Propeller

See
discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/259030167
DEVELOPMENT OF THE OPTIMUM ROTOR

THEORIES On the 100 th Anniversary of Professor
Joukowsky's vortex...
Book · December 2013
CITATION READS
1 929
3 authors, including:
Valery Leonidovich Okulov Gijs van Kuik

Technical University of Denmark retired from Delft University of Technology
134 PUBLICATIONS 1,220 CITATIONS 113 PUBLICATIONS 2,167 CITATIONS
SEE PROFILE SEE PROFILE
Some of the authors of this publication are also working on these related projects:
Development of rotor aerodynamics for solving the wind power problems View project
Wind turbine wake studies using simple vortex methods View project
All content following this page was uploaded by Valery Leonidovich Okulov on 02 June 2014.
The user has requested enhancement of the downloaded file.

The Russian book
Okulov V.L., Sorensen J.N ., van Kuik G.A.M. Development of the optimum rotor theories.
Moscow‐Izhevsk: R&C Dyn., 2013. 120 p. ISBN 978‐5‐93972‐957‐4.
was translated in English of by interpreters of Institute Termophysics, Novosibirsk, Russia
DEVELOPMENT OF THE OPTIMUM ROTOR THEORIES

On the 100th Anniversary of Professor Joukowsky’s vortex theory of screw propeller
V.L.Okulov
(Wind Energy Department, Technical University of Denmark, 2800 Lyngby, Denmark;
Institute of Thermophysics, Siberian Branch of the Russian Academy of Sciences)
J.N.Sorensen
(Wind Energy Department, Technical University of Denmark, DK-2800 Lyngby, Denmark)
Gijs A.M. van Kuik (Delft University Wind Energy Research Institute, Faculty of
Aerospace Engineering, Kluyverweg 1 2629HS Delft, the Netherlands)
The purpose of this study is the examination of optimum rotor theories with ideal load
distributions along the blades, to analyze some of the underlying ideas and concepts, as well as to
illuminate them. The book gives the historical background of the issue and presents the analysis
of the problems arising in the actuator disc theory - the simplest model of the rotor. Authors have
found new analytical solutions for the ideal rotor with a finite number of blades which was used
for a comparison of different rotor models.
CONTENT
1 Introduction
1.1 The development stages of rotor aerodynamics
1.2 The contribution to the development of rotor aerodynamics made by Joukowsky
1.3 The objectives and the structure of the book
2 The Betz–Joukowsky limit: on the contribution to rotor aerodynamics by the British, German
and Russian scientific schools
2.1 Maximal efficiency of wind power utilization
2.2 The actuator disc theory: Retrospect
2.2.1. Contributions of the British School
2.2.2. Contributions the German and Russian Schools
2.3 Corroboration of the actuator disc theory by the vortex theory
2.4 The Betz-Joukowsky limit
2.5 Concluding remarks
2.6 Development of blade element momentum theory (V.L. Okulov’s supplement to Chapter
2)
3 General momentum theory for wind turbines at low tip speed ratios
3.1 Introductory remarks
3.2 General equations
3.2.1. The equation of motion
3.2.2. The disc load
3.2.3. The far wake
3.2.4. The momentum balance
3.3 The disc model of Joukowsky with constant bound circulation
3.3.1. The disc loads
3.3.2. The pressure in the fully developed wake
3.3.3. The momentum and energy balance
3.3.4. Analytical expressions for velocity, power and thrust
3.4 Results and discussion of the Joukowsky’s model
3.4.1. Tendencies in limiting cases
3.4.2. Quantitative results and discussion
3.4.3. Discussion of previous analyses
3.4.4. A likely explanation of the high wind power coefficient at small tip speed ratios
3.5. Simplified models without expansion
3.5.1. The model of Glauert
3.5.2. The model of de Vries
3.5.3. Simplified model of Joukowsky
3.7 Paradoxes and "paradoxes" when extending the one-dimensional momentum theory to
two-dimensional case (V.L. Okulov’s comments to Section 3)
4 Maximum efficiency of wind turbine rotors using Joukowsky and Betz approaches
4.1 Introducing remarks
4.2 Vortex theory for rotors with a finite number of blades
4.3 Solution of Joukowsky rotor
4.4 Solution of the Betz rotor
1

4.5 Results and discussion
4.7 Development of vortex theories of the rotor with a finite number of blades (V.L.Okulov’s
comments to Chapter 4).
4.7.1. Helical vortex theory - the basis for new solutions in aerodynamics
4.7.2. Hypothesis of different models and comparison of proper calculations
4.7.3. The results of experimental tests
5 Conclusion
6 References
2

1 Introduction
1.1 The development stages of rotor aerodynamics

The development of research on rotor aerodynamics (screw propeller, propeller, wind
turbine, helicopter, etc.) has always been associated with the intensive development of the
appropriate branch of engineering. The first attempts to solve the steamship navigation problems
using screw propellers should be considered as starting point for the elementary rotor theory.
This resulted into a very simple Rankine-Froude one-dimensional momentum theory of the screw
propeller, called momentum actuator disc theory (Rankine, 1865; Froude, 1889).
Figure 1.1. N.E. Joukowsky
In the early XX century, the development of rotor aerodynamics was inspired by the rapid
growth of aviation. At that time, all European aerodynamic research schools (English, Russian,
German) studied this, but the Russian and German research schools dominated the development
of the concept of the optimum rotor: N.E. Joukowsky (Fig. 1.1) with his pupil V.P. Vetchinkin
(Fig. 1.2) in Russia, and Prandtl with his pupil A. Betz (Fig. 1.2) in Germany. Their work as well
as the efforts of their contemporaries resulted in the creation of the momentum theory to
calculate the rotor blades; generalization of the momentum actuator disc theory and the
formulation of a new vortex concept of a rotor, quite suitable for the calculation of the screw
propeller with arbitrary finite number of blades (Joukowsky, 1912-1918; Betz 1919)
3

Figure 1.2. Pioneers in rotor aerodynamics in the early XX century:
V.P. Vetchinkin – pupil of Joukowsky, and representatives of the German scientific school –
L. Prandtl and A. Betz (left to right)
The creators of this new vortex theory were able to find proper solutions only for the limiting
case - for uniformly continuous circumferential distribution of vorticity along the permeable
actuator disc replacing the rotor, or, as they called it, for a rotor with an infinite number of
blades. Betz (1920) formulated the conditions for the optimal rotor and Goldstein solved
analyticaly this for 2,3,4 blades only in 1929 (Goldstein, 1929). Though, in terms of calculation,
it appeared to be so time-consuming, that Theodorsen preferred to measure circulation using
electromagnetic analogy (Theodorsen, 1948).
After rapid and fruitful period in the development of aviation rotor aerodynamics, research
activity has somewhat declined due to the emergence of jet propulsion. At this time, the research,
mostly classified, was continued mainly based on the needs of submarine fleet to develop
effective and low-noise screw propellers, as well as by requirements of helicopter engineering.
We will consider further some of the approximation results obtained for marine screw propellers
with a finite number of blades. Significant results were also achieved in the development of the
helicopter lifting airscrew theory, where, in particular, the lifting airscrew theory was further
successfully developed by the students and followers of Joukowsky (B.N. Yuriev, G.I. Maikapar,
A.I. Slutsky, E.S. Vozhdaev, etc.). This important part of the rotor aerodynamics with its specific
streamline modes of flow over the propeller was not directly the subject of commemorative
article of N.E. Joukowsky (1912-1918), and is outside the scope of present book. More details
concerning the advances of this problem can be found in the works of Baskin et al (1973) and
Vozhdaev (2003).
Current developments in rotor aerodynamics are undoubtedly associated with the rapid
development of wind power engineering, with the change of its status from an isolated alternative
energy source into an industrial sector - the most important sustainable component of the global
energy potential. At the beginning of the XXI century, a breakthrough in the development of
wind power engineering is comparable, perhaps, only with the rapid development of nuclear
power engineering in the second half of the last century, and most likely will surpass it. Indeed,
for example, in 2010, in Denmark, the energy produced by wind turbine was 21% of the total
electric energy generated, in Portugal - 18%, Spain - 16%, Ireland - 14%, Germany - 9%. Further
growth is projected up to the level of 50% that brings wind power engineering into line of the
most priority-driven development trends in the world energy production. Respectively, interest of
today’s researchers in the rotor aerodynamics has increased significantly. Today a new stage of
intensive scientific development is outlined again, similar to the productive aviation era of
Joukowsky, Prandtl and their scientific schools. The authors of the current work were involved in
4

this process as well. In particular, in their articles, and first of all, in (Okulov & van Kuik 2012),
was clarified a role of Joukowsky (Joukowsky, 1920) in a derivation of the important result about
maximum wind power conversion efficiency (59.3%) which was wrongly named earlier by Betz
limit as result the one author only (Betz, 1920). As it turned out, this limit has not only great
practical value, but also serves as a great indicator to test the validity of various theories of the
rotor. This value was used in another article (Sørensen & van Kuik 2011) to analyze the general
momentum theory for the rotor. The most important result of the authors' works turned out to be
the analytical solution to the problem on optimum rotor with an arbitrary finite number of blades
(Okulov & Sørensen 2010) obtained in the frameworks of the vortex theory. Indeed, in the
beginning of the last century, both schools were not able to derive analytical solutions for rotors
with a finite number of blades and replaced it by the simpler disk models. The problem of the
helical vortices became a stumbling block. Many outstanding scientists of most famous aero- and
hydrodynamic research schools were not able to deduce the solution to this problem over quite
long time. But recently, substantial progress has been achieved in theoretical studies: the
solutions for the individual helical vortices and their multiplets were obtained in the form
convenient for calculations (Okulov, 2004).
1.2 The contribution to the development of rotor aerodynamics made by Joukowsky
Outside Russia, the name of the great Russian scientist Nikolai Yegorovich Joukowsky (Fig.
1.1) is mainly associated with Kutta-Joukowsky theorem on wing lift and wing profiles designed
by applying conformal mapping according to Joukowsky. Of course, these two facts do not
exhaust a huge heritage of this great scientist, one of the founders of the aerodynamics and the
"father of Russian aviation", whose complete set of works includes nine volumes (Joukowsky,
1935-1937). There may be many reasons why many colleagues rarely refer to the name of
Joukowsky and his scientific achievements. For certain this is caused by the lack of translation of
Joukowsky’s works into English. Indeed, both mentioned facts became known abroad mainly due
to Ludwig Prandtl, who named and described them in his famous textbook on hydrodynamics
(English translation, Prandtl, 1952). Another important reference to Joukowsky’s achievements
in foreign publications was made in a very well-known and still popular review article on the
propeller aerodynamics by Henry Glauert (Glauert, 1935). This review resulted from close
collaboration with the laboratory of Ludwig Prandtl, so it is natural that many issues of rotor
aerodynamics were presented in the interpretation of German school. However, the contribution
of Joukowsky was quite fully portrayed by the author in the introduction, while the section
devoted to the general momentum theory for the rotor, as a matter of fact, was written on the
basis of the translation of the "vortex theory of screw propeller" into French (Joukowsky, 1929).
In particular, in the additional footnote, the author mentioned that Joukowsky was first to derive
the equations of the general theory in 1918 referring to the Paris edition (Joukowsky, 1929).
Because of the limited accessibility of Joukowsky’s work, we now pay some additional
attention to his contribution to rotor aerodynamics. The mentioned series of 4 articles under the
title "Vortex theory of screw propeller", published from 1912 to 1918, is the most significant
work of Joukowsky. Surely, along with these four articles it is worth mentioning three other
works devoted to wind turbines: "Aerodynamic calculation of slow wind mills", "Low-speed
wind mills" and "NEJ type wind mill". The latter article was the last research article of this
scientist (Joukowsky 1920). These articles were published in the "Transactions of the Central
Aerodynamic Institute" in 1920, and the last of them, in addition, was translated into French and
published in the "Bulletin of the Institute of Aerodynamic Research at Kuchino" in 1923. It
seems paradoxical that the last of his efforts was spent by the great scientist for wind power
engineering in the country, where it is still in its infancy because of unlimited natural resources.
5

As if he had foreseen its unprecedented growth occurring in the world today - a century later,
believing that the future is with wind source of energy because this energy is free and
inexhaustible. Today, we are to recognize the exactness of Joukowsky’s scientific forecasting in
this area of knowledge. According to his legendary prediction: "The man has no wings, and in
terms of the ratio between the weight of his body and the weight of his muscles the man is 72
times weaker than a bird ... But I think he will fly not relying on the strength of his muscles but
on the strength of his mind".
Obviously, the special place of these seven articles in the heritage of Joukowsky is explained
by the fact that the theory of screws is the most difficult and important subdiscipline in
aerodynamics, where all accumulated results on the flow around individual bodies and wings
should be extended to rotating objects - blades. Therefore it could be said that the prerequisites
for the development of the theory of screws were outlined in the Joukowsky’s works much
earlier. Of fundamental importance here is the work "Du Bois’s paradox”, written in 1891 and
explaining an interesting phenomenon: the motionless body in the flowing fluid has a greater
hydrodynamic resistance than a body floating in a stationary fluid. Joukowsky has shown that
this is due to the swirling generated in the flowing fluid. In 1898, Joukowsky published an article
"On winged propellers" where he discussed the possibility of creating heavier-than-air aircrafts
and suggested to equip them with flapping wings (similar to the flight of birds). In 1906
Joukowsky published a short article "On the bound vortices", where he was the first to expound
the principle of wing lift and determined the relation of its value with circulation, density and
velocity of the incident fluid. V.A. Kutta attempted to obtain a similar result before Joukowsky
(in 1902) in Germany, but in his dissertation on the lift of the circular arc, he could not
understand the role of circulation at flowing around the arc and instead used its geometric
parameters. It should be noted that the Joukowsky’s work (1906) had not proposed a method for
determining the circulation either, and the theorem on lift definition remained incomplete. Only
in January 1910, Kutta again returned to the subject of his thesis and published the correct
formula for the lift. At the same time, in a series of Joukowsky’s works "On the profiles of
airplane supporting surfaces" (1910), "Geometric study of the Kutta flow" (1911-12), and etc.
and in a number of works by S.A. Chaplygin, the researchers proposed the postulate for
unambiguous definition of circulation and exact calculation of the airplane wing lift. At this
time, based on calculations, Joukowsky suggested the optimal profiles for wings and airplane
screw blades called "Joukowsky’s profiles". The derived formulas allowed finding an optimal
configuration of the profiles, obtaining formulas for the lift and the momentum of these
theoretical profiles.
In fact, the first step toward a theoretical framework for the calculation of the screw was the
work of Joukowsky "Theory of the screw propeller with a large number of blades" published in
1907. In this work the formulas for determining the thrust and power of the propeller and
helicopter rotor were derived. Later in 1912-1920, the vortex theory of propellers and wind
turbines developed by Joukowsky provided a key to solving the problems that arise in the design
and construction of various types of rotors: propellers, rotors for rotary-winged aircrafts, wind
turbines, axial fans and screw propellers. Screws, built on his theory, are called "NEJ screws"
(the initials of the scientist).
It is no coincidence that the milestones in the development of aviation rotor aerodynamics
are associated with the names of Russian scientists and the school of Joukowsky, because at that
time, they were the most advanced in the world. In the early XX century, this fact was generally
recognized, that was proved by the article of V. Margoulis, published in August 1921 in the
French journal “L`Aeronautique” and dedicated to the memory of Joukowsky. The article was
reprinted in the "Bulletin of the Aircraft Fleet" (1923, No. 2), and its English translation was
published in the Technical Memorandum of NACA (Margoulis, 1922). The author analyzes the
works of Joukowsky and his progeny and concludes that the results for the above key theories of
6

the rotor were obtained several years earlier than in other European scientific schools. Since
these results were of a basic character, and further development of these theories included only
improvement and various clarifying amendments, the conclusion of Margoulis appeared to be a
complimentary for the Russian aerodynamic school.
1.3 The objectives and the structure of the book
The core of this book is composed of three recently published articles on the history and
development of the rotor theory as applied to now popular wind turbines. The above-mentioned
articles were published in the leading English-language journals on fluid mechanics (Okulov &
Sørensen 2010) and wind power (Sørensen & van Kuik 2011; Okulov & van Kuik 2012). In
order to describe the rotor theories keeping the logic of presentation from simple to complex, the
material of the articles in this book is provided in reverse chronological order of their publication.
The text is amended with comments written by one of the authors – V.L. Okulov (introduction,
conclusion and the necessary explanations in the form of supplements after each chapter).
This book is dedicated to the centenary of the publication of Professor N.E. Joukowsky’s
novel vortex concept of the rotor theory in the first article of the series "The Vortex theory of
screw propeller", which was reported to the Moscow Mathematical Society on October 1, 1912.
Therefore, our main goal is to restore the history of rotor aerodynamics and to describe its
modern achievements obtained after a century of this landmark date. In this context, the
successful completion of the work and significant contribution to the development of
Joukowsky’s vortex theory of the rotor was a new analytical solution for NEJ rotors with a finite
number of blades and their first comparison with Betz rotor proposed by the German school.
Furthermore, the supplement to the last chapter, first provides a comparison of all major
theories of optimal rotor, including recent solutions, by comparing their results for the wind
turbine modes with the Betz-Joukowsky limit.
7

2 The Betz–Joukowsky limit: on the contribution to rotor aerodynamics by the British,
German and Russian scientific schools
The derivation of the efficiency of an ideal wind turbine has been attributed to famous
scientists of the three aerodynamic research schools in Europe during the first decades of the
previous century: Lanchester, Betz and Joukowsky. However, detailed reading of their classical
papers has shown that Lanchester did not accept Froude’s result that the velocity through the disc
is the average of the velocities far up- and downstream, by which his solution is not determined.
Betz and Joukowsky used vortex theory to support Froude’s result, and derived the ideal
efficiency of a wind turbine at the same time. This efficiency has been known as the Joukowsky
limit in Russia and as the Betz limit everywhere else. Due to the contribution of both scientists,
this result should be called the Betz-Joukowsky limit everywhere.
2.1 Maximal efficiency of wind power utilization

The result mentioned in the title is famous: no more than 59.3% of the kinetic energy of fluid
contained in a stream tube having the same cross section as the area of a rotor disc (or any energy
transformer) can be converted to useful work by the disc. Certainly, this result is right only if the
stream tube is free from external influences, for example, the stream tube should not be inside
diffusers or pipe sections. Indeed it is impossible to stop a wind/fluid motion in the stream tube to
take 100 % efficiency because this implies a full blockage of the flow downstream of the disc.
The flow can only be reduced, not stopped, so there is a rate of deceleration that gives maximum
power conversion: 59.3 % of the kinetic energy. This was established by the actuator disc theory
based on the conservation laws of the flow. In this theory, any device is replaced by a permeable
disc with a distributed load yielding the same overall thrust at the real device. Despite this far-
reaching abstraction this theory is the base of wind turbine aerodynamics. Sometimes this limit is
named as the «Carnot cycle» of wind energy because the value does not depend on the type of
energy transformer.
For a long time, this result about the maximum of kinetic energy which can be utilized by an
ideal wind-driven generator, was known as the Betz limit published in 1920 [1] (Betz 1920). In
1979 Bergey (Bergey 1979) has proposed that Lanchester should be associated with the result too
because of his 1915 publication (Lanchester 1915). In 2007 van Kuik (Kuik 2007) found that
Joukowsky (Joukowsky, 1920) derived the same result in 1920. Taking into account the high
importance of this result, our current study addresses again the history of the discovery.
2.2 The actuator disc theory: Retrospect

The actuator disc (or momentum; or slip-stream) theory is the oldest mathematical
representation of a screw, propeller or wind turbine etc in fluid dynamic calculations. The load
on a real rotor is replaced by a pressure distribution on an infinitely thin, permeable disc with the
same diameter. In its most elementary presentation, this load is uniform and normal, with the disc
placed in an axial flow perpendicular to the flow direction. This actuator disc concept is still used
as an easy qualitative diagnostic model, and any textbook on rotary wing or rotor aerodynamics
starts with it. The three European aerodynamic research schools that were famous in the first half
of the XXth century have contributed significantly to this theory: the British school led by
Froude and Lanchester, the German school led by Prandtl and Betz, and the Russian school led
by Joukowsky and Vetchinkin. Figures 1.1, 1.2 and 2.1 show the photographs of Joukowsky;
Vetchinkin, Prandl and Betz; Froude and Lanchester. These contributions are reviewed in
retrospect.
8

Figure 2.1. Pioneers in the momentum rotor theory:
R.E.Froude, F. Lanchester and B.N.Yuriev (left to right)
2.2.1. Contributions of the British School

The idea to replace a rotor by an actuator disc goes back to the work of Rankine (Rankine
1865). However only in 1889 R.E. Froude (Froude 1889) has for the first time found a correct
dynamic interpretation of the actuator disc action showing that for such a theoretical propeller
one half the acceleration must take place before the propeller and the other half behind it.
Unfortunately, the discussion on the question whether the contraction or expansion of the
streamtube takes place before or behind the disc continued after Froude’s paper, despite his
formal mathematical treatment. In the Appendix to his paper, §27, Froude derives the energy
balance (the work done by the disc is thrust times velocity through the disc, which equals the
change in kinetic energy of the flow in the streamtube times the mass flow in the tube) and the
momentum balance (the thrust equals the mass flow times the change of velocity). This gives the
result that the velocity at the disc is the average of the velocities far upstream and far
downstream.
Professor V.P. Vetchinkin (Vetchinkin, 1918), who was a pupil of Joukowsky, sought an
explanation for the denial of Froude’s result in the misunderstanding of the relation between the
action of a disc and of real rotor blades. Most scientists at that time thought, erroneously, that the
flow before the rotor plane is undisturbed, then receives a full speed alteration when it moves
through the rotor blades, after which the flow behind the rotor is undisturbed too. Another
incorrect interpretation of an ideal rotor, based on Parsons’ remarks to the Froude’s article
(Froude R.E. 1911) assumes that the total flow alteration is received before and during the rotor
passage only, with the flow behind the rotor moving uniform and without disturbances.
Followers of Parsons’ theory said that it is impossible to find any reason or force for the flow
acceleration/deceleration in the free stream tube behind the rotor plane when the rotor action has
ended. Though the assumption of Parsons’ theory results in incompatible equations for the ideal
work and energy in the wake, the theory was very popular among famous English scientists in
beginning of the XX century. From 1910 to 1915 there was a lively discussion on both Froude’s
and Parsons’ theories between professors Henderson, Froude, Parsons and Lanchester in pages of
the popular English edition: «Transactions of the Institution of Naval Architects» (see issues of
52, 53, 55 and 56). An author of the last article is F.W. Lanchester (Lanchester 1915). In the
introduction to this article he wrote: - “The present investigation takes for its starting point the
simplified or idealized conception of propeller due to Mr. R.E. Froude …” – then he remarked on
the discussion mentioned above and concluded – “Without definitely entering or taking part in
the dispute in question, the present contributor proposes to review the theory from its foundation,
in order to make sure of his own ground…”. In his analysis, Lanchester supports the energy and
9

momentum balance as defined by Froude, as is clear from his statements at p. 108. However, he
continues the discussion on the ‘difficulties of regime’ started by Froude, in particular on the
pressure discontinuity at the disc edge. Although Lanchester says that ‘the admitted difficulty
relating to the edge of the actuator is probably more apparent than real’ his next step is to
substitute the continuously operating disc by an intermittendly operating disc shedding vortex
rings into the flow. Lanchester states that a considerable portion of the change in kinetic energy
is now to be found in the outer portions of the vortex rings, so outside the streamtube passing
through the disc. According to modern insights in vorticity dynamics, Lanchester’s statement is
incorrect since he should include also the pressure- and unsteady terms in the energy equation. It
is here where Lanchester deviates from Froude and leaves the possibility open that velocity at the
disc is not the average of the velocites far up- and downstream (see Fig. 2.2 which is copy of Fig.
6 from (Lanchester, 1915)).
а б
Fig.2.2 (a) Сorrect model of the Froude’s actuator disc; (b) Erroneous model by Parsons.
Here the Lanchester’s derivation is presented, showing how close he was to a firm
assessment of the maximum efficiency.
Fragment of the article “A contribution to the theory of propulsion and the screw propeller” by
F.W. Lanchester [3], PROBLEM II. – THE WINDMILL
Et = energy per second in windstream engaged.
mt = mass per second of air engaged.
V = velocity of wind (a constant).
v = residual velocity.
u = velocity when passing through actuator.
w = force on actuator.
wind energy Et
Q or 
work done power developed
(Lanchester has introduced the parameter Q as a measure of the incompatibility for the ideal
work and energy in the wake – rem. by authors)
V 2  v2
E t  mt (2.1)
2
10

