79037
A GUIDE TO FAN SELECTION AND PERFORMANCE 1.
NOTATION
Units A
fan casing dimension, see Sections 4 and 7.2.2: also cross-sectional area of airway according to context
m m2
B
fan casing dimension, see Sections 4 and 7.2.2
m
C
fan casing dimension, see Sections 4 and 7.2.2
m
DT
impeller or rotor diameter at tip, see Sketch 3.7
m
DH
impeller or rotor diameter at hub, see Sketch 3.7
m
d
specific diameter defined Section 3.2
f
frequency
Hz
fB
blade passing frequency
Hz
J
number of sound sources
K
pressure loss coefficient
Kw
fan specific sound power level
k
number of blades on impeller
Lp
sound pressure level
dB
Lw
sound power level
dB
L w∗
base sound power level
dB
N
rotational speed of impeller
rad/s
n
specific speed defined in Section 3.2
rad
p
root mean square sound pressure
Pa
pd
dynamic pressure, see Section 3.1.2 for definition
Pa
p f, d
fan dynamic pressure defined in Section 3.1.2
Pa
ps
static pressure associated with system
Pa
pt
total pressure associated with system
Pa
p0
root mean square reference sound pressure
Pa
∆ pf, s
fan static pressure rise defined in Section 3.1.2
Pa
dB
Issued December 1979 - 83 pages With Amendment A 1
79037 ∆p f, t
fan total pressure rise defined in Section 3.1.2
Pa
P f, s
fluid static power produced by fan defined in Section 3.1.2
W
P f, t
fluid total power produced by fan defined in Section 3.1.2
W
PR
power input to fan impeller
W
q
volume flow rate through fan
m3/s
r
distance
m
T
reverberation time
s
U
mean flow velocity, U = q ⁄ A
m/s
V
volume of reverberant test enclosure
m3
W
sound power
W
WH
width of impeller flow passage at hub, see Sketch 3.7
m
WT
width of impeller flow passage at tip, see Sketch 3.7
m
W0
reference sound power
W
η
fan efficiency expressed as a percentage, defined in Section 3.1.2
per cent
σ
relative density* re 1.2 kg/m3, i.e. σ = ρ ⁄ ρ 0
ρ
density of gas flowing through fan, nominally at inlet
kg/m3
ρ0
density of air at standard conditions, i.e. ρ 0 = 1.2 kg/m3
kg/m3
φ
flow coefficient defined in Section 3.2
rad
ψ
pressure coefficient defined in Section 3.2
rad
*
–1 –2
This definition of relative density should not be confused with the ratio of the density of water vapour to that of dry air sometimes associated with the term "relative density”.
Subscripts
ext
relates to conditions external to system
H
relates to conditions at impeller hub
i
refers to fan inlet quantity
j
quantity associated with j th octave band or measuring position
m
refers to mean value of quantity
max
refers to maximum value of quantity
2
79037 o
refers to fan outlet quantity
s
refers to static pressure or quantity related to fan static pressure rise
t
refers to total pressure or quantity related to fan total pressure rise
T
relates to condition at impeller tip
3
2.
79037 INTRODUCTION This Item contains introductory material to assist the non-specialist with fan selection and tender appraisal. It also presents characteristic properties of the major categories of fans in a non-dimensional form that allows estimates of size, power requirements and noise emission to be made for given combinations of flow rate and pressure rise. To a non-specialist, the particular meanings given to certain terms in the context of fans may be confusing. In addition, use is made in this Item of special non-dimensional groups for generalising the characteristics of fans. The Item therefore first leads the user, in Section 3, through comprehensive definitions and explanations of the conventions and terminology related to fans. For certain applications, e.g. the handling of airborne fibres, certain types of fans are a prerequisite. Section 4 identifies the salient features of the major fan categories and presents their typical characteristics in a non-dimensional form. It will be found, however, that for many applications a choice of fans is available. Here, apart from cost, a fan may be selected (i) according to its performance, e.g. stability of operation, ease of control, power consumption, etc., (ii) according to its mechanical arrangement, e.g. convenience of installation, self-cleaning blade properties, etc., or (iii), because of noise emission advantages. Users familiar with fans, their terminology and characteristics, may refer directly to Section 7 which, for several fan types operating at a given duty, enables impeller powers, speeds and physical sizes to be estimated and compared. Alternatively, where noise emission is the primary concern, Section 9 may be used to provide estimates and comparisons of sound power spectra from several fan types. This section also gives many conversion formulae and other information to enable the diverse presentations of noise data which are found in manufacturers' literature to be compared. For users less familiar with the characteristics of fans, Sections 5 and 6 are necessary preliminary reading. Before selecting a fan, it is essential to have considered stability of operation, flexibility of system operation including growth potential, and special system requirements such as back-up/stand-by capability. Section 5 explains these system characteristics and requirements while Section 6 gives guidance on how they and the characteristics of fans may be successfully integrated. In the majority of applications, fans are driven by electric motors on whose properties Section 8 gives general information. It covers, for example, the starting problems that can arise with certain fan and motor combinations and outlines the methods of speed control that are used for fans.
2.1
Scope The categories of fans dealt with are: axial flow, centrifugal flow, mixed flow and crossflow. The information is limited to fans where the density of the gas varies less than ten per cent between the fan inlet 4 and outlet; typically, this corresponds to a maximum fan total pressure rise, ∆ pf , t , of 10 Pa (approximately 40 inches of water). Compressors and some high pressure ratio fans are thus outside the scope of this Item.
4
3.
79037 FAN TERMINOLOGY AND BASIC RELATIONSHIPS The following paragraphs enlarge on the terms associated with fans that are not otherwise in common usage within internal flow fluid mechanics. Section 3.1 deals with the terms surrounding fan and system performance and Section 3.2 enlarges on the non-dimensional groups used to portray the properties of fans in this Item. Section 3.3 presents the scaling approximations that are known as the Fan Laws and Section 3.4 describes the mechanical conventions for fans including methods for depicting the discharge and motor positions of centrifugal fans.
3.1
Performance Terminology
3.1.1
General terms Characteristic The characteristic of a device or system is the relationship that links important primary variables associated with its operation. The most commonly used fan characteristic is the relationship between pressure rise and volume flow rate for a given impeller speed. Similarly the relationship between pressure loss and volume flow rate is the most commonly used system characteristic. Fan pressure rise and flow rate characteristics are often expressed non-dimensionally in terms of ψ , the pressure coefficient, and φ , the flow coefficient. These coefficients, which are fully defined in Section 3.2, are independent of impeller speed and fan size. Operating Point An operating point is defined as the fan pressure rise/volumetric flow rate condition where the fan and system are in a stable equilibrium. This corresponds to the condition at which the fan pressure rise/flow rate characteristic intersects the system pressure loss/flow rate characteristic.
Sketch 3.1 Definition of operating point On a system or fan that has a variable characteristic, as may be obtained using either system or fan control devices, the operating point will also vary to describe an "operating line" and, where both are independently variable, an "operating region" will result, see Sketch 3.2.
5
79037
(a) Fan controlled, e.g. by a (b) System controlled, e.g. by a (c) Fan and system controlled variable speed motor damper Sketch 3.2 Influence of fan and system controls 3.1.2
Terms related to fan pressure rise Fan total pressure rise, ∆ pf , t , often loosely termed "fan total pressure", is defined as the rise in total pressure produced by the fan between its inlet and outlet. It can be defined with an open inlet, see Sketch 3.3a, or open outlet (Sketch 3.3b) or as an in-duct quantity (Sketch 3.3c). Thus for ∆ pf , t to be consistent for the different fan inlet and outlet arrangements, variations of intake and outlet losses arising from the different geometries are by convention neglected. Fan dynamic pressure, pf , d , is defined as the fluid dynamic pressure corresponding to the mean through flow velocity, Uo , at the fan outlet based on the total outlet area without any deductions for motors, fairings or other bodies. Thus, for the low Mach number flow rates to which this Item applies, the fan dynamic pressure* may be taken as 2
p f, d = ½ρU o .
(3.1)
Fan static pressure rise, ∆ pf , s , often loosely termed "fan static pressure", is defined as the difference between the fan total pressure rise and the fan dynamic pressure, i.e. ∆ pf, s = ∆ pf, t – p f, d .
(3.2)
The fan static pressure rise is thus equal to the gauge static pressure at the fan outlet when the fan draws in air from the atmosphere through a well shaped intake. Note that the definition of this term is peculiar to fans and is not consistent with the normal meaning of static pressure rise. The term is derived from methods of testing the performance of fans. Special care must be taken with calculations involving ∆pf , s since it will not correspond to the actual static pressure rise between the fan inlet and outlet unless the fan draws in directly from the atmosphere or a large plenum, see Sketch 3.3a. When a fan has an inlet and outlet plane of the same through-flow area and discharges directly to the atmosphere, ∆p f , s is equal to the total pressure rise across the fan for negligible density changes, see Sketch 3.3b. Note that when discharge terminals or intakes are coupled directly to a fan, they are, by convention, regarded as an integral part of the fan and not of the system. However, when they are remote from the fan such that the fan inlet and outlet stations are within the confines of the *
2
Strictly, dynamic pressure is the difference between total pressure and static pressure and is thus greater than ½ρU . However, for air at Mach numbers below 0.2, i e. 68 m/s flow velocity at standard conditions, the difference is less than one per cent.
6
79037 connecting duct work, the fan total pressure rise, ∆pf , t , and not fan static pressure rise, ∆p f , s , will equal the static pressure rise between the fan inlet and outlet, again for negligible density changes (see Sketch 3.3c). The relationships in Sketch 3.3 all neglect variations in the inlet losses and the velocity profiles of the flow entering and leaving the fan. They also neglect density changes between inlet and outlet. Although the values of pressure differences calculated using the above relationships may differ slightly from those obtained by integrating the local velocities and pressures measured within the inlet and outlet cross sections, for most practical purposes they are compatible with the pressure rises required to match duct system pressure losses which are also usually calculated making similar assumptions*.
(a)
Free intake Integral diffuser†
( pt )o – ( pt )i = ∆p f , t , 2
( ps )o – ( p s )i = ∆pf , t – ½ ρU o , ( ps )o – ( p s )i = ∆pf , s . Note: ( p s )i = ( pt )i .
(b)
Free outlet
( pt )o – ( pt )i = ∆ pf , t , 2
( pt )o – ( pt )o' = ½ρU o , because fan dynamic pressure is lost at the free discharge, i.e. 2
( ps )o' = ( pt )o' = ( p t ) o – ½ρU o . Hence ( pt )o' – ( p t )i = ∆p f , s .
*
Such duct pressure loss calculations should nevertheless include an allowance for swirl which can be pronounced downstream of tube-axial fans. With the exception of diffusers, the effect of swirl is usually to increase the pressure loss of components. † A diffuser that is directly coupled to a fan may also be regarded as an integral part of the fan in some conventions. Here the fan dynamic pressure is that corresponding to the outlet through-flow area of the diffuser. This is illustrated in Sketch 3.3a.
7
79037 (c)
Ducted inlet and outlet
( pt ) – ( pt ) = ∆ pf , t , o i ( ps ) – ( p s ) = ∆ pf , t , o i ( ps ) – ( p t ) = ∆ pf , s . o i
Sketch 3.3 Fan ducting arrangements Air power, P o , is the power delivered by a fan to the gas and is given by P o = q ∆ pf ,
(3.3)
where the fan pressure rise, ∆ pf , may be taken as either the fan static pressure rise or fan total pressure rise, i.e.
or
P o, s = q ∆ pf, s P o, t = q ∆ pf, t .
(3.4) (3.5)
Note that Po , t and not Po , s is the actual power gained by the fluid. Air static power, P o , s , is an artifice used in the definition of fan static efficiency, η s , described below. Efficiency, η , is the ratio of air power, Po , to the power input to the propeller, PR , expressed as a percentage, i.e. Po η = ------- × 100 per cent. PR
(3.6)
Air power and hence efficiency may be taken as either a static or total quantity, i.e.
or
P o, s η s = ----------- × 100 per cent PR
(3.7)
P o, t η t = ---------- × 100 per cent. PR
(3.8)
The static and total efficiencies are related by η s = η t ∆ pf, s ⁄ ∆ pf, t .
(3.9)
8
79037 Because P o, t is the actual power delivered to the fluid, the fan total efficiency and not static efficiency is the quantity representative of the fan energy conversion efficiency. The fan static efficiency is always less than the total efficiency but this, of course, does not mean that there is any difference in the impeller power required for a given duty; the impeller power, PR , obtained from Equation (3.10) is identical to that obtained from Equation (3.11), i.e.
or
ηs P R = q ∆ pf, s ⁄ --------100 ηt P R = q ∆ pf, t ⁄ --------- . 100
(3.10) (3.11)
Nevertheless, for many systems, a fan offering a high value for η s may be more suitable than another offering a higher η t but lower* η s . This is because many systems include a discharge such that the fan dynamic pressure is lost from the total pressure rise, see Sketch 3.3b. The fan static pressure rise is in such cases the total pressure difference remaining to overcome the system resistance. On systems where the fan is ducted on both the inlet and outlet (see Sketch 3.3c), the fan total pressure rise, ∆ pf , t , should be used to match the fan to the system losses. However, if such a system discharges from a terminal of approximately the same through-flow area as that of the fan exit plane, a fan dynamic pressure will be lost and an optimised fan static efficiency will result in a minimum impeller power, P R . On closed-loop systems or systems incorporating efficient diffusers, the fan total efficiency should normally be optimised. 3.2
Dimensionless Groups Dimensionless groups are used in this Item for characterising the properties of fans of similar fundamental design but of different scale and, to a limited extent, of different proportion. Four groups are used: specific diameter, d , specific speed, n , flow coefficient, φ , and pressure coefficient, ψ . The flow coefficient, φ , has a unique definition, i.e. q -. φ = ---------------------------------3 DT N WT ⁄ DT
(3.12)
Thus, φ is proportional to the ratio of two volume flow rates: that on the numerator is the actual flow rate through the fan while that in the denominator is equal to the product of the impeller tip speed and the impeller surface area circumscribed by the blade tips†.
*
Such a situation is possible when a fan with a large through-flow area produces a certain total pressure rise. A fan of inherently more efficient design (higher ηt ) but smaller through-flow area producing the same total pressure rise at the same volume flow rate may have a lower static efficiency. The static efficiency value is penalised by the higher fan dynamic pressure associated with the smaller through-flow area. † For axial-flow and mixed-flow fans, the impeller surface area is ascribed a nominal value proportional to the disc area such that WT ⁄D T = 1 . See comments later in this Section.
9
79037 The other groups may be defined either on a basis of fan total pressure rise or fan static pressure rise. Thus
or
WT ⁄ D T d s = D T ------------------q
½
WT ⁄ D T d t = D T ------------------q
½
∆ pf, s ------------σρ 0
¼
∆ pf, t -----------σρ 0
¼
(3.13)
;
(3.14)
d is a measure of the compactness of the fan. Generally, large diameter wide impellers are best suited to high flow rates but small narrow impellers are best suited to high resistance systems. q ½ σρ 0 ¾ -----------------------------ns = N WT ⁄ DT ∆ pf, s or
½ σρ 0 ¾ q n t = N ------------------- ------------ ; WT ⁄ DT ∆ pf, t
(3.15)
(3.16)
n is a measure of the impeller speed necessary to produce a fan pressure rise at a given flow rate.
or
∆ pf, s ψ s = --------------------------2 2 DT N σ ρ0
(3.17)
∆ pf, t ψ t = --------------------------- ; 2 2 D T N σρ 0
(3.18)
ψ is proportional to the ratio of two pressures: that on the numerator is the actual fan pressure rise while that on the denominator is the dynamic pressure corresponding to the impeller tip speed. Any of these groups may be expressed in terms of any other two, i.e. ¼
1 1 Ψ = -------- , d = ----------½- = ------------1⁄ 1⁄½ n 3φ 3 φ nψ
(3.19)
½
1 1 φ n = --------- = ----------- = -------- , 3 ½ ¾ d φ dψ ψ
(3.20)
½
1 ψ 2 3 - = n ψ ⁄2 , φ = --------3- = ------2 nd d 2
(3.21)
⁄
1 φ 3 4 2 -. ψ = ------------ = d φ = -------4 2 2 ⁄3 n n d
(3.22)
It may be noted that, contrary to convention, the definitions of d , n and φ used in this Item involve the impeller width to diameter ratio, WT ⁄ D T . The inclusion of this term allows the performance of both double inlet centrifugal fans and cross-flow fans to be correlated in terms of specific speed without having to treat
10
79037 these fans as special cases. Sometimes, however, when it is desirable to illustrate the performance differences between separate categories of fans, conventional definitions of the groups specific diameter, specific speed and flow coefficient, that do not include the term W T ⁄ D T have an advantage. This is because the term WT ⁄ D T , which correlates the effect of different impeller geometries, tends to vary more between categories of fans than within them. By omitting the term, therefore, the difference between, for example, the flow coefficients of two different categories of fans each operating at peak efficiency are exaggerated. This approach is adopted in Sketches 4.2 and 4.3 for illustrating the inherent characteristics of different fan types. The groups specific diameter, specific speed and flow coefficient defined in this way are indicated by a prime ('), e.g. q φ′ = ------------- . 3 DT N The success of conventional definitions in correlating the performance of single-inlet centrifugal fans lies in the range of W T ⁄ DT being limited by mechanical and fluid flow constraints within each category of fan. Therefore, in the definitions of φ' , but especially d ′ and n' (where its influence is smaller), it is an acceptable approximation to absorb the effect of WT ⁄ D T into correlations of, say, d ′ versus n' for each category of centrifugal fan. For axial-flow and mixed-flow fans, W T ⁄ D T is analogous to some function of the impeller hub to tip diameter ratio but, because of design constraints, it is not an independent variable in a fan optimised for a particular duty. Thus, in common with the definitions of d ′ , n' and φ' , the ratio W T ⁄ D T should be assigned a value of unity for evaluating d , n and φ for axial and mixed-flow fans, i.e. d = d ′ , n = n' and φ = φ' .
