Transmission & Distribution of Air: Fan Systems

Chapter 3
Transmission & Distribution of Air
A fan can be used to deliver air to one or more places, to remove air from
one or more areas, or both. Delivery or removal can be achieved by connect-
ing the necessary number of branch ducts to a main duct. In this chapter the
principles of fluid flow are applied to the problem of proportioning the system
so that the required flow will be transmitted through the various duct
branches, and to the equally important problem of achieving the necessary
distribution in the rooms or spaces being served. Additional details of various
systems are given in the chapters dealing with specific applications. Refer to
the chapter on fan systems for information on the relationship of system
characteristics to fan characteristics.
Principles of Duct Design

As in any engineering problem, the best duct design is the one that most
economically produces the desired results. More specifically, in the design of
transmission and distribution systems, the flow must be proportioned as
desired, the combined cost of materials and operation must be minimal, and
no undesirable features should develop during operation.
The location of the fan in any system will be dictated by the direction of
the flow and the desired pressure relations. That is, a supply fan can be used
to pump air into a space, or an exhaust fan can be used to draw air out of the
space. The same through-flow conditions will be obtained, but the pressure
relations will be different. In the first case, there will be a buildup of pressure
in the space, and in the second, a reduction in the space pressure will occur.
Both supply and exhaust fans can be used, in which case the space pressure
will depend on the relative amounts of air handled by each fan; that is, space
pressure will be positive if there is more supply than exhaust, or negative if
there is more exhaust than supply. Assuming the same capacities and end
pressures, the total energy delivered by the fan, or fans, to the air passing
through a given system must be of a certain value whether a supply fan, an
exhaust fan, or both are used.
If one fan is to supply air to several spaces, it must be located upstream
relative to each space. A downstream location relative to each space is
required if a fan is to exhaust air from several spaces.
‹#1999 Howden Buffalo, Inc.

3-2 FAN ENGINEERING
When there must be several branches, the fan should be as centrally

located as possible so that each particle of air will require approximately the
same amount of energy for transport as all others. Only one pressure can exist
at a single point whether it is in free space or in a duct system. This applies at
the junction of any two branches regardless of any differences in branch size,
length, or configuration. Accordingly, the pressure drop along one branch
must equal that along the other. If the branches are not designed to provide
equal pressure drops at the required flow rates, the flow rates will differ from
the design rate. When the available pressure exceeds that required, the flow
through a branch can be reduced to the design value by dampering. This is a
waste of energy but cannot always be avoided. Sometimes, however, the
sizes of the ducts can be changed to balance the pressure drops.
It is often said that the fan must be selected for a pressure sufficient to
overcome the total losses based on the flow through the longest run. This is
just another way of saying that the same pressure must be dissipated by each
portion of the air flowing, regardless of the lengths of the runs. The proper
distribution of airflow can be achieved only by using balancing dampers or by
using appropriate duct sizes in all branches.
There is an optimum duct size that will produce the most economical
balance between owning and operating costs. If owning costs are proportional
to duct weight and operating costs are proportional to pressure loss, each can
be related to velocity. For a given metal thickness, duct weight is
proportional to diameter, which in turn is proportional to the square root of the
velocity V 1 2 , and pressure drop is closely proportional to the square of the
velocity V 2 . If these two costs are totaled at various velocities, a minimum
cost will result at the optimum velocity. The optimum velocity Vopt , or that
velocity which will produce the minimum total of operating and owning costs
for a system as just described, can be determined from
çæ V äã = ç 880x ρ C F ä 13
å 1000â æå fρ t C ãâ
opt m m m
. (3.1)
a 0 p
As expressed by this equation, the optimum velocity is a function of the metal

thickness xm in in., the density of the metal ρ m in lbm/in.3, the first cost of
the metal Cm in $/lbm, the annual owning cost as a fraction of the first cost F,
the Darcy friction factor f , the density of air ρ a in lbm/ft3, the annual
operating time t o in hr, and the cost of power C p in $/hp-hr. The annual
owning cost as a fraction of the first cost can be calculated from
i
F=
1 6
1− 1+ i − n
(3.2)
where i is the annual interest rate and n is the number of years.
© 1999 Howden Buffalo, Inc.

CHAPTER 3 – TRANSMISSION AND DISTRIBUTION OF AIR 3-3
The expression, as given assumes negligible exit loss and perfect fan
efficiency so that it yields only approximate results even for simple straight-
run systems. For more complex systems, it is customary to use different
design velocities in the different branches as required to equalize friction.
The inclusion of fittings and other duct elements such as heaters, etc., which
reduce the fraction of the total system pressure loss due to straight-duct
friction, reduces the accuracy of the results of Equation 3.1.
Usually, the mains and branches will transmit a constant volume of air at
all times. Whenever a system is expected to handle a variable volume, the
duct sizes should reflect the anticipated operation. The fan in such cases must
be capable of delivering the maximum amount even if the ducts are sized for
some lesser amount.
Another special situation where the optimum duct sizes can be based on
reduced volume is that of a double-duct system. Even though each space
served may receive a constant volume of conditioned air (hot plus cold), the
amounts handled by the hot and cold mains individually may fluctuate. As
the number of spaces increases, there is a decrease in the probability that
either main will be required to deliver the maximum amount of conditioned
air at any particular time. The probability may be even further reduced by
using controls to reset the temperature of the air in one or both mains. Ac-
cording to Wilson1 some designers take this into account by arbitrarily sizing
the hot mains for 75% of the total possible capacity. The cold mains are
generally sized to handle 100% of the demand in the extreme downstream
section. Successive upstream sections, each with several branches, can be
sized for a progressively smaller percentage of the demand. Typical values
may be 90%, 80%, and 70% for the third, second, and first quarter sections.
The total pressure requirement must be calculated on the basis of maximum
demand through all sections. The fan and the volume controls must be so
designed that operation under both full and partial load is satisfactory. Refer
to the air-conditioning chapter for a discussion of design and operating
problems.
The optimum velocity based on the owning and operating costs of the
ducts alone is not always the controlling factor in determining duct sizes. In
the so-called conventional systems, which do not use elaborate sound treat-
ment, duct velocities can be limited to rather low values by noise considera-
tions. On the other hand, very high velocities may be justified in cases where
the savings in building costs, etc., due to reduction in duct size more than
offset the increase in power required to move the air through the system.
1
C. M. Wilson, "Handbook on High Velocity Air Distribution Design," Anemostat Corp. of
America, New York, Reprinted from Heating, Piping and Air Conditioning, Chicago,
November 1954, pp. 94-108.

3-4 FAN ENGINEERING
In the design of high-velocity systems it is important to limit the use of

high velocities to those sections where the saving in space justifies the in-
crease in power requirements. Progressively lower velocities will probably be
indicated after each successive take-off. The static-regain method is particu-
larly suitable for sizing the main risers in multi-story building systems. The
equal-friction method or the static-regain method may be indicated for the
horizontal mains on each floor, depending on the particular layouts.
Ranges of design velocities for conventional and high-velocity systems
are given in Table 3.1.
Table 3.1 Design Velocities in fpm

for Conventional and High-Velocity Systems
Conventional High-Velocity
Public Industrial Commercial
Duct Residences Buildings Buildings Buildings
Element Norm. Max. Norm. Max. Norm. Max. Norm. Max.
Main Ducts 700 1200 1000 1600 1500 2200 2500 6000
Branch Ducts 600 1000 800 1300 1000 1800 2000 4500
Outside Air Intakes 500 800 500 900 500 1200 600 1000
Filters 250 300 300 350 350 350 350 350
Heating Coils 450 500 500 600 600 700 600 700
Air Washers 500 500 500 500 500 500 500 500
Cooling Coils 450 450 500 500 500 600 500 500
Adapted from the data of ASHRAE: ASHRAE Guide and Data Book - Equipment, New York,
1969, p. 38.
Although much discussion has been devoted to establishing duct velocities

and duct sizes, the importance of reducing the pressure losses due to elbows
and other fittings cannot be overemphasized. For maximum economy, it is
extremely important that the most direct routes be used in laying out the duct
system. Naturally, where exposed ductwork cannot be tolerated, there must
be some compromise on directness. In any case, when elbows or other
fittings are necessary, easy bends, splitters, or turning vanes should be used.
Duct Design Methods

The common design methods for sizing ducts are constant velocity,
velocity reduction, equal friction, and static regain.
The constant-velocity method is applied in simple systems without
branches and in those exhaust systems where the material transported might
settle if the velocity were reduced. This method can also be used in combina-
tion with others. For instance, in high-velocity design a constant velocity is
often used in the mains up to the point where the friction equals 1 in. wg per
100 ft or so.

An approximate optimum velocity can be determined from Equation 3.1.

