Welding Procedures Specification for FCAW
of Wind Towers
Nelson da Cunha de Matos
Dissertation for the Degree of Master in
Mechanical Engineering
Jury
Chairperson: Prof. Doutor Rui Manuel dos Santos Oliveira Baptista
Supervisor: Prof.ª Doutora Maria Luísa Coutinho Gomes de Almeida Quintino
Co-Supervisor: Eng.º Eduardo Manuel Dias Lopes
Member: Prof.ª Doutora Rosa Maria Mendes Miranda
October 2012
Acknowledgements
The conclusion of this thesis was only possible with the support and guidance of Professor Luísa
Coutinho and Engº Dias Lopes which with all their patience pointed me in the right direction and
shared invaluable knowledge not only for this thesis but for the future also.
Additionally I must congratulate Instituto de Soldadura e Qualidade and its excellent team.
I’m very grateful to Engº Vitor Ferreira for all the effort put into this project, always available to
discuss and explain anything and for sharing the massive volume of welding knowledge.
I would also like to thank welding technicians Carlos Sanches, António Costa and Matias Torrão
for providing a unique field experience, enlightenment about the many and diverse welding
techniques, welding procedures and for sharing a very particular welding vocabulary and popular
sayings.
Liliana Silva from NDT department did an exceptionable job not only by the meticulous analysis
but also by the sheer volume of test that had to be done in a short amount of time and still found time
to explain any clarify doubts that I had.
To all my friends a very special word of appreciation for helping me when moral was low and for
pulling out from me a laugh no matter the situation. A special thank you to my friend Gonçalo Silva
who helped every time possible with his NDT expertise.
At last but not least I would like to thank my family for all the support and patience when things
did not go as they should.
II
Resumo
O objectivo principal deste estudo consiste no desenvolvimento de procedimentos de soldadura
utilizando a tecnologia de fio fluxados de forma a permitir um novo conceito de design e produção de
torres eólicas de grandes dimensões com diâmetros base maiores, levando em consideração as
necessidades dos locais de construção.
Foi desenvolvido um novo conceito de torre eólica permitindo a construção de torres de aço
tubular para conversores de energia eólica em áreas remotas com acesso limitado. A inovação deste
conceito é a substituição do anel de flanges e ligações aparafusadas por soldadura feita no local de
erecção com vista a facilitar o transporte e melhorar a resistência à fadiga permitindo a introdução
aços de maior resistência.
Foram realizados ensaios de soldadura e qualificação de Procedimentos de Soldadura com fios
fluxados, onde uma grande variedade de fios auto protegidos e com protecção gasosa foram testados
e avaliados quanto à sua soldabilidade global.
As soldaduras realizadas usando fios com gás de protecção demonstraram como sendo muito
mais adequado para aplicações de soldadura mecanizada que seus homólogos auto protegidos.
Cobre juntas cerâmicas serão utilizadas, sempre que possível, para a soldadura de passes de
raiz, o que permite soldar apenas de um lado no intervalo de menor espessura em particular na
posição vertical para cima sendo esta a mais relevante para a montagem da torre.
As peças soldadas nos aços S355J2 e S460M foram sujeitas a testes mecânicos e análises
microestruturais de forma a verificar a sua adequabilidade em relação aos esforços esperados para
uma torre eólica.
Palavras-chave
Soldadura com fios fluxados
Fissuração a frio
Protecção gasosa
Auto protegido
Aço S460M
Aço S355J2+N
Torre Eólica
III
Abstract
The main objective of the study concerns the development of a welding procedure using flux
cored wire in order to allow a new design concept for large wind energy towers with increased base
diameters taking into consideration requirements of onsite conditions.
A new tower concept was developed that allows for the erection of high tubular steel towers for
wind energy converters in remote areas with limited accessibility. The innovation of this concept is the
replacement of ring flanges by onsite welds which improves the fatigue resistance and therefore
enables the introduction of higher steel grades
Welding trials and preparation of Welding Procedure Specifications Flux Cored Arc Welding
(FCAW) tests were carried out, a wide selection of self-shielded and gas-shielded wires were tested
and evaluated for their overall weldability. Gas-shielded filler wires proved themselves as being far
more suitable for mechanized welding applications than their self-shielded counterparts.
Ceramic backings will be used wherever possible for the welding of root passes, allowing for
single-side welding in the lower thickness range particularly in the vertical-up position, being the most
relevant for FCAW during the erection of the tower.
The welded samples in S355J2 and S460M steels were subjected to mechanical tests and microstructural analysis in order to ascertain their suitability for the efforts expected in a wind tower.
Keywords
Fluxed Cored Arc Welding
Hydrogen Cracking
Gas-Shielded
Self- Shielded
S460M Steel
S355J2 Steel
Wind Energy Converter
IV
Contents
Acknowledgements ............................................................................................................................... II
Resumo ................................................................................................................................................ III
Palavras-Chave .................................................................................................................................... III
Abstract .................................................................................................................................................IV
Key-Words .............................................................................................................................................IV
Abbreviations ..........................................................................................................................................X
1 Objectives and Motivation ................................................................................................................ 1
2 Thesis Structure ............................................................................................................................. 1
3 State of Art ..................................................................................................................................... 2
3.1 Energy transformation processes and its use............................................................................2
3.2 Renewable energy Technologies ..............................................................................................3
3.2.1 First generation technologies ..............................................................................................3
3.2.2 Second generation technologies .........................................................................................3
3.2.3 Third Generation Technologies ...........................................................................................4
3.3 Wind Energy ............................................................................................................................5
3.3.1 Wind energy transformation ................................................................................................6
3.3.2 Wind energy converters description ....................................................................................7
3.3.2.1 Blades .........................................................................................................................8
3.3.2.2 Nacelle ........................................................................................................................9
3.3.2.3 Towers ........................................................................................................................9
3.3.3 Location ........................................................................................................................... 11
3.3.4 Onsite Build ...................................................................................................................... 12
3.3.4.1 Site preparation ......................................................................................................... 12
3.3.4.2 Foundation construction ............................................................................................. 12
3.3.4.3 Transport ................................................................................................................... 13
3.3.4.4 Erection ..................................................................................................................... 14
3.3.4.5 Complementary constructions .................................................................................... 15
3.3.5 Growth trend..................................................................................................................... 16
3.4 Wind Energy Converter tower design...................................................................................... 17
3.4.1 Tower Weight ................................................................................................................... 18
3.4.2 Plate thickness ................................................................................................................. 18
3.5 Welding technology ................................................................................................................ 19
3.5.1 Welding Technology - Fluxed Core Arc Welding (FCAW) .................................................. 19
3.6 Innovations and Benefits ........................................................................................................ 22
4 Experimental Procedure ................................................................................................................. 24
4.1 Equipment.............................................................................................................................. 24
4.2 Preliminary test of FCAW wires .............................................................................................. 25
V
4.2.1 Wire selection ................................................................................................................... 26
4.3 S460M Weldability test ........................................................................................................... 32
4.3.1 Pre Heating for HSS Steels............................................................................................... 34
4.3.2 Characterization S460M steel welded with different heat inputs......................................... 36
4.4 Mechanized Welding Trials..................................................................................................... 37
4.5 Mechanized Welding ST Samples .......................................................................................... 40
5 Results and analysis ...................................................................................................................... 41
5.1 Preliminary Specification of Welding Procedures .................................................................... 41
5.1.1 Manual Welding tests ....................................................................................................... 41
5.1.2 Welding defects encountered while optimizing and solutions ............................................. 43
5.1.3 Manual Welding Pre-Qualification and deposit rates. ........................................................ 46
5.1.4 Analysis of results............................................................................................................. 50
