Flow in Closed Conduits

Flow in Closed Conduits

6.1 Flow regimes

Figure 6.1-1 Reynolds’ experiment.

Laminar or well-ordered type of flow exists when adjacent fluid layers slide smoothly
over one another. Mixing between layers occurs only on a molecular level. Turbulent
flow exists when packets of fluid particles are transferred between layers, giving the
flow a fluctuating nature. Osborn Reynolds first described the existence of laminar and
turbulent flow quantitatively through his classic experiment in 1883. As shown in Figure
6.1-1, water was allowed to flow through a transparent pipe at a rate controlled by a
valve. Reynolds introduced a dye having the same specific gravity as water into the flow
to observe what was happening. He found that at low flow rates the dye pattern was
regular and formed a single line of color as show in Figure 6.1-1(a). The pressure drop
was also found to directly proportional to the flow rate. As the flow rate was increased a
point was reach where the dye trace was seen to be unstable and it broke up after a
short distance. At still higher flow rates the dye almost immediately dispersed
throughout the pipe cross section. The relationship between pressure drop and flow
rate now became almost quadratic instead of linear.
The stable flow observed initially was called laminar flow. The unstable flow pattern,
characterized by high degree of mixing between the fluid elements, was called turbulent
flow. There is a transition region in between laminar and turbulent where the flow is
unstable but not thoroughly mixed. Laminar flow in a tube persists up to a point where
the value of the Reynolds number is about 2000. Reynolds number is defined as

                        NRe =  =
The Reynolds number is a ratio of the inertial momentum flux (rV2) in the flow direction
to the viscous shear stress or viscous momentum flux in the transverse (mV/D)
direction. Turbulent flow occurs when Reynolds number is greater than about 4000.
Viscous forces are a manifestation of intermolecular attractive forces that stabilize the
flow. Therefore stable laminar flow should occur at low Reynolds numbers where
viscous forces dominate.
6.2 Generalized Mechanical Energy Balance Equation
For laminar flow of a fluid in a cylindrical tube of radius R and length L, the Hagan-
Poiseuille equation provides a relationship between volumetric flow rate and pressure
drop across the tube as follows.

                        Q = tw =  =

Figure 6.2-1 A general piping system.

For a general piping system shown in Figure 6.2-1, we need the generalized
relationship, equation (6.2-1), that can account for the effect of pressure drop on
incompressible fluid flow, changes in elevation, tube cross section, changes in fluid
velocity, sudden contractions or expansions, and friction loss through pipe and fittings
such as valves and flow meters.

                         + gz1 +  + hwp =  + gz2 +  + ef

(6.2-1)
Each term in this equation has units of energy per unit fluid mass flow rate or
(length/time)2.
P = pressure
r = fluid density
g = acceleration of gravity
z = elevation relative to a reference surface
V = average fluid velocity
a = kinetic energy correction factor
a = 2 for laminar flow
a = 1 for turbulent flow
wp = work done per unit mass flow rate
h = pump efficiency (h < 1)
ef = friction loss due to piping and fitting
The friction loss is given by the following equation
                        ef = 4  +
Kfitting,j                                                   (6.2-2)
where

fi =  = friction factor in tube segment i with length Li and diameter Di.

Vi = average velocity within tube segment i.

  Kfitting = friction loss factor or loss coefficient for pipe fittings, some typical values are
given in Table 6.2-1. The velocity Vj in the summation is for the fluid just downstream of
the contraction, expansion, or fitting.
Table 6.2-1 Friction loss factor for various pipe fittings.
Kfitting  
Fitting
Globe valve, wide open 7.5
Angle valve, wide open 3.8
Gate valve, wide open 0.15
Gate valve, half open 4.4
Standard 90o elbow 0.7
Standard 45o elbow 0.35
Tee, through side outlet 1.5
Tee, straight through 0.4
Sudden contraction 0.4
(turbulent flow)
Sudden expansion
(turbulent flow)

