A.Y.

2023-2024 ASE 2231

Laminar Flow and Turbulent Flow

Flow in a conduit is classified as being either laminar or
turbulent, depending on the magnitude of the Reynolds

Flow in Conduits number. The original research involved visualizing flow

in a glass tube, as shown in (a). In the 1880s, Reynolds
injected dye into the center of the tube.

(Laminar)
When the velocity was low, the streak of dye flowed
down the tube with little expansion, as shown in (b).
However, if the water in the tank was disturbed, the
streak would shift about in the tube.

If velocity was increased, at some point in the tube the

dye would all at once mix with the water, as shown in (c).
Justine Clyde Espinosa
When the dye exhibited rapid mixing, illumination with
an electric spark revealed eddies in the mixed fluid (d).

Laminar Flow and Turbulent Flow Laminar Flow and Turbulent Flow
The flow regimes shown are laminar flow (b) and Reynolds number can be calculated with four different equations. These
turbulent flow (c) and (d). Reynolds showed that the equations are equivalent because one can start with one formula and derive
onset of turbulence was related to a 𝜋-group that is now
called the Reynolds number (𝑅𝑒 = 𝜌𝑉𝐷/𝜇) in honor of
the others. The formulas are
Reynolds’ pioneering work.
𝑉𝐷 𝜌𝑉𝐷 4𝑄 4𝑚ሶ
The Reynolds number is often written as 𝑅𝑒𝐷 , where the
𝑅𝑒𝐷 = = = =
𝜈 𝜇 𝜋𝐷𝜈 𝜋𝐷𝜇
subscript “𝐷” denotes that diameter is used in the
formula. This subscript is called a length scale. Indicating
Reynolds discovered that if the fluid in the upstream reservoir was not
the length scale for the Reynolds number is good
practice because multiple values are used. completely still or if the pipe had some vibrations, then the change from
laminar to turbulent flow occurred at 𝑅𝑒𝐷 ∼ 2100. However, if conditions were
ideal, then it was possible to reach a much higher Reynolds number before the
flow became turbulent.
Laminar Flow and Turbulent Flow Laminar Flow and Turbulent Flow
Reynolds also found that, when going from high velocity to low velocity, the The middle range (2000 ≤ 𝑅𝑒𝐷 ≤ 3000) corresponds to a type of flow
change back to laminar flow occurred at 𝑅𝑒𝐷 ∼ 2000. Based on Reynolds’ that is unpredictable because it can change back and forth between
experiments, engineers use guidelines to establish whether or not flow in a laminar and turbulent states. Recognize that precise values of the
conduit will be laminar or turbulent. The guidelines used in this text are as
Reynolds number versus flow regime do not exist.
follows:

Thus, the guidelines given in the tables are approximate, and other
𝑅𝑒𝐷 ≤ 2000 Laminar Flow
references may give different values. For example, some references
2000 ≤ 𝑅𝑒𝐷 ≤ 3000 Unpredictable use 𝑅𝑒𝐷 = 2300 as the criteria for turbulence.
𝑅𝑒𝐷 ≥ 3000 Turbulent Flow

