Skip to main content

Full text of "The Aeronautical Journal 1920-09: Vol 24 Iss 117"

See other formats



(Founpgp 1897 in succession to the ANNUAL Reports). 

Published Monthly at the Offices of the Society, at 7, Albemarle Street, Piccadilly, London, W.1. 
Telephone: “ Gerrard 7373. Telegraphic Address : “ Didaskalos, Piccy, London.” 

Subscriptions per annum, £2 5s.; single numbers, 3s. ad., post free. 
Edited for the Council by J. LAURENCE PRITCHARD, Associate Fellow. 

All communications should be addressed to the Eelitor. 

SEPTEMBER, 1920. Vob, XXIV. 

Notices of the Royal Aeronautical Society. 

Election of Members. 

To prevent disappointment, members are requested to inform applicants 
tor membership whom they may be introducing that the next election will not 
take place until October roth, as there will be no meeting of the Council before 
that date. In aceordance with the Rules, subscriptions of new members elected 
at that meeting will cover the period up to December 31st, 1921. 

Annual Reports. 

In last month’s notices it was stated in error that the first volume of the 
‘ Aeronautical Classics *’ contained a reprint of KF. HH. Wenham’'s Paper 
* Aerial Locometion,’’ whereas this Paper, in fact, forms the second volume ot 
that series. The first volume, which is now out of print, contained the second 
reprint of ** Aerial Navigation,’’ by Sir George Cavley, which was first reprinted 
in the Annual Reports of the Aeronautical Society for 1876, having originally 
appeared in ** Nicholson’s Journal’ during 1809 and 1810. 

The Secretary is anxious to obtain copies of the Twelfth (1877) and Eighteenth 
and Nineteenth (1883-1884, issued together) Annual Reports. 

It is desired to remind the member who borrowed the third volume of the 
Reports (1880-1893) from the Library, that he has omitted to return this up to the 
present. : 

Machine Tool and Engineering Exhibition. 

There are still a few tickets available admitting at half-price to the Machine 
Tool and Engineering Exhibition, being held at Olympia from September 7th to 
September 25th. The Secretary will be glad to forward these to any members 
desiring to avail themselves of the special facilities offered through the courtesy 
of the Machine Tool Trades Association. 


Although a number of the outstanding subscriptions have been received, 
there is. still a considerable amount owing. It has been found impossible to 
continue sending the Journal in such cases, as this involves the Society in heavy 
losses where the subscriptions are not paid. Members are reminded that in all 
cases, whatever the date of election, subscriptions become due on January st. 
Unless subscriptions are paid within a reasonable period of that date, it is 

{78 THE AERONAUTICAL JOURNAL [September, 1920 

exceedingly difficult for the Secretary to form an accurate estimate of the financial 
position on which to base a programme of activities without fear of involving the 
Society in a deficit. 

Annual Dinner. 

In view of the difficulty of making arrangements for a dinner until an 
estimate can be formed of the number of persons likely to attend, those members 
who have not already done so will greatly assist by writing to the Secretary to 
say whether or not they would propose to be present at a dinner towards the end, 
probably in the last week, of October. An estimate of the price cannot be made 
until such rough information is obtained, but every endeavour will be made to 
keep it as low as possible. [t is thought that such a dinner would provide an 
opportunity for exchange of ideas which is impossible during the short time 
tvailable before and after lectures. 

Autumn Session. 

The dates of the following lectures have been fixed during the Autumn and 
Spring Session. They will take place, in the absence of other notification, at 
5-30 pem. at the Roval Society of Arts, John Street, Adelzhi, W.C. Only an 
tbstract of each Paper will be read at the meeting, leaving more time available 
for discussion. Where possible, a brief svnopsis will be published in the Journal 
hefore each Paper, a full report being published afterwards :— 

October 7th.—Major-General Sir F. H. Svkes, G.B.E., ‘* Civil Aviation.”’ 

October 21st.—Squadron-Leader R. M. Hill, R.A.F., ‘* A Comparison 
of the Flying Qualities of Single and Twin-Engined Aeroplanes.”’ 

November 4th.—H. Ricardo, ‘* Possible Developments in Aircraft 

December 16th.—Wing-Commander Flack, ‘* The Human Machine in 
Relation to Flying.” 

January 20th.—-Lord Montagu of Beaulieu, ‘* The Cost of Air-ton Miles 
Compared with the forms of Transport.”’ 

February 17th.—F. Handley Page, C.B.E., ‘‘ The Handley Page Wing.”’ 

March 3rd.—]. W. W. Dyer, ** Airship Fabric.”’ 

March 17th.—Captain D. Nicolson, ‘* Flying Boat Construction.’’ 

W. Lockwoop Marsu, Secretary. 

September, 1920] THE AERONAUTICAL JOURNAL 479 


The Fourteenth Meeting of the Fifty-fifth Session was held in the Hall of 
the Royal Society of Arts, London, Major-General E. L. Ellington, C.B., 
C.M.G., occupying the Chair. 

The CHAIRMAN said Sir Richard Glazebrook required no introduction from 
him. Those present would know him very well as Director of the National 
Physical Laboratory, and until recently as Chairman of the Advisory Committee 
for Aeronautics, of which the late Lord Rayleigh was President. Sir Richard 
Glazebrook had recently been appointed Zaharoff Professor at the Imperial 
College of Science, and had also been appointed Chairman of the Aeronautical 
Research Committee, this taking the place of the Advisory Committee for 
Aeronautics. He was going to tell them something about the work of the late 
Advisory Committee, and the legacy that Committee had left to the new Aero- 
nautical Research Committee, and also about the arrangements that had been 
made in connection with the Zaharoff Professorship for higher education in 
aeronautical science. He thought it would be agreed that there was no one 
more qualified to speak on these subjects than Sir Richard Glazebrook, and he 
would ask him to begin his Lecture. 

Sir R. T. GLazeBrook then delivered the following Lecture :— 


Let me commence with an explanation and an apology. About a year ago: 
we were discussing the Annual Report of the Advisory Committee for Aeronautics 
for 1918-19, and I suggested that in that report was ample material for an 
interesting Paper dealing in great measure with advances during the war; 
material which up to then had been strictly confidential in character. Some few 
months ago an invitation came to me from your Council to read such a Paper, 
and | hastily accepted. 

Six weeks ago | sat down to put my ideas in form and wrote some few pages 
when it suddenly occurred to me that Prof. Bairstow, in his Wilbur Wright lecture 
last year, had dealt with the subject in a most able and complete way and with 
very full knowledge, and that I had but little to add to his admirable exposition. 
Let me repeat what I said in moving a vote of thanks to Mr. Bairstow on that 

‘* At various institutions, such as the Royal Aircraft Establishment and the 
National Physical Laboratory, a band of able workers—to whom great credit is 
cue—have shown how, by means; of experiment, to verify the results of theory to 
a high order of accuracy. Mr. Bairstow had laid stress on the importance of 
work such as he had been describing if in the future aeroplanes were to be 
used with safety by the public and with the assurance that they would comply 
with the conditions laid down to secure that safety. -I could not speak with 
sufficient emphasis of the importance of research and of the necessity that those 
who were responsible for providing the funds, without which research could not 
be prosecuted, should realise the urgency of the position and take steps in the 
immediate future for the continuance of the work to which the position this country 

180 THE AERONAUTICAL JOURNAL [September, 1920 

now occupied was due. An enormous debt was owed by the practical side of 
aeronautics to science. I trust that that debt will largely increase in the future, 
for it was due to the great work of Mr. Bairstow and his fellow investigators 
that this country had arrived at its present foremost position both in theory 
and practice.”’ 

Had I remembered the occasion on which | spoke these words | should 
have explained the position to our Secretary and relieved myself from the 
responsibility of addressing you to-night. 

And there are other reasons for some variation from my original theme. 
The Advisory Committee for Aeronautics has ceased to exist. It has been replaced 
by the new Aeronautical Research Committee. I propose, therefore, with your 
permission, to deal briefly with the history and work of the A.C.A. and then 
devote attention to some of the many problems which still await solution and to 
the task—the heavy task—of the new Committee. 

The Advisory Committee was established in 1909 by Mr. Asquith, then Prime 
Minister. Its inception was due to Mr. Haldane, Minister for War. Aviation 
was in its infancy; the Wrights had flown. Mr. Haldane saw clearly that there 
were many problems which only experiment could solve and which might be solved 
by direct experiment, conducted under proper conditions, much more satisfactorily 
than by attempts at flights in the air on machines built with but little knowledge 
of the aerodynamic forces they would meet or of the stresses which would occur 
in their structure. He appealed to Lord Rayleigh and myself to know if we could 
help at the National Physical Laboratory. A scheme for work was suggested, 
and at a meeting at the Admiralty at which Mr. McKenna, then First Lord, and 
Mr. Haldane were present, the details were agreed upon, and the Prime Minister 
on May 5th, 1909, after explaining to the House of Commons how the work 
was to be shared between the War Office and the Admiralty, made the following 
announcement :— 

‘* With a view to securing that the higher scientific talent shall be brought 
to bear on the problems which: have to be solved in the course of the work of the 
two Departments, the National Physical Laboratory has been requested to 
organise at its establishment at Teddington a special department for continuous 
investigation—experimental and otherwise—of questions which must from time 
to time be solved in order to obtain adequate guidance in construction. 

‘** For the superintendence of the investigations at the N.P.L. and for general 
advice on the scientific problems arising in connection with the work of the 
Admiralty and War Office in aerial construction and navigation I have appointed 
a special Committee,’’ etc. 

We knew but little then. I trust 10 years of strenuous work have added 
considerably to that little, and while there is still much to be achieved, have 
helped in no small degree towards the supremacy in the air won by our British 

It is perhaps of interest to note what were the questions attacked according 
to the first programme. We were to include the determination of the vertical 
and horizontal components of the forces on inclined planes in a horizontal current 
of air, especially for small angles of inclination to the current; the determination 
of surface friction on plates in a current; the position of the centre of pressure 
for inclined planes; the distribution of pressure over the plane; with similar 
investigations for curved surfaces of various forms; all fundamental questions 
about which our knowledge was small, but the answers to which now form part 
of the elementary instruction given to every student. And when dealing with 
aircraft rather than with aerodynamic theory the Committee attacked equally 
fundamental problems; the resistance of the components of an aeroplane, struts, 

September, 1926] THE AERONAUTICAL JOURNAL 48] 

control wires and the like, the problem of stability, the materials used in con- 
struction, the theory of the propeller or airscrew, the efficiency and design of 
motors, and for airships the! study of the shape of least resistance, their stability, 
and the effect of the stabilising planes, the production of hydrogen, the properties 
of fabrics used to contain the gas, their navigation and methods of mooring ; 
all these matters appear in the first programme of work. From the com- 
mencement the importance of meteorology was realised, and Sir Napier Shaw’s 
report on wind structure, dated 3rd June, 1909, was the first of a number of 
Papers which now form a most valuable collection of information on matters of 
the first importance to airmen. 

The first of the wind channels at the N.P.L. was erected. Previously to that 
Dr. Stanton had been experimenting in a short vertical channel 2ft. 6in. in 
diameter. At the present date there is one channel 14ft. by 7ft. almost ready 
for use, three 7ft. by 7ft., two 4ft. by 4ft., and one 3ft. by 3ft. At the Royal 
Aircraft Establishment there are two 7ft. channels, while Messrs. Vickers, the 
Aircraft Co., Messrs. Boulton and Paul, and I believe other makers, have their 
own channels, and in the case at least of those of which I have direct knowledge 
the calls for work are greatly in excess of their capacity. 

A whirling arm for propeller research was also designed, and at an early date 
produced valuable results, while the erection of two towers, each 5oft. high, has 
enabled a number of researches in the open to be carried out. 

Wind channel work depends, as is now fully realised, on the law of 
dynamical similarity. Among the earliest Papers discussed by the Committee was 
one by Mr. Lanchester, ‘‘ Notes on the Resistance of Planes in Normal and 
Tangential Presentation,’’ and to this was added Lord Rayleigh’s ‘* Note as to 
the Application of the Principle of Dynamical Similarity,’’ in which, as usual, 
the author went straight to the root of the problem and laid once for all the 
foundations of our work. It is difficult to express the debt the Committee owes, 
especially for his work during those early years, to their late President. Scale 
effect and its proper valuation is still a difficult question, but Lord Ravleigh’s 
forms the key on which the solution turns. 

The necessity for full-scale work was realised at an early date, and the link 
between the Committee and Farnborough was strengthened in 1910 by the 
appointment of Mr. Mervyn O’Gorman as Superintendent of the Balloon Factory, 
as the Royal Aircraft Establishment was then called. In their report for 1911-12 
the Committee wrote :—Full-scale work was commenced early in 1911, under 
the direction of the Superintendent of the R.A.F. The earlier work was directed 
to the determination of the effect of various modifications in an existing machine. 
An aeroplane of Farman type was available for the purpose, and the altera- 
tions made aimed at diminution of head resistance by various means; the increase 
of mechanical efficiency by improvement of propeller design and correct correla- 
tion of propeller and engine; improvement in the design of the wings; increased 
2ase of control and improved directional stability. After stating that in 
all respects the alterations had proved advantageous, reference is made to the 
““ speed resistance ’’ and ‘‘ speed horse-power *’ curves which have been very 
generally used since to express graphically the properties of an aeroplane. The 
importance of apparatus for recording propeller thrust in flight is noted as well 
as thati of measurements of gliding angle wind velocity. 

In his first Paper on the subject, November, 1911, Colonel O’Gorman wrote : 

‘* Three classes of work must be carried on simultaneously and are naturally 
interdependent :— 


(a) Trials and scientific observations of the actual machine in flight, 
noting difficulties or desiderata. 

482 THE AERONAUTICAL JOURNAL (September, 1920 

** (b) Re-design to secure what is desirable or to overcome troubles so 


(c) Experiments in the Laboratory, both to elucidate the causes of 
abnormal effects seen in (a) and to give guidance and data on which to base 
constructional calculations. 

‘It will be seen that from the design work even so guided there results a 
modified full-size machine which in turn demands trials in the air, observations 
of peculiarities, experiments of a new or modified kind, and so on, the cycle 
being repeated each time with some advance. No one step in the cycle can be 
omitted without rendering the other work too speculative and perhaps futile.’’ 

The Paper from which this is taken is noteworthy as containing one of the 
earliest statements as to full-scale work and its relation to model experiments 
in connection with the development of a machine. It also contains the first 
report on aeroplane research in full-sized machines, while an addition, June, 1912, 
“* Further Notes on Full-scale Experiments,’’ gives the first account of a machine 
afterwards to become famous, the B.E. Class. 

Further work of this class is recorded in later reports, while Mr. Busk’s well- 
known experiments, in the course of which the stability of an aeroplane was for 
the first time thoroughly investigated and demonstrated, are to be found in the 
volume for 1913-14. 

It has sometimes occurred to me that we hardly realise the debt we owe to 
Colonel O’Gorman for the persistent way in which he impressed on the Com- 
mittee the value of full-scale experimental work and the need for increased 
opportunities to carry it out. The work at Orfordness and Martlesham Heath 
was the direct outcome of. these early efforts, and the methods of experiment 
employed were developed naturally from those of Busk and the pioneers of the 
Royal Aircraft Factory. 

To turn to another subject, the stability of aircraft was a matter to which 
the Committee at an early date devoted their attention. It is referred to in their 
first report, and in October, 1909, our indefatigable Secretary (Mr. Selby) 
prepared a summary of Papers relating to the stability of airships and aeroplanes, 
giving an account of the labours of Crocco, Soreau and Faber, with abstracts 
of Papers by Bryan and references to Lanchester’s early work in his volume on 
** Aerodonetics.’”’ 

But our knowledge was small and our ideas vague and indefinite. 

The problem was put on an entirely different footing by the appearance of 
Professor Bryan’s book in 1911. He showed that the nature of the motion was 
defined by the values of certain constants—the resistance and rotary derivatives 
—which themselves depended on the form and configuration of the surfaces 
composing the machine. The work was carried further by Bairstow and Melville 
Jones in their well-known Paper of .March, 1913; they devised experimental 
methods for determining from model experiments the values of the more important 
coefficients and carried out the necessary measurements on a model of the 
Bleriot type. The work was continued a year later in a Paper by Bairstow and 
Nayler, who also showed by a number of beautiful experiments on gliders the 
various manners in which instability might occur and the steps to be taken to 
prevent it. The value of this work has been fully recognised, and arrangements 
have been made for determining by experiments on a complete model the stability 
characteristics of any machine put into public service. 

Much attention has also been given to the strength necessary in the various 
parts of the machine; the stresses brought into play by various manoeuvres can 
in some cases be calculated; in others they have been determined by the aid of 
Dr. Searle’s beautiful Accelerometer. 

September, 1920] THE AERONAUTICAL JOURNAL 483 

Not the least important work done by the Committee is the issue of a 
schedule of load factors. The load factor under given conditions is the ratio 
of the stress in a member under those conditions to the stress in uniform 
horizontal flight. 