5.1.5 Selected Filler Wire........................................................................................................... 54
5.2 S460M Weldability Test .......................................................................................................... 54
5.2.1 Analysis of results............................................................................................................. 59
5.3 Mechanized Welding Trials..................................................................................................... 64
5.3.1 Imperfections detected in mechanized welding by NDT ..................................................... 66
5.3.2 Analysis of Results ........................................................................................................... 68
5.4 Mechanized Welding ST Samples .......................................................................................... 70
5.4.1 NDT imperfections detected in ST test pieces ................................................................... 70
5.4.2 Mechanical Testing Results .............................................................................................. 73
5.4.3 Analysis of Results ........................................................................................................... 74
6 Conclusion ..................................................................................................................................... 78
7 Future work .................................................................................................................................... 80
8 References .................................................................................................................................... 81
9 Annexes......................................................................................................................................... 84
9.1 Pre-heat calculation methods ................................................................................................. 84
9.2 Heat Input Trial Data .............................................................................................................. 86
9.3 Macrographs and Micrographs of Heat Input Trial Samples .................................................... 87
9.4 Hardness Values .................................................................................................................... 93
9.5 Filler Wire Specification .......................................................................................................... 95
9.6 S355J2 and S460M Steel properties ...................................................................................... 97
9.7 Steels certificate ..................................................................................................................... 98
9.8 Charpy test results ................................................................................................................. 98
VI
List of Figures
Figure 1 – Scheme of a conventional hydroelectric dam (left image), and conventional geothermal
system (right image) ........................................................................................................3
Figure 2 - Solar power plant in south Spain (right image), Serpa Solar Park, Portugal (left Image). ......4
Figure 3 - Pelamis power conversion module. .....................................................................................4
Figure 4 - Wave farm device generator, Portugal. ................................................................................4
Figure 5 - Off-Shore Wind Towers. ......................................................................................................5
Figure 6 - Portugal Electricity Generation by Source (2009) .................................................................5
Figure 7 - Portugal Electricity Generation by Source (September 2011) ...............................................6
Figure 8 - Daily Hydro-Wind Complementarity. ....................................................................................7
Figure 9 - Vertical Axis Wind tower. .....................................................................................................7
Figure 10 - Wind Turbine Configuration, Vertical and Horizontal type...................................................7
Figure 11 - Horizontal axis wind turbine breakdown. ............................................................................8
Figure 12 - Blade Cross Section. .........................................................................................................8
Figure 13 - Schematic of multistage gearbox. ......................................................................................9
Figure 14 - Dimension comparison of Wind Tower with a Boeing 747 ................................................ 10
Figure 15 - Left: Rolling, forming and tack welding of the shell. Right: External and internal longitudinal
submerged arc welding. Multi-wired SAW is widely used here. ...................................... 10
Figure 16 - Terminology and a rough component layout of the tower. ................................................ 11
Figure 17 - Image demonstrating size restriction for transport. ........................................................... 11
Figure 18 - Wind tower locations in main land Portugal and islands. (2010) ....................................... 12
Figure 19 - Wind tower foundation. .................................................................................................... 13
Figure 20 - Mountain road transport maneuvers. ............................................................................... 13
Figure 21 - Workers connecting tower section (left) and tower section with foundation (right). ............ 14
Figure 22 - Workers bolting flanges in tower section. ......................................................................... 14
Figure 23 - Cranes lifting blades and rotor. ........................................................................................ 15
Figure 24 - World Wind Power capacity growth from 1996 to 2010. ................................................... 16
Figure 25 - Prognosis of installed wind capacity until 2030. ............................................................... 16
Figure 26 - Assembly overview, circular segments and sectors assembled on site............................. 17
Figure 27 - Tower weight depending on base diameter ..................................................................... 18
Figure 28 - Maximum plate thicknesses depending on base diameter for unreinforced and reinforced
tower. ............................................................................................................................ 19
Figure 29 - Typical setup for semiautomatic FCAW equipment .......................................................... 21
Figure 30 - Experimental Procedure Diagram .................................................................................... 24
Figure 31 - LE Idealarc CV 400-I & LN 742........................................................................................ 24
Figure 32 - LE LN-25Pro & Invertec V350 Pro ................................................................................... 24
Figure 33 - Welding machine, Jigg and ADS for mechanized welding. ............................................... 25
Figure 34 - Ceramic Backing ............................................................................................................. 25
Figure 35 - Welding process variables. .............................................................................................. 28
Figure 36 - Burn-off curves. ............................................................................................................... 28
Figure 37 - Contact Tip to Work Distance. ......................................................................................... 28
Figure 38 - Welding voltage relation with welding amperage. ............................................................. 31
Figure 39 - Torch angle importance. .................................................................................................. 32
Figure 40 - Steel processing routes. .................................................................................................. 33
Figure 41 - Qualitative welding workspace for thick high strength structural steels. ............................ 34
Figure 42 - Calculated preheating. .................................................................................................... 36
Figure 43 - From left to right: HI test layout; Welding machine; Data acquisition system; Print screen of
acquisition; HI test piece; Test weld beads .................................................................... 37
Figure 44 - Fully mechanized welding trials with FCAW selected filler wire PZ6113S: welding of a 10mm thick plate butt joint in the vertical-up (PF) position. ................................................. 38
VII
Figure 45 - Phased array probes are made in a variety of shapes and sizes for different applications.
..................................................................................................................................... 39
Figure 46 - Inspection system during circular welds inspection process in a wind tower. .................... 39
Figure 47 - A test piece being welded during semi-automatic welding tests. ...................................... 41
Figure 48 - Test piece geometry ........................................................................................................ 42
Figure 49 - Weaving technique for PF position. ................................................................................. 42
Figure 50 - Example of trial and error with NR233. ............................................................................ 42
Figure 51 - NR233 Root morphology ................................................................................................. 43
Figure 52 - NR233 Penetration on root pass...................................................................................... 43
Figure 53 - Ropey convex bead......................................................................................................... 43
Figure 54 - Concave bead ................................................................................................................. 44
Figure 55 - Hot Cracking ................................................................................................................... 44
Figure 56 - Standard Front Weld Bead .............................................................................................. 45
Figure 57 - Worm tracking ................................................................................................................. 45
Figure 58 - Weld bead dimensions. ................................................................................................... 45
Figure 59 - Example of a semi-automatic welding joint using a self-shielded wire (NR-203Ni1): root
pass performed on open root, with no backing. Left: front side of welded joint; right: back
side of welded joint. ....................................................................................................... 46
Figure 60 - Example of a semi-automatic welding joint using a gas-shielded wire (PZ6113S): root pass
on performed on ceramic backing. Left: front side of welded joint; right: back side of
welded joint. .................................................................................................................. 46
Figure 61 - General view of test pieces produced throughout the course of manual semi-automatic