Developing and Fully Developed Flow Developing and Fully Developed Flow
Flow in a conduit is classified as either developing flow or The distance required for flow to develop is called the
fully developed flow. For example, consider laminar fluid entry or entrance length (𝐿𝑒 ). In the entry length, the wall
entering a pipe from a reservoir. shear stress is decreasing in the streamwise (i.e., 𝑠)
direction. For laminar flow, the wall shear stress
As the fluid moves down the pipe, the velocity profile distribution is shown in the figure. Near the pipe
changes in the streamwise direction as viscous effects entrance, the radial velocity gradient is high, so the shear
cause the plug-type profile to gradually change into a stress is large.
parabolic profile.
As the velocity profile progresses to a parabolic shape,
the velocity gradient and the wall shear stress decrease
This region of changing velocity profile is called until a constant value is achieved. The entry length is
developing flow. After the parabolic distribution is defined as the distance at which the shear stress reaches
achieved, the flow profile remains unchanged in the 2% of the fully developed value.
streamwise direction, and flow is called fully developed
flow.
Developing and Fully Developed Flow Consider fluid flowing in a round tube of length 1 𝑚 and diameter 5 𝑚𝑚. Classify the flow
as laminar or turbulent and calculate the entrance length for (a) air (50°𝐶) with a speed of
12 𝑚/𝑠 and (b) water (15°𝐶) with a mass flow rate of 𝑚ሶ = 8 𝑔/𝑠. 𝜈𝑎𝑖𝑟 = 1.79 × 10−5 𝑚2 /𝑠 and
Correlations for entry length are 𝜇𝑤𝑎𝑡𝑒𝑟 = 1.14 × 10−3 𝑁 ∙ 𝑠/𝑚2 .
𝐿𝑒𝐷
= 0.05𝑅𝑒𝐷 𝑙𝑎𝑚𝑖𝑛𝑎𝑟 𝑓𝑙𝑜𝑤: 𝑅𝑒𝐷 ≤ 2000
𝐷 𝑉𝐷 4𝑚ሶ
𝐿𝑒𝐷 𝑅𝑒𝑎𝑖𝑟 = 𝑅𝑒𝑤𝑎𝑡𝑒𝑟 =
= 50 (𝑡𝑢𝑟𝑏𝑢𝑙𝑒𝑛𝑡 𝑓𝑙𝑜𝑤: 𝑅𝑒𝐷 ≤ 3000) 𝜈 𝜋𝐷𝜇
𝐷
(12)(0.005) (4)(0.008)
The equations are valid for flow entering a circular pipe
𝑅𝑒𝑎𝑖𝑟 = = 3350 𝑅𝑒𝑤𝑎𝑡𝑒𝑟 = = 1787
1.79 × 10−5 𝜋 0.005 (1.14 × 10−3 )
from a reservoir under quiescent conditions. Other 𝑹𝒆𝒂𝒊𝒓 > 𝟑𝟎𝟎𝟎 (𝒕𝒖𝒓𝒃𝒖𝒍𝒆𝒏𝒕)
upstream components, such as valves, elbows, and
𝑹𝒆𝒘𝒂𝒕𝒆𝒓 < 𝟐𝟎𝟎𝟎 (𝒍𝒂𝒎𝒊𝒏𝒂𝒓)
pumps, produce complex flow fields that require
𝐿𝑒 = 50𝐷
different lengths to achieve fully developing flow. 𝐿𝑒 = 0.05𝑅𝑒𝐷
𝐿𝑒 = 50 0.005 𝐿𝑒 = 0.05(1787)(0.005)
In summary, flow in a conduit is classified into four 𝑳𝒆 = 𝟎. 𝟐𝟓 𝒎 𝑳𝒆 = 𝟎. 𝟒𝟒𝟕 𝒎
categories: laminar developing, laminar fully developed,
turbulent developing, or turbulent fully developed.

Pipe Head Loss

Pipe head loss is associated with fully developed flow in conduits, and
it is caused by shear stresses that act on the flowing fluid. Note that
pipe head loss is sometimes called major head loss, and component
head loss is sometimes called minor head loss.

