole in the cap (which now is at the bottom of the upside down bottle) is
opened a water jet emerges which pushes the rocket into the air (see Figure 4; also check out
YouTube and search for “water rocket”). With the tools we have been discussing so far in the
course, we should be able to (at least) come up with some design rules related to the desired
pressure, geometrical constraints etc.
Application of the Bernoulli equation between a point at the liquid surface in the bottle,
and a point at the bottom where the water flows out learns that the outflow velocity v is
∆p
v = 2 gH + . Note that we (implicitly) assumed here that the liquid surface velocity
ρ
was much less than v, which is OK if d ≪ D . Also we applied the steady state Bernoulli
equation to an unsteady problem (since e.g. the liquid level H is a function of time). You have
seen this quasi steady state approach before.
The momentum balance in z-direction gets a bit complicated: the velocity U of the
bottle is a function of time (the bottle is supposed to accelerate), as is the mass of the bottle,
d π
since water is pouring out continuously: ( mU ) = ρ d 2v 2 − mg . The mass of the rocket is
dt 4
mainly the water in the rocket (the plastic bottle itself has not much mass) so that
π dm π dH π
m = ρ D 2 H . A mass balance relates H and v: = ρ D2 = − ρ d 2 v . Finally (if we
4 dt 4 dt 4
consider the air in the rocket e.g. as an ideal gas), the water level and the pressure in the
rocket are related (thinks this through for yourself). All these equations (Bernoulli, z-
momentum balance, mass balance, (ideal) gas law) are coupled and can be used to determine
U, H, and the trajectory of the rocket as a function of time. This is quite a complicated
exercise that can not be done analytically (computational methods required). We will limit
ourselves here to answering the question what pressure we would need to actually get the
rocket off the ground.
Assume D=15 cm, d=1 cm. Initially assume a liquid level H=30 cm. Given the
d π d
momentum balance ( mU ) = ρ d 2v 2 − mg , for ( mU ) to be positive (then at least
dt 4 dt
π ∆p
initially the rocket goes up) we need ρ d 2v 2 > mg . Since v 2 = 2 gH + , and
4 ρ
π ∆p
m = ρ D 2 H the inequality gets 2d 2 > D 2 Hg − 2d 2 Hg . Since usually D 2 ≫ d 2 we can
4 ρ
D2
simplify this to ∆p > ρ Hg . With our numbers ∆p = 3.3 bar. A good bicycle pump can
2d 2
deliver such pressure; I am not sure if a soda bottle can withstand it.
Notice this interesting design question: Usually we cannot do much about D and H
(largely dictated by the soda bottle geometry). We have some freedom in d however. A small
d is beneficial for it provides less water outflow and thus (potentially) a longer flight. It also
requires more pressure. However, ∆p is limited by the strength of the bottle and/or the pump
we have available.