Alternatively, the velocity can be chosen from Table 3.1, or several trial
values can be selected and evaluated if the design effort can be justified. Duct
sizes are determined from the appropriate capacity and the design velocity
using the equation of continuity or Figure 2.15. Friction losses can be calcu-
lated from the data in the chapter on fluid flow.
The velocity-reduction method can be employed in designing supply or
exhaust systems having numerous branches. For designing supply systems
this method consists of selecting an appropriate velocity as indicated in Table
3.1 for the first section of main and progressively decreasing the velocity at
each take-off. For designing return or exhaust systems, an appropriate
velocity is selected for the duct at the grill or hood and progressively higher
velocities are used at each junction. Subsequently all main and branch sizes
are determined from the appropriate velocity and capacity, using the equation
of continuity or Figure 2.15. Friction losses may be determined from the data
in the chapter on fluid flow.
Since in this method the branch velocity is arbitrarily chosen as a certain
fraction of that in the main, it is necessary to check whether the available
pressure is equal to or greater than the loss for the design flow. Dampers can
be used for balancing when the loss is less than the available pressure. When
the available pressure at any take-off or junction is not enough to produce the
desired flow in the branch, either the available pressure must be increased by
selecting a fan for a greater pressure requirement or the loss in the branch
must be decreased accordingly by increasing duct sizes.
The equal-friction method is also used in designing many supply and
exhaust systems. It is most effective for symmetrical systems or for systems
where the lengths of all the various runs are approximately equal. For such
systems, balancing is automatically achieved by designing for equal friction
per foot of length. Velocity reduction in the direction of flow on supply
systems is also achieved automatically. Figures 3.1 and 3.2 can be used to
size the remainder of the system after the first section of main is sized. For
systems with some unequal branches, the mains and symmetrical branches
can be designed according to the equal-friction-per-foot-of-length method.
The remaining branches can then be sized in accordance with the pressure
available at the takeoff.
The static-regain method can be used for designing supply or exhaust
systems having numerous branches each connected to a single relatively long
main. This method utilizes the increase in static pressure (which accompanies
any velocity reduction) to provide equal static pressures at each takeoff or
junction. If the length of main between branches is either very long or very
short, this method may not be practical.

3-6 FAN ENGINEERING
Figure 3.1
Round Pipe Sizes for the Equal-Friction-per-Foot-of-Length Method -

Figure 3.2
Round Pipe Sizes for the Equal-Friction-per-Foot-of-Length Method

3-8 FAN ENGINEERING
Figure 3.3 Static Regain Method of Sizing Pipes
Adapted from the data of R. Jorgensen: "New Chart Sizes Ducts Directly Static Regain
Method," Heating, Piping and Air Conditioning, Chicago, October 1958. pp. 107-108.

The static-regain method of proportioning duct systems involves equating

the static pressure regain caused by a change in velocity pressure at a takeoff
to the friction loss caused by the subsequent length of main up to the suc-
ceeding takeoff. Assuming a constant coefficient of friction and a constant
coefficient of regain,
f
L2
D2
1
pv 2 = η pv1 − pv 2 . 6 (3.3)
Figure 3.3 is based on this equation using a value of 0.02 for the coefficient of
friction f and a value of 0.5 for the effectiveness η of the recovery process.
The value of 0.5 for the recovery effectiveness is reasonable for most applica-
tions, but the actual value may be as high as 0.7 or 0.8 under best conditions.
The friction coefficient is also subject to considerable variation, but results
obtained by using this chart are usually sufficiently accurate. When elbows
are encountered in any length of main, the appropriate equivalent length of
straight duct can be added to the actual length of straight duct.
Round Ducts
In all methods of design, the equation of continuity can be used to deter-
mine the duct size from the capacity and the velocity. Accordingly, the ratio
of duct sizes Da Db is related to the capacity ratio Qû a Qû b and the velocity
ratio Va Vb by
Da Qû çæ äã çæ V äã
12 12
= a
å â åV â
b
(3.4)
Db Qû b a
in which subscripts a and b are used to denote two different sections of a

duct system. In the constant velocity method of system design the velocity
ratio is, of course, equal to unity, and therefore, the diameter ratio is equal to
the square root of the capacity ratio. In the velocity-reduction method, the
ratio of velocities is a constant with a value less than unity when subscript b
denotes a location farther from the fan than that denoted by subscript a .
For turbulent flow through the straight pipe sections the ratio p La pLb of
the loss in section a to that in section b can be determined from
çæ p äã = çæ L äã çæ V äã çæ D äã çæ ρ äã çæ c äã .
1.0 2 .0 − n 1.4 − n 1.0 − n
å p â å L â åV â å D â å ρ â å c â
La a a b a a
(3.5)
Lb b b a b b
The various symbols have their usual meanings, and the values of n and c
for numerous roughness conditions can be determined from Table 2.5 in the
chapter on fluid flow. This expression yields approximately the same results
as the Colebrook equation and is considerably easier to use. Equation 3.5 is
.

3-10 FAN ENGINEERING
equivalent to Equation 2.78 and further states that the coefficient of friction
f is inversely proportional to a constant times the Reynolds number raised to
the n power and the roughness ratio raised to the 0.4 − 2n power. The
velocity pressure is, of course, proportional to the velocity squared as
indicated by
pVa V çæ äã . 2
pVb
= a
Vb å â (3.6)
For equal friction per foot of length, the expression p L L is a constant.

For the same gas conditions in branches a and b , density ρ and viscosity µ
are both constant. The necessary diameter ratio for equal friction per foot of
length in terms of either the velocity ratio or the capacity ratio can be
determined from Equation 3.5 and written
2.0 − n ( 2.0 − n ) (1.4 − n )

 Da   Va  1.4− n  Qû a 1.0+ 2.0(2.0− n ) (1.4− n )
 =  = û  . (3.7)
 Db   Vb   Qb 
The graphical solution of this equation for average roughness conditions (that
is, n =.16 ) is given in Figures 3.1 and 3.2. Similar charts can be drawn for
other values of n . However, suitable engineering accuracy can usually be
obtained by using these charts. Reasonably accurate results can also be
obtained by reading duct sizes directly from Figure 2.15 along the appropriate
vertical line.
Another convenient relation can be derived from Equation 3.5 for uniform
gas composition and equal friction for unequal lengths L of duct:
çæ L äã = çæ D äã çæ V äã1.4 − n 2 .0 − n
çD ä
=æ ã
5.4 − 3n
å L â å D â åV â åD â
a a b a
. (3.8)
b b a b
Therefore, the appropriate diameter ratio in terms of length ratio is
çæ D äã = çæ L äã 1 1 5.4 − 3n 6
çL ä
≈æ ã
15
åD â åL â åL â
a a a
. (3.9)
b b b
This equation, of course, is applicable only for turbulent flow. The

approximation involving the 1/5 power usually yields reasonably accurate
results.
An expression giving the diameter ratio in terms of the ratio of pressure
loss p Lb pla for constant capacity and uniform gas composition is

This, too, is appropriate only for turbulent flow, and the final approximation
usually gives reasonably accurate results.
The ratio of the pressure losses for any two capacities can also be found
from the appropriate velocity ratios and diameter ratios as indicated by
çæ p äã = çæ V äã çæ D äã 2 .0 − n 1.4 − n
å p â åV â å D â
La a b
. (3.11)
Lb b La
Pressure losses can also be expressed as a certain number of velocity heads

p L pV . In Equation 3.11, substituting the velocity pressure ratio from
Equation 3.6 produces
çæ p pva äã çæ äã çæ D äã
V
−n 1.4 − n
çD ä
≈æ ã
1.4 − n
åp â å â åD â
= a
åD â
La b b
. (3.12)
Lb pVb Vb a a
The final approximation should be used only when the velocities are nearly
constant.
Total pressure losses can also be expressed as a per cent or fraction of a
velocity head pV p L . This is the reciprocal of the number of velocity heads.
For laminar flow
çæ p äã = çæ L äã çæ V äã çæ D äã çæ µ äã . 2
å p â å L âåV âå D â å µ â
La a a b a
(3.13)
Lb b b a b
In laminar flow the friction factor, or coefficient of friction, varies inversely

with the Reynolds numbers and is independent of roughness so that Equation
3.13 is considerably simpler than Equation 3.5. The expressions correspond-
ing to Equations 3.7 through 3.12 are also simpler for laminar flow. For equal
friction per foot of length with uniform gas composition, the diameter ratio in
terms of either velocity or capacity ratio is expressed by
çæ D äã = çæ V äã = çæ Qû äã
12 14
å D â å V â å Qû â
a a a
. (3.14)
b b b
For unequal lengths and equal friction, the length and diameter ratios are
related as indicated in
çæ L äã = çæ D äã çæ V äã = çæ D äã , and
2 4
å L â å D â åV â å D â
a a a a
(3.15)
b b b b
çæ D äã = çæ L äã . 14
åD â åL â
a a
(3.16)
b b