V 2  v2
w u  mt (2.2)
2Q
and momentum/sec. –
w  mt V  v  (2.3)
or by (2.2) and (2.3) –

V 2  v2 V v
u  (2.4)
2QV  v  2Q
If p = work done per unit mass/sec. (per unit mt) –
wu V 2  v2
p or by (2)  (2.5)
mt 2Q
By (2.4) –
du 1
 (2.6)
dv 2Q
By (2.5) –
dp v
 (2.7)
dv Q
By (2.6) and (2.7) –
dp 2Q v
  2v (2.8)
du Q
And when pu is maximum –
dp p

du u
or by (2.4), (2.5) and (2.8) –
V 2  v 2 2Q
  2v
2Q V v
or –
V  v  2v
V  3v (2.9)
Thus the maximum work is got out of the wind when its residuary velocity is one-third of initial
velocity, and this ratio is independent of the value of Q.
V
Assuming best relation v  . By (2.4)
3
V v 4 V 2V
u   (2.10)
2Q 3 2Q 3 Q
11

2 1
The limiting conditions are Q  1 u  V and Q  2 u  V . The first corresponds to the
3 3
Froude condition, the second is analogous to (but in sense of reverse of) the condition attributed
to Parsons, thus (Fig. 2.2).
We will now find an expression to represent the available power, expressing this in terms of
a standard represented by the energy of the free passing per second across an aria equal to that of
the actuator. We will denote the available power so expressed by the symbol .
Assuming best condition, i.e., V  3v . By (2.2)
4mt V 2
Power developed  w u  (2.11)
9Q
2 AV 
where, by (2.10) mt  Au   . (2.11) becomes
3Q
8 AV 3
(2.12)
27Q 2
Power represented by “free wind” on area A
AV 2
(2.13)
2
or
16

27Q 2
For the limiting values –
16
Q  1,   0.6 approximately
27
4
Q  2,   0.15
27
In effect, Lanchester tried to find a compromise between both Froude’s and Parsons’ theories
of the ideal propeller/turbine but his symbiosis of the theories was nearly to Parsons’ point of
view as it was noticed by Vetchinkin (Vetchinkin, 1918). Indeed, Lanchester’s model of the ideal
wind turbine, described in Problem II of [3] and partly repeated in the Appendix, is a symbiosis
of both theories with transition parameter Q (1 < Q < 2). When Q = 1 the symbiosis coincides
with the Froude theory and for Q = 2 it is the Parsons theory (see Fig. 2.2). For the symbiosis
model Lanchester has derived a formula of the wind power coefficient which is equal to 16/27Q2
and tends to the maximum of the Froude’s wind power coefficient if Q = 1. Lanchester suggests
that Q should be around 1.5.
Bergey’s conclusion (Bergey 1979) that Lanchester is co-author of the Betz limit is not valid,
since it assumes that Lanchester has adopted Q = 1. Bergey ignores the Q discussion and
presented the derivations and equations of Lanchester after substitution of Q = 1, whereas
12

Lanchester himself included Q in the equations yielding results for the limit conditions Q = 1
and 2.
In his paper, Bergey discusses the fact that the Prandtl school did not refer to Lanchester’s
work (Lanchester 1915), without finding a satisfactory answer. Indeed Lanchester was well-
known in Germany and in Russia. His article (Lanchester 1915) was well known too. It has been
published in a well accessible edition, and it was referred to by Joukowsky in his famous work
“Vortex theory of screw propeller” (Joukowsky, 1912-1918) (in the last article of this cycle) and
in Vetchinkin (Vetchinkin, 1918). The absence of the reference in Prandtl’s school was because
they could not support it.
As a preliminary conclusion, Lanchester’s name should not be linked to the Betz limit, which
results directly from Froude’s theory. It is noteworthy to realize that he and his biographies never
claimed the rights to this result.
2.2.2. Contributions the German and Russian Schools

In the meantime, more and more scientists in the world became supporters of the Froude’s
theory as Vetchinkin noted it in (Vetchinkin, 1918). In accordance with the Vetchinkin’s remarks
Bendemann (Bendemann 1910) wrote in 1910 that professor Finsterwalder has proved in his
1904 lectures that the induced velocity in the far wake behind a rotor becomes double its value in
the rotor plane. He drew the same conclusion as Froude but it is not clear whether he did know
his article of 1889. In 1912, during the second Russian aeronautic congress, Sabinin and Yuryev
(fig. 2.1) reported this fact too (see the record in sixth footnote of (Joukowsky, 1912-1918)) and
Vetchinkin reproduced Finsterwalder’s proof in (Vetchinkin, 1918). In 1917 Bothezat has
general the result about the doubling of the induced velocity in the wake for actuator discs
producing not only a forward but also a rotary movement in the wake (Bothezat, 1917). Finally
Joukowsky formulated in 1918 the modern state of the Froude’s theory in §6 of (Joukowsky,
1912-1918). This history has been supported and extended by Hoff (Hoff 1921) who indicated
Finsterwalder as the scientist who established the theory too, which was extended afterwards by
Bendemann in (Bendemann 1910) and completed by Prandtl in an appendix to Betz’s paper (Betz
1919). The Russian and German schools knew each other very well and cooperated, e.g. in a
common responsibility for a scientific journal, see fig. 2.3.
Fig 2.3. Title page of a scientific journal, showing the connection

between the Russian and German aerodynamic schools
13

Probably this list of supporters of Froude’s theory may be expanded and each of the scientists
could have derived the limiting value of the Froude’s theory in an application to wind turbine.
Indeed the value consists in Lanchester theory (Lanchester 1915) as a limit at Q = 1. Sabinin
(1927) mentions that Vetchinkin was the first, in 1914, who expanded Froude’s theory of
propellers to wind turbines but this does not become clear from his publications. There may have
been others, as Froude’s theory existed for a long time, since 1889, and if accepting Froude’s
theory, than the derivation of this limit is straightforward and not related to mathematical
complexities.
Though, the problem here is different. In the first two decades of the XX century, there was a
struggle going on between various viewpoints and theories, and yet it was not clearly ascertained
that the Froude’s theory was correct. Although Froude’s theory was accepted by many scientists,
it was not yet possible to show a connection between the abstraction of the actuator disc and the
action of real blades on the flow. Therefore, to establish an abstract limit of Froude’s theory as
the maximum value of wind power coefficient, it was necessary first to show the validity of the
theory itself. This was resolved by rotor vortex theory
2.3 Corroboration of the actuator disc theory by the vortex theory

The first article of Professor Joukowsky from his cycle «Vortex theory of screw propeller»
(Joukowsky, 1912-1918) has been published early in the 1913. First results of this article were
announced on October 1, 1912 as a report to the Moscow Mathematical Society, and although,
according to Vetchinkin, the publication came out at the beginning of 1913, the article was dated
the year of 1912. In this article, he created the vortex model of a propeller based on a rotating
horseshoe vortex, which expanded the first elementary vortex model of a wing with a finite span.
In his vortex theory each of the blades is replaced by a lifting line about which the circulation
associated with the bound vorticity is constant, resulting in a free vortex system consisting of
helical vortices with finite cores trailing from tips of the blades and a rectilinear hub vortex, as
sketched in Fig. 2.4.
Fig.2.4 (a) the first elementary vortex model of a wing with a finite span based on a single horse-shoe
vortex reported by Prandtl at the Gottingen congress and published in 1913;
(b) Joukowski’s vortex model of a propeller based on a rotating horse-shoe vortex in the accordance with
Prandtl’ model for a wing (Prandtl 1913) but with finite vortex core .
Vetchinkin commented in (Vetchinkin, 1918) that Joukowsky’s vortex theory has finally
confirmed Froude’s actuator disc theory. Indeed, an accurate calculation of the velocities induced
by the vortex system of a propeller, presented in the article, confirmed the predictions of the
elementary Froude’s theory. It was the first article enabling to understand the source of velocity
acceleration/deceleration in the wake of a propeller/turbine. This was the vortex system of tip
vortices shedding from the tips of the rotating blades, which induces in the rotor plane an axial
velocity half as low than the velocity induced by the vortex system in the trail at infinity.
14

Though, final approval of the elementary theory of an ideal propeller, in the form proposed by
Froude, succeeded only after publication of the fourth and the final article of this series
(Joukowsly, 1912-1918), which came out in 1918 in the Transactions of the Aviation Design and
Test Bureau. In paragraph 6 of this article, Joukowsky was first to present the theory of ideal
propeller – the general Froude’s theory that was comprehensively explained and presented in the
conventional dimensionless form suggested by Vetchinkin (1913, 1914).
Fig.5 (a) Refined Prandtl’s vortex model of a wing with elliptic load distribution published in 1918
(Prandtl 1918); (b) Betz’s vortex model of a propeller (1919) based on this new Prandtl’s solution.
Simultaneously in Germany A. Betz worked on the creation of the propeller vortex theory.
About the actuator disc theory, he remarked in (Betz 1921): “the point, in which the old propeller
slip-stream theory (the Froude’s theory – rem. by authors) needed to be supplemented, was the
assumption that the thrust could be distributed at will over the surface of the propeller disc… It
may, however be here noted that the difference in comparison with the uniform distribution is not
so great as appears at the first glance.” Then he referred to Föttinger’s propeller horseshoe vortex
model (Föttinger 1918) consisting of infinitely thin vortex lines, which confirmed the main
findings of Joukowsky’s model. As the next step Betz proposed a new model in 1919 (Betz
1919). In Betz’s model of the vortex theory each of the rotor blades is replaced by a lifting line of
which the circulation is associated with bound vorticity, with a free vortex sheet being shed
continuously from the trailing edge of the rotating blade and moving with constant velocity in
axial direction (Fig. 2.5). In contrast with the horseshoe vortex model this alternative propeller
theory used Prandtl’s vortex model of a wing with an elliptic distribution of the loading along the
span (Prandtl, 1918; 1919). Thus, a field of well-defined vortices is connected with the
distribution of the propeller thrust, or the lift of a rotating wing. Betz concluded in (Betz 1921)
that the motion of the fluid is definitely determined by the vortices existing in it, and that the
flow due to the thrust distribution may be calculated by means of this concept of vortices.
The conceptual ideas of Joukowsky’s and Betz’s interpretations of the propeller vortex
theory have allowed proving for Froude’s actuator disc theory as a unique and correct elementary
theory for propellers. In 1912-1918 Joukowsky and in 1919 Betz have shown a connection
between an abstraction of the actuator disc and the real blades action on the flow for the first
time.
2.4 The Betz-Joukowsky limit

In 1920, famous aerodynamists (Joukowsky and Betz) have independently published articles
to develop Froude’s theory to the theory of the ideal wind turbine in which the value of the
maximal ideal work which can be extracted from the kinetic energy of wind was found
(Joukowsky, 1920; Betz 1920). Munk (Munk 1920) did the same, also in 1920. In addition to this
Hoff remarked that his article with the same topic [17] was written somewhat later than articles
(Betz 1920) and (Munk 1920) but was independent of them too. Maybe there are more unknown
15

works. It was a natural result after the confirmation of Froude’s actuator disc theory by vortex
theory in (Joukowsky 1912-1918) and (Betz 1919) (see table 1) because the development of the
actuator disc theory to wind turbines looks simple and ordinary. Nevertheless, we select here
only two names because the two independent publications by Joukowsky and Betz in 1920 are
the result of their great achievements in (Joukowsky 1912-1918) and (Betz 1919). Their
contemporaries made the same choice too to pay attention to only these scientists for their
confirmation of Froude’s theory. Indeed, everywhere in the world the ideal efficiency of a wind
turbine is known as the Betz limit or Betz law without a reference to somebody else. In Russia
this result is known as the Joukowsky limit, with the reference that it is known as the Betz limit
outside Russia. Below we try to find answer for this difference.
The paper of Betz (Betz 1920) has a title that shows the topic clearly: the maximum
efficiency of a wind turbine. Joukowsky’s article (Joukowsky, 1920) has a quite special purpose.
It was a response to inventor Vinogradov, who promised to create a wind turbine with
unprecedented efficiency. Joukowsky answered him by a conclusion about the maximum of wind
energy utilization for the ideal wind turbine described in the first paragraph of the foregoing
article. In the same article, he has expanded the theory for case of the ideal wind turbine with
rotation in wake behind a rotor and with an additional assumption about constancy of the
circulation. Joukowsky was 73 in 1920 and it was the last article in his life. The followed
illnesses and death have given up an opportunity to continue his works on the topic.
In opposite to this Betz has published in 1926 the remarkable book “The Wind Energy and
its utilization by windmills” which made the name of Betz well-known name amongst wind
energy engineers. Furthermore, the paper of Betz has been published in a journal, whereas, the
paper of Joukowsky was part of the Transactions of his scientific institute, with a possibly more
limited distribution. These are reasons why his work went unnoticed for the scientific community
in the world.
Moreover, in 1927 the attention of compatriots has been drawn away by the erroneous theory
of an ideal propeller of Sabinin (Sabinin, 1927). Unfortunately, his wrong theory finds now
followers again since it used in some Russian textbooks and manuals. Sabinin wished to create a
distinct theory from Froude’s theory, for the operating regime with a turbulent wake (fig. 2.6)
where the model of ideal wind turbine is not valid. At first glance, his theory was constructed in
strong accordance with vortex theory unlike the Parsons and Lanchester case with Q > 1.
Nevertheless, he used a wrong idea to calculate the lift forces by a starting-vortex that travels
downstream and keeps the shape of a vortex solenoid.
Figure 2.6. Diagram of rotor operation modes depending on the axial induction coefficient of the
impeller (reproduced from the textbook of Hansen, 2008).
16

The last assumption is wrong because it neglects a roll-up mechanism of the starting vortex,
by which the real flow and pressure are totally different from Sabinin’s model. In his theory, the
maximum efficiency becomes a little larger (68.7%) than the Betz-Joukowsky value. If we write
the Sabinin theory in terms of the Lanchester theory, it is easy to find that Q = 1.137, and we
meet the same contradiction that was in Parsons and Lanchester theories.
Table 1. History of the Betz limit: the main contributions and contributors
Year and scientist of

Contributions British school German school Russian school
1865 W. Rankine
1889 R.E. Froude
1904 Finsterwalder
Actuator disc theory 1910 Bendemann
1912 Sabinin
1913 Vetchinkin
1917 Bothezat
1918 Joukowsky
1919 Prandtl
actuator disc theories 1911 Parsons
with a violation of the 1915 Lanchester
conservation lows 1927 Sabinin
Corroboration of the 1913 Joukowsky
actuator disc theory by 1919 Betz-
the vortex theory
February, 1920
Expansion of the actuator Joukowsky
disc theory to wind August, 1920
turbines and formulation Munk
of the limit for the wind September, 1920
power coefficient Betz
July, 1921
Hoff

Although Lanchester was close to calculate the limit efficiency of a wind turbine, he did not
do so since he did not include Froude’s result that the velocity through the disc is the average of
the velocities far up- and downstream. .
This limit has been known as Betz’s limit everywhere in the world and as Joukowsky’s limit
in Russia only. Due to the contribution of both scientists, this result should be identically called
the Betz-Joukowsky limit everywhere - in Russia and in the world too.
2.6 On the development of blade element momentum (BEM) theory (V.L. Okulov’s
supplement to Chapter 2)
The above Froude’s momentum theory for actuator disc has been considered as the first
elementary one-dimensional model of the rotor which is precise to describe correctly the
averaged and simplified structure of the flow, and established the Betz-Joukowsky limit, but that
17

theory still had not given the method to design the propeller blades. Therefore in parallel of the
Froude’s theory specific methods for the blade designing were developed in blade sections. The
possibility of such a representation of a rotor blade through its separate elements, cut by
cylindrical cross-sections, was proposed in 1892 by Stepan Karlovich Drzewiecki, a native of the
Podolsk province of the Russian Empire (Fig. 2.7).
Figure 2.7. Founders of the blade element method for rotor calculation:
S.K. Drzewiecki, G.A.Bothezat and G. Glauert (left to right)
Figure 2.8. Drzewiecki’s blade element method:

(a) Sketch of the section; (b) the initial velocity triangle by Drzewiecki;
(c) new velocity triangle with correction factor by Yuriev and Sabinin.
According to his theory, the screw is cut by cylindrical sections to blade elements, and
velocity triangles are plotted for each element (Fig. 2.8 a and b). The main value of Drzewiecki’s
theory consists in creation of an efficient method for the blade design. Therefore an improved
version of this method is still widely used when rotors design for different purposes. In his work
(Drzewiecki, 1910), he identified conditions for the maximum performance of airscrew, namely
the attack angle providing best performance, methods of choosing for the size and pitch of the
blades, their cross-sectional sizes, and the formula to determine their number. Drzewiecki’s
theory was easy and convenient for engineering calculations. Nevertheless, his original theory,
even in its final presentation (Drzewiecki, 1910), had a significant defect, because it did not take
into account flow changes due to an influence of the rotor on the flow, that changes the axial
velocity (so called induction factor) in the rotor plane. In his time, not all researchers accepted
the results of the momentum actuator disc theory, because they did not understand the cause of
additional velocity due to the action of the bound vortices of the rotor and the free vortices in the
18

wake downstream the rotor, explained by Joukowsky (Joukowsky 1912-1918). On this occasion
Vetchinkin (1918) wrote that "... the existence of the error remained, and we see that the screws
of S.K. Drzewiecki (manufactured at the Ratmanov factory in Paris), absolutely correctly
designed in accordance with his theory, are replaced in the market by purely empirical screws of
Shovier, Regis, Levasseur, and others, who chose the form and size by thanks exclusively to
tests, without taking any theoretical considerations". More complex situation with the use of the
original Drzewiecki’s theory aroses for the hovering helicopter mode, when there is no main
stream to the rotor. These incongruities had no solution, as noted by Vetchinkin (1918), until
G.H. Sabinin and B.N. Yuriev, the students of the Imperial Technical School, got a fortunate idea
to use Rankine-Froude’s elementary momentum actuator disc theory for correction of the
velocity field in the rotor plane (Figure 2.8, c) for the theory of S.K. Drzewiecki. After that, in
1910-1912, they started designing and studying new theoretical screws. Despite numerous
sources describing this story, including a reference made by Joukowsky (1912-1918) and a
description of their theory in the works of Vetchinkin (1913, 1918), the authors were not able to
publish their work in full at that time, because of their conscription and subsequent stay in a
German captivity. Later B.N. Yuriev (see Fig. 2.1) entirely explained their work in his article
(Yuriev, 1923). It should be noted that in Germany the effect of an additional axial velocity was
introduced to Drzewiecki’s theory only in 1915 by Betz (Betz, 1915). In England, it was taken
into account starting from 1919 after the publication of the book of (Fage & Collins, 1919), while
the correction to the axial velocity of the rotor was taken from the experiment. Obviously,
distrust in Froude’s theory in England continued in 1919 too.
In 1917, Professor Georgy Bothezat (Fig. 2.7), who lived in Petersburg, proposed a more
accurate version of the correction for the additional velocity in the case where the screw provides
to the air not only progressive movement, but also rotational motion. The BEM theory was
corrected by the definition of both induced velocity components (Bothezat, 1917). On the basis
of this completed version of the BEM theory, Henry Glauert (Figure 2.7) conducted a refined
calculation of the Betz-Joukovsky limit in each section of the rotor blade (Glauert, 1935). His
result will be discussed in Section 3.5.1; it proved to be dependent on tip speed ratio, but was
always less than the Betz-Joukovsky limit, determined by slipstream momentum theory.
It should be noted that the correction of Sabinin-Yuriev and its further clarification made by
Bothezat did not take into account an effect of the mutual interaction of the elements of different
blades, i.e. this correction can be applied only for a particular type of screw - with a small
number of thin blades. The exact calculation method by an unwinding of the cylindrical section
of the rotor on the plane and transition to the calculation of planar cascade of the blade profiles
was introduced by Joukowsky (1912-1918) in his third article of this series. Though Joukowsky
addressed to S. Chaplygin’s work (1914), devoted to an abstract analytical solution for an
influence of the cascade wing on two-dimensional parallel flow, it was his achievement to the
rotor case because he successfully used the Chaplygin’s solution to the rotor in order to answer
the question on an influence of number and width of the rotor blades on the screw operation.
Thus, the hydrodynamic theory of the cascades became the finishing point for the refined
designing of the rotor blades using the blade element method for the screws.
A creation of the general momentum theory for actuator disk by Joukowsky became the next
stage in the development of rotor aerodynamics. In 1918 he described it in the last article of his
series. This theory was naturally concluded from his understanding of physical principles of the
rotor operation based on his vortex theory of screw propeller with constant circulation along the
blade. A description of the Joukowsky’s general momentum theory in English was published by
Henry Glauert in his article about the airscrew theory (Glauert, 1935) which is very popular so
far. That became possible due to French translation and publication of all 4 articles of "Vortex
theory of screw propeller" in 1929 in Paris. That happened because Stepan Drzewiecki played
very important role in promoting the works of the Joukowsky’s school abroad. Settling down in
19

