Journal of Energy and Power Engineering 10 (2016) 23-31 doi: 10.17265/1934-8975/2016.01.003

The Mechanism of Flow Field around a Rotating

Cylinder with Fins for High Performance Magnus Wind

Turbine

Nobuhiro Murakami1, Hiroaki Hasegawa2, Toshihiro Haniu1 and Masahide Nakamura1

1. Dept. of Mechanical Engineering Tegatagakuen-machi 1-1, Graduate School of Engineering & Resource Science, Akita Univ.,

Akita 010-8502, Japan

2. Dept. of Mechanical and Intelligent Engineering Yoto, Graduate School of Engineering, Utsunomiya Univ., Utsunomiya-shi 7-1-2, Japan

Received: November 23, 2015 / Accepted: December 02, 2015 / Published: January 31, 2016. Abstract: Spiral Magnus is a unique wind turbine system that rotates with cylinders which have spiral-shaped fins coiled around them (instead of using the more common propeller-type blades). In the present study, three models (cylinder with no fins, cylinder with straight fins and cylinder with spiral fins) were installed, and fluid force measurements were performed by a strain gauge force balance. A PIV (particle image velocimetry) system was used to better understand the flow fields around the cylinder. Considering the results of the experiment, it was confirmed that, the aerodynamic performance of the rotating cylinder can be improved by the fin. However, the straight fin makes the flow close to the cylinder surface ineffective. The rotary cylinder with the spiral fins was able to generate the greatest lift among three models, because the spiral fin effectively influences the vicinity of the cylinder surface.

Key words: Wind turbine, Magnus effect, wind tunnel test, spiral Magnus.

Nomenclature

CL Time-averaged lift coefficient

d Cylinder diameter (m)

Re Reynolds number

U Free stream direction velocity (m/s)

U0 Free stream velocity (m/s)

VS Cylindrical surface speed (m/s)

α Rotational angle (°)

θ Circumferential speed ratio

ν Kinematic coefficient of viscosity (m2/s)

Ω Vorticity

ω Angular velocity (rad/s)

1. Introduction

The development of wind power generation from

large wind turbines has shown excellent profitability,

therefore, this form of renewable energy construction is

accelerating as mainstream around the world. However,

Corresponding author: Nobuhiro Murakami, doctor,

research fields: engineering, aero dynamics and wind turbine.

these businesses must be built in locations away from

residential area, from the point of view of the landscape

and noise. Wind power generation also requires that,

wind conditions be good and the site be near an electric

power-consuming region. However, locations fulfilling

such conditions are few. Therefore, a small wind

turbine with superior quietness and high performance

is demanded. This would allow extensive construction

near residential areas. In addition, development and

maintenance of the transmission line would also be

cheaper.

A Magnus wind turbine uses a rotating cylinder as a

substitution for the blade of a common wind turbine.

When the cylinders rotate, lift is generated. This lift is

sufficient to generate electricity by rotating the wind

turbine to which the cylinders are attached. The wing

of the wind turbine is in a column shape, and, in the

Magnus wind turbine, the power performance rises so

that the lift becomes bigger. After the invention of a

D DAVID PUBLISHING

24

large perform

spiral fin to

patent [1], c

spiral Magnu

even low wi

under stable

wind veloci

high resultin

from the sm

square mete

performance

results, appr

used to rotat

accelerate co

performance

the flow

mainstream

important.

Previous

spiral fin sha

focused on t

shapes, were

computer sim

of several ki

not been eno

mechanism

There are s

improvemen

lift-generatin

rotating cy

mechanism i

the generat

generation w

2. Experim

2.1 Experim

A schem

present stud

an open typ

The rotation

section. The

The

mance enhan

o the colum

commercializ

us wind turbi

ind speeds (F

e output cha

ity fluctuatio

ng amount of

mall wind tur

rs” is shown

e [2]. Accord

roximately 8

te the rotary c

ommercializa

e even more.

mechanisms

and the

studies had t

ape. These in

the comparat

e carried out

mulation stud

inds of mode

ough elucida

of rotation

still many p

nt. This stu

ng mechanism

ylinder with

in the light of

ted flow fie

was revealed.

