Development of new mechanical design

of model turbines

Kristoffer Rundhaug

Mechanical Engineering
Submission date: July 2014
Supervisor: Ole Gunnar Dahlhaug, EPT
Co-supervisor: Torbjrn K. Nielsen, EPT

Norwegian University of Science and Technology

Department of Energy and Process Engineering
Preface

This master thesis has been written at the Waterpower Laboratory, Department of Energy
and Process Engineering at the Norwegian University of Science and Technology (NTNU)
during the spring semester of 2014. The aim of this thesis was to come up with a new me-
chanical design of a Francis model runner.

I would like to thank my supervisor Ole Gunnar Dahlhaug for his guideance and support
during this semester.

A special thanks to the PhD. candidates Bjrn Winther Solemslie and Peter Joachim
Gogstad for great support and discussions. They have also been very helpful with knowl-
edge of the CAD program Creo Parametric and the MATLAB design program Khoj.

In addition I would like to thank the rest of the staff at the Waterpower Laboratory, and
especially Trygve Opland for helping me weld the test rig, and Brd Aslak Brandstr and
Joar Grilstad for helping with parts needed for the testing.

I would also like to thank Anders Flaten at Protolab AS for taking time to meet and discuss
designs and manufacturing methods.

And at last, I would like to thank all my fellow students at the Waterpower Laboratory for
all good discussions and the great working environment.

Trondheim, July 22, 2014

Kristoffer Rundhaug

iii
Abstract

The production of model Francis runner to be used in the Waterpower laboratory at NTNU
is a length and costly process. Due to the production cost there are years between each
time a new Francis model is bought. The Tokke Francis model at Waterpower laboratory,
was given from the industry after they had run model tests in the laboratory when design-
ing the Tokke prototype.

In order to give the students at the Waterpower laboratory an opportunity to design their
own turbines and test, the production cost must be significantly reduced. To be able to do
that, new designs need to be developed and new production method used.

The objective of this thesis is to carry out a new mechanical design of a Francis model
runner and evaluate different mechanical designs of guide vanes and stay vanes. The new
alternative designs were made with a combination of standard production methods and
material, and new plastic materials from rapid prototyping such as selective laser sintering
(SLS) and rapid CNC machining.

To carry out the new mechanical runner design, the MATLAB design software Khoj was
used. From Khoj two sets of streamlines were generated, one set for the sucction side and
one for the pressure side. Then the streamlines were imported into the CAD program PTC
Creo Parametric and a 3D-model of the design was made.

Two test pieces in a polyamide material was made using selective laser sintering. One of
the test pieces was tested for pull-out strength using some self tapping thread inserts, the
results showed that the pull-out strength was more than good enough. The threaded insert
could be used to attach the runner vanes to the hub and ring.

The main focus of this thesis has been the new mechanical design of a model runner and
runner vanes. The model design was not produced and tested in the laboratory due to lack
of time and money.

iv
Sammendrag

Produksjonen av Francis modellturbiner for testing p Vannkratlaboratoriet ved NTNU er

omstendig og kostbar prosess. P grunn av de hye kostnadene er ofte flere r mellom
hver gang en ny Francis modellturbin blir kjpt. Tokke modellturbinen som n brukes ved
Vannkraftlaboratoriet var en gave fra industrien etter at den ble brukt til modelltesting da
de designet Tokke prototype lpehjulet.

For gi studenter ved Vannkraftlaboratoriet mulighet til prve seg p lage egne turbinde-
sign og test, m produksjonskostnadene reduseres kraftig. For kutte kostnadene m nye
design utvikles og nye produksjonsmetoder m tas i bruk.

Denne masteroppgaven har som ml komme opp med et nytt mekanisk design av en Fran-
cis modellturbin, og evaluere forskjellige mekaniske design av ledeskovler og stagskovler.
Disse nye alternative designene var produsert med en kombinasjon av standard materi-
aler og produksjonsmetoder, og ved ta i bruk nye plastmaterialer og rapid prototyping
metoder som selective laser sintering (SLS) og rapid CNC maskinering.

For utvikle det nye mekaniske lpehjul designet, ble MATLAB design programmet Khoj
brukt. Fra Khoj ble to sett med strmningslinjer generert, et sett for trykksiden og et sett
for sugesiden av skovlen. Strmningslinjene ble s importert inn i CAD programmet PTC
Creo Parametric, hvor de ble brukt til lage en 3D-modell av skovlen og det nye designet.

To test klosser laget av polyamid ble laget ved hjelp av SLS printing. En av klossene ble
testet for hvor stor uttrykkskraft som skulle til for dra ut en selvskruende gjengehylse.
Resultatet viste at styrken var mer enn god nok til at hylsene kunne brukes til innfesting
av skovlene mot hub og ring.

Hovedfokuset i denne masteroppgaven har vrt det nye mekaniske lpehjul og skovl de-
signet. Det ble uheldigvis ikke tid til produser det nye lpehjulet med ny hub og ring, og
skovler i plast. Dette var p grunn av mangel p tid og penger.

v
vi
Table of Contents

Preface iii

Abstract iv

Sammendrag v

Table of Contents ix

List of Tables xi

List of Figures xiv

Abbreviations xv

1 Introduction 1
1.1 Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

2 Theory 3
2.1 High head Francis turbine design . . . . . . . . . . . . . . . . . . . . . . 3
2.1.1 Main Dimensions . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.1.2 Energy Distributions . . . . . . . . . . . . . . . . . . . . . . . . 9
2.1.3 Simplified calculations on a high head Francis runner . . . . . . . 10
2.2 Guide vanes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.2.1 Hydraulic force acting on guide vanes . . . . . . . . . . . . . . . 12
2.3 Stay ring and Stay vanes . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.4 CAD-Tool . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.5 Design software - Khoj . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.6 Selective Laser Sintering - SLS . . . . . . . . . . . . . . . . . . . . . . . 14
2.7 Rapid CNC Machining . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.8 Materials used in prototype and model turbines . . . . . . . . . . . . . . 16

vii
3 Mechanical turbine design 17
3.1 New mechanical design of model turbines . . . . . . . . . . . . . . . . . 17
3.1.1 Francis model runner design . . . . . . . . . . . . . . . . . . . . 17
3.1.2 Guide vane design . . . . . . . . . . . . . . . . . . . . . . . . . 22
3.1.3 Stay vane and stay ring design . . . . . . . . . . . . . . . . . . . 25

4 Methode 27
4.1 How to design a Francis runner to be used in the laboratory . . . . . . . . 27
4.1.1 Turbine design in MATLAB program Khoj . . . . . . . . . . . . 27
4.1.2 Drawing of a Francis runner vane in Creo Parametric . . . . . . . 31
4.1.3 Drawing hub and ring in Creo Parametric . . . . . . . . . . . . . 34

5 Experiment 37
5.1 Material tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
5.1.1 Test 1: Pull-out strength test . . . . . . . . . . . . . . . . . . . . 38
5.1.2 Test 2: Compression test . . . . . . . . . . . . . . . . . . . . . . 39
5.1.3 Test 3: Torque test . . . . . . . . . . . . . . . . . . . . . . . . . 39
5.2 Results from material tests . . . . . . . . . . . . . . . . . . . . . . . . . 40

6 Discussion 41
6.1 The new runner and runner vane designs . . . . . . . . . . . . . . . . . . 41
6.2 The new guide vane and stay vane designs . . . . . . . . . . . . . . . . . 42
6.3 Manufacturing methods and materials . . . . . . . . . . . . . . . . . . . 43
6.4 Production of runner blades and model test . . . . . . . . . . . . . . . . 44
6.5 FEM analysis of the new runner design . . . . . . . . . . . . . . . . . . . 44

7 Conclusion 45
7.1 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

8 Further Work 47
8.1 Further Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

Bibliography 49

A Appendix A i
A.1 Guide vane calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . i
A.1.1 Hydraulic force on guide vane in closed position . . . . . . . . . ii

B Appendix B iii
B.1 Price offer from Protolab AS . . . . . . . . . . . . . . . . . . . . . . . . iii
B.1.1 Price on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iii

C Appendix C ix
C.1 Material properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix

D Appendix D xiii
D.1 Drawings and documents from Tokke model turbine . . . . . . . . . . . . xiii

viii
E Appendix E xix
E.1 Product informastion sheets for the fixing parts . . . . . . . . . . . . . . xix

ix
x
List of Tables

4.1 Input data from Tokk used in Khoj [4] . . . . . . . . . . . . . . . . . . . 27

A.1 Data from Tokke model test used in calculations . . . . . . . . . . . . . . i

xi
xii
List of Figures

2.1 Axial view of high head Francis runner [12] . . . . . . . . . . . . . . . . 3

2.2 Velocity Triangles [10] . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.3 Velocity Triangles corrected [12] . . . . . . . . . . . . . . . . . . . . . . 6
2.4 Thoma cavitation coefficient [12] . . . . . . . . . . . . . . . . . . . . . . 7
2.5 The figure are showing U Cu blade loading from inlet to outlet [12]. . . 9
2.6 Principal sketch of high head Francis runner [3] . . . . . . . . . . . . . . 10
2.7 Model of runner for strength calculations [3]. . . . . . . . . . . . . . . . 10
2.8 Principal sketch of high head Francis runner [3] . . . . . . . . . . . . . . 11
2.9 Laser sintering and laser melting; contouring and recoating process [7, p.41] 14