The friction factor for laminar flow (NRe =  < 2000) is given by

f =                                                                                               
(6.2-3)
The friction factor for turbulent flow (Re > 4000) can be estimated by

            f = {- 1.737 ln[0.269  - ln (0.269  + )]}-2                          

(6.2-4)
In this equation e is the surface pipe roughness and D is the inside pipe diameter.
Representative values for surface roughness are given in Table 6.2-2.
Table 6.2-2 Surface roughness
e (ft) e (mm)
Surface
Concrete 0.001-0.01 0.3-3.0
Cast iron 0.00085 0.25
Wrought iron 0.00015 0.045
Galvanized iron 0.0005 0.15
Commercial steel 0.00015 0.046
Drawn tubing 0.000005 0.0015
Equation (6.2-5) developed by Churchill1 adequately predicts the Fanning fiction factor
over the entire range of Reynolds number including a reasonable estimate for the
transition region between laminar and turbulent flow.

f = 2                                                          (6.2-5)

In this equation A =  and B =

If the fluid flows through a noncircular duct, then the equivalent diameter, Deq, can be
used in equations (6.2-2, 3, 4, 5). The equivalent diameter is defined as

                        Deq = 4rH = 4
where              rH = hydraulic radius
Across = cross sectional area of the flow
Pwet = wetted perimeter of the duct

Figure 6.2-2 Flow through an annular tube.

For the flow through an annular tube, the equivalent diameter is given as

                        Deq = 4  = Do - Di

Example 6.2-1. ----------------------------------------------------------------------------------
Water is pumped from the upper reservoir to the lower reservoir through the piping
system shown. Determine the power required for the pump if the water flow rate is 60
kg/s. The fittings from pipe D1 to pipe D2 and from pipe D2 to pipe D3 can be
considered to be standard 90o elbows. Data:
h1 = 10 m, h2 = 3 m, L1 = 50 m, L2 = 300 m, L3 = 2 m, D1 = 0.2 m, D2 = 0.5 m, D3 =
0.03 m, water viscosity = 1 cP = 10-3 kg/m×s, r = 1000 kg/m3. The pipe roughness is
0.05 mm. The pump efficiency is 75%.

Solution ------------------------------------------------------------------------------------------
Applying the mechanical energy balance between (1) and (2) we have

                         + gz1 +  + hwp =  + gz2 +  + ef  

Let the reference level be at (2), the end of pipe 3, the energy equation becomes

                         + g(h1 + L1 - L3) + 0 + hwp =  + 0 +  + ef

                        g(h1 + L1 - L3) + hwp = gh2 +  + ef

D(m) A(m2) V(m/s) NRe e/D f
.2 3.14´10-2 1.91 3.82´105 2.50´10-4 0.00406
.5 1.96´10-1 0.306 1.53´105 1.00´10-4 0.00431
.03 7.07´10-4 84.9 2.55´106 0.0017 0.00600

                        ef = 4  + Kfitting,j
            4  = 2´ 10-3[4.06´  + 4.31´  + 6´ ]
= 5.77´ 103 m2/s2

            Kfitting,j = 0.5´1.912´0.4                       sudden contraction, Kfitting

= 0.4
+ 0.5´0.3062´0.7                       standard 90o elbow, Kfitting = 0.7
                                 + 0.5´0.3062´7.5                       open globe
valve, Kfitting = 7.5
                                 + 0.5´84.92´0.7                         standard 90o
elbow, Kfitting = 0.7