For constant capacity and uniform gas composition, diameters and total
pressure losses are related as indicated by
çæ D äã = çæ p äã 14
åD â åp â
a Lb
. (3.17)
b La
For equal lengths and uniform gas composition, the total pressure loss
ratio can be found from velocity and diameter or from capacity ratios as
indicated by
çæ p äã = çæ V äã çæ D äã = çæ V äã çæ Qû äã .
2 2
å p â å V â å D â å V â å Qû â
La a b a b
(3.18)
Lb b a b a
The same loss in velocity heads is expressed by
çæ p äã çæ äã
pV Qû
åp â å â
= b .
La
(3.19)
Lb pV Qû a
Equations 3.13 through 3.19 are all for laminar flow and, consequently,
limited in application.
Although Equations 3.5 through 3.19 are based on the flow of air through
straight ducts, some can also be used to proportion systems involving elbows.
For easy bends, where it is safe to assume that the loss is equal to a certain
number of diameters regardless of size, the ratio of losses for any two such
bends can be determined from Equation 3.11. It is usually even more con-
venient to add the equivalent length of straight duct for each elbow to the
actual length of straight duct in each case and to proportion the system as if it
consisted entirely of straight duct. When abrupt turns are involved, for which
the equivalent lengths of straight pipe vary appreciably with size, greater
accuracy is obtained by calculating the losses for these turns separately from
the straight sections to which they join. If the loss for one such turn is deter-
mined in velocity heads or per cent velocity pressure, Equation 3.12 or its
reciprocal, respectively, can be used to find the loss for any similar turn
regardless of size.
Rectangular Ducts
All the preceding equations and charts are for round ducts. When rectan-
gular ducts are to be used, it is necessary to determine the diameter of an
equivalent round duct using either Figures 2.14 or 2.16. Figure 2.14 is based
on equal velocities and equal mean hydraulic radii for the rectangular duct and
its equivalent round duct. It is a graphical solution of the equation
2 xy
DV = , (3.20)
x+ y

Figure 3.4
Rectangular Pipe Sizes for Equal-Friction-per-Foot-of- Length

which can be derived by equating the mean hydraulic radii for equivalent
round and rectangular sections.
Figure 2.16 is based on equal capacities and equal friction per foot of
length for the rectangular duct and its equivalent round duct. It is a graphical
solution of the equation
çæ 4 äã 1 xy6 çæå x +2 y äãâ

2 .0 − n 1.4 − n
5.4 − 3n 3.4 − 2 n 5.4 − 3n
DQ =
åπâ 5.4 − 3n , (3.21)
which can be derived from Equations 3.7 and 3.20. Figure 2.16 is drawn for
average roughness (that is, n = .16). Similar charts can be drawn for other
roughnesses. Equation 3.21 is for turbulent flow. For laminar flow
ç 4ä ç 2 äã
= æ ã 1 xy 6 æ
14 12
åπâ å x + yâ
34
DQ . (3.22)
Figure 3.4 can be used to proportion the width of constant-depth rectan-

gular ducts for equal friction per foot of length. This figure is a graphical
solution of the expression
çæ Qû äã 2 .0 − n
ç x y äã çæ x + y äã
=æ
3.4 − 2 n 1.4 − n
å Qû â åx y â åx +y â
a a a b b
, (3.23)
b b b a a
using average roughness (that is, n = .16) and setting ya = yb .

To proportion rectangular ducts by the equal-friction method, Figures 3.1
and 3.2 can be used with the appropriate equivalent diameters (based on
constant capacity as obtained from Figure 2.16) for any case. For one side
constant, Figure 3.4 can be used directly.
To proportion rectangular ducts by the static-regain method, Figure 3.3
can be used with the appropriate equivalent diameters (as determined from
Figure 2.16) and the actual areas.
The following examples illustrate many principles of the various methods
of sizing ducts.
Example 3.1 Round-Duct Design for Exhaust System
Given the system sketched below and the problem of determining duct
sizes and fan requirements, we can calculate the theoretical optimum velocity
from Equation 3.1 at least approximately by using the appropriate rates and
other factors indicated. For this problem let us assume that we use 22 ga
(0.030 in.) steel (0.283 lbm/in.3) duct work costing $4.15 per lbm installed,
that the system operates 2200 hours a year handling standard air (0.075
lbm/ft3), and that the power rate is 4 cents per kilowatt hour. If we assume a
.

life of 10 years and an interest rate of 15%, the annual owning cost as a
fraction of the first cost is
015
.
F= = 0.20.
1 6
1 − 115
.
−10
If we further assume an average friction factor of 0.02, the optimum velocity

is 3978 fpm as indicated by
 880 × 0.030 × 0.283 × 4.15 × 0.20 

13
Vopt = 1000   = 3978 fpm.

 0.02 × 0.075 × 2200 × 0.04 × 0.746 
Here it is not necessary to check the validity of the 0.02 friction factor
since we will assume that the material being handled with the air requires
approximately 4000 fpm transport velocity. The duct sizes can be determined
directly from the capacity and the velocity, but as indicated in the table below,
this procedure yields odd duct sizes. Practical duct sizes giving velocities
within 10% of the required transport velocity will generally be satisfactory.
The duct friction per 100 ft of length p L 100 ft can be found from Figure
2.15.
cfm 500 1000 1500 2000 3000 3500

D for 4000 4.8 6.8 8.4 9.7 11.6 13.0
D practical 5.0 7.0 8.0 10.0 12.0 12.0
V actual 3700 3800 4300 3700 3800 4400
p L 100 ft 4.4 3.1 3.4 1.8 1.6 2.2

The overall pressure drop in in. wg for each of the alternate air paths can
be determined by adding the appropriate individual drops as indicated in the
following table:
Hoods, els, etc. 1.50 2.00 2.00 2.50 2.50 -

Branch duct .88 .62 1.32 .93 .88 -
Main ducts .88 - - - - -
Main ducts .51 .51 - - - -
Main ducts .18 .18 .18 - - -
Main ducts .32 .32 .32 .32 - -
Main ducts .44 .44 .44 .44 .44 -
Collector, etc. 5.00 5.00 5.00 5.00 5.00 -
Overall 9.71 9.07 9.26 9.19 8.82
cfm 500 1000 500 1000 500 3500
The calculations for this example have been greatly simplified since the
losses through hoods, elbows, etc. were predetermined. The losses through
the various ducts are based on average friction, but for most exhaust systems,
smooth or medium-smooth duct friction corrections should be made, depend-
ing upon the material and the number of joints used. As is usually so, the
longest run has the highest pressure drop. With enough experience a designer
can determine by inspection which run will have the highest pressure drop.
However, when there is doubt, all pressure drops should be calculated as in
this example.
The fan requirements can be stated as 3500 cfm and 9.71 in. wg static
pressure if blast gates can be used to balance the friction loss in the runs that
do not have a loss as great as 9.71 in. wg. If blast gates are not used and if the
fan is selected for 3500 cfm at 9.71 in. wg static pressure, the actual capacities
through the various branches will be different from the design values. The
run with the greatest pressure drop will suffer a loss in capacity, and the
shortest run will handle more air than required. The exact values for all runs
could be calculated by a trial-and-error procedure, but the differences will
generally be negligible. The conservative designer might add 5% - 10% to the
fan capacity to ensure that the longest run would not suffer too big a drop in
the amount of air handled.
It is also possible to balance the total pressure losses by reducing the duct
size in those branches with less than maximum pressure drop. In this prob-
lem, for the branch nearest the fan, the overall pressure drop is 9.71 - 8.82, or
0.89 in. wg less than the maximum. The branch loss could theoretically be
0.89 + 0.88, or 1.77 in. wg. With Equation 3.10, the diameter required to
produce this higher pressure drop can be calculated to be 4.34 in. as indicated
below
çæ 0.88äã 15
D = 5.0
å 177
. â
= 4.34 in.

If the ducts could be sized exactly to the required fractions of an inch, the fan
could be selected for 3500 cfm @ 9.71 in. wg static pressure and, theoreti-
cally, no blast gates would be needed. Blast gates are frequently used except
where they may create great hazards so that the inevitable discrepancies
between actual and theoretical requirements can be balanced out.
Example 3.2 Round and Rectangular Duct Design for Supply Systems
The optimum velocity for round duct costing $4.00 per lbm installed when
the power rate is 4.5 cents per kilowatt hour assuming 24 ga (0.024 in.) steel
(0.283 lbm/in. ) and 3500 hours annual operation can be found by means of
Equations 3.1 and 3.2:
015
.
F= = 016
1 6
1 − 115
.
−20
.
ç 880 × 0.024 × 0.283 × 4.00 × 016

= 1000æ
. ä
å 0.02 × 0.075 × 3500 × 0.045 × 0.746ãâ
13
Vopt = 2790 fpm.
A velocity of 2790 fpm is obtained for a 0.02 friction factor and an annual
owning cost equal to 16% of the initial installed cost. The latter is based on
an annual interest rate of 15% and a 20-year life. The validity of the assumed
0.02 friction factor can be determined by calculating the Reynolds number
and referring to Figure 2.8 for the appropriate roughness condition.