3.1 3D-model of the Tokke model runner at NTNU . . . . . . . . . . . . . . 17

3.2 3D-model of the Runner and runner vanes Design 1 . . . . . . . . . . . . 18
3.3 3D-model of the Runner and runner vanes Design 2 . . . . . . . . . . . . 19
3.4 3D-model of the Runner and runner vanes Design 3 . . . . . . . . . . . . 20
3.5 3D-model of the Runner and runner vanes Design 4 . . . . . . . . . . . . 21
3.6 This figure show how the runner vanes in Design 4 will be attached to the
hub and ring. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
3.7 This figure showes the U Cu energy distribution in design 4, the figure is
from Khoj. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
3.8 3D-model of the Guide vane: Design 1 . . . . . . . . . . . . . . . . . . . 23
3.9 3D-model of the Guide vane: Design 2 . . . . . . . . . . . . . . . . . . . 24
3.10 3D-model of the Stay vane: Design 1 . . . . . . . . . . . . . . . . . . . . 25

4.1 View of the Main Dimensions tab in Khoj . . . . . . . . . . . . . . . . . 28

4.2 View of the Axial View tab in Khoj . . . . . . . . . . . . . . . . . . . . . 28
4.3 View of the Distributions tab in Khoj . . . . . . . . . . . . . . . . . . . . 29
4.4 View of the Radial View tab in Khoj . . . . . . . . . . . . . . . . . . . . 29
4.5 View of the Blade Thickness tab in Khoj . . . . . . . . . . . . . . . . . . 30
4.6 View of the Summary tab in Khoj . . . . . . . . . . . . . . . . . . . . . 30
4.7 Making a quilt using streamlines from Khoj and Boundary Blend . . . . . 31

xiii
 4.8 Bondary Blend at leading edge and merging hub and ring . . . . . . . . . 32
4.9 Drawing of the hub and shroud sections . . . . . . . . . . . . . . . . . . 32
4.10 How to make a "slice" of the hub . . . . . . . . . . . . . . . . . . . . . . 33
4.11 Drawings of hub in Creo . . . . . . . . . . . . . . . . . . . . . . . . . . 34
4.12 Finish hub from Design 4 . . . . . . . . . . . . . . . . . . . . . . . . . . 34
4.13 Axial view of the ring sketch in Creo . . . . . . . . . . . . . . . . . . . . 35
4.14 Finish ring from Design 4 . . . . . . . . . . . . . . . . . . . . . . . . . . 35

5.1 Model of the test pieces ordered . . . . . . . . . . . . . . . . . . . . . . 37

5.2 The test rig used for pull-out strength testing. . . . . . . . . . . . . . . . 38
5.3 The figure are showing a 4mm pin attached to the test piece and with the
Ensat S inserted. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
5.4 Pull-out strength testing . . . . . . . . . . . . . . . . . . . . . . . . . . . 40

6.1 Showing pull-out strength for Kerb Konus Ensat with different thread di-
ameter in different plastic materials [9]. . . . . . . . . . . . . . . . . . . 43

B.1 Price offer for a full set of runner vanes to the Tokke model runner . . . . iv
B.2 Price offer for a couple of test vanes to the Tokke model runner . . . . . . v
B.3 Price offer for test piece in SLS material DuraForm HST . . . . . . . . . vi
B.4 Price offer for test pice in SLS material DuraForm PA . . . . . . . . . . . vii

C.1 Material range at ARRK for SLS printing . . . . . . . . . . . . . . . . . ix

D.1 Axial view of the Tokke model runner . . . . . . . . . . . . . . . . . . . xiv

D.2 Drawing of the guide vane . . . . . . . . . . . . . . . . . . . . . . . . . xv
D.3 Guide vane torque for VKL-runner [1] . . . . . . . . . . . . . . . . . . . xvi
D.4 Drawing of the stay vane shapes . . . . . . . . . . . . . . . . . . . . . . xvii
D.5 Drawing of the stay ring . . . . . . . . . . . . . . . . . . . . . . . . . . xviii

E.1 Product info about Kerb Konus threaded inserts Ensat S . . . . . . . . . . xx

E.2 Product info about Kerb Konus threaded inserts Ensat SK . . . . . . . . . xxi
E.3 Product info about Kerb Konus threaded inserts Ensat SI/SBI . . . . . . . xxii
E.4 Product information about the DIN 7 pins . . . . . . . . . . . . . . . . . xxiii

xiv
Nomenclature
Symbols
Symbol Definition Unit
A Area [m2 ]
B Vane hight [m]
D Diameter [m]
F Force [N ]
g Gravity [m/s2 ]
H Head [m]
h Head [m]
IP Polar moment of inertia [m4 ]
Ix Moment of inertia [m4 ]
n Rotational speed [RP M ]
M Moment [m/s]
P Power [W ]
p Pressure [P a]
Q Flow rate [m3 /s]
q Load [N ]
R Radius [m]
r Radius [m]
T Torque [N m]
t Thickness [m]
U Peripheral velocity [m/s]
W Relative velocity [m/s]
Z Number of pole pairs, runner vanes, guide vanes [m/s]

xv
Greek Symbols
Symbol Definition Unit
Guide vane angle [ ]
Blade angle [ ]
Efficiency [ ]
Density [kg/m3 ]
Speed number [ ]
! Angular velocity [rad/s]

Sub-Symbols
Refers to best efficiency point
A Refers to available
atm Atmospheric pressure
e Refers to effective
g Refers to guide vane
h Refers to hydraulic
m Refers to meridional direction
max Refers to maximum value
min Refers to minimum value
P Refers to generator pole pairs
R Refers to required
r Refers to runner
sv Refers to stay vanes
u Refers to peripheral direction
va Vapor
0 Refers to
1 Refers to runner inlet
2 Refers to runner outlet

xvi
Abbreviations
AM = Additive Manufacturing
BEP = Best Efficiency Point
CAD = Computer-Aided Design
CNC = Computer Numerical Control
FEM = Finite Element Methode
NPSH = Net Positive Suction Head
NTNU = Norwegian University of Sience and Technology
OS = Operating System
PA = Polyamide
PTC = Parametric Technology Corporation
RPM = Revolutions Per Minute
SLS = Selective Laser Sintering
VKL = Vannkraftlaboratoriet (Waterpower Laboratory)

xvii
xviii
Chapter 1
Introduction

In Norway about 99% of all the electricity production is coming from hydropower. There
are power plants all over the country, and a lot of them are built back in the 60s and 70s,
the need for upgrading and replacing of old turbines is coming. The low electricity prices
in Norway have made it hard for the energy producers to conduct new profitable develop-
ment projects and larger maintenance project on old power plant. In order to keep the cost
down, better and cheaper technologies and manufacturing methods need to be developed.
To be able to do that a lot of research needs to be done.

When new development projects and larger maintenance projects are being conducted, a
model turbine is often manufacture and model tests are conducted to find the best design
for the prototype runner. Sometimes the industry uses the Waterpower laboratory at NTNU
to run their model tests, and if they do not need the model afterwards they are giving it to
NTNU. The Tokke model runner was given to NTNU from the industry.

1.1 Objectives
The main objectives for this thesis is come up with new mechanical designs of Francis
model runners, guide vanes and stay vanes. The new designs should combined parts made
by traditional manufacturing methods and new methods like rapid prototyping.

1.2 Background
Turbines and pumps consist of complex geometries that are expensive to manufacture. The
production of model Francis turbines to be used in the Waterpower laboratory at NTNU
is a lengthy and costly process. The arrival of rapid prototyping opens for new designs of
pumps and turbine.

1
Chapter 1. Introduction

The Tokke model turbine and prototype have been the basis for the development of the
new mechanical Francis model turbine design. The Tokke model test report with all the
results from tests and drawings of the Francis model runner in the Waterpower laboratory,
have been very helpful.

If the students at the Waterpower laboratory could have design their own turbine and test
it in the laboratory, it would have been very educational for the student. But to be able to
do that the manufacturing cost need to be significantly reduced.

2
Chapter 2
Theory

2.1 High head Francis turbine design

A Francis turbine consitst of a spiral casing, a set of stay vanes, a set of guide vanes and a
runner see figure 2.1. When designing a Francis turbine it starts with the runner and then
the guid vanes, stay vans and spiral casing can be designed.

Figure 2.1: Axial view of high head Francis runner [12]

Francis runners are reaction turbines and there is a pressure difference from inlet to the
outlet of the runner. At the runner inlet part of the specific energy is pressure energy. From
the flow trough the runner the spcific energy is converted to mechanical energy partly
from drop in pressure and partly from the impulse force due to changes in direction of the
relative velocity. The turbine has to be compleatly filled with water to obatain the pressure
drop through the runner and Francis turnbines are therefore also known as full turbines.[2].

3
Chapter 2. Theory

2.1.1 Main Dimensions

Designing a Francis runner starts with geting the hydraulic data from the power plant site,
these are based on the topography and hydrology. From thes data the total net head, Hn ,
total flow rate, Q. Then the next is to start dimensioning the outlet of the runner.

Figure 2.2: Velocity Triangles [10]

Notations used in the figure 2.2:

C Absolute velocity [m/s]
Cm Meridian component of C [m/s]
Cu Tangential component of c [m/s]
U Peripheral velocity [m/s]
W Relative velocity [m/s]
Angle between the relative and peripheral velocit [ ]
1 Inlet [ ]
2 Outlet [ ]

High head Francis turbines have an increasing difference betwen the inlet diameter and the
outlet diameterfor for increasing net head.