ℎ𝑒𝑎𝑑 𝑙𝑜𝑠𝑠 4𝐿𝜏0

ℎ𝑓 = =
𝑖𝑛 𝑎 𝑝𝑖𝑝𝑒 𝐷𝛾
Darcy-Weisbach Equation Darcy-Weisbach Equation
To use the Darcy-Weisbach equation, the flow should be fully There is also another coefficient called the Fanning friction factor,
developed and steady. The Darcy-Weisbach equation is used for often used by chemical engineers, which is related to the Darcy-
either laminar flow or turbulent flow and for either round pipes or Weisbach friction factor by a factor of 4:
nonround conduits such as a rectangular duct.
𝑓𝐷𝑎𝑟𝑐𝑦 = 4𝑓𝐹𝑎𝑛𝑛𝑖𝑛𝑔
4𝜏0 𝑠ℎ𝑒𝑎𝑟 𝑠𝑡𝑟𝑒𝑠𝑠 𝑎𝑐𝑡𝑖𝑛𝑔 𝑎𝑡 𝑡ℎ𝑒 𝑤𝑎𝑙𝑙
𝑓≡ ≈
2
𝜌𝑉 /2 𝑘𝑖𝑛𝑒𝑡𝑖𝑐 𝑝𝑟𝑒𝑠𝑠𝑢𝑟𝑒 The Darcy-Weisbach equation, in relation to pipe head loss, is defined
as
This is known as the Darcy friction factor or the resistance coefficient.
𝐿 𝑉2
ℎ𝑓 = 𝑓
𝐷 2𝑔

Stress Distributions in Pipe Flow Stress Distributions in Pipe Flow

In pipe flow, the pressure acting on a plane that is Apply the momentum equation in the 𝑠 direction. The
normal to the direction of flow is hydrostatic. This means net momentum efflux is zero because the flow is fully
that the pressure distribution varies linearly, as shown in developed; that is, the velocity distribution at the inlet is
the upper figure. The reason that the pressure the same as the velocity distribution at the exit. The
distribution is hydrostatic can be explained with Euler’s momentum accumulation is also zero because the flow
equation. is steady. The momentum equation simplifies to force
equilibrium:
To derive an equation for the shear stress variation,
consider flow of a Newtonian fluid in a round tube that is ∑𝐹𝑠 = 𝐹𝑝𝑟𝑒𝑠𝑠𝑢𝑟𝑒 + 𝐹𝑤𝑒𝑖𝑔ℎ𝑡 + 𝐹𝑠ℎ𝑒𝑎𝑟 = 0
inclined at an angle 𝛼 with respect to the horizontal, as 𝑑𝑝
shown in the upper figure. Assume that the flow is fully 𝑝𝐴 − 𝑝 + Δ𝐿 𝐴 − 𝑊 sin 𝛼 − 𝜏 2𝜋𝑟 ΔL = 0
𝑑𝑠
developed, steady, and laminar. Define a cylindrical
control volume of length Δ𝐿 and radius 𝑟. Let 𝑊 = 𝛾𝐴Δ𝐿 and sin 𝛼 = Δ𝑧/Δ𝐿

𝑟 𝑑
𝜏= − (𝑝 + 𝛾𝑧)
2 𝑑𝑠
Stress Distributions in Pipe Flow Laminar Flow in a Round Tube
The equation shows that the shear stress distribution Laminar flow is a flow regime in which fluid motion is smooth, the flow
varies linearly with 𝑟, as shown in the figure.
occurs in layers (laminae), and the mixing between layers occurs by
Notice that the shear stress is zero at the centerline, it molecular diffusion, a process that is much slower than turbulent
reaches a maximum value of 𝜏0 at the wall, and the mixing.
variation is linear in between.

This linear shear stress variation applies to both laminar Laminar flow occurs when 𝑅𝑒𝐷 ≤ 2000. Laminar flow in a round tube is
and turbulent flow. called Poiseuille flow or Hagen-Poiseuille flow in honor of researchers
who studied low-speed flows in the 1840s.

Equations for Laminar Flow (Round Tube) Oil (𝑆𝐺 = 0.85) with a kinematic viscosity of 6 × 10–4 𝑚2 /𝑠 flows in a 15 𝑐𝑚 diameter pipe at
a rate of 0.020 𝑚3 /𝑠. What is the head loss for a 100 𝑚 length of pipe?