Paris in 1892, he retained the close contact with the Russian scientific school and was a translator
of Joukowsky’s works into French language.
Undoubtedly, the general momentum theory of the rotor, as well as the simplest theory of
Rankine-Froude, is still far from perfection and comprehensiveness due to simplifying
assumptions and conditions. The next Chapter will be devoted right to description of the
problems arising at consideration of this theory as applied to the certain specific limiting case -
the low-speed wind turbines.
20

3 General momentum theory for wind turbines at low tip speed ratios
General momentum theory is employed to study the behavior of the ‘classical’ free vortex wake
model of Joukowsky. This model has recently attained considerable attention as it shows the
possibility of achieving a power performance that greatly exceeds the Lanchester-Betz limit for
rotors running at low tip speed ratios. This behavior is confirmed even when including the effect
of a center vortex, allowing azimuthal velocities and the associated radial pressure gradient to be
taken into account in the axial momentum balance without any simplifying assumptions. It is
shown that the most likely explanation for the anomalous behavior at small tip speed ratios is that
the influence of the lateral component of pressure and friction is neglected in the axial
momentum theorem. A refined model is proposed that remedies the problem of using the axial
momentum theorem and by which the wind wind power coefficient never exceeds the
Lanchester-Betz limit and which tends to zero at zero tip speed ratio.
About a century ago Joukowsky (1912) developed a simple aerodynamic model based on general
momentum theory and the concept of a rotor disc with constant circulation. This model has since
then been the subject of much controversy, since it causes a wind wind power coefficient that,
independent of tip speed ratio, is always greater than 16/27 and that for small tip speed ratios
tends to infinity. In the text book of Glauert (1935) the model was presented as a special case of
the general momentum theory. However, Glauert also stated that the condition of constant
circulation cannot be fully realized in practice, since it implies that near the roots of the blades
the angular velocity imparted to the air is greater than the angular velocity of the propeller itself.
Later investigators, such as de Vries (1979) and Wilson and Lissaman (1978), shared the view
point of Glauert that the solution is unphysical at small tip speed ratios as it results in infinite
values of power and circulation when the tip speed ratio tends to zero. Sharpe (2004) argues that
the theory in principle establishes that there is no loss of efficiency associated with the rotating
wake and that it is possible, at least in theory, to exceed the Lanchester-Betz limit. In a recent
work by Madsen et al. (2007), employing a simple vortex model of Øye (1990) and a numerical
actuator disc model, the computed wind wind power coefficient never exceeded the Lanchester-
Betz limit. Thus, there seems not to be fully agreement in the validity of the model.
The main motivation for the present investigation is to analyze the basic assumptions
behind the derived equations in order to come up with an explanation for the apparent
contradiction. Thus, instead of claiming that in principle it is possible to exceed the Lanchester-
Betz limit at small tip speed ratios, we seek an explanation for why this happens. The paper is
organized as follows. In Section 3.2 we derive the basic equations that form the basis for the
analysis. In Chapter 3 we introduce the disc model of Joukowsky with constant bound
circulation. In Section 3.4 we solve the equations and discuss some representative results to
understand the physics. The results clearly show unphysical behavior, such as infinite expansion
and an infinite wind wind power coefficient when the tip speed ratio goes to zero. This issue has
been discussed in several papers using various approximations. We include the most important
contributions in our discussion. In the classical derivation of the equations, the influence of
friction and pressure on the lateral boundary on the control volume is ignored; although it is clear
that this is needed in order to limit the unphysical area expansion. In Chapter 4 we show that the
unphysical behavior at small tip speed ratios can be completely remedied by including the lateral
force component. In Section 3.5 we introduce a simplified version of the model ignoring
expansion which avoids the problem of using the axial momentum theorem. Finally, in Chapter 6
we conclude the work.
21

3.2. General equations
3.2.1. The equations of motion

The flow is considered to be inviscid, incompressible and adiabatic. The equation of motion in its
most simple form:
1
 f  p   v  v (3.1)

will be used, as well as the equivalent form:
f   H  v   (3.2)
in which f is the force density [N/m ], p the static pressure [N/m ], v the velocity vector [m/s]
3 2
and H  p  ½  v  v the total pressure [N/m ]. The two expressions are related by the vector
2
identity v  v = (½ v  v) - v    v and the expression for the total pressure. In a cylindrical
reference system ( x , r ,  ) with the positive x coinciding with the downwind wake axis, r and 
the radial and azimuthal coordinate, (3.1) becomes :
1  v2   vv 
 f  p   e x v vx  er  v vr    e  v v   r  (3.3)
  r   r 
where e x is the unit vector across the infinitesimal thin disc, er is the unit vector in the radial
direction and e is the unit vector in the azimuthal direction. Only the pressure and the azimuthal
velocity will be discontinuous, so integration of the axial and azimuthal component of (3.3)
across the disc gives:
1 p
F  ex  e v x v (3.4)
 

 H 1 
 ex   v2   e vx v (3.5)

  2 
where F denotes a surface load [N/m ],  the difference between the down- and upwind side of
2
the disc and where

1
p  H   v2 (3.6)
2
is used.
3.2.2. The disc load

The local power converted by the force field f is f  v which has to be equal to the local
contribution to the torque, rf , times rotational speed  . From (3.2) the converted power f  v
becomes
f  v  rf   v   H (3.7)
22

Figure 1: The control volume CV, coordinate system and direction of forces and velocity.
This shows that the work done by the force field is expressed in a change in the total pressure or
Bernoulli constant H 
Integration of (3.7) across the thickness gives:
r
H  F (3.8)
vx
or, with the azimuthal component of (3.5):
1
H  r v  (3.9)

Using this, the disc load (3.5) can be expressed entirely in kinematical terms,
 
1 1 1
p  Fx   r  v  v (3.10)
  
 2 

1
F  vx v (3.11)

In the remainder the index 0 is used for flow variables in the undisturbed, upstream flow. The
fully developed far wake is indicated by the index 1, see Figure 3.1. If there is no index, the
variables are taken at the position of the actuator disc.
3.2.3. The far wake

With the conservation of circulation
rv  r1v 1 (3.12)
the Bernoulli equation (3.9) is written as:
1 1
 p0  p1    vx21  U 02  v21   r1v 1 (3.13)
 2
Differentiating with respect to the radial coordinate in the far wake and combining it with the
radial pressure equilibrium in the far wake:
p1 v21
 (3.14)
r1 r1
we get the following relation between the axial and azimuthal velocities in the far wake:
23

v    r1v 1 
 
1 vx1 
2
     1    (3.15)
2 r1  r1  r1
From this it is immediately seen that a one-dimensional flow with constant vx1 can be established
in two ways, either as a rotating body ( v 1  r1 ) or as a free vortex flow (v 1   (2 r1 )) where
 is the vortex strength.
3.2.4. The momentum balance

A second relationship between axial and azimuthal velocities can be established by applying the
axial momentum equation on integral form along a surface enclosing the disc. As illustrated in
Figure 3.1, it is common to use a control volume going along the stream surface passing through
the tips (CV). In general, shear stress, lateral pressure or velocity of shear layer between the wake
and the exterior are not known. However, at high Reynolds numbers these quantities are small
and in all text books on propeller aerodynamics they are ignored, and the momentum equation is
written as
T 
A1
 p1  p0  dA1   A vx1  vx1  U 0  dA1
1
(3.16)
where the thrust T   pdA Using (10), (12), (13) and the conservation of mass, vx dA  vx1dA1 
A
this can be rearranged into the following relationship:
 v v ,1 
   2 
r1 
2
2
r
A1  v x1  U 0  dA1  A1 r1v 1v x1  v x  v x1 dA1 
   (3.17)
 
 
This equation was almost a century ago derived by Joukowsky (1912). However, as will be
demonstrated later, ignoring the lateral stress term in (3.16) may have a significant impact on the
results at small tip speed ratios.
3.3. The disc model of Joukowsky with constant bound circulation

We now consider the special case of a far wake consisting of an axial centre vortex and a circular
vortex cylinder originating from the edge of the actuator disc with total axial circulation opposite
to the concentrated centre vortex. The induction from this system results in an azimuthal velocity
that is inversely proportional to the radial distance from the centre and a constant axial flow. As
shown previously this flow is a special case that obeys the kinematical wake relation given by
(3.15). This concept for the disc flow was introduced by Joukowsky (1912), and enables a full
analytic treatment of the problem. The free vortex at the axis of the wake induces infinite
azimuthal velocities when the vortex core diameter  is infinitely small, leading to infinitely low
pressure. For that reason, the analysis starts with a finite value for  , after which the limit   0
is applied. The vortex core is assumed to be a Rankine vortex core, which implies a solid body
rotation: v r .
24

3.3.1. The disc loads
By the jump in azimuthal velocity across the disc, v , the disc becomes a bound vortex sheet
with strength:
  er v (3.18)
According to Saffman (1992) the velocity of an infinitely thin vortex sheet is the average of the
velocity at both sides, so the azimuthal velocity at the disc itself is:
1
v  x 0  v (3.19)
2
The azimuthal velocity field is induced by a discrete vortex  with kernel thickness  at the axis
of the wake. The vortex kernel is a Rankine vortex core, giving:

v  v  for r   (3.20)
2 r
2
 r
   for r   . (3.21)
2 r   
Equation (3.9) becomes
1 
H   constant for r   (3.22)
 2
2
  r 
   for r   (3.23)
2   
which shows that a disc with a constant jump in H is equivalent to a disc with constant bound
circulation. Eq. (3.22) also shows that for a wind turbine, having H  0 ,  and  must have
different signs. We assume that   0 , so:
0 for a propeller (3.24)

  0 for a wind turbine
The axial load becomes, using (3.10) and (3.11)
1   1    2     r  2 1   r  2 
F        (3.25)
 x  2 2  2 r     r  R  2    2  2  2   0 r 
Xiros & Xiros (2007) have shown that the axial velocity through the disc is constant when the
azimuthal vorticity   0 everywhere except at the wake boundary. This applies to the
Joukowsky model of constant circulation, so vx is constant for r   . Across the boundary of the
vortex kernel, r   , the Bernoulli constant H is continuous, as is p , v and vr . Consequently,
vx is also continuous. Inside the kernel, (3.15) shows that vx is constant, which leads to the
conclusion that the axial velocity is constant for 0  r  R .
Then the azimuthal load is, using (3.11):
25

1      r 
2
F  v  v  (3.26)
   x 2 r    r  R  x 2 r     0 r 
Integration of (3.25) and (3.26) gives the thrust T :
  2  2   2  dr 1 
R R
1 1

T
  Fx 2 rdr 
0

2 
R 
2  4   r 4 
     (3.27)
the torque Q :
  2  2 
R
1 1
Q  F 2r dr  v x R 
2
(3.28)
  0
2  2 
and power P :
1 1   2  2 
P Q  vx R   (3.29)
  2  2 
Comparison of (3.27) with (3.10) shows that the term containing  originates from H , whereas
the term between square brackets in (3.27) originates from ½ v2 .The latter term makes the
thrust to become infinitely large in the limit   0 while the torque and power remain finite:
T  TH  T 


1  2 
TH  R  HR 2 
 2 


1 2   1  
T    ln   
 
 4  R 4   for 0 (3.30)
 R
1  
Q  vx R 2 

 2 

1  2 
P  vx R  v x HR 2

 2 
Comparison of the expressions for thrust and power shows that T , independent on the choice
of vortex core, does not convert power, so does not change H . This can be explained by
substituting H  0 in (6). Then Fxv  p   12 v2 , showing that the pressure jump and
the jump in azimuthal velocity balance each other, without a change in H . A second explanation
is provided by considering the disc as a bound vortex sheet. Integration of (3.2) across the disc
gives, using (3.18) and (3.19):
F  e x H   v   (3.31)
The last term is the Kutta-Joukowsky load FKJ    v    It performs no work, since the load is
perpendicular to the local velocity. Evaluation of FKJ with v  e x vx  e r vr  e 12 v gives:
1 2
FKJ  x     v   (3.32)
2
26

Integration for 0  r  R gives the expression forT , so TKJ  x  T which confirms that only a
part of the axial load on the disc performs work. In Section 3.3.3 where the momentum and
energy balance are evaluated, this is confirmed.
3.3.2. The pressure in the fully developed wake

The pressure distribution in the far wake is given by (3.13). Using (3.20) the pressure gradient for
1  r1  R1 is:
p1 1 v 
   v21  vx 1 x 1  (3.33)
r1  r1 r1   r  R
1 1 1
This gradient has to deliver the centripetal acceleration v2 / r . Substitution of v 1    2 r1 

shows that this is satisfied when vx ,1 / r1  0 . Consequently vx ,1 is constant in the fully
developed wake for r1  1 . Following the arguments given between eqs. (3.25) and (3.26), this
also holds inside the kernel of the vortex, so vx ,1 is constant for 0  r1  R1 .
All terms at the right-hand side of (13) except 1
2  v21 are constant, so this can be written as:
1
p1  p0    v21  p (3.34)
2
 
2
At the wake boundary the pressure has to be undisturbed ( p0 ), so p  12  v2 R1  12  
2 R1 and:
1       
2 2
p1  p0 1 2 2 
  v  R1  v ,1       (3.35)
 2 2  2 R1   2 r1  
  1  r1  R1
The pressure for 0  r1  1 is derived from (3.13), using (3.21) and (3.23):
2
  r1 
v 1    for r1  1 (3.36)
2 r1  1 
2
1   r1 
H1    for r1  1 (3.37)
 2  1 
giving:
 2 2 4 
p1  p0 1    r1      r1  
 
       p  (3.38)
 2    1   2 r1   1  
  0  r  1 1
where p is determined by equating (3.35) and (3.38) for r1  1 , giving


2
   
p    (3.39)
 2 R1  
Integration of (3.35) and (3.38) across the wake cross-section gives
27

R1 R1
p1  p0 2 2 dr1  2

0

2 r1dr1  
16 4 
 r1

4
1  (3.40)
1
For 1  0 the following remains:

R1 R1
p1  p0 2 2 dr1

0

2 r1dr1  
16 4 
 r1
(3.41)
1
3.3.3. The momentum and energy balance

The axial momentum balance is given by (3.16) or (3.17). Substitution of (3.30) and (3.41) in
(3.16) gives:
 1     2  dr 1 dr1 
2 R R 2
  R1 
2 2  2 R  4   r 1 r1 
        vx 1  vx 1  U 0    

(3.42)
R
The diameter of the vortex core and the wake radius is dictated by the continuity equation. Since
vx as well as vx1 do not depend on r , it follows
2 2
vx  R1   1 
    (3.43)
vx1  R    
so R  1R1 . Consequently the term between square brackets is:
1
d  r R  1 d  r1R1 

R r R
 
1R1 r1R1
0 (3.44)
and the axial momentum balance simplifies to:

2
    vx  vx 1 
  2   1  (3.45)
 U 0  2 RU 0 
2
U0  U0 
This shows that the divergent pressure integral in the expression for the thrust, (3.30), is
cancelled by an equal term in the far wake, by which it does not contribute to a change of
momentum. The second term in the LHS of (3.45) is the result of p  in (3.34) required to have
undisturbed pressure at the wake boundary.
The Bernoulli eq. (3.13) combined with (3.35), expressed in far-wake values becomes:
2
     vx 1 
2
      1  (3.46)
 U 02  2 R1U 0   U 02 
As for the momentum balance, there are no divergent terms since these contributions to ( p1  p0 )
and  v2,1 cancel each other. In (3.45) as well as (3.46) the terms with  2 make the difference
1
2
with respect to flows without swirl.

The combination of (3.43), (3.45) and (3.46) gives:

vx 1  vx ,1  4  R 2
   1 (3.47)
U0 2  U0  4  
R12
28

According to (3.24)   0 for a propeller, and R  R1 , whereas for a wind turbine   0 with
R  R1 . In both cases vx  1
v  U 0   It should be noted that if expansion is neglected then the
2  x 1
two terms become equal.

Now we know that the vortex kernel has no influence on the momentum balance, we could have
derived the momentum equation (3.45) easier by substituting (3.35) and the first term at the right-
hand side of (3.25) in (3.16)
The combination of the mass balance (3.43), momentum balance (3.45) and energy balance
(3.46) provides expressions for the unknowns v x , v x ,1 and R1  for the independent parameters
,  and U 0 .
3.3.4. Analytical expressions for velocity, power and thrust

To simplify the equations we write q   / 2RUo and introduce the tip speed ratio   R / U o ,
so  / U o2  2q . The minus sign is added to q in order to adopt to wind turbine practice
where both  and  are taken positive, in contrast to (3.24). Furthermore the axial induction
coefficient at the disc is defined as a  1  vx U 0 and in the wake it is given as b  1  vx 1U 0 .
With this, the momentum balance (3.45) becomes
q 2  2 q  2b(1  a )  0 (3.48)
and the Bernoulli equation (3.46) together with the mass balance  RR1   1  b   1  a  
2
1 b 2
q  2q  b(2  b)  0 (3.49)
1 a
Writing (3.48) plus (3.49) gives

ba 1 1 
b  2 q    2 qvx ,1  
2
 (3.50)
 1 a   vx vx ,1 
and the area expansion is given as
2
 R1  vx 2 q
     (3.51)
 R  vx 1 2 q  b
2
Elimination of (1  a) in (3.49) by substitution of (3.48), and multiplying the result by (q  2 )

expresses b in the independent parameters  and q :
 2  q  b 2  4  b  2  q  2   q   0 (3.52)
giving:
2  4 2  2 q  q 2  4 2 
b (3.53)
q  2
or
vx1 q  42  2q  q 2  42 
 (3.54)
U0 q  2
The flows defined by q  2 are excluded from the solution. Since q  2 is a propeller flow
29

state, this is not a problem. For q  2 direct substitution in (3.48) and (3.49) gives
a  0 and b  q 2 for q  2 .
Then vx is solved with (3.48):
2 q  q  4 
 2 2
vx q 2  2 q
1
   (3.55)
U0 2b 2  4 2  2 q  q 2  4 2 
With the power expressed in dimensionless coefficients, C P  P / ½ U o3R 2 , and using (3.30) we
obtain :
v q 2  2 q
C p  2 q x  2 q (3.56)
U0 2b
With the thrust coefficient defined as CT  T / ½ U o2R 2 we arrive at results for a b CT and C p
expressed in the independent variables q and  :
2  4 2  2 q  q 2  4 2  

q  2 
q  2 
b  

q2 q  2 




 (3.57)
q  2 q
2
a  1 
2b 


  1 
CT  CTH  CT  2 q  2q  ln   
 R 4


Cp 2 q (1  a ) 
 
3.4. Results and discussion of the Joukowsky’s model

For given tip speed ratio  and circulation q , the system of equations (3.57) determines the axial
interference factors and the wind wind power coefficient. In particular, the system can be utilized
to optimize the theoretical power output. In the following we first show some general tendencies
in the case of infinite wake expansion and next we show the results from the full range of
governing parameters.
3.4.1. Tendencies in limiting cases

In Figures 3.2 and 3.3 we show C p - and vx U 0 values for 0  q  5 and 0    5 .
30

Figure 3.2: Cp as a function of λ and q = - Figure 3.3: Vdisc/U0 as a function of λ and q = -
Г/(2πRU0). Г/(2πRU0).
The following observations are made:

1. The momentum balance (45) fails when there is no stream tube anymore, so vx1 has
to be greater than zero. For vx1  0 R1   by which (46) gives  / U o2  1
equivalent with 2q  1 or H   12 U 02  From (3.57) it follows that
1 2 
a  1  q  
 
2 
1 2  H
Cp  1  q   for  1 or 2q  1 (3.58)
2 U 0
  2
2 
1
1 
 
2q 
For q   we see that C P   .
2. For q  2 we have a  0 and b  q 2  These flow states combine an undisturbed disc
velocity with a decelerated wake velocity. The white line through the origin of the C P
figure separates the region C P  1 from the region C P  1 . For   q / 2 the disc flow
is higher than the undisturbed one. For q  1 and   12 , C p  1 This particular flow
case was first calculated by Lam (2006).
3. Figure 3.2 indicates that for   2 the classical one-dimensional results are
reproduced with C p max  16  27 . This is confirmed by the more detailed analysis in the
next Section.
3.4.2. Quantitative results and discussion

In Figures 3.4 to 3.9 the results are presented in more detail. Figure 3.4 gives the wind wind
power coefficient as function of the axial wake interference factor. For tip speed ratios less than
2, the wind wind power coefficient starts to increase for b -values greater than 0.5 and for   1 ,
C p attains a maximum value at b  1 that greatly exceeds 16/27 with C p max   for   0 .
31

Figure 3.4: Wind wind power coefficient as function of axial wake interference factor
at various tip speed ratios (TSR =  = 0.2, 0.5, 1.0, 2.0 and 10.0).
If we look at the corresponding axial interference factors a and b, we see that there is no
longer a simple linear relationship (see Figure 3.5). From Figure 3.5 it is observed that the
interference factor in the rotor plane becomes negative for   0.5 . This implies that the flow is
accelerated when passing through the rotor plane, indicating that the flow emanates from an
upstream area that is bigger than the area of the rotor plane. In Figure 3.6 we depict the
dimensionless circulation that was calculated from (3.57). As in the case of the wind wind power
coefficient the circulation increases and obtains its maximum value for b  1 a value that tends to
infinity for   0 . From (3.51) the area expansion may be determined. This is shown in Figure
3.7 which depicts the area expansion as function of axial wake interference factor. For small tip
speed ratios we see that the maximum area increases dramatically, tending to infinity for   0 .
Figure 3.8 shows the wind power coefficient as function of axial interference factor. For small tip
speed ratios we here observe a rather peculiar behavior where C p increases nearly linearly for
decreasing a -values. In Figure 3.9 we summarize the computed results by showing the
maximum wind power coefficient as function of tip speed ratio. It is here seen that the maximum
C p -values tend to infinity for   0 and that the maximum C p -value always is greater than the
Lanchester-Betz limit at 16/27. This behavior has been the subject of many considerations and
explanations. However, before going into a discussion on earlier works, it is instructive to
analyse if this seemingly unphysical behavior is possible in practice. To this end we verify if the
lift distribution achieved at small tip speed ratios can be obtained in practice.
32

Figure 3.5: Axial interference factor in rotor plane as function of axial wake interference factor
at various tip speed ratios  (TSR).
Figure 3.6: Dimensionless circulation as function of axial wake interference factor