4
 2.1 High head Francis turbine design

Dimensions at runner outlet

The outlet is a critical point for cavitation, and it is appropriate to start the dimensioning
at the outlet. First the peripheral velocity, U2 , and the outlet angle, 2 , are choosen from
empirical data. The values can be found in the interval [2]:

15 < 2 < 22 (2.1)

32m/s < U2 < 43m/s (2.2)

The higher the velocity, U2 , is, the higher head and the smaller the outlet angle, 2 is, the
higher head. When dimensioning the outlet it is assumed best efficiency point (BEP), and
at BEP there is no swirl in the draft tube. It can then be assumed that:

CU2 = 0 [m/s] (2.3)

Further Cm2 is found by using the known values in the velocity triangles (figure 2.2), see
equation 2.4.

Cm2 = U2 tan 2 [m/s] (2.4)

Now the outlet diameter, D2 , can be found by using the ralations between D2 and Cm2 ,
equation 2.5.
r
4Q
D2 = [m] (2.5)
Cm 2
By knowing the outlet diameter, D2 , the rotational speed, n, can be calculated using equa-
tion 2.6:

U2 60
n= [RP M ] (2.6)
D2
The number pole pairs in the generator, ZP , depends on the rotational speed, n, and the
grid freqaency, fgrid . In Norway the grid frequency is 50Hz and the number of pole pairs
is found by using equation 2.7.

fgrid 60
ZP = [ ] (2.7)
n
There are requirements that the turbine need to have a synchronous rotational speed to be
attached to a gridd with constant frequency, and therefor the number of pole pairs,ZP , has
to be an integer. In order to get synchronous speed, the value of ZP has to be round up or
down. From the new ZP the new corrected synchronous speed, nCorr , is calculated. To
keep the outlet blade angle, 2 , for this geomtry, a new corrected outlet diameter, D2,Corr ,
is calculated from the synchronous speed [2].

Keeping flow rate,Q, and outlet angle, 2 the same and CU2 = 0 the new corrected and
the old velocity triangle at the outlet will have geometric similarity, see figure 2.3.

5
Chapter 2. Theory

Figure 2.3: Velocity Triangles corrected [12]

4Q 4Q
Cm2 Cm2,Corr D22
2
D2,Corr
tan 2 = = = nD2
= nCor D2,Corr
[ ] (2.8)
U2 U2,Corr 60 60
Rearranging the equation 2.8 giving:
3
nCorr DCorr = nD23 (2.9)
Now the new corrected outlet diameter, D2,Corr , can be calculated using equation 2.10.
s
nD23
(2.10)
3
D2,Corr = [m]
nCorr

Submerging
When designing high head Francis turbines the level of submergence need to be calculated.
If the turbine is not sufficiently submerged cavitation may occure. To avoid cavitation
the water pressure in the runner need to be higher than the vapor pressure. The level of
submergence is expressed as Net Positive Suction Head (NPSH), and is found by using the
equation 2.11.
2
Cm2 U2
N P SHR = a +b 2 [m] (2.11)
2g 2g
The parameters a and b are depending on the speed number, , and based on empirical
data they are found in the interval [2]:

1.05 < a < 1.15 (2.12)

0.05 < b < 0.15 (2.13)

6
 2.1 High head Francis turbine design

The speed number,, are an dimensionless number that is an expression for the rotational
speed at a given head at BEP, and it can be found from the equation 2.14:
q
=! Q [ ] (2.14)

NPSH are dependent of the runner geometry and need to fulfill the following requierment
to avoid cavitation:

N P SHR < hatm hva HS = N P SHA (2.15)

Using the Thoma cavitation number, , is an other method to calculate how much the
turbine need to be submerged. The Thoma cavitation coefficient is found by calculating
the speed numer, for the runner and then use the graph in figure 2.4 to get [12].

Figure 2.4: Thoma cavitation coefficient [12]

When is found, the level of submergence is calculated using the equation 2.16.

Hs < 10 He (2.16)
If the turbine not is sufficient submerged and there is not possible to move the turbine,
then it is necessary reduce the rotational speed, n. Reducing n will need a generator with
a larger numer of pole paris, ZP , and due to that the size of the runner will increase.

7
Chapter 2. Theory

Dimensions at runner inlet

When the dimensioning the inlet the inlet diameter, D1 , inlet height, B1 , and the inlet
angle, 1 are calculated. For the dimensioning of the inlet empirical values are needed.
Reduced values are introduced in the design process, and they are indicated with a under-
line. The dimensioning starts with the Euler turbine equation:

U1 Cu1 U2 Cu2
h = [ ] (2.17)
gHe
By introducing the reduced values and rearranging the equation 2.17, assuming BEP and
no rotation in the outlet, Cu2 = 0, the Euler equation is reduced to:

h = 2U 1 C u1 [ ] (2.18)
The hydraulic efficiency, h , is normaly set to h = 0.96. The relation between U1 and
Cu1 are choosen to minimize losses by precussion. For high head Francis runner the re-
duced peripheral velocity, U1 in the range 0.71 0.73 will meet the requirement. The
equation 2.18 is used to find Cu1 and then the equation 2.19 is used to find U1 [2].

p
U1 = U1 2 g He [m/s] (2.19)
Now the inlet diameter can be found using equation 2.20:
2U 1 U1 60
D1 = = [m] (2.20)
! n
Using the velocity triangle (see figure 2.2) for the inlet, an expression for the inlet angle,
1 , can be found.

Cm1
tan 1 = [ ] (2.21)
U1 Cu1
In order to calculate the inlet angle, 1, Cm1 need to be found. Using the continuity
equation is giving:

Cm1 A1 = Cm2 A2 (2.22)

Based on experiences from turbine manufacturers in Norway, an acceleration of 10%
through the runner are desirable. [2]. That gives:

Cm2 = 1.1 Cm1 [m/s] (2.23)

The inlet height, B1 , can now be found using equation 2.24

1.1 D2 2
B1 = [m] (2.24)
4 D1

8
 2.1 High head Francis turbine design

2.1.2 Energy Distributions

The energy distribution are showing how the energy is distributed along the blades through
the runner. The distribution describes the transformation from pressure energy to ratational
energy.

There is two ways to calculate the distribution, one way is to choose the U Cu distribution,
and calculate the distribution. The other way is to choose and then calculate U Cu
[8]. The calculations are done using the equation 2.25.
Cm1
= arctan [ ] (2.25)
U1 Cu1
For high head Francis turbines, it is desirable to utilize most of the energy at a relatively
large diameters. The blades are thickest at the leading edge to withstand the high tensions
due to the large pressure differens. The blades becomes thinner towards the trailing edge
due to lower pressuer differences and less tension. Avoiding a larege slip angle is desirable
to reduce the risk for cavitation. For a good blade design, the Cu component in one point
along a streamline should never be lager than the U component [5].

Figure 2.5: The figure are showing U Cu blade loading from inlet to outlet [12].

9
Chapter 2. Theory

2.1.3 Simplified calculations on a high head Francis runner

In a Francis runner the strength calculations is due to the complex geometries and load
conditions very difficult to determine exactly. Therfore simplifications is necessary to be
able to do some calculations.

Figure 2.6: Principal sketch of high head Francis runner [3]

In high head Francis runners most of the torque are transmitted to the hub at reltively
large diameters. For simplifying the calculations it is assumed that the whole torque is
transformed outside RT , see figure 2.6.
R 1 + RT
Rm = , and a = R1 RT [m] (2.26)
2
The load, p, can be calculated from the equation 2.27, by assuming the torque, Mr , is
uniformly distributed on the runners inlet part, from RT to the inlet.
P
Mr = Z a b Rm p= [N m] (2.27)
!

Figure 2.7: Model of runner for strength calculations [3].

10
 2.1 High head Francis turbine design

Assuming that the pressure load, p, is uniformly distributed over a vane gives:
P 30 P
p= = [P a] (2.28)
Z a b Rm ! Z a b Rm n
The load, q, over an radiell extension, r, on the vane:

q= p r [N/m] (2.29)
From a simplified method for calculating bending stress it can be assumed that the stiffnes
of the hub and ring are very large compared to the vane. Assuming this the bending stress
can be calculated using the formulas for a beam that is fixed at one end, see figure 2.8. The
equation for the torque is then:

b2
M =q [N m] (2.30)
3

Figure 2.8: Principal sketch of high head Francis runner [3]

The maximum bending stress is then:

Mt
= [P a] (2.31)
Iz 2
3 0
Where Iz = 12 rt
, and t = sint is the vane thickness in tangential direction. Inserting the
equations 2.29 in the equation 2.30 and then into 2.31, the equation for maximum bending
stress becomes:

6M 2 b2 p
= 2
= [P a] (2.32)
rt t2

This equation can be used to roughly give an approximation of the stresses in the runner.

11
Chapter 2. Theory

2.2 Guide vanes

The guide vanes consit of a number of adjustable blades, they are adjusted to regulate the
flow direction and the flow rate into the runner wheel. The guide vanes are adjusted with
a servo motor that by the regulating ring, lever arms and links are turning the guid vanes
into the desired position. The guide vanes also need to withstand the bending moment and
the torque from the hydraulic forces [3].

To minimize the pressure pulsation due to the blade pasing frequence, the number of guide
vanes has to be choosen according to the equation 2.33 [8]
N umber of Guide vanes
6= Integer (2.33)
N umber of Runner V anes

The Francis model test rig on the Waterpowerlaboratory at NTNU has 28 guide vanes.