mental App

mental Setup

matic of the

dy is shown in

pe, with an o

n cylinder mo

DC motor (S

e MechanismHig

ncement in 2

mn, and after

zation is bein

ne obtains sig

Fig. 1). It can

aracteristics e

ons. Therefor

f cumulative

rbine “swept

as superior p

ding to the act

% of the gen

cylinder. How

ation, it is des

Therefore, th

that occu

spiral cylin

the purpose o

nvestigations,

tive analysis

t in wind tun

dies, and actu

els [3, 4]. Ho

ation of the fl

around the

possibilities

udy intended

m in the flow

fin. By

f the state of th

eld, the me

paratus and

measuremen

n Fig. 2. The

outlet of 300

odels were p

S12280-240,

m of Flow Fieldgh Performan

2004 by addin

r applying fo

ng advanced.

gnificant lift f

n be seen to h

even under r

re, the relati

generated po

area below

power genera

tual measurem

nerated powe

wever, in orde

sirable to impr

he elucidatio

ur between

ndrical wing

of optimizing

, which prima

of various sp

nnel experime

ual on-site tes

owever, there

low field and

finned cylin

for performa

d to clarify

w field aroun

considering

he fluid force

echanism of

d Methods

nt section of

wind tunnel

mm × 300 m

placed in the

Special Elect

d around a Rce Magnus W

ng a

for a

The

from

have

rapid

vely

ower

200

ation

ment

er is

er to

rove

on of

the

g is

g the

arily

piral

ents,

sting

e has

d the

nder.

ance

the

nd a

the

e and

lift

f the

was

mm.

test

trical

Fig.(Ma

Fig.

Co.

the

bala

plac

unit

gen

the

visu

pro

the

mod

vert

(cha

1,00

lase

gen

Rotating CylinWind Turbine

. 1 The spiraarch 2011 start

. 2 Schematic

) rotated the

model and th

ance about t

ced on the fro

t moved ab

nerated as the

strain gauge

ualization an

cess, the tran

mirror. Relat

del was esta

tical. The P

arge-coupled

08 × 1,024 p

er (Solo 120X

nerator (Quan

nder with Fins

al Magnus wint-up).

c of experiment

cylinder. Th

he motor, wa

the fulcrum,

ont side of th

out the fulc

flow passed

e indirectly m

nd PIV (part

nsmitted lase

tive to the lon

ablished as e

IV system m

d device) cam

pixels, Koda

XT, New Wav

ntum Compos

s for

nd turbine at

tal setup.

he rotation un

as structured

therefore, a

e fulcrum. A

crum, owing

over the rotat

measured the

ticle image

er sheet was

ngitudinal dir

xtending fro

mainly consi

meras (Mega

ak Co. Ltd.),

ve Research,

sers, Inc., 93

Akita marina

nit, including

to move as a

weight was

s the rotation

g to the lift

ting cylinder,

e lift. In the

velocimetry)

deflected by

rection of the

m the upper

ists of CCD

plus ES1.0,

, a Nd-YAG

Inc.), a pulse

14), a rotary

a

g

a

s

n

t

,

e

)

y

e

r

D

,

G

e

y

The Mechanism of Flow Field around a Rotating Cylinder with Fins for High Performance Magnus Wind Turbine

25

encoder (Ono Sokki LG-919) and a smoke generator

(Kano Max, 8304). Tracer using an oil mist, the

particle size is about 1 μm. Rotational speed of the

cylinder with blades was measured using a reflection

tachometer. The timing of the laser irradiation is

synchronized by a pulse generator.

2.2 Experimental Methods

The origin of the coordinate system was the center

of gravity of the experimental model. The main flow

direction was X, the measurement unit span direction

was Y, and the height direction was Z. The free stream

Reynolds number (Re) of flow around a cylinder, with

a typical dimension for the column diameter d = 0.07 m,

and the representative speed is the mainstream

velocity U0 = 4 m/s, was from Re = 1.9 × 104. The

mainstream and the ratio of the rotary speed of the

column θ are the important parameters for showing

the state. The flow field around the cylinder, the outer

flow 1/2dω, and the mainstream U0 due to the rotation

of the cylinder. Eq. (1) shows the velocity ratio as a

circumferential speed ratio θ.