Figure 3.7: Area expansion as function of axial wake interference factor

33

Figure 3.8: Wind power coefficient as function of axial interference factor
Figure 3.9: Maximum wind power coefficient as function of tip speed ratio  (TSR)
From Joukowsky’s theorem we have the following expression for the lift:
L  Vrel , (3.59)
where Vrel is the relative velocity in the rotor plane:
Vrel  U 02 (1  a ) 2  (r  ½ v ) 2 . (3.60)
Defining the lift coefficient as
L
CL  , (3.61)
½ Vrel2 2 r
we get
2q
CL  . (3.62)
2
r
2
 r 
2
  (1  a )       ½ q 
2
R  R 
At the root of the blade CL  4 , after which it decreases monotonously to a smaller value at the
edge of the actuator disc. Although a value of CL  4 at the root may seem somewhat high, it
should be noted that the denominator in eq. (3.61) is based on the local perimeter of the disc
which goes to zero at the root. Even in the limiting case where vx ,1  0 a finite CL distribution is
obtained. Inserting the limiting values from eq. (3.58), we get
4
CL  , (3.63)
2
r
2
 3  r 2 
  (2 2
 ½) 2
  4     
R  R 
34

which gives for   1 a CL distribution from 4 at the disc center to 0.73 at the disc edge.
3.4.3. Discussion of previous analyses

Glauert (1935) stated that the condition of constant circulation cannot be fully realized in practice
since it implies that near the roots of the blades the angular velocity imparted to the air may be
greater than the angular velocity of the propeller itself. This indicates that Glauert did not believe
in the model as such, since the angular velocity for any tip speed ratio at some radius always will
be greater that the angular velocity of the propeller itself. Other investigators, such as de Vries
(1979) and Wilson and Lissaman (1978), shared the view point of Glauert that the solution is
unphysical as it results in infinite values of wind power coefficient and circulation when the tip
speed ratio tends to zero. In a recent work, Sharpe (2004) argues that the theory in principle
establishes that there is no loss of efficiency associated with the rotating wake and that it is
possible, at least in theory, to exceed the Lanchester-Betz limit. Comments on Sharpe’s analysis
were published by Lam (2006) and Xiros & Xiros (2007). Furthermore, Wood (2007) published
an approximate analysis of the same problem including also the core radius of the hub vortex.
From their analyses, they all conclude that the maximum wind power coefficient always is
greater than the Lanchester-Betz limit of 16/27, and that it increases towards infinity with
decreasing tip speed ratio. However, in recent numerical studies on optimum actuator discs by
Madsen et al. (2007, 2010) the optimum wind power coefficient did not exceed the Betz limit.
Thus, there seems not be fully agreement in the validity of the model. In Section 3.5 the
discussion on some of the previous models is continued.
Figure 3.10: Maximum wind power coefficient using the modified momentum equation,
as function of tip speed ratio  (TSR).
It is believed that the present analysis solves some of the major problems encountered in
the previous analyses, where the authors had to make simplifying assumptions. The effect of the
infinite azimuthal velocities and the associated negative-infinite pressures indeed makes the
thrust to become infinite for   0 , but this singular behavior is shown to have no direct effect
on the momentum and energy balance. The same holds for the contribution of the kernel of the
vortex itself. However, the behavior shown in figs. 3.9 and 3.10 is clearly in contradiction to
intuition and other models dealing with optimum rotors. Logically, there must be a problem in
the model or the used equations, which is the topic of the next subsection.
3.4.4. A likely explanation of the high wind power coefficient at small tip speed ratios
The main motivation for carrying out the present investigation was to analyze the basic
35

assumptions behind the derived equations in order to come up with an explanation for the
apparent contradiction of an infinite wind power coefficient. Thus, instead of claiming that it in
principle is possible to exceed the Lanchester-Betz limit at small tip speed ratios, we seek an
explanation for why this happens. We therefore take a critical look at the equations derived in
Section 3.2. In all cases the Bernoulli equation is valid along a stream surface, hence the
derivation of eqs. (3.1) - (3.15) is correct. Thus, it is more likely to search for the problem in the
use of the axial momentum equation, eq. (3.16). A more complete version of this equation also
includes pressure and friction terms, as shown below
T   pdA   vx v  dA   pdA  e   τdA  e

x x . (3.64)
CV CV CV
Recalling that the influence of friction and pressure on the lateral boundary on the control
volume was ignored, it is likely that this is needed in order to limit the big area expansion seen in
Fig. 3.5. The question is if the swirl is responsible for an additional suction force that makes it
possible to achieve a solution with infinite wake expansion. To verify if the strange behavior at
small tip speed ratios is related to the lack of the influence of the lateral pressure, we try to
include this term in the equations. The problem is that the term is not known a priori, which is
also is the reason that it is neglected in the usual analysis. However, although we do not know the
exact expression, it is most likely proportional to the area expansion and the pressure drop over
the rotor at the edge of the disc. Thus, we propose to model it using the following expression,
 pdA  e
lb
x  pr  R ( A1  A) , (3.65)
where pr  R is the pressure drop at the edge of the rotor and  is a small coefficient giving the
net influence of the integrated pressure acting on the lateral boundary of control volume CV (see
fig. 3.1). With pr  R given by (3.10), introduction of this term into the axial momentum balance
gives:
T   pdA    vx ,1 (vx ,1  U o ) dA1   ( p1  p0 ) dA1  pr  R ( A1  A) (3.66)
from which it follows (see also eq. (3.50)),
1 1
b 2   2 q (1   )   q 2 )  vx ,1 (  ) (3.67)
vx vx ,1
Introducing dimensionless variables, the area expansion is now given as (see also eq. (3.51)),
2
 R1  v 2q
   x  . (3.68)
R v x ,1  q 
2q1   (1  )  b 2
 2 
From eq. (3.49) the circulation is determined as
36

2 2
R R R
q           2  b(2  b) , (3.69)
 R1   R1   R1 
and from the eq. (3.56) the wind power coefficient reads
2
R 
CP  2 q  1  (1  b) . (3.70)
R
Solving eqs. (3.68)-(3.70) for different values of  we observe a remarkable change of the
solution. For all  -values the wind power coefficient becomes zero at zero tip speed ratio. Even
utilizing as small  -values as 10 7 , the wind power coefficient becomes exactly zero for   0 .
In Fig. 3.10 we depict the maximum wind power coefficient as function of tip speed ratio for
different  -values and compare to both the original curve and to the optimum curve of Glauert. It
is here clearly seen that the influence of the lateral pressure has a big impact on the behavior of
the maximum wind power coefficient. For   0.1 it still has a hump, but for larger  -values it
goes monotonously towards the Betz limit. Thus, a likely explanation for the large increase in
C P at small tip speed ratios is the lack of the lateral pressure or friction effects on the control
volume
3.5. Simplified models without expansion

The model presented in the previous Section requires the unknown parameter  for a closed form
solution. In this chapter we discuss various models in which this parameter may be neglected by
assuming no wake expansion.
Figure 3.11: Maximum wind power coefficient as function of tip speed ratio  (TSR) for four different
models. Full line: Free vortex model without expansion; Dashed line: Glauert model; Dashed-dotted line:
Optimized free vortex model; Dotted line: Erroneous model of de Vries (3.79).
3.5.1. The model of Glauert

In the rotor model of Glauert (1935) the following assumptions were applied:
v2  v2,1 , p o  p1 and v x  ½(U o  v x ,1 ) .
37

Inserting these assumptions into eq. (3.13), we get
(v x2,1  U o2 )  2rv  v2  0 . (3.71)
It is readily seen that it is impossible to keep both the circulation and the axial velocity constant
along the rotor blade, since there is no longer a pressure difference to outbalance the difference,
as in eq. (3.13). Introducing the circulation, (r )  2rv , and defining it in dimensionless form

as q   , eq. (3.66) is written as
2RU o
q 2  2 2 q  4a(1  a) 2  0 , (3.72)
where a  1  v x / U o is the axial interference factor and   r / R . Solving this equation with
respect to the dimensionless circulation results in
q   2     2  CT /  2  , (3.73)
 
where CT  4a(1  a) is the thrust coefficient. From Euler’s turbine equation the power is obtained
as follows,
R 1
P   2 r 2  vxv dr  2R3Uo2  (1  a)q d , (3.74)
0 0
or in dimensionless form,
1
CP  4  (1  a )q d  . (3.75)
0
By assuming that the different stream tube elements behave independently of each other, Glauert
used this equation to optimize the integrand for each x separately, resulting in the well-known
relationship for the optimum rotor,
1  3a
q  2 2 a  2 2 for C P  C Pmax , (3.76)
4a  1
v
where a   is the azimuthal interference factor. Combining eqs. (3.73) and (3.76) it is
2r
possible to determine a and q at optimum conditions. Inserting the results into eq. (3.75) and
integrating, the maximum wind power coefficient may be obtained as a function of tip speed
ratio. The result is the medium curve shown in Fig. 3.11.
3.5.2. The model of de Vries

In the report by de Vries (1979) it was proposed to seek an optimum by assuming constant axial
induction. This is a natural extension of Joukowsky’s rotor model, since this model assumes
38

constant circulation and axial induction. However, with the introduced approximations resulting
in eq. (3.73), it is only possible to keep one of the two variables constant. Thus, assuming
constant axial induction, eq. (3.75) is written as,
1
CP  4(1  a)  q d . (3.77)
0
Inserting eq. (3.73) into eq. (3.77), we get

1
 1  1  C /( ) 2  d .
CP  4(1  a ) 
2 3
(3.78)
0
 T

Integrating this equation, the following expression is found for the wind power coefficient,
(1  a )  2 C 2     CT 
2
CP     CT  (2  ½CT )  3  T ln . (3.79)

  2 C 
 T 
For a given tip speed ratio  , the optimum wind power coefficient is determined by solving eq.
(3.79) for different a -values. It is interesting that the result is completely identical to the one
obtained from the analysis of Glauert. This is in contrast to the achievements of de Vries (1979),
who determined a maximum wind power coefficient that was markedly lower than the one from
the Glauert analysis. Looking more closely into the equations derived by de Vries, however, we
found an error in the expression of the wind power coefficient. It seems that (besides a typing
error) the factor 2 was not included in the denominator of the last term in eq. (3.79). In Fig. 3.11
we depict the Glauert curve together with the erroneous curve obtained by de Vries. As can be
seen, the maximum value of eq. (3.79) corresponds exactly to the curve of Glauert. Thus eq.
(3.79) is a closed form of the optimum Glauert curve, which uniquely describes the optimum
wind power coefficient as a function of tip speed ratio and axial interference factor. The last
curve in Fig. 3.11 is based on a modified Joukowsky model, which will be derived in the next
subsection.
3.5.3. Simplified model of Joukowsky

The assumptions stated in the optimization models derived above are quite strong, and may result
in severe constraints of possible solutions. As an alternative we here present a modified model in
which we neglect the influence of area expansion. Assuming further that v x  ½(U o  v x ,1 ) , from
eq. (3.48) we get
q 2  2q  4a (1  a )  0 , (3.80)
Solving eq. (3.80) with respect to the dimensionless circulation, we get
q     2  4a (1  a ) . (3.81)
From Euler’s turbine equation the following expression determines the wind power coefficient
C P  2q (1  a )  2 (1  a )( 2  4a (1  a )   ) . (3.82)
39

For any given values of the tip speed ratio,  , the optimum of this function is determined
numerically. The outcome is shown as the solid curve in Fig. 3.11. It is interesting that this curve
for any value of  is greater than the optimum curve of Glauert and that it never exceeds the Betz
limit. Thus, this model confirms that the rotation indeed may increase the performance at low tip
speed ratios, as compared to the optimum of Glauert, but that it never becomes greater than
16/27. The main advantage of the modified model, as compared to the original model of
Joukowsky, is that it is not needed to solve the momentum equation in order to determine the
wind power coefficient. This has the obvious advantage that we do not need to estimate the
unknown lateral pressure contribution. An alternative interpretation of the model is that it
corresponds to the induction of a vortex sheet located midway between the outer free stream flow
and the wake, assuming the so-called roller-bearing analogy (see Øye 1990), in which the
convective velocity of the sheet is given as the average between the free stream velocity and the
wake velocity. Since the induced velocity is perpendicular to the relative velocity at the location
of the sheet, we get the following relationship:
½(U o  v x ,1 )  /( 4R )
 , (3.83)
R   /( 4R ) ½(U o  v x ,1 )
which simply can be rewritten as eq. (3.80). An alternative way of writing equation (3.80) is
4a(1  a)
q . (3.84)
q  2
Inserting this expression into eq. (3.82) gives

CP  4a (1  a ) 2 . (3.85)
  ½q
This equation shows clearly that the wind power coefficient tends to the Betz limit for high tip
speed ratios and that it for small tip speed ratios tends to zero. One can interpret the added
rotational term as a modification to 1-dimensional momentum theory, as in Madsen et al. (2007).
It should be mentioned that eq. (3.85) is identical to eq. (3.82) and that a similar equation can be
found in the report by de Vries (1979) and in the text book of Burton et al. (2001).
3.6. Concluding remarks

The equations forming the general momentum theory of wind turbine rotors were analyzed in
order to understand anomaly behavior at small tip speed ratios. The analysis is divided into three
parts: 1) an exact solution within the Euler flow regime assumptions, 2) adaptations of this
solution to account for effects of pressure and friction at the wake boundary, and 3) the
derivation of simplified mathematical models. The respective conclusions are:
1. The azimuthal velocity in the wake is induced by a discrete vortex at the wake axis.
Previous publications on this topic made assumptions to deal with or to avoid the infinite
azimuthal velocity and associated negative pressure near the axis. In the present analysis
the azimuthal velocities and the associated radial pressure gradient have been taken into
account in the axial momentum balance without any assumptions. For all  -values the
wind power coefficient, C P , attains values above the Betz limit. For   0 the wind
power coefficient tends to infinity and for high  it tends towards the Betz limit.
40

2. Most likely it is the influence of neglecting the contribution from the lateral pressure and
friction forces in the axial momentum theorem that causes the anomaly. Including this
effect into the equations show that the wind power coefficient, in contrast to what is
found from the momentum theory results of 1), indeed tends to zero at small tip speed
ratios.
3. Another part of the investigation concerns optimum behavior of rotors with constant axial
induction. We here found that, in contrast to the results shown in the report of de Vries
(1979), optimizing a rotor assuming constant axial induction give the same power
performance as in the analysis of Glauert (1935).
4. Finally, we show that neglecting expansion both remedies the problem of using the axial
momentum equation and increases the wind power coefficient at small tip speed ratios, as
compared to the model of Glauert (1935). Thus, assuming constant circulation and
neglecting expansion results in a rotor design that performs better the Glauert model and
that is physically consistent
3.7 Paradoxes and "paradoxes" when extending the one-dimensional momentum theory to
two-dimensional case (V.L. Okulov’s comments to Chapter 3)
First of all, let us return to Froude’s theory and try to understand why it gives correct results.
Let us remind to the reader that this is a one-dimensional theory which describes a variation of
some integral flow characteristics along a single axis. However, this is done in strict accordance
with the one-dimensional conservation laws. That gives a correct estimation for the Betz-
Joukowsky limit or the acceleration/deceleration of the averaged axial flow with doubling of the
axial induction factor in far wake. Moreover, such simplified model of the turbine in the form of
a pressure jump describes a propagation of long-wave pressure perturbations in pressure pipes
correctly, and even predicts their resonant properties (Okulov, Pylev, 1996). The resonant
frequencies of pressure systems at hydroelectric power plants were calculated by using of the
simple rotor theory and were found in a good agreement with the test data of the plant.
Undoubtedly the one-dimensional flow is a very rough approximation which could not
describe any radial distribution of forces or velocities in the rotor plane. A variation of radius of
the control stream tube along the flow axis can be defined by this theory and continuity equation
that sometimes is erroneously perceived visually as two-dimensionality (Fig. 2.2), but the data is
useless for the rotor design. The use of blade element momentum theory agrees to profile design
but in the different blade sections without an interaction between them which forms a distribution
of characteristics along the rotor blades. The distributions should be found by the vortex theory
of the rotor, formulated at the beginning of the XX century, but at that time the scientists could
not solve the task and instead they return back to consideration of the rotor as an actuator disc.
As a result a general momentum theory for the rotor, first presented by Joukowsky (1912-
1918) in his last article of this series, came into the scientific world. His understanding of the
relationship between the actuator disc theory with the distribution of the pressure jump, and his
own vortex theory with replacement of the rotor by the equivalent "vortex" disc with a
continuous distribution of the circulation (see the conclusion after equation (3.23)) served a
starting point in creation of the general momentum theory. However, here we must realize fully
that, in general, the transition to the two-dimensional model assumes a solution of the differential
axisymmetric Euler or Navier-Stokes equations that is possible only numerically in contrast the
integral equations. But at that time there were neither contemporary numerical methods nor
computers for their solution, while the application of the integral analogues, i.e. use of
conservation laws, that were so effective in the one-dimensional theory, in order to obtain the
distributed characteristics for general case is very difficult. Therefore, to obtain the desired radial
distributions, one had to make simplifying assumptions, for example, assuming the flow axial
41

induction coefficient, or circulation, or both, to be constant. Typically, these assumptions brought
to a good approximation, though they were not the solutions themselves that sometimes,
especially for limit cases, gave absurd, unreal results that resemble paradoxes.
The original article that gave the basis to this Section, was just devoted to a systematic
analysis of the paradox on unlimited increase of the power coefficient Cp for low-speed wind
turbines, arising in Joukowsky’s general momentum theory under the assumption of a constant
distribution of circulation along the disk radius. It is necessary to note that Joukowsky, when
describing this theory, noted a strange fact at the end of the paragraph 7 of his 4th article
(Joukowsky, 1912-1918) that at the transitional mode, when the screw starts working as low-
speed mill, the relation for doubling the flow axial induction in the far wake is lost. This was
shown in Fig. 3.5, and as shown above, implies unlimited growth of Cp. To resolve this paradox,
one should take into account that the distribution with constant circulation is not an exact
solution. In respect to the propeller, this fact was actually established and eliminated in the great
work by Keldysh and Frankl (1935). In this article, the authors indicated that Joukowsky, having
formulated his theory for vortex disk, has found the solution in the approximation of just
cylindrical configuration of the vortex sheet in the wake, without consideration of its deformation
(in accordance with Section 3.5.3). The wake deformations give the exact solution to the screw
problem, which reduces to a system of integro-differential equations which was solved by the
method of successive approximations. Joukowsky’s simplified solution (Section 3.5.3)
corresponds to one of the first members of the successive approximations, which, as it was
proven, converges to the exact solution. However, the exact solution for the distribution of bound
vortex strength r  in case of a vortex wake deformation (or deformation of the control stream
tube, as discussed in Sections 3.3 and 3.4) should not be constant and relate to the class providing
the finiteness of the functions along the radial direction
r  R  r
and r 3 R  r , (3.86)
r
that authors also used it in a form of a requirement of an analyticity from the functions
r  r 
and (3.87)
r2 Rr
Both conditions (3.87) correspond to zero strength of the bound vortex when approaching the
hub ( r  0 ) and on the edge of the disk ( r  R ). The above analysis considered in 3.4.4 to
address the paradox on unlimited growth of Cp actually leads to the same conclusion. Since the
requirement of zero values of the circulation on the edge of the disc looks like an equivalent to
the introduction of shear layer within the expanding boundary of the control stream tube. In this
respect, we can say that the above analysis of general momentum theory for wind turbines with
an expanding control stream tube is confirmed by the evidences of the classical work by Keldysh
and Frankl (1935), applied to the analysis of the screws. Of course, in the future, we would like
to find closest fit between these theories, but even now it can be stated that the explanation of the
paradox about abnormal Cp growth for slow-rotating wind turbines in Joukowsky’s general
momentum theory has been already found.
We should note an existence of another type of "paradoxes" that arise when extending the
momentum actuator disc theory to the two-dimensional case. In contrast to the Joukowsky’s
model above analyzed, the "paradoxes" of many other theories are caused by their own inherent
errors, rather than by assumed simplifying approximations. A striking example of such an error is
the so called GGS limit (Gorban, Gorlov, and Silantyev, 2001), which, according to the authors,
42

should replace the Betz-Joukowsky limit for more general case of the non-uniform pressure
distribution in the actuator disk. In fact, the authors present the solution of a very well-known
problem for optimization of filtration through a porous plate using a modified Kirchhoff method,
as the desired result, and even do not make any efforts to extend it to the disc. But it is obvious,
that the main problem here is caused by incomprehension of the rotor operation principle, shown
by the authors. It is clear that the vector of bound vorticity, producing the work, should be
oriented along the blade. That means for wind turbine with a horizontal axis it should be directed
radially while for the wind turbines with a vertical axis it aligns to the axis with a vertical
direction. In both cases that could not be an azimuthal direction, as it was put in GGS case, when
vorticity appears due to the flow around the disc/plate. Even with the reasonableness of the
author comment that whole fluid can not flow freely through the rotor disk, and should be
rejected by the blades radially, the authors have lost in their GGS model the main component of
the vorticity on the disk, which produces useful work. By focusing on the description of a minor
detail the authors were not able to determine the optimal wind power coefficient. They just did
not have the active forces on the disc, and their GGS limit has nothing to do with the Betz-
Joukowsky limit.
Another example is when at the transition to the general two-dimensional model the pressure
jump or bound circulation of the main nature of the loaded rotor has been neglected completely.
The example is related to the modeling of wave propagation in pressure systems of the
hydroelectric power plants. In the previously mentioned article (Okulov, Pylev, 1995, there is a
misprint in the characteristic equation (6): the minus sign should be replaced by plus sign -
commented by V.L. Okulov) the resonant frequency of 1.584 Hz calculated by one-dimensional
model with the pressure jump as the turbine action correlates well with the typical frequencies of
1.4-1.45 Hz of pulsation growth detected by commissioning tests in the pressure system of the
Sayano-Shushenskaya hydro-electric power plant. On the other hand for more complex two-
dimensional model when the impeller was modeled by a circular cascade of profiles the pressure
jump for the wave development was ignored because it seems quite reasonable to neglect its
effect, taking an assumption that the small-scale obstacles do not affect on long waves
propagation (Kurzin, 1993). Quite reasonable approximation of the long-wave theories leads here
to the infeasible paradoxical result that one of the resonant frequencies of the pressure system in
such a model is the same as the rotational speed of the impeller equal to 2.5 Hz. This "paradox"
as well as the previous one, is caused by neglecting the effect of impeller work as the loaded
cascade, though for unloaded small-scale obstacles of the cascade-type the model seems to be
valid.
It should be emphasized again that the one-dimensional Froude theory is well known exactly
by the fact that it described the useful pressure jump on the rotor as a cornerstone and its main
properties. It does not allow one to introduce and accept the secondary flow characteristics as the
main ones, as was done in both above attempts.
Even more difficult to understand the situation in which an author announces the power
coefficient greater than the ideal limit of Betz-Joukowsky, for example, equal to 72% in his
experimental study (Gorelov, 2012). It seems that the author describes the ideal method for
determining of this limit for Darrieus wind turbine model with vertical axis. The experiment was
carried out for a turbine model rotating and moving uniformly in horizontal direction into
immovable water. The power was measured for both rotating rotors with blades and without
them. Further, based on the viewpoint, quite reasonable at first sight and valid for any linear
system, the authors define the power coefficient through a difference between these two powers,
believing that this difference is exactly equal to the power developed just by rotor blades without
loss on rim wheels and the rotor shaft. However, the flow around the impeller with blades and
without them has a complete different nature, and their linear composition or decomposition is
not acceptable. It is well known that the flow between two coaxial rotating disks has a very
43

complex vortex nature accompanied by the vortex breakdown on the axis (Okulov, Sørensen,
Voigt, 2005). This phenomenon induces an additional vortex resistance of the system. However,
small perturbations of this flow completely reconfigure its structure, and the vortex breakdown
disappears or turns into other type. The introduction of the blades into the two-disc system,
certainly, is a very strong perturbation, not only adding to the system useful work from the flow
around the blades, but also strongly affecting the central vortex by changing the vortex resistance
of the disk- disc system.
In fact, for a correct estimation of the power coefficient depending only on the effect of
blades the power characteristics of each blade must be measured, or more correctly the power
characteristics of the whole impeller should be determined. In fact, for ideal turbines with a
vertical axis, the limiting value of the power coefficient is a little higher, around 64% (Madsen,
1983), but this is not due to inaccuracy of the Betz-Joukowsky limit. This is due to the fact that
fluid particles of the same stream tube influence on the blade twice during a full rotor turn in
such turbine. The Betz-Joukowsky limit was derived only for a single impact. Note that both
limiting values are valid only for ideal impellers. If we take into consideration the fact that for
turbine with a vertical axis the blades move against the initial flow much during the operation
with a considerable resistance, than the actual (experimental) values of the maximal power
coefficient for them should be always essentially lower than it for the turbines with a horizontal
axis. That explains the absolute predominance of the horizontal design in wind power industry.
Certainly, many other attempts to revise and change the Betz-Joukowsky limit have been
made for nearly a century since it was introduced. Nowadays, it is even impossible to list all of
them and all the more it is impossible to describe all these supposed "paradoxes" of erroneous
one-dimensional or two-dimensional disc and cascade rotor theories. In the previous Section, we
described two very old misapprehensions concerning the definitions based on one-dimensional
momentum theory, and here we have reviewed three new examples that are most popular now,
and today require explanations of their erroneous points.
Undoubtedly, more accurate results should be expected from the vortex theory of a rotor with
a finite number of blades, as it described below.
44

4 Maximum efficiency of wind turbine rotors using Joukowsky and Betz approaches
Based on the concepts outlined by Joukowsky nearly a century ago, an analytical

aerodynamic optimization model is developed for rotors with a finite number of blades and
constant circulation distribution. In this Chapter we show the basics of the new model and
compare its efficiency with results for rotors designed using the optimization model of Betz.