2.2.1 Hydraulic force acting on guide vanes

The hydraulic pressure after the guide vanes changes with the RPM of the runner and the
pressure increases with increasing of RPM. The pressure upstream increases with decreas-
ing guide vanes opening. The highest pressure acting on the guid vanes is when they are
in colsed possision.

Guide vanes in closed position

The force acting on each guide vane when they are in closed possision can be calculated
using the following equations.
gHB0 D0
F0 = gHB0 l = [N ] (2.34)
z0
The torque, T0 , on the guide vane tap is then:
gHB0 D0 a0
T 0 = F0 a 0 = [N m] (2.35)
z0
The polar moment of inertia, IP for the guide vane shaft:

r4
IP = [m4 ] (2.36)
2
The maximum shear stress, max , due to torque can then be found:
T0
max = max = r [P a] (2.37)
IP

12
 2.3 Stay ring and Stay vanes

The stress in the transition between the shaft and the guide vane blade could be controlled
by doing som estimate calculations :
M r
b = [P a] (2.38)
IX
The bending moment at the transition between the guide vane blade and the shaft can set
as:

g H0 B 0 2 D 0
M= [N m] (2.39)
8 z0
The lages pricipal stress can be calculated by:
r
2
b b
max = + 2 [P a] (2.40)
2 4

2.3 Stay ring and Stay vanes

The stay vanes are design to have no hydraulic effect, they have the main purpose of
keeping the spiral casing together. The dimensions have to be given due to stress in the
spiral casing.

2.4 CAD-Tool
During this thesis the CAD program PTC Creo Parametric 2.0 has been used to draw all
the different designs of the runner vanes, guide vanes and stay vanes. The program is made
by the computer softwear company Parametric Technology Corporation (PTC), which has
specializing in 2D and 3D design software. The program PTC Creo was earlier known as
Pro/Engineer. Creo Parametric runs on Microsoft Windows operating systems (OS) [13].

2.5 Design software - Khoj

The design software Khoj is a MATLAB program that was developed for designing Fran-
cis turbines. This program was developed by Kristin Gjster in her project and master
thesis [8]. Khoj has been further improved by Peter Joachim Gogstad in co-operation with
Gjster during his master thesis [10]. This design software has a graphical user interface
(GUI) and this makes it easy to access all the different design parameters. For more de-
tails about how this program is working and how it was developed take a closer look at
Gjster and Gogstads master theses [8] [10].

13
Chapter 2. Theory

2.6 Selective Laser Sintering - SLS

Selective Laser Sintering is a "3D-printing" method. The sintering process does not need
any support structure during the build, because the unsintered powder surrounds and sta-
bilised the part during the build process.

Laser sintering or selective laser sintering are terms that are used for printers that process
plastic materials. There are many different material types from flexible "rubber like" ma-
terials to strong composite plastics. The types of materials vary from printer manufacture.
The production method is commercialised by the U.S. company 3D Systems Inc. and the
German company EOS GmbH.

Figure 2.9: Laser sintering and laser melting; contouring and recoating process [7, p.41]

The technologies used in the machines from the two manufacturers are almost the same.
The machines consist of a build chamber that is filled with powder with grain size smaller
than 50m and a laser unit on top that melts the contour of each layer in the X- and Y-
direction. The build chamber is designed as a piston that can be adjusted in the Z-direction
and after solidification of the contour of one layer the piston is moved down by one leyer
thickness. The space on the top of the powder is then filled with new powder that is ejected
from a feed chamber by a roller. To spread the powder uniformly the roller rotates counter
clockwise to its linear movement and the procedure of loading a new layer is often called
recoating. An illustration of the process can be seen in figure 2.9. To minimise the power
consumption of the laser and to avoid oxidation the build chamber is preheated and flooded
by a shielding gas [7, p.40-42].

The layer thickness in SLS printers are normally around 0.1mm and therefore post-processing
can be necessary to achieve a good surface quality. Sanding, coating or spray painting can
do this. In Appendix C see figure C.1 a material data sheet for SLS material delivered by

14
 2.6 Selective Laser Sintering - SLS

Arrk Europe Ltd.

15
Chapter 2. Theory

2.7 Rapid CNC Machining

Rapid CNC machining starts with a solid block in desired material, the CNC machine are
removing material layer by layer until the shape of the model is finished. The machines
are using 3D CAD data to make the pattern for machine; the pattern for each layer is the
contour of the 3D model. It is almost exactly the opposite of 3D printing where the part is
build layer by layer instead of removed. Since the machines are starting with a solid block
of material, almost every possible materials could be chosen [6].

Protolab AS are using Arrk Europe Ltd as their manufacturer of prototypes. Arrk have their
rapid CNC machines facilities in UK and the Far East. Arrk offers a range of machines
with 5 and 3 axis capability and a wide range of standard materials [11].

2.8 Materials used in prototype and model turbines

Over the past decades the development of stronger and lighter materials, new production
method, and new turbine designs have made the turbines larger, lighter and stronger. This
has made it possible to utilize higher head for Francis turbines; now high head Francis
runners are used where only Pelton turbines were used earlier. This sets high demands to
the materials used. The turbine components have different requirements for materials due
to different working environment.

In large high head prototype turbines the guide vanes and runner with labyrinth sealing are
made of stainless steel, often steel grades such as 13Cr4N i and 16Cr5N i. These parts
can be exposed cavitation and erosion due to water flow with high velocity and therefore
high quality steel is used [3].

In the Francis turbine rig on the Waterpower Laboratory at NTNU the runner hub and ring
are made of Bronze JM-7 and runner vanes are made of Bronze JM-3. The guide vanes are
made of stainless steel SIS 2343 and the stay vanes are made of aluminum EN-AW 6082
[1].

The Bronze JM-7 alloy used in the model runner consist of 80% copper, 10% aluminum,
5% iron and 5% nickel. The Bronze JM-3 alloy used in the runner vanes is a mixture of
88% copper and 12% tin. The JM-3 alloy is a bit softer than the JM-7 and is therefore also
used in the upper- and lower labyrinth sealing.

16
Chapter 3
Mechanical turbine design

3.1 New mechanical design of model turbines

In this section different new mechanical design of Francis model runners, guide vanes
and stay vanes are described. The designs are based on the Tokke model turbine at the
Waterpower laboratory at NTNU and the Tokke prototype runner. All designs are made so
they would fit the Francis model test rig at NTNU.

3.1.1 Francis model runner design

In these subsections four different runner designs are described. All the designs are com-
bining old and new manufacturing methods and different types of material.

Figure 3.1: 3D-model of the Tokke model runner at NTNU

17
Chapter 3. Mechanical turbine design

Runner and runner vanes Design 1

The first design considered was to use the existing Tokke turbine model, and just replaces
the runner vanes with some made of a plastic material. The plan was to manufacture the
runner vanes by rapid prototyping or rapid CNC machining. The Tokke model runner
consists of 30 vanes, where 15 of the vanes are half-vanes and the other 15 are normal
vanes.

(b) Exploded view of the whole runner De-

(a) Runner vane Design 1 sign 1

Figure 3.2: 3D-model of the Runner and runner vanes Design 1

To attach the parts together there were two alternatives; the first one is to produce the run-
ner vanes with the existing holes a bit smaller and use self-tapping screws to attach it to
the hub and ring. The second alternative is to make the holes bigger to insert self-tapping
threaded inserts with internal M5 threads and use the existing M5 bolts to attach it to hub
and ring.

The prototype company Protolab As in Trondheim was contacted to get a price estimate
for manufacturing of all the 30 runner vanes. A price offer for manufacturing a couple of
test vanes was also given. The price was given for SLS printing and using the PA plastic
material DuraForm HST. The price offers can be seen in the figures B.1 and B.2 in the
Appendix B.

18
 3.1 New mechanical design of model turbines

Runner and runner vanes Design 2

The second design was to manufacture a new hub and ring out of bronze or stainless steel
and the runner vanes of plastic. The runner is designed with 17 vanes, the same as the
Tokke prototype runner. The shape of the blades is designed in Khoj, and is the same as
Tokke. The hub and ring are made with a smooth surface with no grooves for the vanes,
only holes at the leading edge for throug bolts. The outside of the runner is made to fit the
test rig at NTNU.

(b) Exploded view of the whole runner; De-

(a) Runner vane Design 2 sign 2

Figure 3.3: 3D-model of the Runner and runner vanes Design 2

To hold the parts there will be a through bolt at the leading edge of the vanes and through
the hub and ring. This will provide a good strength to hold the parts together. To attach the
vanes to the hub and ring, it will be used glue and some small pins may be used to avoid
slippage.

19
Chapter 3. Mechanical turbine design

Runner and runner vanes Design 3

The third design is almost the same as the Design 2 but to get better support for the runner
vanes a groove is machined around the hub and ring. All the runner vanes have a section
that is 360
17 and it will fit the groove and provide good support, see figure 3.4b.

(b) Exploded view of the whole runner; De-

(a) Runner vane Design 3 sign 3

Figure 3.4: 3D-model of the Runner and runner vanes Design 3

To hold the parts together there will be used glue along the blades and in the grooves. In
this case it would be enough to use pins on just a couple of vanes to avoid slippage because
they are supporting each other.

20
 3.1 New mechanical design of model turbines

Runner and runner vanes Design 4

The fourth and last design is to make the runner vanes as slices with a section of the hub
and ring attached to it 3.5. The hub and ring should be made of stainless steel or bronze
and the runner vanes of a plastic material. This way CNC milling would be used to man-
ufacture the hub and ring and the runner vanes could be manufactured by SLS printing or
by rapid CNC machining.