02

d

U

(1)

Lift measurements were conducted using three

types of experimental models for θ = 0-1.1. The strain

gauge, using a two-component force meter that was

proven in previous experiments [5, 6], using that single

component force. The measurement data, 5,000 data

points acquired at a sampling speed of 500 μs, was

carried out six times, and is determined as an average

value. The error in this measurement is about 1.9%.

The flow velocity distribution of the flow field around

a rotating cylinder was determined by PIV. In addition,

the peripheral speed ratio at this time is θ = 1. The

irradiation laser was carried out in two consecutive

passes, the irradiation interval, and the 500 μs. The

width of the laser sheet was 5 mm. We analyzed the

images in the PIV software (FtrPIV, FLOWTECH

RESEARCH Corp.). However, the laser irradiated

from the upper side of the model, and hence, it was

not possible to photograph the entire flow field.

Therefore, to complete the model, the data from the

upper half of the surface were reversed for the model

surface of the lower half, and the upper and lower data

were then synthesized by the X-axis reference.

2.3 Experimental Models

Fig. 3 shows in three experimental models. Fig.3a is

the reference made no cylinder of the fin. Fig. 3b is a

cylinder with straight fin cylinder (dual). Fig. 3c is a

cylinder with a spiral fin (dual). Straight fins with 10 mm

width and 18 mm height is placed in parallel to the Y

axis, and the other one was placed similarly in a

(a) No fin

(b) Straight fins (dual)

(c) Spiral fins (dual)

Fig. 3 Three types of test models (dimensions in mm).

26

position wh

left-handed

height of 18

cylinder en

similarly in

3. Experim

3.1 Lift Mea

Fig. 4 giv

models. In a

the potential

reference. In

spiral fins a

without fins

that was add

speed ratio θ

2πθ of an id

spiral fins a

θ > 1, lift

However, th

When the va

increase the

occur in the

described in

section of th

there is actu

pulsation of

the vicinity

0.9-1.1 for t

little change

wind turbine

The range o

leading to a

However, w

tip and the

distances fro

in an actual

that, if a tri

main flow a

turbine, it is

In the rotati

The

hich is shifte

in a 35° and

mm as a sta

d face. And

a position wh

mental Resu

asurement

es the lift me

addition, it sho

l theory, with

n the θ = 1 n

and straight

. In particular

ded to the spi

θ = 1 showed

deal fluid. In

lso increased

began to co

he model with

alue of θ was

lift. Changes

e vicinity of

n the “PIV me

his paper. W

ually always

f wind change

y of the circ

the spiral fin.

e in the lift v

e is insensitiv

f the optimum

an increase in

when the actua

root of the

om the center

l wind turbin

iangular vect

and the flow d

not the same

ing cylinder,

e MechanismHig

ed 180°. Sp

d its width is

arting point to

d the other

hich is shifted

ults and Co

asurement re

ows the CL =

h the lift of a

neighborhood

fins became

r, the rotating

iral fins had

d high lift con

addition, the

d lift with inc

onverge to a

hout fins conti

s less than 1,

s in the work o

θ = 1. The r

easurement ar

When the wind

a value of θ

es. In Fig. 4,

umferential

. It was show

values for θ.

ve to the pulsa

m power cond

n the wide p

al wind turbi

cylinder blad

of the wind tu

ne, it is neces

tor is genera

due to the tur

e near the root

the velocity

m of Flow Fieldgh Performan

piral fins, wa

s 10 mm, the

o any point of

one was pla

d 180°.

onsideratio

sults for the t

2πθ derived f

an ideal fluid

d, the lift for

larger than

cylindrical b

a circumferen

nforming to C

straight fins

creasing θ. W

constant va

inued to incre

the fins acte

of the fin beg

reason for th

round a cylin

dmill is runn

θ for receivin

the lift is hig

speed ratio

wn that, there

. This means

ation of the w

dition is a fea

ower generat

ne is turning

des are diffe

urbine. Theref

ssary to cons

ated between

rning of the w

t as it is at the

triangle of f

d around a Rce Magnus W

as a

e fin

f the

aced

on

three

from

as a

r the

that

blade

ntial

CL =

and

When

alue.

ease.

ed to

in to

his is

nder”

ning,

ng a

gh in

θ =

was

s the

wind.

ature

tion.