In the history of rotor aerodynamics two ‘schools’ have dominated the conceptual
interpretation of the optimum rotor. In Russia, Joukowsky (1912-1918) defined the optimum
rotor as one having constant circulation along the blades, such that the vortex system for an Nb-
bladed rotor consists of Nb helical tip vortices of strength  and an axial hub vortex of strength -
Nb  . A simplified model of this vortex system can be obtained by representing it by a rotating
horseshoe vortex (see Figure 1, a). The other school, which essentially was formed by Prandtl
and Betz (see Betz 1919), assumed that optimum efficiency is obtained when the distribution of
circulation along the blades generates a rigidly helicoidal wake that moves in the direction of its
axis with a constant velocity. Betz used a vortex model of the rotating blades based on the lifting-
line technique of Prandtl in which the vortex strength varies along the wingspan (Figure 1, b).
This distribution, usually referred to as the Goldstein circulation function, is rather complex and
difficult to determine accurately (Goldstein 1929). In both cases only conceptual ideas were
outlined for rotors with finite number of blades, whereas later theoretical works mainly
concerned actuator disk theory. Hence, in practice, the blades are modelled using Blade–Element
Momentum (BEM) theory, corrected by the tip correction of Prandtl (see e.g. Glauert 1935).
Figure 4.1. Sketch of the vortex system corresponding to lifting line theory of the ideal propeller
of Joukowsky (a) and Betz (b).
Recently in (Okulov & Sørensen 2008a, b), we have derived an analytical solution for rotors
with Goldstein distributions of circulation along the blade (Betz rotor) using a new analytical
model of the velocity field induced by helical vortices (Okulov 2004). In the present work we
exploit the analytical model further to develop a vortex theory of an ideal rotor based on the
concepts outlined by Joukowsky using constant circulation along the blades (Joukowsky rotor).
Both solutions enable for the first time to compare the theoretical maximum efficiency of wind
turbines with Betz and Joukowsky rotors
4.2. Vortex theory for rotors with a finite number of blades

In the vortex theory each of the blades is replaced by a lifting line on which the radial
distribution of bound vorticity is represented by the circulation   (r ) which is a function of
45

the radial distance along the rotor blade. This results in a free vortex system consisting of helical
trailing vortices, as sketched in Figures 4.1a and 4.1b. Using vortex theory, the bound vorticity
serves to produce the local lift on the blades while the trailing vortices induce the velocity field in
the rotor plane and in the wake. As illustrated in Figure 4.2 the velocity vector in the rotor plane
is made up by the rotor angular velocity  0 , the undisturbed wind speed U  , the axial and
circumferential velocity components u z0 and u0 , respectively, induced at a blade element in the
rotor plane by the tip vortices, and v0 , the circumferential velocity induced by the hub vortex.
The fundamental expressions for the forces acting on a blade (Figure 4.2) is most conveniently
expressed by the Kutta–Joukowsky theorem, which in vector form reads
Figure 4.2. Velocity and power triangles in the rotor plane

of Joukowsky rotor (a) and Betz rotor (b).
The fundamental expressions for the forces acting on a blade (Figure 4.2) is most
conveniently expressed by the Kutta–Joukowsky theorem, which in vector form reads
  
dL  V0   dr , (4.1)
where dL is the lift force on a blade element of radial dimension dr , V0 is the resultant relative
velocity and  is the density of the air.
From equation (4.1), we can write the local torque dQ of a rotor blade as follows
 
dQ    U   u z0 rdr . (4.2)
Integrating equation (4.2) along the blades and summing up, we get the following expression for
the power output, P  0Q ,
 
R
P  N b  0   U   u z0 rdr , (4.3)
0
where R is the radius of the rotor.

To determine the theoretical maximum efficiency of a rotor the wind power coefficient is
introduced as follows
46

C P  P /( 12 R 2U 3 ) . (4.4)
The maximum power that can be extracted from a stream of air contained in an area equivalent to
that swept out by the rotor corresponds to the maximum value of the wind power coefficient.
This is determined as function of the tip-speed ratio
 0  0 R /U  . (4.5)
To determine the velocity field v0 , u z0 and u0 induced at a blade element in the rotor plane
the free half-infinite helical vortex system behind the rotor is replaced by an associated vortex
system that extends to infinity in both directions. Neglecting deformations or changes in the wake
the vortex system is uniquely described by the far wake properties in the so-called Trefftz plane,
which per definition is the plane normal to the relative wind far behind of the rotor. Thus, in
accordance with Helmholtz’ vortex theorem, the bound circulation  about a blade element is
uniquely related to the circulation of a corresponding vortex in the Trefftz plane. By symmetry, it
is readily seen that the induced velocities at a point in the rotor plane (Figure 4.2) equals half the
induced velocity at a corresponding point in the Trefftz plane (see e.g. Joukowsky 1912-1918;
Betz 1919).
v0  12 v ; u0  12 u and u z0  12 u z . (4.6)
4.3. Solution of Joukowsky rotor

In the vortex theory of the Joukowsky rotor (Joukowsky 1912-1918) each of the blades is
replaced by a lifting line about which the circulation associated with the bound vorticity is
constant, resulting in a free vortex system consisting of helical vortices trailing from the tips of
the blades and a rectilinear hub vortex. The vortex system may be interpreted as consisting of
rotating horseshoe vortices with cores of finite size, as sketched in Figure 4.1, a which is
reproduced from the original drawing Joukowsky. The associated vortex system consists of a
multiplet of helical tip vortices of finite vortex cores (   R ) with constant pitch h and
circulation  . The multiplet moves downwind (in the case of a propeller) or upwind (in the case
of a wind turbine) with a constant velocity U  1   in the axial direction where  denotes the
difference between the wind speed and axial translational velocity of the vortices. Denoting the
angle between the axis of the tip vortex and the Trefftz plane as  (see Figure 4.2, a), the helical
pitch of the multiplet is given as
h  2R tan  , or l R  h 2R  tan  . (4.7)

The free vortex lines are made up by vortex cores of finite size in order to avoid singular
behavior. The vortex cores are collinear to the axes of the helical lines and their vorticity is
assumed to be uniform and densely distributed across the core cross-section. According to
(Okulov 2004), in cylindrical coordinates (r, θ, z), the components of fluid velocity induced by
N b helical vortices in the domain outside the vortex cores are given as
N b  1 R Nb   I m mr l K m mR l 

u z r ,       m  cosm n  , (4.8)
2l 0 l 2 n1 m1  I m mR l K m mr l 
N  0 R N b   I m mr l K m mR l 
u r ,    b     m  cosm n  , (4.9)
2r 1 rl n 1 m 1  I m mR l K m mr l 
47

z 2 n  1
where I m  x  and K m x  are modified Bessel functions;    
and  n    . When
l Nb
the first two dominant singularity terms are extracted, Equations (4.8) and (4.9) are reduced to
the following rough-and-ready formulas (Fukumoto & Okulov, 2005)
   e i n 
  N b  4 l 2  R 2 l  3r 2  2l 2 3R 2  2l 2 
 
Nb
u z r,       Re     ln 1  e i n 
2l   0  4 l 2  r 2 e  e
i n
24 l  r  2 l  R  2
 2 2
3
2 2
3
  

n 1
   
(4.10)
   e i n 
   0  4 l 2  R2 l  3r 2  2l 2 
3R 2  2l 2
  
Nb
u r,     ln 1  e i n
 
2r   N b  4 l 2  r 2
 Re e   e 
 i n
 
24  l 2  r 2 3 2 l 2  R 2 3 2  

n 1
   
(4.11)
  
r 1  1  R 2 l 2 exp 1  r 2 l 2  (unfortunately, this expression was printed with an
where e  
R 1 
1 r 2 l 2 exp 1  R l 
2 2
error in articles by Fukumoto & Okulov, 2005 and Okulov & Sørensen 2008a, b and eqs (4.10-
4.11) was wrote with an incorrection in Okulov & Sørensen 2010). Here we use notations “  ”
and “ : ” in which the upper sign or symbols in brackets corresponds to r  R and the lower to
r  R . The velocity component in the  - direction is given as (Okulov 2004)
r
u  u  u z . (4.12)
l
2
1
2 0
Introducing the azimuthally averaged induced axial velocity as u z 
 u z d , from
Equation (4.8) we get
Nb
uz  0 for r  R and u z 
 const for r  R . (4.13)

2l 
It should be mentioned that the dimensionless averaged induced axial velocity in the wake
( 0  r  R ), which is identical to the total axial wake interference factor a , takes the same
constant value
Nb
aU   u z.  (4.14)
2l  0 r  R
The vortex system also includes a rectilinear hub vortex of strength  N b  , resulting in a
simple formula for the additional induced velocity that only consists of the circumferential
component,
Nb
v   , (4.15)
2r
Defining the azimuthally averaged azimuthal velocity induced by the helical multiplet as
2
1
2 0
u   u  d , and combining equations (9) and (15), we get
48

v r R
  u  rR
. (4.16)
To eliminate the singularity of the induced velocity field in the vicinity of the vortex
filament the vortex system is represented by a set of helical vortices with finite core. For an
unexpanded wake originating from a rotor with infinitely many blades, the convective velocity of
the vortex system equals half the averaged induced axial velocity in the wake. This is sometimes
referred to as the ‘roller-bearing analogy’. Although this approximation cannot be rigorously
justified for a vortex system consisting of a finite number of vortices, we employ the same
analogy by assuming that the helical vortices are transported with a relative axial speed,  , that
corresponds to half the averaged induced velocity,
1 R   
 a , (4.17)
2 R
where a correction of small expansion of the cross-section of the wake is made in order to
include the radius,  , of the vortex cores. Thus, the vortices are assumed to translate in the bi-
normal direction to the helical axis of the tip vortices with the velocity, u b , (Figure 4.2, a)
a R    a R    R
u b   cos   cos   . (4.18)
2R 2R R l2
2
The problem of determining the induced equilibrium motion of a multiple of helical vortices in
an unbounded domain was solved in (Okulov 2004) but here we will use a more suitable notation
introduced in (Okulov & Sørensen 2007). The motion of the tip vortices in the  - direction can
be described by Equation (2.13) of (Okulov & Sørensen 2007) in the form
4R 1  2 N b
u     u       2 
  
(4.19)
1  N 1   2 3 2 1    4 3   3
  ln b
      3   2
2
1   2    4  1   2 7 2 
12
8  Nb
where   l R and    R are the non-dimensional pitch and radius of the vortex core,
respectively, and (3) = 1.20206… is the Riemann zeta function. Introducing the bi-normal
velocity component, u b , as
l
u b  u  , (4.20)
R l2
2
the conditions for equilibrium motion of the far wake is now determined in accordance with the
‘roller-bearing analogy’ of (4.18). Finally, for this wake motion the dimensionless radius  of
the tip vortex core must satisfy the equation
N
u      2b 1    . (4.21)

For any given value of pitch l and number of rotor blades N b , we compute the radius of the
tip vortex core by solving Equation (4.21). Figure 3a shows the “total” core size, which is equal
to the vortex core radius multiplied by the number of blades, as function of the tip vortex pitch
for different numbers of blades. It may be noted that the limit of small pitch has the same
asymptote for all number of blades and the “total” core tends to zero faster than the pitch with a
h
fixed ratio  6 . This implies that the distance between the tip vortices, independent of the
N b
value of the pitch, always is greater than 4 core radii, as depicted in Figure 4.3. This also implies
that it is impossible to reach a dense cylindrical vortex surface as used in actuator disk theory. In
49

the other limit, when the pitch tends to infinity, the vortex core disappears which shows that an
equilibrium vortex multiplet motion is achieved when the vortices become rectilinear.
Figure 4.3
Joukowsky rotor (a): The vortex core radius for equilibrium motion of tip vortex multiplet as function of
helical pitch for different numbers of blades (here and hereby on the next figures the number on the solid-
line curves refer to the number of blades); Doted line indicates asymptotic behaviour of the core radius for
small pitch and the sketch shows the helical tip vortices representing the limit case of the wake;
Betz rotor (b): Examples of Goldstein function for 3 blades and different values of helical pitch.
The axial velocity field induced by the tip vortices can subsequently be determined in all
points of the Trefftz plane because we have defined the finite radius of the tip vortex core by the
‘roller-bearing analogy’. The velocities outside the vortices are determined by (4.8) and inside
the vortex cores by taking the average value of the velocity on the boundary of the vortex cores.
The axial velocity, made dimensionless with the azimuthally averaged induced axial velocity
u z  of (4.13) in the n-blade direction (   2n N b , z = 0), takes the form:
  2n 
 u z  r ,  if r  R   and r  R  
~  2n  1   N b 
u z  r ,    (4.22)
 N b  aU   R    r u  R  , 2n   R    r u  R  , 2n  if R    r  R  
z
N b 
z
 2  2  N b 
From Equation (4.14) we get the following relation between the bound circulation and the
interference factor,
N b   2laU  . (4.23)
From simple geometric considerations in the rotor plane (Figure 4.2a), using Equations (4.16)
and (4.17), the angular pitch is given as
U   u z0 U  1   U  1  12 a 1    l
tan  r  R  r R
     .(4.24)
 0 R  u 0  v0 0 R 0 R R
r R r R
Equation (4.24) can be also written as
 0 l  U   12 aU  1    . (4.25)
Inserting Equations (4.6), (4.22), (4.23) and (4.25) into Equation (4.3), the power can be
determined from the following integral
50

 
1
 a
P   R 2 U 3 a1  1   1  a  u~z  x,0xdx  . (4.26)
 2  0 
Performing the integration and introducing the dimensionless wind power coefficient (4.4), we
get
C P  2a 1  12 a J 1 1  12 a J 3  , (4.27)
1
where J 1  1   and J 3  2  u~z x,0 xdx . For a given helicoidal wake structure, the wind power
0
coefficient is seen to be uniquely determined, except for the parameter a . Differentiation of CP

with respect to a yields the maximum value, C , resulting in
P , max
a(C P  C P ,max ) 
2
3J1 J 3
 
J 1  J 3  J 12  J 1 J 3  J 32 . (4.28)
4.4. Solution of the Betz rotor

To compare the efficiency of the Joukowsky rotor with the Betz rotor, we here outline the
main points of the derivation of the aerodynamics of the Betz rotor (for more details we refer to
Okulov & Sørensen 2008a). In this model, which is based on Lanchester-Prandtl wing theory, the
vortex strength of the lifting line varies along the blade span, following the so-called Goldstein
distribution. This results in a vortex sheet that is continuously shed from the trailing edge (Figure
4.1b). Betz (1919) showed that the ideal efficiency is obtained when the distribution of
circulation along the blade produces a rigidly moving helicoidal vortex sheet with constant pitch,
h , that moves downwind (in the case of a propeller) or upwind (in the case of a wind turbine) in
the axial direction of its axis with a constant velocity U  1  w . The associated vortex system to
the wake consists of a regular helical sheet extended to infinity in both directions. Denoting the
angle between the vortex sheet and the Trefftz plane as  (see Figure 4.2, b), the pitch is given
as
h  2r tan  or, l r  h 2r  tan  (4.29)

where r is the radial distance along the sheet. Since the sheet is translated with constant relative
axial speed, wU  , the induced velocity comprises only the component wU  cos  that is
‘pushed’ normal to the screw surface (Figure 4.2b). The axial and circumferential velocity
components u z and u induced by the infinite sheet at the sheet itself are therefore given as
u  wU  cos  sin  and u z  wU  cos 2  . (4.30)

From simple geometric considerations these equations are rewritten as
xl x2
u  wU  and u z  wU  2 , (4.31)
l 2  x2 l  x2
where x  r R is the dimensionless radius.
Goldstein (1929) was the first who found an analytical solution to the potential flow problem
of the moving associated vortex system consisting of an infinite helical vortex sheet. In his model
a dimensionless distribution G  x, l  of circulation was introduced as follows
51

N b x, l   2l wU  G  x, l  . (4.32)
Using infinite series of Bessel functions, Goldstein (1929) succeeded in obtaining an analytical
solution to the problem, but for N b  2 and 4 only. For any given value of the wake pitch l and
number of rotor blades N b , the Goldstein circulation function G  x, l  , shown on Figure 4.3b, has
been determined in (Okulov & Sørensen 2008a, b).
To compute the wind power coefficient we employ the same procedure as outlined in the
previous Section, that is we integrate Equation (4.3) using Equation (4.6), and the lift distribution
in Equation (4.32). In addition to this, from geometric considerations in the rotor plane (Figure
4.2b), using Equations (4.29) and (4.30), the angular pitch is given as
U   12 u z U  1  12 w l
tan     . (4.33)
 0 r  12 u  0r r
The relation given in Equation (4.33) can be deduced as follows:
sin  U   ½u z U  (1  ½ w cos 2  )
tan     
cos   0 r  ½u  0 r  ½U  w cos  sin 
 0 r sin   ½U  w cos  sin 2   U  cos  (1  ½ w cos 2 ) 
 0 r sin   U  cos  (1  ½ w(cos 2   sin 2  )) 
sin  U  (1  ½ w)
tan    .
cos  0r
From which it follows that
U   ½u z U  (1  ½ w)
 .
 0 r  ½u 0r
Equation (4.33) can be also written as
0l  U  1  12 w . (4.34)
Inserting Equations (4.31), (4.32) and (4.34) into Equation (4.3), the power can be determined
from the following integral
 w
1
 w x2 
P  R 2U 3 w1    2G x, l 1   xdx . (4.35)
 2 0  2 x 2  l 2 
Performing the integration and introducing the dimensionless wind power coefficient (Equation
4.4), we get
CP  2w1  12 wI1  12 wI 3  , (4.36)

where
1 1
x 3 dx
I1  2  G  x, l xdx and I 3  2  G  x, l  2 2 .
0 0
x l
The coefficients I1 and I 3 are usually referred to as the mass coefficient and the axial energy
factor, respectively.
52

For a given helicoidal wake structure, the power and thrust coefficients are seen to be
uniquely determined, except for the parameter w. Differentiating of CP , Equation (36), with
respect to w yields the maximum value, C , resulting in
P , max
w(C P  C P ,max ) 
2
3I 3
 
I1  I 3  I12  I1 I 3  I 32 . (4.37)
4.5 Results and Discussion

In the following we present some representative results from the new model. To compare the
performance of rotors of constant circulation (Joukowsky rotor) with rotors optimized using the
Goldstein circulation distribution we show results from both models. To compare the efficiency
of the two rotor concepts it is needed to use some unambiguous parameters. As usual in rotor
aerodynamics, we employ the axial interference factor, a , and the tip-speed ratio,  0 . However,
since a does not appear explicitly in Equations (4.36) and (4.37) and  0 does not appear
explicitly in any of the equations, it is needed to derive some additional relations. In the case of a
Betz rotor,  0 is connected to the helical pitch l and the generic parameter w through Equation
(4.34), resulting in the following relationship,
0 R R  w 
0   1   . (4.38)
U l  2
For a Joukowsky rotor a similar dependency can be found from Equation (4.25),
0 R R  a   
0   1  1    . (4.39)
U l  2  R 
Figure 4.4 presents the optimum wind power coefficient of both models for different number of
blades as function of tip speed ratio. From the plots it is evident that the optimum wind power
coefficient of the Joukowsky rotor for all number of blades is larger than that for the Betz rotor.
The difference, however, vanishes for    or for N b   , where both models tend towards
the Betz limit.
Figure 4.4. Wind power coefficients, CP, of an optimum rotor as function of tip speed ratio
and number of blades. Left: Joukowsky rotor; Right: Betz rotor.
53

In the Betz model an expression for the axial interference factor can be obtained by
combining Equations (4.8) and (4.32),
1
a  w G x, l dx . (4.40)
0
In Figure 4.5 we display the axial interference factor of the two rotors as function of tip-speed
ratio and number of blades. Comparing the two plots it is readily seen that the Betz rotor for the
same tip speed ratio decelerates the flow less than the Joukowsky rotor. As a consequence, if we
employ the axial interference factor as independent variable, an optimum Betz rotor can produce
more power than a Joukowsky rotor under the same deceleration of the wind (see Figure 4.6).
Figure 4.5. Axial interference factor as function of tip speed ratio and number of blades.
Left: Joukowsky rotor; Right: Betz rotor.
Figure 4.6. Wind power coefficient, CP, of an optimum rotor as function of axial interference factor
and number of blades. Left: Joukowsky rotor; Right: Betz rotor.