(b) Exploded view of the whole runner; De-

(a) Runner vane Design 4 sign 4

Figure 3.5: 3D-model of the Runner and runner vanes Design 4

To keep the runner together there are threaded rods at the leading edge with a threaded
insert at the hub and ring side. The threaded inserts are screwed inside the hub and ring
sections of the runner vanes, and they are mounted with a fixed distance between each
other to keep a constant inlet height. At the ends of the threaded rods there are M5 nuts to
tightening together the hub and ring, see figure 3.6.

21
Chapter 3. Mechanical turbine design

Figure 3.6: This figure show how the runner vanes in Design 4 will be attached to the hub and ring.

For this design a new blade shape was made. The new shape was designed by setting the
U Cu energy distributions to a linear distribution, where the energy is equally distributed
along the blade. This was to get a unstable runner and try to provoke some vibrations [4].
The energy distribution can be seen in figure 3.7.

Figure 3.7: This figure showes the U Cu energy distribution in design 4, the figure is from Khoj.

3.1.2 Guide vane design

In this section some new mechanical design of guide vanes are described. The goal was to
come up with a new design that reduces the manufacturing and material cost.

22
 3.1 New mechanical design of model turbines

Guide vane: Design 1

The first guide vane design is to manufacture the guide vane blade in a plastic material
and the guide vane tap and shaft in stainless steel or bronze to keep a good strength and
withstand the torque. The guide vane shaft and tap are produced with a "disc" at the end,
with five holes. There are also five holes on the top and bottom of the guide vane blade;
the holes are 3mm in diameter and 3mm deep. Inside the holes on the shaft and tap pins
are glued, and then glued to the guide vane blade. These pins are used to make the guide
vanes withstand the hydraulic force and torque acting from the water pressure.

(a) Guide vane parts Design 1 (b) The whole guide vane

Figure 3.8: 3D-model of the Guide vane: Design 1

23
Chapter 3. Mechanical turbine design

Guide vane: Design 2

The second guide vane design is the same as Design 1 with plastic vane and steel shaft
and tap, but here the "disc" that is a part of the shaft and tap in Design 1 is here a part of
the vane. This is to make a stronger transition that is absorbing the forces from the water
pressure. On the top and bottom of the vane there is a hole for the shaft and tap, and it is
shaped almost like a keyhole, see figure 3.9a. The purpose of it is prevent the shaft and tap
from spinning when the guide vanes are adjusted or when there is a torque acting on the
guide vanes due to the flow.

(a) Guide vane parts Design 2 (b) The whole guide vane

Figure 3.9: 3D-model of the Guide vane: Design 2

24
 3.1 New mechanical design of model turbines

3.1.3 Stay vane and stay ring design

In this section one alternative design of the stay ring and stay vanes are described.

(a) Stay vane Design 1 (b) Turbine with guide vanes and stay vanes

Figure 3.10: 3D-model of the Stay vane: Design 1

Stay vane: Design 1

The new mechanical design of the stay vanes have the same shape as the existing vanes
in the Francis test rig at the Waterpower laboratory. Normally the vanes are welded to the
spiral casing and the stay ring, and hold the spiral casing together. This design are using
stay vanes made of plastic and welding will not be possible, therefore large steel bolts
through the stay vanes, stay ring and the spiral casing will be used. There are threaded
inserts at the top and bottom of the stay vanes to keep a constant inlet height on leading
edge of the stay vanes and preventing the spiral casing from collapsing. To avoid leakage,
the stay vanes are also glued to the stay ring as a sealing.

25
Chapter 3. Mechanical turbine design

26
Chapter 4
Methode

4.1 How to design a Francis runner to be used in the lab-

oratory
In this section it is described how a Francis model runner can be designed by using the
MATLAB program Khoj to get two sets of streamlines, one for the pressure side and one
for the suction side. These streamlines are used to make a CAD model of the runner vane.
The Tokke prototype runner was used as a reference design in Khoj and scaled down by a
factor of 5.1 in Creo to get the size of the model runner.

Table 4.1: Input data from Tokk used in Khoj [4]

Tokke input data

Head H = 377[m]
Flow rate Q = 377[m3 /s]
Number of pole pairs ZP = 8
Rotational speed n = 375[rpm]
Inlet hight B1 = 0.3[m]
Outlet diameter D2 = 1.78[m]
Number of blades 17

This section describes how the model turbine in in Design 4 can be drawn.

4.1.1 Turbine design in MATLAB program Khoj

To star the program Khoj the file RunMe.m is run in MATLAB. After running the file Khoj
will start and it is shown with a GUI that make it easy to use. In the Intro tab it is possible
to choose between starting a new project, load an earlier project or compare with CFX. To
make a new design just push the New Project button and select the tab Main Dimensions.

27
Chapter 4. Methode

Main Dimensions
In the graphical view of the Main Dimension tab the main dimension of the turbine is
inserted, see figure 4.1. Here the values for Head, H, Flow, Q, number of pole pair in
generator, Zp , outlet diameter, D2 , inlet height, B1 and reduced inlet peripheral velocity,
U1 is submitted. The reduced inlet peripheral velocity is changed by sliding the bar, see
figure 4.1. The velocity triangles for inlet and outlet is shown in Main Dimensions and
they are changing by changing values.

Figure 4.1: View of the Main Dimensions tab in Khoj

Axial View
In the graphical view of the Axial View tab the axial view of the streamlines are shown.
Here the number of streamlines is chosen, the height difference from the leading edge to
the trailing edge is set, b. There is four control points with X and Y coordinates to adjust
the shroud line data, and four points to adjust the trailing edge. Thes point can be changed
by typing in the X and Y values for the points or by moving the points around, see figure
4.2. There are also two tab for resetting changes on the trailing edge or the shroud line. If
the streamlines will be used to make a 3D-model the number of lines should be less than
ten to get a smooth surface of the runner vanes.

Figure 4.2: View of the Axial View tab in Khoj

28
 4.1 How to design a Francis runner to be used in the laboratory

Distributions
In the graphical view of the Distributions tab the energy distributions on the runner vane
is shown. There is possible to choose between U Cu distribution and Beta, distribution.
The distributions can be changed by moving the three points around, see figure 4.3.

In this case the U Cu distribution was chosen and the distribution line was set to be an
almost linear line, see figure 4.3. This was to get a model with runner vanes that may
create vibrations. The distribution shown is from the Francis runner vane: Design 4.

Figure 4.3: View of the Distributions tab in Khoj

Radial View
In the graphical view of the Radial View tab the blade leaning is set by moving three points
on the graph up in the left corner. The number of runner blades is chosen by typing in the
number up in the right corner, see figure 4.4. A radial view of the blade is also shown and
it is possible to change what the figures on the right side are showing by choosing on the
drop menu Colormap an blades.

In this case the number of blades was set to 17, the same as the Tokke prototype runner.
The blade leaning was set to zero degrees to be able to have through bolts at the leading
edge.

Figure 4.4: View of the Radial View tab in Khoj

29
Chapter 4. Methode

Blade Thickness
In the graphical view of the Blade Thickness tab the blade thickness at leading edge and
trailing edge are set, and the shape of the leading edge geometry are chosen.

Figure 4.5: View of the Blade Thickness tab in Khoj

Summary
In the last tab Summary an table with overview of all the velocity components, character-
istic parameters and the turbine dimensions are shown, see figure 4.6.

Figure 4.6: View of the Summary tab in Khoj

In the GUI in Khoj there is also a tabs for the Labyrinths, the Guide Vanes and one for the
Runner Cascade. In this case nothing were changed in these tabs because the guide vane
profile from Khoj are not the same type as the one on the Francis test rig at the Waterpower
Laboratory. The only thing used from Khoj is the shape of the runner blade.

30
 4.1 How to design a Francis runner to be used in the laboratory

4.1.2 Drawing of a Francis runner vane in Creo Parametric

In this section it will be described how to draw a 3D-model of a Francis runner vane in the
CAD program Creo Parametric (Pro/Engineer). The focus of this section will be how to
draw a runner vane ("Slice") like the one described in the section Design 4.

1. Import the following curve files from Khoj; bladess.ibl and bladeps.ibl. This files
contains a set of streamlines for the pressure- and the suction side of the runner vane.
See figure 4.7a

(a) Streamlines imported from Khoj (b) Boundary Blend between the streamlines

Figure 4.7: Making a quilt using streamlines from Khoj and Boundary Blend

2. If the streamlines is not oriented the way that is desired, a new coordinate system
can be made. This is done by using the function Coordinate System, then select the
existing coordinate system as References just by click on it in the left menu. Now
the new Coordinate System can be oriented by selecting the Orientation tab and then
rotate it about one of the coordinate axis. In this thesis the runner vanes is drawn in
a Coordinate System that is rotated 90 about the X-axis.
3. Use Boundary Blend between the curves from bladesps.ibl to make a quilt at the
pressure side see figure 4.7b, repeat for the suction side. (If the thickness of the
vane is wants to be changed, it can be done by Offset the two quilts).
4. Select the leading edge on the suction side and use the function Extend and remove
3 r. Repeat the procedure on the pressure side and remove 1 r.
5. Create the leading edge by using Boundary Blend between the pressure- and the
suction side. Select Constrains and set Conditions to Tangent and set the stretch
Value to 3 for suction side and 1 for the pressure side see figure 4.8a.
6. Use Boundary Blend the edges of the pressure- and suction side at the trailing edge.
7. Select the suction side edge and use Extend to remove 23 of the thickness towards
the pressure side. Then select the quilt edge on the suction side and use Extend to
remove 3sin30
2
thickness. Then use Boundary Blend to close the gap.