2 𝑄 0.02 0.02
𝑟02 − 𝑟 2 𝛾Δℎ 𝑟 𝑉= = = = 1.13 𝑚/𝑠
Velocity Profile 𝑉 𝑟 =− = 𝑉𝑚𝑎𝑥 1 − 𝐴 (𝜋𝐷2 )/4 𝜋 0.15 2 /4
4𝜇 Δ𝐿 𝑟0 𝑉𝐷 (1.13)(0.15)
𝑅𝑒𝐷 = = = 283
𝐷2 𝛾Δℎ 𝑉𝑚𝑎𝑥 𝜈 6 × 10−4
Mean Velocity 𝑉ത = − = 𝑅𝑒𝐷 < 2000 (𝐿𝑎𝑚𝑖𝑛𝑎𝑟)
32𝜇 Δ𝐿 2
32𝜇𝐿𝑉 32𝜌𝜈𝐿𝑉 32𝜈𝐿𝑉
32𝜇𝐿𝑉 𝐿 𝑉2 ℎ𝑓 = = =
Head Loss ℎ𝑓 = = 𝑓 𝛾𝐷2 𝜌𝑔𝐷2 𝑔𝐷2
𝛾𝐷2 𝐷 2𝑔 −4
32(6 × 10 )(100)(1.13)
ℎ𝑓 =
32𝜇𝐿𝑉 𝐷 2𝑔 64𝜇 64 9.81 0.15 2
Friction Factor 𝑓= = = 𝒉𝒇 = 𝟗. 𝟖𝟑 𝒎
𝛾𝐷2 𝐿 𝑉 2 𝜌𝐷𝑉 𝑅𝑒𝐷
A.Y. 2023-2024 ASE 2231

Velocity Distribution
The time-average velocity distribution is often described using an

Flow in Conduits
equation called the power-law formula.
𝑚
𝑢 𝑟 𝑟0 − 𝑟
=

(Turbulent)
𝑢𝑚𝑎𝑥 𝑟0

where 𝑢𝑚𝑎𝑥 is velocity in the center of the pipe, 𝑟0 is the pipe radius,
and 𝑚 is an empirically determined variable that depends on 𝑅𝑒

Justine Clyde Espinosa

Velocity Distribution
An alternative approach to the equation is to use the turbulent
boundary-layer equations in surface resistance. The most significant
of these equations, called the logarithmic velocity distribution, is given
by

𝑢(𝑟) 𝑢∗ (𝑟0 − 𝑟)
= 2.44 ln + 5.56
𝑢∗ 𝜈

Where 𝑢∗ is the shear velocity and is given by 𝑢∗ = 𝜏0 /𝜌

Friction Factor, 𝒇 Friction Factor, 𝒇
The equation for the friction factor 𝑓 in turbulent flow is defined as Nikuradse’s data shows the friction factor 𝑓 plotted as a
function of Reynolds number for various sizes of sand
the following grains. To characterize the size of sand grains, Nikuradse
used a variable called the sand roughness height with
1 the symbol 𝑘𝑠 . The 𝜋-group, 𝑘𝑠 /𝐷, is given the name
= 2.0 log10 𝑅𝑒 𝑓 − 0.8 relative roughness.
𝑓
In laminar flow, the data in the figure show that wall
roughness does not influence 𝑓. In particular, notice how
The equation, first derived by Prandtl in 1935, gives the resistance the data corresponding to various values of 𝑘𝑠 /𝐷
coefficient for turbulent flow in tubes that have smooth walls. To collapse into a single blue line that is labeled “laminar
determine the influence of roughness on the walls, Nikuradse, one of flow.”
Prandtl’s graduate students, glued uniform-sized grains of sand to the
inner walls of a tube and then measured pressure drops and flow
rates.

Friction Factor, 𝒇
In turbulent flow, the data in he figure show that wall
roughness has a major impact on 𝑓. When, 𝑘𝑠 /𝐷 = 0.033,
then values of 𝑓 are about 0.04. As the relative
roughness drops to 0.002, values of 𝑓 decrease by a
factor of about 3.