, the

erent

fore,

sider

n the

wind

e tip.

flow

Fig.

flow

is a

win

infl

in th

3.2

F

circ

obta

flow

by

indi

wel

velo

axe

are

Add

resu

ima

rota

VS

dire

bott

indu

the

port

Rotating CylinWind Turbine

. 4 Coefficien

wing in each r

a complex flo

nd turbine, i

luence of the

he radial dire

PIV Measure

For the lift me

cumferential

ained for eac

w field straigh

PIV measu

icate the dire

ll as the dist

ocity U as a

es in the figure

dimensionl

ditionally, the

ult of the PI

ages. Consid

ating cylinder

generated by

ection and ma

tom of the c

uced by rotat

other hand,

tion, the reve

nder with Fins

t of lift.

radial position

ow. Hence, w

it is necessa

circumferent

ection of the r

ement of arou

easurement at

l speed ratio

h model. To

ht fins and the

rements. Th

ection and spe

tribution of

gradient. Th

e, each with a

ess by main

e speed of th

IV processin

dering the po

r in Fig. 5, the

y the rotatio

agnitude is th

ylinder. Cyli

tion, is gener

the cylindric

erse side, is

s for

n is different.

when designi

ary to consid

tial speed rati

rotating cylin

und a Cylinde

t the time of c

o, a differen

test the cause

e spiral fin we

he measurem

eed of the flo

the main flo

he vertical an

a diameter of

nstream win

he data is ave

ng of 40 set

otential flow

e cylindrical s

on of the cyl

he symmetry a

indrical surfa

rated in the A

al surface sp

-VS. On the

Therefore, it

ing an actual

der that, the

io is different

nder.

er

hanges in the

nt trend was

e of this, two

ere examined

ment results

ow vector, as

ow direction

nd horizontal

f the cylinder,

d speed U0.

raged for the

s of particle

w around the

surface speed

linder as the

at the top and

ace speed VS,

A section. On

peed of the B

reverse flow

t

l

e

t

e

s

o

d

s

s

n

l

,

.

e

e

e

d

e

d

,

n

B

w

Fig. 5 Situa(α = 0°).

side of the c

interference

this rotation

Fig. 6 sho

mainstream,

Fig. 6 Flow

The

ation of flow

cylinder, a st

between the

.

ows a straight

, a flow field

past around ci

e MechanismHig

past around

tagnation reg

main flow an

finned rotatin

image taken

(a)

(c)

ircular cylinde

m of Flow Fieldgh Performan

circular cyli

ion is formed

nd the flow du

ng cylinder in

at the momen

er with straigh

d around a Rce Magnus W

inder

d by

ue to

n the

nt of

eac

defo

with

in a

the

sho

cyli

is th

sim

fins

mov

that

low

form

a la

the

ht fins: 4 angle

Rotating CylinWind Turbine

h rotation

formed by th

h straight fins

a far area from

lift of Fig. 4

w a high lift.

indrical surfa

he same. In t

milar, which is

s through the

ve synchrono

t the fins exe

west state. F

mation of the

arge area, as

stagnation r

Stagnation are

conditions (θ =

nder with Fins

angle α. Th

e rotational a

s affect to the

m the cylinde

, in the case

In the part A

ace speed VS

the forward f

s a unique sta

e forward flo

ously at the s

ert on the per

urthermore,

stagnation re

shown in Fig

region, in wh

(b)

ea (d)

= 1).

s for

he flow is

angle. Rotati

e flow velocity

er. In the me

of θ = 1, finn

A of Fig. 5, th

and mainstre

flow side, the

ate. According

ow side of th

same speed, t

riphery of th

when we f

egion, we see

g. 6d. Thus, t

hich U0 and

27

periodically

ing cylinders

y distribution

asurement of

ned cylinders

he flow of the

eam speed U0

ese flows are

gly, since the

he main flow

the influence

e flow is the

focus on the

e that it forms

the impact of

-VS is in the

7

y

s

n

f

s

e

0

e

e

w

e

e

e

s

f

e

28

reverse dire

become a m

the reasons t

relatively low

Figs. 7 a

indicating th

and showing

For Fig. 4, w

of the lift fo

seen when t

This is an ar

lift line of an

the circumfe

shows the vo

angles at: (a

α = 135°.

