R. A. Wallis (Auth.) -Axial Flow Fans. Design and Practice-Elsevier Inc (1961)

R. A. Wallis (Auth.) -Axial Flow Fans. Design and Practice-Elsevier Inc (1961)

(Parte 1 de 5)


R. A. WALLIS M.E., A.S.T.C, A.I.M.E.Aust., A.F.R.Ae.S.

Sen/or Scientific Officer

Aeronautical Research Laboratories (Melb.)

Department of Supply Australia

© R. A. WALLIS, 1961

U.S. Edition published by


First published 1961

FOREWORD by PROFESSOR A. R. COLLAR, University of Bristol

The design of axial flow fans is a branch of engineering science which has been the subject of a large amount of ad hoc study, but in view of the very wide use of the fan, surprisingly little really systematic scientific attention has been devoted to it. This has probably been due in part to the fact that pro pellers and fans are usually quite remarkably efficient, even when the design is rudimentary, so that refinement of design has been considered not worth the effort involved. The ab sence of systematic work has, however, resulted in the absence of a satisfactory handbook, and this in turn has had a feedback effect on design philosophy. It is true that a good many scientific papers have derived from the ad hoc studies mentioned above, but these have been related to special topics and have not attempted to view the subject as a whole. In the early days of my own interest in this field there were two reference volumes most commonly employed: Design of Screw Propellers for Aircraft by H. C. Watts, and Aerofoil and Airscrew Theory by H. Glauert. The former contained many experimental data ; the latter was almost wholly theoretical. "Glauert", in fact, was the aerodynamicist's vade-mecum, and was rightly regarded as indispensable, despite its foundation in inviscid theory and its consequent absence of interest in the effects of Reynolds number. These two volumes could provide basic data and method for a problem, but thereafter one was on one's own, especially in relation to ancillary problems involving ducting. It was a recognition of the paucity of published information that prompted me, two decades ago, to record my own experience relating to the powering of wind tunnels in some half dozen papers in the R. & M. Series, on fan, straightener, and cascade design and the like, papers which, I think, explain (if not justify) the invitation to me to write this Foreword, which I am privi leged to do.


In view of the situation outlined above, I am very glad that

Mr. Wallis has recognized the need for an authoritative book on fans ; moreover, I think his book will quickly become the stan dard work in this field. Many will note with approval the compendium of boundary layer theory contained in Section 3, which indicates plainly the author's view that this is a subject of major importance for scientific fan design ; the insistence on assessment of the characteristics of ducts is also noteworthy. This is to mention only two of the preliminary subjects studied before the book embarks on careful and systematic investigations of fan design, with detailed illustrations.

Of the author, it need only be said that Mr. Wallis had already established a high reputation in Australia before coming to work for a time in England. In this country he quickly won golden opinions for his aerodynamic work, which was both soundly based and original. I am sure that the present volume will add considerably to his reputation.

A. R. COLLAR Bristol 1960.

The design of a ducted axial flow fan using sheet metal blades was one of my first tasks as an aerodynamicist. I used the simple design method developed at the Aeronautical Research Laboratories by Dr. G. N. Patterson, but experienced difficulty in obtaining reliable information on the aerodynamic character istics of cambered plate fans. In consequence, a limited research programme on the subject was undertaken. These were the origins of a personal interest in axial flow fans.

Dr. Patterson's pioneer work at A.R.L. on fan design has become widely known, and as a result many inquiries for further information have been received by the Laboratories. At the same time, the need for a modern comprehensive treatment of the aerodynamics of ducted axial flow fans has long been recognized, and the present book seeks to fulfil that need.

In view of the fact that the main potential user of the data contained in this work is probably the ventilating engineer, much thought has been given to the question of suitable presentation. Contacts with members of the ventilating industry and study of professional journals in the field suggest that the industry has perhaps been a little slow to assimilate the great advances which have been made in general aerodynamics. Since ventilation engineers will be compelled to make everincreasing use of these advances, it has appeared logical to present the material of this book in a modern aerodynamic manner. No apology is offered for the discarding of the tradi tional hydraulic notation, since I consider this step to be both justified and rendered inevitable by the trend of development. My views in the matter have been reinforced by the ability to deal with a hitherto unfamiliar design concept shown by a group of practising engineers who attended a series of lectures which I delivered on the subject. vii

No attempt has been made to present the historical and classical development of the momentum theory as applied to fan design. Instead, emphasis has been placed on the deriva tion of the required equations from well-established basic principles and design assumptions. One of the major aims has been the development of a simply-followed design method, supported by a sufficiency of background data which can be applied in making a rational appraisal of the design. The work should also make it possible to handle any problem of design, testing or analysis by means of either a direct application or a suitable rearrangement of the basic data presented.