Eventually wall roughness does not matter, and the

value of 𝑓 can be predicted by assuming that the tube
has a smooth wall. This latter case corresponds to the
blue curve in that is labeled “smooth wall tube.”
Friction Factor, 𝒇 Friction Factor, 𝒇
In the Moody diagram, the abscissa is the Reynolds To provide a more convenient solution to some types of
number 𝑅𝑒, and the ordinate is the resistance coefficient problems, the top of the Moody diagram presents a
𝑓. Each blue curve is for a constant relative roughness scale based on the parameter 𝑅𝑒𝑓 1/2 . This parameter is
𝑘𝑠 /𝐷 and the values of 𝑘𝑠 /𝐷 are given on the right at the useful when ℎ𝑓 and 𝑘𝑠 /𝐷 are known but the velocity 𝑉 is
end of each curve. To find 𝑓, given 𝑅𝑒 and 𝑘𝑠 /𝐷 one goes not. Using the Darcy-Weisbach equation and the
to the right to find the correct relative roughness curve. definition of Reynolds number, one can show that
Then one looks at the bottom of the chart to find the
given value of 𝑅𝑒 and, with this value of 𝑅𝑒, moves 𝐷3/2 1/2
vertically upward until the given 𝑘𝑠 /𝐷 curve is reached. 𝑅𝑒𝑓 1/2 = 2𝑔ℎ𝑓 /𝐿
𝜈
Finally, from this point one moves horizontally to the left
scale to read the value of 𝑓. If the curve for the given
value of 𝑘𝑠 /𝐷 is not plotted in the Moody diagram, then
one simply finds the proper position on the graph by
interpolation between the 𝑘𝑠 /𝐷 curves that bracket the
given 𝑘𝑠 /𝐷.
Friction Factor, 𝒇 Solving Turbulent Flow Problems
In the Moody diagram curves of constant 𝑅𝑒𝑓 1/2 are To recognize problems that require iterative approaches or computer solutions,
plotted using heavy black lines that slant from the left to engineers classify problems into three cases based on the goal of the problem
right. For example, when 𝑅𝑒𝑓 1/2 = 105 and 𝑘𝑠 /𝐷 = 0.004, and based on what information is known.
then 𝑓 = 0.029. When using computers to carry out pipe-
flow calculations, it is much more convenient to have an
equation for the friction factor as a function of Reynolds Case 1: is when the goal is to find the head loss, given the pipe length, pipe
number and relative roughness, which is defined as the diameter, and flow rate. This problem is straightforward because it can be solved
following. using algebra
Case 2: is when the goal is to find the flow rate, given the head loss (or pressure
𝑓=
0.25 drop), the pipe length, and the pipe diameter. This problem usually requires an
2
log10
𝑘𝑠
+
5.74 iterative approach or solver program
3.7𝐷 𝑅𝑒 0.9 Case 3: is when the goal is to find the pipe diameter, given the flow rate, length of
This equation predicts friction factors that differ by less
pipe, and head loss (or pressure drop). This problem usually requires an iterative
than 3% from those on the Moody diagram for 4 × 103 < approach or a solver program
𝑅𝑒 < 108 and 10–5 < 𝑘𝑠 /𝐷 < 2 × 10–2 .

Solving Turbulent Flow Problems Water (𝑇 = 20°𝐶) flows at a rate of 0.05 𝑚3 /𝑠 in a 20 𝑐𝑚 asphalted cast-iron pipe. What is
the head loss per kilometer of pipe? 𝜈 = 1 × 10−6 𝑚2 /𝑠, 𝑘𝑠 = 0.12 𝑚𝑚