3.3
Fan Laws In Section 3.2, the non-dimensional groups ψ and φ are assembled from a combination of variables governing the performance of fans, i.e. D T , N , W T and σ , and variables that reflect their performance, i.e. q and ∆pf , s (or ∆pf , t ). Thus, by assigning a particular value to ψ and knowing the values of D T , N and ρ associated with a fan and its operation, the fan pressure rise is defined. Similarly, by assigning a value to φ and knowing D T , N and WT ⁄ D T , the flow rate is defined. Thus a fan characteristic consisting of a function of ∆pf versus q can, for known values of D T , N , W T ⁄ D T and σ , be alternatively represented by a function of ψ versus φ . The function, ψ ( φ ) , representing the characteristic of a given fan may be used to represent the characteristic of any geometrically similar* fan. If the system characteristic is known, the operating point may also be defined as a pair of co-ordinates, ψ and φ . An operating point defined in this way is known as a "point of rating". Geometrically similar fans having the same point of rating (i.e. operating at the same values of ψ and φ ) have properties that are related by expressions known as the "Fan Laws". For example, if the subscript "1" is used to denote one fan and the subscript "2" another geometrically similar fan and if both operate at the same point of rating, the following expressions apply:
*
Two fans are geometrically similar when one fan and a scale factor completely defines the other.
11
79037 q2 q1 ------------------------------------------ = φ 1 = φ 2 = -----------------------------------------WT WT 3 3 ( D T ) 1 N 1 -------( D T ) 2 N 2 -------DT DT 1
and
(3.23)
2
( ∆ pf ) ( ∆ pf ) 1 2 ---------------------------------------- = ψ 1 = ψ 2 = ---------------------------------------- . 2 2 2 2 ( DT )1 N1 σ1 ρ0 ( DT )2 N2 σ 2 ρ0
(3.24)
For geometrical similarity, ( W T ⁄ DT )1 = ( W T ⁄ D T )2 and the expressions may be rearranged into the form 2
N2 ( ∆ pf ) = ( ∆ pf ) -----2 1 N 1 and
( DT ) 2 σ 2 2 --------------- -----( DT ) σ1
(3.25)
1
3 N2 ( DT )2 . q 2 = q 1 ------ --------------N1 ( DT )
(3.26)
1
The fan pressure coefficient, ψ , and pressure rises, ( ∆pf ) and ( ∆p f ) , in Equations (3.24) and (3.25) may 1 2 relate to either the fan static pressure rise or fan total pressure rise. Equations (3.25) and (3.26) can be combined to give a relationship between the impeller powers, ( P R ) 1 and ( PR ) 2 , i.e. N2 ( P R ) = ( P R ) -----2 1 N 1
3
( DT ) 2 --------------( DT )
5
1
σ2 ------ . σ1
(3.27)
Using a similar argument regarding the governing variables of fan noise, a relationship can be derived between the sound power levels, ( L w ) 1 and ( L w ) 2 , i.e. ( DT ) σ2 N2 2 --------+ 50 log10 + 70 log10 --------------- . ( L w ) = ( L w ) + 20 log10 2 1 ( DT ) σ1 N1 1
(3.28)
Equation (3.28) is derived from a combination of other fan laws (Equations (3.25) and (3.26)) and the function used in the noise estimation procedure that relates a change in pressure rise or flow rate to a change in sound power level. Equation (3.28) is therefore subject to the same conditions of applicability as the fan laws and the noise estimation method (given in Section 9.6). Furthermore, its use for predicting sound power level changes within individual octave bands, resulting from speed or diameter changes, is limited to those changes that are sufficiently small to maintain the blade passing frequency (see Equation (9.13)) in the same octave band. Sketch 3.4 demonstrates one application of the fan laws. The pressure and volume flow rate characteristic for a particular fan is known at a rotational speed N1 , and it is required to predict the characteristic for the fan at a new speed N 2 . Using Equations (3.25) and (3.26), point 1 of Sketch 3.4 is transformed, moving along a parabolic path of constant ψ and φ due to the increase in speed, to a new point 2. Similarly every other point on the original performance characteristic may be mapped onto the new characteristic.
12
79037
Sketch 3.4 Effect of change of speed on fan characteristic In this particular case, the parabolic path linking point 1 to point 2 coincides with the system characteristic. This is because the effect of speed implies a fan pressure rise that is proportional to q2 , a characteristic that follows most system characteristics. The variation of density also affects the fan characteristics in a way that approximately balances the effect on most systems without changing the fan point of rating as illustrated in Sketch 3.5.
Sketch 3.5 Effect of change of density on fan characteristic Sketch 3.6 however demonstrates another application of the fan laws for predicting, from the known characteristic of one fan, the characteristic of a similar but larger fan operating at the same impeller speed. The paths representing constant points of rating no longer coincide with the system characteristic. It is thus not possible to predict operating point 2 from operating point 1 alone; instead a knowledge of a range of the original fan characteristic is necessary
13
79037
Sketch 3.6 Effect of change of size on fan characteristic Note that the fan laws are strictly only applicable for changes that involve exact geometrical similarity and for incompressible flows with constant Reynolds numbers. However, in the context of fans, the effect of Reynolds number variations is usually small and it will often be found that fans of different sizes available within a manufacturer's line may, for practical purposes, be regarded as geometrically similar. 3.4
Mechanical Conventions
3.4.1
Major impeller dimensions The major impeller dimensions referred to in this Item involve diameter and width. They are illustrated in Sketch 3.7.
3.4.2
Direction of rotation The standard25 method of specifying the direction of rotation for centrifugal fans is to view the fan from the drive side and indicate whether the rotation required is to be clockwise or anti-clockwise. Note that the drive side of a centrifugal fan is considered to be the side opposite the inlet even when the actual position is on the inlet side. The standard25 method of specifying the direction of rotation for axial and mixed-flow fans is to view the fan from the discharge end and indicate whether the rotation required is to be clockwise or anti-clockwise.
3.4.3
Discharge and motor position As with the direction of rotation, the specification of discharge position is normally only necessary with centrifugal fans. Sketch 3.8 shows the usual method of indicating discharge position and rotation for a centrifugal fan25. The position of the motor for centrifugal fans should also be specified, i.e. either directly coupled or according to Sketch 3.9.
14
79037
Sketch 3.7 Major impeller dimensions
Sketch 3.8 Rotation and discharge designations
15
79037
Sketch 3.9 Standard motor positions for centrifugal fans Axial-flow fans may be driven either by a directly coupled motor or via vee belts to a motor mounted on the outside of the casing.
16
4.
79037 MAJOR CATEGORIES OF FANS There are four main categories of fans: axial flow, centrifugal flow, mixed flow and cross flow. These categories are defined by the nature of the flow through the impeller blades, see Sketch 4.1.
Sketch 4.1 The major fan categories Within each of the major fan categories, the following sub-categories are covered.
Axial Flow
Centrifugal Flow
Propeller
Forward-curved
Tube-axial
Radial-bladed
Guide-vane-axial
Backward-bladed
Mixed Flow
Cross Flow J-casing
Axial-casing
S-casing U-casing
Each type of fan has a different characteristic that may, or may not, be suited for the required duty. For example, a fan intended to blow cooling air slowly over electronic components will have a relatively large volume flow rate capacity but relatively low pressure gain. By contrast, a fan required to blow air through a filtration system offering a high flow resistance will have a relatively small volume flow rate capacity but relatively high pressure rise. The characteristics of pressure rise and volume flow rate can be represented non-dimensionally in terms of ψ , the fan pressure coefficient*, and φ , the volume flow rate coefficient*. Fans of the same geometric proportions, though of different size and running at different speeds, when represented in terms of these coefficients can be compared directly. Another advantage of representing fan characteristics in terms of pressure and flow coefficients is for the comparison of fans from different categories although, for this purpose, it is customary to use the more conventionally defined flow coefficient†, φ' .
* †
Section 3.2 gives full definitions of these terms. The definitions of the non-dimensional groups flow coefficient, φ' , specific speed, ns' , and specific diameter, ds' , employed in Sketches 4.2 and 4.3 do not include the term ( WT ⁄D T ) . Elsewhere in this Item, however, the groups are defined with a ( WT ⁄DT ) term. See notes in Section 3.2 for a full explanation..
17
79037
Sketch 4.2 Optimum efficiency fan static pressure coefficients for various types of fans Sketch 4.2 compares the Cordier line with the pressure coefficient and flow coefficient†, φ' , corresponding to optimum efficiency for the categories of fans dealt with in this Item. The Cordier line15 is intended to represent pressure coefficient and flow coefficient combinations that are produced by the most efficient designs of turbomachines themselves operating at optimum efficiency. It agrees acceptably well with most fan categories except the forward-curved and cross-flow types whose geometry is not compatible with the Cordier classes of turbomachines. Sketch 4.2 also shows that axial-flow fans can only be designed to produce relatively low pressure coefficients. Centrifugal-flow fans, however, produce higher coefficients. Within the envelope for axial-flow fans, the wide range of pressure coefficient is achieved by varying the solidity, e.g. a small number of blades gives a low pressure coefficient and vice-versa. Within the centrifugal fan envelope, the range of pressure coefficient is achieved by varying the impeller blade angles, i.e. backward-curved blades give the lowest pressure coefficients, radial blades give intermediate values and forward-curved, the highest. In addition, forward-curved blades produce the highest flow coefficients. In the same way that ψ and φ' relationships can be used to denote the pressure rise and flow rate characteristics of a type of fan, other non-dimensional groups, namely ds' , the specific diameter*, and ns' , the specific speed*, may be used to typify the impeller diameter and operating speed relationship for each fan category.
*
The definitions of the non-dimensional groups flow coefficient, φ' , specific speed, ns' , and specific diameter, ds' , employed in Sketches 4.2 and 4.3 do not include the term ( WT ⁄D T ) . Elsewhere in this Item, however, the groups are defined with a ( WT ⁄DT ) term. See notes in Section 3.2 for a full explanation..
18
79037
Sketch 4.3 Diagram illustrating specific speed versus specific diameter for optimum efficiency conditions of various fans Sketch 4.3 illustrates the relationship between specific speed and specific diameter for the different fan categories when each fan within each category runs at maximum efficiency. As with Sketch 4.2, it is noticeable how different fan categories monopolise different regions on the graphs. The sketch shows that, for a given impeller diameter, axial-flow fans run at higher impeller speeds than centrifugal fans. It should be noted that Sketches 4.2 and 4.3 do not merely display patches of characteristics from a single fan in the vicinity of its maximum efficiency. Instead the curves in the sketches connect the ψ and φ' conditions of several fans in each category each operating at maximum efficiency. Detailed design variations between fans in the same category allow individual fans to be optimised for different duties within the ranges indicated in Sketches 4.2 and 4.3. Suitability of pressure rise and volume flow rate is not the only criterion for selecting a fan. Consideration must be given to the shape of the pressure rise/flow rate characteristic, the efficiency, shape of the power characteristic, speed, size, inlet and outlet configuration and also the power and spectrum shape of the noise emitted from different fan types. In addition, the use of a certain type of fan may be determined by its mechanical suitability, e.g. for propelling airborne particulate matter. The following Sections, 4.1 to 4.4, comment in some detail on general performance and mechanical features of different types of fans. Guidance on their noise characteristics is given in Section 9. Typical fan performance characteristics are illustrated in terms of ψs , the static pressure coefficient, and φ , the flow coefficient. Since these coefficients are directly proportional to the fan static pressure rise*, ∆ pf , s , and the volume flow rate, q , the ψs versus φ charts are merely convenient methods of giving a more general presentation than their dimensional counterparts. Also included on the same charts in the following sections are typical fan static efficiencies and typical impeller power characteristics. It should be noted that although the characteristics illustrated in Sections 4.1 to 4.4 give an indication of the behaviour of each category of fan, they are not wholly general within their category due to detailed design variations.
*
Note that ∆ pf , s is not necessarily equal to the change in static pressure between the inlet and outlet of a fan. For a full explanation of fan static pressure rise, see Section 3.1.2
19
4.1
79037 Axial-Flow Fans Within the major category of axial-flow fans, there are four sub-categories: propeller fans, tube-axial fans, contrarotating fans and guide-vane-axial fans. Most axial-flow fans are available with many blade angle settings that in some cases may be adjusted, when stationary, by slackening a clamping mechanism in the impeller hub. Other, more sophisticated fans, have a variable pitch facility that can alter the impeller blade angles while the fan is in operation. Changing the blade angles predominantly affects the flow coefficient, φ . Fans optimised to produce high flow coefficients are set with large blade angles that can give rise to stalling if the flow is over-throttled. Stalling is manifested by a rapid decrease in both the fan pressure rise and the efficiency as the flow rate is reduced, see for example, Sketch 4.11. Whilst the blade angle is the chief design variable governing flow coefficient, the impeller solidity is the main variable governing the fan pressure coefficient. Generally, high solidity impellers (large number of blades and, or, high hub to tip diameter ratios) generate high pressure coefficients more efficiently than low solidity impellers which operate more efficiently at low pressure coefficients.
4.1.1
Propeller fan
For approximate dimensions A, B, and DT see Section 7.2
Sketch 4.4 Propeller fan Applications for this fan mostly involve moving air through a partition from one open space to another such as the supply or removal of air through the roofs or walls of buildings or through the cabinets of electronic equipment. The shape of the lips of the mounting ring has a marked effect on the efficiency. Small sizes of these units, e.g. D T < 0.5 m, often have impellers moulded from plastic but the larger sizes usually have impellers of pressed steel. Except for very large units, e.g. DT > 1.5 m, most are supplied with integral electric motors and are restricted to speeds near the synchronous pole speeds (see Section 8).
20
79037
Sketch 4.5 Pressure coefficient, flow coefficient, efficiency, and impeller power characteristics of a typical propeller fan 4.1.2
Tube-axial fan
For approximate dimensions A , B and D T , see Section 7.2
Sketch 4.6 Tube-axial fan Applications for this fan are found in many air-conditioning systems that require a high volume flow rate but small pressure rise. In contrast to centrifugal flow fans, this type of fan can be installed directly in a duct. Design variations involving the blade geometry and number of impeller blades can significantly affect both the pressure rise and flow rate. The swirling motion of the discharged flow dissipates energy that reduces the efficiency of the fan and can increase the pressure loss of at least the first component downstream. Fans of this type are frequently coupled directly to electric motors and thus have a restricted range of operating speeds but some versions are driven through vee belts by an electric motor mounted on the casing.
21
79037
Sketch 4.7 Pressure coefficient, flow coefficient, efficiency, and impeller power characteristics of a typical tube-axial fan 4.1.3
Contrarotating fan
For approximate dimensions A , B and D T ,see Section 7.2
Sketch 4.8 Contrarotating fans These fans are used for duties that require higher pressure rises than those obtainable from single axial-flow fans. The units are normally constructed from a pair of tube-axial fans so that the swirling motion imparted to the flow by the upstream fan is countered by the second fan. Thus the overall efficiency of the combination is higher than the efficiency of either component fan in isolation. Although the second fan is often set with smaller blade angles than the first because of the swirling inlet flow, the maximum pressure coefficient obtainable from the combination can be more than twice the maximum pressure coefficient obtainable from either component fan in isolation.