welding tests of each wire chosen. ................................................................................ 46
Figure 62 - Effect of Tip-To-Work distance, WFS and deposition rate on welding current. .................. 47
Figure 63 - Wires used in weighing.................................................................................................... 47
Figure 64 - Deposition rates for one-side butt welding, 12-mm thick plate, vertical-up (PF) position after
welding process optimization. ........................................................................................ 48
Figure 65 - Welding joints, sequences and parameters for vertical-up (PF) position. .......................... 48
Figure 66 - Welding joints, sequences and parameters for flat (PA) position. ..................................... 49
Figure 67 - Welding joints, sequences and parameters for horizontal (PC) position............................ 49
Figure 68 - Variation ratio between stickout and wire diameter. ......................................................... 52
Figure 69 - Parameters optimization summary................................................................................... 53
Figure 70 - Test specimens obtained................................................................................................. 55
Figure 71 - Weld diluition areas ......................................................................................................... 56
Figure 72 - Hardness Test Results .................................................................................................... 56
Figure 73 - Maximum Hardness for each heat input value tested ....................................................... 57
Figure 74 - 2D and 3D flow................................................................................................................ 58
Figure 75 - Steel welding temperature over time for different HI @ r=3mm. ....................................... 59
Figure 76 - HAZ width of each HI. ..................................................................................................... 59
Figure 77 - Max. hardness relation with cooling times. ....................................................................... 59
Figure 78 - C-Mn Steel Weld CCT Diagram of S460M. ...................................................................... 61
Figure 79 - Cap parameters tryout upward and downward angle ....................................................... 64
Figure 80 - Cap parameters tryout 90º angle ..................................................................................... 64
Figure 81 - Mechanized root pass welded on ceramic backing. Left: front side of welded joint; right:
back side of welded joint immediately after removal of backing. ..................................... 65
Figure 82 - Completed welding joint. Left: front side of welded joint; right: back side of welded joint. .. 65
Figure 83 - On site UT testing ........................................................................................................... 65
Figure 84 - Macrographs of two cross sections of perfect welded joint. .............................................. 66
Figure 85 - Shielding gas flow regions. .............................................................................................. 68
Figure 86 - Laminar and turbulent flow. ............................................................................................. 69
Figure 87 - Radiography of ST005, 0 to 165mm ................................................................................ 71
Figure 88 - Radiography of ST005, 170mm to 335mm....................................................................... 71
Figure 89 - Radiography of ST005, 335 to 500mm ............................................................................ 71
VIII
Figure 90 - Macrograph of ST005 weld bead showing a slag channel ................................................ 71
Figure 91 - ST005 removed defect sample and air carbon arc thinning .............................................. 72
Figure 92 - Difference between initial preparation (left) and correct preparation (right) ....................... 72
Figure 93 - ST008 Charpy test. ......................................................................................................... 73
Figure 94 - ST004&ST007 Charpy test. ............................................................................................. 73
Figure 95 - ST013&ST014 Charpy test. ............................................................................................. 74
Figure 96 - ST012&ST015 Charpy test. ............................................................................................. 74
Figure 97 - Marangoni flow. ............................................................................................................... 75
Figure 98 - Lorentz flow..................................................................................................................... 75
Figure 99 - Anisotropic mechanical properties. .................................................................................. 77
List of Tables
Table 1 - Reference tower details ...................................................................................................... 17
Table 2 - Selected Self-shielded wire for testing FCAW-SS ............................................................... 27
Table 3 - Selected gas-shielded wire for testing FCAW-GS ............................................................... 27
Table 4 - Pre-heat calculation results ................................................................................................ 35
Table 5 - Current, stickout and diameter variations. ........................................................................... 52
Table 6 - Heat input test table ........................................................................................................... 55
Table 7 - Microstructures................................................................................................................... 56
Table 8 - Weld Dilution ...................................................................................................................... 56
Table 9 - Cooling rate variables and values. ...................................................................................... 58
Table 10 - Dc value for each HI. ........................................................................................................ 58
Table 11 - Cooling times, rates and peak temperature @ r=3mm for different HI. .............................. 59
Table 12 - Welding speed and oscillation of ST018 ........................................................................... 70
Table 13 - Welding parameters, HI and Deposit rate calculation ........................................................ 70
Table 14 - Weld bead measurements and corresponding ISO 5817 level .......................................... 70
Table 15 - ST samples tensile testing ................................................................................................ 73
Table 16 - Average values of Charpy Test of ST008 in S460M steel(welded transversely to rolling
direction) ............................................................................................................................ 73
Table 17 - Average values of Charpy Test of ST004 & ST007 in S460M steel (welded parallel to rolling
direction) ............................................................................................................................ 73
Table 18 - Average values of Charpy Test of ST013 & ST014 in S355J2 steel(welded transverse to
rolling direction).................................................................................................................. 74
Table 19 - Average values of Charpy Test of ST012 & ST015 in S355J2 steel (welded parallel to
rolling direction).................................................................................................................. 74
Table 20 - Heat Input Trial Data ........................................................................................................ 86
Table 21 - Filler wire specification- SS. .............................................................................................. 95
Table 22 - Filler wire specification- GS. ............................................................................................. 96
Table 23 - HSS properties ................................................................................................................. 97
Table 24 - Mechanical properties of S460M plates ............................................................................ 98
Table 25 - Nominal composition of S460M plate ................................................................................ 98
Table 26 - Mechanical properties of S355J2 plates............................................................................ 98
Table 27 - Nominal composition of S355J2 plates ............................................................................. 98
Table 28 - Charpy results for ST008. ................................................................................................. 99
Table 29 - Charpy results for ST004&ST007. .................................................................................. 100
Table 30 - Charpy results for ST013&ST014 and ST012&ST015. .................................................... 100
IX
Abbreviations
AWS
CR
CTOD
EC3
FCAW
GMAW
HAZ
HSS
IIW
NDT
NORSOK
PAUT
SAW
SMAW
ULS
UT
WEC
pWPS / WPS
WM
BM
CE / CET
EWF
WFS
WS
HI
WM
PM
RT
WEC
American Welding Society
Computer radiography
Crack tip opening displacement
Eurocode 3, EN 1993
Flux cored arc welding
Gas metal arc welding
Heat affected zone
High Strength Steel
International Institute of Welding
Non-destructive testing
Standards Norway, with assistance of Norwegian petroleum industry
Phased Array Ultrasonic Testing
Submerged arc welding
Shielded metal arc welding
Ultimate Limit State
Ultrasonic testing
Wind Energy Converter
Preliminary / Welding Procedure Specification
Weld Metal
Base Metal
Carbon Equivalent, according to EN1011-2
European Welding Federation
Wire Feed Speed
Welding Speed
Heat Input
Weld Metal
Parental Material
Radiography Testing
Wind Energy Converter
X
1 Objectives and Motivation
Nowadays, conventional towers have reached all transportation limits in dimension and weight by
road. Limitations are even more restrictive for onshore wind generation as the best wind conditions
exist in remote areas which mostly have very limited accessibility conditions therefore the need for
larger and more efficient Wind Energy Converters (WEC) can only be achieved by innovation in
manufacturing and transportation capacity as well as assembling and erection methodologies.
Onsite manufacturing of the lower tower sections in a mobile factory including correct positioning
and onsite welding will overcome transportation restrictions and therefore will allow larger bottom
diameters, which take advantage of best wind conditions that exist in remote areas with limited
accessibility conditions and improve WECs overall behavior to fatigue.
The main objective of this thesis was the optimization of welding procedures using Fluxed Core
Arc Welding (FCAW) in wind energy converters towers replacing bolted flanges, factory production
and Submerged Arc Welding (SAW) by onsite fabrication with FCAW, using two steel grades, S355J2
and S460M. This welding technology must ensure the mechanical properties required for this type of
construction, be appealing price-wise and be compatible with the new construction and erection
method of the towers.
2 Thesis Structure
The thesis is divided in nine chapters:

The first chapter is the objectives and motivation of this thesis.

The second is the thesis structure where the different chapters are explained.

The third chapter is the state of art. In this chapter it is presented an overall view of wind
energy converters and competing technologies. Moreover, there is a brief description of the
welding technology for this project.

The fourth chapter is the experimental procedure chapter. In this chapter there is a
description of all the procedure since the selection of consumables, S460M heat input tests
and methodologies used to obtain the optimized parameters for mechanized welding.

The fifth chapter covers the results and analysis of all experimental procedures described in
the previous chapter.

The sixth chapter contains the thesis conclusions.

The seventh chapter suggests possible future work for this thesis subject.

The eight chapter shows the citations and literature survey used in the research and
execution of this thesis.

Finally, the last chapter is the appendix that covers additional information.
1
3 State of Art
3.1 Energy transformation processes and its use
More and more power is needed. This is a simple statement that should make everyone think
about the future of our planet.
With the incredible technological evolution came the need to have more power, the consumption
worldwide has increased in such a way that the word sustainability is being taken in consideration
more and more nowadays.
The conventional methods for producing electrical power use natural gas, coal, petroleum,
nuclear and hydropower. From these five methods only hydropower is a pollution free form of energy
production since nuclear produces nuclear waste and high danger in case of malfunction.
Due to environmental constraints, especially greenhouse gas emissions, values of pollutants like
NOx and carbon dioxide must be reduced according to the Kyoto Protocol [1]. At the same time
research to develop methods to gather energy with lower levels of pollution and reduce the carbon
footprint was greatly supported, the new technologies emerging from this effort are called Green
Energy [2].
Renewable technologies are essential contributors to sustainable energy as they generally
contribute to world energy safety, reducing dependence on fossil fuel resources, and providing
opportunities for mitigating greenhouse gases.
It is also important to say that the levels of pollution emitted by coal and natural gas power plants
has been greatly reduced none the less they still exist.
Green energy includes natural energetic processes e.g. geothermal, wind, small-scale hydro,
solar, biomass, tidal and wave power. It is important to remember that no power source is entirely
impact-free. In order to gather energy from any source it is require to spend energy and give rise to
some degree of pollution by manufacturing the necessary structures however by assessing
environmental impacts associated with all the stages of a product's life from-cradle-to-grave (i.e., from
raw material extraction through materials processing, manufacture, distribution, use, repair and
maintenance, and disposal or recycling) the benefit of using “green methods” is obvious. Life Cycle
Assessment can help avoid a narrow outlook on environmental concerns [3].
The unpredictable factor of nature does not ensure a continuous supply of wind or sun, other
methods like tidal and wave are difficult to implement therefore is impossible at this time to rely only on
power provided by green energy methods.
A symbiosis between Green energy and coal/natural gas power plants and hydro is very
important since when weather conditions do not allow the correct operation of green power generators
and conventional thermal power plants must step in to assure the power supply but when weather
conditions are propitious, power plants can initiate a stand-by state saving precious fossil resources
[3].
2
3.2 Renewable energy Technologies
The type of renewable energy technology that can be implemented is directly influenced by the
geographical and environmental aspects therefore a thorough analysis must take place to ensure
maximum efficiency.
Renewable technologies are divided in three generations.
3.2.1 First generation technologies
The first generation is most competitive were the natural resources are abundant, this generation
is composed by hydroelectric plants, which are structures of long life with few emissions however
dislocation of people and ecosystems changes may occur. Biomass combustion and geothermal
power plants that can operate round the clock ensure a base load capacity but are only accessible in
limited areas of the world [6].
Figure 1 – Scheme of a conventional hydroelectric dam (left image), and conventional
geothermal system (right image). [Wikipedia & Climatepedia]
3.2.2 Second generation technologies
Wind energy conversion is the main technology in this generation of green technologies due to
high potential and relative low costs making it a very attractive investment however some issues due
to aesthetic or environmental reasons and connection to electricity grids may difficult a wind mill
implementation.
Solar heating systems cannot produce electricity, solar thermal collectors heat a fluid and the
heat is then storage in a reservoir or tank for subsequent use. This system is mostly used in domestic
and small scale system like swimming pools or space heating.
Besides solar heating systems there are, in less number, solar power plants that are able to
convert solar energy to electrical energy. The normal operation of this method is subordinate to
weather since a cloud cover can disrupt solar radiation and halt all production [6].
3
Figure 2 - Solar power plant in south Spain (right image), Serpa Solar Park, Portugal (left
Image). [Solarthermalmagazine.com&Wikipédia]
In this generation there is also transformation of biomass into oil subprodutcs like fuel ethanol
which is another important technology. Brazil has reached complete self-sufficiency in oil thru the
production of ethanol from sugar cane together with domestic exploitation of oil resources
demonstrating the importance as backup to other means of energy production [6].
3.2.3 Third Generation Technologies
Third generation technologies are still in research and development although they already
demonstrate a comparable potential to other renewable technologies.
These newest technologies include advanced biomass gasification, biorefinery technologies,
solar thermal power stations, hot dry rock geothermal energy, and ocean energy (Tidal and
Wave).An interesting fact is that Portugal is home of the first wave farm located near Póvoa de
Varzim, Porto [5].
Figure 3 - Pelamis power conversion module.
[greentech.co.uk]
Figure 4 - Wave farm device generator,
Portugal. [energyinformativ.org]
4
3.3 Wind Energy
In order to seize this energy it is necessary to convert wind energy to mechanical energy thru
wind turbines. Wind energy is an attractive alternative to fossil fuels mainly because is a clean,
renewable, plentiful and widely distributed source of energy.
According to some tests the maximum amount of wind energy that can be converted is 59.3%,
also known as the Betz Limit, however realistically the efficiency achieved is about 25%.
Some studies calculate that the amount of power capable of extraction from wind on land and
near-shore is around 72 TW, equivalent to 54,000 MToE (million tons of oil equivalent) per year, over
five times the world's current energy use in all forms [4,6]
.
Figure 5 - Off-Shore Wind Towers.
[meteorologynews.com & windtaskforce.org]
In Portugal the stationary electricity provided by wind power is becoming very meaningful, a huge
increase can be seen since 2009 (19%) until 2011 (57,6%).
Hydroelectric
13,3%
Nuclear
5,0%
Coal
15,4%
SRP
35,5%
Other
7,1%
32,4%
Other
Cogeneration
Wind
Natural Gas
29,4%
Natural Gas
Fuel Oil
1,3%
54,5%
Coal
Hydroelectric
SRP
6%
*SPR – Special Production Regime
Figure 6 - Portugal Electricity Generation by Source (2009)
[European Commission’s Directorate-General for Energy]
5
It is worth mentioning that 5.0% of nuclear power as part of the electricity is imported from Spain
by the international grid, this energy incorporates a portion of nuclear energy. This production
generates radioactive waste however they are treated at the origin country.
Figure 7 - Portugal Electricity Generation by Source (September 2011)
[European Commission’s Directorate-General for Energy]
3.3.1 Wind energy transformation
Major problem with wind energy is its intermittency, this issue rarely creates problems when
supply is up to 20% of total electric production but a failsafe technique such as pumped-storage
hydroelectricity, grid upgrade or lowered ability to supplant conventional production will increase costs
[4].
Power grid management when wind energy is integrated in the national electricity production is
very challenging, several techniques for the excess or lack of wind power production are needed, for
example exporting or storing excess power, complement production with backing supply such as
natural gas, are very important to maintain the desired level of electric power.
A very important technique to storage unused power is called Hydro-Wind Complementarity,
hydropower plants with pumping ability can compensate wind power fluctuations or take advantage
when over-production occurs [8].
At night time the power demand is low but if wind is available it makes no sense stopping the
wind generators and electrical over production occurs. This over production is used to pump water and
increase the water storage in upstream reservoirs of dams.
During the day the inverse may occur, the power demand increases above wind production which
is lower at daytime (because at night the land cools off more quickly than the ocean when the
temperature onshore cools below the temperature offshore, the pressure over the water will be lower
than that of the land establishing a land breeze), the water storage during the night can now be used
do increase the hydro production and compensate the power output [7,8].
Hydroelectric are not dedicated to this type of operation, they also supply directly to the grid
allowing the use of its reserves at any time of day to ensure variability of the production by other
means.
6
Figure 8 - Daily Hydro-Wind Complementarity. [EDP]
3.3.2 Wind energy converters description
Wind energy converters are nowadays manufactured in a range of vertical and horizontal axis
types.

Horizontal upwind: The generator shaft is positioned horizontally and the wind hits the blade
before the tower.

Horizontal downwind: The generator shaft is positioned horizontally and the wind hits the
tower first then the blade.