An analytical model has been developed for a rotor with a finite number of blades and
constant circulation (“Joukowsky rotor”). The method is based on an analytical solution to the
problem of equilibrium motion of a helical vortex multiplet in a far wake. The vortex system
behind the rotor is represented by a set of helical vortices with finite core to eliminate the
singularity of the induced velocity field in the vicinity of each filament. The finite core radius is
determined in the framework of an ideal fluid by assuming that the relative wake motion is
governed by a constant axial speed equal to half the averaged induced velocity in the wake. The
54

main achievement of the model is that it eliminates the singularity of the solution at all operating
conditions. In contrast to earlier models, the new model enables for the first time to determine the
theoretical maximum efficiency of rotors with constant circulation and an arbitrary number of
blades.
Optimum conditions for finite number of blades as function of tip speed ratio were compared
for two models: (a) Joukowsky rotor with constant circulation along the blade, and (b) Betz rotor
with circulation given by Goldstein’s function. For all tip speed ratios the Joukowsky rotor
achieves a higher efficiency than the Betz rotor, but the efficiency of the Betz rotor is larger if we
compare it for the same deceleration of the wind speed.
4.7 Development of vortex theories of the rotor with a finite number of blades
(V.L. Okulov’s comments to Chapter 4)
The analytical solution of the NEJ rotor with a finite number of blades based on the vortex
theory of Joukowsky was perhaps firstly presented in the above research, surely, if we do not
referred to the recent and ongoing attempts of numerical and approximate methods and models.
A century ago in concluding part to the first article of the series (Joukowsky, 1912-1918) where
this problem was firstly formulated Joukowsky wrote: “The main idea of the bound vortices,
which forms the basis for this article, may permit performing all calculations on the basis of the
correct velocities of the relative liquid motion, but this analysis would be very complex”. It is
interesting to note that Ludwig Prandtl in his review (Prandtl, 1923), after a brief description of
the Betz vortex model, concluded: “In order to obtain a quantitative statement, we shall for
simplicity’s sake next think of a screw having a large (infinite – commented by V.L. Okulov)
number of blades”. In other words, both leading schools declined searching the solutions for
rotors with a finite number of blades. As a result, Joukowsky has created the general momentum
theory for wind turbines (Section 3), and Prandtl has found an approximate correction for a
transition from the disc theory to the rotor with a finite number of blades based on the solution to
the auxiliary problem about the flow around a cascade of semi-infinite plates. Until now this
Prandtl’s tip correction, written by him in appendix to the Betz article (Betz, 1919), is still the
main approach when designing the blades for the Betz rotors.
Figure 4.7. The diagram of Vetchinkin’s screw vortex system (1913).
Yuriev (1923) notes that the works with the Betz screw theory of the Prof. Prandtl’s school
came to Russia just after the First Civil War. On the other hand before this time V.P. Vetchinkin
55

(1913) has suggested the screw which was a quiet different from the NEJ rotor because it has an
arbitrary non-constant distribution of circulation along the blade. His diagram of the screw vortex
system is shown in Fig. 4.7. He used a variation method for calculating circulation at each vortex
element on the diagram. Certainly, the piecewise approximation by Vetchinkin directly does not
lead to the Betz’s optimal screw condition (Betz, 1919) formulated for continuous vortex sheets.
Note that the study of Betz was a logical extension of the Munk’s dissertation (another
postgraduate student of Prandtl), who has shown for the first time that an elliptical distribution of
load (circulation) along the wing of a finite span gives a minimal induction resistance (Prandtl
1918; 1919). Extending their results to the rotating rotor blade, Betz gave just the correct
formulation of a theorem about the optimal screw, though its proving was incomplete which was
explained in the article of (Frankl, 1939). N.N. Polyakhov has succeeded in correct and complete
proving of this theorem (1937); in the same research he has suggested to use the Ritz method for
numerical optimization of the Betz rotor.
Fig. 4.8. Goldstein S. and a fragment of the page with his solution.
Figure 4.9. Theodorsen prepares his equipment to demonstrate an electromagnetic analogy to measure the
Goldstein circulation function (left) and a fragment of the wake model mades by convolute metal foil (right).
Sidney Goldstein was the first who derived an exact analytical solution for the Betz rotor
with 1, 2, 3 and 4 blades by description of equilibrium motion of infinitely thin helical vortex
sheets (Goldstein, 1929). He found for the circulation distribution along blades in complex form
via trigonometric series based on modified cylindrical functions which links as Kapteyn type.
56

Later this circulation was named by Goldstein factor or Goldstein circulation function (Fig. 4.8).
Though the analytical solution has been derived, that was very difficult to use for calculations
when designing a concrete blade. Thus, instead of the complex calculation, T. Theodorsen began
measuring the Goldstein factor using an electromagnetic analogy (Theodorsen, 1948). Figure 4.9
shows, how he prepares the measurement equipment to demonstrate his method, and represents
also the electromagnetic model of the helical "wake" made of convolute metal foil.
We should also note an existence of another approach (V.S. Vozhdaev and E.S. Vozhdaev,
1997; 2001 and references therein) for description of non-uniform velocity field caused by the
finite number of blades. Instead of exact simulation of a vortex wake behind the rotor, the authors
upgraded the disk vortex theory of the screw by adding azimuth harmonics multiple of the
number of blades. The each mode was analytically integrated in close form along the semi-
infinite wake and in this case a classical axisymmetric disk solution considered as some zero
term of the expansion. The authors of this approach wrote that just a few first harmonics were
significant in the expansion, though they did not prove and test it. They have compared with the
disk theory only, i.e. with the first term of the same expansion and dropped any comparisons with
other well-known solutions for the helical vortex sheet (Goldstein, 1929) and single helical
vortex filament (Hardin, 1984) moreover, even references to these well-known solutions were
lacking in their works.
A complete analytical investigation of the original problem formulated by Joukowsky (1912-
1918) for the NEJ rotor with a finite number of blades and finite size of a core of the tip vortices
has been performed, perhaps, only in the article of (Okulov, Sørensen, 2010). The results of this
study are presented above in Chapter 4. An explanation of the long story was indicated above in
the citation of Joukowsky about would very complex analysis of the velocity induced by helical
structure of the rotor wake. Indeed before recently there was neither theory, nor analytical
solutions for the helical vortices with a finite radius of core, which, according to Joukowsky, are
the basic elements of the vortex wake in his scheme for the rotor (Fig 4.1a).
4.7.1. Helical vortex theory is a basis for new solutions in aerodynamics

Research on helical vortices has a long history, which goes back to the famous work of Lord
Kelvin (Kelvin, 1880), who considered the helical perturbations of the columnar vortex, i.e.
helical vortices with a large pitch. The helical vortices have a fundamental importance for the
fluid mechanics because they describe one of the main states of swirling flows. Examples of
vortex structures with a helical form described widely in the literature: tip vortices in the wake of
the rotors (propellers, screw, turbines and wind turbines), concentrated vortices in rotating tanks
and vortex devices; tornadoes; funnel in the fluid flowing out of the vessel; vortex structures
caused by the vortex breakdown behind the delta wing and in pipes, etc. Certainly, helical
vortices and their multiplets play a special role in the vortex theory of the rotor with a finite
number of blades. An accurate estimation of velocities in the rotor plane requires knowing
velocity field induced by the vortices of the helical form. These velocities are significant and
become a third from velocity of free stream at an optimal operating regime. In the retrospective
review on the airscrew theory developed in TsAGI, G.I. Maykopar (1969) pointed out that at the
increased speed of the flight, the mean velocity estimated by disk theories should be replaced by
a correct distributed velocity for a rotor with a finite number of blades, though a calculation of
that velocity depending from blades represents a most difficult part of the screw calculation. In
this context, the helical vortex theory is of special interest here. Therefore, starting with the
seminal work of Joukowsky (1912-1918), helical vortices have been always actively studied in
the applications of rotor aerodynamics.
57

Figure 4.10. Elementary vortex structures: a straight line
or a point vortex, vortex ring and a helical vortex (left to right).
Elementary models of helical vortex structures include endless canonical helical vortices or
vortex filaments associated with these structures, similarly to other fundamental objects of vortex
dynamics: a rectilinear (sometimes point) vortices and vortex rings (Fig.4.10). Among them,
infinitely long rectilinear vortex with zero value of curvature and torsion of vortex lines is the
simplest geometric object. Vortex ring is more complex object, since it has non-zero value of
curvature, whereas helical vortex has non-zero torsion in addition to the non-zero curvature.
These elementary structures also differ from the hydrodynamic point of view. Indeed, isolated
rectilinear vortex even with infinitely thin core lies without moving in an infinite undisturbed
space because its elements induce velocities in parallel planes having no effect on each other. For
vortex rings, the planes perpendicular to the circular vortex filament are crossed with a ring,
vortex elements affect to each other causing self-induction motion with velocity which is
perpendicular to the ring plane and proportional to the ring circulation and depends on the
curvature radius and vortex core thickness (Alekseenko, Kuibin, Okulov, 2003). In addition to
the straight line motion, the self-induced motion of the helical vortex is accompanied by rotation
as well, and its self-induced velocity additionally depends also on the torsion. Unlike the theory
of point vortices and vortex rings, the helical vortex theory was not systematically described in
the literature and was not considered at textbooks and monographs on classical fluid mechanics.
This was due to a lack of a simple closed theory of helical vortices that could be used to analyze
various hydrodynamic problems effectively. An attempt to fill this gap was made in a recently
published monograph (Alekseenko, Kuybin, Okulov, 2003). However, at a time of monograph
publication, the studies had not yet been finally completed, that did not allow authors to
formulate the basic theory of helical vortices. This was completed later in subsequent works of
(Okulov, 2003; Okulov 2004; Okulov, Fukumoto, 2004; and Fukumoto, Okulov, 2005)
Figure 4.11 The geometrical parameters of helical filament.

58

The infinitely thin helical vortex filament (Fig. 4.11), being the simplest model, is the
fundamental singular object of the theory of vortices in inviscid incompressible fluid, similarly to
a straight line filament or a point vortex and infinitely thin vortex ring. However, in contrast to
these objects, the complexity of its mathematical description is that the Biot-Savart integral for
helical filament cannot be integrated in the final form. It cannot be represented as a simple pole
like that for a point vortex, or written down by means of the complete elliptic integrals, as in the
case of an infinitely thin vortex ring (Alekseenko, Kuybin, Okulov, 2003). Therefore Maykopar
(1969) notes that in the case of helical vortices the induced velocity was calculated by special
functions either through their integral representation, arising from the Biot-Savart integral, or by
means of series of Bessel functions. Results of attempts to conduct such direct calculations of
induced velocity happen to be not very accurate. In TsAGI, these results were compiled in
diagrams and tables (G.I. Maykopar et al, 1940). However, this calculation technique revealed
insufficient accuracy, for example, that become apparent in the asymmetry of the calculated
velocity profiles along the circular sections of the rotor plane (Figure 4.12, a).
а б
Figure 4.12. Examples of calculating the azimutal velocity, induced by helical vortex filament along
different circules: (a) direct calculation with the incorrect asymmetric profile (Maikapar, 1969);
(b) correct calculation taking into account the singularity separation (Fukumoto, Okulov, 2005).
а б в
Figure 4.13. Example of the flow calculation, inducing a helical vortex filament
in cross section using the presentation through series of the Kapteyn type:
(a) growthing coefficients of the series with increasing of harmonic
number m at limit to the position of the filament at r/R  x = 1;
(b) inaccurate flow calculation by a finite segment of the series;
(c) calculation after separation of the sinularities (Okulov, 2004).
59

Note that the problems concerned to a correctness of the calculations arise even after the
conversion of Biot-Savart integral into trigonometric series (4.8) and (4.9) of the Kapteyn type
(Hardin, 1984) similar to the classic Kapteyn series in an approximation of the rotor induction
factor by Lerbs (1952) or the Goldstein solution for the vortex sheet (Goldstein, 1929). The
classic Kapteyn series arise when the trigonometric functions in (4.8) and (4.9) are equal to
unities that correspond to calculation of the velocities along the radial direction passing through
the vortex filament in one-dimension radial direction. When computation point tends to the
vortex filament (( r  R or x  r R  1 )) the coefficients of the series (4.8) or (4.9) growth
with an increase of the harmonic number, providing a singular behavior of the solution in the
point of the filament (Figure 4.13, a). In the classic case the singularity was separated by Lerbs
(1952) and Morgan & Wrench (1965) though it gives a correct calculation of the induced
velocity just in one radial direction, passing through the vortex filament which gives a correct
simulation of the induction factor of an rotor with one blade. It is clear that a smooth regular
character of the solution along other directions that do not pass through the filament is provided
by alternating harmonics in the trigonometric series (4.8) and (4.9) of the Kapteyn type. To be
precise, apparently, the series should contain whole infinite multitude of harmonics since the
calculation of the velocity field by final segment of series gives pseudo mounts on a cylinder
which forms the filament (Fig. 4.13, b). An estimation of the velocity behavior in all direction
from filament is important for a transition from the filament to helical vortex with a finite core
and for correct estimations of an induction factor of several blades. The last reason restricted
applications of the radial approximation by Lerbs (1952) and his followers returned to direct
simulations of the Goldstein function. The high-precise tables of this function were counted by
IBM 7090 via direct calculation of the Goldstein solution in the form of Kapteyn series (Tibery,
Wrench, 1964). The cited work was related to design efficient and low noise marine propellers,
so this solution was tabulated for sufficiently low tip speed ratios, typical for shipbuilding, i.e.,
the solution was not complete, it was suitable not for the whole range of operating parameters of
the rotors used in other technical fields.
It is interesting to note that an incorrect calculation of induced velocities is caused by specific
behavior of the solution near the filament (Fig. 4.14), found yet by Joukowsky (1912-1918).
Figure 4.14. Fragment of the first Joukowsky’s article (1912-1918) with the asymptotic representation
of the singular behavior of the velocity in the vicinity of the helical vortex filament.
60

This asymptotic representation of the solution in the vicinity to a vortex filament via a sum of
pole, logarithm and regular component is usually associated with Batchelor (Batchelor, 1967).
However Ricca (Ricca, 1996) has found that this development was first obtained by his
countryman Da Rios in 1911; though the fragment represented in Fig. 4.14 shows that it was
derived independently by Joukowsky as well as an outcome when he considered helical vortices
in his first article of the series (Joukowsky, 1912-1918). Indeed, if we do not take into account
these singularities then, as shown above on Fig. 4.13, b, the calculation of the trigonometric
series (4.8) and (4.9) of the Kapteyn type would present some difficulties. On this occasion, in
1997, the mathematical problem No 97-18 was formulated and given the status of important
problem which needed to be resolved. Its solution was found in the following year (Boersma,
Yakubovich, 1998) by separation of the singularities from the series in a form of the solution for
the simple point vortex plus a logarithm from a Cartesian distance. With this separation, the
regular remainder is significant; therefore, the authors replaced it by an equivalent integral
representation and then calculated the integral numerically within a higher accuracy of eighth
significant digit (Boersma&Wood, 1999; Wood&Boersma, 2001). It should be pointed out that
more efficient solution to this problem is the selection of the singularity in a special form (4.10)
or (4.11), which takes into account the helical structure of the flow as well as depends on the
curvature and torsion. Otherwise, it can be said that the pole and the logarithm were not written
in the usual Cartesian system like it was done in (Boersma, Wood 1999), but in some distorted
coordinates. Note that such a possibility to solve the problem has been found by one of the
authors of the present book long before the formulation of the problem in 1997 (Okulov, 1993;
Okulov, 1995). A prototype for the solution can be found in Lerbs (1952) but he could not
describe of the behavior for all directions from helical vortex filaments. As a result of Okulov’s
approach, the approximation by ordinary functions of (4.10) and (4.11) was obtained to include
all an information about the solution, while the regular reminder was so small that it could be
ignored without harm or in seldom cases it may be also corrected by an additional two terms of
poli-logarithm types, as it was shown by Okulov (2004). A calculation of the velocity field
induced by the helical filament (Fig. 4.13c) eliminates errors, obtained by direct calculation based
on special functions (Fig. 4.12 b). It is important to note that this approach is universal. The
approximation in form of sum of the dipole, pole and the logarithm for the helical filament with a
continuous distribution of dipoles along axis takes the same form in the same distorted
coordinates (Okulov, Fukumoto, 2004). Both solutions are shown in Fig. 4.15.
Figure 4.15. Cross section of stream tubes for flows induced by monopole (top row) and dipole (bottom
row) helical filaments for different pitch values : h / a = 8; h / a = 2; h / a = 1 (left to right)
61

Figure 4.16. Comparison of the axial (left) and azimuthal (right) velocity profiles induced by helical
vortices: thin lines – infinetely thin vortex filament, thick lines - the vortex with the core of 0.3R.
Figure 4.17. Comparison of axial induced velocity uz: (symbols) analytical solution based on the two
terms of the Dyson’s development; (lines) high precision calculations based on the vortex method with
using of 67,000,000 vortex elements (Walther, at al 2007).
The next important question is how the velocity field induced by vortex with finite core can
be approximated by using the approximations (4.10) and (4.11) for an infinitely thin filament.
The helical vortex with finite core is important because it uses in the Joukowsky’s vortex model
of a rotor. The answer was found by Dyson’s development that was applied by him to describe a
vortex model of Saturn's rings (Dyson, 1893). For both cases of vortex ring and helical vortex the
solution outside of the finite vortex core can be written in form of a superposition of singular
solutions for vortex filaments with different distributions of multipole singularities along them
(Fukumoto & Okulov, 2005). Figure 4.16 shows a comparison between the velocities induced by
helical vortex with finite core (thick lines) and the infinitely thin helical filament (thin lines). The
difference is not very significant even for a very large vortex core selected. It should be noted
that the difference disappears completely as far as the core size decreases. Thus, solutions for the
filaments representing by formulas (4.10) and (4.11) are quite enough as good approximation for
a rotor optimization. A numerical calculation (Walther, at al 2007) served an additional
verification of above conclusion, or, to be more specific, in this work a conclusion about using of
just two terms in the Dyson’s development for helical vortices has been tested by a calculation
based on vortex technique when one turn of helical vortex was replaced by 67 million vortex
elements. The results of both calculations completely coincided (Fig. 4.17).
It should be noted that the solution in the form of (4.10) and (4.11) for a single helical
filament has been successfully applied to determine the Goldstein circulation also (Okulov,
Sørensen, 2008; Okulov, Sørensen, 2008). A rotor vortex wake in the Betz model corresponds to
the helical vortex sheets moving in equilibrium.
62

Figure 4.18. Goldstein circulations for different pitch (left, h = 1, right, h = 1/12) and number of vortex
sheets (indicated by curve number); lines: calculation by representation of the helical vortex sheets using
100 vortex filaments; symbols: direct calculations by the Kapteyn series (Tibery & Wrench, 1964).
Figure 4.19. Goldstein circulations for rotors with rectileanar (symbols) and srew (lines) blades:
for different pitch (left, h = 1, right, h = 1/12) and number of vortex sheets (indicated by curve number).
Symbols is direct calculations by the Kapteyn series of Tibery & Wrench (1964).
The vortex sheet was discretized by 100 uniformly spaced helical filaments, provided the
condition of its motion together with fluid. The results completely coincided with the tabular data
calculated by the Goldstein solution (Fig. 4.18). New efficient algorithm developed for solving
the Betz rotor for any tip speed ratio. In addition, the discretization technique made it possible to
calculate curved blades often used for marine propellers. Comparison of Goldstein functions for
the curved blades and straight-line blades (Fig. 4.19) shows that the curvature is effective only
for the wakes with large pitch e.g. for ship propellers with a low rotation speed. The optimization
of the Betz rotor for curved blades can be now precisely counted by the vortex method. Semi-
empirical models and approximate approaches, such as, for example, presented in work by
Achkinadze (2003), has completely eliminated.
The final element of the theory of helical vortex, which is required for obtaining solutions for
NEJ rotor, is a definition of already mentioned self-induced motion velocity of the helical vortex
with a finite core. Self-induced motion of the helical vortices in practice is known as precession
of helical rope or the vortex core. Although helical vortex filaments, which make up the vortex
with a finite core, are a singular object, their singularity is integrable for a dense distribution of
the filaments in the core. In contrast to other elementary vortices the helical vortex solution is
impossible to be integrated in a closed analytical form. Therefore for a long time, various
asymptotic considerations for small and large values of the torsion or helical pitch (Kelvin 1880;
63

Levy, Forsdyke 1928; Widnall 1972; Kuibin, Okulov 1998; Boersma, Wood, 1999; etc) or
approximate approaches of desingularization, for instance, by simple cutting out the feature, etc.
(Thomson, 1883; Rosenhead, 1930; Crow, 1970; Batchelor, 1967; Widnall et al, 1971; Widnall,
Bliss, 1971; Moore, Saffman, 1972; Callegari, Ting, 1978; Fukumoto, Miyazaki, 1991; etc)
dominated here. Another technique based on an explicit separation of the singularity. The
simplest way is to separate a pole as for a point vortex and a logarithm from Cartesian distance
before integrating and the large regular remainder in this case has been determined by
numerically (Boersma, Wood, 1999). The next logical technique: the separation of the singularity
can be carried out by deduction of more complex object - associated vortex ring which self-
induced motion is well described (see Fig. 4.20).
Figure 4.20 Determination of the self-induced velocity of the helical vortex by vortex ring separation.
Traditionally it is considered that this approach was proposed by (Moore, Saffman, 1972),
while it was made in 1912 Joukowsky in his first paper of this series (Joukowsky, 1912-1918).
He demonstrated clearly that a formal representation for the self-induced velocity of the helical
vortex by Biot-Savart integral after some neglecting in the integral core is transformed into self-
induced velocity of associated vortex ring. We regret that the Western reader was not familiar
with such famous work of the Russian scientist because the approximation of the helical vortex
by vortex ring was offered by Joukowsky before 60 years then it was repeated by other scientists.
At first sight, this approach is more accurate than the first one, though the second sizeable regular
integral term still remains with very complex area of the integration. This term was neglected by
Joukowsky when he tried to obtain the analytical form of the solution by strong simplifications of
the integral core. The neglected regular term is not small as it was found by direct numerical
calculation (Ricca, 1994).
An interesting property was also found in the calculations by Ricca (1994). For all calculated
values of torsion or helical pitch of a helical vortex with a relatively small core of circular cross-
section, the self-induced velocity differs from the average velocity induced by vortex filament
concentrated along the vortex axis on the core border about the same constant. Later it was
theoretically proofed that this difference is exactly equal to one-quarter under certain velocity
normalization (Boersma, Wood, 1999). The same property is well known for vortex rings, though
there the difference value is equal to three quarters (Kuibin, Okulov 1998) at the same velocity
normalization. So a determination of the self-induced velocity of helical vortex can be reduced to
task already solved and based on determination of the velocity induced by infinitely thin vortex
filament on the core border (Okulov, 2004). The solution was reproduced above by (4.20) with
substitution of (4.19).
64

Figure 4.21 Comparison of the exact solution for the self-induced rotation velocity of helical vortex (solid
line) with its approximation through a vortex ring (dotted line), found by Joukowsky in 1912.
Figure 4.21 shows a comparison between the exact solution calculated by (4.20) and (4.19)
and Joukowsky’s vortex ring approximation (1912-1918). The difference can be explained easily
for example in limit of small pitch, when helical vortex turns into cylindrical vortex sheet, while
in the Joukowsky’s approximation the vortex structure remains ring with the same radius of the
vortex cylinder. In other words, it is clear, that two different objects should induce different
speeds.
4.7.2. Hypothesis of different models and comparison of proper calculations

It should be noted that the vortex approach for the rotors with a finite number of blades,
proposed by Joukowsky and Betz, are not sufficient for complete formulation of the problems.
Some additional assumptions need to close a formulation of both problems. First of all a correct
choice of the helical pitch in the vortex wake for both rotors should be done. There are three
reasonable assumptions for the wake:
1) the pitch is independent from velocities induced by the wake

U l
tan     ,
0 R R
2) the pitch depends on the velocities induced in the far wake u z , u :
 
U   u z l
tan     ,
 0 R  u  R
3) the pitch depends on the velocities induced in the rotor plane u z , u : 0 0
U   u z0 l
tan      .
 0 R  u0 R
Table 4.1 presents the rotor models which were based on these different assumptions. The
simplest first model was used in the first calculations of rotor. For some reason, this model was
considered as a good approximation for low loaded rotor, though it was used only to simplify the
rotor solution, since if the pitch of the screw would depend on the induced velocity, which itself
depends on the pitch, and the development of such solutions involves an iterative process, that
complicates a solution of the problem.