There are several approaches that sometimes eliminate the need for an iterative
approach. For case 2, an iterative approach can sometimes be eliminated by using 𝑉=
𝑄
=
0.05
= 1.59 𝑚/𝑠
an explicit equation developed by Swamee and Jain 𝐴 𝜋/4 0.02 2
𝑉𝐷 1.59(0.2)
𝑅𝑒 = = = 3.18 × 105
𝑘𝑠 1.78𝜈 𝜈 10−6
𝑄 = −2.22𝐷 5/2 𝑔ℎ𝑓 /𝐿 log + 𝑘𝑠 0.00012
3.7𝐷 𝐷 3/2 𝑔ℎ𝑓 /𝐿 = = 0.0006
𝐷 0.2

Using the equation is equivalent to using the top of the Moody diagram, which From Moody Diagram: 𝑓 = 0.019
presents a scale for 𝑅𝑒𝑓 1/2 . For case 3, one can sometimes use an explicit equation 𝐿 𝑉2
ℎ𝑓 = 𝑓
developed by Swamee and Jain and modified by Streeter and Wylie 𝐷 2𝑔
1000 1.592
ℎ𝑓 = 0.0019
4.75 5.2 0.04 0.2 2(9.81)
𝐿𝑄 2 𝐿
𝐷 = 0.66 𝑘𝑠1.25 + 𝜈𝑄 9.4 𝒉𝒇 = 𝟏𝟐. 𝟐 𝒎
𝑔ℎ𝑓 𝑔ℎ𝑓
 Water (𝑇 = 20°𝐶) flows from a tank through a 50 𝑐𝑚 diameter
steel pipe. Determine the discharge of water. 𝜈 = 1 × 10−6 𝑚2 /𝑠,
Minor Loss Coefficient, 𝑲
𝑘𝑠 = 0.046 𝑚𝑚
When fluid flows through a component such as a
partially open value or a bend in a pipe, viscous
𝑘𝑠 0.000046 Iteration 1: 𝐺𝑢𝑒𝑠𝑠 𝑓 = 0.020 Iteration 2: 𝑈𝑠𝑒 𝑉 = 10.7 𝑚/𝑠 effects cause the flowing fluid to lose mechanical energy.
= = 9.2 × 10−5 For example, the figure shows flow through a “generic
𝐷 0.5
𝑉𝐷 10.7(0.5) component.” At section 2, the head of the flow will be
1/2
2(9.81)(20) 𝑅𝑒 = = = 5.35 × 106
𝑉=
𝑝1 𝑉12 𝑝2 𝑉22 1 + 200(0.02) 𝜈 10−6 less than at section 1. To characterize component head
+ + 𝑧1 = + + 𝑧2 + ℎ𝐿 From Moody Diagram: 𝑓 = 0.012
𝛾 2𝑔 𝛾 2𝑔 𝑉 = 8.86 𝑚/𝑠 loss, engineers use a 𝜋-group called the minor loss
0 + 0 + 60 = 0 +
𝑉22
+ 40 + 𝑓
𝐿 𝑉22
𝑅𝑒 =
𝑉𝐷 8.86(0.5)
= = 4.43 × 106 Since 𝑓 = 0.012 is unchanged, no need for coefficient 𝐾
2𝑔 𝐷 2𝑔 𝜈 10−6 more iterations
1/2
2𝑔 × 20 Δℎ Δ𝑝
𝑉= From Moody Diagram: 𝑓 = 0.012 𝐾≡ =
1 + 200𝑓 1/2 𝑄 = 𝑉𝐴 𝑉 2 /2𝑔 𝜌𝑉 2 /2
2(9.81)(20) 𝜋 2
𝑉= 𝑄 = 10.7 0.50
1 + 200(0.012) 4
𝑸 = 𝟐. 𝟏𝟎 𝒎𝟑 /𝒔 where Δℎ is drop in piezometric head that is caused by a
𝑉 = 10.7 𝑚/𝑠
component, is the pressure drop that is caused by the
component, Δ𝑝 and 𝑉 is mean velocity

Minor Loss Coefficient, 𝑲

Thus, the head loss across a single component or
transition is ℎ𝐿 = 𝐾 𝑉 2 / 2𝑔 , where 𝐾 is the minor loss
coefficient for that component or transition.