Vertical axis: The generator shaft is positioned vertically with the blades pointing up with the
generator mounted on the ground or a short tower.
Figure 9 - Vertical Axis Wind tower.
[motionsystemdesign.com]
Figure 10 - Wind Turbine Configuration,
Vertical and Horizontal type.
[motionsystemdesign.com]
The most common style, large or small, is the "horizontal axis design" (with the axis of the blades
horizontal to the ground). On this turbine, two or three blades spin upwind of the tower that it sits on
[9,10].
This type is composed by 3 major modules: the blades, nacelle and tower.
7
Figure 11 - Horizontal axis wind turbine breakdown.
[wind-energy-the-facts.org]
3.3.2.1 Blades
The components which a blade is made of are:
Root of the blade - made from a metallic cylinder which has bolts in order to connect the blade to
the rotor hub.
Blade core - or spar is normally made of balsa wood or foam, the core provides the blades shape.
Shell - upwind and downwind shell are made of fiberglass and epoxy resin, each shell goes from
the leading to the trailing edge and are glued to the spar with a special adhesive.
Safety systems - lightning receptors all along the length of the blade and sensors in the blade to
monitor stress, strain, acoustic emissions and other signals are installed to prevent damage when
weather conditions are unfavorable.
Figure 12 - Blade Cross Section.[4]
WEC blades are very similar to airplane wings both in its aspect and aerodynamics however a
very important difference is the capability to continue operation under stall conditions between rated
and cut-out wind speeds, they are also thinner and longer in order to have enhanced performance
8
even in low wind speed but in order to have a high performance and avoid damaging the blades they
should be frequently clean from dust, dead insects and other debris.
The dimension of the blade is directly related to the height of the wind tower and nacelle installed.
The larger mass-produced blade is 61.5m long with a weight of 6 tons, this weight was achieved by
the combination of different blade components of carbon and glass fibers [4,11].
3.3.2.2 Nacelle
The nacelle is where the heart of the energy conversion occurs, motion provided by the blades is
transformed into electricity and the tower provides structural stability several meters above ground to
both the nacelle and the blades.
The components of nacelle are the main shaft, bearings, gearbox, generator, brake, nacelle
frame, hydraulic systems for brakes and lubrication, and cooling systems.
Gearbox – this component purpose is to increase the rotational speed in order to receive the rotor
low rpm and output to the generator high rpm (for example hub = 24 rpm and speed of a generator =
1,800 rpm. This speed conversion will require a gearbox of 1:75 ratio, which can be accomplished in a
three-stage gearbox.)
Figure 13 - Schematic of multistage gearbox.[4]
Housing and frame of the nacelle are made of glass-reinforced plastic and protects all
components inside the nacelle from the elements of nature.
The type of generator has a significant impact on the efficiency of the turbine rotor. The generator
is responsible to convert the gear box mechanical energy to electrical energy and generators can be
divided in three types: synchronous, permanent magnet, and asynchronous.
For WEC application variable-speed generators are more efficient at capturing wind energy over
a wider range of wind speeds therefore, the utility-scale wind turbine market has moved to this type of
generator [4].
3.3.2.3 Towers
The tower is perhaps one of the most important parts of a WEC. It can also be well over half the
cost of a system overall. In order to assure non turbulent air, thus work efficiently, a wind generator
must be placed at least 10 meters above any surrounding trees or buildings and the tower must be
able to withstand efforts, shaking, and the weight of the whole structure.
9
Figure 14 - Dimension comparison of Wind Tower with a Boeing 747
Nowadays almost every weld on WEC tower is made on the factory using SAW and the preassembled modules are then transported to the erection site. The thickness varies from 12 to 75
millimeter depending on the specific tower design, the plates are rolled into cylinders and then SAW is
used to make all weld seams in the construction (it is estimated that more than 90 percent of the total
welding for tubular wind turbine towers is performed by submerged arc welding), FCAW is used for
door frames, internal fittings (man-lift system as means of transportation for repair crews) and
platforms [12].
Figure 15 - Left: Rolling, forming and tack welding of the shell. Right: External and internal
longitudinal submerged arc welding. Multi-wired SAW is widely used here. [12]
The connection at both cylinder ends is assured by flanges also welded by SAW inside the shell,
for posterior onsite assembling and bolting. Normally each wind tower consists of two up to five
sections.
10
Figure 16 - Terminology and a rough component layout of the tower. [12]
With the a increase of tower diameter thinner walls can be used, the maximum diameter of the
section it is limited to 4.3 meters and 40 meter of length due to transport restrictions, the minimum
height of an overpass is the restrictive factor. This restriction implies limitations to the total height,
weight of the nacelle, rotor, tower itself and in consequence energy conversion efficiency.
Figure 17 - Image demonstrating size restriction for transport. [geograph.org.uk]
The tower height is a very important factor as increasing the tower height by 40 meters can double
the wind energy available.
A very pragmatic cote illustrates very well the importance of the tower height decision:
"A tower too short is like putting a solar system in the shade"
3.3.3 Location
The location where a WEC is installed is a key factor.
What makes a good location? There are many factors to consider, the following is a list of
specially good locations: on the shore of large bodies of water, a western exposure is frequently best,
within a wind tunnel that runs East/West, preferably with a clear Western exposure, an area where
11
wind is funneled and concentrated and high areas, especially high areas with an open view to the west
clear of obstructions.
On the other side tall and irregular forest canopy or other irregular obstructions surrounding the
location and in a North/South running valley or behind a hill to your West should be avoided.
Figure 18 - Wind tower locations in main land Portugal and islands. (2010)
[Energias de Portugal- EDP]
3.3.4 Onsite Build
3.3.4.1 Site preparation
The first step to erect the wind tower is the site preparation, this task includes:

Upgrade of public roads

Wind farm land preparation - clearing of brush and trees, leveling land, construction of
access road, and other tasks to make the entire wind farm easily accessible to earthmoving
equipment, section transport and cranes.

Wind turbine land preparation - crane pads for both main and tail crane, tower, nacelle, and
blades staging area, rotor assembly area, storm water drainage, foundation excavation and
compaction.

Temporary storage area - created to store items like cables, rebar and other material.
3.3.4.2 Foundation construction
The next step is the construction of the foundation.
12
The soil must be excavated to the projected depth then an outer form is placed followed by rebar,
bolt cage and cables conduits, the concrete is then poured and let to cure (may take up to a month).
Figure 19 - Wind tower foundation. [sunjournal.com]
3.3.4.3 Transport
Third step is to transport the tower sections, blades and nacelle from where they were built to the
erection location. This is a very complicated task due to the size of the cargo and the difficult access of
erection locations.
In case vessels can be used WEC components are transported from the production shop to the
closest location possible of the erection site due to its high cargo capability and lack of course barriers.
From the harbor to the erection site WEC components must be transported by road using trucks.
Mountain roads, small villages, overpasses, bridges represents serious complications, added costs to
the project and schedule delays since precise maneuvers and a big transport convoy is needed to
ensure the success of the operation.
Figure 20 - Mountain road transport maneuvers. [offshore-dialog.de]
13
The transportation of WEC components is a planning challenge, with complex logistics and
elaborated means of transport. All transport stages must be planned and verified very carefully since
there is no room for error.
If a route is miscalculated, during the transport components will not pass thru an area of the route
and the hole project will alt and it will be a major setback since it is necessary to remove all transport
equipment and components from the public street, furthermore future transportation will not be able to
use that route and a new one must be plotted [13,14].
3.3.4.4 Erection
WEC erection is a process that can take two or three days and requires two cranes, the main
crane with a capacity of 500 to 650 tons is required to lift the tower sections, nacelle and blades, the
auxiliary crane or tail crane has a capacity of about 90 tons.
First step is to assemble the tower, which consists of two or five sections, with a precise effort of
the two cranes. The main crane lifts the section while the auxiliary crane controls the section swing
and position.
Figure 21 - Workers connecting tower section (left) and tower section with foundation (right).
[brightdirections.co.uk]
With each section stacked in position they now must be bolted to the foundation, in case of the
first section, or to the previous section. The outside surface of the tower is smooth and conical and in
the inside flanges allow then to be bolted together.
Figure 22 - Workers bolting flanges in tower section.
14
After the tower sections have been stacked and bolted, the nacelle is lifted. Depending on the
turbine manufacturer, there are two options for lifting of nacelle:

Single lift of nacelle with generator

Two lifts: first, the nacelle is raised without a generator, and then the generator is lifted
and placed in the nacelle.
The WEC will be complete after the rotor and blades are lifted and assembled. The lift of the
blades represents a very precise operation in order to prevent the blades from swinging and hitting the
tower during the lift. For this operation the usual strategy is the use of two cranes working in tandem.
Figure 23 - Cranes lifting blades and rotor. [AMEC Wind]
The method of joining all components is mainly bolted, foundation to tower, between tower
sections, blades to hub, hub to generator, generator to nacelle and others. A major concern in the use
of this method is to assure that all the bolts are properly tightened.
Bolts must be subject to adequate tension in order to withstand a high fatigue life. Correctly
tensioned bolts are subjected to a small change in tension as external loads are applied. Insufficient
tightening of bolts has been a significant cause of failures. Torque-based methods for tightening of
bolts have been a source of problem.
The torque method isn’t accurate due to variations in friction between bolt and nut causing
misleading values. Hydraulic tensioning of bolts is an alternate method of tightening, in which the bolts
are tensioned to an appropriate level (desired tension plus load transfer relaxation) and then the nut is
turned down [4,15].
3.3.4.5 Complementary constructions
With the completion of the construction of the wind tower itself it is now necessary to implement
electrical systems in order to connect to the grid such as Collection system, Substation and
maintenance building construction, SCADA Systems (control and communication).
15
3.3.5 Growth trend
According to WWEA (World Wind Energy Association) in 2010, more than half of all new wind
power was added outside the traditional markets of Europe and North America, new construction in
China, accounted for nearly half the new wind installations (16.5 GW).
Figure 24 - World Wind Power capacity growth from 1996 to 2010. [WWEA]
Although the wind power industry was affected by the global financial crisis in 2009 and 2010, a
BTM Consult five year forecast up to 2013 projects substantial growth. Over the past five years the
average growth in new installations has been 27.6 percent each year. In the forecast to 2013 the
expected average annual growth rate is 15.7 percent.
More than 200 GW of new wind power capacity could come on line before the end of 2013, wind
power market penetration is expected to reach 3.35 percent by 2013 and 8 percent by 2018 [16].
Figure 25 - Prognosis of installed wind capacity until 2030. [WWEA]
Based on the yearly forecasted wind energy investment shall reach approximately 20 billion in
Europe by 2025, as renewable are very high in the EU political agenda, as shown by the recently
approved European legislation on emissions reduction [17,18].
16
3.4 Wind Energy Converter tower design
The aim of the project is to design a new wind energy converter with a rated power of 4-5 MW
and a hub height of 120 meters with a modular tower of cylindrical closed-walled tubes with variable
diameter and wall thickness and implement a completely different erection method (field-assembled
tower panels carried out by onsite welding with FCAW) thus allowing larger base diameters than 4.3m,
more efficient steel use and installation in remote areas which have mostly the best wind conditions by
eliminating the transportation restrictions.
Table 1 - Reference tower details
Transportation
For the selected concept, the top part of the tower is expected to be transported once assembled
on shop floor. The remaining bended sections of the tower are to be assembled on site. Thus, they
must be transported from the shop floor to the placement site of the tower by means of trucks
equipped with adequate tooling and handling systems. The restriction shall be again related with the
limits of the road (size, length and weight).
Note: The tower design will be based on the GL Guideline for the Certification of Wind Turbines. The
structural verification will be carried out according to the series of standards of EN 1993 (Eurocode 3) and
IEC 61400.[25]
Assembly
For the onsite assembly, four semi-circular tiles must be welded to each other onsite becoming a
tower section, this newly created section will be positioned by cranes on its final location in the tower
assembly and finally welded to previously placed sections.
Figure 26 - Assembly overview, circular segments and sectors assembled on site. [19]
17
Improvements are also expected by the use of high strength steel namely S460M leading to
smaller thickness and increased buckling resistance. Bolted connections are replaced by onsite
welding will also lead to better fatigue behavior.
Flanged joints enable a short onsite erection time but are very costly due to the number of bolts
needed (120 to 150 high strength bolts per flange and 4 to 5 flanges per tower), correct pretension in
order to guarantee fatigue strength for the entire WEC service life is very difficult to oversee.
The use of bolted rig flanges also introduces geometrical effects due to their massive L-shape
which are critical for both the ultimate limit state (ULS) and the fatigue limit state (FLS) [19].
From recent studies it is known that with the use of proper welding technologies and high strength
consumables a superior fatigue performance can be achieved in welded joints of HSS thus creating an
alternative to the traditional bolted connections and allowing a weight reduction due to the use of
reduced thickness [19].
Thinner steel plates provide a more homogeneous grain structure through thickness, higher yield
strength, lower residual stresses and improve fatigue strength as it depends directly on the wall
thickness as their higher yield strength limits the extension of the plastic zones associated to notch
areas, which delays the fatigue crack initiation [20,21,22].
3.4.1 Tower Weight
The only possible solution, due to transport restrictions, to withstand an increase in axial and
bending loads is to increase the wall thickness however thickness increase has much less influence
on the bending load capacity than increasing the tower diameters. Besides, very large increase of the
plate thickness has further disadvantages, like increasing weights, therefore there was scope to
develop a tower design that allows for an increased base diameter.
According to the project simulation the optimal tower design is for the given situation a 120 m hub
height and 3.75 m top flange diameter, consisting of 5 sections with 20 m to 25 m. A tower with larger
base diameters than 9500 mm is not reachable without stiffeners, provided that strong oversizing
should be avoided. The gross load weight has its minimum at a base diameter of 8000 mm [19].
Figure 27 - Tower weight depending on base diameter [19]
3.4.2 Plate thickness
A study to relate the wall thickness to the base diameter and tower height was also carried out.
18
With a base diameter of 4500mm the maximum wall thickness was around 200mm, a reference
tower with 100m and 4.5MW has about 65mm of wall thickness and its base diameter is around
7500mm.
Figure 28 shows the dependence of the maximum thicknesses from the base diameter [19].
Figure 28 - Maximum plate thicknesses depending on base diameter for unreinforced and
reinforced tower. [19]
3.5 Welding technology
Welding is a process that is used to join two or more components, usually for metallic material.
The welding process is of great importance and is used when complex shapes or very big structures
are needed. Instead of making one big part, it is possible to join smaller parts which facilitate e.g.
transportation and production.
3.5.1 Welding Technology - Fluxed Core Arc Welding (FCAW)
Process principles
FCAW is a fusion welding achieved by an electric arc produced between a continuous filler metal
electrode and the weld pool.
In this method the filler metal of tubular shape is continuously fed and has a fluxed core which
provides shielding capabilities to the welding process with or without additional shielding from an
externally supplied protection gas. The core is mainly formed by slag formers, deoxidizers, arc
stabilizers, and alloying elements.
This process has two major variations that differ in their method of shielding the arc and weld pool
from atmospheric contamination. One type consists of self-shielded FCAW, also known as FCAW-SS,
and the other type is gas-shielded FCAW, or FCAW-GS.
Welding Protection
Arc shielding
When joining metals with the use of heat the high temperature makes the metals chemically
reactive with oxygen and nitrogen present in the atmosphere, promoting the formation of oxides and
nitrides that weaken the strength and toughness of the metal.
19
Covering the molten pool with a protective shielding gas, vapor or slag decreases the negative
impact of the oxygen and nitrogen while welding by minimizing the contact of the molten metal with the
atmosphere.
The welding protection against nitrogen and oxygen found in the atmosphere can be achieved by
two processes:
Self-shielded (FCAW-SS) - this process requires that all the protection of the weld metal against
some elements of the atmosphere must be delivered only by the core ingredients without any external
shielding assistance.
The elements in the core that provide protection to the weld create their own shielding gas
through the vaporization and decomposition of core ingredients, as the wire is consumed air is
displaced and slag is formed which enables the protection of the weld metal drop since it melts until it
reaches the molten weld pool by slag covering.
In order to achieve improved quality welds many self-shielded electrodes use deoxidizers and
denitrifying compounds, the use of arc stabilizers and alloying elements also contribute to sound
welding. The attraction of self-shielded method relies in the fact that it is portable and generally has
good penetration into the base metal. However, this process can produce excessive, harmful smoke
(making it difficult to see the weld pool)
Gas-shield (FCAW-GS) – in this method a shielding gas is provided externally and added to the
existing protection delivered by the flux (“dual shield” welding). The used gas depends on the material
to weld usually CO2 and Argon, in pure state or mixes.
This particular style of FCAW is preferable for welding thicker and out-of-position metals, this
method allows a higher production rate. However it cannot be used in a windy environment without
any additional protection of the welding area, as the loss of protection of the gas from air flow will
result in porosity in the weld metal, often visible on the welding surface.
Application
While both variants are usable indoors for field use the self-shielded process is preferred due to
the possibility of drafts disrupts the gas protection of the gas-shielded process.
Fluxed core is a very versatile process utilized in a variety of fabrication shops due to its wide
range of applications and capability to weld thin metals as that used in vehicles bodies and as thick as
structural steel in buildings.
Main process variables are:

Wire feed speed (and current)

Arc voltage

Electrode extension

Travel speed and angle

Electrode wire type and diameter

Shielding gas composition (if required)
20
Equipment
FCAW process relies on semiautomatic, mechanized and fully automatic systems.
The basic equipment includes a power supply, wire feed system and a welding gun. In case the
use of shielding gas or a higher level of automation is needed additional equipment will be required
like gas shielding system and also fume removal equipment in most applications especially indoors.
Figure 29 - Typical setup for semiautomatic FCAW equipment
The power sources generally recommended for FCAW are direct current constant voltage type,
both rotating and static generator types are used. Welding guns may be either air-cooled or watercooled.
Mechanized and automatic FCAW
The equipment used in mechanized and automatic is not substantially different from that used in
the semiautomatic FCAW process. Various travel mechanisms are used, depending on the
applications. These mechanisms include side-beam carriages, tractor-type carriages, and robots.
Amongst the advantages of FCAW it can be said that has high-quality weld metal deposit (high
current density),excellent weld appearance (smooth, uniform welds), relatively high electrode deposit
efficiency and higher tolerance to contamination that may cause weld cracking. On the other side
production of a slag covering which must be removed, increased FCAW electrode wire cost compared
to solid electrode wire, the wire feeder and power source must be fairly close to the point of welding
and more smoke and fumes are generated compared to GMAW and SAW.
General description

Self-shielded FCAW: easiest equipment arrangement

All-position welding (with the proper choice of consumables)

Easy to weld in vertical-up position (with the proper choice of consumables)

Designed for single and multiple pass welding

Semiautomatic fully mechanized and automatic applications

Current type DC+ for gas shielded wires, DC- for self-shielded wires (with exceptions)

Good impact and CTOD toughness
21
Filler Wires
Self- shielded flux-cored wires
Considering the general requirements of the project, self-shielded FCAW would be a convenient
and obvious choice to address the welding to be performed on site, since a great deal of positional
work is expected to be done outdoors during tower erection.
Gas-shielded flux-cored wires
Anticipating any possible shortcomings of self-shielded wires to address some welding issues
expected, as the matching of mechanical properties of S460M steel or the actual fittingness of
particular types for the mechanized welding that is to be implemented during tower erection, gasshielded wires were also be considered.
Although the use of gas-shielded wires requires protection from wind and draughts in the
immediate vicinity of the welding area when compared to their self-shielded counterparts, average
gas-shielded wires will more readily meet higher requirements for properties of weld metal.
All-position rutile types will be preferred over basic or metal cored types, due to their far better
aptitude for positional welding. The fast freezing slag supports the molten pool and allows for higher
currents in positional work, increasing productivity in a very effective way [28,29,30].
3.6 Innovations and Benefits
FCAW was first developed in the early 1950s and it is commonly use in structural welding around
the world. Gas-shielded (FCAW-GS) wires were introduced to the market around 1957 and selfshielded (FCAW-SS) wires were introduced to the market later, around 1961 this might explain some
resistance from welders to use self-shielded wires together with a more demanding welding technique.
In many outdoor structural jobs, such as buildings, shipyards and bridges, welders prefer the
extra work of protecting the workspace from nature and additional equipment required to weld with
gas-protected wire instead of it self-shielded counterpart.
The present work tested a wide range of wires, gas protected and self-shielded, with the goal of
optimize the welding parameters for future mechanization which isn’t that common in FCAW especially
in vertical up position, most of mechanizations are orbital. Self-shielded wire testing is very important
since it has a huge potential but because prejudice and habit, on behalf of welders, the wire
development is longstanding since there is no feedback from the end user to the manufacturer. A
more frequent use of this kind of wire by the welders will inevitably lead to an acceleration of
development and continuous use of self-shielded wires.
Welding in vertical up position is extremely important for the new WEC in question. Nowadays
WEC towers are manufactured in factories, transported to the desired location and assembled thru
bolted connections.
Two other types of towers are the Hybrid Concrete/Steel and with friction connection tower, both
are relative new tower options a still maintain the problem of transportation to erection site. If success
22
is achieved mechanizing FCAW along with the use of higher steel grades, a fourth construction option
can be added. The ability of onsite mechanized welding simplifies tower components and machinery
transportation as well as better fatigue behavior moreover the use of High Strength Steels will
translate in a smaller necessity of steels ton when compared with conventional design.
After having an overall understanding of the welding process and wind energy converters
technology indispensable for this study the next chapter describes the approach and procedures that
will enable to obtain conclusive results about the feasibility of this project.
23
4 Experimental Procedure
The experimental procedure can be divided in four stages. In the first stage manual trials selected
the appropriate FCAW wire, the second stage is related to the base materials to be weld, metallurgical
properties and the need for pre-weld heat treatment. Third and fourth stage are related to the FCAW
mechanized trials and welding the mechanized samples for mechanical testing respectively.
Figure 30 - Experimental Procedure Diagram
4.1 Equipment
Diverse equipment was used in order to evaluate all wires and preform the mechanized welds.
For manual trials two welding couples were used: for self-shielded wires welding machine Lincoln
Electric LN-25 Pro and feeder Lincoln Electric Invertec V 350 Pro; for gas-shielded wires and Lincoln
Electric Idealarc CV 400-I with Lincoln Electric LN 742.
The equipment for the mechanized trials consisted of a Tractor Gullco Gk-197-FO/036, a welding
machine Kemppi Mig 500 + Pro 5000, a one axis rotating positioning Table/Jig – ABB MTB250 and a
Data Acquisition System - DEMTEC – D-ADS to record all welding parameter for each individual weld.
Figure 31 - LE Idealarc CV 400-I & LN 742
Figure 32 - LE LN-25Pro & Invertec V350 Pro
24
Figure 33 - Welding machine, Jigg and ADS for mechanized welding.
4.2 Preliminary test of FCAW wires
General process assessment and preliminary specification of basic welding procedures were
carried out in order to select the most appropriate gas-shielded or self-shielded wire while having the
highest deposition rate possible and ensuring all necessary mechanical properties and zero welding
imperfections.
Due to project restrictions the joint can only be welded from one side (for 10-12mm thickness)
therefore as a direct consequence the only type of joint recommended is a single V-shaped butt weld,
for 30mm thickness a double V joint will be used (shape and dimensions of test pieces meet minimum
requirements of standard EN 15614-1), moreover ceramic backing were tested in self-shielded wires
although they are optional, on the other hand ceramic backing are mandatory for gas-shielded wires.
Ceramic backing enables the deposit of the first layer of molten metal preventing it from escaping
through the root of the joint resulting in a sound first pass without addition of elements from the
backing to the weld bead [28].
Figure 34 - Ceramic Backing
25
The goal is to mechanize the fulfillment of all the welds necessary to the erection of the tower
however by initiating the process assessment and preliminary specification manually it becomes
easier and quicker to test different welding setup and execute a trial and error test type.
After obtaining a well specified manual welding procedure the parameters such as torch position
and angle, welding speed, stickout and type of weaving can be exported to the mechanized trials
saving time, man hours, steel and consumables.
From a welding perspective the most complex welding position, due to operator skill, welding
physics and type of electrode required, is the vertical up (PF), moreover this position will be the most
relevant for FCAW during the erection of the tower.
The application range foreseen for SAW has been put forward as a likely option to cover
horizontal position girth welding during erection, as well as assuring most of the on-site fabrication
therefore the PC position will take a secondary role in this work.
Taking into account the joint characteristics (design and position), mentioned above, suitable
wires must be selected. Wire choice is based in welding economy together with matching structural
requirements and type of joint, wires can be expressed in four groups targeted for the specific job:

Fast-fill – suitable when a large amount of weld is needed to fill the joint.