The Committee are well aware that in this matter there is still much to be 
learned, and in the light of future knowledge the specification will no doubt 
require modification, but at any rate it represents the unanimous conclusion of 
a representative body. And here let me interpolate a few words as to the method 
in which the Committee has worked. As various problems arose for. solution 
Sub-Committees were appointed to deal with them, and we were fortunate in 
securing the ready help of the men most capable of dealing with them. The 
number of persons engaged has been very large and the variety of subjects very 
great. For some years past Sub-Committees dealing with Aerodynamics, Engine 
Problems, Meteorology, Light Alloys, Inventions, and Accidents have sat at 
definite intervals and reported regularly to the main Committee. The work of 
editing the reports has been no light task, and for this and many other services 
the Committee is specially indebted to its Secretary (Mr. Selby) and his assistant 
(Mr. Nayler). For the experimental results they have depended mainly on the 
staffs at Farnborough and at the N.P.L., who have worked with the greatest 
devotion and skill. 

And now let us turn to the future. After 126 meetings the Advisory 
Committee for Aeronautics has finished its work. Its place is taken by the new 
Committee for Aeronautical Research, performing functions somewhat different 
from those of the Advisory Committee, undertaking certain Executive duties and 
constituted with the following terms of reference :— 

(1) To advise on scientific and technical problems relating to the con- 
struction and navigation of aircraft. 

(2) To undertake or supervise such research or experimental work as is 
proposed to the Committee by the Air Ministry and to initiate any re- 
search work which the Committee consider to be advisable; to carry out 
such work itself or to recommend by whom the work should be carried out. 

(3) To take over complete responsibility for the Air Inventions Committee 
and for the Accidents Committee. 

(4) To promote education in Aeronautics by co-operating with the 
Governors of the Imperial College. 

(5) To assist the Aeronautical industry of the country by scientific advice 
-and research and to co-operate with any Research Association that may be 

(6) To prepare for the approval of the Air Council a scheme of work 
and an estimate of expenditure for the year and to administer the funds 
placed at its disposal by the Air Council. 

(7) To make reports from time to time to the Air Council. 

The reference is sufficiently comprehensive and gives ample opportunity for 
aseful endeavour. While the Committee will work, I have no doubt, in cordial 
co-operation with the industry and will realise that those who have to make 
their productions pay can teach them much, the industry, I trust, will appreciate 
the fact that for ultimate success the fullest scientific investigation of many 
difficult questions is necessary. It does not appear probable at present that any 
outstanding discovery is immediately in front of us, progress will rather be 
through minor improvements in many directions. The advent of the petrol motor 
and the reduction thereby caused in weight per horse-power of the engine made 
flying possible, a similar revolution is hardly likely to occur again just yet, but 
improvement in many directions is certainly to be hoped for. Let us consider 

484 THE AERONAUTICAL JOURNAL [September, 1920 

the reference more fully and first as to education. The education in question 
is mainly that of the Engineer, the Designer or Constructor, or other scientific 
worker, the Meteorologist, the Navigator or Electrician, rather than that of the 
Pilot or the Mechanic. The Mechanic or artisan will be provided for naturally 
by special classes at Technical Schools in the neighbourhood of aircraft works ; 
the Pilot will learn and take his certificate at one of the flying schools; but 
the men who are to be Designers or Constructors need training in the theory of 
Aerodynamics, the principles of design, the properties of the materials used, the 
theory and construction of engines; and for the Meteorologists and Navigators 
the laws of the weather and the theory and use of many complex instruments. 
Moreover, access is needed to a Laboratory with wind channels and the other 
apparatus for model tests and to an Aerodrome with aircraft, machinery and staff 
for full-scale research. To such a specialised course an undergraduate training 
in Mathematics, Physical Science or Engineering forms a necessary introduction. 
For men who have passed through such training there should be provided :— 

(a) Courses for those who have attained the highest standards of 
University training at the various Technical Schools. 

(b) Courses based on a sound general Engineering education, but not 
involving too high a standard of mathematical knowledge, intended for 
men of special practical ability. 

(c) Opportunities for experimental work both in an Aeronautical Labora- 
tory and a Research Aerodrome. 

Now all this implies a heavy expenditure ; moreover, the number of such students 
must at present, in view of the opportunities open to them on completion of their 
course, be small. It was necessary, therefore, to concentrate, and in view of the 
generous foundation by Sir Basil Zarahoff of the Zarahoff Professorship of 
Aviation—a Professorship of the University of London held at the Imperial 
College—it has been decided to make the Imperial College the Central School 
for Advanced Study in Aeronautics, and funds have been provided by the 
Government for this purpose. 

There is much that can be done elsewhere. Aeronautics will, it is hoped, 
find some place in the curricula of many of our Engineering Schools; the 
principles concerned are the same whether applied to the structure of aircraft 
and the theory of the petrol engine, or to the building of a bridge or a locomotive, 
and there are many special problems which may well be studied wherever there 
is a man capable of the work and suitable apparatus for the investigator, but at 
present it does not seem possible to inaugurate complete courses of advanced 
construction at more than one centre, particularly when regard is had to the 
elaborate character and cost of the practical work. To secure this for the 
Central School arrangements have been made with the authorities at the Air 
Ministry and at Teddington by which students wili have access to Farnborough 
or other Air Stations or to the N.P.L. Details are hardly complete, but qualified 
students are to be admitted as student assistants, or in some such capacity, to 
these institutions and will learn by taking their share in the test work and the 
researches in progress. It is hoped in this manner to train a succession of 
Designers and Constructors fully qualified to carry on the brilliant work in- 
augurated during the last few vears. And we trust we shall not be dependent 
solely on the Government institutions for opportunities of practical experience 
for our students, but look forward to firms and private works for assistance in 
this scheme of education. The Governors of the Imperial College have appointed 
a representative Committee to advise them, and a scheme satisfactory to all 
concerned should be the result. In giving effect to such a scheme they will have 
the co-operation of the Research Committee, and for less advanced work we may 
look for help to many Engineering Schools. Aeronautical Engineering in its 

September, 1920] THE AERONAUTICAL JOURNAL 485 

principles does not differ from Engineering generally. No complete special 
curriculum is required. 

Mathematics and mechanics, some branches of physics and chemistry, machine 
design, strength of materials, and the theory of the heat engine are all common 
to the whole range of Engineering training. The Aeronautical Engineer will pay 
attention to the thermodynamics and efficiency of the internal combusion engine 
rather than of the steam engine. A high-speed petrol motor will interest him 
more than a marine turbine. The properties of wood or of aluminium alloys will 
concern him more than those of mild steel. In dealing with the structures and 
with the propelling machinery of aircraft the relation of strength to weight will 
be of prime importance, and no doubt at the Imperial College, as well as at other 
centres of Technical Education, opportunities will be given for such study to the 
undergraduate who proposes to join the Aeronautical industry, while for his 
practical training he will go to an Aircraft Factory rather than to an Engine 
Works or a Machine Tool Shop. 

Let me conclude this part of the subject with a reference to one other 
duty which I trust will fall to the staff of the Central School. It should serve 
as a clearing house for the co-ordination and dissemination of Aeronautical know- 
ledge in all its branches. To quote from a recent report: ‘‘ The functions of 
the teaching staff of the School may be stated under four distinct though closely 
related purposes :— 

‘* (a) To study, co-ordinate, summarise, apply and extend the knowledge 
derived from the experimental work carried out by the individual workers at 
various experimental stations in this country and abroad. 

‘*(b) To stimulate research by indicating what information is most 
urgently required and what line of attack is likely to prove most profitable. 

‘*(c) To guide and encourage research by constructive criticism based 

on the careful study of past and current work in this country and abroad.’’ 

After referring to the utility of a clearing house in any subject, the report 
continues :— 

‘* In Aeronautics the facts are the result of the work of the last five or 
ten years, and the framework uniting them exists only in the minds of the 
few men who have been personally connected with the process of develop- 

Before the war the total available knowledge was small, and it was possible 
for the members of the Advisory Committee to keep all the facts in mind while 
devoting the majority of their time to other duties. They then provided the 
necessary co-ordinating factor. This is no longer possible, and the function 
could best be discharged by the staff of the School working under their Director 
with a view to co-ordinating and making available all the knowledge in each 
branch of the work as existing at the moment. 

But I am wandering too far from the work of the Advisory Committee and 
of its successor; let me return to the latter. We have discussed somewhat fully 
the manner in which it may promote education in Aeronautics. Its main work, 
however, is to advise on scientific and technical problems and to undertake or 
supervise research and experimental work in Aeronautics. Let us consider its 
procedure and some of the more important problems calling for solution. As 
in the past the work will be carried on mainly through the Sub-Committees, 
and the legacies left by the Advisory Committee are by no means inconsiderable. 
Recently some general questions relating to fluid motion have been under the 
consideration of the Advisory Committee. When fluid is flowing steadily along 
a tube of diameter d and at a uniform mean speed v thence up to a definite value 
of the quantity wvd/v where v is the coefficient of kinematic viscosity, 

186 THE AERONAUTICAL JOURNAL [September, 1920 

the fluid is at rest along the walls of the tube, its velocity increases as you 
move away from the walls according to a parabolic law and the friction between 
the fluid and the tube is proportional to v. As the quantity vd/v increases beyond 
the critical value this law breaks down, and ultimately the friction reaches a 
value approximately proportionate to v®. It is of importance to know what exactly 
is the state of the motion at the surface. Is the fluid there still at rest, or does 
the lamina motion still persist close to the boundary ? 

Dr. Stanton has recently made some experiments on the flow close to the 
wall in tubes of various diameters and for values of vd/v of 460. to 
325,000. The measurements were made with a special Pitot tube, one side 
of which was the wall of the tube, while the other was a kind of small lip 0.05 mm. 
in thickness, which could be screwed outwards from the wall. By this means 
it was possible to make measurements of the friction on the walls and_ the 
velocity of flow for very small openings of the Pitot tube. In one series of 
measurements with a 12.7 cm. pipe the position of the centre of the Pitot tube 
ranged from 0.013 mm. to 0.178 mm. From the results it appears certain that 
there is always between the limits of wd/v indicated a very thin layer 
along the walls of the tube which is in a state of laminas flow. The velocity 
is zero over the walls and the friction is given by the limit of the quantity udv/dx 
when x is zero, x being measured at right angles to the surface. This result is 
one of the very greatest importance in the theory of fluid motion applied to 
Aeronautics, establishing as it does the conditions which must be assumed to 
hold at the surface of aircraft. Along with this, perhaps, should be classed as 
fundamental for aerodynamical theory some recent work of Messrs. Cowley and 
Levy and some investigations on which it is understood that Professor Bairstow 
is engaged on the equations of motion of a viscous fluid. 

The theory of the airscrew offers a wide field for investigation. Recent 
experiments of Mr. Fage have shown that the well-known Froude momentum 
theory when supplemented by the ‘‘ inflow ’’ effect gives a very good general 
account of the behaviour of an airscrew, and that the aerofoil theory when 
corrected for the interference of the blades in accordance with some very in- 
genious experimental work at the R.A.E. enables the form of a propeller to be 
designed with fair accuracy. But before a further advance can be made it is 
necessary to know more about the distribution of the air pressure over the 
blades. Experiments to determine this in the case of a model airscrew have been 
made by Mr. Fage, and the results will be before the Committee at an early 
date: there is little doubt that they will permit a further step in theory of great 
value. Meanwhile at the R.A.E. the apparatus for plotting the pressure over an 
airscrew in flight is well advanced, as also is the apparatus for measuring 
the thrust of the screw under the same conditions. More recently still a 
programme of tesis on a family of propellers of varying pitch diameter values 
and varying aspect values has been laid down by a panel of the Aerodynamics 
Committee appointed for the purpose, and it has been arranged to give this 
programme precedence over all other airscrew tests in one of the 7ft. channels 
at the N.P.L., while experiments on a reversible propeller are in progress. Among 
other airscrew questions is the effect of the high tip speed which is now reached. 
With various engines such experiments as) have been made have shown that with 
a stationary airscrew, as the tip speed reaches the velocity of sound, the character 
of the air flow round the propeller entirely alters; the slip stream disappears ; 
the air appears to be drawn in at the centre and driven radially outwards at the 
tips of the blades. Probably the thrust had almost entirely disappeared, and 
the propeller became very inefficient. 

In wind channel work perhaps the most interesting investigation in the 
immediate future will be a series of comparative tests which is being arranged 
‘between the national channels in this country, America, France and _ Italy. 
Model aerofoils of standard section will be tested in all the channels, and it is 

September, 1920] THE AERONAUTICAL JOURNAL 487 


hoped to include a test on at least one complete model. Details are now under 
discussion between the representatives of the various channels. M. Toussaint 
has made the very valuable suggestion that among the models should be at least 
one which has already been tested on the aerodrome track at St. Cyr. 

Methods for facilitating or improving the accuracy of channel work are 
continually under discussion, and in connection with the demand for a full 
investigation into the stability of any new type of aeroplane methods for measuring 
the rotary derivatives have become increasingly important. Here again is ample 
room for research and investigation. Some further experience will show how, 
given the values of the coefficients, the necessary calculations can be best effected. 
Recent Papers by Miss Cave-Browne-Cave and by Mr. Relf have thrown much 
light on this somewhat intricate matter. The importance of stability is generally 
recognised, and the researches now in progress as to the stability characteristics 
of many of the best-known types of aircraft are full of interest. 

Another matter calling for attention is the investigation of the aerodynamic 
properties of special forms of wings, especially high lift wings. A number of 
interesting results have been attained. It remains to compare these, and possibly 
to extend them, in directions which may offer promise of advance. Aerofoils 
suitable for airscrew design also offer a field of useful investigation. 

A problem to which the A.C.A. have recently devoted considerable attention 
is that of the prevention of fire on aircraft. It appears from the records of the 
Accidents Department that fires in the air are very rare; five were recorded 
in a period prior to December 31, 1918, during which over 500,000 hours were 
flown, while in the next six months the figures were four and 36,o00 hours. 
The fires on crash vary greatly with the type of the machine, ranging from 
one in 35 crashes in one type to one in 4.4 crashes in another. A _ striking 
difference between rotary and stationary engined types of machines was noted, 
the latter firing four times as often as the other. Investigation showed that with 
rotary engines a fire-resisting bulkhead had almost always been inserted between 
the engine and the rest of the machine. This led the Committee dealing with 
the matter to recommend the insertion of such a bulkhead as a general rule. 
They have made other preliminary recommendations and are continuing experi- 
ments with a view to improved safety. One obvious precaution is the use of a 
fuel having a much higher flash point than petrol. Thus there would be less 
tendency to catch fire through the-tank bursting on crash and the fuel splashing 
over the engine or otherwise coming in contact with something capable of igniting 
it. A number of such fuels have been prepared and tested. They are all somewhat 
heavier than petrol and less efficient for equal weights, but the experiments are 
still in progress. 

The question how fires occur on crash is one of difficulty. The tank usually 
bursts and the petrol is splashed about. Sparks from the magneto may be a 
possible cause, and accordingly steps have been taken to secure that the high 
tension system should be reasonably safe and the magneto fireproof. Again, 
the liability to burst differs greatly in various types of tanks, and recent experi- 
ments have shown a reasonable probability of designing a tank which will, not be 
unduly heavy in proportion to its contents and yet will remain intact when a 
crash occurs. 

A still other line of inquiry has been the attempt to provide a system of jets 
through which some fire-quenching liquid could be spraved did a fire occur, and 
this, though not completed, has met with a fair amount of success. With a 
view to showing how effect could be given to the suggestions of the Committee 
an aeroplane at Farnborough has beet modified so as to be practically immune 
from many of the risks common to other aircraft. The importance of all this to 
the safety of future passengers is very obvious and there is ample scope for 
further work. 

488 THE AERONAUTICAL JOURNAL (September, 1920 

When dealing with fuel and engines note should be taken of the experiments 
on engines burning other fuels than petrol. The phenomenon known as detona- 
tion is of importance in this connection, and the Engine Sub-Committee have in 
hand a series of experiments aimed at elucidating its cause. 

Reference has already been made to the report of the Load Factor Com- 
mittee. This raises many questions which will need careful study and_ in- 
vestigation. The endeavour to secure ample safety in all parts of the machine 
and the effort to reduce the dead load and increase the useful load pull in 
opposite directions, and are not always easy to adjust. Our knowledge is 
increasing continually, new materials become available, new methods are devised 
of distributing the material so as to reduce weight without impairing strength. 
All this means that if we are to progress we must investigate continually and 
be prepared as new facts appear to modify our former views and specifications. 