In the forw

Fig. 7 Flow aand (d) α = 13

The

ection flow,

major factor of

that high lift

w Reynolds n

and 8 exhib

he direction

g the distribut

we discussed t

or the straight

the value of θ

rea that exhib

n ideal fluid.

erential speed

orticity for th

a) α = 0°, (b

ward flow see

around rotatin35° (θ = 0.2).

Flow

e MechanismHig

shows that,

f lift generatio

is generated

number.

bit the meas

and speed of

ion of vorticit

the results of

t fin. A rapid

θ is in the vi

bits lift values

Fig. 4 shows

ratio θ = 0.2.

he case of θ =

b) α = 45°,

en in Fig. 7d,

(a) α =

(c) α =

ng cylinder wit

m of Flow Fieldgh Performan

, side collis

on. This is on

from the tim

surement res

f the flow ve

ty as the grad

the measurem

increase in li

cinity of 0.2-

s greater than

s a high lift w

Therefore, F

= 0.2, for rota

(c) α = 90°,

when α = 135

= 0°

= 90°

h straight fins.

d around a Rce Magnus W

sions

ne of

me of

sults,

ector

dient.

ment

ift is

-0.4.

n the

when

ig. 7

ation

, (d)

5°, a

stro

(blu

in w

nea

add

the

mai

dire

(ye

that

vici

the

is g

forw

the

imp

the

. Velocity distr

Rotating CylinWind Turbine

ong vorticity

ue—clockwise

which α = 45

ar, the vorticit

dition, in the r

rotation an

instream and

ections, an

llow—count

t, the vorticity

inity of α = 4

straight fin, a

generated in

ward directio

opposite dir

pact is small.

vortices is e

ribution and vo

nder with Fins

y near the

e vorticity) is g

5°, when the b

ty is peeled a

reverse direct

ngle α = 0°

d the fin are

nd vorticity

terclockwise

y is peeled aw

45° (Fig. 7b)

a large effect

n the vicinity

n. On the con

rection is fa

The positioni

estimated by

(b) α = 45°

(d) α = 135°

orticity for: (a)

s for

backward c

generated. Th

backward cy

away from the

ion flow, in th

°, seen in F

e in flowing

y from th

vorticity) o

way as the fin

. In the flow

occurs becau

y of the cyl

ntrary, since

ar from the c

ing of the lift

its effects. O

) α = 0°, (b) α =

cylinder fin

hen, in Fig. 7b,

ylinder comes

e cylinder. In

he vicinity of

Fig. 7a, the

in opposite

he fin tip

occurs. After

s move to the

field around

use the vortex

linder in the

the vortex in

cylinder, the

generated by

On the other

= 45°, (c) α = 90

n

,

s

n

f

e

e

p

r

e

d

x

e

n

e

y

r

0°

Fig. 8 Flow 0.6 and (d) θ =

hand, Fig. 4

0.2-0.4. For

plateau. At

system has

phenomenon

edge vortex

vicinity of th

the calculati

angle α = 13

and 8d: θ =

series of flow

the blade su

large lift [7]

strongly in t

cylinder. Th

rotation ang

The

around rotatin= 0.8 (α = 135°

shows high l

r 0.6-0.8, the

the low circ

exceeded th

n is considere

x that occurs

he rotation an

ion results of

5° (Figs. 8a:

0.8). In stud

w field leadin

urface, to con

]. Vorticity (

the forward d

his is means

gle α = 135°

Flow

e MechanismHig

(a)

(c)ng cylinder wit°).

life circumfer

e increase in

cumferential

he lift of an

ed the influen

s at the back

ngle α = 135

f the vorticity

θ = 0.2, 8b: θ

dies of conven

ng edge vortic

ntribute to the

(blue—clockw

direction flow

s that, in the

°, the leading

m of Flow Fieldgh Performan

θ = 0.2

θ = 0.6 th straight fins

rential speed r

n lift becom

speed ratio,

ideal fluid.