22
79037
Sketch 4.9 Pressure coefficient, flow coefficient, efficiency and impeller power characteristics of typical contrarotating fans 4.1.4
Guide-vane axial fan
For approximate dimensions A , B and D T , see Section 7.2
Sketch 4.10 Guide-vane axial fan These fans are used for duties involving high volume flow rates with modest pressure rises. This makes them suitable for use in high velocity air-conditioning systems and also in certain forced draft combustion applications. Generally, these fans are well suited to applications where high efficiency is important and also where swirl downstream of the fan must be controlled such as in wind tunnels. The construction of guide-vane-axial fans is similar to tube-axial fans with the addition of inlet or outlet guide vanes. The guide vanes improve the efficiency by countering the swirl produced by the rotor. In more sophisticated versions, the angle of the rotor blades may be varied while the fan is operating to effect an efficient flow control. Some versions of these fans are driven through vee belts by an electric motor mounted on the casing but it is more usual for the impeller to be directly coupled as depicted in Sketch 4.10. The vee-belt drive however offers a convenient means of speed changing. Large versions of these fans are sometimes shaft driven.
23
79037
Sketch 4.11 Pressure coefficient, flow coefficient, efficiency and impeller power characteristics of a typical guide-vane-axial fan 4.2
Centrifugal-Flow Fans Within the major category of centrifugal fans, there are three sub-categories that relate to the blade shapes of the impeller, namely: forward curved, radial bladed and backward bladed. In all cases the flow enters the impeller axially through a hollow hub (termed "eye") from both or, more usually, one end and is discharged with a radial component into the casing (termed "volute", "scroll" or "shroud").
For approximate dimensions A , B , C , D T and W T , see Section 7.2
Sketch 4.12 Centrifugal fan 4.2.1
Forward-curved centrifuga1 fan
Sketch 4.13 Forward-curved impeller
24
79037 This type of fan, sometimes referred to as a "volume blower" or "multi-vane" fan, is best suited to applications requiring a high volume flow rate at low to medium pressure rises. It therefore competes with tube-axial and guide-vane-axial fans for some duties. Unlike axial fans, however, it cannot normally be mounted in-line within the ductwork although some manufacturers produce a unit enclosed in a rectangular casing that can be installed in-line between rectangular ducts. The impeller accelerates the flow to a high velocity while rotating at a speed that is usually low compared with other fans. The dynamic pressure is converted to static pressure by the diffusing action of the scroll whose design is therefore important to the efficiency of the fan. With well-designed scrolls, the maximum static efficiencies obtainable are comparable with those achieved by tube-axial designs. A noteworthy feature of this fan is the shape of the power curve. If, by chance, during operation the flow resistance unexpectedly drops, for example due to an inspection door being opened, the power absorbed by the impeller will rise possibly to an extent that could overload the electric motor.
Sketch 4.14 Pressure coefficient, flow coefficient, efficiency and power characteristics of a typical forward-curved centrifugal fan 4.2.2
Radial-discharge centrifugal fan
a. curved heel blades
b. straight blades
Sketch 4.15 Radial-discharge impellers The chief application of radial-discharge impellers is the handling of airborne particles. The blades tend to be self cleaning in moderately dirty conditions and the more efficient unit with curved-heel blades is thus often used for draught induction in boilers. The fans are capable of very high pressure rises and, because of their tolerance to particulate matter, are suited to filtration duties. The simple straight-bladed radial impeller can be made from thick steel to withstand impact from large 25
79037 particles and may be constructed with the flat blades attached directly to a spider hub. Low efficiency and noisiness restricts the general application of this type of fan. The shape of the power curve, see Sketch 4.16, is similar to that of forward-curved impellers in that it rises to a maximum when the fan static pressure rise is zero. Care must therefore be taken to ensure that the drive motor is adequately rated for contingencies, such as inadvertent filter removal, that might cause the system flow resistance to be reduced. For operation in dusty environments where deposits may build up on the blades, an allowance for out-of-balance forces should be made by specifying special bearings. In such cases, it is also usual to specify a vee-belt drive because the reduction in performance caused by deposits on the blades may be compensated for by changing the pulleys in order to increase the impeller speed. However, because the efficiency also deteriorates under such conditions, it is necessary to specify an electric drive motor with an adequate power margin.
Sketch 4.16 Pressure coefficient, flow coefficient, efficiency and impeller power characteristics of a typical radial-bladed centrifugal fan 4.2.3
Backward-bladed centrifugal fan
Sketch 4.17 Backward-bladed impeller Depending primarily on the ratio DH ⁄ D T , the pressure coefficients of these impellers can be varied to suit high pressure rise/low flow rate requirements or medium pressure rise/medium flow rate applications. The passages between the blades diffuse more than half of the dynamic pressure into static pressure so that
26
79037 energy losses are reduced when the remaining dynamic pressure is converted to static pressure in the scroll. Efficiencies are generally high but the operating speeds are higher than those of other comparably sized centrifugal fans. In order to exploit the high efficiency characteristics, the inlet can be equipped with variable angle guide vanes to control the flow rate. Such guide vanes, by pre-swirling the inlet flow in the direction of the impeller, provide a more efficient means of control than throttling, see Section 6.3.2. The impellers may be constructed with aerofoil-shaped blades, constant thickness curved blades or straight blades.
Sketch 4.18 Blade types used with backward-bladed impellers In principle there are advantages of efficiency with aerofoil blades and advantages of cost and toughness with straight blades. Curved blades offer a frequently-used compromise. Two further categories of backward-bladed fans are often quoted, that of wide backward-bladed and that of narrow backward-bladed. There is no firm definition of these categories but "wide" fans may normally be taken to mean those with W H ⁄ DT ≥ 0.3 and "narrow", those with W H ⁄ DT < 0.3 .
Sketch 4.19 Pressure coefficient, flow coefficient, efficiency and impeller power characteristics of a typical backward-bladed centrifugal fan
27
4.3
79037 Mixed-Flow Fans
Sketch 4.20 Mixed-flow fan Mixed-flow fans have characteristics that overlap those of axial-flow fans and those of centrifugal-flow fans. Outwardly, the most common type resembles an axial-flow fan. They are normally applied to duties suited to characteristics between those of axial-flow and centrifugal-flow fans. 4.3.1
Axial-casing mixed-flow fan
For approximate dimensions A , B and D T , see Section 7.2
Sketch 4.21 Axial-casing mixed-flow fan These fans are frequently used when a fan characteristic approximately that of a backward-curved centrifugal fan is required but the installation dictates an axial inlet and outlet configuration.
Sketch 4.22 Pressure coefficient, flow coefficient, efficiency and impeller power characteristics of a typical axial-casing mixed-flow fan
28
4.4
79037 Cross-Flow Fans
For approximate dimensions A , B , C , D T and W T , see Section 7.2
Sketch 4.23 Cross-flow fan These fans are used where the convenience of the slot-like aspect of the inlet and outlet planes is more important than efficiency and where low pressure rises are sufficient. Typical applications include domestic fan-assisted heaters, hand-held hair dryers, and air curtains. The fans are also known as "transverse-flow fans" and "tangential blowers". All have an impeller similar in cross section to those of forward-curved centrifugal impellers but, unlike centrifugal fans, the flow enters the impeller through a peripheral segment rather than through the hub. Thus the width of the impeller is not restricted by flow considerations through the "eye" of the impeller. In practice, impeller widths between about 0.7D T and 10D T may be specified. The operation of these fans is very sensitive to the shape of the casing and is easily disturbed by components placed too close to the inlet or outlet. 4.4.1
Casing configurations of cross-flow fans
Sketch 4.24 J, S and U configurations of cross-flow fans Cross-flow fans are available with three casing configurations that determine the relative positions of the intake and discharge planes. The J casing is the most common and results in a higher performance than the U casing.
29
79037
Sketch 4.25 Pressure coefficient, flow coefficient, efficiency and impeller power characteristics of a typical J-casing cross-flow fan
30
5.
79037 SYSTEM CHARACTERISTICS AND REQUIREMENTS It is seldom that a system can operate continually at a single operating point*. For example, filter blockage in a filtration system will markedly vary the system pressure drop while the opening and closing of discharge terminals in a ventilation system will significantly vary the system volume flow rate. Atmospheric temperature, humidity and pressure variations will also affect the air density. The system resistance, which may be calculated as a static pressure loss, ∆ ps , or total pressure loss, ∆ pt , is obtained by summing the pressure losses due to the individual components in the system. The pressure 2 2 loss due to a component can be expressed as K s ½ρU or K t ½ρU , where K s and Kt are the static-pressure and total-pressure loss coefficients for the component respectively; hence 2
∆ ps = Σ½ρU K s + ( ∆ ps ) ext and
2
∆ pt = Σ ½ρU K t + ( ∆ pt ) . ext
(5.1) (5.2)
Values of K s and K t may be obtained from sources such as Reference 8. The term ( ∆p s )ext or ( ∆pt )ext represents an externally applied pressure drop which would arise, for example, in a boiler where significant buoyancy effects would be expected. In such a case, the applied pressure drop would be negative. Similarly, particular wind directions may cause applied pressure drops (either positive or negative) to systems installed in buildings with intakes and outlets located in different positions. For most systems, variations in ambient temperature, humidity and pressure, whilst having no effect on the volume flow rate, have an effect on the pressure drop that is represented by Equation (5.3): ( ∆p s ) – ( ∆p s ) ( ∆p t ) – ( ∆p t ) σ2 2 2, ext 2 2, ext --------------------------------------------------- = -------------------------------------------------- = ------ , ( ∆p s ) – ( ∆p s ) ( ∆p t ) – ( ∆p t ) σ1 1 1, ext 1 1, ext
(5.3)
where subscript 1 corresponds to a condition in which the relative density† at a particular point in the system, e.g. the fan inlet, is σ 1 . Subscript 2 corresponds to a condition giving the same volumetric flow rate but where the relative density† at the same point in the system is σ 2 . 5.1
Fixed Systems For the high Reynolds number flow regime to which this Item is applicable, the pressure loss will, for many purposes, be approximately proportional to the square of the flow rate. This assumes that there are no automatic control devices interfering with the natural characteristics of the system, i.e. the loss coefficients K s or Kt remain constant. If this is true, then where a system pressure loss, ( ∆p s )1 or ( ∆p t ) 1 , corresponding to a flow rate, q1 , is known, the pressure loss corresponding to a different flow rate, q 2 , is given by σ 2 q 2 2 = ------ ----- [ ( ∆ ps ) – ( ∆p s ) ] + ( ∆ ps ) 2 1 ext ext σ 1 q 1
(5.4)
σ 2 q 2 2 ( ∆ pt ) = ------ ----- [ ( ∆ pt ) – ( ∆ pt ) ] + ( ∆ pt ) . 2 1 ext ext σ 1 q 1
(5.5)
( ∆ ps )
or * †
Operating point is defined in Section 3.1.1. Relative density, σ , as defined in this Item is the ratio of the actual flow density, ρ , to a standard flow density, ρ equal to 1.2 kg/m3. 0 It should not be confused with the special meaning of relative density, i.e. ratio of water vapour density to air density, adopted by some terminologies.
31
79037 The relationships given in Equations (5.4) and (5.5) enable the system operating line to be drawn, see Section 3.1.1, that describes the system pressure loss variation with volume flow rate. Note, however, that Equations (5.4) and (5.5) are valid only while the loss coefficients, K s and K t , remain unchanged - hence the term "fixed system".
5.2
Systems with Varying Loss Coefficients In many systems, the loss coefficients vary either by design, e.g. through the use of dampers and controllable discharge terminals, or as a consequence of operation or ageing, e.g. due to the blocking of filters or other in-duct devices. This means that the system operating line will depart from the fixed characteristic represented by Equations (5.4) and (5.5). This is illustrated in Sketch 5.1.
Curve No. (1)
Corresponds to blocked filters/closed dampers/closed terminal,
(2)
corresponds to conditions mediate beween (1) and (3),
(3)
corresponds to clear filters/open dampers/open terminal.
inter-
Sketch 5.1 Characteristics of system with varying loss coefficients The passive characteristics of a system such as depicted in Sketch 5.1 should not be regarded in isolation from the requirements for the system. For example, although the loss coefficients in a filtration system will vary, ideally the volume flow rate through the system should not be affected. Alternatively, although the volume flow rate through a system with controllable terminals will clearly vary, the flow rate through those terminals that are open should ideally be maintained constant by an unchanging pressure drop. These contrasting requirements are illustrated in Sketch 5.2. (a)
Constant volume flow requirement for filtration system.
(b)
Constant pressure drop requirement for a short ventilation system.
For curves 1, 2 and 3 see Sketch 5.1.
Sketch 5.2 Constant volume flow and constant pressure drop system requirements Although it is possible to meet very closely such requirements, by matching the system characteristics to controllable fans, it is often possible to find satisfactory solutions without recourse to controls by relaxing the precision with which the ideal requirements are met. This is discussed in Section 6. 5.3
System Requirements for Reliability (Back-up Systems) The previous sections have dealt with systems and their requirements in the context of performance. The following notes, however, consider systems that are configured in certain ways to meet reliability criteria. Depending on the criteria set, which can frequently be governed by codes of practice, systems may be 32
79037 designed to be driven by more than one fan operating in parallel, or more than one fan operating in series. Sections 5.3.1 and 5.3.2 describe some of the special considerations that must be given to these systems.
5.3.1
Systems with fans in parallel Where failure is unacceptable, a system may be designed to be powered by two or more fans operating in parallel. In this configuration, certain flow stability problems can arise and the guidance given in Section 6.1 should be carefully noted. To cover the event of one fan failing, the fans must be equipped with devices for closing the ducts to prevent short circuiting of the flow through the failed fan. In addition, unless an effective back-flow damper is provided, it may be necessary to specify a rotor lock to avoid high starting currents from unstarted fans windmilling in the wrong direction.
5.3.2
Systems with fans in series A system employing two fans in series with one fan operating and the other idling is a common requirement for certain ventilation applications. This configuration known as "series standby" is less efficient than single fan configurations due to the pressure drop caused by windmilling of the stand-by fan. This extra resistance must be taken into account when calculating the system pressure drop. Series configurations may also be used in systems where total failure is unacceptable but partial failure is acceptable. In this case the normal operating point will be achieved only when both fans are operating but an acceptable reduced flow condition will result when one fan fails. Section 6.1 gives guidance on how the characteristic of a two-fan system can be calculated.
33
6.
79037 FAN AND SYSTEM MATCHING In each of the following sections, the fan and system characteristics used to find the operating point are mostly portrayed in terms of ∆ p ⁄ σ . The term ∆p is left without a subscript because it may represent either a fan static pressure rise and a compatible system pressure drop, or a fan total pressure rise and system total pressure drop. Guidance on these aspects is given in Section 3.1.2 The relative density, σ , appears because, in the majority of applications, for a fixed volume flow rate, both the pressure loss exhibited by the system and the pressure rise produced by the fan are directly proportional to the density of the flow. Thus density changes arising from temperature, pressure and humidity variations may be taken into consideration. It should be noted, however, that the fan pressure rise being proportional to density is conditional upon the fan running at constant speed. With the type of electric motor commonly used for driving fans, minor variations in impeller power (which is also proportional to density) will normally lead to negligible speed variations but it is nevertheless important to check that the motor is adequately rated for the maximum expected flow density. Further guidance on electric motor specification is given in Section 8. The operating point should not only be positioned at the required system flow rate but situated on parts of the fan and system characteristics that provide stable operation. In order to achieve this, it may sometimes be desirable to employ more than one fan. Guidance on this is given in Section 6.1. Section 6.2 explains why matching a fan to a fixed system requires a careful assessment of the uncertainties surrounding system resistance estimation. For systems with varying loss coefficients, it is necessary to decide whether control devices are necessary to meet the system requirements. Section 6.3 gives guidance on this and also on the choice of the several means of control that are available. Section 6.4 gives guidance on system trouble shooting.
6.1
Multi-Fan Arrangements Situations can arise where the system pressure drop for the required flow rate exceeds that available from a single fan. Similarly, because of availability or other constraints, situations can arise where the required system flow rate cannot be met by a single fan. In these cases, illustrated by Sketches 6.1a and 6.1b respectively, solutions may be found by employing more than one fan.
Sketch 6.1 System requirements incompatible with characteristics of single fans There can be other reasons dictating the use of more than one fan, e.g. (a)
two fans may fit the available space more easily than a single larger fan,
34
6.1.1
79037 (b)
volume flow rate control using multiple fan combinations may be more economical than other control techniques,
(c)
fans positioned near the system inlet and near the system exhaust are often used in ventilation applications to avoid excess pressure in the space being served,
(d)
where total failure is unacceptable.