I have shown, | hope, that from the standpoint of design and construction 
there is ample opportunity for investigation, and that the new Committee will 
not lack material for its work. But there are other questions of a different class. 
Let me mention one or two in conclusion. Can anything be done to reduce landing 
speed without reducing the top speed? The advantages are obvious. Professor 
B. M. Jones has recently called attention to these and made some valuable 
suggestions for the investigation of the subject. Air brakes brought into action 
on landing are very inefficient; a reversible propeller otherwise satisfactory fails 
at the critical moment if engine failure compels a landing ; some means is needed 
by which control may be maintained when the plane is at or close to the stalling 
angle, and this at present appears perhaps the most hopeful avenue 
of advance. Stability is also an important subject of discussion, and 

one on which skilled pilots differ. Its desirability for many purposes 
is admitted, and as I have already stated the stability characteristics 
of all new types are to be determined. In the case of many machines a 

reduction in load factor is permitted if the machine is stable, and there is no 
doubt that in comparing certain classes of machines the accidents are much greater 
among unstable than among stable machines. It appears probable that an unstable 
machine will usually have a stable position when flying inverted, but the whole 
question is a complex one and calls for fuller investigation. A stable machine 
has a will of its own, and the characteristics which give it this are in some 
degree opposed to those desirable for rapid manoeuvrability. Accordingly among 
Service pilots there are many who prefer the unstable machine, and the decision 
as to the amount of stability to specify for fighting aircraft will need much 
consideration. A third and somewhat similar matter which needs discussion 
relates to the number of engines on a plane and the responsibility for running 
them. In most cases, I understand, the pilot has almost everything to do him- 
self. With a single-engined machine the actions required are sufficiently complex. 
When he has to attend to four, in addition to steering the machine, the task is 
often almost beyond his power. On a submarine or destroyer the officer in 
command communicates his orders through the telegraph to the engineer or the 
steersman, and it may be that some system such as this may be required for 
aircraft; at any rate, the suggestion is one for the Committee to consider. 

I have spoken long enough, and if in much that I have dealt with I may 
seem to have erred beyond the points of importance in the work of the Advisory 
Committee I hope I have said sufficient to show that their work has been 
important and that the legacy they are leaving to their successors is full of 
promise for fruitful work in the future. ; 

September, 1920] THE AERONAUTICAL JOURNAL 489 



Sir JoserpH Peraver thanked the lecturer for his interesting discourse. He 
wondered why he had chosen the title ‘* Some Points of Importance in_ the 
Work of the Advisory Committee for Aeronautics,’’ instead of simply ‘* The 
Work of the Aeronautical Committee.’’ He ventured to think he was led to that 
choice because he could hardly separate his identity from the work of the 
Committee, of which work he had been the heart and soul. One could say in 
two words that the work of the Committee had been the ** building-up ”’ of a 
new science—or an entirely new branch of science. There was already a science 
of hydraulics, of physics and of chemistry, but no science of aerodynamics. In 
building-up a new science somebody to lay the foundation stone was needed, 
somebody of deep thought and real genius, and that person in this case was 
Rayleigh. Somebody with imagination, vision and invention was needed to 
sketch the new building. It would be remembered that Lanchester’s book 
formed the first outline of the theory of the subject. Then there must be 
somebody to solve the mathematical problems which arose; Bryan's work gave 

the key to such solutions. Theories alone were useless. In building there 
must be an engineer, an architect, to work out the details and make the theory 
practical. That man was Prof. Bairstow. For the new science a mass of 

information, both theoretical and practical, had to be accumulated and great 
experimental abilitv, hard work and consistent patience were needed, and, in 
the case of the full-scale experiments in a science of that kind, fearless courage 
on the part of the experimental pilots. The first systematic and scientific full- 
scale experiments were organised by Col. O’Gorman at the Royal Aircraft 
Factory. He thought tribute should be paid to Col. O’Gorman, under whose 
inspiration a group of able and fearless experimental pilots sprang up. All had 
been ready to run great risks in the interests of the work in hand, and several 
had given their lives. During the past twelve vears the new building—the new 
branch of science—had been truly and completely built, and its result had been 
not a small factor in winning the war. As Sir Richard Glazebrook had said, a 
new organisation was ready to impart the knowledge that had thus been 
systematised, and a new Committee would carry on the work. It was interest- 
ing to inquire what the present objects in view should be. He suggested that 
Aeronautics was a subject of the highest importance to the entire British Empire, 
in knitting together the too widely spread elements of the Empire. It was 
needed commercially, and it would be needed some day from a military point 
of view. The work of the new Committee would therefore be to accumulate 
an armament of fact and new knowledge with which to fight the battle of 
commerce or industry, and possibly, at some later date, of real war. 

Prof. L. Barrstow said he regarded the formation of the original Advisory 
Committee for Aeronautics in 1908 as one of the most important marks in the 
history of aviation in this country, and, he thought it fair to. say, in the history 
of the world. The example had been followed in America, where there was now 
a National Advisory Committee on the same general lines. He regarded the 
Committee, which had no very general executive powers, as having had its chief 
value, in the first place, in encouraging experimental work and research. During 
the whole time he was connected with the National Physical Laboratory, under 
the auspices of the Advisory Committee, ample funds were placed at the disposal 
of the Aeronautics Department, and work, not always of too excellent a quality, 
Was criticised sympathetically, while general lines of policy were laid out for 
the staff of the Laboratory to follow. As Sir Richard Glazebrook had_ told 
them, the programme put forward in the report of the Committee in 1908 con- 
tained a long list of problems of great importance which had not yet been 


exhausted, and the remarks of the members of the Committee indicated the 
state of knowledge at that time. The first slide used by the lecturer showed a 
state of knowledge which was a long way ahead of that at the time of the 
original reports by Mr. Lanchester and the late Lord Rayleigh. Lord Rayleigh 
put his note on one page of the first blue book, and he (Prof. Bairstow) 
remembered taking the blue book home for the summer holiday and failing to 
see the meaning of that page. Another article of Lord Rayleigh’s gave a clue 
to the whole matter, and for two years afterwards the work of the N.P.L. was 
fundamentally affected by that brief note. The assumptions as to what happened 
in the fluid dealt with in the motion of aeroplanes through the atmosphere were 
mathematically crude. In particular the assumption was made that the properties 
of the air on which the resistance depended were known, and accepting this 
Lord Rayleigh showed that the form of the law of resistance must be known. 
It, however, needed an experiment such as that described by the lecturer to 
establish the assumption. The relation between model and full-scale experi- 
ments is involved, and wherever the law can be used the position is much 
better than would have been the case if no law had been stated. The law 
(of similitude) also brought out the fact that one could not depend on model 
results alone, but must go for finality, to experiments such as those carried out 
at the Aircraft Factory by Busk, who, unfortunately, lost his life in the early 
B.E.2, and Keith Lucas, who also gave his life to aeronautics. Many other 
pilots were willing to take up aeroplanes known to be prone to accidents, and 
investigate the cause of the accidents. The Advisory Committee for Aeronautics 
learnt of all these full-scale experiments, not because of its authority, but 
because the people who were doing experimental work had a great trust in it, 
and were anxious to submit their work. The N.P.L. was directly controlled by 
the Advisory Committee; the Royal Aircraft Factory was not, but there was no 
evidence of that fact in the way reports were presented and discussed. Not only 
did the official establishments make communications, but private individuals 
communicated Papers and were usually invited to be present at the discussion 
which ensued. The result has been that the reports of the Committee have 
become a general collecting ground for the whole of British aeronautical in- 
formation. During the war aeronautical research suffered to some extent because 
a result was wanted quickly, and the tendency was to expend a considerable 
amount of energy in the production of results which were of little permanent 
value. Since the war there was evidence of a return to sounder experiment and 
research, and he hoped that one of the chief functions of the new Committee 
would be the co-ordination of experimental work, particularly at its inception. 
A great deal of data was being collected, and they knew, better than was 
formerly the case, where they were going. Some experiments were not worth 
doing—comparing their relative importance to others—and it was in regard to 
such questions that the old Committee did its best work; he hoped the new 
Committee would do equally good work in this direction. 

Commander LAxp said when he recently had the honour of being elected a 
Foreign Member of the Society he was glad to take advantage of the courtesy, 
and he was desirous of joining for three reasons: First, on account of the honour ; 
second, for professional reasons; and third, because he thought he would be 
absolved from being called upon to make speeches about subjects of which he 
knew very little. He was once told by a man who thought he was a speaker that 
the only way to hold the attention of an audience was to tell a funny story or insult 
their intelligence, but most of the stories he knew were not of the dinner-table 
variety (unless coffee had been served), and he did not know enough about this 
subject to insult the intelligence of the audience. 

The work done and being done by the Advisory Committee of Great Britain 
had been a tremendous asset to the aeronautical science, not only of Great Britain, 
but also of the world. As a previous speaker said, a similar Committee had been 

September, 1920] THE AERONAUTICAL JOURNAL 49] 

formed in the United States, and he understood it had co-operated and co-ordinated 
with the British Committee. He felt sure this was true so far as the Air Ministry 
and the Navy Department of America were concerned. There was an interchange 
of results of tests which was mutually advantageous, and possibly prevented 
covering the same ground on both sides, unless it was considered necessary to 
confirm the evidence. A previous speaker said the results obtained in model 
experiments could not always be obtained in the life-size results. He knew that 
was true in regard to some designs, such as ship designs, but on the other hand, 
one obtained a great deal of knowledge, and the law of similitude followed fairly 
closely the model basin results. He had one comment which he hoped was con- 
structive, i.e., that it was essential for the Advisory Committee, engine designers 
and hull designers to co-operate with the operating personnel. The latter were 
usually willing, but did not always take the trouble to keep research committees 
and designers informed of the results obtained in actual operations, and he was 
glad to hear there was that co-operation here. It was sometimes difficult to get 
the co-operation of the operators unless it was negative. If anything was wrong 
they would give criticism, but frequently things of importance on the constructive 
side went by the board, and the designers did not know whether they should be 
repeated or not. In case of accidents it was difficult to determine the cause. He 
thought many operators, pilots and observers did not like to put their ideas to 
pen and paper and really put down what they thought was the cause of the trouble 
and what were the good and bad points of a machine. He suggested to the 
Advisory Committee that they should keep more in touch with the people who 
used the machines, no matter of what type, through the people who worked in the 
laboratory and the model basin, so that the Advisory Committee would be the 
co-ordinating link between the designers and operators. Certainly in the Navy 
it was an advantage for the people on the beach who produced the ships to get 
the best ideas of the people who operated them. Some propaganda work might 
be done to bring them more closely together, so that the operating results might 
be better known in the laboratory or model basin. 

Captain G. T. R. Hit said,the lecturer had given an account of the work 
of the Advisory Committee from the point of view of Commander-in-Chief. He 
would like to put forward the view of a private who worked for some time for the 
Committee. For about eighteen months after the beginning of the war he was 
connected with the Royal Aircraft Establishment; later he was there again in 
the Experimental Squadron, and afterwards at the Air Ministry under the inspiring 
influence of Prof. Bairstow. After the war he (Capt. Hill) became connected 
with Messrs. Handley-Page, Ltd., on experimental work. 

At the beginning of the war flights were great occasions. The Royal Aircraft 
Factory had only two aeroplanes that would fly—one of which was the B.E.9, 
whose stability record had just been shown to the meeting. 

When he was back with the Experimental Squadron in 1917 they had quite 
a number of machines which could fly, something like 20 or 30. In his opinion 
they were not so short of aeroplanes as of measuring instruments, particularly 
instruments for measuring control and stability. A record shown by the lecturer 
of the variations of loading on an aeroplane during a mock flight was one of the 
first successful ones taken with the accelerometer. Capt. Noakes flew the aero- 
plane on which the accelerometer was mounted and he flew the other machine. 

He always thought it would have been much more interesting if they had 
both carried accelerometers, and co-ordinated the two records, which would have 
shown up the hottest moments of the flight. He thought they were still short of 
instruments. Towards the end of the war machines became plentiful enough, 
and he believed the number now at Farnborough was nearly a hundred. 

Turning to what he had seen of research in a civilian firm, at the end of 
the war not many orders were coming in, but he was very greatly encouraged 

$592 THE AERONAUTICAL JOURNAL [September, 1920 

by the way research was pushed on, though, of course, on a smaller scale 
compared with what he had been used to seeing. In the present state of the 
industry, and in view of the large expense incurred in full-scale experimental 
work, he felt that they would have to look for the great majority of their 
experimental results to the Government establishments. He never realised until 
it became a question of accounting for the money, what a great expense aero- 
nautical research involved. They were pushing on with full-scale work, and 
were held up partly by lack of instruments and partly by lack of funds, and it 
was to Farnborough they would have to turn for most of their full-scale data. 

With regard to the training schemes proposed, he felt that there was a 
tendency to choose a man to be a pilot and train him at a flying school, and at 
the same time to choose other men and train them to be constructors and de- 
signers. Perhaps it was the simplest way, but he thought there would be a 
definite loss of efficiency if the constructors and designers of the future were not 
pilots themselves. He felt that this had been one of the reasons why progress 
had been slower than it otherwise might have been. Speaking as a pilot, he 
felt that pilots were rather inarticulate—not very good at expressing their 
opinions and writing their conclusions down clearly—and that this was partly 
due to lack of scientific education. He thought everyone would agree that very 
few pilots were expert mathematicians and physicists, and that few mathematicians 
and physicists were good pilots. It seemed very useful for a man to combine 
mathematical and physical ability with skill in the handling of aircraft. Un- 
fortunately, it appeared that opposite types of men were expert at those two sides 
of the question, but the fact that they wanted people familiar with both should 
not be lost sight of. How they were to be obtained he would ask others to 
decide. Now, in time of peace, the putting in of hundreds of hours necessary 
to make a first-class pilot cost so much and would take so long with the few 
machines available, that it would be difficult to get expert pilots who were also 
expert engineers. 

In conclusion, he would like to mention that he had always been met with 
the greatest courtesy at the N.P.L. and the R.A.E. since he left the Air Force, 
and he looked forward with pleasure to his visits to those places. Those who 
went there for information, in his experience, came away with more than they 
expected to learn. 

Major A. R. Low said that in the early days of Aeronautics he had been 
a stout defender of the A.C.A.’s policy of scientific research and of the Royal 
Aircraft Factory’s co-operation, when the Press and manufacturers objected to 
the absorption of a large proportion of the available funds in ways that brought 
no direct returns to the imdustry. He therefore felt himself able to offer dis- 
interested criticism from outside the official circle of research. 

There was a tendency, especially on the part of junior members of that 
circle, to hold a complacent belief in their supremacy in aeronautical research. 
The U.S.A. were cited as mere understudies, the French received little mention, 
the Germans less. Yet many of the most fruitful ideas, as distinct from mere 
routine measurement, had come from outside. The late Lord Rayleigh’s remark 
on aerodynamical similarity had controlled the whole comparison of model 
with full-scale work, and though a member of the Committee he had attained 
fame and knowledge in other places and times. 

Bryan had solved the problem of stability fore and aft in 1903, and his 
aper to that Society read as fresh as if written yesterday. -As the lecturer 
had told them, Bryan completed his placing of the whole question on a mathe- 
matical basis in 1911. Several men had risen to fame on applications of Bryan’s 
methods, but he would like to ask if full use was being made of Bryan’s powers. 

Alexandre Sée had given a theory of aeroplane dimensions far more complete 
than anything developed in this country, but his work was ignored. 



Toussaint’s name he would always associate with a system of testing which 
was in many ways superior to those standardised here. 

In the U.S.A. they had installed a high altitude laboratory far in advance 
of anything done here, and their published engine research generally was ahead 
of ours. 

Aeroplane engine design offered the most conspicuous target for criticism, 
for we had gone into the war without an aeroplane engine industry, and without 
French engines could not have carried on the war in the air in the early years. 

Durand had carried out a monumental research on airscrews of different 
forms, preferring a wind channel of Eiffel to N.P.L. type, and we were preparing 
to follow. 

Drzewiecki’s analogy between parallel strips of aerofoils and coaxial strips of 
airscrew blades had proved one of the most fruitful and powerful methods in 
aeronautics, yet when he had presented his view of it to that Society in 1913 the 
N.P.L. representative had criticised it as a mere hypothesis unsupported by 
experimental evidence, a statement which showed lack of knowledge of foreign 
research. He had at that time pointed out that if the coaxial cylinder containing 




. § 21 | 

3 ‘\ | 

qx ‘ 

220 MeanSpeed tor Both 
s Blades in Complete 

% « \Circle. 



218 ~ 

a) One Blade inSemi-~~~3 
BS 7 circle occupied by 

= Blade, 

} -- 


Steady Wind Speed. 

the spiral paths of blade elements were cut along a generator and developed on a 
plane it would be seen that blade elements followed each like aerofoils in tandem 
with a path distance determined by elementary kinematics. He was interested to 
see that the idea was now claimed by the official staff as put forward by-them 
in 1916. He would not be surprised to find that the idea was as old as marine 
screws. The N.P.L., however, had the credit, as far as he knew, of being the 
first to carry out this important piece of research, which would complete the 
Drzewiecki theory and would reduce the problem of finding the surface integrals 
of thrust and torque over the blade of an airscrew to experimental determination 
cf the reactions on a series of aerofoils of different sections corresponding to blade 
sections at a sufficient number of points, several aerofoils of each section being 
arranged in cascade, as it was called. 