nce of the lead

kward fin in

°. Thus, it sh

y for the rota

θ = 0.4, 8c: θ =

ntional airfoi

ces are attache

e generation

wise vorticity

w of the rota

e vicinity of

g edge vorte

d around a Rce Magnus W

s. Velocity dist

ratio

es a

this

This

ding

n the

hows

ation

= 0.6

ils, a

ed to

of a

y) is

ating

f the

ex is

esti

rear

the

lead

blad

W

wid

by

redu

four

is ro

Fig

fin

ups

poin

in t

Rotating CylinWind Turbine

tribution and v

imated to imp

r fin behind w

fins. In addi

ding edge v

des increases

When photog

dened. This o

the shape of

uce the blind

r synthesized

otated forwar

. 9) the upstr

was in reve

tream lower

nt, the windin

the downstrea

nder with Fins

(b) θ = 0.4

(d) θ = 0.8 vorticity, for: (

pact due to th

when the main

ition, owing

vortex, circu

, which also i

graphing a sp

occurs becaus

f the fins. Th

d area by in

d four, as show

rd when takin

ream upper s

erse rotation

surface (area

ng direction o

am of the sh

s for

4

(a) θ = 0.2, (b)

he fact that, o

nstream is to

to the occur

ulation aroun

increases lift.

piral fin, the

se a shadow

herefore, we

ndividually ph

wn in Fig. 9.

ng pictures o

surface. Subs

when photo

a ② of Fig. 9

of the fin is le

hooting, the f

29

θ = 0.4, (c) θ =

occurs in the

pass through

rrence of the

nd the rotor

.

blind area is

is generated

attempted to

hotographing

The cylinder

of (area ① of

equently, the

graphing the

9). Up to this

eft. However,

fin is in front

9

=

e

h

e

r

s

d

o

g

r

f

e

e

s

,

t

30

of the laser s

So, we cont

the direction

basis of the

area ① of F

rotation and

Fig. 9), and

surface (area

Fig. 10 is

with spiral

the flow fie

show the re

circumferen

Fig. 10 for th

the straight f

the main flow

Fig. 9 Methrotating cylin

Fig. 10 FlowVelocity distr

The

sheet. Theref

tinued taking

n to the righ

e line that se

Fig. 9. Inciden

d downstream

a reverse ro

a ④ of Fig. 9

s the flow fie

fins in the

eld at each r

esults for rot

tial speed rat

he spiral fin, w

fin result. We

w direction o

od of shootingnder with spira

w around rotribution (α = 0°

e MechanismHig

fore it is not p

g photograph

ht winding o

parates the a

ntally, we the

m upper surf

otation and do

9), which we

eld around a

mainstream.

rotation angl

tational angle

tio θ = 1. W

we compared

e examined th

of the flow rat

g photography al fins.

tating cylinder°).

m of Flow Fieldgh Performan

possible to sh

hs after chan

of the fin on

area ③ from

en had a forw

face (area ③

ownstream lo

synthesized.

rotating cylin

For spiral

le is similar,

e α = 0°, an

With the result

d it to Figs. 6a

he distributio

te in the grad

when flow aro

r with spiral

d around a Rce Magnus W

hoot.

ging

n the

m the

ward

③ of

ower

nder

fins,

, we

nd a

ts in

a-6d,

on of

dient.

ound

fins.

The

dist

the

the

rota

in t

(Y-

norm

of t

cyli

diff

add

perp

cha

the

con

4. C

In

lift-

win

the

the

resu

F

lift

1. T

redu

dec

F

the

sma

gen

con

and

lift

flui

0.2-

that

rota

T

com

Rotating CylinWind Turbine

e straight fins

tribution in a

cylinder. Th

flow to diffu

ation is gener

the flow aff

axis direction

mal shape of

the fins is con

indrical surf

ference occur

dition, since

pendicular to

anged intermi

spiral fin,

ntinuous.