Two fans in parallel With two fans connected in parallel by symmetrical ductwork, equilibrium occurs when the value of the total pressure rise* from both fans is identical. Thus the combined fan characteristic for two fans in parallel may be constructed from the individual fan characteristics by summing, for a series of values of fan total pressure rise*, the values of the volume flow rate for each fan at that pressure. In Sketch 6.2 the individual characteristic for the two fans (which happen to be identical) is curve A-A, whilst the characteristic for the two fans in parallel is the curve C-C. However, there are complications to this simple procedure.
Sketch 6.2 Combined performance characteristic for two fans in series and parallel In Sketch 6.3, the efficiency and pressure characteristics of each of two fans operating in parallel are illustrated. Their combined characteristic is also shown and is drawn by adding all the possible combinations of volume flow rate at different pressures. The maximum in the individual pressure characteristics, a feature of many fan types, causes a loop in the combined characteristic which, in the case of forward-curved fans, is likely to occur near the condition of maximum efficiency. If the combination is operated at capacities corresponding to the loop, each fan may switch from either of the two conditions that satisfy the system characteristic. This cyclic switching, apart from causing noise annoyance, can cause damage to the fans, other items in the system and to the fan transmission.
*
If both fans, which need not be of the same size or type, discharge into a large plenum, equilibrium will occur when each fan produces the same fan static pressure rise, ∆ pf , s .
35
79037
Sketch 6.3 Unstable operation of fans in parallel In order to avoid such problems, it is necessary to select fans with pressure rises that are sufficiently high for the operating point to occur on the steep portion of the combined characteristic. However, even if under normal operation there are no problems of instability, starting may pose problems if the fans are not equipped with control devices. Where the fans have electrical starting devices or speed controls (see Section 8), they should be started in a progressive sequence. Further, if the fans are of an axial-flow type, the backflow through the unstarted fan may induce a stall from which it may not recover. When such fans are employed, therefore, it is usual to equip one fan with a throttling device to prevent backflow while that fan is started. If forward-curved or radial-bladed fans are operated in parallel, the electrical system should be protected (see Section 8.3) against the failure of one fan. This is because such fans have a power characteristic that increases as the fan flow rate increases. 6.1.2
Two fans in series The information given in this section applies only to configurations where both fans are operating* and where there is insignificant interaction between the fans†. Fans in series all pass the same volume flow, q , assuming no gains or losses in the duct system. The combined fan total pressure rise, ∆p f , t , will approach the sum of the individual fan total pressure rises. The fan static pressure rise for the combination is defined by Equation (3.2) and is thus not equal to the sum of the individual fan static pressure rises. Normally the fan dynamic pressure for the combination is taken to be that at the outlet of the most downstream fan. The combined fan total pressure rise versus flow rate characteristic for two fans in series may be constructed from the individual fan characteristics by summing, for a series of values of q , the values of the total pressure rise for each fan at that flow rate. In Sketch 6.2 the individual characteristic of the two fans is the curve A-A whilst the characteristic for the two fans in series is the curve B-B.
* †
For stand-by systems, see Section 5.3. This will usually be a reasonable assumption for centrifugal fans irrespective of how closely coupled they may be. However, tube-axial flow fans produce swirl which, unless damped by flow straighteners, can persist for many duct diameters. See Section 4.1.3 on contrarotating fans.
36
6.2
79037 Matching Fans to Fixed Systems Because a fixed system has a single characteristic, the operating point is determined solely by the fan characteristic as depicted in Sketch 3.1. If it is necessary to vary the flow rate through a fixed system, some means of controlling the fan must be provided either through speed control, adjustable guide vanes, variable pitch rotor control or adjustable dampers on either the inlet or outlet of the fan. The operating line resulting from the variety of characteristics offered by such a fan is shown in Sketch 3.2a.
Sketch 6.4 Choice of fans for required flow rate It is normally possible for more than one type of fan or fan combination to produce a satisfactory operating point. Sketch 6.4 illustrates the characteristics of a forward-curved centrifugal fan and an axial-flow fan intersecting the system characteristic to secure the same operating point. Provided there are very close tolerances in either the system characteristic or the fan characteristic, the choice of fan is not obvious and may be swayed by considerations of efficiency, cost, ease of installation, etc. However, if the system characteristic had been predicted from loss coefficients that were underestimated, the choice of fan would have been important, see Sketch 6.5.
Sketch 6.5 Choice of fan influenced by potential for error in system calculations Sketch 6.5 illustrates that the flow rate is less susceptible to estimation error in the system characteristics if the operating point is situated on the steep portion of the fan characteristic. However, it should be recognised that fan characteristics, although based on test measurements*, can themselves be susceptible to error due to manufacturing tolerance and, more importantly, differences between the installation and the *
Derivation 12 gives three classes of tolerance, A, B, and C. Manufacturers of small composite motor driven fans, e.g. most propeller fans, usually quote to class C tolerance which specifies that the volume flow rate, q , for a specified duty shall not be more than 7 per cent below the quoted value and the fan pressure rise shall not be more than 14 per cent below the quoted value. Class B tolerance, usually adopted for other types of fans, specifies corresponding tolerances of 5 per cent and 9 per cent for flow rate and pressure rise respectively.
37
79037 test set up. The fan and system tolerances together define a region of probable operating points. Again, as illustrated in Sketch 6.6, a smaller tolerance on the flow rate, q , results from the fan with the steeper characteristic.
Sketch 6.6 Matching fan to tolerance requirements Note that although the shallow and steep fan characteristics used in Sketches 6.4 to 6.6 have been attributed to forward-curved centrifugal fans and axial fans respectively, no general conclusions should be drawn about their relative shapes. The local gradients of fan characteristics are by no means constant and depend greatly on the flow rate. The sketches in Section 4 may be used as a guide but before any design is finalised, the manufacturer's data should be used. 6.3
Matching Fans to Systems with Varying Loss Coefficients In the systems with more than one characteristic described in Section 5.2, the operating line resulting from an uncontrolled fan may or may not meet the required operating conditions. The operating line depicted in Sketch 3.2b is one such example. Within certain tolerances, however, it is sometimes possible to select a fan with a characteristic that intersects the system characteristics to produce an acceptable operating line. This is discussed in Section 6.3.1. However where that is not practicable, it is necessary to resort to fan controls and they are discussed in Section 6.3.2.
6.3.1
Meeting system requirements with fixed characteristic fans A requirement for constant pressure rise is illustrated by line B in Sketch 5.2. This can be approximated by selecting a fan, or fan combination, with a flat characteristic in that range of volume flow, see Sketch 6.7.
Sketch 6.7 Matching constant pressure requirement
38
79037 Similarly, the requirement for constant volume flow rate illustrated by line A in Sketch 5.2 can be approximated by selecting a fan or combination of fans with a steep characteristic of negative slope in that range of volume flow, see Sketch 6.8.
Sketch 6.8 Matching constant flow rate requirement The approximation to constant flow rate is not as good as that to a constant pressure and in some cases it may be necessary to resort to a controllable fan. 6.3.2
Meeting system requirements with control devices Fan controls are necessary when the system requirements cannot be adequately matched to the natural fan characteristic. A controllable fan offers a flexibility of operation such that, over limited ranges of flow rate and pressure gain, most system requirements can be met. For example, a control characteristic that gives a truly constant volume flow rate with increased system resistance can be achieved, see Sketch 6.9.
Sketch 6.9 Use of fan controls to achieve constant volume flow system requirement However, if such a control characteristic is to be maintained automatically, a stability analysis should be undertaken. The starting procedure should also be modelled. Five types of control are commonly used: speed control, variable pitch rotor control, inlet guide-vane angle control, bleed control and throttle control. Speed control, which can be achieved either by mechanical or electrical means, is suitable for all categories of fans and is aerodynamically the most efficient means of flow control. For cross-flow fans, it is the only reliable means of flow control. Mechanical controls depend on friction and require regular maintenance. Electrical means of control are detailed in Section 8.2. 39
79037 The consequences of speed control can be investigated using the fan laws given in Section 3.3. Variable pitch rotor control of axial-flow fans provides a large range of flow control with little sacrifice of efficiency. Such fans, however, are costly and require more frequent maintenance than those with fixed rotor blades. The normal operating range variations available with such fans are illustrated in Sketch 6.10. Inlet guide vane controls can be fitted to backward-bladed centrifugal fans. The vanes act by pre-swirling the inlet flow in the direction of the impeller rotation to provide a more efficient means of flow control than throttling. However, with radial-bladed and forward-curved centrifugal fans, although it is also possible to govern the flow with inlet guide vanes, the power characteristics of these fans are such that more simple throttling devices can be used efficiently, see Sketches 4.14 and 4.16. The normal operating range variations available with such devices are illustrated in Sketch 6.11. Bleed or bypass control is a stable means of controlling the flow into a system from any fan and is often used in laboratory test rigs. In commercial operations, however, a bleed or bypass control is normally only applied to axial-flow and backward-bladed fans because they alone have characteristics that avoid power consumption penalties as the flow is diverted from the system.
Sketch 6.10 Rotor pitch control of an axial-flow fan (arbitrary units)
Sketch 6.11 Inlet guide-vane control of a backward-bladed centrifugal fan (arbitrary units)
40
79037 Throttle control is an inexpensive and often-used means of control that can be applied to all fans except cross-flow fans. Throttling devices, often referred to as "dampers", should normally be positioned downstream of the fan so that the flow distortion that they inevitably cause does not affect the fan. Throttling devices are inefficient when used with axial-flow and backward-bladed centrifugal fans because their power characteristics are such that little power will be saved, in the normal operating range, as the flow is throttled. With axial-flow fans, stalling problems may arise if the flow is over-restricted. However, when used in conjunction with forward-curved centrifugal fans, throttling devices offer a less costly alternative to other means of control without incurring power consumption penalties. The range of control available with throttling devices can be investigated by adding a pressure loss proportional to q2 to the system characteristic as illustrated in Sketch 6.12. The constants of proportionality, k1 to k3 , depicted in Sketch 6.12 are related to the loss coefficients and thus the position of the throttle. Item No. 690228 gives approximate data on the variation of loss coefficients with throttle position.
Sketch 6.12 Modification of system characteristic by throttling 6.4
System Trouble Shooting Due to uncertainties in estimating system resistance and the effects of installation on fan performance, it is inevitable that some systems do not meet their specifications but, just as there is often more than one cause of a problem, there are often several solutions.
6.4.1
Causes of low volume flow rate If the system resistance is underestimated and/or the fan performance overpredicted, a low volume flow rate will result. However, if following a check it is found that errors of estimation are within normal uncertainties, there are two possible causes. Either the fan has a flat characteristic near the operating point as illustrated in Sketch 6.5 or some components in the fan and system are not functioning as intended. It should be noted that, because of differences between the performance of a fan as measured under standard test conditions12 and that occurring in an installation, it is often difficult to apportion blame on the fan for an inadequate performance. Difficulties of obtaining representative measurements of pressure and volume flow rate in an installation compound the uncertainties. A frequent cause of trouble is an installation that distorts the flow entering the fan. Axial-flow fans are
41
79037 especially sensitive to inlet distortion on which Derivation 24 gives some practical guidance. Another cause of trouble is stalling in diffusers which prevents the conversion of dynamic pressure into static pressure. Item No. 760278 gives guidance on how to correct diffuser problems. Remedies that boost fan performance are offered in Section 6.4.2.
6.4.2
Boosting fan performance If the fan inlet flow distribution and diffuser performance are satisfactory, there remains the possibility of boosting the fan to increase the system volume flow rate. Where fans are driven by vee belts, this is most easily achieved by changing the pulley wheels to increase the impeller speed. The fan laws given in Section 3.3 indicate the amount by which the speed should be changed to give the required boost to the fan performance. The corresponding increase in rotor power should also be calculated and checked against the rated power of the motor. If the fan is driven directly by an electric motor, changing the speed is impracticable. Then it is necessary to increase the flow and pressure coefficients of the fan. For tube-axial fans, this is most conveniently achieved by adding guide vanes. Many manufacturers supply guide-vane units that may be attached directly to the fan casing. For guide-vane-axial fans, the effect of changing the number of blades and their pitch angles should be investigated using manufacturer's data. If these measures are insufficient, an additional booster fan* can be considered as an alternative to an increase in fan size which the fan laws in Section 3.3 may be used to calculate. For centrifugal fans, a variety of impellers may be available that run in the same housing but here manufacturer's data must be used to ascertain the effect of an impeller change.
*
If the booster fan is installed near the original fan, the directions of rotation of the two impellers should be contrarotating.
42
7.
79037 FAN SELECTION In selecting the best fan, the mechanical suitability, physical dimensions and performance characteristics of the available fans should be established. The mechanical suitability of the fans should be checked against their intended application. For example, for duties involving dust laden air, certain fan types are unsuitable. Guidance on this is given in Section 7.1. The primary physical criterion is that the fan can be installed in the available space - convenience of installation will, in many designs, be the deciding factor for choosing between axial and centrifugal fans. A procedure for obtaining approximate dimensions and approximate impeller speeds based on manufacturers' data (Section 11.3) for the different fan types is given in Section 7.2. The primary performance criterion is that the fan should supply the required volume flow rate against the system pressure drop at an acceptable efficiency. Selection on this basis is explained in Section 7.3. Guidance on noise, which can also influence selection, is given in Section 9. These procedures are illustrated by the examples in Section 10.
7.1
Mechanical Suitability For duties involving clean and dry air at temperatures typically between 4°C and 40°C, and at absolute inlet pressures of typically between 700 mbar and 1300 mbar, most fans from manufacturers' standard ranges will be mechanically suitable. Outside those ranges of temperature and pressure, the manufacturer should be consulted to see if alternative materials or finishes are necessary for the fan, its bearings and its electric motor. Similarly, for duties involving corrosive, wet or inflammable gases, manufacturers' guidance should be sought before any design is finalised. There are no simple rules to imply that one fan type is necessarily more suited than another to such arduous conditions. The choice will often depend on the availability of sufficiently robust fans and those of designs that allow convenient impeller removal and replacement. When a fan is to be used for duties involving particulate matter, centrifugal fans are normally employed. Straight radial-bladed centrifugal fans with open impeller construction are in general better suited to handling flows containing large particles or fibres, while the more efficient shrouded impeller types may be used for less arduous dust control duties, see Section 4.2.2. Backward-bladed impellers with straight or curved blades may also be used in the less severe environments. With inlet flow distortion, the loss in performance and increase in vibration will be less severe from a centrifugal fan than from an axial-flow fan. When such a performance handicap is expected, a fan should be selected that gives a generous estimated performance margin, see Section 6.2.
7.2
Approximate Physical Dimensions Because it is usually possible for more than one type of fan, or combination of fans, to provide a satisfactory system operation, the choice of fan may be governed by physical constraints. Where this is not so, a fan, or combination of fans, may be selected on the basis of maximum efficiency, see Section 7.3. The physical proportions of each axial-flow and mixed-flow fan type are approximately constant. In centrifugal and cross-flow fans, the only governing proportion that varies significantly is the width to diameter ratio of the impeller, W T ⁄ D T , which determines the proportions of the casing. Fundamental therefore to the estimation of the physical dimensions of a fan is knowledge of the impeller diameter, DT , and, in the case of a centrifugal or cross-flow fan, the width to diameter ratio, W T ⁄ D T . By correlating values of specific diameter*, ds , against specific speed, ns , coinciding with high efficiencies
*
See Section 3.2 for definitions and explanation.
43
79037 for each fan type of known W T ⁄ D T , it is possible to deduce an approximate relationship between the impeller diameter and the impeller speed, N , that is best suited to the required flow rate and pressure rise. Using this relationship between the impeller diameter and impeller speed, the diameter can be estimated. Notes detailing the procedure for estimating D T are given in Section 7.2.1 and the remaining fan dimensions can be estimated using the typical proportions listed in Section 7.2.2.
7.2.1
Procedure for estimating impeller diameter, D T The following notes describe the diameter estimation procedure for a given fan type. Following the diameter estimation, it may become apparent that the choice of fan type should be reviewed and/or impeller width/diameter ratio changed and/or impeller speed changed. It may become evident that two or more fans would be better suited to the required duty; in that case, the fan pressure rises and volume flow rates referred to in the notes and charts should be taken as those appropriate to each fan in the series or parallel arrangement chosen.
Step
Procedure
1.
Select a fan type. If there are no overriding mechanical constraints, the qualitative information given in Section 4 may be helpful.