65

Table 4.1 Main assumptions underlying various models
Determination of the vortex Distribution of the circulation
Number of
Theory system pitch along the blade
blades
Betz-Joukowsky limit Actuator disc not specified not specified
Glauert calculation by
not specified not specified* not specified
section method
without correction for flow

Joukowsky rotor infinite constant
axial induction

Betz rotor infinite Betz distribution
axial induction

Prandtl correction finite Betz correction for distribution
axial induction

finite
Goldstein solution axial induction Goldstein’s function

Theodorsen solution finite
axial induction Goldstein’s function

Novel solution for NEJ finite constant
axial induction
rotor (Section 4.3)
Novel solution for Betz finite
axial induction Goldstein’s function
rotor (Section 4.4)
* axial and azimuthal induction factors are introduced in the rotor plane
According to the second model, the pitch was determined by the velocities in the far wake,
because some time ago investigators believed that the steady vortex structure with constant pitch
may form behind the rotor in the far wake only. The third and final model describes by (4.24) and
(4.33) and was introduced in (Okulov & Sorensen, 2008 and 2010).
For all the basic models listed above it was possible to introduce a single-valued parameter
for the rotor optimization. For the NEJ rotor it is an axial induction factor а which bases on the
azimuthally averaged flow. In the case of the Betz rotor it is w - additional axial velocity of a
moving of the vortex sheet which induced by this sheet. Both parameters are constant and
independent from radial position in the wake. All rotor models could be optimized by to these
parameters for each fixed pitch of the vortex wake. On the next step a transition from the pitch to
the operating parameter - tip-speed can be done in according with formulas (4.38) for the Betz
rotor, and (4.39) for the NEJ rotor. Results of the optimization of wind turbines for these three
models of the wake behind the Betz rotor are presented in Figs. 4.22 (the first model), 4.23
(second model) and 4.4 (third model). Comparison of their results with the Betz-Joukowsky limit
indicates absolute unacceptability of wake models 1 and 2. The third model considered above
confirmed by the Betz-Joukowsky limit without a glitch and also very well correlated with the
Glauert optimization based on blade element momentum theory (Glauert 1935), set above in
Section 3.5.1.
It is interesting to note that results of the first and second models, when rotor operates as
propeller, does not look so absurd it was indicated in the case of wind turbines (Figs 4.22 and
4.23). Since the rotor optimization for propeller regimes bases on analyzing of the ratio CT C P .
66

Figure 4.22. Optimization according to the model of weakly loaded rotor (wake model 1) introduced by
Betz and Goldstein. Horizontal dotted line - the Betz-Zhukovsky limit. Curcles - Glauert calculation
(Glauert 1935).
Figure 4.23. Optimization according to the wake model 2, introduced by Teodorsen. Horizontal dotted
line – the Betz-Zhukovsky limit. Curcles - Glauert calculation (Glauert 1935).
Indeed, if we would consider the ratio of CT C P and eliminate an abnormal behavior of the
coefficient CT at low tip speed ratios for the first model (Figure 4.22), then the ratio is roughly
the same for all three wake models. Since in the first model both coefficients are about two times
higher than the true values, and in the second model both ones are twice as little. Thus, when the
propeller was optimized by the ratio, all three ideal models give approximately the same value,
so it was quite difficult to identify the erroneous model for propeller operation modes.
When the rotor operates in the mode of a wind turbine, Betz-Joukowsky limit restricts values
of power coefficient, which should not be exceeded. That rule allows us to select the third wake
model (Figure 4.4) as a correct one. On the one hand, calculations according to this model never
exceed the Betz-Joukowsky limit, and on the other hand the model gives the values which are
very close to this limit by BEM theory (Glauert 1935). The choice of the third model completely
closes the problem for the Betz rotor, though for the NEJ rotor, the selection of wake pitch is not
sufficient for a complete statement of the problem.
Returning back to the NEJ rotor model, we should remind that Joukowsky considers tip
vortices with the finite core size by using of an ideal fluid model. In this regard, the following
questions remain open: how the tip vortex is formed on the blade, where exactly it is formed, and
what is a size of the vortex core. For the system associated with wake and consisting of doubly-
infinite helical vortices (item 4.2), the questions can be reformulated as follows: how should the
67

radius of the vortex multiplet coincide to the rotor radius, and what should the size of the core for
the tip helical vortices be chosen.
The first question can be reduced to a simple approximation with a search between two basic
positions, e.g.
A) the multiplet radius is smaller than the rotor radius by the vortex core radius;
B) the multiplet radius coincides with the rotor radius.
This problem is not a principal hypothesis because it just slightly specifies an actual radial size of
the multiplet. In fact, the proposed options differ from each other by a small amount, equal to the
radius of the tip vortex which size is significantly smaller than the multplet size.
The second question should not be answered in an ideal fluid model and for this reason it was
reduced to postulating the velocities of tip vortex moving and the core radius of the helical vortex
in the multiplet was chosen to which coinciding to this motion of the tip vortices by using the
formula (4.19) for self-induced and induced motion of helical multiplets. After substituting
postulated velocity of the multiplet in (4.20) instead of the self-induced velocity the vortex core
size was determined by solution of equation (4.21). The following reasonable assumptions were
tasted in respect to the multiplet motion:
а) the azimuthal component of the self-induced velocity of the tip vortices in the multiplet
completely compensate by the azimuthal velocity swirled in the other direction by hub
vortex, see the condition (4.16);
б) the relative motion of the multiplet in the axial direction flows in accordance with the
"bearing" analogy under a half of the axial induction factor.
The idea to define the vortex core by self-induced velocity of the multiplet again goes back to
Joukowsky (1912-1918). In the early XX century this idea could not properly rise due to the lack
of experimental information and the rough approximation of the helical vortex motion by the
formula for vortex ring (Fig. 4.21). Indeed, Joukowsky (1912-1918) proposed a model of the
"frozen-in" wake based on the photos of the wake behind screw propellers made by Flamm. That
suggests a rotation of the wake together with the rotor to maintain the steady-state helical shape.
The assumptions a) and b) stated above supports an opposite hypothesis to Joukowski’s proposal
about the strong rotation of the free tip vortices in the wake. They allow the free vorticity leaving
from tip to move in axial direction with flow velocity without rotation like bound vorticity on the
blades. According to b) these free vorticity is drifting downstream with a constant undisturbed
velocity (wind speed) minus the half-rate of the flow deceleration in the wake. That motion
results the tip vortices in a fixed helical form that looks similar to the "frozen-in" wake model on
photos. Though the wake approximation of the free vortex system by a) and b) did not include a
wake expansion recorded in the experiments. Just below, in the description of our recent
experiments, we will show that the conditions a) and b) are much more close to reality than the
"frozen-in" model. As the first step we will examine an effect of various deviations from the
proposed assumptions on the power characteristics of the rotor solutions.
Testing of the different assumptions for the NEJ rotor model, as it was made in the case of
the pitch definition between 1-3 models, was arranged by comparing with the Betz-Joukowsky
limit. As a result of the searching through various choices, it was found that any deviations from
the assumptions a) and b) for the multiplet velocity and option B) of its radial size conflicts with
the Betz-Joukowsky limit and give higher values for the power coefficient (Fig. 4.24). Thus, to
be agreeing with the Betz-Joukowsky limit the additional assumptions should be choiced in the
form:
3) The pitch of the wake vortex structure depends on the velocities induced in the rotor
plane: u z , u
0 0
U   u z0 l
tan     ;
 0 R  u0 R
68

Б) the multiplet radius coincides with the rotor radius;
а) the azimuthal component of the self-induced velocity of the multiplet is completely
amortized by the azimuthal velocity, swirled in the other direction and induced by the central
vortex, the condition (4.16);
б) the relative motion of the multiplet in the axial direction is taken in accordance with the
"bearing" analogy at the deceleration half-rate.
Figure 4.24. Examples of the test for the additional assumptions of the NEJ rotor at:
(top row) - calculation with the radial size of the multiplet by option A
and a deviation of its velocity from assumptions a and b;
(lower row) - calculation in accordance with the assumptions a and b and option B.
Figure 4.25. Numerical investigation of an influence of the wake extension on power coefficient: for
different number of blades: 1 (·), 3 (+) and 7 (×). (Left) Forms of the wake extenses in the axial direction
z. (Right) The difference between numerical calculation with the wake extension (Segalini & Alfredsson
2013) and a the exact solution of the Section 4.3 without the wake expansion.
69

The last question that should be considered for the analytical model 4.3 is estimation of an
influence of the wake expansion on the power coefficient. We can analyze this question using
numerical vortex model (Segalini & Alfredsson, 2013). Their result presented in Fig. 4.25
indicates very slight influence of the wake expansion on the power coefficient of the wind
turbine with derivation about 1-2%. Thus, the analytical model of Section 4.3 is quite correct for
calculations of the optimal NEJ rotors. The calculations give some justification for the additional
assumption to different rotor theories, though the final criterion here should be based on
experimental results.
4.7.3. The results of experimental tests
It should be noted that the experimental aerodynamics of the rotor developed more slowly
than the theories described above. Sometimes it served an obstacle in approving valid theories,
and often, on the contrary, an ambiguity of experimental data or an incorrect interpretation of
experimental observations, resulted to development of erroneous conceptions. As already
mentioned, the wrong perception and interpretation of experimental results by Parsons did not
allow asserting Froude theory about 30 years. The flow downstream the impeller is very complex
and requires special diagnostic techniques and methods. Nevertheless, first experiments were
performed using simple techniques, such as visualization of the vortex structure in the water
behind a screw by a small air bubbles, and determination of changes in flow structure by the
paper strips or hairsprings in the rotor wake. Flamm’s bubble visualization behind the screw and
Riabouchinsky’s hairspring visualization of the flow directions behind propeller gave Joukowsky
(1912-1918) the basic arguments to develop his vortex theory of screw propeller.
Then a long time average characteristics of swirling flow behind rotor have been studied by
applying various contact methods such as Pitot tube or hot wire anemometers and in modern
stage by using the non-contact velocimeters (LDA and PIV) (Ciocan, 2007). The “average” stage
of flow diagnostics had a positive effect to estimate the acceleration/deceleration velocity in the
wake for development of a classical disk theories of the rotor though an important information
about helical form of the vortex structure was lost the average approach if it compares with the
first visualizations. Knowledge about the average flows has stimulated a development of some
semi-empirical and engineering methods for rotor optimizations, but it certainly has been limited
for using in the development of correct theories and a verification of numerical simulations.
Perhaps that is why in the recent European project «MEXICO» (Snel, Schepers, Montgomerie,
2007) the attempt was made again to carry out a full-scale investigation of the flow downstream a
rotor model in wind tunnel. Although a processing of the experimental data is not yet completed
in full, the first comparisons of the measurements with numerical results (Shen, Zhu, Sørensen,
2011) did not show a very good agreement. Another recent and most comprehensive study of the
flow structure downstream the ship's propeller in the water was based on air-bubble
visualizations combined with LDA measurements of the flow field with a phase averaging in
several cross sections of the wake (Felli, Camussi, Di Felice, 2011). At present, the authors of
this book have initiated such comprehensive experimental study for a water model of wind
turbine. The first results have already been obtained and published (Naumov et. al., 2012) in
which three-dimensional distributions of the instantaneous velocity fields were obtained (Fig.
4.26). Stereo PIV system was used to determine the velocity field behind the wind turbine model
in a water channel. Tribladed rotor was designed to provide tip-speed ration λ = 5. Three images of
the Cartesian velocity components u, w, v are shown at the level of 13 lines with a uniform pitch.
Fig. 4.27 shows average profiles of the same velocities in the cross section at a distance of 0.4R
in front of the rotor (the squares), and at a distance of 2.5R behind the rotor (circles). Dots show
the profiles that correspond to half the sum of velocities in front of the blade and behind it.
70

Figure 4.26. Distribution of velocity components behind a rotor for the tip-speed ration λ = 5:
axial - u, azimuthal – w, and radial - v components (top to bottom).
1
u/V w/V v/V
0.4 0
0.667
0.2  0.2
0.333
0  0.4
0
r/R r/R r/R
0 0.5 1 0 0.5 1 0 0.5 1
Figure 4.27. Profiles of axial - u, azimuthal – w, and radial - v components of the velocity in front of
the rotor (squares), during motion of the rotor (dark circles), and in the far wake (light circles) for the tip
speed ratio λ = 5.
Surely, the investigations should be continued and completed, but some conclusions can be
summarized now. The measured data indicates a complex vortex system in the wake clearly. First
a distinct set of three tip vortices of the helical shape can be found on the plots. For the velocity
components u and v in the plane of a laser light sheet the cross-sections of vortex cores are well
defined by chain of dipoles, following from the blade tip downstream in the axial direction. Very
weak appearance of the vortex cores and sometimes their total absence for the third component w
connects with a disappearance of azimuthal velocity at a peripheral boundary of the wake where
the tip vortices exists (middle plot of Fig. 4.27). In additional to tip vortices for the distributions
of v and w components also a vortex sheet drifting down from the blade edge along all length is
well pronounced. That confirms by an explicit periodic variation of the velocities in axial
direction with a period equal approximately to pitch and inicates by pronounced radially
extended tongues. There is a concentration zone of flow rotation along the rotor axis that
corresponds to concentrated central columnar vortex in the wake behind the rotor.
71

It should be pointed out that the experimental results confirm many from main assumptions
of classical rotor theories. At first that concerns to the simplest Froude theory (Froude 1889),
which for the wind turbine model under the optimal operation at the tip speed ratio equal to 5
should provide a double deceleration of the axial velocity in the far wake as compared with the
deceleration at the rotor plane. Indeed, in Fig. 4.27, u/V at the rotor plane (dark circles) become
slower at 1/3, and in far wake - at 2/3 as it is predicted by Froude theory. The next element of the
classical theories is a prediction of a wake expansion behind wind turbines approximately in 1.22
times at the optimal modes (Hansen 2008). Indeed, in Fig. 4.26 a radial distance from the rotor
axis of the tip vortices increases up to the lower horizontal limit of the observation domain which
has a size equal to 1.28R. In Fig. 4.25 the wake extension in simulations (Segalini & Alfredsson
2013) has approximately the same value.
An important question for the rotor theories with a finite number of blades is a validation for
an assumption of a constant pitch between turns of the tip vortex in axial direction. That could
not predict theoretically because the axial velocity in the wake changes in two times as it was
mentioned above. We have a confirmation of the assumption in the experiments because the axial
distance between the cores of the two nearest tip vortices does not undergo such a significant
change and remains almost constant (Fig. 4.26). This fact, which is important for the approval of
the vortex theories with a finite number of rotor blades, was also observed in air measurements in
a wind tunnel behind the wind turbine model in terms of the «MEXICO» project (Nilsson et al,
2011). Another important assumption of the analytical model considered in this Section made
about an absence of the tangential rotation of the tip vortices as a contradictory concept to the
assumption by Joukowsky (1912-1918) about "frozen-in" wake rotating together with rotor. This
section assumption is confirmed by direct measurements of the velocity field for the optimal
operating mode (see azimuthal velocity in Figs. 4.26 and 4.27).
Thus, the instantaneous structure of the three-dimensional velocity field in the longitudinal
section of the flow behind the tri-blade rotor model of this study allowed verifying and validating
the main assumptions and hypotheses of different classical rotor theories.
5 Conclusion
This book is dedicated to a special occasion, a remarkable 100th Anniversary of publishing
the first article "Vortex theory of screw propeller" of the famous N.E. Joukowsky.
The introductory part of the book gives the retrospective view on the development of key
historical theories of optimal rotor with ideal load distributions along the blades and describes the
current status of these theories to show an importance and relevance of the theory developed by
Joukowsky. The analysis of primary sources resulted in an acceptance conclusion, that at the time
the leading role in the evolution of rotor aerodynamics belonged to the Russian scientific school
headed by Professor Joukowsky. The results of Joukowsky and his pupils were several years
ahead of the results of foreign scientific schools. In particular the concept of the vortex theory of
a rotor was proposed by Joukowsky seven years earlier than by Betz from German aerodynamic
school of Professor Ludwig Prandtl. The main result of the retrospective analysis presented in the
introduction and the first chapter of the book was a restitution of the principal result in wind
power engineering to the Russian science school; that is a limiting value of energy which can be
extracted from full kinetic energy of wind. Henceforth this result is to be called the Betz-
Joukowsky limit, since it was simultaneously with A. Betz published by Joukowsky in 1920 in
the third article of the series dedicated to wind turbines (Joukowsky 1920). The article turned out
to be the last in his life and long remained unknown to the world scientific community. It seems
paradoxical that the last of efforts of the great scientist was spent to development of wind power
72

engineering in the country, where the energy resource is still in its infancy because his country
has unlimited natural fuels. As if he had foreseen unprecedented growth of the wind energy
occurring in the world today - a century later, knowing that his results would again be in demand
after a century.
The second chapter examines a paradoxical result of the general momentum theory of a rotor
proposed by Joukowsky in the last article of the series "Vortex theory of screw propeller". That
general momentum theory of actuator disc was formulated by Joukowsky to obtain analytical
solutions, considering the disc as a limiting case of the vortex theory for a rotor with an infinite
number of blades. However, in the early XX century, a solution for a rotor with finite number of
blades has failed due to its complexity, and Joukowsky at the end of his first article gave up his
search, noting that "... this analysis would prove to be very complicated". However, even in this
extreme case, an adequate solution for the general momentum theory was not easy to deduce,
because for the operation of wind turbines at low tip speed ratios this would lead to an unnatural
infinite growth of power generated by the flow, or to an unlimited increase of the wind power
coefficient. This paradox of the ideal model for many years has stimulated numerous attempts to
refute the exact value of the Betz-Joukowsky limit, obtained from a simple momentum actuator
disc theory, when the type of energy utilizer is not specified yet. However in the general theory,
it has been already introduced as an infinite number of blades. The blades induce additional
losses responsible for the swirling of the flow in wake which leads to losses and disappearing
wind power and makes wind power coefficient lesser than the Betz –Joukowsky limit always.
Indeed when calculating by alternative theories such as BEM the power coefficient grows
together with increase of the tip speed ratio monotonically from zero to the Betz-Joukowsky limit
but always remains below it. The author's analysis of the causes of abnormal growth of solution
in the general momentum theory has shown that it is determined by the limitations of ideal fluid
model. The adaptation of solution by introducing the effect of pressure and friction on the
boundary of the control stream tube allowed authors to eliminate the infinity. As revised, the
behavior of the maximum power coefficient was to fit in with the solutions for other rotor models
and determined its value as a lesser than the Betz-Joukowsky limit, as it actually should be when
the utilizer of the flow energy specifies in the form of rotating blades as this was done in the
general momentum theory.
Derivation of an analytical solution for the rotor with the finite number of blades and the
finite size of the core of tip vortices within the vortex concept of NEJ rotor is the concluding
result of the present work. The model of NEJ rotor was proposed a century ago, though never
solved analytically by Joukowsky or followers. The sticking point here was a lack of a solution
for helical vortices with a finite vortex core. It was the reason that Joukowsky had referred to the
complexity of any analysis when deriving solution. It should be noted that he propesed the
approximation of helical vortex by vortex ring in his first article of "Vortex theory of screw
propeller” about 60 years earlier than this approximation was found by another authors.
However, accuracy of his approximation was still not sufficient to solve the problem. The
analytical solution of the problem on helical vortex with finite core was obtained only in the
beginning of the XXI century that allowed the authors of this work to be the first who have
brought the solution of Joukowsky’s rotor (NEJ) to the final stage. Moreover, the approach
suggested by the authors to determine an optimal load along the blade proved to be quite suitable
for the study and for comparison of other rotor models. For the basic concepts of optimal rotor
proposed by Betz, Goldstein and Teodorsen, the solutions were obtained and the results were
analyzed and compared. In particular it was found out that the model of NEJ rotor was more
effective than other models operating in the mode of wind turbine.
Thus, this work justifies the priority of the Russian school headed by Professor Joukowsky in
the development of the vortex theory of a rotor in early XX century and restitutes to the Russian
science the non-trivial result on the maximum value of power produced from the kinetic energy
73

of wind. The work has studied and eliminated the problems that arose in Joukowsky’s general
momentum actuator disc theory in the operating modes of a low-speed wind turbine. The
analytical solution for the ideal NEJ rotor with a finite number of blades and finite core of tip
vortices was found for the first time and became a significant contribution to the further
development of the vortex theory of a rotor by Professor Joukowsky.
This work was partially funded by the projects of the Russian Ministry of Education and
Science (GK No. 14.740.11.0144 and GK No. 11.519.11.6022), grants of the Russian Foundation
for Basic Research (RFBR) 10-08-01093, and the Danish Energy Technology Development and
Demonstration Program (EUDP) (No. 64011-0094).