31
Chapter 4. Methode

(b) Merging hub and ring with the runner

(a) Creating the leading edge vane

Figure 4.8: Bondary Blend at leading edge and merging hub and ring

8. Use the function Merge to merge all the 5 quilts together.

9. Close the vane by make some Boundary Blend between the pressure- and the suction
side at the top and bottom of the vane. The use the Merge function and merge all the
quilts together.
10. Select the last Merge in the list to the left and use the function Solidify to solidify
the vane.
11. Start a Sketch in the Front Plane and draw the left side of the axial view of the runner
hub and ring. Make sure to extend the sketch so it cuts off the top of the vane. Exit
the sketch and use Revolve around the center axis and select the tab Revolve as
surface.
12. Select the hub and ring surfaces and use Merge and select Options tab and use Inter-
sect. This way the surfaces cuting the hub and ring will close the vane and the rest
of the hub and ring surfaces will be cut away. See the area inside the orange lines in
figure 4.8b

(a) Drawing of the hub section (b) Drawing of the shroud section

Figure 4.9: Drawing of the hub and shroud sections

32
 4.1 How to design a Francis runner to be used in the laboratory

13. Select the last Merge and use Solidify to make the runner vane a solid part.
14. Then start a Sketch in the Front Plane and draw the hub section of the runner vanes,
see figure 4.9a. When the sketch is finish use Revolve, and revolve the section around
the center axis.

15. To make the ring section start a Sketch in the Left Plane and draw the section, see
figure 4.9b. Then repeat the same procedure as for the hub section.
16. Make a Plane and Offset it 300mm up from the Top Plane. Start a sketch in the
new Plane and use Project on the outer edge of the hub and the edges on hub side
of the runner vane. Then use Construction Mode and make a Line along the center
of the vane. Then use the function Rotate Resize and rotate the centerline 360 32 in
each direction. Connect these two lines with the outer edge of the hub by using Line
and make a tangent from the line edges to the hub line. Close the lines like figure
4.10a and exit the Sketch. Select Extrude and use the tab Remove Material. Select
Extrude to intersect with selected surface and select the inner surface of the hub.
This section is now a "slice" that is 17
1
of the hub, see figure 4.10b. Make new Plane
300mm Offset from the Top Plane in the opposite direction and repeat the procedure
for the ring side.

(a) Section cut through the hub (b) Finish "Slice" on the hub side

Figure 4.10: How to make a "slice" of the hub

17. Now the runner vane ("slice") is completed. Then make the holes for attachment,
the holes are made in the center of each ledges. The three holes on the top of the
vane for Ensats insert have a diameter from 7.3-7.5mm, see data sheet for Ensats
figure E.1 in appendix E. At the trailing edge side of the ring sectin ther is a hole for
DIN 7 pin 4mm, and ther three other holes on ring side er the same at for the hub
section. At the leading edge there is a 5mm hole through for the M5 threaded rod.

33
Chapter 4. Methode

4.1.3 Drawing hub and ring in Creo Parametric

Hub
1. Start a Sketch in the Front Plane and draw the hub as shown in the Tokke model
drawing, see figure D.1 in Appendix D. Then to make the runner vanes fit, the hub
section of the vanes needed to be removed from the hub, see figure 4.11a. Exit the
sketch and use Revolve around the center axis.

(b) Sketch of the shaft keyway on

(a) Axial view of hub sketch the hub

Figure 4.11: Drawings of hub in Creo

2. The next is make the keyway by starting a Sketch on the top of the hub and draw the
keyway, see figure 4.11b. Then exit the sketch and Extrude a cut all the way trough
the hub.
3. The holes for attachment are made by using the same distances from the center axis
that was used for the runner vanes. The holes for bolts are made with countersink
for the bolt head. When the holes for one vane are made, Geometry Pattern are used
and an pattern of 17 for all the vanes are made 360 around the center axis.

(a) View of hub seen from above (b) View of hub seen from below

Figure 4.12: Finish hub from Design 4

34
 4.1 How to design a Francis runner to be used in the laboratory

Ring

Figure 4.13: Axial view of the ring sketch in Creo

1. The ring are made the same way as the hub, the ring section of the runner vanes
needed to be removed from the ring, see figure 4.13. All the rest are made the same
way as the hub.

(a) View of ring seen from above (b) View of the ring seen from below

Figure 4.14: Finish ring from Design 4

35
Chapter 4. Methode

36
Chapter 5
Experiment

5.1 Material tests

To find out if the polyamide (PA) material DuraForm HST-SLS from Arrk Europe Ltd. has
good enough strength to hold the runner hub and ring together with Ensat threaded inserts,
pins and M5 bolts, some test need to be done. Two test pieces were drawn in Creo and
then ordered from Protolab As to test the strength of the material. The two pieces were
produced by SLS printing and one of them was sprayed with some coating to show how a
smoother surface could be achieved, the test pieces are shown in figure 5.1.

(a) The test piece for pull-out (b) The test piece for the compres-
strength and torque testing sion test

Figure 5.1: Model of the test pieces ordered

37
Chapter 5. Experiment

5.1.1 Test 1: Pull-out strength test

The Ensat S threaded inserts were mounted into the cube test piece by using a homemade
installation tool. The tool was a M5 bolt, where the head was cut off, then two nuts and
a washer was screwed on and the Ensat S. The tool was mounted in a drill press, and the
inserts were screwed into the test piece by turning the chucks with one hand and gently
push down the drill press with the other hand.

To be able to test the SLS material piece with the Kerb Konus Ensat S self-tapping threaded
inserts a test rig was made. The test rig was made of a square beam that was approximate
550-650mm long. At the end of the beam a square plate was welded to hold the test piece,
and in the center of the plate a 18mm hole was drilled. At the opposite end a threaded rod
was bent into a hoop and welded to the beam to be able to hang it up in a pulley. On one
side of the beam a square hole was cut to be able to place the test piece into the rig. The
test rig can be seen in figure 5.2.

Figure 5.2: The test rig used for pull-out strength testing.

To be able to load weight on the test rig was hung up in a pulley, and a M8 nut was welded
to a M5 bolt, then a M8 eyebolt was screwed into the nut and fasten to the test piece. In the
eyebolt a shackle was attached and the weight "disc" for the generator torque calibration
was used to load weights, see figure 5.2.

38
 5.1 Material tests

5.1.2 Test 2: Compression test

In order to find out if a threaded rod with threaded inserts in the hub and ring section of
the runner vane, are enough to keep a constant inlet height, a compression test need to
be done. The test can be done by using the test piece seen in figure 5.1b, one threaded
insert at each end and a threaded rod throug. Then measure the distance between the two
inserts and put the piece into a hydraulic press. The load added need to be measured and
the distance between the inserts. The objective is to find out how much load in can handle
before the distance is changed.

5.1.3 Test 3: Torque test

A pin would be used to attach the runner vanes to the ring and need to be tested. The cube
test piece used in the pull-out strength test, was also design to be used for the torque test.
At one side of the cube there was made two holes that are 4mm in diameter, and are made
to fit a 4mm pin. To test the torque a pin is inserted in one of the two holes and a Ensat S
is mounted in the 7.4mm center hole, see figure 5.3. The next is to use a steel plate with
one 4mm hole for the pin and one hole for a bolt. The test piece is placed on the plate and
the bolt is screw into the Ensat S and tightens to the plate. Then the plate is attached to
something heavy that not will move and the test piece are exposed to a torque until the pin
or the plastic breaks. The force needed before it breaks is then measured.

Figure 5.3: The figure are showing a 4mm pin attached to the test piece and with the Ensat S
inserted.

39
Chapter 5. Experiment

5.2 Results from material tests

Due to lack of time and not having all the equipment needed, only the pull-out strength
test was done.

When testing the pull-out strength it started with the thread inserts in 7.5mm holes, then
the 7.4mm holes. For the two 7.5mm holes and the first of the 7.4mm holes, 160 kg was
loaded onto the disc before the Ensat S inserts were pulled out. For the next 7.4mm hole
all the calibration weights was used (about 230kg) and it was still not enough to pull it out.
To gain more weight scrap metal were weighed and used as weights, after loading all the
scrap metal that it was room for, the total weight was up in about 330kg, see figure 5.4a.

(a) Loading with scrap metal to (b) Using a crane weight to meas-
gain weight sure the load

Figure 5.4: Pull-out strength testing

After all the metal scrap was used, a new method had to be used. Then a crane weight
was borrowed on the Thermal Energy workshop, the crane weight had a limit of 500kg.
The limit was reached without problem. After that there was not found any measuring
equipment with a higher limit, therefore the maximum pull-out force was not found. The
two 7.3mm holes was either not tested, and they would probably withstand even more than
the 7.4mm hole.

40
Chapter 6
Discussion

6.1 The new runner and runner vane designs

The first alternative considered was the Francis Design 1 where the existing Tokke model
runner hub and ring are being used, and replacing the runner vanes with someone made of
plastic. This solution is a very good alternative when it comes to testing if the strength and
stiffness in the material is good enough, but this gives small or almost no opportunities to
change the design of the runner vanes. There is no trough bolts at the inlet and the plastic
vanes are the only structure keeping the distance between the hub and ring. If the stiffness
and strength in the plastic are not sufficient to keep a constant distance between the hub
and ring the lower labyrinth sealing could touch each other and lose a lot of energy due to
friction. This could also lead to leakage in the labyrinth sealing.