Because the horizontal branches are of equal length, a well-balanced

system will result if we design them for equal friction per foot of length and if
we design the main risers by the static-regain method so that the static pres-
sures at each takeoff will be equal. It is impossible to use either of these
methods with constant velocity, so some compromise on the optimum veloc-
ity is required. The procedure followed below utilizes the optimum velocity
for the 4000-cfm branches. This establishes the size and friction loss per 100
ft according to Figure 2.15. The other branches are then sized for the same
friction loss per 100 ft by noting the corresponding diameters and velocities at
the appropriate capacity on the friction chart.
cfm 4000 cfm 6000 5000 3000

V 2800 p L 100 ft .62 .62 .62
D 17 D 19 18.5 14.5
pL 100 ft .62 V 3100 2930 2650
The first section of main riser can be sized for any reasonable velocity.
The static-regain method produces a considerable reduction in main-riser
velocity so that, for an average riser velocity approaching the optimum, the
velocity in the first section must exceed the optimum. A value of 3500 fpm
requires an area of 11.72 ft2, which roughly corresponds to a 46-in. diameter
and a 0.27-in. wg pressure loss (or friction) per 100 ft.
The remaining main-riser sizes can be obtained directly from Figure 3.3.
The most convenient way of using this chart is to tabulate the information
indicated below using the subscript 2 to indicate the portion of main riser
under consideration and the subscript 1 to indicate the preceding, or upstream,
portion of main riser. For the second portion of main riser, which must handle
30000 cfm, we have already established that the velocity in the upstream
portion will be 3500 fpm. From 30000 divided by 3500 a figure of 8.57 is
obtained and at the intersection of this value of Qû 2 V1 with the appropriate
length L2 of 15 ft, a diameter of 42 in. can be read directly from the chart.
The velocity in a 42-in. pipe carrying 30000 cfm is 3120 fpm. Similar calcu-
lations are required for the remaining sections of main riser.
Qû 2 V1 Qû 2 V1 L2 D2 V2
30000 3500 8.57 15 42 3120
23000 3120 7.37 10 38 2920
16000 2920 5.48 10 33 2690
9000 2690 3.48 10 26 2260

At this point, the design of the transmission system is essentially com-

pleted. For this example the calculations were simplified somewhat by using
predetermined values for the losses in elbows, outlets, etc. The static pressure
required of the fan can be determined as indicated by the following tabulation
and calculations.
2.00 Terminal Requirement

60/100 × 0.62 = 0.37 Horizontal Branches
(3100/4000)2 × 0.52 = 0.31 Takeoff (worst 3500 to 3100)
151/100 × 0.27 = 0.04 Main Riser
1.00 Coils, etc.
3.72 in. wg Overall
The procedure for sizing rectangular ducts is basically the same. The
installed cost of rectangular duct is slightly higher than that of round duct so
that, assuming all other conditions to be the same, the optimum velocity for
rectangular duct costing 10% more would be 2900 fpm as indicated by
çæ 110
. ä
ã
13
Vopt = 2810
å 100
. â
= 2900 fpm.
Once again sizing the 4000 cfm branch for the optimum velocity, the
required area is 1.38 ft2, which can be satisfied with rectangular dimensions of
16 in. × 12.4 in. The equivalent diameter based on constant friction per foot
of length as determined from Figure 2.16 is 15 in., and the corresponding
friction per 100 ft is 0.92 in. wg. The remaining branches are easily sized by
noting the appropriate equivalent diameter opposite the necessary cfm for the
same friction per 100 ft on Figure 2.15. The corresponding rectangular
dimensions can be found in Figure 2.16 and the actual velocity calculated as
indicated.
cfm 4000 cfm 6000 5000 3000

V 2900 p L 100 ft .92 .92 .92
x× y 16 × 12.4 DQ 17.5 16.5 13.5
DQ 15 x× y 16 × 16.7 16 × 14.4 16 × 9.8
p L 100 ft .62 V 3240 31200 2750
Sizing the first section of main for 3800 fpm means that 10.79 ft2 are
needed. This requirement is matched by a 24-in. × 65-in. rectangular cross
section. The equivalent diameter based on constant capacity is approximately
42 in., and the corresponding friction is 0.32 in. wg per 100 ft of length.
Using 3800 fpm in the first section of main and 30000 cfm in the second, the
Qû 2 V1 equals 7.90, and for a 15-ft L2 , the equivalent diameter D2 is 39.8 in.
.

From Figure 2.16, this corresponds to a rectangular section 24 in. × 59 in., and
the actual velocity here is 3050 fpm. Similar calculations must be made for
the remaining sections of main as indicated below.
Qû 2 V1 Qû 2 V1 L2 D2 x× y V2
30000 3800 7.90 15 39.8 24 × 59 3050
23000 3050 7.54 10 38.3 24 × 54 2650
16000 2650 6.03 10 34.5 24 × 44 2180
9000 2180 4.13 10 28.7 14 × 7 1860
The fan should be selected for a capacity of 41000 cfm and a static
pressure of at least 3.97 in. wg as indicated by the following tabulation.
2.00 in. wg Terminal Requirement

60 100 × 0.92 = 0.55 Horizontal Branches
(3120/4000)2 × 0.61 = 0.37 Takeoff (Worst 3800-3120)
15/100 × 0.32 = 0.05 Main Riser
1.00 Coils, Etc.
3.97 in. wg Overall
In both the round- and rectangular-duct examples, the velocities in some

of the horizontal mains exceeded the velocity in the main riser immediately
upstream. This condition is no worse than that caused by balancing dampers,
which would otherwise be required. Also, note that the design velocities are
in the high-velocity range of Table 3.1, so some kind of sound treatment will
be needed.
Examples 3.1 and 3.2 both led to the determination of fan static pressure
even though total pressure losses were calculated. This is so because the
kinetic energy loss at the exit or exits was neglected. In Example 3.1, the exit
loss is equal to the fan velocity pressure so the fan static pressure was calcu-
lated exactly. In Example 3.2, the exit loss was assumed to be equal to the fan
velocity pressure but this really should be verified.
Construction Details
The Sheet Metal and Air Conditioning Contractors National Association
has established standards for duct construction based on experience and tests.
Table 3.2 lists sheet metal thicknesses and some of the joint and reinforcing
details found in those standards. The joint details for low-velocity ducts are
keyed to the sketches given in Figure 3.5. The joints indicated provide
sufficient rigidity except that additional transverse stiffening is necessary
when the length of a duct section exceeds the centers listed. No special
provisions for air tightness are required if the joints are made in a workman-
like manner.
Adapted from the data of SMACNA: Low Velocity Duct Construction Standards, Washing-
ton, D.C., Fourth Edition, April 1969, pp. 11-39, and High-Velocity Duct Construction
Standards, Washington, D.C., Second Edition, January 1969, pp. 6, 14-17.Copper data added.

Table 3.2 Construction Details for Ducts

Low Velocity-2000 fpm max. Low Pressure - 2 in. wg max.
Rect. Sheet Metal Transverse Long- Transverse
Duct Thickness Joints itudinal Reinforcing
Longest Cop- H- Joint
Side Galv. Alum. per Types dim. Types Angle Centers
3"-12" 26 ga .020" 16 oz A,B,K 1" N,O,Z - -
13"-l 8" 24 ga .025" 24 oz A,B,K 1" B,O,Z - -
19"-30" 24 ga .025" 24 oz C,E,K 1" N,O,Z 1" x 1"x 1 8 " -
31"-42" 22 ga .032" 32 oz E,G,K 1" N,I,Z1 1" x 1"x 1 8 " 5'
43"-54" 22 ga .032" 32 oz E,G,K 1½" N,I,Z l½" x l½" x 1 8 " 5'
55"-60" 20 ga .040" 36 oz E,G,K 1½" N,I,Z l½" x l½" x 1 8 " 5'
61"-84" 20 ga .040" 36 oz F,G,H,J,L 1½" N,I,Z l½" x l½" x 1 8 " 2'6"
85"-96" 18 ga .051" 48 oz H,J,LM 1½" N,I l½" x l½" x 1 8 " 2'6"
over 96" 18 ga .051" 48 oz H,J,L:M 2" N,l 2" x 2" x ¼" 2'6"
Round Galvanized Sheet Metal Thickness