            Kfitting,j = 2.52´ 103 m2/s2

Therefore                  ef = 5.77´103 + 2.52´103 = 8.29´103 m2/s2

                        g(h1 + L1 - L3) + hwp = gh2 +  + ef

                        9.81(10 + 50 - 2) + 0.75wp = 9.81´3 +  + 8.29´103

wp = 1.51´104 m2/s2
The power required for the pump is

                         =  wp = 60´1.51´104 = 9.08´105 W = 1220 hp

Note: 1 hp = 746 W

Resistance formula: Hagen-Poiseuille equation, Darcy-Weisbach

formula, Colebrook –White
In nonideal fluid dynamics, the Hagen–Poiseuille equation, also known as the Hagen–
Poiseuille law, Poiseuille law or Poiseuille equation, is a physical law that gives
the pressure drop in an incompressible and Newtonian fluid in laminar flow flowing
through a long cylindrical pipe of constant cross section. It can be successfully applied
to air flow in lung alveoli, or the flow through a drinking straw or through a hypodermic
needle. It was experimentally derived independently by Jean Léonard Marie
Poiseuille in 1838[1] and Gotthilf Heinrich Ludwig Hagen,[2] and published by Poiseuille in
1840–41 and 1846.[1] The theoretical justification of the Poiseuille law was given
by George Stokes in 1845.
The assumptions of the equation are that the fluid is incompressible and Newtonian;
the flow is laminar through a pipe of constant circular cross-section that is substantially
longer than its diameter; and there is no acceleration of fluid in the pipe. For velocities
and pipe diameters above a threshold, actual fluid flow is not laminar but turbulent,
leading to larger pressure drops than calculated by the Hagen–Poiseuille equation.
Poiseuille's equation describes the pressure drop due to the viscosity of the fluid; other
types of pressure drops may still occur in a fluid (see a demonstration here). [4] For
example, the pressure needed to drive a viscous fluid up against gravity would contain
both that as needed in Poiseuille's law plus that as needed in Bernoulli's equation, such
that any point in the flow would have a pressure greater than zero (otherwise no flow
would happen).
Another example is when blood flows into a narrower constriction, its speed will be
greater than in a larger diameter (due to continuity of volumetric flow rate), and its
pressure will be lower than in a larger diameter (due to Bernoulli's equation). However,
the viscosity of blood will cause additional pressure drop along the direction of flow,
which is proportional to length traveled (as per Poiseuille's law). Both effects contribute
to the actual pressure drop.

Equation

In standard fluid-kinetics notation:

where:

 Δp is the pressure difference between the two ends,

 L is the length of pipe,
 μ is the dynamic viscosity,
 Q is the volumetric flow rate,
 R is the pipe radius,
 A is the cross sectional area of pipe.
The equation does not hold close to the pipe entrance.
The equation fails in the limit of low viscosity, wide and/or short pipe. Low viscosity
or a wide pipe may result in turbulent flow, making it necessary to use more complex
models, such as the Darcy–Weisbach equation. The ratio of length to radius of a
pipe should be greater than one forty-eighth of the Reynolds number for the Hagen–
Poiseuille law to be valid.[9] If the pipe is too short, the Hagen–Poiseuille equation
may result in unphysically high flow rates; the flow is bounded by Bernoulli's
principle, under less restrictive conditions, by
because it is impossible to have negative (absolute) pressure (not to be confused
with gauge pressure) in an incompressible flow.

where Re is the Reynolds number, ρ is the fluid density, and v is the mean flow velocity,
which is half the maximal flow velocity in the case of laminar flow. It proves more useful
to define the Reynolds number in terms of the mean flow velocity because this quantity
remains well defined even in the case of turbulent flow, whereas the maximal flow
velocity may not be, or in any case, it may be difficult to infer. In this form the law
approximates the Darcy friction factor, the energy (head) loss factor, friction loss
factor or Darcy (friction) factor Λ in the laminar flow at very low velocities in cylindrical
tube. The theoretical derivation of a slightly different form of the law was made
independently by Wiedman in 1856 and Neumann and E. Hagenbach in 1858 (1859,
1860). Hagenbach was the first who called this law Poiseuille's law.
The law is also very important in hemorheology and hemodynamics, both fields
of physiology.[10]
Poiseuille's law was later in 1891 extended to turbulent flow by L. R. Wilberforce, based
on Hagenbach's work.

Relation to the Darcy–Weisbach equation

Normally, Hagen–Poiseuille flow implies not just the relation for the pressure drop,
above, but also the full solution for the laminar flow profile, which is parabolic. However,
the result for the pressure drop can be extended to turbulent flow by inferring an
effective turbulent viscosity in the case of turbulent flow, even though the flow profile in
turbulent flow is strictly speaking not actually parabolic. In both cases, laminar or
turbulent, the pressure drop is related to the stress at the wall, which determines the so-
called friction factor. The wall stress can be determined phenomenologically by
the Darcy–Weisbach equation in the field of hydraulics, given a relationship for the
friction factor in terms of the Reynolds number. In the case of laminar flow, for a circular
cross section:

Derivation
The Hagen–Poiseuille equation can be derived from the Navier–Stokes equations.
The laminar flow through a pipe of uniform (circular) cross-section is known as Hagen–
Poiseuille flow. The equations governing the Hagen–Poiseuille flow can be derived
directly from the Navier–Stokes momentum equations in 3D cylindrical
coordinates (r,θ,x) by making the following set of assumptions:

1. The flow is steady ( ∂.../∂t = 0 ).

2. The radial and azimuthal components of the fluid velocity are zero ( ur = uθ = 0 ).
3. The flow is axisymmetric ( ∂.../∂θ = 0 ).
4. The flow is fully developed ( ∂ux/∂x = 0 ). Here However, this can be proved via
mass conservation, and the above assumptions.
Then the angular equation in the momentum equations and the continuity equation are
identically satisfied. The radial momentum equation reduces to ∂p/∂r = 0, i.e.,
the pressure p is a function of the axial coordinate x only. For brevity, use u instead
of The axial momentum equation reduces to
where μ is the dynamic viscosity of the fluid. In the above equation, the left-hand side is
only a function of r and the right-hand side term is only a function of x, implying that both
terms must be the same constant. Evaluating this constant is straightforward. If we take
the length of the pipe to be L and denote the pressure difference between the two ends
of the pipe by Δp (high pressure minus low pressure), then the constant is simply
defined such that G is positive. The solution is

Since u needs to be finite at r = 0, c1 = 0. The no slip boundary condition at the pipe wall
requires that u = 0 at r = R (radius of the pipe), which yields c2 = GR2/4μ. Thus we have
finally the following parabolic velocity profile:

The maximum velocity occurs at the pipe centerline (r = 0), umax = GR2/4μ. The average
velocity can be obtained by integrating over the pipe cross section,

The easily measurable quantity in experiments is the volumetric flow rate Q = πR2 uavg.
Rearrangement of this gives the Hagen–Poiseuille equation

Elaborate derivation starting directly from first principles

Startup of Poiseuille flow in a pipe
When a constant pressure gradient G = −dp/dx is applied between two ends of a long
pipe, the flow will not immediately obtain Poiseuille profile, rather it develops through
time and reaches the Poiseuille profile at steady state. The Navier–Stokes
equations reduce to with initial and boundary conditions,
The velocity distribution is given by

where J0(λnr/R) is the Bessel function of the first kind of order zero and λn are the
positive roots of this function and J1(λn) is the Bessel function of the first kind of
order one. As t → ∞, Poiseuille solution is recovered.[11]
Poiseuille flow in an annular section[edit]

Poiseuille flow in annular section

If R1 is the inner cylinder radii and R2 is the outer cylinder radii, with constant applied
pressure gradient between the two ends G = −dp/dx, the velocity distribution and the
volume flux through the annular pipe are

When R2 = R, R1 = 0, the original problem is recovered

Poiseuille flow in a pipe with an oscillating pressure gradient

Flow through pipes with an oscillating pressure gradient finds applications in blood
flow through large arteries. The imposed pressure gradient is given by

where G, α and β are constants and ω is the frequency. The velocity field is given by

where

where ber and bei are the Kelvin functions and k2 = ρω/μ.

Plane Poiseuille flow

Plane Poiseuille flow

Plane Poiseuille flow is flow created between two infinitely long parallel plates,
separated by a distance h with a constant pressure gradient G = −dp/dx is applied in the
direction of flow. The flow is essentially unidirectional because of infinite length.
The Navier–Stokes equations reduce to

with no-slip condition on both walls

Therefore, the velocity distribution and the volume flow rate per unit length are

Poiseuille flow through some non-circular cross-sections

Joseph Boussinesq derived the velocity profile and volume flow rate in 1868 for
rectangular channel and tubes of equilateral triangular cross-section and for elliptical
cross-section. Joseph Proudman derived the same for isosceles triangles in
1914.] Let G = −dp/dx be the constant pressure gradient acting in direction parallel to the
motion.
The velocity and the volume flow rate in a rectangular channel of height 0 ≤ y ≤ h and
width 0 ≤ z ≤ l are

The velocity and the volume flow rate of tube with equilateral triangular cross-section of
side length 2h/√3 are
The velocity and the volume flow rate in the right-angled isosceles triangle y = π, y ± z =
0 are

The velocity distribution for tubes of elliptical cross-section with semiaxes a and b is

Here, when a = b, Poiseuille flow for circular pipe is recovered and when a → ∞, plane
Poiseuille flow is recovered. More explicit solutions with cross-sections such as snail-
shaped sections, sections having the shape of a notch circle following a semicircle,
annular sections between homofocal ellipses, annular sections between non-concentric
circles are also available, as reviewed by Ratip Berker.