The second alternative with new hub and ring in bronze and runner vanes in plastic. This
alternative has a trough bolt at the inlet, glue along the blades and pins to hold the parts
together and lock it in place. The third alternative is almost the same as the second one,
only with a groove around hub and ring, and a section of the hub and ring on the runner
vanes. This sections will prevent slippage and give better support for the vanes since the
vanes are supporting each other. Both design 2 and 3 may become unstable due to poor
attachment, with only proper mounting at the leading edge, this would probably not be
enough keep the runner firmly in place.

The last alternative considered is the Design 4 where the runner vanes have a section of the
hub and ring. This alternative with the large sections is having a lot of possibilities when it
comes to fixing. With all the fixings, it should be enough to keep the runner in place. The
disadvantages with this are that the volume of the vanes is quiet large, and that increases
the production cost, espcially with SLS printing.

41
Chapter 6. Discussion

With the design 4 the hub and ring could be used for different design just by replacing the
slices. It may be necessary to replace some of the holes to attach all the parts together.
This way a new model runner with different vanes could be made only by changing the
design of the vanes and make it fit the hub and ring.

The blade shape in design where the U Cu energy distribution was set to be almost linear
from inlet to outlet. This was made provoke some vibrations in the runner. The relative
velocity, W , at the hub side was going up and down through the runner, and close to the
trailing edge the velocity increases up 1.4 times the inlet velocity. On the center of the
blade the relative velocity is almost constant the first 34 of the blade length and the 14 of
the blade the velocity increases to over 2 times the inlet velocity. At the shroud side of the
blade the velocity is constant the first half og the length and increases to almost 3 times
the inlet velocity over the second half of the blade. For a blade that is designed good, the
Cu component should never be larger than the U component, at the hub side of this design
the Cu component is larger than U along the whole blade.

6.2 The new guide vane and stay vane designs

The firs guide vane alternative was the one with a "disc" and pins and where the guide
vane shaft and tap are glued to the guide vane blades. With this alternative all the forces
and torque would be absorbed in the transition between shaft and the blade, and this is the
weakest point.

The second guide vane with connection between shaft and blade almost like a keyhole. In
this design the blade in would absorb all the forces and bending moment from the water
pressure and the transition between shaft and blade only need to withstand the torque.

The during the Tokke model test at the Waterpower laboratory most of the tests was run
with head, H = 30 m, and rotational speed, n = 540 RP M . According to the test
results the maximum torque on the guide vanes was at approximate 1 opening and then
the torque was just over 4 N m, see figure D.3 in appendix D. According to some simple
calculations done, the torque in closed position is about 15Nm and the bending stress was
calculated to about 38MP. Due to this the new guide vane design at least need to withstand
15Nm torque. The calculations could be seen in Appendix A.

The new stay vane design where plastic vanes and trough bolts are replacing the aluminum
vanes welded to the spiral casing would probably not be economically profitable. That is
because the shape of the stay vanes are not that complex and there are probably small
differences from making the vanes of plastic by rapid prototyping, or standard material
and production method.

42
 6.3 Manufacturing methods and materials

6.3 Manufacturing methods and materials

The two rapid prototyping methods considered for manufacturing the parts for the new
designs are selective laser sintering (SLS) and rapid CNC machining. Both methods have
advantages and disadvantages.

Using SLS gives possibilities for more complex designs and with internal structures, such
as channels for wiring and grooves for sensors. One of the disadvantages with SLS is that
tolerance is not that good. When ordering SLS printed parts from Protolab AS and Arrk
Europe there are four different plastic materials to choose between, see figure 2.9.

When using rapid CNC machining the quantity has a grate influence on the manufacturing
price. If there is a production series of only a couple of parts the rapid CNC machining
can not compete with SLS printing when it comes to price., but when there are larger se-
ries and large volume the CNC may be the cheapest method [6]. CNC machining from
Arrk Europe gives a lot of different materials to choose from, there are over 35 standard
materials and if thats is not enough other materials can be ordered [11].

Most of plastic materials used in rapid prototype machines could absorb a small amount
of water, even if it are just 0.05% that could be enough to change the distance between
the hub and ring with a tenths of a millimeter. The tolerance is so fine that this could
be enough to make the labyrinth sealing and runner to touch and loose efficiency due to
friction, and therefore is it important to have a design where the vanes not are able to move.

Figure 6.1: Showing pull-out strength for Kerb Konus Ensat with different thread diameter in dif-
ferent plastic materials [9].

43
Chapter 6. Discussion

When testing the SLS material and threaded inserts for pull-out force, there was some
quiet large differences in the force for the same hole diameter. That could be due to not
a perfect mounting of the inserts or because of roughness in the holes. According to the
manufacturer the threaded insert Ensat with M5 internal threads should hold a pull-out
force of approximate 10kN in PA plastic, see figure 6.1. That may be correct, because the
Ensat in the 7.4mm was loaded with 500kg without giving any sign of giving in.

6.4 Production of runner blades and model test

In the start of this semester the plan was to produce new runner blades for a Francis model
turbine and run model tests if they were available before end of the semester. Everything
did not go as planned and due to lack of time and money the runner vanes was not manu-
factured. The was no available time in the laboratory before the end of May, and therefore
it would not have been time to run model tests even if the runner vanes was manufactured.

6.5 FEM analysis of the new runner design

At the beginning it was planed to do an Finite Element Method (FEM) analysis of the new
runner vane and guide vane designs to compare the results with empirical data and some
simplified calculations done on the runner vanes and guide vanes. This was not done since
there was not sufficient time.

44
Chapter 7
Conclusion

7.1 Conclusion
It is possible to reduce the production time and cost of production of Francis model tur-
bines, by combining new and old production methods.

Comparing the four different designs of the Francis runner vanes, the Design 4 is the best
alternative both when it comes to strength and possibilities to change the design. The quiet
large section of each vane gives better fixing opportunities and is more stable the other.

The guide vane Design 2 are the best alternative because in this design the vane with
the "discs" is one solid part, and are absorbing all the bending moment from the water
pressure. The connections between the vane, the guide vane shaft and tap only need to
withstand the torque. That should not be any problem, but the connections could be rein-
forced with some steel bushings.

To change the stay vane design, and replace the vanes with someone made of plastic with
trough bolt would probably not be a very good solution, that is because the production cost
probably will be higher than for a regular method. It is also a risk of leakage under and
over the vanes.

When it comes to manufacturing methods Selective laser sintering (SLS) would be the best
method for producing the runner vanes, especially when the geometries are complex, that
is because the support from the unsintered powder surrounding the part during build. If
the runner vanes are designed with channels for wiring inside the blades and grooves for
sensor, SLS printing is the only method. All parts with requirement to tolerance produced
by SLS printing, need to be coated to get a smooth surface finish. The CNC machining
are giving a smoother surface finish, so if the production cost is lower for CNC than SLS,
then CNC should be used.

45
Chapter 7. Conclusion

The runner vanes are produce with SLS printing or CNC and guide vanes and stay vanes
are manufactured by CNC machining as long as they are not designed with slots for sen-
sors or channels for wiring, then they are also SLS printed.

The hub and ring will be produced with standard CNC milling by a Norwegian mechanical
workshop. These parts should be produced in bronze-JM7, the same material as the Tokke
model runner hub and ring are made of.

If the Kerb Konus self-tapping threaded inserts are used to attach the runner vanes to the
hub and ring, they must be properly installed. If they are not properly installed, they may
not withstand the pull-out force needed, and therefore proper installation tools should be
used.

46
Chapter 8
Further Work

8.1 Further Work

As further work a FEM analysis of the new runner vane and guide vane designs could be
done to control if they will withstand the forces from the water pressure in the Waterpower
Laboratory Francis rig. The analysis should control for both the highest torque in open
position and the highest force acting on the guide vanes when they are in closed position.

Continue the runner design process and do the changes needed before the runner vanes are
redy to be manufactured. Manufacture the runner vanes, new hub and ring. Assemble the
parts and run some model test in the rig at the Waterpower laboratory at NTNU.

Another possibility for further work could be to design a runner vane for testing with strain
gage. It could be designed with a channel for wiring inside the blade and up through the
hub. The vane could be designed with a groove for the strain gage so it not disrupts the
flow. Just to test this method a runner vane for existing Tokke model runner could be made
in plastic with channel for the wiring and replace one of the bronze vane.

As a start just to test the guide vane design, just a couple of existing guide vanes could be
replaced with the new design. To control the new guide vanes a channel for wiring and a
groove for strain gage could be added in the 3D-model before production.

Use the two test pieces and do the torque test and the compression test, described in the
Experiment chapter. This is to find out if the fixing and material are strong enough.

As further work different plastic material could be analysed to see if there is some ma-
terials that are scalable compare to the different stainless steel grades used in prototype
runners.

47
Chapter 8. Further Work

48
Bibliography

[1] Tokk model turbine test report, 2007.