Duct Spiral Lock Longitudinal Round Duct
Diameter Seam Duct Seam Duct Fittings
3"-l2" 28 ga 26 ga 26 ga
13"-l8" 26 ga 24 ga 24 ga
19"-28" 24 ga 22 ga 22 ga
29"-36" 22 ga 20 ga 20 ga
37"-52" 20 ga 18 ga 18 ga
High Velocity - over 2000 fpm

Rect. Medium Pressure - to 6 in. wg High Pressure - to 10 in. wg
Duct Transverse Reinforcing Transverse Reinforcing
Longest Galv. Tie- Cen- Galv. Tie Cen-
Side Sheet Rods Angle ters Sheet Rods Angle ters
3"-12" 24 ga - - - 22 ga - - -
13"-18" 24 ga 1 - 48" 22 ga 1 - 40"
24 ga - 1" x l"x 16 ga 48" 22 ga - 1" x l"x l6ga 48"
19"-24" 22 ga 1 - 48" 22 ga 2 - 40"
22 ga - 1" x 1"x 1 8 " 48" 22 ga - 1" x 1"x 1 8 " 48"
25"-36" 22 ga - 1" x 1"x 1 8 " 32" 22 ga - 1¼" x 1¼" x 1 8 " 32"
37"-48" 22 ga - l½" x l½" x 1 8 " 30" 22 ga - 2" x 2" x 1 8 " 30"
49"-60" 20 ga 1 l½" x l½" x 1 8 " 24" 20 ga 1 l½" x l½" x 1 8 " 24"
20 ga - 2" x 2" x 1 8 " 24" 20 ga - 2" x 2" x 1 8 " 24"
61"-72" 20 ga 1 l½" x l½" x 1 8 " 24" 20 ga 1 l½" x l½" x 1 8 " 24"
20 ga - 2½ " x 2½" x 3 16 " 24" 20 ga - 2½ " x 2½" x 3 16 " 24"
73"-84" 18 ga 1 l½" x l½" x 18" 24" 18 ga 1 l½" x l½" x 18" 24"
18 ga - 2½ " x 2½" x 3 16 " 24" - - - -
85"-96" 18 ga 1 l½" x l½" x 1 8 " 24" 18 ga 1 l½" x l½" x 1 8 " 24"
over 96" 18 ga 2 2" x 2" x 1 8 " 24" 16 ga 2 2" x 2" x 1 8 " 24"
Round Galvanized Sheet Thickness Girth Reinforcing

Duct Spiral Lock Longitudinal Round-Duct
Diameter Seam Duct Seam Duct Fittings Angle Centers
3"-8" 26 ga 24 ga 22 ga - -
9"-22" 24 ga 22 ga 20 ga - -
23"-36" 22 ga 20 ga 20 ga - -
37"-50" 20 ga 20 ga 18 ga 1¼" x 1¼" x 18" 72"
51"-60" - 18 ga 18 ga 1¼" x 1¼" x 18" 72"
61"-84" - 16 ga 16 ga l½" x l½" x 18 48"
See footnote on page 3-20 for reference.

Figure 3.5 Low-Velocity Duct Joints
Adapted from the data of SMACNA: Low- Velocity Duct Construction Standards, Washing-
ton, D.C., Fourth Edition, April 1969, pp. 13 and 15.

Round duct is generally less expensive but more space-consuming than

rectangular duct. A round duct requires the least metal for a given cross-
sectional area and the least reinforcing to prevent vibration. Flat, oval duct is
frequently used in high-velocity systems because it can be more economical
than rectangular duct and will fit into tighter spaces than round duct. No
details are given in Table 3.2 because the combination of sheet metal thick-
ness and reinforcing will vary with the manufacturer.
The joints in high-velocity systems must be sealed to provide adequate
air-tightness. Welded or gasketed flanged joints can be used, and liquid or
mastic sealants can be applied to slip joints and standing seams. Tapes are not
recommended. Tie rods can be used at joints as well as at intermediate
locations. Complete details are given by SMACNA.
If abrasive action is expected, ducts for industrial exhaust systems should
be made from thicker sheets. A rule of thumb is to use two gauges heavier for
slight abrasive action and four gauges heavier for greater abrasive action.
Round ducts are commonly used. SMACNA has prepared standards for
round, industrial duct construction that contain information on the minimum
thicknesses required for a wide variety of materials, as well as on the rein-
forcing members required, the connections, and the anchors and supports.1
The usual material in a heating, air-conditioning, or ventilating system is
galvanized steel. However, aluminum or copper can sometimes be justified.
Various materials or protective coatings can be used on industrial exhaust
systems depending on the nature of the gas or material being handled. The
weight per linear foot of galvanized-steel rectangular ducts, including an
allowance for the recommended standing seams, can be determined from
Table 3.4. But additions must be made for any extra reinforcing. For round
galvanized ducts, see Table 3.3; for black-steel ducts, consult Table 3.5.
The surface area S of the material in a 90° elbow of either round or
rectangular cross section can be calculated from the cross-sectional dimen-
sions D , or x and y and the center line radius r as indicated by
S=
1 2
2
1
π rD = πr x + y .6 (3.24)
For the same center line radius, it is immaterial whether the aspect ratio is
x y or y x . Table 3.6 can be used together with Tables 3.3 through 3.5 to
determine the weight of a galvanized or black-steel elbow of either rectangu-
lar or round cross section. The thickness of the material in an elbow or other
fitting should be equal to or greater than that of the connecting straight pipe or
duct. Under highly abrasive conditions, rectangular-section elbows are
generally used since their wearing surfaces can be replaced more easily.
1
Round lndustrial Construction Standards, SMACNA, Vienna, Virginia,1977.

Table 3.3 Weights of Galvanized-Steel Round Ducts

In pounds per lineal foot
Diam. of ft2 per U.S.S. Gauge

Duct Running ft 26 24 22 20 18 16
4 1.13 1.13 1.47 1.69 1.97 2.56 3.10
5 1.39 1.39 1.80 2.08 2.43 3.19 3.82
6 1.65 1.65 2.14 2.47 2.89 3.79 4.54
7 1.91 1.91 2.48 2.86 3.34 4.39 5.25
8 2.18 2.18 2.83 3.27 3.81 5.01 6.00
9 2.44 2.44 3.17 3.66 4.27 5.61 6.71
10 2.70 2.70 3.51 4.05 4.72 6.21 7.42
11 2.96 2.96 3.85 4.44 5.18 6.80 8.14
12 3.22 3.22 4.18 4.83 5.63 7.40 8.85
13 3.48 3.48 4.52 5.22 6.09 8.00 9.57
14 3.74 3.74 4.86 5.61 6.54 8.60 10.28
15 4.01 4.01 5.21 6.01 7.01 9.22 10.86
16 4.27 4.27 5.55 6.40 7.47 9.82 11.74
17 4.53 4.53 5.85 6.79 7.92 10.42 12.45
18 4.87 4.87 6.33 7.30 8.51 11.18 13.36
19 5.14 5.14 6.68 7.71 9.00 11.80 14.11
20 5.40 5.40 7.02 8.10 9.45 12.42 14.85
21 5.59 5.59 7.26 8.39 9.78 12.85 15.36
22 5.92 5.92 7.70 8.88 10.35 13.60 16.25
23 6.18 6.18 8.04 9.27 10.81 14.40 17.00
24 6.45 6.45 8.38 9.67 11.30 14.84 17.71
25 6.71 6.71 8.72 10.06 11.74 15.41 18.41
26 6.97 6.97 9.05 10.45 12.20 16.00 19.15
27 7.33 7.33 9.40 10.85 12.67 16.62 19.87
28 7.50 7.50 9.75 11.27 13.13 17.26 20.60
29 7.75 7.75 10.07 11.63 13.58 17.81 21.30
30 8.10 8.10 10.54 12.17 14.20 18.62 22.25
31 8.36 8.36 10.87 12.54 14.63 19.20 23.00
32 8.62 8.62 11.20 12.93 15.10 19.84 23.70
33 8.88 8.88 11.56 13.34 15.56 20.42 24.40
34 9.15 9.15 11.90 13.73 16.00 21.08 25.18
35 9.41 9.41 12.23 14.10 16.48 21.65 25.85
36 9.67 9.67 12.57 14.50 16.91 22.22 26.60
37 9.93 9.93 12.91 14.90 17.40 22.84 27.30
38 10.19 10.19 13.25 15.29 17.81 23.40 28.00
39 10.46 10.46 13.60 15.60 18.31 24.02 28.70
40 10.72 10.72 13.95 16.08 18.76 24.68 29.50
41 10.98 10.98 14.27 16.47 19.20 25.25 30.20
42 11.24 11.24 14.60 16.86 19.61 25.86 30.90