Poiseuille flow through arbitrary cross-section

The flow through arbitrary cross-section u(y,z) satisfies the condition that u = 0 on the
walls. The governing equation reduces to

If we introduce a new dependent variable as

then it is easy to see that the problem reduces to that integrating a Laplace equation

satisfying the condition

on the wall.
Poiseuille's equation for an ideal isothermal gas
For a compressible fluid in a tube the volumetric flow rate Q(x) (but not the mass flow
rate) and the axial velocity are not constant along the tube. The flow is usually
expressed at outlet pressure. As fluid is compressed or expanded, work is done and the
fluid is heated or cooled. This means that the flow rate depends on the heat transfer to
and from the fluid. For an ideal gas in the isothermal case, where the temperature of the
fluid is permitted to equilibrate with its surroundings, an approximate relation for the
pressure drop can be derived. Using ideal gas equation of state for constant
temperature process, the relation Qp = Q1p1 = Q2p2 can be obtained. Over a short section
of the pipe, the gas flowing through the pipe can be assumed to be incompressible so
that Poiseuille law can be used locally,

Here we assumed the local pressure gradient is not too great to have any
compressibility effects. Though locally we ignored the effects of pressure variation due
to density variation, over long distances these effects are taken into account. Since μ is
independent of pressure, the above equation can be integrated over the length L to give

Hence the volumetric flow rate at the pipe outlet is given by

This equation can be seen as Poiseuille's law with an extra correction

factor p1 + p2/2p2  expressing the average pressure relative to the outlet pressure.

Darcy–Weisbach equation

In fluid dynamics, the Darcy–Weisbach equation is an empirical equation that relates

the head loss, or pressure loss, due to friction along a given length of pipe to the
average velocity of the fluid flow for an incompressible fluid. The equation is named
after Henry Darcy and Julius Weisbach. Currently, there is no formula more accurate or
universally applicable than the Darcy-Weisbach supplemented by the Moody
diagram or Colebrook equation.
The Darcy–Weisbach equation contains a dimensionless friction factor, known as
the Darcy friction factor. This is also variously called the Darcy–Weisbach friction
factor, friction factor, resistance coefficient, or flow coefficient
Pressure-loss equation
In a cylindrical pipe of uniform diameter D, flowing full, the pressure loss due to viscous
effects Δp is proportional to length L and can be characterized by the Darcy–Weisbach
equation
where the pressure loss per unit length Δp/L (SI units: Pa/m) is a function of:
ρ, the density of the fluid (kg/m3);
DH, the hydraulic diameter of the pipe (for a pipe of circular section, this equals D;
otherwise DH = 4A/P for a pipe of cross-sectional area A and perimeter P) (m);
v, the mean flow velocity, experimentally measured as the volumetric flow
rate Q per unit cross-sectional wetted area (m/s);
fDthe Darcy friction factor (also called flow coefficient λ).

For laminar flow in a circular pipe of diameter  , the friction factor is inversely

proportional to the Reynolds number alone (fD = 64/Re) which itself can be expressed in
terms of easily measured or published physical quantities (see section below). Making
this substitution the Darcy–Weisbach equation is rewritten as

where
μ is the dynamic viscosity of the fluid (Pa·s = N·s/m2 = kg/(m·s));
Q is the volumetric flow rate, used here to measure flow instead of mean velocity
according to Q = π/4Dc2<v> (m3/s).
Note that this laminar form of Darcy–Weisbach is equivalent to the Hagen–Poiseuille
equation, which is analytically derived from the Navier–Stokes equations.

Head-loss form
The head loss Δh (or hf) expresses the pressure loss due to friction in terms of the
equivalent height of a column of the working fluid, so the pressure drop is

where:
Δh = The head loss due to pipe friction over the given length of pipe (SI units: m);
[b]

g = The local acceleration due to gravity (m/s2).

It is useful to present head loss per length of pipe (dimensionless):

where L is the pipe length (m).

Therefore, the Darcy–Weisbach equation can also be written in terms of head loss: [6]
In terms of volumetric flow
The relationship between mean flow velocity <v> and volumetric flow rate Q is
 where:
Q = The volumetric flow (m3/s),
A = The cross-sectional wetted area (m2).