The lecturer had stated that Froude’s method of measuring the increase of 
momentum in the well-defined column of flow was being investigated, and that a 
combination of the two methods would complete the:solution. He had already 
argued that the solution was completed by the cascade tests, and, further, he. did 
not admit that the measurement of the velocities across suitably chosen surfaces 
cutting the column of flow threw much light on the surface conditions over the 
blade. Even the complete exploration of the whole column of flow would not give 

494 THE AERONAUTICAL JOURNAL [September, 1920 

a complete account of the reactions unless the tangential forces at the blade surface 
were determined by some means or other. Of course the volume integral must 
theoretically agree with both Froude’s and Drzewiecki’s surface integrals, but 
they must here consider what they were actually measuring, not what they 
would like to measure. Drzewiecki argued that Stanton was measuring the 
average momentum across the column of flow, and the former had designed a 
Pitot tube with an interrupter gear which measured the instantaneous velocity 
at every point round the circle relatively to the blade positions. The results 
(shown in Fig.) confirm the prediction based on the aerofoil analogy that 
the flow is of a highly pulsating nature. The upper full horizontal line shows 
the mean flow behind a blade due to both blades, and the lower the mean flow 
due to one blade in the semi-circle occupied by the other blade. Without attempt- 
ing to give a final opinion on the subject, it is obvious that Drzewiecki could not 
be merely ignored, and on the general subject of airscrew research he had stated 
that Eiffel’s laboratory had furnished, beyond dispute, the most valuable experi- 
mental data, which might be set against the English claim that the A.C.A. had 
furnished by far the greatest volume of aeronautical data. 

Major Hill had referred to the present lack of funds for research, but if this 
gave more time for reading up the literature of other countries and for thinking 
about the subject there might be gain, and not loss, although the piling up of 
tabuiated measurements might be reduced in quantity. 

These small criticisms must not be taken as carping, for no one was more 
keenly aware how great was the work done by the A.C.A. in putting at least 
that branch of engineering on a sound technical basis in this country. He hoped 
that other branches of the engineering industry would follow the brilliant example 
set by Sir Richard Glazebrook. 

Major D. H. Kennepy said he wished with diffidence to refer to two points 
near the end of the Paper. The lecturer sketched out an alluring educational 
programme which made older men envy the aeronautical student of the future. 
In connection with that programme he thought the need for popular aeronautical 
education should be considered. Aeronautical science was now going through 
the stage which the electrical science passed through thirty or forty years ago, 
and wonderfully good work was done by pioneers in popular electrical education. 
He was afraid much of the valuable work done now by Research Committees 
did not get down far enough—did not reach the rank and file. Referring again 
to electrical education, he remembered going to lectures by Prof. Garner in the 
Durham College of Science, and those lectures were listened to by thousands of 
people. They must realise that the air way was the way of the future, and there 
must be a great programme of popular education. He would be glad if Sir 
Richard Glazebrook would put before the Air Council the desirability of inducing 
members of their staff to take advantage of such opportunities as were to be 
had of seeing full-scale experiments. He knew many people at the Air Ministry 
with whom he came into contact would welcome such opportunities if they were 
put before them in the way of a really strong invitation. In regard to the 
reference at the end of the Paper to the engine room, he thought that up to now 
our pilots—or pilots plus observers—had been asked to do far too much. Some 
single-seaters had about twenty instruments in front of the pilot. Some work 
should be done quickly in the direction pointed out by Herbert Spencer when he 
said that progress depended upon differentiation of function. Something should 
be done in the way of dividing into the maximum instead of the minimum 
number of functions the work of steering and navigating an aeroplane. 

The Lecturer, replying briefly to the discussion, said he took the title of 
‘* Some Points of Importance,’’ etc., because an attempt to give an account of the 
whole work of the Advisory Committee was beyond the time and space of a 
single Lecture. He was grateful for much that Sir Joseph Petavel and Prof. 

September, 1920] THE AERONAUTICAL JOURNAL 495 

Bairstow said in regard to the Committee’s work. He assured Commander 
Land that he and his colleagues on the Committee realised the need for securing 
the fullest co-operation between operators and designers and constructors. The 
fact that the Committee almost always had at its meetings some of those 
directly concerned as pilots or as engineers in the construction of aircraft was 
of the greatest assistance. If Capt. Hill could assist in what he suggested— 
encouraging the engineer and constructor to acquire at least the rudiments of 
the pilots’ craft—he would have done a valuable piece of work. It was a 
matter he (the Lecturer) had discussed with Prof. Bairstow a few days ago, and 
it was a part of their scheme of education for engineers at the College to arrange 
for their securing some knowledge of the work of the pilot. In regard to Major 
Low’s remarks, he hoped that nothing in his Paper claimed infallibility for the 
work of the Advisory Committee, or even, in many respects, superiority over 
other work. He realised all they owed to investigators elsewhere, and that with- 
out the work of Drzewiecki on propellers their knowledge would be much inferior 
to what it was at the moment. He would not go into the two views of the 
action of the aerofoil propeller, as it would take him too long and too far into 
mathematics. They must keep the matter of popular education before them. It 
was only by such means the man in the street could be made to appreciate the 
importance of the advance of the scientific and research work which it was hoped 
to carry out. 

The CHAIRMAN, in proposing a vote of thanks to the lecturer, said it was 
impossible to over-estimate the importance of the work of the Advisory Com- 
mittee in the past. It was not possible in a short Lecture to describe the whole 
of their work, but the work since 1909, when they started, had been very great. 
Now Lord Rayleigh had passed from them Sir Richard Glazebrook was entitled 
to the greatest honour in connection with the work, and he had added to their 
debt of gratitude by coming there and giving them such an interesting and 
instructive Lecture. 

A vote of thanks was also accorded the Chairman, on the motion of Wing 

Commander CAave-BROwNE-CaAvE, who said the Council had been glad to get such 
a distinguished officer to consent to preside at the meeting. 



The following extremely important table, compiled by Major Linton Hope, 
Was inadvertently omitted from the reprint of his lecture in the August issue of 
the Journal. It should also be noted that the last set of Flying Boat lines, on 
page 470, is the design which Major Linton Hope made in 1916, after the trials 
of the 41ft. design, Fig. 2, and to which he refers in his reply to Mr. Baker. It 
should be noted that the angles of the step tangents to the foreplaning bottom 
have been increased and the back step reduced in size and moved aft, both of 
which features appear in C.E.1. designed some considerable time afterwards. 




({September, 1920 





| ft. 
1 2 Curtiss, 90) 32.96 
h.p. | 

2, Austro- 24.50 
Daimler, | 
120 h.p. 
_ 24.00 
4} Mono-Gnome,} 26.16 
4 100 h.p. 
22 320 h.p. |56.00 
1-300 h.p. 
6/9 Rolls-Royce,| 42.25 
7 300 h.p. 
“ 2 Sunbeams, | 45.00 
320 h.p. 
89 Rolls-Royce,| 45.00 
. 360 h.p. 
9 Hispano- 30.00 
200 h. . 
10 Ditto. 30.00 
11 9 Hispanos, | 34.7 
200 h.p. 
2) pitto. 134.70 
13, 200 hp. | 23.99 
14 Hispano- 23.90 
200 h.p 
15 Ditto 23.90 
16; 900 h.p.? | 28.00 
17 9 Rolls-Royce,| 45.00 
360 h.p. 
18)9 Rolls-Royce,| 45.00 
360 h.p. | 
19) 1,875 bp. | 60.00 

20° 4600 R.R. | 64.00 





3.00 | 


5.66 | 

5.66 | 













of Planing 

I B. 
ft ft. 
12.37 | 8.00 
12.50 | 4.00 
12.50 | 4.00 
9.00 3.25 
24.00 | 14.00 
30.00 | 10.00 
30.00 |.10.00 
| 30.00 | 10.00 
{13.50 | 4.00 
113.50 | 4.00 
12.00 | 8.00 
12.00 | 8.00 
12.33 | 8.00 
9.50 | 4.00 
| - 
; 9.50 | 4.00 
112.95 | 4.90 

120.25 | 7.50 









y. ft. 






660.0 | 

of Skin. 

5/16 in, spruce 
and — fabric 

2 skins, carvel 

Top sides, 
3/32 in diag. ; 
11/64 in. FL & 
A, Bottom, 2 
skins cedar, 
5/32 in. diag. ; 
316in. F.& A. 


3/32 in. maho- 
gany diag.; 
3/16 in. cedar, 

FL& A. 

Carvel, 1 skin, 
3/16 in. maho- 
gany, fabric. 

Carvel, in. 
mahogany, 1 

3/32 in. diag. ; 
hin F.&A., 

3/32 in. 

in. I. 

& A. 

3/32 in. maho- 
gany stringer 


gany, single 
skin, 3/16 in. 

2 skins, maho- 
| fabric, be- 


| tween 5/64 in. | 

diag., 5/32 in. 

F.& A. out. 
}9/32 in. at 
| 2.036 lbs. per 
| sq. ft. 
| skins) 

(2 | 

Timbers. Striugers. a 
s z 
gin. x 2 in, Ribbands, 1 | _ 
ash,spaced | in. x1} in. | 
63 ins. ash. 
gin. x5/l6in.| 2 in x ¢ in. | 
elm,spaced spruce | 
1 ins. 
= _ 520.0 
7/16 in. x 5,16) lf im. x fin. | - 
in. R. elm. to ¢ in 
spaced 2) spruce. 
_ _ 1,585.0? 
— 1,560.0? 
ees _ 1,608.02 
tin.x5)l6in. Ijin. xdin.| 346.5 
elm, spaced spruce. taper-| 
1} ins. ing F.& A. 
Ditto. Ditto. 333.0 
7/16in. x 3 in. -- 
spaced, 4} | 
ins. | 
Ditto. - 
tin.x3/16in. i im. x Gin | _ 
elm,spaced spruce. | 
2} ins. 
eg | - 
3116in.x iin. 14 in. xljin. 165.0 
elm,spaced| X {i in 
13 ins. spruce. | 
Ditto. Ditto. | 165.0 
: : | 
5/16in.x fin. Heavy spruce,| = — 
R. elm, 1) in. x ¢ in. 
spaced 1; to 516 in. 
in. (?). (?) | 
5/16in.xdin.,! 1,in.x 2 in | Estd. 
spaced, [} | spruce. 1,160.08 
in. | 
. ° ! ° 
Ditto. Ditto. | Ditto. 
oe = an 
5/16in.x gin. | 2.25in. x5 ad | 2,505 

x 2in. 

in., I. fillets 
5/16 in. and | 

per sq. ft. 








* F.or A. indicates fore or aft of main step. 

t Indicates weight without 

September, 1920] THE AERONAUTICAL JOURNAL 497 
50 Total 3 | & a | 
Planing Extensions. Pal Wei Z = | a | Oy | OQ 
A _ Total os | Displace- S| et | 
=e oo jE! «CDisplace- | 68 | mentin Fe ca | 2 Remarks. 
en | mentof TZ | Flying 7 3s | & 
w 2 (Hullinibs. 2] Trim = | sa | ss 

Area. Thickness. o| we ie a ne | 

ze = : | ,e) ar | 
=| = p | 
sq. ft. 5 p.c ft Ibs. 

123 | 3/16 in, maho- | — ~— ‘ 21,256 6.0007 301.3, 75 TF 1 step. No 
; gany diag., | ee 
| 5/16 in. maho- keelson or 

gany FL & A. deep strin- 
| | gers. 

= ! —_ = a — _— “ 2.25 F | 1 step. 
ea 160 9,150 0503 2931 212 295 F | 245) Hull, C.G. 

2.625. Im- 
| | proved ty pe. 
| | 1 step. 

} | 
| _ 3 320 7,800 0410 2,260 : 33 A | 22.6 1step. 

201.9 13 skins—2 in- | - = ne 3,000 114,200 | 0262 21,000 (453.8) 66 F | 22.5 , Total area 
| ner diag.,skin, | | P. coeff. 
|i in. and 1 | 4. Felix- 

outer F. & A,, | stowe  de- 
|> 16 in. maho- | | sign. 1 step. 

a ae Estd. -- 1,775¢ 538,800 | .0530 11,544 2.98 F | Ditto, 2 step. 

192 | 
—-} — | Estd. = 1,760 57,558 | 0514 12,800 | 348.0 | 3.33 F | 19.7 Ditto, 2 etep 
192 | 
| - | Estd. _ 1,800 _ -- ae = | - Ditto, 2 step 
\ 192 , | 

59.65 | 1/16 in. cedar 12 - | 450 136 | 15,040 0304 3,967 321.6 1.16 F | 17.8 2 steps. 
idiag., } in. ma- Error 7 Ibs. 
jhogany F.& A. | | or 1.5 p.c. | 

59.65 ! Ditto. 2) — | 430 110 15,010 029 — = | 116r - 2steps. 

Error 41bs. 
| or 1.6 p.c. 

$21 | 3/32 in. diag.|24; — | — 25,900 0456 86.650 | 281.0 | 16.6 step. I. & 
jmahogany, a5 | | A. keelson 
}o'32in. F.& A., | | Ibs. or 15 and. strin- 
| $an.t. & A. | .c. * | gers 

321 Ditto. 24) — | 1,004 1,021 25,5: 0100 6,500 | 286.2 16.5 
| Error 17 lbs. | 

| Or 1.7 pc. | 

48.0 | 3 skins -- 5/52) — _- 275 7,082 0384 2,600 170.14 | 13.0 | 2-seater 
lin.and 11 64in. | | (? Seout. 
| cedar, and | in. | | 
' mahogany. | 2 | i )( 

37.5 (4/16. in. diag.| 12) — 216 210 7.351 9280 2359: 19133 | 11.7 Total area 
| cedar, {in F..& Error 4 lbs. | P. coef. 

' A, mahogany. | or 1.8 p.c. 4. Single 
; | | seater. 2 
' | steps. )( 
37.5 (1/16 in. diag. | 1.2 | 212 209.5 7,351 | Q280 - oh - Total area 
cedar, Lin. F & | j Error 23 Ibs. P. coef. 
A. mahogany. | or 1.1 p.e. 54. Single 
| seater, 2 
13/32 in. diag.| — : 503 527 2 4.912 ca Le AY -— ae hs: vexs 
| mahogany ? ; | trial cess due 
) 5/32 in. FL & A. | 4,600 to heayy 
' mahogany ? ; | fesid. spruce c.g. 
}3in. FL. & A. | of hull 53 jn. 
mahogany ? F. of step, 

211 |2 skins maho-| 1.2; Estd. | 1,290$ | 1,395 18,250 — | 12,000 301.9 16.7 Total area 
lgany, fabric 176 | Error 103 ¥ P. coef. )( 
i between 3.16 Ibs, or 8 50. Fr. 3 

in. F. & A., 3/32 | wing strue- 
in. diag., 552 ture. 
in. F.& A., 5/61 
| in. diag. (Aft). 
241 Ditto. 1.2] Ditto. | 1,202§. 1,281 48,230 — 12,000 [301.9 16.7 Redesigned 
Error79 lbs . ¢ Wing struc- 
or 65 pe. . ture. 
{—- _ 3,106 32,400 _ 4.75 ¥ =e =< 
max. load 
{ 21,000 
— _ _ _ 2,970 2,982 |Error121bs. 52,000 _ — _— aa 
! or 0.4 p.c. 

fabric. § ludicates weight without gun mounting and side forts. 

498 THE AERONAUTICAL JOURNAL [September, 1920 




The resistance offered by bodies to motion through fluids has for many years 
been a problem of the first importance, and one which has received a corre- 
spondingly large degree of attention both from scientific investigators and those 
commercially interested in the subject. A considerable amount of mathematical 
investigation has been carried out; but a detailed study of fluid motion has only 
proved possible in the case of a perfect fluid, i.e., one in which viscosity is not 
present. Assumptions have been made as to the quantities upon which the motion 
of the fluid and the body depends, and formule have been prepared expressing 
the laws connecting force and speed for dynamically similar bodies. 

In the present note attention will be directed chiefly to the case where the 
only properties of the fluid upon which the motion depends are the viscosity and 
density, though brief mention may be made of instances where the acceleration 
due to gravity is also important. 

The familiar example in which the action of gravity affects the motion is that 
of a naval vessel which sets up surface waves. In such a case the force (F) 
acting upon the body will be equal to 

pol? F (el/v, v*/1g) 

where v is the speed, | a dimension of the body, y the acceleration due to gravity, 
and p and v the density and kinematic viscosity respectively, of the fluid. The 
frictional resistance is a function of vl/v, while that due to wave making is a 
function of v*/ly. 

Aircraft, like submarine naval vessels, may be regarded as completely sub- 
merged in a fluid so that there is no wave making effect at the free surface of 
the fluid. It appears probable, therefore, that under these circumstances there 
will be no gravitational effect and that changes of v*/lg may be neglected* ; this 
conclusion is supported by the experiments which have been carried out on aircraft 
and models. 