Table 3.4 Weights of Galvanized-Steel Rectangular Ducts

26 U.S.S. Gauge
Duct Size 6 7 8 9 10 11 12
6 2.00 2.17 2.34 2.50 2.67 2.84 3.00
7 2.17 2.34 2.50 2.67 2.84 3.00 3.17
8 2.34 2.50 2.67 2.84 3.00 3.17 3.34
9 2.50 2.67 2.84 3.00 3.17 3.34 3.50
10 2.67 2.84 3.00 3.17 3.34 3.50 3.67
11 2.84 3.00 3.17 3.34 3.50 3.67 3.84
12 3.00 3.17 3.34 3.50 3.67 3.84 4.00
14 3.34 3.50 3.67 3.84 4.00 4.17 4.34
16 3.67 3.84 4.00 4.17 4.34 4.50 4.67
18 4.00 4.17 4.34 4.50 4.67 4.84 5.00
24 U.S.S. Gauge
Duct Size 18 20 22 24 26 28 30
8 5.65 6.07 6.50 6.94 7.37 7.80 8.24
9 5.87 6.29 6.72 7.15 7.59 8.02 8.45
10 6.08 6.50 6.94 7.37 7.80 8.24 8.67
11 6.30 6.72 7.15 7.59 8.02 8.45 8.89
12 6.52 6.94 7.37 7.80 8.24 8.67 9.10
14 6.95 7.37 7.80 8.24 8.67 9.10 9.54
16 7.38 7.80 8.24 8.67 9.10 9.54 9.97
18 7.82 8.24 8.67 9.10 9.54 9.97 10.42
20 8.25 8.67 9.10 9.54 9.97 10.42 10.85
22 8.68 9.10 9.54 9.97 10.42 10.85 11.29
24 9.12 9.54 9.97 10.42 10.85 11.29 11.72
26 9.55 9.97 10.42 10.85 11.29 11.72 12.13
28 10.00 10.42 10.85 11.29 11.72 12.13 12.56
30 10.43 10.85 11.29 11.72 12.13 12.56 13.00
22 U.S.S. Gauge
Duct Size 32 34 36 38 40 42
8 10.00 10.50 11.00 11.50 12.00 12.50
9 10.25 10.75 11.25 11.75 12.25 12.75
10 10.50 11.00 11.50 12.00 12.50 13.00
11 10.75 11.25 11.75 12.25 12.75 13.25
12 11.00 11.50 12.00 12.50 13.00 13.50
13 11.25 11.75 12.25 12.75 13.25 13.75
14 11.50 12.00 12.50 13.00 13.50 14.00
16 12.00 12.50 13.00 13.50 14.00 14.50
18 12.50 13.00 13.50 14.00 14.50 15.00
20 13.00 13.50 14.00 14.50 15.00 15.50
22 13.50 14.00 14.50 15.00 15.50 16.00
24 14.00 14.50 15.00 15.50 16.00 16.50
26 14.50 15.00 15.50 16.00 16.50 17.00
28 15.00 15.50 16.00 16.50 17.00 17.50
30 15.50 16.00 16.50 17.00 17.50 18.00
32 16.00 16.50 17.00 17.50 18.00 18.50
34 16.50 17.00 17.50 18.00 18.50 19.00
36 17.00 17.50 18.00 18.50 19.00 19.50

Table 3.4 (continued) Weights of Rectangular Ducts

22 U.S.S. Gauge
Duct Size 44 46 48 50 52 54
8 13.00 13.50 14.00 14.50 15.00 15.50
9 13.25 13.75 14.25 14.75 15.25 15.75
10 13.50 14.00 14.50 15.00 15.50 16.00
11 13.75 14.25 14.75 15.25 15.75 16.25
12 14.00 14.50 15.00 15.50 16.00 16.50
14 14.50 15.00 15.50 16.00 16.50 17.00
16 15.00 15.50 16.00 16.50 17.00 17.50
18 15.50 16.00 16.50 17.00 17.50 18.00
20 16.00 16.50 17.00 17.50 18.00 18.50
22 16.50 17.00 17.50 18.00 18.50 19.00
24 17.00 17.50 18.00 18.50 19.00 19.50
26 17.50 18.00 18.50 19.00 19.50 20.00
28 18.00 18.50 19.00 19.50 20.00 20.50
30 18.50 19.00 19.50 20.00 20.50 21.00
32 19.00 19.50 20.00 20.50 21.00 21.50
34 19.50 20.00 20.50 21.00 21.50 22.00
36 20.00 20.50 21.00 21.50 22.00 22.50
38 20.50 21.00 21.50 22.00 22.50 23.00
40 21.00 21.50 22.00 22.50 23.00 23.50
42 21.50 22.00 22.50 23.00 23.50 24.00
44 22.00 22.50 23.00 23.50 24.00 24.50
46 22.50 23.00 23.50 24.00 24.50 25.00
48 23.00 23.50 24.00 24.50 25.00 25.50
20 U.S.S. Gauge
Duct Size 64 68 72 76 80 84 88
10 21.60 22.80 23.90 25.10 26.30 27.40 28.60
12 22.10 23.40 24.50 25.70 26.70 28.00 29.20
14 22.80 23.90 25.10 26.30 27.40 28.60 29.80
16 23.40 24.50 25.70 26.70 28.00 29.20 30.40
18 23.90 25.10 26.30 27.40 28.60 29.80 30.90
20 24.50 25.70 26.70 28.00 29.20 30.40 31.50
22 25.10 26.30 27.40 28.60 29.80 30.90 32.10
24 25.70 26.70 28.00 29.20 30.40 31.50 32.70
26 26.30 27.40 28.60 29.80 30.90 32.10 33.30
28 26.70 28.00 29.20 30.40 31.50 32.70 33.90
30 27.40 28.60 29.80 30.90 32.10 33.30 34.40
32 28.00 29.20 30.40 31.50 32.70 33.90 35.00
34 28.60 29.80 30.90 32.10 33.30 34.40 35.60
36 29.20 30.40 31.50 32.70 33.90 35.00 36.20
38 29.80 30.90 32.10 33.30 34.40 35.60 36.80
40 30.40 31.50 32.70 33.90 35.00 36.20 37.30
42 30.90 32.10 33.30 34.40 35.60 36.80 37.90
44 31.50 32.70 33.90 35.00 36.20 37.30 38.50
46 32.10 33.30 34.40 35.60 36.80 37.90 39.00
48 32.70 33.90 35.00 36.20 37.30 38.50 39.60
50 33.30 34.40 35.60 36.80 37.90 39.00 40.20
52 33.90 35.00 36.20 37.30 38.50 39.60 40.80
54 34.40 35.60 36.80 37.90 39.00 40.20 41.40

Table 3.5 Weights of Black-Steel Round Ducts

Duct ft2 per U.S.S. Gauge

Diam. Running ft 24 22 20 18 16 14 12
4 1.13 1.30 1.58 1.86 2.43 2.99 3.62 5.08
5 1.39 1.60 1.95 2.29 2.99 3.68 4.45 6.25
6 1.65 1.90 2.31 2.72 3.54 4.36 5.28 7.42
7 1.91 2.20 2.67 3.15 4.10 5.05 6.11 8.58
8 2.18 2.50 3.05 3.60 4.68 5.77 6.97 9.80
9 2.44 2.80 3.42 4.03 5.25 6.47 7.80 10.98
10 2.70 3.10 3.78 4.45 5.80 7.15 8.64 12.15
11 2.96 3.40 4.15 4.88 6.36 7.85 9.47 13.31
12 3.22 3.70 4.50 5.31 6.91 8.52 10.30 14.48
13 3.48 4.00 4.88 5.74 7.48 9.21 11.15 15.66
14 3.74 4.30 5.23 6.17 8.03 9.90 11.97 16.84
15 4.01 4.61 5.61 6.61 8.61 10.61 12.83 18.03
16 4.27 4.91 5.97 7.04 9.16 11.29 13.65 19.17
17 4.53 5.21 6.35 7.48 9.74 12.00 14.49 20.40
18 4.87 5.60 6.81 8.03 10.45 12.89 15.55 21.90
19 5.14 5.91 7.20 8.48 11.04 13.60 16.42 23.10
20 5.40 6.21 7.56 8.90 11.60 14.30 17.26 24.30
21 5.59 6.43 7.83 9.22 12.00 14.80 17.87 25.10
22 5.92 6.80 8.28 9.75 12.70 15.65 18.90 26.60
23 6.18 7.11 8.66 10.20 13.29 16.38 19.80 27.80
24 6.45 7.41 9.04 10.63 13.85 17.08 20.65 29.00
25 6.71 7.71 9.40 11.06 14.40 17.75 21.50 30.20
26 6.97 8.01 9.75 11.48 14.96 18.41 22.30 31.30
27 7.23 8.31 10.11 11.93 15.51 19.12 23.10 32.50
28 7.50 8.62 10.50 12.38 16.10 19.87 24.00 33.75
29 7.75 8.91 10.85 12.78 16.67 20.50 24.80 34.90
30 8.10 9.32 11.34 13.37 17.40 21.45 25.90 36.40
31 8.36 9.61 11.70 13.80 18.00 22.15 26.75 37.60
32 8.62 9.92 12.07 14.25 18.52 22.83 27.60 38.80
33 8.88 10.21 12.45 14.66 19.10 23.50 28.40 40.00
34 9.15 10.53 12.81 15.10 19.68 24.43 29.30 41.20
35 9.41 10.82 13.18 15.51 20.20 24.90 30.10 42.30
36 9.67 11.11 13.54 15.95 20.78 25.60 30.90 43.50
37 9.93 11.42 13.90 16.40 21.38 26.30 31.80 44.70
38 10.19 11.71 14.28 16.80 21.90 27.00 32.60 45.80
39 10.46 12.03 14.65 17.27 22.50 27.74 33.50 47.10
40 10.72 12.33 15.00 17.70 23.01 28.40 34.30 48.25
41 10.98 12.62 15.38 18.11 23.60 29.10 35.10 49.40
42 11.24 12.93 15.75 18.55 24.20 29.80 36.00 50.60