2.
Referring to Figure 1, decide on a value for ns using the graph on the far left of the chart. A value of n s corresponding to the middle-range value of 1⁄ d s for the chosen fan type will result in a good compromise between impeller diameter and speed. However, a smaller diameter can be obtained by specifying a higher value for ns .
3.
Again referring to Figure 1, it can be seen that the lower portion of the graph on the far right of the chart requires a value for W T ⁄ D T . For axial-flow fans and mixed-flow fans, take W T ⁄ D T = 1 . For single inlet centrifugal fans, any number for W T ⁄ D T between approximately 0.12 and 0.65, as typified by the values quoted on the chart*, may be chosen. For cross-flow fans, impellers with W T ⁄ D T ranging between 0.7 and 10 are commonly available. Note that the selection of a wide impeller reduces the diameter but increases the speed. For applications where a high pressure rise is required, considerations of speed may thus restrict the practical width of the impeller.
4.
Evaluate σ , the relative density of the flow entering the fan. For sea-level installations running at outside ambient conditions, σ may be taken as unity for the purposes of this estimation procedure.
5.
For the maximum flow rate required, qmax , calculate the necessary fan pressure rise. For many systems, the pressure rise will be conveniently related to the fan static pressure rise (see Section 3.1.2) but, where this is not so, the fan total pressure rise, ∆ pf , t , may be calculated and converted to ∆ pf , s using the correlation presented in Figure 4.
6.
Evaluate ∆p f , s ⁄ σ ..
7.
Pinpoint the co-ordinate on the upper far right-hand graph of Figure 1 corresponding to the selected value of W T ⁄ D T and q max and ∆p f , s ⁄ σ . The point marked A exemplifies the use of the grid for a value of W T ⁄ D T = 0.5 , qmax = 8 m3/s and ∆pf , s ⁄ σ = 500 Pa.
For footnotes see end of procedure.
44
79037
8.
From the far left-hand graph of Figure 1, draw a horizontal line from the chosen l⁄ ds and n s co-ordinate across to one of the two Reference Lines. The point marked B exemplifies this for a l⁄ ds versus ns combination selected for a forward-curved centrifugal fan. Where this horizontal line intersects the appropriate Reference Line, draw a line parallel to the inclined lines on the centre graph.
9.
Draw a horizontal line from point A ascertained in Step 7 until it intersects with the inclined line drawn in Step 8. (This is the point marked C in the running example.) This point of intersection lies directly above the scale giving DT which can now be read.
10.
To find the corresponding impeller speed turn to Figure 2. Again pinpoint point A on the far right-hand graph using the same approach as in Step 7.
11.
Using the same value of ns as before, locate point B on one of the two Reference Lines on the centre graph and, again following the same approach as in Step 8, draw a line parallel to the inclined lines.
12.
Draw a horizontal line from point A to intersect with the inclined line at point C. If the right-hand Reference Line was used, the intersection lies directly above the scale giving the impeller speed in revolutions per minute. Alternatively, if the left-hand Reference Line was used, the scale gives the impeller speed in radians per second.
13.
If the fan is directly coupled to the motor, the impeller speed will be governed by the pole speed of the motor. Then the nearest pole speed must be selected† and the corresponding value for n s ascertained by working backwards through Figure 2. Using this revised value for ns , the corresponding diameter can be found by repeating Steps 2 to 9.
For footnotes see end of procedure. *
For double inlet centrifugal fans, the limits of this range may be doubled.
†
See Section 8.
If the impeller speed and diameter are satisfactory, the remaining physical dimensions can be estimated using the procedure of Section 7.2.2. If the impeller speed is very high, this may indicate the need for a change in fan type or for more than one fan to be employed in series. If the impeller speed is low, parallel operation of two smaller fans should be considered although a double inlet centrifugal fan may suffice. Guidance on the use of fans in series and fans in parallel is given in Section 6.1. 7.2.2
Estimation of fan external dimensions The table below gives approximate ranges for the dimensions A , B and C as detailed on the Sketches in Section 4.
7.3
Approximate Power Requirements It is possible to estimate the impeller power of a fan using a similar approach to that given in Section 7.2. For power estimation, the crucial variable is the fan efficiency, which is correlated in terms of the specific speed, ns . The procedure given in Section 7.3.1 requires a value for the specific speed, ns , for which a provisional estimate should be available from the procedure in Section 7.2.1.
45
79037
Fan
Sketch No.
A ⁄ DT
B ⁄ DT
Propeller
4.4
1.70 – 0.50D T for D T ≤ 0.9 m 0.77 – 0.30D T for D T ≤ 0.9 m 1.25 for D T > 0.9 m 0.5 for D T > 0.9 m
Tube axial
4.6
1.25 – 0.15D T for D T ≤ 1.0 m 1.1 for D T > 1.0 m
Con-rotatng
4.8
Guide-vane
4.10
Fwd-curved
4.12
Rad-bladed
4.12
Back-bladed
4.12
Mixed-flow
4.21
Cross-flow J, S and U
4.24
7.3.1
as for tube axial
C ⁄ DT
1 to 0.4 0.5 to 2.0
1.28 – 0.20D T for D T ≤ 0.9 m 1.1 for D T > 0.9 m
0.6 to 1.8
1.4 to 2.5
1.3 to 2.1
1.20W T ⁄ D T + 0.1 to 1.20W T ⁄ D T + 0.9
1.2 to 2.1
1.2 to 1.6
0.5 to 1.0
1.6 to 2.0
1.5 to 2.0
1.65W T ⁄ D T + 0.20 to 3.33W T ⁄ D T + 0.35
1.1 to 1.3
1
1.3 to 1.65
1.75
W T ⁄ DT + 0.25
Procedure for estimating impeller power, P R The following notes describe the impeller power estimation procedure for a given fan type. From the power estimate, it may become apparent that the choice of fan type should be reviewed and/or specific speed varied. This will influence the impeller speed and diameter and will necessitate repeating the procedure given in Section 7.2.1.
Step
Procedure
1.
Referring to the far left-hand graph of Figure 3, find the fan static efficiency, ηs , corresponding to the value of ns used for estimating the impeller diameter for the fan type selected.
2.
Again from Figure 3, pinpoint the co-ordinate on the far right-hand graph corresponding to the values of qmax and ∆ pf , s . (The point marked A exemplifies the use of the graph for qmax = 8 m3/s and ∆p f , s = 500 Pa.)
46
79037
3.
From the far left-hand graph, draw a horizontal line from the value of ηs obtained in Step 1 across to intersect with the correct Reference Line for the fan being treated (point B). At this intersection, draw a line parallel to the inclined lines on the centre graph.
4.
Draw a horizontal line from point A ascertained in Step 2 to intersect with the inclined line drawn in Step 3. (This is point C.) This point of intersection lies directly above the scale giving PR which can now be read.
5.
If the efficiency is satisfactory, the procedure is complete. If, however, a higher efficiency is required, the consequences of changing n s , and/or the fan type, to achieve higher values must be re-investigated through the procedure in Section 7.2.1.
47
8.
79037 GUIDANCE ON ELECTRIC MOTOR SPECIFICATION FOR FANS Almost all fans are powered by alternating current induction motors of squirrel-cage rotor construction, most of which are three-phase machines to which Sections 8.1 to 8.4 are directed. Small fans, e.g. those absorbing less than 1 kW, however, are usually powered by single-phase machines which are discussed in Section 8.5. Motors should primarily be chosen with respect to their power output and operating speed, i.e. full-load rating*. Normally a 20 per cent maximum power margin between the fan and motor should be allowed for an indirectly driven unit and a 10 per cent margin for a directly coupled unit. For fans operating in dusty environments, however, more generous margins should be used to allow for performance and efficiency deterioration, see Section 4.2.2. The fan maximum power should be calculated with consideration given to (i)
the shape of the impeller power versus flow rate characteristic so that if the operating point varies, for example due to controls or modification of the system, adequate power is available,
(ii)
contingencies such as the blocking of or inadvertent removal of filters, or the breakdown of another fan in the system,
(iii)
the maximum density of the flow arising from, for example, extremes of ambient temperature, pressure and humidity. This is in contrast to the environmental conditions that should be checked in relation to the maximum allowable motor operating temperature but see Section 8.4.
Induction motors, although not directly synchronised to the supply frequency, are normally rated to run at speeds fairly close to the synchronous speed. Three-phase induction motors at their full rated loads normally run at approximately 2 per cent slip†, i.e. 0.98 times synchronous speed, but this value for slip may be less for large motors, i.e. > 40 kW, or more for small motors, i.e. < 10 kW. The corresponding rotor speeds are given by N = 0.98 × 120 × f /no. of poles where N is the motor speed in rev/min and f is the supply frequency in Hz. The number of poles is typically, 2, 4, 6 or 8. Two pole machines, however, are only available for low power duties. Because the speeds of induction motors are related to the supply frequency, efficient speed control is difficult. However, controls employing frequency regulation are available and these and other means of control are discussed in Section 8.2. 8.1
Starting In general, squirrel-cage induction motors are used wherever practicable due to their low cost, high efficiency and good reliability. However, whilst giving satisfactory running performance, such motors do present starting problems. The speed/torque characteristic of a typical squirrel-cage induction motor is illustrated in Sketch 8.1. In general, a high efficiency motor, i.e. one with low resistance windings, has a low starting torque because, at the stalled condition, the nearly purely inductive rotor impedance gives it a transformer-like characteristic. Two kinds of problems arise from low starting torques: prolonged electrical heat generation and slow system response‡. In most applications, however, it is only the first of these that requires attention. In addition, the
*
Motors intended for speed regulation should be selected with respect to the complete fan power versus speed characteristic after consultation with the manufacturers. This is because speeds less than the maximum speed will give rise to higher motor temperatures, see Section 8.2. † Induction motors designed for voltage regulation speed control, see Section 8.2.2, have high resistance windings that run with typically ten per cent slip at their full-load rating.
48
79037 associated high starting currents, which can amount to ten times the full-load current, require special consideration. The heat generation rate is directly related to the rotor slip, i.e. the difference between the actual rotor speed and its synchronous speed. Clearly the longer the run-up time (which will be prolonged if electrical starting devices are used to limit the starting currents) the higher the temperatures will rise in the motor. Thus it is essential to select a motor with consideration given to the polar moment of inertia of the fan and its transmission, see Table below. The time intervals between starts must also be considered, see Section 8.3.
Sketch 8.1 Torque versus speed characteristic of a squirrel-cage motor and its fan To protect the electrical system from high starting currents, it is necessary with motors of powers typically greater than 5 kW to reduce the voltage on starting. The most common electrical starting devices are star/delta starters and autotransformer* starters. Normally, only two settings, i.e. "start" and "run", are offered by either device although autotransformers can be provided with several voltage tappings. A star/delta device is less costly but depends on both ends of each motor winding being made available for connection. If either starting device is feasible, the choice may be swayed by the cost of the extra wiring associated with star/delta starters versus the cost of autotransformer starters.
‡
Some fan bearings and particularly vee-belt transmissions may have high breakaway torques. This torque can be overcome by employing a wound-rotor motor whose effective resistance is varied externally through slip ring connections, see Section 8.2.2. * An autotransformer has a primary winding that is common to a portion of the secondary winding. It is thus less costly than a conventional transformer whose windings are mutually isolated.
49
79037 The following table may be used as a guide to the relative suitabilities of different starting techniques.
Power ratings 0 – 1 kW
Suitable starting arrangement Single-phase, no external starting arrangement
0.5 kW to 5.0 kW
Three-phase, no external starting arrangement
5.0 kW to 37.5 kW
Three-phase, star/delta starter
above 30 kW up to 75 kW
Three-phase, wound rotor motor with external resistance. Used only if star/delta and autotransformer are unsuitable for torque reasons
above 37.5 kW
Three-phase, autotransformer starter
The above values should be taken only as a guide and it should be noted that in many cases, the choice of starting arrangement is governed by electricity supply mandates.
Sketch 8.2 Fan motor starting characteristics The problem of insufficient starting torque has several solutions: (i)
an aerodynamic solution in which the fan torque curve is flattened by the use of a variable pitch fan, the temporary provision of a bypass, or temporary throttling,
(ii)
the specification of a larger size of electric motor,
(iii)
the use of an autotransformer starter with several voltage tappings,
(iv)
the use of a wound-rotor induction motor, see Section 8.2.1,
(v)
the use of a slipping coupling, for example, a fluid or powder coupling.
A frequently-used rule of thumb9 to check that starting will not cause overheating is to specify a motor that has a run-up time within those obtained from the following table. However, longer run-up times may be satisfactory in many installations but where this is envisaged the motor manufacturer should be consulted.
50
79037 Starting method
Maximum allowable run-up time (in seconds)
Direct on-line
Rated Power ( kW ) 3 + -----------------------------------------------7.5
Star/delta or autotransformer with approximately 60 per cent full voltage
Rated Power ( kW ) 9 + -----------------------------------------------2.3 Rated Power ( kW ) 8 + ------------------------------------------------ for powers up to 37.5 kW 3 Power ( kW ) or 18 + Rated ------------------------------------------------ for powers above 37.5 kW 15
Wound rotor motor
The actual starting or run-up time may be calculated from the following expressions. 2
–3
IN × 10 T start = ------------------------------------------------------τ × Rated Power (kW) where
8.2
(8.1)
N
is the motor speed (rad/s) at full load,
I
is the polar inertia of the fan, drive and motor (referred to motor) in kg m2 and
τ
iS the average torque available for acceleration divided by the full-load torque. Typically τ = 0.25 for star/delta starting and 1 for direct on-line starting.
Speed Control There are five means of electrical speed control that may be applied to fans. They are:
8.2.1
(i)
control employing a wound-rotor induction motor,
(ii)
control employing voltage regulation to an induction motor,
(iii)
control employing frequency regulation to an induction motor,
(iv)
control employing a two speed induction motor,
(v)
control employing a direct current motor.
Wound-rotor induction motors The resistance, and hence current and torque, of a wound-rotor induction motor can be varied by connections, made through slip rings, between the motor windings and external resistances. This enables the speed to be governed by the torque versus speed characteristic of the fan, see Sketch 8.3.
51
79037
Sketch 8.3 Torque/speed characteristics of wound-rotor motor Due to the increase in slip inevitable with reducing the speed of any induction motor with a fixed frequency supply, the efficiency deteriorates as the speed is lowered. However, when the fan power characteristic is taken into account, power savings nevertheless result. 8.2.2
Voltage regulation Like the method employing wound-rotor motors, this is essentially a means of torque control that enables the speed to be governed by the torque/speed characteristic of the fan, see Sketch 8.4. To alter the voltage, thyristors are often employed in the supply to the induction motor.
Sketch 8.4 Torque/speed characteristics with induction motor voltage regulation The slip associated with such means of control leads to low efficiencies at low speeds but, as with the wound rotor induction motor control, power savings result. Unlike the wound-rotor motor, however, the wasted power, which reaches a peak at approximately 75 per cent of the maximum speed, is dissipated as heat within the motor rather than in external resistances. The provision for such heat dissipation should be arranged in consultation with the motor manufacturer. Generally, as motor sizes rated at three times the fan maximum power are usually specified, cost considerations rule out the application of this form of speed control for powers in excess of approximately 30 kW. The motors are normally equipped with high resistance windings in order to widen the stable operating speed range.
52
8.2.3
79037 Frequency regulation Because the speed of an induction motor, when operating at its rated condition, is closely related to the supply frequency, it may remain largely at its rated condition yet vary in speed if the supply frequency varies. A typical torque/speed characteristic is illustrated in Sketch 8.5.
Sketch 8.5 Torque/speed characteristics with induction motor frequency regulation This type of control is very much more efficient than one involving high slip. However, as the effectiveness of the motor cooling may decrease as the speed is reduced, the provision of adequate cooling should be checked with the motor manufacturer. The high costs of this type of control normally restrict the application to high power duties, e.g. above approximately 40 kW. 8.2.4
Two-speed induction motors In many fan applications, a limited range of speed control is sufficient. There are three types of two-speed motors currently in production that offer a stepped means of speed control: dual-winding motors, pole changing (or Dahlander) motors and pole amplitude modulation (PAM) motors. The Dahlander winding is a special case of the PAM winding that offers a speed ratio of 2:1 only. A PAM winding can be designed to produce a speed ratio corresponding to any number of poles combination. For example, a 6 pole PAM motor can be wound to be switched from a 6 pole speed to a 4 pole speed. The traditional dual-winding motor has the same flexibility but suffers from problems of space that can compromise the magnetic circuits. Efficiency is thus often a little lower than Dahlander or PAM motors operating at high speed for which their windings are optimised. However, star/delta starting on pole changing motors* is not possible without special designs. Therefore for medium to high power requirements, e.g. above about 5 kW, dual-winding motors are often specified. Although two-speed motors are more expensive than single-speed motors, the additional cost will usually be significantly less than other means of speed control. In addition, the control does not suffer from the efficiency and heat dissipation problems associated with slip.