It should be noted that in the flight of an aeroplane the force due to gravita- 
tion is opposed by the lift, and in order to maintain horizontal flight of the 
aeroplane on an increase in the value of g it would be necessary to reduce propor- 
tionally the mass (without change in the form of the machine), or increase v* by 
an amount inversely proportional to the change in q; the attitude of the machine 
is, under the condition of similarity, supposed constant. It therefore appears 
that changes in the magnitude of gy are compensated by methods which, as would 
be expected from the previous statement that the motion is independent of changes 
in v?/lg, only modify the *‘ flow pattern ’’ due to change in vl |v. 

The forces on an aeroplane would be directly calculable from experiments 
on a model if the latter were carried out at the same value of vl/v as for the 
aeroplane. For aircraft this is rarely possible and a knowledge of the form of 
the function f is therefore necessary in order to make accurate predictions from 

* An additional term should be introduced for compressibility of the air, but this does not appear 
to be of importance at present flying speeds. 

September, 1920] THE AERONAU TICAL JOURNAL 499 

In experiments on models of naval surface vessels, however, it is possible 
to secure the value of v*/ly appropriate to full-scale. Thus, so far as the major 
portion of the resistance (that due to wave making) is concerned, naval architects 
have an advantage over designers of aircraft in the prediction of performance from 
models. The prediction of the remaining skin-friction resistance, which may 
amount to as much as 4o per cent. of the total, is carried out by an empirical 
process based on Froude’s experiments with flat boards. The air resistance due 
to the superstructure of the vessel is ignored. 

In aeronautics perhaps the most important practical application of informa- 
tion as to the resistance of bodies of approximately streamline form is in connec- 
tion with the design of airship hulls, since such resistance represents a large 
proportion (perhaps 60 per cent. in a rigid airship) of the total resistance of the 
aircraft. In such a case the large dimensions of the hull render experiments at 
the same value of tl/v extremely difficult; in a wind channel observations would 
be taken at approximately the same speed and the model would be about 1 150th 
full size. Under these circumstances it is obvious that the prediction of full-scale 
airship resistance from experiments on models is rendered uncertain unless, at the 
highest value of vl/v at which experiments can be carried out in a wind channel, 
the resistance be found to vary accurately as the square of the wind speed, and 
the ‘‘ square law’ holds up to the full-scale value of vl/v; accuracy is here a 
matter of importance since if only a slight variation is taking place an increase 
in vl/v to 150 times its wind channel value offers ample scope for an important 
change in the resistance coefficient. If the resistance vary as the square of the 
speed a non-dimensional coefficient, F'/pv7l?, may be employed to calculate the 
forces on the actual airship. 

Experiments on models show that on many airship forms there are marked 
departures from the ‘‘ square law ’’ though fortunately in wind channel experi- 
ments on many low resistance forms the change in the coefficient becomes less 
rapid as the wind speed is raised, and it is possible that a comparatively small 
increase in the speed, though not affording a value of vl/v materially nearer full 
scale, would carry the results to a region where the coefficient was_ sensibly 


The remaining portion of this note will be devoted to a consideration of 
experiments carried out at the N.P.L. on a series of ‘* streamline *’ bodies. 

Resistance Experiments on Streamline Bodies. 

Experiments were carried out at N.P.L.* in a seven-foot wind channel on a 
composite model having a maximum diameter of six inches. There were four 
alternative forms of head, five forms of tail, and a numbey of cylindrical portions 
which could be inserted between the head and the tail. Measurements were also 
made on a model of the envelope of airship ‘* S.S. 60,000,’’ and on four models 
sent by the Admiralty, one of which proved to have a lower resistance coefficient 
on a unit volume basis than any other model examined at N.P.L. In this connec- 
tion it may be noted that a ‘* one-piece ’’ model is likely to have a lower resistance 
than a composite model of similar form due to slight unavoidable irregularities in 
the surface of the latter. The composite model is, however, believed to be satis- 
factory for indicating the general lines upon which low resistance bodies should 
be designed. 

These experiments were carried out in connection with the design of airship 
hulls, and the results expressed in terms of a coefficient C for comparison on an 
equal volume (i.¢., equal lift) basis. The coefficient is caiculated frora the fol- 
lowing formula :— 

Ute F /pv?l? 

* Report of the Advisory Committee for Aeronautics R. & M. 607—‘‘ The Effect of Form on 

500 THE AERONAUTICAL JOURNAL [September, 1920 

where F is the force on the model, v the wind speed, | the cube root of the volume 
and p the density of the fluid. It should be borne in mind that the body giving 
the lowest value of ( will not form the best shape for fairing a dise of the same 
maximum diameter in normal presentation; in this case a much shorter form will 
give the optimum result. Throughout the present note resistance on a volume 
basis only will be considered. 

Fig. 1 depicts four of the more important combinations of the composite 
model and **S.S. 60,000°’ for purposes of comparison. By an oversight the 
curve of head H.33 at the after end was not made parallel to the axis, and H.33b 
was designed to provide a somewhat similar head free from that defect. In 
designing a head it was discovered that an ellipse gave a form approximating to 
those of the heads illustrated. The tail T.S.L.M. consists of a series of truncated 
cones. The form of U.721, which was found to have such a low resistance co- 
efficient, approximates to that of the combination H.33b, T.E,; but with a some- 
what reduced fineness ratio. 

Before considering the numerical values it may be noted that the results 
show very clearly the important errors which may arise if experiments are carried 
out on small models at low speeds. For many combinations of head and tail (for 
instance H.33b, T.E,) the effect of introducing cylindrical body was at low values 
of vl negligible, and in some cases resulted in a decrease of resistance; at the 
higher values of vl the introduction of cylindrical body of a length greater than 
one half the diameter caused an increase of resistance which grew rapidly with 
the length of the cylinder. 

A typical set of curves is reproduced in Fig. 2 in which values of the resist- 
ance coefficient are plotted on a vl base. Head H.33b was used throughout these 
experiments, and each group of curves corresponds with a particular form of tail. 
Each curve represents the results on a model with a particular length of evlin- 
drical body varying from zero (C,) to a length equal to twice the diameter (C,). 
With certain heads (viz., H.33a) evlindrical portions as much as five diameters 
long were inserted; the resistance coefficient increased continuously, but at a 
less rapid rate, for the longer cylindrical portions. A curve in the lower group 
marked ** L.33°’ represents the results for the ‘‘ one piece ’’ model, containing 
approximately 1? diameters of cylindrical body, tested in R. and M. 541.* It will 
be noticed that these results do not fall in their proper sequence, the probable 
reason being that the slight irregularities present in composite models increase 
their resistance. The present experiments indicate that a form of appreciably 
lower resistance than L.33 (which is a model of the hull of the German airship of 
that name) could be produced by removal of the cylindrical body. 

In Fig. 3 the results are plotted on a base of fineness ratio, and for each 

group of curves the value of vl is constant. 

The relative lateral displacement of the curves in this figure is caused by thes 
different lengths of head and tail for the various models. On each curve the 
point at the lowest fineness ratio corresponds with the combination of head and 
tail alone, and increase of fineness ratio is effected by the introduction of cylin- 
drical body. At a first examination the curves at low values of vl appear entirely 
erratic; but on close inspection a process of evolution may be traced as the value 
vl is increased. This change is particularly marked in the curves for the model 
with tail T.33. The change in the curve for the model with tail T.S.L.M. from 
convex upwards to concave is also interesting. 

The increase of resistance due to introduction of cylindrical body is shown 
more clearly in these figures. The model with tail T.33 actually shows a decrease 
for certain increases in the length of cylindrical body ; but this effect diminishes as 
the value of vl is increased until it has disappeared at vl = 60. 

* Report of the Advisory Committee for Aeronautics. 




September, 1920] 

a) | 

~~ WIS 1 


502 THE AERONAUTICAL JOURNAL [September, 1920 


HEAD H33b 

Resistance Coefficient C. 

FIG. 2. 

September, 1920] 



Head _H 33b. 

w(x =SLM oHE£.5 
Vl = 15. Wh Tas =ES3A je vVl= 40 

0-015 + 6 7 8 0-018, — if meas =: aeaet I aes 
& ~~ G } | 
ON nt eg 0-0/2 - Sa : 
3 eh FES LES 8 | | 
S00) |--k— Slo-01 —— — — 
& T,S..m*x “ 3 | 
§ 0-010 : — 0-010 - , 
Sy | ® EL 
&| 0-009 |} \! }—- 8! 0.009 —- 
: al Ez: : 
3] 0-008 }—_—__;____ et —__+— 0-008} 
oe | | € 

0-007 |---~}- --— | {+ 0-007 

o-006|____ ¥ Be 0-006, 1 = ee 

ne, Ve % i. eee 
Fineness Ratio (Lengthy, da) Fineness Ratio (49 Vina eia,) 

oosf——§¢—T- #-§ vos 
: | & L 
Gp 9-0/2 |--——— = sie Vjo-o2 
R | § 
S| 0-07 | = 9] 0-017 
& . s 
< 8 
& 0-0/0 | 0-0/0 
: : 
8| 0-009 §|°009 
S a | 
4S KY 
“~~ + — — | 533 | ———— & 0-008 ;— 
€ 0-007|-- — | 1. —_—---+—____+ 0-007 —+ 
| | | | 
0-006 5 } ay aa ae — 0006 eat “ystgr he  ——e 
Fineness Ratio (4 er, ao ) Fineness Ratio Ratio —_ dia) 
0.0137 & 8 2 
Ul 6. ee Se eee 
n 0-012 | | v 
s | 8 
be < 
8] 0-010 |—— resis i ~~ % 
© | * | Oo 
£| oot 8. oS a : — 
3 | SS | 
i i] H 
3 { + jer Nb oc ae 3 
| 2-008; i | 2 
g | 7,£3 TV 3a | é 
0-007;—_- — T Sonal | — mae 
| | ] 
L | | = oa 
a On ee eee ole ee 
Fineness Ratio (2°96 04 cha) Fineness Katio (Corie dia) 

FIG. 3. 


As regards the numerical value of the coefficient C it will be noticed that in 
Fig. 2 the lowest value (with tail T.E,) is 0.0075 and the curve appears to be 
approximately horizontal. For model S.S. {see Fig. 1) at a similar value of vl 
the coefficient is 0.0132 and at the highest value of vl 0.0125. The results there- 
fore show that a considerable reduction of resistance is possible over that of a 
model of the S.S. form; but it should be observed that the coefficient is falling 
for S.S. at the highest speed attained, while the value for H.33b, T.E, is constant 
for the higher speeds, and the tendency for greater values of v is unknown. The 
minimum resistance coefficient for the *‘ one piece ’’ model U.721 (similar to 
H.33b, T.E,) is 0.0068. : 

The main conclusion arrived at in these experiments may be briefly sum- 
marised as follows :— 

(1) A tail 2.5 diameters long can be made to give as low a resistance co- 
efficient as any tail vet examined. 

(2) A form 4.6 diameters long can be produced which will give a resistance 
coefficient of 0.007, which compares with 0.013 for ‘‘ S.S.’’ 5.2 diameters long. 
The forward curved portion of such a body may be elliptical, and must be at least 
2 diameters long. There should be no cylindrical portion. 

(3) For all the models examined, except certain ones with T.S.L.M., the 
introduction of cylindrical body causes an increase of resistance coefficient at the 
higher values of vl; with T.S.L.M. the introduction of more than one diameter 
also causes an increase. 

(4) In forms which include more than one diameter of cylindrical body, a 
tail 2.75 diameters long, composed of a series of truncated cones, may be used 
without appreciable increase in the resistance coefficient at the high values of vl. 

In conclusion, it may be stated that at present there appears to be no reliable 
method by which the resistance of a streamline body may be calculated. It has 
often been suggested that the skin friction resistance may be calculated from a 
coefficient determined on a plane surface; but the skin friction coefficient applicable 
to streamline body varies markedly for bodies of similar surface roughness, but 
of different form. 

A calculation of the resistance coefficient of model U.721, using a skin friction 
coefficient recently obtained at N.P.L. in a preliminary experiment on plate glass, 
resulted in a value of the order of twice the measured resistance. Recent experi- 
ments at N.P.L. on low resistance forms* indicate that the force obtained by 
integrating the normal pressures over a model in the wind channel is zero, to 
the order of accuracy of the experiments, so that in those cases the resistance may 
be regarded as due entirely to the tangential force. 

* Report of the Advisory Committee for Aeronautics. R. & M. 600. ‘‘ Distribution of Pressure over 
the surface of Airship Model U.721.’’ 

September, 1920] THE AERONAUTICAL JOURNAL 505 




The inclusion of incidence wires in strength calculations, by the method of 
strain energy, makes very little difference in normal flight conditions, but shows 
large changes in wing loads in nose diving and in centre section loads with a 
broken flying wire. There is a serious increase of load in the bottom plane in 
nose diving. 

An aeroplane is a braced structure with many redundant members, which are 
usually ignored in strength calculations. This is a simple way of getting a first 
approximation as a basis of design and gives a fair idea of the worst loads likely 
to occur in most of the members. 

It is quite easy to go on to a second approximation, taking into account 
some, at least, of the redundant members by the method of strain energy. This 
is by no means accurate, even in theory; and owing to uncertainty of physical 
data and lack of time for arithmetic it has to be limited by many approximate 
assumptions. A complete treatment would include the energy due to end load, 
bending and shear of every member and fitting from wing tip to elevator, with 
due allowance for initial tensions and subsequent slackening due to settling down 
of the fittings. In any practicable form most of this has to be left out. In the 
simplest form of all, the only one considered in this paper, we deal with the one 
most important item in the strain energy, that due to tension in the wires. 

Further, we break up the aeroplane into convenient units, by supposing the 
centre section of the fuselage to be rigid. In certain cases we can adopt the 
bay-to-bay method proposed by Mr. Case (‘‘ Aeronautics,’? December, 1918), 
treating each bay in turn as if rigidly held to the rest. This is useful and saves 
much labour in the conditions of normal flight and nose diving, but it cannot be 
applied to the case of a broken flying wire. 

Thus we are calculating the stresses, not in a given aeroplane, but in an 
imaginary craft with perfectly elastic wires, perfectly rigid spars, struts and 
fittings, perfectly trued up. Stress methods which neglect strain energy make 
all these assumptions in obtaining the end loads, with the additional one that all 
redundant wires are slack all the time; so we are going one step, if a small one, 
nearer the truth, and we can go as many more steps as data and time permit. 

Much that is omitted here mav prove of importance and more work is wanted 
on all sides of the question. In a composite structure, one chief source of error is 
the settling down of metal fittings on wooden members. This is equivalent to 
an initial slackness in the wire, and if this can be measured, it is simple to allow 
for it. With all-metal construction the effect is probably less important, and 
the method of strain energy will be the more useful. The general effect of the 
movement of fittings is to equalise stress and to increase rather than decrease the 
loads which are at present neglected. 

The general result of strain energy calculations is to confirm the simpler 
methods in most, but not all, essential features. Where the two differ much, the 
effect of the incidence wires is almost always to relieve loads, and an appreciable 
saving of structure weight is indicated. But there are members in which the 
load is increased and these should be watched. The loads found cannot be 
accurate and may even be as wide of the mark as the first approximation; but 
at any rate they show in which sense a correction is needed and much may be 
learnt from the comparison. 

506 THE .AERONAUTICAL JOURNAL [September, 1920 

The method of calculation for redundant wires is simple and straightforward 
and not nearly so long as is often supposed. Much of it can be done graphically 
and the rest is arithmetic and algebra to simple equations. There is one physical 
principle involved, namely, that for strains within the elastic limits, the loads in 
the redundant members take up such values as make the total strain energy a 
minimum. There is one mathematical theorem used, namely, that the conditions 
for a minimum are the vanishing of all the partial differential coefficients of first 
order. The proofs are to be found in the text books. 

The simplest possible redundant system consists of a weight hung by duplicate 
wires; each takes half the load. If the wires are not of equal strength the load 
in each is proportional to the product of the sectional area and modulus of elas- 
ticity. In more complicated systems the redundant wires tend to equalise loads, 
with due regard to strength and length; that is to say, they come into play so as 
to reduce the higher loads, especially in the weaker or longer members, at the 
expense of increasing some smaller loads. If we are finding the load T in a 
particular redundant wire, then everv wire whose load is relieved by the existence 
of T is an argument for an increase of T, and every wire whose load is increased 
is an argument for the decrease of T; the redundant member itself always belongs 
to the second class. The value of T adjusts itself so that, for an infinitesimal 
increase of T, the resulting increases and decreases of the strain energy of the 
two classes of members balance exactly. 

The strain energy of a wire of length 1, uniform sectional area A, modulus 
of elasticity EF and tension P is }(l/EA) P?. We need to tabulate 1/EA and P 

for each wire. The total strain energy is then 
4 (l/EA) P?. 