Table 3.6 Weights of Elbows

Multiply value from Tables 3.3-3.5 by factor below
Center Line Radius in Inches

0 1 2 3 4 5 6 7 8 9
0 0.14 0.28 0.43 0.57 0.71 0.85 0.99 1.13 1.27
10 1.42 1.56 1.70 1.84 1.98 2.12 2.27 2.41 2.55 2.69
20 2.83 2.98 3.12 3.26 3.40 3.54 3.68 3.82 3.97 4.11
30 4.25 4.39 4.54 4.67 4.81 4.96 5.10 5.24 5.38 5.52
40 5.66 5.81 5.95 6.09 6.24 6.37 6.52 6.66 6.80 6.94
50 7.08 7.22 7.36 7.51 7.65 7.79 7.94 8.07 8.22 8.35
60 8.50 8.64 8.78 8.92 9.06 9.21 9.35 9.49 9.63 9.77
70 9.92 10.1 10.2 10.3 10.5 10.6 10.8 10.9 11.1 11.2
80 11.3 11.5 11.6 11.8 11.9 12.0 12.2 12.3 12.5 12.6
90 12.7 12.9 13.0 13.2 13.3 13.5 13.6 13.7 13.9 14.0
100 14.2 14.3 14.4 14.6 14.7 14.9 15.0 15.2 l5.3 15.4
Principles of Distribution
Distribution requirements vary considerably from one application to
another. Some systems may require a concentrated supply of air at high
velocity, but in others, the objective may be uniform distribution without
objectionable drafts over a broad area.
Air can be circulated and distributed within a space by utilizing the kinetic
energy of one or more streams of air issuing into the space. This energy must
be supplied, directly or indirectly, by a fan.
As mentioned under "Principles of Duct Design," supply and exhaust can
be accomplished with a single fan in either position. A supply fan must
develop the necessary total pressure to produce the kinetic energy required at
the duct opening into a room, plus that necessary to overcome the losses due
to friction, etc. in the duct system, plus that necessary to produce a static
pressure in the space sufficient to force air through the available exhaust
openings. An exhaust fan, however, must develop the necessary total pressure
to overcome the losses due to friction, etc. in the duct system, plus that
necessary to create a negative static pressure in the space sufficient to produce
the required kinetic energy at the duct opening into the room.
Both supply and exhaust openings are necessary for through-circulation.
Air discharging from an opening, like any moving object, tends to continue at
its exit velocity along a straight-line path. Various effects tend to slow down
and deflect the stream. The effects of entrainment and temperature difference
are discussed below. The mutual effect of the supply and exhaust openings
may also be important.
Air entering an exhaust opening tends to approach equally from all
directions. For an opening located in the plane of the wall, the velocity at
one-diameter distance from the opening will be approximately 10% of that at
the opening itself. The velocity decreases rapidly with distance so that the
.

effect of an exhaust opening in producing air motion is limited to the immedi-

ate area of the opening.
Short circuits do develop, to the detriment of room circulation, when
supply and exhaust openings are directly in line at relatively short distances.
The seriousness of short circuiting can be greatly reduced by changing the in-
line relationship or by increasing the distance. But the mutual influence of
supply and exhaust openings on each other is usually of far less importance
than, for example, the effect of cold window panes in producing drafts by
convection. The effect of opening a door may also be more severe than short
circuiting. Frequently, the supply and exhaust openings can be located in
such a way as to counteract these effects. Devices combining both supply and
exhaust in a single unit have been successfully used. The location of supply
and exhaust openings is frequently limited by architectural and functional
requirements. Whenever uniform distribution is required in a large space, the
use of multiple openings should be considered for both supply and exhaust.
Throw of Isothermal Air Jets

The kinetic energy of a jet can be utilized to provide air motion at a
considerable distance from the point where the jet originates, to promote
mixing of the jet of supply air with the room air, or both. Neither effect can
be achieved to the exclusion of the other.
The center line velocity of a jet issuing from a plain round opening will
persist for the first four diameters of throw. This maximum velocity will
decrease as the square root of the distance over the next four diameters or so,
after which the velocity will be inversely proportional to the distance from the
outlet. When the residual velocity falls below about 500 fpm, the decrease in
velocity is more than proportional to the total throw or distance from the
outlet. Room air will be entrained in gradually increasing amounts.
Similar effects will occur with openings of other shapes. A rectangular
opening will produce a jet the maximum velocity of which will not change for
a throw of about four times the short dimension. Thereafter, the maximum
velocity will decrease as the square root of the total throw for a distance of
about four times the aspect ratio times the short dimension. Even with an
aspect ratio of 40 or 50, this stream eventually will become an expanding cone
with a solid angle of 20° to 24°. The point beyond which the maximum
velocity decreases directly with the increase in total throw is about 20 effec-
tive diameters, regardless of the opening shape. Closely spaced multiple
openings produce streams similar to that of a single opening of the same total
area.
The effective area Ae of any outlet is the total area at the vena contracta
or the net free area if there is no contraction. The corresponding effective
diameter De is
De = 113
. Ae (3.25)

The effective diameter of any square-edged orifice, whether round, square,

or rectangular, can be found from its basic dimensions using Figure 3.6.
Figure 3.6
Effective Diameters of Round and Rectangular Openings
Adapted from the data of R. D. Madison and W. R. Elliot: "Throw of Air from Slots and
Jets," ASHVE Journal Section of Heating, Piping and Air Conditioning, Chicago, November
1946, pp.108 to109.

This chart is based on a ratio of free area to gross area R A of 1.00, a

coefficient of discharge CD of 0.6, and a velocity-of-approach factor ϕ i of
1.00. For any other conditions the square-edge value De.S .E . must be modified
according to
R Aϕ i CD
De = De.S .E . . (3.26)
0.6
The values of R A and ϕ i C D for several typical outlets are given in Table
3.7.
The effective velocity of the jet issuing from the outlet Vo is that at the
vena contracta. This can be calculated from the capacity Qû and the effective
area Ae using the equation of continuity or from the gross area Ag of the
outlet using
Qû Qû
Vo = = . (3.27)
Ag R Aϕ i CD Ae
The average residual velocity at any distance from the outlet, beyond
approximately 10 effective diameters, is about 1/3 of the maximum residual
velocity. The maximum, or centerline, residual velocity Vr relative to the
effective velocity at the outlet Vo is a function of the throw X and the
effective diameter De . Conversely, the throw in diameters X De is
proportional to the velocity ratio Vr Vo . as indicated by
X V
=K o. (3.28)
De Vr
The proportionality constant K for orifices and nozzles can be determined

directly from Table 3.8. These values should be modified for other types of
outlets as indicated in Table 3.7.
The entrainment ratio RQ , or the ratio of the total moving quantity Qû t to
the primary or jet quantity Qû , is also a function of the velocity ratio V V as
j r o
indicated by
Qû t
2
Vr  Vo 
RQ = û = 0.314  1.12 + 0.35 K  − 1 . (3.29)
Qj Vo  Vr 
Various solutions of Equations 3.28 and 3.29 are presented graphically in

Figure 3.7.