Tribute to the author's wife for aid in preparing the manu script is a familiar feature of prefaces to technical books. The present work is no exception, and I should like to acknowledge my wife's assistance, particularly in the typing of the manu script. Mr. L. P. Coombes, the Chief Superintendent and Mr. F. W. David, the Aerodynamics Superintendent at A.R.L. have been most helpful during the preparation of the work. I am indebted in particular to Mr. W. Howard, and also to Mrs. B. L. Cumming, for pointing out errors and suggesting improve ments. The great assistance given by Mr. N. Ruglen during the proof-reading stage is acknowledged with gratitude.

I am most grateful to Mr. V. J. Smith, who has kindly assisted me in the preparation of the Appendix dealing with free fans.

This book contains material used in a thesis submitted to the

University of New South Wales, and I should like to record my appreciation of the readiness with which the University's permission for its publication was given.

The assistance gained from various publications in the pre paration of certain of the illustrations is much appreciated ; acknowledgments of the sources of the figures concerned will be found in the text.

Finally, the permission given by the Department of Supply for the publication of the book and for the use of photographic material in the possession of the Department is gratefully acknowledged. R. A. W. Melbourne

Between pages 182 and 183

Plate l.(a) Fan unit of wind tunnel at A.R.L. (Melb.) (b) Industrial air-moving fan unit

2.(a) Wind tunnel fan unit employing cambered sheet metal blades

(b) Sheet metal blade rotor

3. Construction of 12-bladed fan from wooden laminations

4.(a) Straighteners and tail fairing of A.R.L. (Melb.) wind tunnel fan

(b) Straighteners and drive fairings of industrial fan unit

LI. General

Propeller, or axial flow, type fans have been in use for many decades but it is only during the last two or three that much attention has been paid to their technical refinement. For the first part of this century, centrifugal fans of both the impeller and multi-vaned types enjoyed popularity owing to their greater operating flexibility, efficiency and quietness. The axial flow fan is however now rapidly gaining ground, as is indicated, for example, by the present preference for such fans in large mine ventilating installations. In the compressor field, too, the axial flow machine has virtually superseded the centrifugal compressor in gas turbine engines. This progress is due to the experimental and theoretical studies which have ensured the attainment of very high efficiencies together with satisfactory operation of the fan within the desired range. Marked reductions in noise level have also been achieved.

Important factors affecting the choice of fan unit are (a) manufacturing simplicity, (b) cost of manufacture, (c) fan efficiency, and (d) flow reversal in an emergency. The two former often favour the centrifugal while the two latter favour the axial flow type.

The uses to which fans are put are self-evident and hence will not be outlined here. It is important, however, in choosing or designing a fan to have a precise knowledge of the requirements, as otherwise the fan installed may be completely unsuited for the task. To avoid this predicament, which unfortunately occurs far too often, a basic approach to fan and duct problems is essential. Rule-of-thumb methods cannot take into account all the eventualities likely to be met.


1.2. General Types of Axial Flow Fan Axial flow fans can be placed in three main categories :

(i) Free fan. A free fan is one which rotates in a common unrestricted air space (Fig. 1.1). In special cases a shroud ring attached to the blade tips may be employed.

Fig. 1.1. Air circulator, free fan

(i) Diaphragm mounted fan. This type of fan transfers air from one relatively large air space to another (Fig. 1.2). A shroud ring may be attached to the blades or, alternatively, to the partitioning structure.

(i) Ducted fan. A fan is ducted (Plate 1 (a) and (b)) when the enclosing duct constrains the air to enter and leave the blading of the fan unit in an axial direction. The minimum duct length to satisfy this condition will be in excess of the distance between inlet to and outlet from the blading. When addi tional fan stages are fitted in series, the pressure rise of which the unit is capable increases. In the extreme case, a multi stage unit becomes a compressor.