*
It may nevertheless be acceptable, provided high current transients are acceptable, to start such motors at their lower speeds before switching to their higher speeds.
53
79037
Sketch 8.6 Torque/speed characteristics of a pole amplitude modulated motor 8.2.5
Controls employing direct current motors The efficient control of d.c. motors is relatively simple compared with a.c. machines and starting problems are obviated because the control range extends from the maximum to zero speed. When series wound, d.c. motors have a limited power consumption which makes them especially suitable for marine and other installations where the power supply may have a limited capacity. However, d.c. machines, for a given rated power, are more costly and require more maintenance than squirrel-cage induction motors. Although d.c. motors are conventionally controlled by external resistors, an increasingly popular method uses thyristor rectifiers. Here, the harmonics generated by the controls may induce extra motor heating and reduce brush life and, although most modern motors are designed appropriately, the suitability of a particular motor for this means of control should be checked with the manufacturer. This means of control is, in general*, less expensive than frequency regulation for motors rated below about 40 kW.
8.3
Overload Protection Modern electric motors are rated to operate at high temperatures and, because of their compact size, are more susceptible to overheating than their earlier counterparts. Simple fuses or circuit breakers which must withstand high starting currents will not in general provide adequate protection against overheating. Whether or not overheating protection is an economic proposition largely depends on the size and therefore cost of the motor. Thus on larger motors temperature transducers are often embedded in the windings. Such protection is particularly desirable for forward-curved and radial-bladed centrifugal fan motors which absorb more power as the system resistance decreases. Overloading may therefore be caused by the accidental opening of an inspection door or a runaway in an automatic control system. The high polar inertia of large versions of these types of fans means that temperature transducers may also be necessary to protect the motor from overheating as a result of too frequent starting.
8.4
Motor Construction Motors are available with various enclosures and classes of insulation to suit the environment within which they will be operated. These are defined in the appropriate British Standards but are summarised as follows:
*
Due to the high costs of d.c. motors suitable for use in hazardous environments, frequency regulation may be more economic under such conditions for powers less than 40 kW.
54
79037 Enclosure
Screen-protected – for use in relatively clean and dry locations, Drip-proof
– for use as above, but where liquid is likely to drip onto the motor,
Totally-enclosed
– for use in areas where dirt, water or vapour makes enclosure desirable,
Flameproof
– for use in a hazardous area where there are inflammable gases or risks of explosion,
Weatherproof
– for use without further protection from weather conditions.
Insulation The standard ambient temperature for motor design is 40° C . Associated with this are seven classes of insulation, summarised below, that classify the maximum temperature rise at which normal service life can be expected.
Class
Temperature rise above 40 °C
Y
50° C
A
65° C
E
80° C – in normal use
B
90° C – in normal use
F
115° C – in normal use
H
140° C
C
> 140 ° C
For motors operating at high altitude, the allowable temperature rise will be exceeded if the motor is loaded to produce its rated power because the cooling effectiveness is reduced by the lower air density. However, if the fan maximum power as ascertained in note (iii) in Section 8 is related to the flow density at sea level, the motor will not overheat at any higher elevation because the impeller power requirement, for a given point of rating* decreases faster than the motor capability. Similarly, a motor selected for a fan whose maximum power is related to the flow density at 40° C will not overheat at lower temperatures provided the motor is also subject to the lower temperatures. This is a useful feature when a high temperature system is started cold. 8.5
Single-Phase Motors Single-phase squirrel-cage motors are almost always used for fan applications requiring less than 500 watts. The properties of the most frequently used types are summarised in the following table. Other types of single-phase motors include a.c. series motors (universal motors) which are used where speeds in excess
*
See Section 3.3 for definition.
55
79037 of the maximum a.c. synchronous speeds are required and split-phase induction motors which have characteristics somewhat similar to capacitor-start motors. Single-Phase Motors for Fans
Motor type
Rated Rated speeds output 50 Hz (W)
Description
(rev/min)
Shaded
Usually of open
pole
construction,
induction
starting torque is
motor
provided
Rated Starting speeds Efficienc torque 60 Hz y (per cent) (per cent
by
to
permanently short-circuited
of full load)
(rev/min)
0.75
up 875
1050
1300
1550
2600
3100
200
Application Disadvantag Advantages s es
30
to
to
Low
power
fans,
Low efficiency,
e.g.
multispeed
low
propeller and
capability,
torque.
cross-flow
quite small.
requiring 40
Inexpensive,
starting
no
maintenance.
80
auxiliary winding. Capacitor
Usually
start,
enclosed
of
induction
construction,
run
auxiliary
has
starting
winding in series with
external
capacitor. This is isolated
by
centrifugal switch
Fans in small
High starting
Non-adjustable
commercial
torque.
speed.
equipment, 40 to
35 950
1140
1425
1725
2850
3450
800
suited to belt
165
to
to
50
240
drives to
owing high
starting torque.
as normal running speed
is
approached. Capacitor
Of
start,
construction,
capacitor
two windings, one
run
that
enclosed
is
has 90°
20
electrically phased from other
to
by capacitor. 2000
45 900
1075
1350
1625
2700
3250
to 60
30
Medium
High starting
Speed control is
to
directly
torques
not
250
driven fans.
are
possible
possible. Can
with
(high value
be
two-capacitor
only
efficiency and
versions.
additional
fitted
with
Single capacitor
starting
speed
and
capacitor
reversal
low
with
controls.
torque and load
with
centrifugal switch)
56
size
of
high
versions
have starting
sensitive speed.
79037
9.
GUIDANCE ON NOISE
9.1
Introduction There is an increasing emphasis placed on acceptable noise levels so that, because the noise depends on the system configuration, the design of duct systems should not be undertaken in isolation from noise considerations. To enable duct system designers to understand more fully the principles of noise generation and propagation within duct systems this Section provides an introduction to acoustic terminology and relationships used in fan noise estimation. A means of estimating the fan sound power level, and general guidance on the selection and operation of fans for minimum noise, is also provided. The noise within a duct system can originate from several sources other than the fan. Noise will be generated by the fan drive, bearings and gearing but for other than very low speed fans these noise sources are likely to be dominated by aerodynamic noise generated by the fan. Noise will also be generated by turbulent air flows resulting from obstructions, such as bends and dampers within the duct, and branching of the duct system; additional noise may be generated due to structural resonance effects induced by periodic forces associated with the blade passing frequency or vortex shedding. Noise may also be introduced into the system from external sources; for example, when the duct system passes through an area with high ambient noise levels. The noise emitted from the system may be transmitted to the serviced areas, or other areas through which the ducting system passes, through the air, the system structure or directly via the duct. Airborne noise principally emanates from the plant room and can be attenuated (i.e. reduced) by silencing the duct downstream of the fan and acoustically insulating the plant room. However, if the system is poorly designed, the flow within the duct may cause the duct panels to vibrate so providing an additional source of airborne noise. Structure borne noise may be reduced by isolating the driver and fan from the system by the use of anti-vibration mounts and by using flexible couplings at suitable positions in the duct system. Direct ductborne noise is introduced into the serviced area through louvres and may be reduced by the use of special acoustic louvres or by noise suppressors between the fan and ventilators. Although knowledge of the fan noise alone will clearly not be sufficient to determine the noisiness of a system, the fan sound power level, L w , is an important input to such calculations. Procedures for determining the noisiness of ventilation systems are available in References 3 and 6 and a method of rating industrial noise and its effect on persons living in the vicinity is given in Reference 2.
9.2
Noise Units and Basic Relationships – 12
8
Sound powers radiated in air have a practical range of about 10 W to 10 W. Sound powers for some familiar noise sources are tabulated in this Section. Because this range of powers is so large it is convenient to express sound powers in decibel units (dB). This unit may be used to compare the relative magnitudes of any quantity related to power. The number of bels (1 bel = 10 decibels) is the logarithm to the base 10 of the ratio of the powers. The two quantities used in this Item that are measured in dB are sound power level, L w , and sound pressure level, Lp . These quantities are defined as follows:
and
W Lw = 10 log10 ------- W0
(9.1)
p Lp = 20 log10 ----- , p0
(9.2)
where W 0 and p0 are reference sound power and root mean square sound pressure respectively. The preferred* reference power, W 0 , is 10–12 W and the preferred reference sound pressure in gaseous media
57
79037 is 20 µ Pa*; these values are used in this Item. In air, sound pressure level is the easiest quantity to measure so that the noise of fan and duct systems is measured in terms of L p .
The quality of noise is dependent both on its volume and frequency content (i.e. its spectral composition). For a full description of a noise source, it is therefore necessary to quote sound power (or pressure) levels within specific frequency bands. The frequency bands used for fan noise description are octave bands (i.e. bands of frequency having an upper frequency twice that of the lower frequency). The preferred frequency limits for each octave band, and their nominal centre frequencies, are tabulated for the eight octave bands of interest. *
The preferred reference values correspond to the youthful hearing threshold and are expressed in SI units in this Item. It should be stressed that for all quantities expressed in dB the reference value used should be stated when absolute values are given. –5 –4 * 1 Pa is 1 pascal = 1 N/m2; therefore 20 µPa = 2 ×10 N/m2. The equivalent value of p 0 may be quoted in other units such as 2 ×10 –7 2 2 dyn/cm or 4.18 ×10 lbf/ft .
58
79037
Lower band limit, Hz
44
88
177
355
710
1420
2840
5680
Centre frequency, Hz
63
125
250
500
1000
2000
4000
8000
Upper band limit, Hz
88
177
355
710
1420
2840
5680
11 360
It should be noted that the band centre frequencies, by which the bands are usually referred to, are the nominal geometric mean of the upper and lower band frequency limits, i.e. fcentre = ( f upper × flower )½ . 9.3
Reaction to Noise Values of L p may be quoted for each octave frequency band but it is often more convenient to quote a single number. In order that the number meaningfully represents both the absolute sound pressure level and frequency content of the sound it is necessary to apply a correction, based on the observer response, to assign to individual octave bands an equivalent loudness. This corrected sound pressure level is said to be weighted. Since the human ear is most sensitive to frequencies in the higher range of audible frequencies, to make a subjective assessment of noisiness it is necessary to reduce the sound pressure levels of the lower octave frequency bands before combining them with the higher octave frequency bands. Corrections for the octave bands of interest for the A, B and C weighting curves are given in Table 9.1. In using these curves the relative sound level is added (arithmetically) to the band sound pressure level to give the weighted band level. The weighted band levels are then summed to give the weighted sound pressure level using Appendix A or Equation (9.3). L p1 L p2 L p8 L p = 10 log10 antilog 10 --------- + antilog 10 --------- + ... + antilog 10 --------- , 10 10 10 8
i.e.
L p = 10 log10
∑ 10
L p j ⁄ 10
,
(9.3)
(9.4)
j=1
where L p is the weighted sound pressure level and Lp1 , Lp2 , ... Lp8 are the weighted band sound pressure levels in each of the octave bands. After summation of the band levels, the resulting noise level is known as dBA, dBB and dBC respectively for the A, B and C curves. The most commonly used sound pressure levels are A-weighted. As an alternative to weighted sound pressure levels, the annoyance of fan system noise may be assessed using noise criteria or noise rating curves. Noise assessment on the basis of these curves is more influenced by the spectral content of the noise than a weighted sound pressure level. Details of noise criteria and noise rating curves are given in Derivation 10 and Reference 5 respectively. 9.4
Summation of Sound Power Levels The total sound power level from more than one source may be found by first converting the individual sound power levels to sound power, then summing* (arithmetically) to obtain the total sound power (in
*
The sound power from contrarotating axial fans may not be obtained in this way because the closely spaced impellers interfere. On average a two-stage axial fan with contrarotating blades will be 5 to 6 dB noisier than a guide-vane fan working at the same duty.
59
79037 watts) and finally reconverting to a total sound power level. This procedure is expressed mathematically as total
sound
power
L w1 L w2 L wJ = 10 log10 antilog 10 --------- + antilog 10 --------- .... + antilog 10 --------- , 10 10 10 J
i.e.
total sound power level = 10 log10
∑
10
L w j ⁄ 10
level (9.5)
,
(9.6)
j=1 th
where J is the number of sources and Lwj is the sound power level of the j source. A graphical procedure is given for this summation in Appendix A. The summation should be carried out separately for each octave band; however, if a single overall sound power level is required the octave band power levels may be summed using the same procedure. 9.5
Fan Noise Testing The three preferred methods for measurement of fan noise, in-duct, free-field and reverberant-field methods, are fully described in Derivation 16. The primary objective of fan noise testing is to ascertain values of sound power level in each of the octave frequency bands of interest. It should be noted that in Derivation 16 fan manufacturers are only required to provide noise data for the octave bands in the range of band centre frequencies from 125 Hz to 4000 Hz. Many manufacturers also provide data for the bands centred at 63 Hz and 8000 Hz but, due to difficulties in measuring noise levels in these latter bands, the data provided may not be guaranteed to the same accuracy as those given for the other octave bands. Sound pressure levels are first measured in each octave band and the equivalent sound power level estimated using appropriate relationships*. Measurements to determine either the inlet or outlet fan sound power levels may be made. Depending upon the test method used, either inlet sound power levels or open-inlet sound power levels are determined, and similarly for outlet and open-outlet sound power levels. These sound power levels are associated with the appropriate sound powers. The inlet sound power of a fan is the rate at which sound energy enters a uniform airway connected to the fan inlet; the open-inlet fan sound power is the rate at which sound energy is radiated into the ambient atmosphere from the open-inlet and the casing of the fan. Similar definitions apply to the outlet and open-outlet sound powers. A means of conversion between inlet (or outlet) sound power levels and open-inlet (open-outlet) sound power levels, for ducts of uniform cross section terminating in open space, is provided in Figure 5.
9.5.1
In-duct fan testing For in-duct testing the relationship between measured sound pressure level and sound power level is given by L w = L p + 10 log10 A ,
(9.7)
where A is the cross-sectional area† of the test airway. *
All equations given in this Item relating sound pressure level to sound power level are approximate since these relationships are dependent on the temperature and pressure at the time of measurement. For general engineering applications within the usual range of temperature and pressure, this dependence can be ignored. Further, it is assumed that sound pressure level measurements are taken at a distance remote from the sound source (i.e. acoustic far-field) such that particle velocities are predominantly in the direction of propagation of the sound and the acoustic intensity is proportional to sound pressure squared. It is usual to assume that far-field conditions exist at distances more than one sound wavelength away from the source or at 3 times the fan diameter, whichever is the greater.
60
9.5.2
79037 Free-field fan testing For free-field testing the fan assembly may be mounted such that sound radiates freely in all directions or it may be mounted on a flat sound reflecting surface within a free-field test facility. For the first case sound pressure levels are measured at points on a hypothetical sphere centred on the sound source and for the latter case the sound measuring points are located on a hypothetical hemisphere centred on the sound source. For either testing method, the sound pressure level within an octave frequency band, Lpm , is the logarithmic mean of the j values ( Lp1 , L p2 , ... Lpj ) measured on the sphere, or hemisphere, for that frequency band, i.e. L p1 L p2 L pj 1 L pm = 10 log10 --- antilog 10 --------- + antilog 10 --------- + ... + antilog 10 ------- . 10 10 10 j
(9.8)
The open-inlet or open-outlet sound power level is then estimated using one of the following relationships: for spherical sound radiation L w = L pm + 20 log10 r + 11 ,
(9.9)
and for hemispherical sound radiation L w = L pm + 20 log10 r + 8.0 ,
(9.10)
where for both expressions r is the radial distance of the measuring points from the sound source. 9.5.3
Reverberant-field fan testing For reverberant-field testing, sound pressure levels are measured at a number of locations in the diffuse sound field. In each octave frequency band the mean reverberant-field sound pressure level, Lpm , is determined from the averaged value of the several band sound pressure levels ( L p1 , Lp2 , ... L pj ,) measured within the test chamber, using Equation (9.8) where j is the number of measurement locations. The open-inlet or open-outlet sound power levels in each octave band are then estimated using the relationship L w = L pm – 10 log10 T + 10 log10 V – 14 ,
(9.11)
where V is the volume of the test enclosure and T is the reverberation time. The reverberation time in an enclosure, for sound of a given frequency, is the time required for the mean square sound pressure in the enclosure to decay 60 dB from its initial steady state after the source is stopped. 9.6
Understanding Fan Noise Data In providing noise data for their fans, manufacturers usually quote sound power (or pressure) levels for various operating conditions. They also provide a spectrum shape so that the sound levels within each octave band may be deduced from the overall levels. Particularly when noise data are provided as sound pressure levels, attention should be given to the test conditions under which those data were measured. Sound pressure levels measured in free-field or reverberant-field conditions are not generally representative
†
In using the relationships in this Item for estimating values of L w from measured values of L p , the SI system of units is assumed. With alternative unit systems these relationships have additional constant terms.