74

References
Alekseenko S.V., Kuybin P.A., Okulov V.L. 2003. Introduction to the Theory of Concentrated
Vortices - Novosibirsk: Nauka, Institute of Thermal Physics. 504 p.
Achkinadze A.Sh. 2006. General optimum condition for screw propeller with finite number of
blades, operating behind the hull // “Morskoy Vestnik”, No 3 (19). Pp. 69-76.
Baskin V.E., Vildgrube L.S., Vozhdaev E.S., Maikapar G.I. 1973. Theory of the lifting airscrew.
– M.: Mashinostroenie.
Batchelor G.K. 1964. Axial flow in trailing line vortices // J Fluid Mech. Vol. 20. P. 645–658
Batchelor G.K. 1967. An introduction to fluid dynamics. Cambridge: Cambridge University
Press. 615 p.
Bendemann F. 1910. Luftschraubenuntersuchenden. Zeitschrift für Flugtechnik und
Motorluftschiffahrt. Vol. 7. P. 177-198
Bergey K.H. 1979. The Lanchester-Betz limit // Journal of Energy, Vol. 3. P. 382–384.
Betz A. 1915. Zeitschrift für Flugtechnik und Motorluftschiffahrt. Vol. 6. Göttingen: R.
Oldenbourg. P. 97.
Betz A. 1919 Schraubenpropeller mit пeringstem Energieverlust: mit einem Zusatz von L.
Prandtl. Gottingen Nachrichten, Gottingen.
Betz A. 1920. Das Maximum der theoretisch moglichen Ausnutzung des Windes durch
Windmotoren // Zeitschrift fur das gesamte Turbinenwesen. Vol. 26. P. 307–309.
Betz A. 1921. The Theory of the Screw Propeller, NACA Technical Notes. Vol. 83. P. 1-19
(translation from German in: Die Naturwissenschaften 1921; 18)
Betz A. 1926. Wind-Energie und ihre Ausnützung durch Windmühlen. Göttingen: Bandenhdect
& Ruprecht. 64p.
Boersma J., Yakubovich S.B. 1998. Solution to problem 97–18*: the asymptotic sum of a
Kapteyn series // SIAM Rev. Vol. 40 P. 986–990
Boersma J., Wood D.H. 1999. On the self-induced motion of a helical vortex // J Fluid Mech.
Vol. 384. P. 263–280
Bothezat G. 1917. Research of work phenomenon for propeller with blades. Petrograd. English
translation in: The General Theory of Blade screw, chapter 3. NACA Report 1920; 29:
198-225
Burton T., Sharpe D., Jenkins N., Bossanyi, E. 2001 Wind Energy Handbook. John Wiley &
Sons. Ltd, England.
Callegari A.J., Ting L. 1978. Motion of a curved vortex filament with decaying vortical core and
axial velocity // SIAM J Appl Math. Vol. 35(1). P. 148–175
Chaplygin S.A. 1914. Theory of cascade wing // Mathematical Collection, Vol. XXIX.
Ciocan G.D., Iliescu M.S., Vu T.C., Nennemann B., Avellan F. 2007. Experimental Study and
Numerical Simulation of the FLINDT Draft Tube Rotating Vortex // J. Fluids Eng. Vol.
129. P. 146-158.
Crow S.C. 1970. Stability theory for a pair of trailing vortices // AIAA J. Vol. 8. P. 2172–2179
De Vries, O.1979 Fluid dynamic aspects of wind energy conversion, Appendix C, section C4
Momentum Considerations, AGARDograph 243, AGARD, Brussel, Belgium.
Drzewiecki S.K. Theory of air propellers and the method of their calculation / foreword by N.B.
Delone. - Kiev: Publ. of R.K. Lubkovsky, 1910. – 62 p.
Dyson F. W. 1893. The potential of an anchor ring. Part II // Philos. Trans. R. Soc. London, Ser.
A 184. P. 1041.
Fage A., Collins H. E. 1919. Applied Aerodynamics. London: Sir J. Causton & Sons, Ltd.
Felli M., Camussi R. , Di Felice F. 2011. Mechanisms of evolution of the propeller wake in the
transition and far fields // J. Fluid Mech. Vol. 682. P. 5–53.
75

Föttinger, H. 1918. Neue Grundlagen für die theoretische und experimentelle Behandlung des
Propellerproblems. Jahrbuch STG. Vol. 19. P. 1385-472.
Frankl F.I. 1939. To one rough mathematical error, widespread in aerodynamic literature and its
correction // Aircraft Fleet Engineering. No 7/8.
Froude R.E. 1889. On the part played in propulsion by differences of fluid pressure //
Transactions of the Institute of Naval Architects. Vol. 30. P. 390–405
Froude R.E. 1911. The acceleration in front of a propeller. Transactions of the Institute of Naval
Architects. Vol. 53. P. 139–182.
Fukumoto Y., Miyazaki T. 1991. Three-dimensional distortions of a vortex filament with axial
velocity // J Fluid Mech. Vol. 222. P. 369–416
Fukumoto Y & Okulov V.L. 2005 The Velocity Field Induced by a helical Vortex Tube. Phys.
Fluids. 17(10), 107101 (1–19).
Glauert H. 1935 Airplane propellers. Division L in Aerodynamic Theory, vol. IV, (ed. Durand
W.F.). Springer: Berlin. P. 169–360.
Goldstein S. 1929 On the vortex theory of screw propellers. Proc R Soc London A, Vol.123. P.
440–465.
Gorelov D.N. 2012. Aerodynamics of wind wheels with vertical-axis / Omsk Branch of the S.L.
Sobolev Institute of Mathematics. - Omsk: KAS Printing Center, 2012 – 68 p.
Gorban A.N., Gorlov A.M., Silantyev V.M. 2001. Limits of the Turbine Efficiency for Free Fluid
Flow // Journal of Energy Resources Technology, Vol. 123, 311-317
Hansen M.O.L. 2008. Aerodynamics of Wind Turbines: second edition. // In series: Earthscan.
181р.
Hardin J.C. 1982.The velocity field induced by a helical vortex filament // Phys. Fluids. Vol. 25,
No 11. – P. 1949–1952.
Hoff W. 1921. Theory of the ideal windmiil, NACA Technical Notes. Vol.46. P. 1-17.
Joukowsky N.E. 1912-1918. Vortex theory of screw propeller, I-IV. Trudy Otdeleniya
Fizicheskikh Nauk Obshchestva Lubitelei Estestvoznaniya: part I - 1912, vol. 16(1): 1–31;
part II - 1914, vol. 17(1): part III - 1–33; 1915, vol. 17(2): 1–23; Trudy Avia Raschetno-
Ispytatelnogo Byuro: part IV - 1918; 3: 1–97 (in Russian).
Joukowsky NE. 1920 Aerodynamic calculation of slow wind mills (article 1). Low-speed wind
mill (article 2). Windmill of the NEJ type (article 3). In: Transactions of the Central
Institute for Aero-Hydrodynamics of Moscow. And it was reprinted in: Joukowsky N.E.
Collected Papers Vol. VI. Moscow-Leningrad, Russia: ONTI, 1937; vol. VI, pp. 387 –
424 (in Russian).
Joukowsky N.E. 1935-1937. Complete Works in 9 Volumes (S.A.Chaplygin, Nekrasov V.A.,
Archangelsky V.P., Vetchinkin V.P. and Kotelnikov A.P. - Eds.). ML: ONTI.
Joukowsky N.E. 1929. Théorie tourbillonnaire de l’hélice propulsive. Paris: Gauthier-Villars,
198p.
Keldysh M., Frankl F. 1935. Rigorous justification of the Joukowsky’s theory of screws //
Mathematical Collection. Volume 42, No 2. Pp. 241-273.
Kelvin, Lord. 1880. Vibrations of a columnar vortex // Philos. Mag. Vol. 10. P. 155–168.
Kuibin P.A., Okulov V.L. 1998. Self-induced motion and asymptotic Expansion of the velocity
field in the vicinity of helical vortex filament // Phys. of Fluids, - Vol. 10, N 3. - P. 607-
614.
Kuik, G.A.M., van. 2007. The Lanchester–Betz–Joukowsky Limit // Wind Energy Vol. 10. P.
289–291
Kurzin V.B. 1993. Low-frequency acoustic self-oscillations in the wheel space of hydraulic
turbines // Applied Mechanics and Technical Physics. No 2. Pp. 96-106.
Lam, G.C.K. 2006 Wind energy conversion efficiency limit, Wind Engineering Vol. 30(5) . P.
431.
76

Lanchester F.W. 1915. A contribution to the theory of propulsion and the screw propeller //
Transactions of the Institute of Naval Architects. Vol. 57. P. 98–116
Lerbs H. 1952. Moderately loaded propeller with a finite number of blades and an arbitrary
distribution of circulation // Trans. SNAME, v. 60, pp 73-123. N.Y.: SNAME.
Levy H., Forsdyke A.G. 1928. The steady motion and stability of a helical vortex // Proc. Roc.
Soc. Lond. – Vol. A120. - P. 670-690.
Madsen H.Aa. 1983. The Actuator Cylinder – A Flow Model for Vertical Axis Wind Turbines //
Ph.D. thesis. Aalborg University, Denmark.
Madsen H.Aa., Bak C., Døssing M., Mikkelsen M., Øye S. 2010 Validation and modification of
the Blade Element Momentum theory based on comparisons with actuator disc
simulations. Wind Energy, 13, pp. 373-389.
Madsen H.Aa., Mikkelsen R., Øye S., Bak B., Johansen J. 2007 A Detailed investigation of the
Blade Element Momentum (BEM) model based on analytical and numerical results and
proposal for modifications of the BEM model. The Science of Making Torque from Wind,
Journal of Physics: Conference Series 75, 012016 doi:10.1088/1742-6596/75/1/012016.
Margoulis W. Propeller theory of professor Joukowsky and his pupils, translated from
‘L’Aeronautique’ August 1921, Technical Memorandums . Vol. 79. P. 1-12
Maykopar G.I. 1969. Studies of air propellers // Materials to the History of TsAGI. M:
Publishing Department of TsAGI. 13 p.
Maykapar G.I., Lepelkin A.M., Khalezov D.V. 1940. Aerodynamic calculation of screws based
on the theory of blade// Transactions of TsAGI, issue 529.
Morgan B.M., Wrench J.W. (Jr) 1965. Some computation aspects of propeller design // Methods
in Computational Physics. Vol. 4. P. 301-331.
Moore D.W., Saffman P.G. 1972. The motion of a vortex filament with axial flow // Phil. Trans.
R. Soc. Lond. - Vol. A272. – P. 403-429.
Munk M. 1920. Wind driven propeller. NACA Technical Memorandums. Vol. 201. P. 1-12
(translation from German in: Zeitschrift für Flugtechnik und Motorluftschiffahrt August
15, 1920. P. 220-223).
Nilsson K., Shen W.Z., Sørensen J.N., Ivanell S. 2011. Determination of the tip vortex trajectory
behind the MEXICO rotor // Abs. Wake conference, Gotland University, Visby, Sweden,
8-9 June, (eds Sørensen J.N. & Ivanell S.) p. 94-99
Naumov I.V., Rahmanov V.V., Okulov V.L., Velte K.M., Meyer K.E., Mikkelsen R.F. 2012.
Flow diagnostics downstream of a tribladed rotor model // Thermophysics and
Aeromechanics. Vol. 19. No 3. p. 267-278.
Okulov V.L. 1993. Resonance hydro-acoustic processes in the flow section of the machines and
units with intensive flow swirling // Abstract of the Doctor Dissertation in Physical and
Mathematical Sciences. Novosibirsk, Institute of Thermophysics SB RAS. 34 p.
Okulov V.L. 1995. The velocity field induced by vortex filaments with cylindrical and conic
supporting surface // Russ. J. Eng. Thermophys. Vol. 5(2). P. 63–75.
Okulov V.L. 2003. Generalization of the problem on stability of the polygonal configuration of
point vortices with reference to helical vortex filaments //In the book: "The Fundamental
and Applied Problems in the Theory of Vortices" (edited by Borisov A.V., Mamaev I.S.,
and Sokolowsky M.A.) Moscow-Izhevsk: Institute of Computer Science. Pp. 392-413.
Okulov V.L. 2004 On the Stability of Multiple helical Vortices. J. Fluid Mech. 521, 319-342.
Okulov V.L., Fukumoto Ya. 2004. Helical dipole // DAN. Vol. 399, No 1, Pp. 56-61.
Okulov V.L., Pylev I.M. 1995. Instability of pressure systems // DAN, Vol. 341, No 4. Pp. 470-
473.
Okulov V.L., Sørensen J.N. 2007 Stability of helical tip vortices in rotor far wake. J. Fluid Mech.
576, 1-25.
77

Okulov V.L., Sørensen J.N. 2008a Refined Betz limit for Rotors with a Finite Number of Blades.
Wind Energy, 11(4), 415-426.
Okulov V.L. & Sørensen J.N. 2008b An ideal wind turbine with a finite number of blades.
Doklady Physics 53(6), 337-342.
Okulov V.L., Sørensen J.N. 2010. Maximum efficiency of wind turbine rotors using Joukowsky
and Betz approaches J. Fluid Mech. Vol. 649. P. 497-508
Okulov V.L., Sørensen J.N., Voigt L.K. 2005. Vortex scenario and bubble generation in a
cylindrical cavity with rotating top and bottom // European J. Mechanics B, Vol. 24, no.
1, pp. 137-148
Okulov V.L., van Kuik G.A.M. 2012. The Betz–Joukowsky limit: on the contribution to rotor
aerodynamics by the British, German and Russian scientific schools. Wind Energy, Vol.
15(2). P. 335-344
Øye, S. 1990 A Simple Vortex Model Proceedings of the third IEA Symposium on the
Aerodynamics of Wind Turbines, ETSU, Harwell, 1990, pp. 4.1-4.15.
Polyakhov N.N. 1937. On the most optimal screw // Transactions of TsAGI, Vol. 455. Pp. 1-30.
Prandtl L. 1913. Ergebnisse und Zielte der Göttinger Modellversuchsanstalt, Zeitschrift für
Flugtechnik und Motorluftschiffart, 1913, No3.
Prandtl L. 1918. Tragflügeltheorie. Part I. Mitteilung, Nachrichten der Gesellschaft der
Wisseneschaften zu Gottingen, Math. Physik K1. P. 151–177.
Prandtl L. 1919. Tragflügeltheorie. Parts II. Mitteilung, Nachrichten der Gesellschaft der
Wisseneschaften zu Gottingen, Math. Physik K1. P. 107–137
Prandtl L. 1923. Applications of modern hydrodynamics to aeronautics. NACA Report. Vol. 116.
P. 1-215.
Prandtl L. 1952. Essentials of Fluid Dynamics. London-Glasgow: Blackie & Son Limited. 452p.
Rankine W.J.M. 1865. On the mechanical principles of the action of propellers // Transactions of
the Institution of Naval Architects. Vol. 6. P. 13-39.
Ricca R.L. 1994. The effect of torsion on the motion of a helical vortex filament // J Fluid Mech.
Vol. 273. P. 241–259
Ricca R.L. 1996. The contributions of Da Rios and Levi-Civita to asymptotic potential theory
and vortex filament dynamics // Fluid Dynamics Res. Vol. 18. P. 245–268
Rosenhead L. 1930. The spread of vorticity in the wake behind a cylinder // Proc R Soc London
A. Vol. 127. P. 590–612
Sabinin G.H. 1927. Theory of the ideal wind turbine. Transactions of the Central Institute for
Aero-Hydrodynamics of Moscow, 32: 1-27 (in Russian).
Saffman, P.G. 1992 Vortex Dynamics, Cambridge University Press, USA
Segalini, A., Alfredsson P.H. 2013. A simplified vortex model of propeller and wind turbine
wakes // J Fluid Mech. (in press).
Sharpe, D.J. 2004 A general momentum theory applied to an energy-extracting actuator disc,
Wind Energy, 7, 177.
Shen W.Z., Zhu W.J., Sørensen J.N. 2011. Actuator line/Navier-Stokes computations for the
MEXICO rotor: comparison with detailed measurements // Wind energy, DOI:
10.1002/we.510.
Snel H., Schepers J.G., Montgomerie B. 2007. The MEXICO (Model Experiments in Control
Conditions): the database and first results of the data process and interpretation // Journal
of Physics: Conference Series. Vol. 75. P. 012014
Sørensen J.N., van Kuik G.A.M. 2011. General momentum theory for wind turbines at low tip
speed ratios. Wind Energy, Vol. 14(7). P. 821-839
Tibery C.L., Wrench J.W. (Jr) 1964 Tables of Goldstein factor // Report no. 1534 of Department
of Navy, Washington. P. 1-69.
Theodorsen T. 1948. Theory of Propellers. New York: McGraw-Hill.
78

Thomson J.J. 1883. A Treatise on the Motion of Vortex Rings. Mac-Millan, London
Yuriev B.N.1923. The present status of the screw propeller theory // Aircraft Fleet Bulletin. No 5,
Pp. 324.
Vetchinkin V.P. 1913 Calculation of screw propeller, Part I. Buleteni Politekhnicheskogo
obshestva, no. 5.
Vetchinkin V.P. 1914. On invariants of the screw propeller. // Transactions of the Physical
Sciences Division of the Natural Science Society, Vol. 17 (1).
Vetchinkin V.P. 1918. Calculation of screw propeller, Part II. Trudy Avia Raschetno-
Ispytatelnogo Byuro; no. 4: 1-129 (in Russian).
Vozhdaev V.S., Vozhdaev E.S. 1997. To N.E. Zhukovsky’s vortex theory of screw propeller. //
Transactions of TsAGI, Vol. XXVIII, No 1.
Vozhdaev V.S., Vozhdaev E.S. 2001. Analytical solutions for the the velocity field harmonics
induced by the lifting airscrew //Transactions of TsAGI, Vol. XXXII, No 3-4.
Vozhdaev E.S. 2003. Aerodynamics and flight dynamics of helicopters. Chapter 5 of the book
"TsAGI - the Main Stages of Scientific Activity, 1993-2003".M.: Nauka-Fizmatlit.
Walther J.H., Guénot M., Machefaux E., Rasmussen J.T., Chatelain P., Okulov V.L., Sørensen
J.N., Bergdorf M., Koumoutsakos P. A numerical study of the stabilitiy of helical vortices
using vortex methods // Journal of Physics: Conference Series 2007. Vol. 75, pp. 012034
Wilson R., Lissaman P.B.S. 1974 Applied Aerodynamics of Wind Power machines, Oregon State
University. Research Applied to National Needs (RANN) under grant No. GI-41840.
Widnall S.E. 1972. The stability of helical vortex filament // J. Fluid Mech. - Vol. 54. – P. 641 -
663.
Widnall S.E., Bliss D.B, Zalay A. 1971. Theorical and experimental study of the stability of a
vortex pair // Proc. Symp. on Aircraft Wake Turbulence, Seattle, Washington, Plenum –
P. 305–338.
Wood, D.H. 2007 Including swirl in the actuator disk analysis of wind turbines, Wind
Engineering, 31, no. 5, p. 317-323.
Wood, D.H. & Boersma, J. 2001 On the motion of multiple helical vortices. J. Fluid Mech. 447,
149–171.
Xiros, M.I. and Xiros, N.I. 2007 Remarks on wind turbine power absorption increase by
including the axial force due to the radial pressure gradient in the general momentum
theory, Wind Energy, 10, 99.
79

View publication stats

Propeller

Uploaded by

Copyright:

Available Formats

You might also like

Propeller

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Propeller

Uploaded by

Copyright:

Available Formats

See

DEVELOPMENT OF THE OPTIMUM ROTOR

Book · December 2013

Valery Leonidovich Okulov Gijs van Kuik

SEE PROFILE SEE PROFILE

The user has requested enhancement of the downloaded file.

DEVELOPMENT OF THE OPTIMUM ROTOR THEORIES

1.1 The development stages of rotor aerodynamics

Figure 1.1. N.E. Joukowsky

1.2 The contribution to the development of rotor aerodynamics made by Joukowsky

1.3 The objectives and the structure of the book

2.1 Maximal efficiency of wind power utilization

2.2 The actuator disc theory: Retrospect

2.2.1. Contributions of the British School

or by (2.2) and (2.3) –

2.2.2. Contributions the German and Russian Schools

Fig 2.3. Title page of a scientific journal, showing the connection

2.3 Corroboration of the actuator disc theory by the vortex theory

2.4 The Betz-Joukowsky limit

Year and scientist of

2.5 Concluding remarks

Figure 2.8. Drzewiecki’s blade element method:

3.1 Introducing remarks

3.2.1. The equations of motion

the disc and where

3.2.2. The disc load

3.2.3. The far wake

3.2.4. The momentum balance

3.3. The disc model of Joukowsky with constant bound circulation

Equation (3.9) becomes

0 for a propeller (3.24)

The axial load becomes, using (3.10) and (3.11)

Integration of (3.25) and (3.26) gives the thrust T :

3.3.2. The pressure in the fully developed wake

This gradient has to deliver the centripetal acceleration v2 / r . Substitution of v 1    2 r1 

where p is determined by equating (3.35) and (3.38) for r1  1 , giving

Integration of (3.35) and (3.38) across the wake cross-section gives

For 1  0 the following remains:

3.3.3. The momentum and energy balance

and the axial momentum balance simplifies to:

with respect to flows without swirl.

two terms become equal.

3.3.4. Analytical expressions for velocity, power and thrust

Writing (3.48) plus (3.49) gives

Elimination of (1  a) in (3.49) by substitution of (3.48), and multiplying the result by (q  2 )

2  4 2  2 q  q 2  4 2  

3.4. Results and discussion of the Joukowsky’s model

3.4.1. Tendencies in limiting cases

The following observations are made:

3.4.2. Quantitative results and discussion

Figure 3.6: Dimensionless circulation as function of axial wake interference factor

Figure 3.7: Area expansion as function of axial wake interference factor

where Vrel is the relative velocity in the rotor plane:

Vrel  U 02 (1  a ) 2  (r  ½ v ) 2 . (3.60)

Defining the lift coefficient as

3.4.3. Discussion of previous analyses

T   pdA   vx v  dA   pdA  e   τdA  e