The first two types are relatively long-established and are probably the most common fans in use. The task which they perform is, generally speaking, not one for which centrifugal

GENERAL TYPES OF AXIAL FLOW FAN 3 fans are suitable. Although improvements have been made in the design of these unducted fans, there are still many of a relatively crude type in existence.

r j

v 1 ■ o

1 4v 1Λ

Ψ ¥ A Fig. 1.2. Diaphragm mounted exhaust fan

One of the notable improvements of recent years has been the design of exhaust fans which operate in short compact ducts. These fans are superseding the cheap but crude type of dia phragm mounted fan illustrated in Fig. 1.2. It might be concluded, therefore, that the second type of fan listed above

4 INTRODUCTION represents a passing phase in the development of quiet, high efficiency exhaust fans.

The ducted fan has undergone a rapid evolution and is now the most refined of the axial flow types. This is due in no small measure to the opportunity which the designer usually has to provide favourable aerodynamic working conditions for the fan. The present work will be concerned almost exclusively with this type of fan ; some useful data on the other distinct type of fan, namely the free fan, are presented in an appendix.

1.3. Elements of a Ducted Fan Unit

The various components which go to make up a ducted fan unit are indicated in Fig. 1.3. Rotor blades are a series of aerofoils which, owing to their relative motion with the air, add total head to the air stream. It is desirable that this function should be discharged with minimum losses through friction, flow separation and secondary flows.



C" 1 L 1 Π


PREROTATOR-' - STRAIGHTENER Fig. 1.3. Components of ducted fan unit

Stationary vanes, known as stators, are often located upstream and/or downstream of the rotor for reasons which will be elaborated on in Section 1.6.

In well-designed units the rotor boss diameter will normally be within the range 40-70 per cent of the rotor diameter. Near the axis of the rotor, both the blade velocity and swept areas are very small; as a result, the potential work output here is

GENERAL METHODS OF MANUFACTURE 5 very small. The larger boss diameters are usually associated with fans designed for high pressure rises.

Suitably shaped fairings upstream and downstream of the rotor boss are an essential part of good axial flow fan design.

In a multi-stage unit of co-rotating rotors there is a row of stators between each rotor stage.

1.4. General Methods of Manufacture

The method of fan construction often has a marked influence on the aerodynamic design and final blade geometry. A very common type of fan employs sheet metal blades which normally possess some form of local or general cross-sectional camber. The design of such fans is sometimes crude and of an ad hoc nature and in consequence they operate with a relatively low efficiency. When design is based on modern aerodynamic principles, however, efficiencies can be high, as demonstrated in later sections. A modern sheet metal bladed wind-tunnel fan is illustrated in Plate 2(a). The blade twist and camber of a similar fan can be observed in Plate 2(b).

High efficiency fans are often made of wood since the material can be readily and accurately worked to the desired shape. The blades and boss are built up in an integral manner out of wooden laminations which are subsequently glued together (see Plate 3) ; the difficulty of this method increases with the number of blades required. When the laminations are carefully designed, considerable strength can be built into the boss for the purpose of resisting hoop and tensile stresses. Metal reinforcing plates or rings on the upstream and down stream faces of the bosses are, however, usually essential. This method is normally used only for fans in limited production.

The casting method of manufacture is a very common one.

Early fans consisted of a few large blades cast integral with the boss, and bore traces of the influence of marine screw design. These fans were relatively crude by present-day standards. The trend is now towards separate blades produced by a modern casting process which provides a smooth surface finish without resort to expensive machining. High efficiency, high pressure rise fans possessing a large number of blades are often produced by this method.


Sufficient has been said to show that the type of construction adopted can exert a definite influence on the aerodynamic design of the fan.

1.5. Ducted Fan Duty

The duty of a ducted fan is specified by the amount of air which it forces through the system against resistance comprising skin friction, flow separation, secondary flow and discharge loss.

In many commercial duct systems, the fan exhausts directly to the atmosphere. In such cases the customer is mainly concerned with the pressure losses upstream of the fan and as a consequence manufacturers specify their fans in terms of their ability to overcome a particular upstream loss. The fan static pressure head, as this loss is called, is related to the useful total head rise across the fan by the relation

(Parte 1 de 5)