61
79037 of those achieved in practice since conditions in most buildings fall between those two extremes. Sound pressure levels must be converted to sound power levels before system noise calculations can be carried out. Noise data given in terms of sound pressure levels for in-duct or free-field noise tests are referenced to a particular distance from the sound source (usually three fan diameters or metres). For free-field conditions, sound pressure levels may be corrected to any required reference distance since for that condition sound pressure level is inversely proportional to the square of the distance to the noise source. Fan manufacturers generally quote either the fan inlet or fan outlet sound power levels. If total sound power level is quoted, it is the arithmetic average of the total inlet sound power level and the total outlet sound power level. In the absence of further data it is usual to assume that fan sound power is emitted equally from the fan inlet and outlet. When comparing the noise specifications for two fans, it is important to ensure that the data provided have a common basis and have been measured under similar conditions. For example, outlet noise data from one fan should not be compared with open-outlet data from another, since end reflections in the duct termination on the former fan will attenuate the sound radiated into open space. A means of conversion between inlet (or outlet) and open-inlet (or open-outlet) sound pressure levels, based on duct area, is provided in Figure 5 for the range of octave frequency bands of interest. This figure shows that the greater attenuation due to duct end reflections occurs at the lower frequencies. In using this Item to estimate fan sound power levels it should be noted that in the prediction procedure it is assumed that the fan is working at, or close to, its maximum efficiency. Fans are at their quietest when operating near peak efficiency and noisiest when working at, or near, their stalled condition. For a particular installation, it is therefore necessary to select the type of fan on the basis of the performance requirements, discussed elsewhere in this Item, before considering noise criteria. Because the prediction procedure provided considers the fan to be working at its maximum efficiency it is better, where possible, to obtain fan noise data from manufacturers' catalogues for the fan operating at the specific duties required in the particular installation. It should be noted that fan generated noise will increase if the airflow into the fan is unsteady. Since this increase may be considerable, to ensure its quietest operation it is essential that the air passing into the fan is uniform and turbulence free. The noise generated by the basic types of fan may be compared by considering the specific fan power levels, K w , given in Table 9.2. It is evident that centrifugal fans generate most of their noise at low frequencies while axial-flow fans generate high noise levels at high frequencies. Since the human ear is more sensitive to high frequency noise, axial fans will be noisier than centrifugal fans operating under similar flow conditions. However, it should be noted that it is easier to suppress high frequency noise than low frequency noise. It therefore follows that a high-speed axial-flow fan fitted with an appropriate attenuator may be less noisy, and cheaper to install, than a slower speed centrifugal fan giving the same performance. When selecting a fan on the basis of noise it is important to realise that, although the human response to noise is frequency dependent, it is not, in general, possible to distinguish between noise levels differing by less than about 3 dB. Also because of the inherent difficulties in both fan noise measurement and prediction, differences in noise levels of less than 2 dB are usually considered insignificant.
9.7
Estimation of Fan Noise The in-duct sound power levels of fans within ducts may be estimated using the procedure described here. For the case of propeller fans the predicted sound powers are the open-inlet (or open-outlet) sound power levels. It is implicit in this prediction procedure that the fan is working at, or near, its peak efficiency.
62
79037 The in-duct sound power levels include, at the lower frequencies, the sound energy within the duct system which does not emerge into the occupied space owing to reflections at the outlet. The reduction in sound power levels at the outlet (i.e. the difference between open-inlet (or open-outlet) sound power level and inlet (or outlet) sound power level) is plotted against duct cross-sectional area for each of the octave bands in Figure 5. The sound power level of a fan in each octave frequency band, in the range of centre frequencies from 63 Hz to 8000 Hz, is given by L w = K w + L w∗ .
(9.12)
The base sound power level, L w∗ , which is independent of frequency, is found from the fan operating condition. Values of Lw∗ are plotted against q for a range of values of ∆ pf , t in Figure 7. The specific fan sound power level, K w , accounts for spectral variations in the sound power level and is independent of the fan operating conditions. Values of K w for each octave frequency band are given in Table 9.2 for the various types of fan. It is necessary to apply a correction to the sound power level in the octave band containing the blade passing frequency, fB . The blade passing frequency is given by fan speed (rev/min) × number of blades f B = ------------------------------------------------------------------------------------------------ . 60
(9.13)
The octave band containing fB may be obtained from Figure 6. Having identified the octave frequency band containing fB , the sound power level in that frequency band is augmented by the blade frequency increment given in Table 9.2. For guidance in the initial system design, typical numbers of fan blades for the various fan types considered are given in Table 9.2. Overall sound power levels and octave band sound power levels predicted by the method in this Item have been compared with the equivalent in-duct sound power levels given in fan manufacturers’ catalogues for fans working at their peak efficiency. On the basis of that comparison, the majority of predicted overall sound power levels are within ± 7 dB of the catalogue levels. The scatter of compared data in the individual octave bands is, in most cases, greater than that for the overall level; for example, in the octave band centred at 250 Hz, the majority of predicted levels are within ± 9 dB of the catalogue levels. TABLE 9.1 A-weighting relative sound pressure level dB
B-weighting relative sound pressure level dB
C-weighting relative sound pressure level dB
63
– 26.2
– 9.3
– 0.8
125
– 16.1
– 4.2
– 0.2
250
– 8.6
– 1.3
0
Octave centre frequency Hz
63
79037 TABLE 9.1 – 3.2
– 0.3
0
1000
0
0
0
2000
+ 1.2
– 0.1
– 0.2
4000
+ 1.0
– 0.7
– 0.8
8000
– 1.1
– 2.9
– 3.0
500
TABLE 9.2 K w dB Fan type
Axial
Centrifugal
63
125
250
500
1000
2000
4000
Propeller
51
48
49
47
45
45
43
31
5–7
Tube axial
44
44
44
44
41
38
32
25
6–8
Guide-vane axial
43
42
42
43
41
38
34
26
6–8
Forwardcurved
43
41
36
32
28
24
20
15
2
32 – 64
Radial blade
46
47
47
43
39
32
25
19
2–5
5 – 10
Backwardbladed
39
36
34
32
28
24
18
12
3
8 – 16
46
43
43
38
37
32
28
25
4–6
8 – 15
Mixed flow Axial casing *
Octave band centre frequencies Hz
Blade* Typical frequency number increment of fan dB blades 8000 3–7
7 – 16
When a range is given for the blade frequency increment, the higher increments occur for fans having the lower number of blades and the lower increments for fans having the higher number of blades.
64
79037
10.
EXAMPLES
10.1
Fan Selection and Size and Power Estimation A fan is required to supply air to a small, fixed, ventilation system installed at sea level. The required flow rate is 8 m3/s for which a gauge static pressure of 500 Pa is needed.
Sketch 10.1 Gauge pressure, ps = 500 Pa required Select a suitable fan, obtain its approximate dimensions and estimate the power required. Within Section 7 on fan selection, Section 7.1 gives guidance on the mechanical suitability of fans for different duties. Since the conditions in this example are not mechanically arduous, the choice of fans is not restricted by such considerations. Section 7.2 explains the principles for sizing the various types of fans and Section 7.2.1 details the procedure as follows. . Step
Procedure
1.
The notes in Section 4 and in particular Section 4.2.1 state that forward-curved centrifugal fans are well suited to applications requiring a high volume flow rate at low to medium pressure rises. Thus, this type of fan appears appropriate especially as it draws in air directly from the atmosphere so that no special inlet connections have to be made.
2.
A middle-range value from Figure 1 of the parameter 1⁄ ds for forward-curved fans is 1.2. The corresponding value of ns = 2.7 .
3.
The notes on Figure 1 advise that a value for W T ⁄ DT of approximately 0.5 is typical for forwardcurved fans.
4.
The relative density, σ , has a value of unity since the fan is to be installed at sea-level.
5.–7.
Following the dashed line representing this example in Figure 1, a value for ∆ pf , s ⁄ σ of 500 and qmax = 8 m3/s locates point A on the upper right-hand graph. Note that the fan static pressure must be equal to 500 Pa for the reasons set out in Section 3.1.2 and, in particular, Sketch 3.3a.
6..
Point B is located on the Reference Line for centrifugal fans.
7.
Point C is located on the central inclined grid and the corresponding value for D T may be read as D T = 0.74 m.
8.
Point A is pinpointed on Figure 2.
65
79037
Step
Procedure
9.
Using n s = 2.7 as before, point B is located on the Reference Line for the rev/min speed scale.
10.
Point C is located on the central inclined grid and the corresponding impeller speed may be read as 606 rev/min.
11.
The fan will be belt driven and so will not be restricted to electric motor pole speeds.
As the impeller speed and diameter appear reasonable at this stage, the fan external dimensions will be estimated. Section 7.2.2 tables approximate dimensions of the fan casing. From Section 7.2.2 and using typical middle-range values, the estimated fan dimensions are as follows.
The overall height,
A,
is 1.95 × 0.74 = 1.4 m,
the overall depth,
B,
is 1.7 × 0.74 = 1.2 m,
and the casing width, C ,
is ( 1.2 × 0.5 + 0.5 )0.74 = 0.8 m.
Section 7.3 explains the principles for estimating the power requirements for the various types of fans and Section 7.3.1 details the procedure as follows.
Step
Procedure
1.
From Figure 3, for a value of n s = 2.7 , the fan static efficiency may be read from the far left-hand graph as 42 per cent.
2.
Point A is located on the far right-hand graph corresponding to ∆ pf , s = 500 Pa and q max = 8 m3/s.
3.
Point B is located on the Reference Line for centrifugal fans.
4.
Point C is located within the central inclined grid and the corresponding value for P R may be read as PR = 9500 W.
5.
A higher efficiency could have been obtained from the same type of fan (forward curved) if the specific speed were reduced. However, as an inspection of Figure 1 reveals, this would have resulted in a larger fan. The fan total pressure rise may also be estimated using the correlation given in Figure 4. For this example with n s = 2.7 , the value of ∆ pf , s ⁄ ∆ pf , t for a forward-curved fan is approximately 0.7. Thus the fan total pressure rise is given by ∆ pf , t ≈ 500 ⁄ 0.7 = 715 Pa.
10.2
Fan and System Noise Estimation
10.2.1 Fan noise estimation For the fan of the previous example running at the same conditions, estimate the in-duct sound power level in each of the eight octave frequency bands and the overall sound power level. The fan has 40 blades.
66
79037 Using Figure 7 with q = 8 m3/s and ∆ pf , t = 715 Pa. L w∗ = 51.5 . From Figure 6, for a fan with 40 blades rotating at a speed of 606 rev/min, the blade passing frequency occurs in the octave band centred at 500 Hz. For a centrifugal fan having forward-curved blades, the blade frequency increment, given in Table 9.2, is 2 dB. The sound power level for each octave band is found using the relationship L w = L w∗ + K w and applying the appropriate correction to the frequency band containing the blade passing frequency. Values of K w for the particular fan type are found from Table 9.2 and the summation for each octave band is tabulated.
Octave band centre frequency
Hz
63
125
250
500
1000
2000
4000
8000
Lw∗
dB
51.5
51.5
51.5
51.5
51.5
51.5
51.5
51.5
Blade frequency increment
dB
Kw
dB
43
41
36
32
28
24
20
15
In-duct band sound power level
dB
94.5
92.5
87.5
85.5
79.5
75.5
71.5
66.5
2
To obtain the overall sound power level, the eight octave band levels are combined by the method given in Appendix A or by using Equation (9.6). Using the method in Appendix A successive pairs of sound power levels are combined until a single value is obtained. The combination of levels is tabulated here and the detail of part of this combination is given in the example in Appendix A.
67
79037
The estimated overall sound power level in the duct is 97.5 dB. 10.2.2 System noise The fan of the previous example services a space through a length of uniform duct of cross-sectional area 0.02 m2 which terminates in the centre of a wall. Neglecting duct losses, other than those due to reflections at the duct termination, estimate the sound pressure level in dBA at a distance of 3 m from the duct outlet. Assume that free-field conditions exist in the serviced area and that the sound radiation pattern is non-directional. Firstly the sound power radiated into the space is estimated by correcting each octave band level for losses due to end reflections at the duct exit. These losses are obtained from Figure 5 for the duct exit area of 0.02 m2 . Since the duct exit is in the centre of a wall, and the radiation pattern is non-directional, hemispherical radiation can be assumed and the octave band sound pressure levels at a distance of 3 m (i.e. r = 3 ) may be found using Equation (9.10), L w = L pm + 20 log10 r + 8.0 , i.e.
L pm = L w – 20 log10 r – 8.0 .
(10.1)
The sound pressure levels in each octave band are then weighted using the A-weighting relative sound pressure levels given in Table 9.1. Finally the weighted band levels are summed, using the method in Appendix A or Equation (9.4), to obtain the single-value sound pressure level in dBA. The calculation steps are tabulated.
68
79037
Octave band frequency
centre
Hz
63
125
250
500
1000
2000
4000
8000
dB
94.5
92.5
87.5
85.5
79.5
75.5
71.5
66.5
dB
see below
14.7
9.2
4.6
1.2
0
0
0
Sound power level radiated into serviced area
dB
–
77.8
78.3
80.9
78.3
75.5
71.5
66.5
Sound pressure level at 3 m from exit. Equation (10.1)
dB
–
60.3
60.8
63.4
60.8
58.0
54.0
49.0
dB
–26.2
–16.1
–8.6
–3.2
0
+1.2
+1.0
–1.1
dB
–
44.2
52.2
60.2
60.8
59.2
55.0
47.9
In-duct sound power level from previous Example End correction from Figure 5
A-weighted relative sound pressure level from Table 9.1 A-weighted band sound pressure level
The end correction for the 63 Hz centred band falls outside the range of Figure 5. Since the correction in this frequency band will reduce the band sound pressure level well below the maximum band level, the contribution of the 63 Hz centred band to the dBA sound pressure level is negligible. The summed A-weighted sound pressure level is 65.6 dBA.
69
79037
11.
REFERENCES AND DERIVATION
11.1
References The references given are recommended sources of information supplementary to that in this Item (listed in chronological order).
11.2
1.
–
Classification of insulating materials for electrical machinery and apparatus on the basis of thermal stability in service. British Standards Institution, B.S.2757, 1956.
2.
–
Method of rating industrial noise affecting residential and industrial areas. British Standards Institution, B.S.4142, 1967.
3.
SHARLAND, I.
Woods practical guide to noise control. Woods of Colchester Ltd, 1972.
4.
–
Motor starters for voltages up to and including 1000 V a.c. and 1200 V d.c. Part 1 - Direct online (full voltage) a.c. starters. British Standards Institution, B.S.4941, Pt 1, 1973.
5.
–
Handbook of noise and vibration control. Edited by R.H. Warring. Trade and Technical Press Ltd, 1973.
6.
IQBAL, M.A. WILLSON, T.K. THOMAS, R.J.
The control of noise in ventilation systems - a designer's guide. E. and F.N. Spon, London, 1977.
7.
OSBORNE, W.C.
The selection and use of fans. Engineering Design Guides, No.33, Design Council, 1979.
8.
ESDU
Fluid mechanics, Internal Flow sub-series, Vols 1-4, Engineering Sciences Data Unit, 1980.
Derivation The derivations given are sources of information, excluding manufacturers' catalogue data, employed in the production of this Item (listed in chronological order).
9.
–
Motor starters and controllers. British Standards Institution, B.S.587. 1957
10.
–
Noise reduction. Edited by L.L. Beranek. McGraw-Hill, 1960.
11.
–
Woods practical guide to fan engineering. Edited by W.C. Osborne and C.G. Turner. 2nd Edition, Woods of Colchester, 1960.
12.
–
Methods of testing fans. Part 1 – Performance. British Standards Institution, B.S.848, Pt 1, 1963.
13.
HUGHES, E.
Electrical technology. 4th Edition, Longman, 1969.
14.
–
Fan engineering. Edited by R. Jorgensen. Seventh Edition, Buffalo Force Company, USA, 1970.
70
79037 15.