We are limiting the sum to the flying, drag and incidence wires of both wings, 
including the wires attaching the planes to the fuselage. 

Usually all the P’s are unknown; the independent variables are the loads 
in the redundant wires. The choice of certain wires as redundant is arbitrary; 
flight is possible with an incidence wire slack, but so it is with a flying wire 
broken. We take the incidence wires in the wings, and the flying wires, if any, 
in the centre section, as redundant, and denote the load in one of these by T. 
If there are » redundancies, there are n independent variables T; all the other 
loads can be written down in terms of the 7T’s by means of the conditions of 
equilibrium at the nodes. The tension P in any wire is the sum of a constant 
term and constant multiples of the T’s, say, 

ot a n-ne | 

Here P, is the load induced by the external forces, when all the redundant wires 
are slack; P, is the load in the same wire that would be induced by unit load in 
the first redundant wire when the external forces are absent and all the other redun- 
dant wires are slack; and so on for the other coefficients. All these can be found 
by drawing-board methods. 

It is more convenient to express each load as a multiple of the length of the 

wire in which it acts, say, 
Pang, 7, @ 61... 
This often saves calculating a different set of coefficients for the two crossed 
wires of an incidence panel, which are of different lengths if the planes are 
staggered. Relations similar to (1) hold between these load coefficients, say, 
o=—_,4+ pt, +... pd. : : ; : (2) 

The expression for the strain energy becomes 

e 2 (8 /2E A) (Po + pt, see ete 5) La 

September, 1920] THE AERONAUTICAL JOURNAL 507 

wheret, . . . t, are the unknown load coefficients and all the other letters stand 
for known numerical constants. Unless there are variations of stagger or gap, 
most of the p;’s have one of the values 0, 1, — 1. All the equations of equilibrium 
at the nodes are satisfied by virtue of (1) or (2), and the variables t are independent. 
The conditions that { is a minimum are that each of its partial differential 
coefficients, with regard to t, . . . t, vanishes :— 

= (12 /EA) p, (Po + Pit, +... + Ptr) = 0, 
& (1? /EA) Pa (Po + Pit, +... + Pata) = 0, 
Here are n equations for the n unknowns t, from which all the loads can be found. 

The writer has, at different times, made calculations of this sort for several 
aeroplanes and parts, and seen others’ results. The effect of various dimensions, 
particularly of stagger, is too great to allow of any definite numerical results; 
but the common general features are very marked. All that follows refers to 
single engine biplanes only; for triplanes the results are likely to be still more 

In low speed flight, with the centre of pressure forward, the incidence wires 
transfer load from the more highly stressed front truss to the rear. But long 
before the truss loads are equalised the rate of increase of strain energy in the 
drag and incidence panels overcomes the rate of decrease due to equalising the 
flying wire loads, and only a small fraction of the difference of lift is carried 
across. All the higher stresses are somewhat relieved. 

Much the same holds for high speed flight, with the centre of pressure near 
the middle of the chord. These are very satisfactory confirmations of current 
methods. The same conclusions were drawn by Mr. Pippard from a series of 
destruction tests, with and without the incidence wires in position. The details 
were issued by the Admiralty in C.I.M. No. 20, February, 1917. 

In a nose dive, though the stresses are very important, the calculations are 
less so, owing to the uncertainty of the distribution of air pressure, which probably 
changes rapidly with a small change of attitude. In this condition the unit 
stresses are much nearer to the elastic limits than in normal flight and the bedding 
of fittings becomes of more importance. The incidence wires relieve both the 
main trusses, so there is a strong argument for expecting a considerable load, 
and to neglect it is to ask for a heavier front anti-flving wire than is really 
necessary to meet the nose diving tension. In the bottom drag bracing there 
is a decided increase of tension, due to the back stagger of the incidence wires, 
and even more important is the increased load this throws on the bottom rear 
spar. Fortunately most aircraft have more than their required factor in the 
bottom plane in nose diving, but the point requires attention and might become 
a source of danger. 

It is in the less important case of a broken flying wire that strain energy 
methods have the greatest advantage over others. One incidence wire ceases to 
be redundant and its load is determinate, or nearly so, and determines all the 
loads in the adjacent bay, which are the most important wing loads for this 
condition and are not affected by the redundancies. But another effect of the 
broken wire is to cause a discrepancy of end load between corresponding spars 
of the two wings. If we assume that all the redundant incidence wires are idle, 
this discrepancy has to be balanced entirely by large loads in the centre section 
drag and flying wires. Now there may be six or eight wing drag wires helping, 
which reduce the centre section loads to a small fraction of their first estimate. 
This requires tensions in the redundant wires and modifications of the other loads, 
especially in the undamaged wing; but these are unimportant. For the centre 
section in this condition the method of strain energy gives us a reasonable instead 
of an unreasonable estimate. 


EXAMPLE.—In order to bring out these points without burdening the reader 
with masses of detail, the method will now be applied to a wing structure sim- 

plified to the last degree. Though the numerical results are worthless, from a 
practical point of view, the general features are very similar to those of an actual 

Assume all the wires of the same length, size and material. Imagine a 
cubical bay on each side of a cubical centre section, the whole held rigidly at the 
base of the centre section, this square representing the fuselage. The gap is 
1oft. and the total load 1,200lbs. 

To suggest low speed flight, let equal lift forces of r1oolbs. act at all the 
nodes of the front truss and half these lifts at the rear nodes; neglect drag. 
Without the incidence wires the front flying wire load is 283lbs. and its length 
is 14.1ft., the load coefficient p is 20, and for the rear flving wire it is 10. These 
are the only wires in play. 

Now introduce a tension ¢l in each outer incidence wire from top front to 
bottom rear. This relieves the tension in the front flying wire, increases the rear 
load, and calls into play the top drag, bottom anti-drag and centre incidence wires 
on each side, which now become necessary for flight. We suppose the wires 
crossing these to slacken and the other centre section wires to be absent (see 
Fig. 1). The strain energy for one side, which is half that of the whole wing 
structure, is & (l*/2E A) p*, a sum of six terms, where p has the following values :— 

Wire. p. dp /dt. 
Outer incidence T.F. to B.R. ... t I 
Front flying si re as 20—t —I 
Rear flying eis es ape 10 + t I 
Top drag ... Spe a ce t I 
Bottom anti-drag o ae t I 
Centre incidence T.R. to B.F. ... t I 

The equation for minimum strain energy is 
> (l3/ EA) p (dp dt) = O, 
where |*/EA is a constant factor for all six terms and can be dropped. Writing 
the sum in full, we have :— 
t—(z0o—t) + (10 + t) + 3t =o 
= 6 — 10 
i= 127. 

The front flying wire is relieved of 1/12 of its load, and the front and rear 
loads, instead of being in the ratio 2:1, are as 1.57: 1, being equalised to this 
extent only. The loads themselves are shown in the diagram. 

With the centre of pressure back, exchanging the front and rear lifts, we get 
the same tension in the other incidence wire. 

Now suppose that the flving wires are twice as strong as any others, then 
1/A is no longer a constant factor and must be retained. The effect is to halve 
the terms arising from the flying wires, and the equation for t becomes :— 

1 (2t— 10) + gt =o, t=1, 
and the lift transferred is considerably less than before. The stronger wires 
can take the same load with less storing of energy, and there is less to be saved 
by tending to equalise their loads. 

Next, instead of varying the strength of the flying wires, vary the incidence 
wire itself. If we put in a weaker wire it takes less load, but the load does not 
drop nearly as fast as the strength. Let the area of the outer incidence wire be 
ax that of any other; then one term is multiplied by 1/a, and 

t(5 + 1/a) = 10. 

September, 1920] THE AERONAUTICAL JOURNAL 509 

_ Ifa =1, then ¢ = 1.67, as we had just now; but if a = .5, t = 1.43, a wire 
of half the strength takes 85 per cent. of the load; and if a = .1, t = .67, a wire 

of one tenth the strength still tries to take more than a third of the load. The 
fact that a wire is redundant is no guarantee that it will not get more load than 
it can stand. It would be wrong to suppose that, when a load divides itself 
between the wires, the ratio depends only upon their sizes and angles; it may 
depend even more on members some way off. In the case considered, the size 
of the centre incidence wire has exactly as much influence on t as that of the outer 
wire to which ¢ actually belongs. 

200 200 100 100 
\ 4 ills A K 
<5 i 
%. | j Ry 
“aR WN 
*S Ny 
Sp r x 
. TT 
t R F R 

%. VY 

& Y 

z KX Ky, 
“x RS 
. ~ bi 
: _\ 
FIG, F:; 

Now suppose a forward stagger equal to the gap. ‘The wires are of different 
lengths, and we retain 1°, which has a marked effect. The drag bracing in each 
plane carries stagger loads, and the centre ingidence wire is in play from top rear 
to bottom front independently of the redundancy. The parts of the load co- 
efficients which depend on t are the same as before, though this is not true of 
the loads themselves. 

Wire. lL. p. lp (dp /dt). 
Outer incidence T,F. to B.R. 22.4 t 11,180 t 
Front flying a 1763 20—t 5,200 (t — 20) 
Rear flying 29.3 io +t 5,200 (t + 10) 
Top drag 14.1 is+t 2,830 (t + 15) 
Bottom drag 3 see is eet 15 —t 2,830 (t — 15) 
Centre incidence T.R. to B.F. ... 10 30 +t 1,000 (t + 30) 

> = 28,240 t — 22,000. 
From this we get t = .78, very much less than before. 

o10 THE AERONAUTICAL JOURNAL [September, 1920 

With the centre of pressure back, the opposite incidence wire is in play in 
the outer panel; as its length is different, the equation is altered, but the load 
coefficients are as before except for signs. The centre incidence is not turned 
over, but its stagger load is relieved instead of aggravated. We find t = 4.5, 
nearly six times its value with the centre of pressure forward. The ratio of front 
and rear flying wire loads is 1.07, and the equalisation is nearly complete. A 
stagger of 45° is excessive, but a moderate amount still has a great effect. It is 
this which prevents any generalisation as to the fraction of the difference in lift 
which we may expect to be transferred. 

Pass to the nose diving case for the unstaggered wings. There are up loads 
on the rear truss, down loads on the front truss and considerable drag loads; 
we take them to be 3o00lbs., 20o0lbs., 1oolbs., at the respective struts. The outer 
redundant wire comes into play, from top rear to bottom front. It will probably 
turn over the top drag wire, but we do not assume this as vet. 

Estimates of Load. 

Wire. p. p (dp/dt). First. Second. 
Outer incidence T.R. to B.F. t t oO 165 
Front anti-flving ... pe .. 20—t t— 20 283 117 
Rear flying me sig .. 30—t t — 30 425 259 
Top drag ... ies ie .. 1o—t t— 10 141 — 24 
Bottom drag sic soe - £O-+¢ t+ 10 14! 307 
Centre incidence T.R. to B.F. ..... 2o—t t — 20 283 117 
> = 6t— 70, t = 11.7. 

Here is a negative tension in the top drag, showing that we have made the 
wrong assumption. We should have had the anti-drag wire, with 
p = t—u1o0, dp/dt = +1, p(dp/dt) = t—1o, 

giving exactly the same term in the equation for t. If the crossed wires had been 
of different sizes, or if we had been including the energy of the spars, there 
would have been more changes and some correction to ft. 

The last two columns in the table give the loads as estimated without and 
with the effect of the incidence wire. The cruder method gives all the loads 
much too high except two—the incidence wire itself and the bottom drag. The 
latter is important as the revised load is more than double the first approximation 
and there is an added compression of 117lbs. in the bottom rear spar. 

A forward stagger equal to the gap, which shortens the incidence wire in 
play, increases its load coefficient to 16, and the lift transferred and the effect 
on the other wires are increased in the same ratio; the actual load in the redun- 
dant wire is hardly 2ltered. An equal back stagger reduces the load coefficient 
to 9.7. 

Suppose the front flying wire broken on the left wing when the centre of 
pressure is forward. The condition is unsymmetrical and there are different load 
coefficients in the two outer incidence wires, sav t on the damaged side and ft’ on 
the undamaged side, which must be considered as well. On the damaged side 
t is not unknown, but is given by the fresh condition, that the broken wire takes 
no load, t—20=o0. This incidence wire is not redundant, we have one un- 
known ¢t’; and we have first to decide which wire to assume in play on the 
undamaged side. 

Due to normal lift, there is an inward thrust of 20olbs. on the top front spar 
centre ‘section, produced by the flying wire. Due to the incidence wire, this is 
relieved by 10 t, and there is also a component pull rot from the drag wire. 
The thrust on the centre section becomes 200 — 20 t; with t = 20, this is negative 
and represents a pull. On the other side we have a thrust 200 — 20t’; the 
resultant is an unbalanced end load 20(f—t’), tending to move the top front 

September, 1920] THE AERONAUTICAL JOURNAL 511 

spar to the left, taken up by the centre drag wire with a load coefficient 2 (t — t’), 
if we assume for the present that there are no centre section flying wires. The 
system is completed by the centre incidence wires, both assumed in play from 
top rear to bottom front, the load coefficients being 3t—2t’ on the damaged 
side and 3 t’— 2¢t on the other. 

If we have t = 20, t’ = o, these three centre section coefficients are 40, 60, 40, 
greater than for any wing wire. We assume that the redundant wire relieves 
these heavy loads, and that t’ has the same sign as t, quite apart from the question 
whether this increases or decreases the strain energy of the undamaged wing by 

Damaged side. Undamaged side. 
Wire. p. dp/dt’. p. dp /dt’. 

Outer incidence T.F. to B.R. ... t O t’ I 
Front flying ee oe .. 20—¢t oO 20— t’ I 
Rear flying a za we 10+ O 10 + ¢’ I 
Top drag . : ae nu t oO C I 
Bottom anti- drag a t oO t’ I 
Centre drag oe Jn bore —2 _ —_ 
Centre incidence T.R. to B.F. ... 3t—2t) —e2 3t'—a2t 3 

Now t = 20 = = constant, and we only differentiate with regard to the unknown 
t’. The damaged wing contributes nothing to the equation, the undamaged wing 
5t’—10, and the three centre section wires 17 t/— 320, being much more 
influential than all the other ten wires. Altogether, 
aa — 490 = 9, of = 16, 
The centre load coefficients are reduced from 
; , , 
40, 60, — 40 with t’ = o, 
to 10, 30, — 5§ with t’ = 15. 
Now introduce the centre section flying wires. One of these is redundant ; 
let its load be ul. The expressions for the wing load coefficients are unaltered, 
but the centre section terms become :— 

Wire. p. bp /at’. bp /du. 
Top drag sp te we 2t—2—U —2 —I 
Front flying sat re ve u oO I 
Rear flying te — se . oO I 
Incidence, damaged side 3t— —u —2 —I 
Incidence, undamaged side... 3 t’/— ae +u 3 I 

Putting t = 20, and differentiating partially with regard to t’ and u, we get 
the simultaneous equations :— 


peo ta eat 30 =0 if 11 
, 33 \ whence; 
7t' + 5u—140 = oO) i = 62.6 

and the centre drag coefficient = 2t— 2t’/—u = 5.4. 

The old assumption used to be that the centre drag and flying wires shared 
the unbalanced end load equally. In our case the flying wire takes about 2} x the 
load in the drag wire, on the assumption of equal sizes. Even if the drag wire 
were infinitely strong, the ratio would still be about 1}. 

Now pass to a two-bay aeroplane, still composed of cubes. In a nose dive 
(see Fig. 2) there are two redundant incidence wires each side in play from top 
rear to bottom front, with load coefficients, say ¢ in the outer and w in the inter- 
mediate panel. The other coefficients are as follows, where we write s = t + u, 

512 THE AERONAUTICAL JOURNAL (September, 1920 

which is proportional to the total lift transferred by the two bays together. This 
simple change of variable saves much labour :— 

Load Coefficients. 

Wire. Outer bay. Inner bay. Centre section. 
Incidence mt Bie t s—t 30-—s 
Front anti-flying .. 20—€ 40 — 8 — 
Rear flying... .. 30—¢ 60 — 8 — 
Top anti-drag ... .. t—10 8 — 20 — 
Bottom drag ... .- Ott 20 + 8 — 
200 200 200 300 300 300 
Ny A X . K 
iS) i 9 Y 
2 ¢ 93 x yy 
Ny N Y b Y 
RS SS y y 
F R F R F R 
AY gy S 
yy, Y N 
y Y > 
N Ww 
R y R 
Oy, ie ia sy 
9 o S x 
wy Roy wy Ry, 
x x 
i yy aS, »y 
ii; aN IK N F 
100 100 100 100 100 100 
FIG. 2. 

Differentiating the strain energy partially with regard to t and s, we find 

6t—s—s50=0| aoa 
—t+68s—130=0j' baa. ip 

As before, all the wires are much relieved except the bottom drag, whose 
loads are more than doubled. The load in the outer incidence is not very different 
from the one-bay case. 