Most values of 𝐾 are found by experiment. For example,

consider the setup shown. To find 𝐾, flow rate is
measured and mean velocity is calculated using 𝑉 = 𝑄/𝐴.
Pressure and elevation measurements are used to
calculate the change in piezometric head.

𝑝2 𝑝1
Δℎ = ℎ2 − ℎ1 = + 𝑧2 − + 𝑧1
𝛾 𝛾
 Combined Head Loss Equation
The combined head loss equation is defined as the following

𝐿 𝑉2 𝑉2 𝑉2 𝐿
ℎ𝐿 = ෍ 𝑓 + ෍ 𝐾 = ෍ 𝑓 + ෍ 𝐾
𝐷 2𝑔 2𝑔 2𝑔 𝐷
𝑝𝑖𝑝𝑒𝑠 𝑐𝑜𝑚𝑝𝑜𝑛𝑒𝑛𝑡𝑠 𝑝𝑖𝑝𝑒𝑠 𝑐𝑜𝑚𝑝𝑜𝑛𝑒𝑛𝑡𝑠

To apply this equation, follow the same approaches that were used
for solving pipe problems. That is, classify the flow as case 1, 2, or 3
and apply the usual equations: the energy, Darcy-Weisbach, and flow
rate equations.

Nonround Conduits Nonround Conduits

When a conduit has a section area that is noncircular, then engineers modify the The hydraulic diameter of a round pipe is the pipe’s diameter 𝐷. When the Darcy-
Darcy-Weisbach equation to use hydraulic diameter 𝐷ℎ in place of diameter. Weisbach equation is used to calculate head loss, the resistance coefficient f is
found using a Reynolds number based on hydraulic diameter. Use of hydraulic
𝐿 𝑉2 diameter is an approximation. This approximation introduces an uncertainty of 40%
ℎ𝐿 = 𝑓 for laminar flow and 15% for turbulent flow.
𝐷ℎ 2𝑔
4 × 𝑐𝑟𝑜𝑠𝑠 − 𝑠𝑒𝑐𝑡𝑖𝑜𝑛 𝑎𝑟𝑒𝑎
𝐷ℎ = 64
𝑤𝑒𝑡𝑡𝑒𝑑 𝑝𝑒𝑟𝑖𝑚𝑒𝑡𝑒𝑟 𝑓= ± 40% 𝑙𝑎𝑚𝑖𝑛𝑎𝑟 𝑓𝑙𝑜𝑤
𝑅𝑒𝐷ℎ
where the “wetted perimeter” is that portion of the perimeter that is physically 𝑘
touching the fluid. The wetted perimeter of a rectangular duct of dimension 𝐿 × 𝑤 is 𝑓𝑀𝑜𝑜𝑑𝑦 𝑅𝑒𝐷ℎ , ± 15% (𝑡𝑢𝑟𝑏𝑢𝑙𝑒𝑛𝑡 𝑓𝑙𝑜𝑤)
𝐷ℎ
2𝐿 + 2𝑤. Thus, the hydraulic diameter of this duct is:

4 × 𝐿𝑤 2𝐿𝑤
𝐷ℎ ≡ =
2𝐿 + 2𝑤 𝐿 + 𝑤
Nonround Conduits
In addition to hydraulic diameter, engineers also use hydraulic radius, which is
defined as

𝑠𝑒𝑐𝑡𝑖𝑜𝑛 𝑎𝑟𝑒𝑎 𝐷ℎ
𝑅ℎ ≡ =
𝑤𝑒𝑡𝑡𝑒𝑑 𝑝𝑒𝑟𝑖𝑚𝑒𝑡𝑒𝑟 4

Notice that the ratio of 𝑅ℎ to 𝐷ℎ is 1/4 instead of 1/2. While this ratio is not logical, it
is the convention used in the literature and is useful to remember.