CSANADY, G.T.
Theory of turbomachines. McGraw-Hill Book Co., USA, 1964.
16.
–
Method of testing fans. Part 2 - Fan noise testing. British Standards Institution, B.S.848, Pt 2, 1966.
17.
–
Design for sound. Woods Fans, Toronto, April 1966.
18.
KENNY, R.J.
Fans and blowers. Machine Design, Vol.40, No.6, pp.l52-173, March 1968.
19.
BUSH, E.H.
Crossflow fan - history and recent developments. Fan Technology and Practice, Paper 4, pp.50-66, Inst. mech. Engrs, April 1972.
20.
GRAHAM, J.B.
How to estimate fan noise. Sound and Vibration, Vol.6, No.5, pp.24-27, May 1972.
21.
ECK, B.
Fans - design and operation of centrifugal, axial-flow and cross-flow fans. 1st English edition, Pergamon Press, 1973.
22.
FINKELSTEIN, W.
Preliminary computations of fan noise. Bldg Services Engr, Vol.41, pp.268-275, March 1974.
23.
SEARLE, D.G.
Pole amplitude modulated motors for fan and pump drives. Electrical Review, Vol.194, pp.407-409, 19th April 1974.
24.
–
Fan application guide. Heat. Vent. Air Conditioning Manufacturers Assn, 1975.
25.
–
Fan terminology. Eurovent 1/1, Paris, March 1975.
26.
–
Electrical motors and controls, Section 1 Motors. Machine Design, Vol.48, No.10, pp.8-42, 1976.
27.
ASTROM, L.
Fans for VAV systems. Flakt Review, Vol.13, No.2, pp.18-22, 1978.
28.
GORDON, C.G.
Fan noise and its prediction. Internoise 78, pp.154-165, San Francisco, May 1978.
29.
PAPAMARCOS, J. BAESEL, H.D. et al.
Choosing fans for power plants. Power Engineering, Vol.82, No.6, pp.46-60, June 1978.
30.
GERRARD, G.W. MESSER, M.G.M.
Designing with moving air. Original Equipment Manufacture Design, Parts 1 to 5, pp.24-26 (June), pp.36-41 (July), pp.37-41 (August), pp.50-55 (September), pp.53-59 (October), 1978.
31.
McFARLAND, J.
Fans for filtration. Filtration and Separation, Vol.16, No.2, pp.l44-149, March/April 1979.
32.
–
Specification for quantities, units and symbols. Part 7 - Acoustics. British Standards Institution, B.S.5775, Pt 7, 1979.
33.
BUSH, E.H.
Private Communication, Airwheel Ltd, Holton Heath, Dorset, August 1979.
34.
CAMPBELL, J.
Private Communication, Ove Arup Partnership, November 1979.
35.
WOODS-BALLARD, W. Private Communication, Woods of Colchester Ltd, December 1979.
71
11.3
79037 Manufacturers' Catalogue Data The following manufacturers, listed in alphabetical order, supplied catalogue and other data employed in the production of this Item.
C1.
AIRSCREW
Mixflo fans. Airscrew Howden Ltd, Weybridge, Surrey, UK.
C2.
AIRWHEEL
L, S and U flow fans. Leaflets on tangential fans. Airwheel Ltd, Holton Heath, Dorset, UK.
C3.
BLACKMAN
Series KB 24 centrifugal fans. Publication 65, 2nd Edition. Keith Blackman Ltd, London N17, UK.
C4.
DAVIDSONS
Standard fan catalogues on Sirocco axial fans and Sirocco centrifugal fans. Davidson and Co. Ltd, Belfast, Northern Ireland.
C5.
ENGART
Technical literature for HD and HV series of centrifugal fans. Engart Fans Ltd, Aberdare, UK.
C6.
FLAKT
FTBA axial-flow fans and HCL, HCM and HCH centrifugal fans. Flakt Ltd, Staines, Middlesex, UK.
C7.
HALIFAX
Paddle, backward inclined, backward curved and multivane fans. Halifax Fan Manufacturing Co. Ltd, Salterhebble, Halifax, UK.
C8.
ITT
Medium pressure tangential blowers. ITT, Thornton Industrial Estate, Milford Haven, Dyfed, UK.
C9.
JOY
Joy axivane fans. Catalog J-610, Joy Manufacturing Co., Ohio, USA.
C10. MATTHEWS & YATES
Centrifugal fans. Catalogue F/7, Matthews and Yates, Swinton, Manchester, UK.
C11. MYSON
Axial-flow fans and propeller fans. Myson Fans Ltd, Colchester, UK.
C12. SMITHS
Propeller fan PFD catalogue and leaflets on cross-flow fans. Smiths Industries Precision Fan Co., Witney, Oxon, UK.
C13. STANDARD POCHIN
AND Centrifugal fan catalogue. Standard and Pochin Ltd, Leicester, UK.
C14. TROX
Axial fan data sheets. Trox Bros Ltd, Thetford, Norfolk, UK.
C15. WOODS
Aerofoil axial fans: performance charts and data tables, Publication AFI, Woods of Colchester Ltd, UK.
72
Reference line for axial and mixed-flow fans
Reference line for centrifugal and cross-flow fans
For guidance on the use of this chart, see Section 7.2.1
2
va pa
mf
!
0.8
1 / ds
ta pa
1.0 0.6
va ta
0.4
∆p f,s /σ (Pa)
!
cr mf ch ba, bc, bi
!
fc
xf pb
1
500 200 100 50
!
2
0.8 0.6
1 / ds
ba, bc, bi
!
xf
0.4
!
1000 2000 5000 10000
pb 0.2
73 0.6 0.8
1
2
4
ns
6
8
10
20
40
10
2
-2
4
6
10
-1
2
4
6
1
2
D T (m)
increasing speed
increasing dimeter
4
6
10 10 8
Legend
Section
6
pa propeller fan ta tube-axial cr contrarotating va guide-vane axial
4.1.1 4.1.2 4.1.3 4.1.4
4
fc forward-curved centrifugal pb radial-discharge, paddle bladed ch radial-discharge, curved heel ba backward-bladed, aerofoil bc backward-bladed, curved bi backward-bladed, straight
4.2.1 4.2.2 4.2.2 4.2.3 4.2.3 4.2.3
mf mixed-flow, axial casing
4.3.1
xf cross-flow, J,S and U types
4.4.1
!
2
WT / DT Note:
1 0.8 0.6 0.4
!
0.2
0.1 -3 10
2
4
6
10
-2
2
4
6
10
-1
2
4
6
2
1
6
4
3
q max (m /s)
FIGURE 1 ESTIMATION OF IMPELLER TIP DIAMETER, DT
10
2
4
6
10
2
2
4
6
10
3
2
4
6
79037
For all axial and mixed-flow fans, take W T / D T = 1 For single-inlet centrifugal fans, typical ranges of W T / D T are as follows: forward-curved 0.40 < W T / D T < 0.65 radial discharge 0.40 < W T / D T < 0.55 backward bladed 0.12 < W T / D T < 0.30 For double-inlet, double the above values. For cross-flow fans, 0.70 < W T / D T < 10
!
peak efficiency fans available off-peak efficiency but in normal operating range
20
Reference line for rev/min impeller speed
Reference line for rad/s impeller speed
For guidance on the use of this chart, see Section 7.2.1
∆p f,s /σ (Pa)
20 10 50
!
B
20 10
10 00 50 0 00 00
00
0
0
0
50
74
!
A
!
C
! !
0.6 0.6
1
4
2
ns
6
8
10
(Note value of n s must be the same as that used in Figures 1,3 and 4 )
20
2
4
6
10 10 2 10 3 Impeller speed rad/s or rev/min 40
2
4
6
2
6
4
10
10 4
2
8 6 4
Note:
!
2
WT / D T
1
!
0.4
forward-curved 0.40 < W T / D T < 0.65 radial discharge 0.40 < W T / D T < 0.55 backward bladed 0.12 < W T / D T < 0.30 For double-inlet, double the above values. For cross-flow fans, 0.70 < W T / D T < 10
0.2
0.1
10 -3
2
4
6
10 -2
2
4
6
10 -1
2
4
6 8
1
2
4
6 8
3
q max (m /s)
FIGURE 2 ESTIMATION OF IMPELLER SPEED, N
10
2
4
6
10 2
2
4
6
10 3
2
4
6
10 4
79037
0.8 0.6
!
For all axial and mixed-flow fans, take W T / D T =1 For single-inlet centrifugal fans, typical ranges of W T / D T are as follows :
pa ta cr va
propeller fan tube - axial contra - rotating guide - vane axial
4.1.1 4.1.2 4.1.3 4.1.4
fc pb ch ba bc bi
forward - curved centrifugal radial - discharge, paddle bladed radial - discharge, curved heel backward - bladed, aerofoil backward - bladed, curved backward - bladed, straight
4.2.1 4.2.2 4.2.2 4.2.3 4.2.3 4.2.3 4.3.1
xf
4.4.1
cross - flow, J,S and U types
Reference line for axial and mixed - flow fans
For guidance on the use of this chart, see Section 7.2.1
∆pf,s (Pa) 10 0 50 0 0 0 0 20 0 0 10 0 0 50 0 20 0 10 0 50
mf mixed - flow, axial casing
Reference line for centrifugal and cross - flow fans
Section
Legend
Peak efficiency fans available Off - peak efficiency, but in normal operating range 100 cr
ηs 40 per cent
va ta
C
pa mf
75
20
mf
va cr
A
!
80 60
pa ta
!
100 80 60
ba pb
B
!
fc xf
20 pb
bc bi ch
!
!
ηs per cent
40
!
1
2
4 6
10
20
40 10 2
4 6
2 10 2
PR
(watts)
ns (Note value of ns must be the same as that used in Figures 1, 2 and 4)
4 6 103 2
4 6 104 2
4 6 105 2
4 6106 2
10-3 2 4 6 10-2 2 4 6 10-1 2 4 6 1 2 4 6 10 2 4 6 102 2 4 6 103
FIGURE 3 ESTIMATION OF IMPELLER POWER, PR
qmax (m3/s)
79037
0.5
1.0
cr xf
pb
ba,bc,bi mf
Legend
fc
ta
ch va
0.9
pa
0.8
0.7 ∆p f,s / ∆p f,t
Section
pa ta cr va
propeller fan tube - axial contrarotating guide-vane axial
4.1.1 4.1.2 4.1.3 4.1.4
fc pb ch ba bc bi
forward-curved centrifugal radial-discharge, paddle bladed radial-discharge, curved heel backward-bladed, aerofoil backward-bladed, curved backward-bladed, straight
4.2.1 4.2.2 4.2.2 4.2.3 4.2.3 4.2.3
m f mixed-flow, axial casing
4.3.1
xf cross-flow, J,S and U types
4.4.1
peak efficiency fans available fc
0.6
off-peak efficiency but in normal operating range
76 0.5
mf
ba, bc, bi
va cr
0.4
pa
xf pb
ch
0.3 ta
0.5
0.6 0.7 0.8 0.9
1
1.5
2
2.5
3
4
5
6
7
8
9
10
ns
∆p f, s FIGURE 4 CORRELATION OF -------------∆p f, t
15
20
25
30
40
50
( N ote value of n s must be the same as that used in Figures 1,2 and 3 )
79037
0.2
79037
20
Inlet (or outlet) sound power level open inlet (or open outlet) sound power level. dB
18 16 14 Octave band centre frequency Hz
12
63
10 125
8 250
6 500 4 1000 2
2000 4000
0 2
3
4
6
8 10-2
2
3
4
6
8 10-1
2 2
Duct area m
FIGURE 5
77
3
4
6
8
100
2
3
4 5
79037
Octave bands and centre freq. Hz
Number of blades 6050 40
2
10
3
2000
30
1000
20 15
8
10
6
fB Hz
500
7
250
5 4
4 3
3
2
2
125 10
2 8 6
63 6
8
102
2
3
4
6
Fan speed rev/min
FIGURE 6
78
8
103
2
3
4
100 4
10 8 6
90
4
∆pf,t Pa 2
80
3
10 8 6
4
70
2 2
10
60 L w* dB
50
79 40
30
20
10
2
3
4
6
8 10-1
2
3
4
6
8
100
2
3
4
6
8
101 q m3/s
FIGURE 7
2
3
4
6
8
102
2
3
4
6
8
103
2
3
4
6
8
104
79037
0 10-2
79037
APPENDIX A COMBINATION OF LEVELS IN dB*
A1.
A2.
NOTATION
L1
higher level
dB
L2
lower level
dB
L
level due to combination of L 1 and L 2 , L 1 + ∆L 1
dB
∆L 1
increase in level above L1 due to L2
dB
NOTES A method of determining the combined sound level resulting from two or more sources of known levels is presented in this Appendix. The method is applicable to sound power levels and pressure levels or any other parameter expressed in dB provided that all levels to be combined are with respect to a common reference value. In the Figure A1, ∆L 1 is plotted against L1 – L2 . Clearly the combined level from any number of sources may be obtained by repeated combination of pairs of levels. The method may be used to obtain the overall level for a single source by combination of the levels in particular frequency bands. The method applies strictly to statistically independent sources, however. Figure A1 shows that ∆L1 is negligible when L 1 – L 2 is large, rising to a maximum of 3 dB when the levels are the same.
A3.
DERIVATION ( L2 – L1 ) ∆L 1 = 10 log10 1 + antilog 10 ------------------------- . 10 1.
A4.
–
Handbook of noise control. Edited C.M. Harris. McGraw-Hill, New York, 1957.
EXAMPLE It is required to determine the overall sound power level resulting from three sources with individual levels of 94.5, 92.5 and 89.6 dB with respect to a common reference power. To combine the first pair of levels, choose L 1 = 94.5 dB, so
*
L 2 = 92.5 dB,
L 1 – L 2 = 2.0 dB.
This Appendix is also available as Item No. 66017 in the Noise Sub-series and in the Acoustic Fatigue Sub-series.
80
79037 From Figure A1 Therefore
∆L 1 = 2.1 dB. L = 94.5 + 2.1 = 96.6 dB.
Now combining this with the remaining original level, choose L 1 = 96.6 dB, so From Figure A1 Therefore
L 2 = 89.6 dB,
L 1 – L 2 = 7.0 dB. ∆L 1 = 0.8 dB. L = 96.6 + 0.8 = 97.4 dB,
which is the overall level for the original three sources.
81
3.0
2.5
2.0
1.5
82 ∆ L1 dB 1.0
0.5
0
2
4
6
8
10
12
L1 - L2 dB
FIGURE A1 COMBINATION OF LEVELS IN dB
14
16
18
20
79037
0.0
79037
THE PREPARATION OF THIS DATA ITEM The work of the permanent professional staff of the Engineering Sciences Data Unit on this particular Item was monitored and guided by the following Working Party. Mr A.R. Green – Prof. R.I. Lewis – Mr J. McFarland – Mr E.J. Perry – Dr D. Pollard – Mr W.R. Woods-Ballard–
Trox Brothers Ltd Newcastle University Matthew & Yates Ltd Atkins Research and Development GEC Power Engineering Woods of Colchester Ltd.
on behalf of the Internal Flow Panel which has the following constitution: Chairman Mr N.G. Worley
– Babcock Power Ltd
Members Mr J. Campbell Dr D. Chisholm Dr D.J. Cockrell Dr R.B. Dean Mr D.H. Freeston* Dr G. Hobson Prof. J.L. Livesey Mr D.S. Miller Mr B. Payne Dr D. Pollard Mr J.A. Ward
– – – – – – – – – – –
Ove Arup Partnership National Engineering Laboratory Leicester University Atkins Research and Development Auckland University, New Zealand GEC Turbine Generators Ltd. Rugby Salford University British Hydromechanics Research Association Kellogg International Corporation GEC Power Engineering Ltd. Whetstone Atomic Energy Technology Unit.
The Internal Flow Panel has benefited from the participation of members from several engineering disciplines. In particular, Dr G. Hobson has been appointed to represent the interests of mechanical engineering as the nominee of the Institution of Mechanical Engineers and Mr B. Payne has been appointed to represent the interests of chemical engineering as the nominee of the Institution of Chemical Engineers. The work on this Item was carried out in the Internal Flow and Physical Properties Group of the Engineering Sciences Data Unit. The members of staff who undertook the technical work involved in the initial assessment of the available information and the construction and subsequent development of the Item were Mr C.J.T. Clarke Mr R.F. Lambert Mr M. Bolton
*
– Group Head, Internal Flow and Physical Properties – Group Head, Noise and Structural Dynamics – Senior Engineer, Internal Flow and Physical Properties Group.
Corresponding Member
83