Conclusion

n this stu

-generating m

nged rotating

three types o

flow field w

ults are summ

For the rotatin

continues un

Then, it slow

uced at θ > 1

rease gradual

For the rotati

forward flow

all. However

nerates a lar

ntributes to an

d circulation,

measuremen

id at the time

-0.4. This is

t occurs wh

ating cylinder

The cylinder

mpared with t

nder with Fins

s appeared to

far area from

his occurs wi

fuse the norm

rated. In com

fected by the

n) componen

f the diffusion

ncentrated in

face, and

rs in the upp

the straig

o the main

ttently by the

the change

ns

dy, we ai

mechanism in

cylinder. Flu

of models we

was analyzed w

marized as fol

ng cylinder w

ntil the circum

wly declines

, because the

lly.

ng cylinder w

w side of the

r, the fin on

rge stagnatio

n increase in

which incre

nt results exce

e the circum

the effect of

hen the mai

r with a straig

with spiral

the cylinder

s for

o affect the f

m the cylinde

ith a straight

mal shape wh

mparison, for

e fin, the sp

nt is added. T

n is suppresse

n a narrow reg

a large flo

per and lower

ght fins are

flow, the f

e rotation. Me

e in the fl

med to el

n the flow fi

uid force mea

ere performed

with PIV pro

llows:

winged fins, a

mferential sp

s. Then, lift

e role of the f

with fins, the

fin in the cas

n the revers

on area. As

n the flow ra

eases the lift.

eeded the lift

mferential spe

f the leading

instream ove

ght fin.

fins showed

with a straig

flow velocity

er away from

t fin because

hen given the

the spiral fin

pan direction

Therefore, the

ed, the action

gion near the

ow velocity

r streams. In

e positioned

flow field is

eanwhile, for

ow field is

lucidate the

eld around a

asurements of

d. Moreover,

ocessing. The

an increase of

eed ratio θ =

is gradually

fins begins to

e function of

se of θ = 1 is

e flow side,

a result, it

ate difference

. Straight fin

ft of the ideal

eed ratio θ =

edge vortex

ercomes the

d a high lift

ht fin. When

y

m

e

e

n

n

e

n

e

y

n

d

s

r

s

e

a

f

,

e

f

=

y

o

f

s

,

t

e

n

l

=

x

e

t

n

The Mechanism of Flow Field around a Rotating Cylinder with Fins for High Performance Magnus Wind Turbine

31

seen from the results of the PIV processing, the

intermittent flow field of the straight fin is changed by

the rotation angle, and the range of the velocity

fluctuation, and reaches a relatively far area from the

cylinder.

Meanwhile, the change of the spiral fins around the

flow field is continuous. In the area near the cylinder,

a large speed difference has been shown to occur. This

is the main cause of the high lift of the spiral fin.

References

[1] Nobuhiro, M. 2004. Magnus type wind power generation. Patent 2005-517,614, file June 14, 2004, and issued June 14, 2004.

[2] Michihiko, N. 2015. “Spiral Magnus Windmill.” Japan Wind Energy Association 39 (113): 14-7.

[3] Shigemitu, S., Jun, I., Nobuhiro, M., Hiromichi, K., and

Shinichi, K. 2004. “A Study on the Magnus Effect in a Rotating Cylinder with a Load Structure.” In Proceedings of the 26th Wind Energy Utilization Symposium, 155-8.

[4] Akiyosi, I., Chisachi, K., Nobuhiro, M., and Tadahiko, N. 2010. “Evaluation of Aerodynamic Properties of Magnus Wind Mill with Spiral Fins.” Transactions of the Japan Society of Mechanical Engineers B 76 (763): 379-82.

[5] Hiroaki, H., Yusuke, O., and Kazuo, M. 2006. “Relationship between Dynamic Lift and Streamwise Vortices Acting on a Disk Simulating a Hand of Swimmer.” Journal of Japan Society of Fluid Mechanics 25 (4): 357-66.

[6] Hiroaki, H., Jun, W., and Kazuo, M. 2010. “Effect of 3-Dimensional Airfoil Shape on Unsteady Fluid Forces in Pitching Motion.” Journal of Fluid Science and Technology 5 (2): 270-80.

[7] Tatjana, Y. H., and Cameron, T. 2010. “The Importance of Leading Edge Vortices under Simplified Flapping Flight Conditions at the Size Scale of Birds.” The Journal of Experimental Biology 213 (July): 1930-9.