[2] Hermod Brekke. Pumper & turbiner. Waterpower Laboratory, NTNU, 2003.
[3] Hermod Brekke. Konstruksjon av pumper og turbiner. Waterpower Laboratory,
NTNU, 2008.
[4] Ole Gunnar Dahlhaug. Personal conversation, Spring 2014.
[5] Mette Eltvik, Pl Henrik Finstad, Grunde Olimstad, Eve Walseth, and Zhao Wei.
High pressure hydraulic machinery. Technical report, Waterpower Laboratory,
NTNU, Autum 2009.
[6] Anders Flaten. Conversation, Spring 2014.
[7] A. Gebhardt. Understanding Additive Manufacturing: Rapid Prototyping, Rapid
Tooling, Rapid Manufacturing. Carl Hanser Verlag, 2011.

[8] Kristine Gjster. Hydraulic design of francis turbine exposed to sediment erosion.
Masters thesis, NTNU, Spring 2011.
[9] Kerb Konus GmbH. Kerb konuss webpage. http://www.kerbkonus.de/,
2014.
[10] Peter Joachim Gogstad. Hydraulic design of francis turbine exposed to sediment
erosion. Masters thesis, NTNU, Fall 2011.
[11] Arrk Europe Ltd. Arrk europes webpage. http://www.arrkeurope.com/,
2014.
[12] Torbjrn K. Nielsen. TEP4195 Turbomachinery, Lecture, Spring 2013.

[13] Inc. PTC. Ptcs webpage. http://www.ptc.com/product/creo//, 2014.

49
50
Appendix A
Appendix A

A.1 Guide vane calculations

In this appendix some simple calculations on the guide vanes and runner vanes will be
shown.

Data from Tokke model turbine test

Head H = 30 [m]
Rotationalspeed n = 540 [RP M ]
Number of guide vanes Z0 = 28 [ ]
Inlet diameter D1 = 3.17 [m]

Table A.1: Data from Tokke model test used in calculations

i
A.1.1 Hydraulic force on guide vane in closed position
Some of the values used in the calculations are taken from the 3D model in Creo; B0 =
0.05938[m], a0 = 0.01015[m], l = 0.07921[m], D0 = 0.70595[m], d = 0.014/0.020, .
The total force on each vane:

gHB0 D0 1000 9.81 30 0.05938 0.70596

F0 = gHB0 l = = = 1384.2 [N ]
z0 28
(A.1)
The torque, T0 , on one guide vane tap is then:

T0 = F0 a0 = 1384 0.01015 = 14.05 [N m] (A.2)

The polar moment of inertia, IP for the guide vane shaft:

d4 0.0144
IP = = = 3.77 10 9 [m4 ] (A.3)
32 32
The maximum shear stress, max , due to torque is then found:
T0 14.05 0.007
max = max = r = = 26.08 [M P a] (A.4)
IP 3.77 10 9
Bending moment:

g H0 B 0 2 D 0 1000 9.81 30 0.059382 0.70596

M= = = 10.3 [N m]
8 z0 8 28
(A.5)
Moment of inertia:

d4 0.0144 9
IP = = = 1.89 10 [m4 ] (A.6)
64 64
Bending stress:
M r 10.3 0.007
b = = = 38.15 [M P a] (A.7)
IX 1.89 10 9
The lages pricipal stress is then calculated by:

r r
2 38.15 38.152
b b
max = + 2 = + + 26.082 = 51.4 [M P a] (A.8)
2 4 2 4

ii
Appendix B
Appendix B

B.1 Price offer from Protolab AS

In this appendix there is a review of the different price offer received from Protolab AS. T

B.1.1 Price on

iii
 Figure B.1: Price offer for a full set of runner vanes to the Tokke model runner

iv
Figure B.2: Price offer for a couple of test vanes to the Tokke model runner

v
 Figure B.3: Price offer for test piece in SLS material DuraForm HST

vi
Figure B.4: Price offer for test pice in SLS material DuraForm PA

vii
viii
Appendix C
Appendix C

C.1 Material properties

Protolab AS use ARRK Europe LTD as their manufacturer of prototypes. Here is a
overview of the material properties for the material from different

Figure C.1: Material range at ARRK for SLS printing

ix
 Rapid CNC Machining
ARRK are able to oer customers a
wide range of CNC machining op ons
from our new rapid CNC centre at
Gloucester, plus larger volumes
manufactured out of ARRK facili es
in the Far East.

Rapid CNC machining -

available here in the UK as well
as in the Far East
If you are looking for a rapid turnaround of
machined parts within a matter of days our new
UK Rapid CNC Cell in Gloucester could be the
perfect solution for you.

With a range of Mazak and DMG machines,

offering 5 and 3 axis capability, our new centre
has been set up to satisfy our customers
increasing demand for speedy delivery. Whether
you need just a single component or a number of
machined parts we can help.

CNC machined components - through to

Block Modelling, Texturing and Finishing

Alternatively, where customers are looking for

more complex machining or post machining
options, our facilities in the Far East provide
even greater CNC capacity and services. The
ARRK team in the Far East has built up an
impressive reputation for taking complex,
intricate projects and making them a reality.

prototyping
 ARRK are able to cut from a wide range of plas cs and metals, from
standard materials to the more unusual.
Out standard range includes:
Material Trade Name

ABS Tecaran
Acetal Tecaran
Acrylic Tecaform AH
Cast Nylon Tecast
Delrin Tecanyl
Nylon 6 Tecamid 6
Nylon 66 Tecamind 66
Nylon 12 Tecamid 12
Peek Tecapeek
Petg Petg
Polycarbonate Tecanat
Polyester Tecapet
Polyethersulpone Tecason E
Polyphenylenesulphone Tecason P
Polyethylene Tecane PE
Uhmw PE Tecane PE10
Polyimide Tecasint
Polypropylene Tecane PP
Polysulphone Tecason S
PPS Tecastron
PTFE Tecason PTFE
PVC Tecavinyl
PVDF Tecaon PVDF
Tufnol Tufnol
Ultem Tecapei
AL/AL Plate 6082
AL/AL Plate 5083
AL/AL Plate 7075
Brass Plate
Brass Bar
Copper Plate
Why use ARRKs CNC Service
Copper Bar
Zinc Plate
Mild Steel Plate Speed of delivery
Stainless Steel Plate 304/316 Sheared Huge capacity
Stainless Steel Plate 304/316 Plasma We can start your project quickly
Stainless Steel Plate 304/316 Sawn
Great part quality
Wide range of material options
Vast experience
Models that are simply stunning
Highly competitive service and rates

For more information, please contact:

Gloucester Technical Centre. Olympus Park,
Quedgeley, Gloucester. GL2 4NF, UK.

Tel: +44(0)1452 727700

Fax: +44 (0)1452 727755 www.arrkeurope.com
E-mail: [email protected]
xii
Appendix D
Appendix D

D.1 Drawings and documents from Tokke model turbine

In this appendix the drawings of Tokke model runner, guide vanes, stay vanes and stay
ring used to make the new designs fit the Francis model rig at the Waterpower laboratory
at NTNU.

xiii
 Figure D.1: Axial view of the Tokke model runner

xiv
Figure D.2: Drawing of the guide vane

xv
 Figure D.3: Guide vane torque for VKL-runner [1]

xvi
Figure D.4: Drawing of the stay vane shapes

xvii
 Figure D.5: Drawing of the stay ring

xviii
Appendix E
Appendix E

E.1 Product informastion sheets for the fixing parts

In this appendix the product information sheets for the parts used to fasten the runner vanes
to the hub and shroud are shown.

xix
 Figure E.1: Product info about Kerb Konus threaded inserts Ensat S

xx
Figure E.2: Product info about Kerb Konus threaded inserts Ensat SK

xxi
 Figure E.3: Product info about Kerb Konus threaded inserts Ensat SI/SBI

xxii
 GOUPILLES CYLINDRIQUES
DOWEL PINS

GOUPILLES CYLINDRIQUES
ISO2338 - NFE27484
DOWEL PINS ISO2338 NFE27484

Dsignation : Goupilles cylindriques ISO2338 [forme] D x L

Description: Dowel pins ISO2338 DxL
Forme A : D = tolrance m6 / Forme B : D = tolrance h8 /
Forme C : D = tolrance h11
A shape: D = tolerance m6 / B shape: D = tolerance h8 / C shape: D = tolerance h11

ATTENTION : Bien choisir la forme des extrmits et la tolrance

Please note: choose the right shape and tolerance for your needs

Matire = 9SMnPb28K et inox

Material: 9SMnPb28K and stainless steel

D L

1 3 4 6 8 10 12 14 16 18 20
1,5 3 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32
2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 36 40
2,5 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 36 40
3 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 36 40 45 50
4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 36 40 45 50 55 60
5 6 8 10 12 14 16 18 20 22 24 26 28 30 32 36 40 45 50 55 60
6 6 8 10 12 14 16 18 20 22 24 26 28 30 32 36 40 45 50 55 60
8 10 12 14 16 18 20 22 24 26 28 30 32 36 40 45 50 55 60 70 80
10 12 14 16 18 20 22 24 26 28 30 32 36 40 45 50 55 60 70 80 90 100
12 12 14 16 18 20 22 24 26 28 30 32 36 40 45 50 55 60 70 80 90 100 120
Botes assortiments goupilles voir p. 25
Boxed sets of pins, see p. 25

SUR DEMANDE 8
Dimensions et tolrances spciales

ON REQUEST Other tolerances and dimensions
8
Matires spciales notamment en inox
Special materials, particularly stainless steel

16
Gardette Industrie S.A - Tl : +33 (0)4 74 02 94 94 - Fax : +33 (0)4 74 65 35 05 / +33 (0)4 74 07 05 87
Mail : [email protected] - www.gardette.fr

Figure E.4: Product information about the DIN 7 pins

xxiii