With a wire cut (see Fig. 3) there are four different incidence loads connected 
by one relation, expressing that there is no load in the broken wire. The value 
of either t or s is known on one side. The unbalanced end load 1s 
20 (t — t’) + 20(s —s’), and we expect, and find, that the signs of the redun- 
dancies are always such as to reduce this. 

September, 1920] 

200 200 200 200 200 200 
A A A A A 
$ Xo, /2 
aa, S- | © 
7. RF RF R Ff RF RF R 
my, 2 
< 5 of 
LS) dy ote) n> » <0 
ay, % 
ty RY 
100 100 100 100 100 100 
A A A | A A A 
, Y. 5 
re x x 9 
2 5 x 
S ne 
R R 
x | 
ra e 6 2 
F t F 
is) R 
% 5 Na £0 
F Dd F 


FIG. 3. 

514 THE AERONAUTICAL JOURNAL (September, 1920 

If the front out wire is cut, t = 20, so u is negative and t’ wu’ positive. 
The load coefficients are given in the diagram. We form the simultaneous 

$2 /3s = o, 53/ dt’ = o, 83/88’ = 0, 

put ¢ = 20 and solve. The result is 

t = 20 f’ = 9.6% 
t= is vse 

u = — 18.8 cs 


Of the lift transferred from the damaged bay, nearly the whole is brought 
back at the next incidence panel. There is a great disturbance of load on the 
undamaged wing, but we need not pay attention to these members as no worse 
case arises. The striking thing is that, by the action of the three redundant 
incidence wires the load in the centre drag wire is reduced in the ratio 80: 7. 

With the front inner wire cut, s = 40, and t is unknown and also uw; we 
discard the equation 6X/6s = o and use instead 6%/ét = 0, the other two equa- 
tions standing. The solution is 

t = (—)3.2 t’ = 16.3 
s’ = 40 3’ = 18.3 
l= ~as2 i =z 

The sign of ¢t is the more remarkable, because in normal flight it is positive. 
Also, if we consider the wing by itself, a definite amount of lift has to be trans- 
ferred by the two incidence wires, and we should find that they shared it, each 
taking a definite amount greater than in normal flight. It is the claim to relief 
of the centre section that reverses the outer incidence load. The intermediate 
incidence wire carries slightly more than the whole lift on the truss up to that 

The next table compares the centre section loads for low speed flight with 
a broken flying wire, without and with the action of the redundant wires. It is 
clear that the first approximation does not deserve the name. 

Outer wire cut. Inner wire cut. 
Wire. Ist 2nd Ist 2nd 
estimate. estimate. estimate. estimate. 
Centre incidence— 
Damaged side... 1,410 120 1,700 625 
Undamaged side... 1,130 — 40 1,130 — 200 
Centre drag oe 1,130 100 1,130 60 

The next table compares the bay-to-bay method with the more elaborate 
calculations of the chief load coefficients for two bays. In a nose dive the errors 
in the two incidence loads tend to correct each other, but in the case of a cut wire 
the effects on the centre section are cumulative. 

Load coefficient. 

Condition. Bay. Wire. Bay to bay. Complete. 
Nose dive os Outer Front anti-flying 10 rf 
Inner = se 18 16 
Outer Bottom drag... 20 22 
Inner ws re xe 2 44 
F. outer wire cut Centre Top drag ee 43 44 
F. inner ,, = - a = is 67 4 

In conclusion, the following figures are a few of the results of calculation for 
an existing design of a two-bay seaplane with a forward stagger, and are quoted 
by permission. 

September, 1920] THE AERONAUTICAL JOURNAL 515 

Unit tension with 
redundant wires. 

Condition. Bay. Wire. Neglected. Included. 
Nose dive ei Outer Front anti-flying 1,160 450 
Inner x - 2,070 930 
Outer Rear flying... 1.730 970 
Inner 5 AA ch 2,920 770 
Outer Bottom drag ... 350 965 
Inner 3 Sl aie 530 1,400 
F. outer wire cut Centre Top drag we 3,470 380 
F. inner _,, - Re a i ws 2,960 210 


It is generally allowed that a bulkhead wire in a fuselage gets a considerable 
tension only if a wire in an adjacent panel is broken or absent. The bulkheads 
also act so as to distribute an unsymmetrically applied load. For example, if a 
horizontal rudder force comes wholly on to the top face the first few bulkheads 
transfer part to the lower face, but as this sets up stresses in the vertical sides, 
which absorb energy, less than half the load is transferred. 

These observed facts can be supported by calculation, the principle of mini- 
mum strain energy determining the tensions in the redundant bulkhead wires. 
The chief cause of inaccuracy as already remarked in this method is the settling 
down of fittings (especially with timber longerons), which is difficult to measure. 
If we assume the method to apply, it is easy to calculate the loads in any given 
fuselage. When the dimensions of the bays and sizes of the members follow 
regular laws we can write down general equations for each type of loading, and 
in some cases obtain a complete algebraic solution. The results of calculation 
are given below for the simplest case of a straight fuselage with square bays, 
all the wires being of the same length, size and material. The energy of the 
longerons and struts, the movement of the fittings and the initial tensions are 
neglected. The fuselage is held rigidly at one end and loaded at the other. 

1. Unsymmetrical Loading.—Let unequal vertical loads P, Q be applied at 
the rear struts. The solution depends on a recurring series, and each bulkhead 
load is strictly the sum of two terms, one of a descending and one of an ascending 
geometrical progression, but the second is negligible. The total load transferred, 
from the side with the heavier load P to the other, is practically one quarter of 
the difference of loads, as follows :— 

.207 (P—Q) at the end bulkhead, 
.0355 (P—Q) ,, ,, 2nd 93 
.oo61 (P—Q) ,, ,, 3rd i etc., 

forming a G.P. with common ratio .172. 

At the fixed end the longeron loads are nearly equal on the two sides. 

If Q = 0, the fuselage is loaded on one side only; three-quarters remain 
there and one-quarter travels across. 

If Q=—P, we have pure torsion; the load transferred is 4 P. Half the 
couple on the structure is transferred from the vertical to the horizontal faces. 
The transference often takes, place in the tail bracing, before the fuselage itself 
is reached ; then the bulkhead wires are idle. In any case the adjustment is prac- 
tically complete after a couple of bays. 

516 THE AERONAUTICAL JOURNAL [September, 1926 

If the wires in the vertical side panels are twice as strong as any others, 
the total load transferred drops to 4 (P-—Q), namely :— 

.132 (P—Q) at the end bulkhead, 
.0275 (P—Q) ,, ,, 2nd ms 
.0057 (P—Q) ,, ,, 3rd 5 etc., 

the common ratio of the G.P. being now .21. 

2. Broken Wire.—If the applied load is symmetrical, P = Q, and if the 
wire in one vertical face is broken, the whole of P has to be transferred. by the 
bulkheads between: the tail and the break. If all the wires are of equal strength, 
they share it as follows :— 

.828P at the bulkhead next to the break, 
ae a5 ye . next behind, 
-024 sin - cs ‘5 Pic., 

the common ratio .172 as before. 
By far the greatest part of the load is carried by the nearest wire. 

If the break is in the bay next the tail, the end bulkhead has to take the 
whole load; if there are two bulkheads available, they take 5/6 P and § P 
respectively. Between the break and the fixed end practically the whole load 
travels back by the same instalments as it came, if there are at least three bays. 
Even with only one opportunity, 4,5 P returns; the effect of the break is entirely 
local, and the undamaged side does not carry the double load all the rest of 
the way. 

If the bulkhead wire is lighter than in the sides it takes still less load, and 
the effect is felt further along the fuselage. It is found that the consistent 
strength for the bulkhead bracing, still on the assumption of cubical bays, is 
exactly one-third that of the adjacent side wire, except for the two end panels. 
The effect of taper is marked. If it is greater in elevation than in plan, the 
rear bulkhead wires are at a worse angle for vertical loads, and, compared 
with a straight fuselage, the transference is thrown slightly forward, with the 
reverse effect for horizontal forces. A straight tapered square fuselage of 
geometrically similar bays would give almost the same results as the untapered 

An actual fuselage has not much resemblance to a row of cubes,’and seldom 
follows a regular law. This note is offered as a curiosity of some interest, 
showing the considerable agreement between elaborate and rule of thumb methods 
of treating bulkhead loads. 

(The effect of incidence bracing, and redundant bracing generally, is one 
of considerable importance in aeroplane structures, and if any reader has any 
figures based on practical experience which he may care to have published, 
and which would prove of value to designers, the pages of the JOURNAL are 
open to him.—EpirTor.) 

- GRR 

September, 1920] THE AERONAUTICAL JOURNAL 517 


BY MAJOR G. W. C. KAYE, O.B.E., M.A., D.SC., R.A.F., 


1... TESTS. 

During the war a heavy demand was made on the supplies of aeronautical 
timber, and in order to allow of the use of smaller sized timber, it became necessary 
to permit the introduction of splices into aeronautical construction. We knew of no 
means to obtain accurate comparative data, and the purpose of the present inquiry 
(with which Captain K. Robertson was also associated) was that of finding a 
method of splicing timber beams which should prove to be the most effective for 
withstanding bending moment (as in aeroplane main plane spars) and which should 
at the same time be simple to produce and easy to inspect. 

As a preliminary step we invited suggestions from aircraft constructors and 
others as to the forms of splices which, in their opinion, were likely to give the 
most useful results. A large amount of helpful information was thus obtained. 
In some cases actual examples were submitted, and from these and others expressly 
devised at the time, a number of types were selected. Where only sketches or 
drawings were available, full-size spars were made to these selected designs. 

All the spliced test-spars which were specially made were constructed to a 
standard section. The size adopted for the test beam was 5 feet long by 3 inches 
deep, by 1} inches broad. The length selected was determined by the capacity 
of the testing machine. The depth and breadth were chosen as the average size 
of the spars used on a number of aeroplanes of recent design. The best cake glue 
was used for all the specimens (except where otherwise stated), and precautions 
were taken to ensure good gluing and clamping. <All the spars were rectangular 
in section and not spindled to ‘‘1’’ section. (See, however, Example No. 26 in 
the table below.) 

The chief objects of the tests were (a) to ascertain the breaking load of ‘the 
splice ; and (b) to observe critically the behaviour of the splice under progressive 
localised loading, having regard to both the design of the component parts and 
the degree of interlocking of the glued surfaces. In other words, to ascertain by 
careful scrutiny during increasing loading which particular feature of the design 
was first responsible for the ultimate collapse. : 

With this in mind, Young’s modulus became of secondary importance ; and 
in those cases where deflection-readings are given in the following tables, they 
are not inserted with the object of demonstrating the stiffness-value of the particu- 
lar splice, but rather to show the range of distortion that the splice permitted 
while still maintaining a degree of structural intactness. 

The experiments were carried out during 1917. The test-pieces were placed 
upon supports 4 feet 6 inches apart centre to centre. The load was applied from 
above, at the centre of the length of the beam, by means of a dual plunger, having 
centres two inches apart. Each point of contact with the beam, both above and 
below, was made through self-adjusting shoes, fitted into semi-cylindrical bearings. 
Automatic seating against the faces of the beam was thus obtained. The loading 
was applied at an approximately uniform rate of 3,ooolbs. per minute. 

518 THE AERONAUTICAL JOURNAL [September, 1920 

Particulars are given below of the tests on 25 types of splices, which have 
been selected as typical of the many types tested. The breaking loads shown for 
each type represent the average of the results on several specimens. 

The average breaking load for a solid unspliced spar of silver spruce, made 
to the above dimensions, is shown opposite the first diagram (No. O). For ease 
of comparison, in addition to the column giving the actual breaking load in each 
case, a column ts added giving the comparative percentage value of each splice 
as compared with the strength of the standard unspliced beam, which is taken 
as 100. : 

Such of the beams as were not of the standard section or length, have had their 
results reduced, so far as is possible, to the standard dimensions. 

In General. 

In attempting to estimate the comparative values of the several types of 
splices, it is recognised that the specimens employed in the present experiments 
were probably more skilfully made than could be expected from average workshop 
practice. This is specially the case with the splices of complicated design, which 
although they gave high results under test (see Nos. 13, 25, etc., in the accom- 
panying table of results), cannot be regarded as commercial propositions ; their 
involved character would make production costly and inspection difficult. 

The following points * arise out of the present and other tests :— 

(1) The importance, from the point of view of good gluing, of ‘* blending ”’ 
the end-grain with as large a proportion as possible of ‘‘ side-grain ’’ in 
any splice, and so of reducing the proportion of end-grain to a practical 
minimum. This expansion of gluing surface can be achieved in two 
ways :— 

(a) By increasing the Jength of the splice. 
(b) By increasing the effective width of the face of the splice, e.g., by 
serrating, convoluting, etc. 

(2) The importance of avoiding splices containing abrupt steps or cross-cuts 
at right angles to the grain of the wood. 

(3) The importance of disturbing as little as possible the continuity of the 
wood fibres of the upper and lower faces especially when employing 
supplementary securing devices such as bolts, pegs, etc. 

Plain Sloped Splice. 

On a consensus of the above points, the uniformly reliable and high average 
results yielded by plain sloped splices of adequate length, impressed themselves 
early on our notice (see Nos. 22 and 23). Such splices have the additional 
advantage of ease of manufacture and inspection ; and it may be said at once that 
we were ultimately led to recommend this splice. 

The best disposition of the plane of the splice in this form of joint depends 
upon circumstances. In cases where the depth of the beam exceeds the breadth, 
the plane of the splice should be vertical. Where the breadth exceeds the depth, 
the splice may, with advantage, be ‘‘ horizontal.’’ If the depth and breadth are 
equal, the vertical splice is preferable. 

Length of Splice. 

With reference to 1 (a) the greater the length of a given type of splice, the 

* These remarks refer to splices designed to resist bending moment, and do not necessarily refer 
to splices which are to be subjected to pure endwise tension or compression. 

September, 1920] THE AERONAUTICAL JOURNAL 519 

greater its reliability and efficiency (see for example Splices 2 and 3, 4 and 22, and 
6 and 23). 

In the present tests, short splices were invariably found to give way at low 
loads, and usually quite suddenly without warning. ‘This is doubtless due to the 
fact that in splices of short length the glued-surface connection is largely confined 
to the end-grain of the wood. In making the joint in such cases glue is quickly 
absorbed into the end-grain, and is thus rendered non-effective. In point of fact 
(as in the case of No. 2), subsequent examination revealed that in some of the 
shorter splices the glue had practically disappeared, and a slight shock was 
sufficient in these cases to break the splice. 

The value of a long splice is especially demonstrated by the examples having 
a number of interlocking wedge-shaped ** fingers’’ or tongues (see Nos. 20 and 
21). These may virtually be regarded as a variety of very long straight scarf, 
which permits from 7 to 10 times as much glued surface as in other splices of the 
same over-all length. Further, apart from gluing, the longitudinal serrations also 
afford valuable mechanical gripping. Unfortunately, until suitable machinery is 
available, difficulties of manufacture appear to militate against the adoption of 
this form of splice, which on test gave the highest results. 

Serrated Splice. 

With reference to 1 (b), a considerable amount of investigatory work has been 
done during the war in the United States and Canada resulting also in the ultimate 
choice of a plain vertical scarfed splice, but with the added improvement of corru- 
gating or serrating each of the gluing faces, the two sets of corrugations exactly 
meshing into each other. Fig. A shows the details. 

= tite, ee = 
= | Sa 
= So 4 ° = Px 
= ° ae. S —— 
Details of American 
Saw-tooth Splice 
fig A 

Fic. A. 

In addition to the actual mechanical interlocking (which is pronounced), the 
glued area is by this means more than doubled. The Americans have gone in 
strongly for this splice; and the more extensive use of precision woodworking 
machinery in that country has enabled a production of the highest efficiency to be 
obtained. If a similar condition of things obtained in this country we should have 
no hesitation in recommending this splice before all others. 

Splices Containing Abrupt Steps. 

With reference to (2) splices constructed with an abrupt step at right angles 
to the grain (equivalent at the point to a butt joint) were invariably found to be the 
starting point of rupture. These initial ruptures had the immediate effect of 
reducing the effective cross-section of the beam by the amount of the step. This 
weakness was particularly noticeable when the step occurred on a face of the 
beam, but particularly on the tensional side, as in Splices Nos. 1, 2, 3, 8, 14 and 16. 

520 THE AERONAUTICAL JOURNAL [September, 1920 

Steps which occur at the sides of spars (as in Nos. 9, 10, 13 and 25) were 
generally found to be less harmful; but it should be noted that the jointing faces 
in these examples are mostly set at a small angle or parallel to the side of the 
spar, thus affording a good area of side-grain for gluing. 

Bolts and Dowels. 