Table 3.7 Distribution Factors for Various Outlets
Approx. Approximate†
Outlet RA φ i CD * K
Rounded-entrance
nozzles 1.00 0.99 Use Table 3.8 Values Directly
Square-edged
orifices 1.00 0.60 Use Table 3.8 Values Directly
Plain straight-duct
openings 1.00 1.00 Multiply Table 3.8 Values by 1.05
Bar grilles 0.84 0.66 Multiply Table 3.8 Values by 0.86
Perforated panels 0.40 0.83 Multiply Table 3.8 Values by 0.89
Effect of high-aspect
ratios - - Multiply Table 3.8 Values by 0.86
Effect of adjacent wall
Parallel to axis
of flow - - Multiply Table 3.8 Values by 1.42
Adapted from the data of A. Koestel, P. Hermann, and G. L. Tuve: "Comparative Study of
Ventilating Jets from Various Types of Outlets," Trans. ASHVE, vol. 56, New York, 1950,
pp. 459-478.
*Velocity-.of-.approach factors corresponding to 1400 fpm ahead of bar grilles are included
in the table.
†The value of K increases for throw calculation., but decreases for entrainment calculations.
This chart can be used to find the entrainment ratio and maximum residual
velocity at any distance from a given outlet. It can also be used to determine
the necessary combination of effective diameter and original velocity for a
given throw and a specified residual velocity. Example 3.3 illustrates a
typical use of this chart.
Example 3.3 Isothermal Air Jet
Given a 42-in. × 6-in. bar grille discharging 2000 cfm, find the residual
velocity and the total quantity of moving air at a throw of 30 ft.
From Table 3.7,

R A = 0.74,
φ i CD = 0.78, and
K = Table 3.8 values × 0.72.
From Equation 3.27,

Table 3.8
Proportionality Factors K for Throw from Nozzles and Orifices
Vo -fpm
Vr -fpm 1000 2000 3000 4000 5000
500 up - 5.3 5.5 5.7 6.0
400 - 5.0 5.2 5.5 5.8
300 4.4 4.6 4.8 5.0 5.3
200 4.1 4.2 4.4 4.6 4.8
100 3.3 3.3 3.4 3.5 3.5
Adapted from the data of G. L. Tuve and G. B. Priester: "Control of Air Streams in Large
Spaces," Trans. ASHVE. vol. 50, New York, 1944, pp 153-172.
2000 2000
Vo = = = 1980 fpm, and
. × 0.74 × 0.78 101
175 .
From Figure 3.6 and Equation 3.26,
DeS . E . = 14 in. = 1.17 ft, and

0.78
De = 117
. 0.74 × = 1.14 ft.
0.6
30
Using Figure 3.7 with V = 1980 and X De = = 26.3,
114
.
Vr = 430. (first approximation)
Since Figure 3.7 is based on Table 3.8 values and since Table 3.8 values must
be modified for bar grilles, adjustments must also be made according to
Tables 3.7 and 3.8, and Equation 3.28:
Vr = 0.72 × 430 = 310 and K = 46 , (second approximation)

K = 0.72 × 4.6 = 3.3, and
1980
Vr = 3.3 × = 249. (final approximation)
26.3
The average residual velocity is 1/3 of this value, or 83 fpm.
Using Equation 3.29 and the definition of entrainment ratio,

RQ = 0.314 ×
249
1980
1 6
. + 0.395 × 3.3 × 7.96 − 1 = 5.21 − 100
112
2
. = 4.21.
Therefore, the total quantity of air put into motion is

4.21 × 2000 = 8420 cfm.

Note: The successive approximations of Vr are required since Figure 3.7 is

based on Table 3.8 values of K . Wherever Table 3.8 can be used directly,
Figure 3.7 can also be used directly.
Figure 3.7 Throw of Air from 0penings

Adapted from the data of R.D. Madison and W.R. Elliot: "Throw of Air from Slots and Jets,"
ASHVE Journal Section of Heating, Piping and Air Conditioning, Chicago, November 1946,
pp. 108-109.
Throw of Heated or Cooled Air Jets

If the primary air in a jet is heated above the temperature of the air in the
room, the stream will tend to rise as it proceeds away from the outlet.

The vertical rise Y of the center line of a horizontally projected stream is

a function of the throw X De , the original velocity Vo and temperature To ,
the ambient temperature Ta , and the effective diameter De as indicated by
Y Xçæ äã çæ T − T äã çæ gD äã .
3
= 0.065
å â å T âå V â
o a e
2
(3.30)
De De a o
Similarly, if the primary air is cooled below the ambient temperature, the
stream will fall.
The maximum possible vertical downward projection X max of a heated
stream or the maximum upward projection of a cooled stream can be calcu-
lated from
X max çæ
Ta äã çæ V äã − 2.85.
2
= 3.4
å â å gD â
o
(3.31)
De To − Ta e
Equations 3.30 and 3.31 are adapted from the data of Koestel1, which also
give the results of temperature-distribution studies.
Example 3.4 Heated Air Jet
Given a 1.0 ft2 squared-edged orifice with a plenum approach projecting 3000
cfm of 180°F air through ambient air at 70°F, find the maximum throw for
downward projection and the vertical rise of the center line of a horizontally
projected stream at that same throw.
From Table 3.7,

R A = 1.00,
φi CD = 0.6
Using Equations 3.27 and 3.25,
3000
Vo = = 5000 fpm,
1.0 × 0.6
De = 113
. 1.0 × 0.60 = 0.875 ft, or
1
A. Koestel, "Paths of Horizontally Projected Heated and Chilled Air Jets," Trans. ASHVE,
vol. 61, New York, 1955, pp. 213-232, and "Computing Temperatures and Velocities in
Vertical Jets of Hot or Cold Air," Trans. ASHVE vol. 60, NewYork, 1954, pp. 385-410.

Using Equation 3.31,

ç 3.4ç 530 ä ç 5000 ä
æå æå 640 − 530ãâ æå 32.2 × 3600 × 0.875äãâ − 2.85ãâ .
2
X max = 0.875
X max = 0.8753 3.4 × 4.81 × 247 8 − 2.85 = 53 ft.
Using Equation 3.30,

ç 53 äã çæ 1 äã çæ 1 äã = 10.6 ft.
Y = 0.875 × 0.065æ
3
å 0.875â å 4.81â å 247 â

Transmission & Distribution of Air: Fan Systems

Uploaded by

Copyright:

Available Formats

Transmission & Distribution of Air: Fan Systems

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Transmission & Distribution of Air: Fan Systems

Uploaded by

Copyright:

Available Formats

Chapter 3

Transmission & Distribution of Air

Principles of Duct Design

‹#1999 Howden Buffalo, Inc.

When there must be several branches, the fan should be as centrally

As expressed by this equation, the optimum velocity is a function of the metal

where i is the annual interest rate and n is the number of years.

© 1999 Howden Buffalo, Inc.

‹#1999 Howden Buffalo, Inc.

In the design of high-velocity systems it is important to limit the use of

Table 3.1 Design Velocities in fpm

Although much discussion has been devoted to establishing duct velocities

Duct Design Methods

‹#1999 Howden Buffalo, Inc.

An approximate optimum velocity can be determined from Equation 3.1.

‹#1999 Howden Buffalo, Inc.

‹#1999 Howden Buffalo, Inc.

‹#1999 Howden Buffalo, Inc.

Figure 3.3 Static Regain Method of Sizing Pipes

© 1999 Howden Buffalo, Inc.

The static-regain method of proportioning duct systems involves equating

in which subscripts a and b are used to denote two different sections of a

‹#1999 Howden Buffalo, Inc.

For equal friction per foot of length, the expression p L L is a constant.

2.0 − n ( 2.0 − n ) (1.4 − n )

Therefore, the appropriate diameter ratio in terms of length ratio is

This equation, of course, is applicable only for turbulent flow. The

© 1999 Howden Buffalo, Inc.

Pressure losses can also be expressed as a certain number of velocity heads

In laminar flow the friction factor, or coefficient of friction, varies inversely

‹#1999 Howden Buffalo, Inc.

The same loss in velocity heads is expressed by

‹#1999 Howden Buffalo, Inc.

‹#1999 Howden Buffalo, Inc.

çæ 4 äã 1 xy6 çæå x +2 y äãâ

Figure 3.4 can be used to proportion the width of constant-depth rectan-

using average roughness (that is, n = .16) and setting ya = yb .

Example 3.1 Round-Duct Design for Exhaust System

‹#1999 Howden Buffalo, Inc.

If we further assume an average friction factor of 0.02, the optimum velocity

 880 × 0.030 × 0.283 × 4.15 × 0.20 

Vopt = 1000   = 3978 fpm.

cfm 500 1000 1500 2000 3000 3500

© 1999 Howden Buffalo, Inc.

Hoods, els, etc. 1.50 2.00 2.00 2.50 2.50 -

‹#1999 Howden Buffalo, Inc.

ç 880 × 0.024 × 0.283 × 4.00 × 016

Vopt = 2790 fpm.

‹#1999 Howden Buffalo, Inc.

Because the horizontal branches are of equal length, a well-balanced

cfm 4000 cfm 6000 5000 3000

‹#1999 Howden Buffalo, Inc.

At this point, the design of the transmission system is essentially com-

2.00 Terminal Requirement

cfm 4000 cfm 6000 5000 3000

‹#1999 Howden Buffalo, Inc.

2.00 in. wg Terminal Requirement

In both the round- and rectangular-duct examples, the velocities in some