In a full-flowing, circular pipe of diameter  ,

Then the Darcy–Weisbach equation in terms of Q is

Shear-stress form
The mean wall shear stress τ in a pipe or open channel is expressed in terms of the
Darcy–Weisbach friction factor as[7]

The wall shear stress has the SI unit of pascals (Pa).

Colebrook-White Formula

The Colebrook-White formula:

 
1/sqrt(f) = 1.14 – 2 log10 [e/D + 9.35/(Re x sqrt(f))]
 
f = friction factor
 
e = internal roughness of pipe
 
D = internal diameter of pipe
 
Re = Reynolds number
 
Friction factors for turbulent flow calculated by Pipe Flow Expert are based on the
Colebrook-White formula.
 
The friction factor for Laminar flow is calculated from f = 64/Re

Minor losses in pipes

Minor losses in pipe flow are a major part in calculating the flow, pressure, or energy
reduction in piping systems. Liquid moving through pipes carries momentum and energy
due to the forces acting upon it such as pressure and gravity. Just as certain aspects of
the system can increase the fluids energy, there are components of the system that act
against the fluid and reduce its energy, velocity, or momentum. Friction and minor
losses in pipes are major contributing factors

Friction Losses[edit]
Before being able to use the minor head losses in an equation, the losses in the system
due to friction must also be calculated.
Equation for friction losses:

HLf= Frictional head loss

v= Downstream velocity
g= Gravity of Earth
Rh= Hydraulic radius
∑
❑
❑=Total length of piping

f = Fanning friction factor

Total Head Loss

After both minor losses and friction losses have been calculated, these values can be
summed to find the total head loss.
Equation for total head loss, HL, can be simplified and rewritten as:

HL= Frictional head loss

v= Downstream velocity
g = Gravity of Earth
Rh = Hydraulic radius

 =Total length of piping

f = Fanning friction factor

= Sum of all kinetic energy factors in system

Once calculated, the total head loss can be used to solve the Bernoulli Equation and
find unknown values of the system.
Losses due to friction
The flow of fluid through a pipe is resisted due to the viscous shear stresses within the fluid and the
turbulence. These turbulences occur along the internal pipe wall, which will be dependent on the
roughness of the pipe material. Friction losses are a complex function of the system geometry and it
corresponds to the fluid properties and the flow rate in the system. By observation, we may see
that this loss is roughly proportional to the square of the flow rate in most engineering flows. This
topic will deal with friction loss concepts and friction loss formula with examples. Let us learn it!

What is Friction Loss?

Friction may be observed as the resistance required for moving a body through the external surface.
But the friction loss is related to the flow of liquid through some pipe. Thus, it is a kind of energy
loss due to the friction inside the tube. It is therefore related to the velocity and viscosity of the fluid.
Friction loss can be computed as hl as friction loss is nothing but the energy loss. This resistance is
termed as pipe friction and is measured in meters head of the fluid. Many types of research have
been carried out to establish various formulae to calculate this loss.
The Darcy formula or the Darcy-Weisbach equation mainly tend to be referred for this computation.
And is now accepted as the most accurate pipe friction loss formula. Although it is more difficult to
calculate and use than other friction loss formulas, with the introduction of computers, it has
become the standard equation for hydraulic engineers.

Formula for Friction Loss

Friction loss formula is: hl=f×(L/D)×(v2/2g)

Where,

 f is the friction factor

 L is the length of the pipe

 D is the inner diameter of the pipe

 V is the velocity of the liquid

 g is the gravitational constant

 hl is the friction lost

Reynolds Number is the basic dimensionless group in viscous flow. Velocity times Length Scale
divided by Kinematic Viscosity. Relative Roughness is related to the height of a typical roughness
element to the scale of the flow.

Example: 1: Find out the friction loss, if the inner diameter and length of the pipe are 0.3 m and 30
m respectively. Also, the friction factor and velocity of the liquid are 0.4 and 25 m per sec?

Solution: Given parameters,

 Length of the pipe, L = 30 m

 internal diameter, D = 0.3 m

 velocity of the liquid, v = 25 m per sec

 friction factor, f = 0.4

 g=9.8mpersqsec2
Now, friction loss formula is,

hl=f×L/D×v2/2g
=0.4×300.3×2522×9.8
=1275.51m.