The advantage of pegs, dowels or bolts, when properly disposed, is abundantly 

displayed by the tests. They are indispensable in the case of the straight sloping 
scarf. The disposition of the pegs, dowels, etc., is very important. As already 

remarked, any vertical peg, bolt, wedge, or other piercing element that breaks 
the continuity of the top or bottom surfaces of the beam has a deleterious effect. 
This was noticeably the case in Nos. 15 and 22, and in several other examples 
not included in the diagrams. Transverse fractures invariably first arise in the 
vicinity of one or more of the pegs in the lower or tension surface of the beam. 
On the other hand, horizontal pegs or bolts placed approximately in the neutral 
axis of the beam interfere very little, if at all, with the strength. 


The importance and value of good taping was repeatedly demonstrated. This 
appears to arise not so much from the actual added strength afforded by the tape 
(though this is considerable), but because of its usefulness in preventing incipient 
separation of the glued joint near the ends of the splice and so of checking any 

tendency to abrupt fracture. The fact that in several instances taped spliced 
spars yielded higher results than the solid unspliced spar itself can fairly be 
attributed to the taping. This feature has also been observed in connection with 

similar experiments upon taped laminated spars. 


1. \ ‘straight ’’ splice, with the contact faces either flat or alternatively 
serrated and meshed, is recommended as the best form of scarf for splicing timber 
which is to be subjected to bending moment. 

2. If the depth of the timber part is equal to or exceeds the breadth, the 
plane of the splice should be vertical. If the depth is less than the breadth, the 
plane of the splice may, with advantage, be ‘* horizontal.”’ 

3. Consistent with the limitations of design, the splice should be as long as 
possible. A steeper slope than 1 in g is, in our opinion, not advisable. 

4. After gluing, the splice should be reinforced with bolts or dowels. When 
the plane of the splice is vertical the bolts should be staggered along the neutral 
axis of the spar. The spacing or ‘‘ pitch ’’ of the bolts should be approximately 
twice the thickness of the spar. The distance between either tip of the splice and 
the nearest bolt should not be less than the thickness of the spar. Adjacent bolts 
should be arranged on alternate sides of the neutral axis, but always within the 
middle third of the depth of the spar. The stagger of the bolts should be at least 
twice the diameter of the bolts. 


When the plane of the splice is Horizontal, the bolts will necessarily be vertical, 
and (with a view to severing as few of the flange-fibres as possible) should be 
arranged, without stagger, along the centre line of the spar. 

5: The splice should be well taped with not less than two layers of tape, well 
glued on. The taping should be continued for the whole length of the splice, and 
at least one inch beyond each end. When a splice is dowe'led, the dowels should 
be inserted and cleaned off before the taping is applied. When bolts are used they 
should be inserted after taping, washers being placed under the heads and nuts. 

September, 1920] THE 




Description of Splice. 

No. oO 
No. I 
Glued & Foned CLiovid olue) 
No. 2 
2 | ~~ $¢ 
a ee 
No. 3 
Glued, screwed & toned 
No. 4 Zz = z 
, ie 
a eae: 
lo & r ) 
No. 5 a 
| ‘2 
joie at 
Glued & taped 6 Ras. 
No. 6 or Ss 4 
+ + 






Solid | 



ce i | 



Breaking Load. | 




Solid silver spruce spar made to 
standard section of 3in. by rin. 


Splice parted suddenly along 
glued faces at load shown. 


Broke suddenly along the glued 
faces of joint. Liquid glue used. 

Splice like No. 2, but longer. 
Broke along surfaces of splice. 
Cake glue used. 

Splice opened at bottom and 
then fractured along grain of 

wood at centre of depth. 

Plain glue splice, taped, but not 

pegged or — screwed. After 
severe deflection parted along 

glued surfaces. 

Splice finally gave by shearing 
along fibres longitudinally, the 
wood breaking up into tough 
shreds. Glued _ surfaces held: 

522 THE AERONAUTICAL JOURNAL [September, 1920 
| Breaking Load. 
Description of Splice. | = Remarks. 
| | Solid 
Lbs. Spar 
| 100. 
S = . Result similar to No. 6, but less 
No. 7 a —<= 806 ., | Shearing. Under severe deflec- 
, + | Be ‘~ | tion the splice finally parted 
| along glued surface, though not 
, | . 5 5 . . ’ £ 
f ” | throughout. 

\ dor e 

Joint opened at bottom, following 

No. 8 20 ? 
No. r+ 209 2 J 
S a + the glued faces and breaking 
| oe = SE | away the steps at central wedges. 

Gl . wedoed. neaged § taped 

No. 9 L — 73 The steps at central wedges 
g ss : e A | z i . sheared, and transverse fractures 

i - 

- occurred at pegs. 

G\ A E pe 
as 2 ee = 806 | 80 | Same splice as No. 9, but longer. 
. 7. “ee | | Correspondingly higher breaking 
: load. 

Dovetail s hice. qlued iq | | E 
plice.glu Etsped | [Taped and glued only when 
No. 11 ze 896 | 80 | tested. (This splice was 
3 ea OE 2 | designed to be strengthened by 
ee | 

light metal straps, but these 
‘Zo Lonqued. screwed & Caped 
No. 12 ei <= 9 

were not used during the test.) 

g- | After severe deflection _ finally 

52 : 
# | broke out at screws on underside. 
hosed cena | 
Compare with Nos. g and 1o. 
No. 13 1,133 101 This has greater length, and 

central hardwood kev added, 
but no wedges or pegs. 

» GPT 

wa = 


ae a A, ae 

September, 1920] 







Description of Splice. 

ves 4 thick 

Breaking Load. 





| Solid 
| Spar 








Broke at lower inlaid hardwood 

key at end of splice, which 

dragged out; splice then broke 
at this point. 

Itractured under very — shght 
deflection and load. Fractures 
chiefly followed vertical pegs. 

Plywood connecting pieces broke 
suddenly at central butt joint. 

Broke suddenly at central butt 

joint. Compared with No. 16, 

this splice had an extra layer of 
plywood and was taped. 

Broke suddenly at ends of fingers. 
This joint not symmetrical above 
and below. 

This joint stood well, and at 

maximum load had a deflection 

of rlin. Spar fractured at ends 
of tongues. 

After severe deflection gave way 

by slipping of lower layers and 

shear along solid wood of spar 
near centre of depth. 

524 THE AERONAUTICAL JOURNAL [September, 1920 
Breaking Load. | 
Description of Splice. Solid Remarks. 
Lbs. Spar 
= 100. 

Carried a steady increasing load 
up to maximum, when deflection 
> a 103 Aig eet ais 
No. 21 55 ~” was 23 inches. At finish fingers 

began to draw and_ fracture 
occurred on underside. 
Plain spliced glued, pegaed & taped | 
Placed Horizon ESily S li I 7 | 
. ey YS Splice gave way at lower en 
No. 22 ee ee pegs | .9g | Ten Sere ee) : a or 
ee silt slipping ; then broke at vertical 
= 3 | a 
gp oes os nanan | 
Pian splice figses vertically 
ved pegged © toped. 8 Reached maximum load at ?in. 
No. 23 1,097 | 9° | deflection. Still carried more 
Sacaihinesinin aaah than half maximum load at 44in. 
ae sae deflection. 
Wice. alas 
Not screw r pegged. Maxi- 
No. 24 1,097 8 ot screwed or pegg 
re > 1 | mum load was reached at deflec- 
— | . . 
manent tion of 1} in. 
~— 16% ——+} | 
} epped fingeredglued& (no pegs) 
: Broke across at end of splice 
J 2 * 1,086 ¢f ei ee a : 
No. 25 Z i ot: 97 through fingered portion. 
See © ——+ 
| | Spindled unspliced spar to stan- 
| | . 717-1. ° ° 
| Sane | dard section. Thickness of web 
| om | ; Pele “e 
| | din. rhickness of flange $in. 
No. 26 

Entered here for comporison only 

1,064 | 95 (on face of spar). Shape of 
spindling between flange and 
web: 3in_ radius. Compare 
strength with No. o. Untaped. 





6. The form of splice that is recommended for positions where the depth of 
a spar is greater than its breadth is shown in Fig. B. 

® @ |i 

W Recommenoeo SPrice 
Fic. B. 


7. Finally, we would wish to record our opinion that notwithstanding the 
fact that the efficiency of splices may equal, or on occasion surpass that of solid 
timber, they should only be resorted to when solid timber of suitable quality and 
of adequate length for the purpose in view cannot be obtained. The problems of 

glue and gluing alone introduce a number of factors of uncertainty. 

Notr.—It may be added that the above splice has now been practically 
universally adopted for use in all machines in connection with which splices are 
authorised. It is laid down that a splice should, if possible, be situated at a point 
of inflection, and in any event should be confined to a region where the calculated 
bending moment is never more than half the maximum. 

526 THE AERONAUTICAL JOURNAL [September, 1920 


To the Editor of the AERONAUTICAL JOURNAL. 

Sir,—-On my return from abroad, a paragraph on page go of the AERO- 
NAUTICAL JOURNAL for March, 1920, which has reference to my lecture on metal 
construction, has been brought to my notice. 

The Steel Wing Company is perfectly correct in pointing out that no state- 
ment was made in my lecture to the effect that the omission of mention of the 
Company's work was due to my inability to secure their permission for such 

Before preparing my lecture, I addressed a letter to the Steel Wing Co. 
identical with those addressed to some twenty other firms in the country, 
requesting permission to mention their work and asking them to send me such 
particulars as could be incorporated in my lecture, ?.e., without infringing regula- 
tions then in force. 

With two exceptions, every firm to which I had applied gave me _ the 
necessary permission and information. One firm courteously replied that they 
did not think that their work was sufficiently advanced for publication. The 
Steel Wing Co., the second exception, replied requesting an interview with 
their director to ** discuss several matters in connection with metal construction.”’ 

For the purposes of my lecture I was glad to receive such information as 
various firms might wish to publish in this connection. Owing to my official 
position during the war I was naturally in possession of considerable information, 
and only wished to include in my lecture such matter as the firms concerned 
might select. Any discussion on general matters of metal construction seemed 
irrelevant, and, as more than ample materia! had been already placed at my 
disposal, I took no steps at the time to seek an interview with the Steel Wing 

In spite of this, being anxious to meet every reasonable request from 
interested firms, I attended at a place and time selected by a representative of the 
director of the Steel Wing Co., ostensibly for the purpose of an interview with 
the same director. I waited on two successive days for some considerable time, 
but the director of the Steel Wing Co. did not put in an appearance. Having 
thus done my best to meet him, and having nevertheless no information for 
publication placed at my disposal by this firm, obviously I did not feel justified 
in including mention of this firm’s work in my lecture. 

I] may perhaps mention that on the second occasion on which I attended 
and waited for the director of the Steel Wing Co., his representative read me a 
letter expressing regret at his inability to attend. This letter contained references 
to matters concerned in my official confidential reports on technical points made 
while I was holding an appointment at the Air Ministry during the war, which 
obviously are irrelevant and outside the region of discussion. 

With regard to the last sentence of the paragraph, it is well known that 
the services of the Technical Staff at the Air Ministry during the war were at 
the disposal of all firms working to improve and develop aircraft for the purpose 
of winning the war, and all such firms were at liberty to profit by information 
pooled at the Air Ministry.—Yours truly, 

A. P. Tuurston. 

September, 1920] THE AERONAUTICAL JOURNAL 327 

To the Editor of the AERONAUTICAL JOURNAL. 

Sik,—As you invite discussion on the subject of Dr. H. C. Watts’ recent 
paper in your July issue, on the subject of the theories of screw propulsion, I 
should like to make a few remarks. 

In the first place I am in entire agreement with the view held by your 
contributor with reference to the inconsistency of régime of the modern so-called 
‘‘inflow ’’ theory of the airscrew. As Dr. Watts points out, it is the inter- 
ference flow which constitutes the only physical inflow, and the neglect of 
emphasising this cardinal fact has led many writers on the subject into lines 
of thought which cannot be considered otherwise than as hopelessly irrational. 
At the same time, however, do not let us overlook an important factor in the 
evolution of any theory of screw propeller action. I refer to the precise quantity 

of fluid ‘** handled’? by each blade element of each propeller blade. = This 
quantity is up to the present quite unknown w priorl, and some assumption has to be 
made for it before any theory can be successfully employed in practice. This 

being so, we find ourselves in very much the same position as the schoolboy 
who makes two mistakes in his arithmetic and obtains the correct answer, 
through one mistake cancelling the other. It is perhaps somewhat pedantic to 
insist too forcibly on the irrationality of the ‘* inflow ’’ theory and to suggest a 
substitute theory which, whilst possessing a perfectly rational basis, equally fails 
in practical application with the first on account of a lack of suflicient knowledge 
in the assessment of one of the fundamental factors of all screw propeller theory. 
Considered as a rigid theory, the *‘ inflow ’’ theory is undoubtedly both irrational 
and inconsistent. Considered purely as a first tentative guess, as an empirical 
process entirely, the ‘‘ inflow’’ theory is not open to the same objection. On 
the other hand any new theory, which is to be rational in basis, will have to be 
evolved along lines closely following the conception of a periodic flow set up and 
maintained by blade interference. Before such a theory can be usefully 
employed in practice either further knowledge of the subject in general will be 
necessary or certain assumptions (analogous to those used in the “‘ inflow ”’ 
theory) will have to be made. Probably the latter course will be followed, at 
any rate at the beginning. It is then perhaps too early to begin to congratulate 
ourselves on a new and fundamental discovery which is so obviously an advance 
on all former theories, particularly the ‘‘ inflow ’’ theory, for the reason that we 
shall still find ourselves handicapped by the necessity for making some assumption 
of a character not materially different from that employed in the despised 
‘‘inflow ’’ treatment. When further knowledge is reached, the ‘‘ inflow ’’ theory 
will naturally fall into oblivion, its place having been taken by a more general 
statement of the problem evolved upon a basis of interference flow as suggested 
by Dr. Watts. 

I am not quite clear as to the exact manner in which Dr. Watts deduces his 
figures 3, 4 and 5, from’ the (original figure 2 which is taken from S. 
Drzeweicki’s latest work. Perhaps he would be kind enough to publish this in 
greater detail.—I am, yours, etc., 





M. A. S. RIACH. 


‘‘ Australian Meteorology.” By Griffith Taylor. 

Perhaps we expected of this book more than what it pretends to be, but 
the subject being one about which little has been written, and which, from its 
nature, should be of considerable interest, we had hoped for a fairly thorough 
description of Australian meteorology, especially as we are told in the preface 
that the book is developed from a course of lectures given at the University of 
Melbourne. On reading the book, we cannot help feeling disappointed, as the 
early chapters deal with extremely elementary meteorology, such as is described 
in many other text books, and which is in no way peculiar to Australia. At the 
same time, many important matters, such as the ** variation of barometric pres- 
sure with height,’’ are passed over, while much space is devoted to such subjects 
as how the atmospheric pressure may be measured by a mercury barometer. One 
cannot help being surprised to find the only suggested cause for the existence 
of the stratosphere, is a theory concerning the effect of atmospheric dust, which 
is briefly touched upon, while no mention of Gold's radiation theory is found at 
all. The complete absence of all mathematical symbols throughout the book 
will no doubt rejoice the non-mathematical reader. The Author has largels 
used American publications, and references to European work are comparatively 
few. The chapters which deal with pressure gradient and wind do not carry the 
reader very far—indeed, we have considerable doubt if the uninitiated would have 
much idea as to what the ** gradient wind"? really is, if this book were his only 
source of information. We should have thought a short description of this, 
such as is found in the Meteorological Office ** Glossary,’’ might well have had 
a place in the book. 

Considering its importance to Australia, the subject of “ rainfall’? rightly 
occupies a prominent place; one thing which will be noticed by the European 
reader is the much greater regularity in the movements of the centres of high and 
low pressure than that occurring in the North Atlantic. In view of the elemen- 
tary nature of the rest of the book, cight pages devoted to the Author’s own 
theory as to the origin of tropical cyclones seems rather out of place, especially 
as few meteorologists will, we think, agree with this ** convection dome *’ theory. 

The chapter on upper air research describes the instruments used very fully, 
even to details of the method of calibration, but practically the whole of the 
information which has been obtained by these methods, and which is now 
considerable, is passed over with but the briefest notice. 

The chapter on ** hurricanes and tornadoes’? of Australia, together with 
that on ‘* long-range forecasting,’’ will probably be of interest to the reader. 

Perhaps the least satisfactory part of the whole book is that on ‘* aviation 
and meteorology ’?; one cannot help feeling that the Author’s contact with 
aviation has been somewhat limited. Thus in reading the account of a 
meteorological flight by the Author, we find difficulty in understanding such 
phrases as ‘‘ adjusting the speed of the propeller to suit the conditions of a 
particular patch of air.’’ It appears that much of the information given in this 
chapter is derived from essays, written by the officers at the: Commonwealth 
Flying School, to whom the Author was teaching meteorology. 

The book is well illustrated throughout, and the sliding diagram at the 
beginning skows the movement of the anti-cyclonic and rainfall belts to the 
north and south, according to the time of year, in an excellent manner.