Fast-freeze – are the best option when welding out-of-position (overhead and vertical) and
quick solidification is key. They can also be a fill-type or penetration type but fast freezing is of
paramount importance.

Fast-follow – ability of the molten metal to follow the arc at rapid travel speed usually used in
single pass.

Penetration – used when deep penetration welding must provide adequate mixing of weld and
base metal.
The obvious choice for this welding task is the fast-freeze wire type. The deposit weld metal
solidifies rapidly after being melted, supporting the molten pool, allowing higher currents in positional
work and increasing productivity in a very effective way.
Welds made by this wire type are slow (low travel speeds and low voltage compared to other
positions) and required high skill on behalf of the operator, deep penetration, maximum admixture and
a flat weld bead with distinct ripples are also trademarks [32].
The ASME/AWS classifications adequate for the desired welding operations are:
E71T-1 (FCAW-GS): Fast Freezing, highest deposition rates out-of-position.
E71T-8 (FCAW-SS): Fast freezing rutile, highest deposition rates out-of-position without a shielding
gas.
4.2.1 Wire selection
After a market research it was decided to select welding consumables from Lincoln Electric and
ESAB, two leading companies in the welding area.
26
The following filler wires were chosen for the self-shielded FCAW tests.
Filler wire
NR-233
Manufacturer Lincoln Electric
NR-203Ni1
NR-Offshore
Coreshield 8
Coreshield 8Ni1 H5
Lincoln
Electric
Lincoln
Electric
ESAB
ESAB
Shielding
type
Self-Shielded
Self-Shielded
Self-Shielded
Self-Shielded
Self-Shielded
Classification
E71T-8 (*)
E71T8-Ni1 (§)
NA
E71T-8 (*)
E71T8-Ni1-J
(§)
Chosen
diameter[mm]
1.6
2.0
2.0
1.6
1.6
Table 2 - Selected Self-shielded wire for testing FCAW-SS
(*) - AWS A5.20/ASME SFA-5.20 & (§) - AWS A5.29/ASME SFA-5.29
For the gas shielded filer wires the following table contains the chosen ones.
Filler wire
Filarc PZ6113
Filarc
PZ6113S
Filarc
PZ6114S
Filarc
PZ6116S
Filarc PZ6138
Manufacturer
ESAB
ESAB
ESAB
ESAB
ESAB
Gas - CO2
Gas - CO2
Gas - CO2
Gas - Ar/CO2
Shielding
type
Classification
Chosen
diameter[mm]
Gas CO2
Gas Ar/CO2
E71TE71T-1M1C-H4 (*)
H8 (*)
1.2
1.2
E71T-9CH4 (*)
E71T-1C-JH4 E81T1-K2C- E81T1-Ni1M(*)
JH4 (§)
JH4 (§)
1.2
1.2
1.2
1.2
Table 3 - Selected gas-shielded wire for testing FCAW-GS
(*) - AWS A5.20/ASME SFA-5.20 & (§) - AWS A5.29/ASME SFA-5.29
Full wire specifications can be consulted in Annex 9.5
In order to select appropriate wires the classification, typical applications and mechanical
properties stated by the manufacturer were consulted.
The NR-Offshore (Code 54) is a prototype wire kindly provided by Lincoln Electric for the tests
however none of its mechanical properties were available.
A special requirement of the project was to assess the welding capability of electrodes containing
Ni, a weld bead with Ni will have an improved corrosion resistance capability important when a WEC
service time is over 20 years.
The electrode diameter chosen was the smallest available for best off position welding especially
for vertical up and for the particular joint geometry.
All wires have mechanical properties higher than those of S355J2 steel however none of the selfshielded wires (with exception of Offshore prototype which information is unavailable) and gas
shielded PZ6113 with CO2 can reach the yield strength and tensile strength of the S460M.
In this case, all-position rutile types were preferred over basic or metal cored types, due to their
far better aptitude for positional welding.
27
Parameter optimization
In order to overcome some problems during parameter optimization a unique solution is not
possible. A parameter modification will echo on the other parameters therefore when tweaking a
parameter a broad view of the modifications necessary is important to understand the changes.
Figure 35 - Welding process variables.
Wire Feed Speed (WFS) - (or welding current)
Wire feed speed is directly related to welding current (if the wire extension beyond the guide tip is
constant). As the wire-feed speed is varied, the welding current will vary in the same direction, in other
words, an increase (or decrease) in the wire-feed speed will cause an increase (or decrease) of the
current. This relationship is commonly called the ”burn-off’ characteristic. Figure 36 also shows that
when the diameter of the wire electrode is increased (or decreased), at any wire-feed speed, the
welding current is higher (or lower).
One important fact is the shape of each burn-off curve. In the lower current range for each wire
size, the curve is nearly linear. In other words, for every addition to the current, there is a proportional
(and constant) increase in the melt off. However, at higher welding currents, particularly with small
diameter wires, the burn-off curve becomes non-linear. In this region, higher welding currents cause
larger increases in the burn-off.
Figure 36 - Burn-off curves. [39]
Figure 37 - Contact Tip to Work Distance. [39]
28
This is due to resistance heating of the wire extension beyond the guide tube known as Joule
heating. This resistance heating is known by Equation 6, the greater the welding current, the greater
the resistance heating.
The resistance (R) is measured in ohms (Ω) of a material depends on its length (L) expressed as
meters, cross-sectional area (A) expressed as square meters, and the resistivity (
in ohm-metres:
(1)
Resistance also depends on temperature, usually increasing as the temperature increases. For
reasonably small changes in temperature, the change in resistivity, and therefore the change in
resistance, is proportional to the temperature change. This is reflected in the equations:
[
(
]
(2)
and equivalently
[
(
]
(3)
The parameter α is called the temperature coefficient of resistance, an empirical parameter fitted
from measurement data, coefficient α is typically 3×10−3 K−1 to 6×10−3 K−1 for metals near room
temperature. R0 is the resistance at temperature T0.
Electric power is given by the equations:
(4)
(5)
(6)
Where P is the power in watts, V is the voltage across an element measured in volts and I is
welding current in amperes.
Summing up
If arc voltage, travel speed and Contact Tip to Work Distance (CTWD) are held constant, WFS
variations have the following major effects:
1. Increasing the WFS increases melt-off and deposition rates.
2. Excessive WFS produces convex beads. This wastes weld metal deposited in excess and
badly distributed.
3. Increasing WFS also increases the maximum voltage which can be used without porosity.
Lowering the WFS requires lowering the voltage to avoid porosity. As the WFS is increased,
the arc voltage must also be increased to maintain proper bead shape.
Contact Tip to Work Distance (CTWD)
Contact Tip to Work Distance or ”stickout” is the distance between the last point of electrical
contact, usually the end of the contact tip, and the end of the wire electrode, seen in Figure 37, it is in
this area that resistance preheating effect occurs.
The contact tip-to-work distance affects the welding current required to melt the wire at a given
feed speed as can be seen by Equations1 and 6.
29
Basically, as the tip-to-work distance is increased (L), the value of Equation 1 will also increase
leading to an increase of the amount of I2R heating (Joule Effect) seen on Equation 6 and
consequently the welding current required to melt the wire is decreased since the welding power is
constant. The converse is also true.
There is also a significant effect on another parameter, weld penetration. Long extensions result
in excess weld metal being deposited with low arc heat. This can cause poor bead shape and low
penetration. In addition, as the tip-to-work distance increases, the arc becomes less stable. It is very
important that the wire extension be kept as constant as possible during the welding operation.
Summing up
If the voltage and wire feed speed setting and the travel speed are held constant, variations in CTWD
have the following major effects:
1. Increasing CTWD reduces the welding current. Decreasing CTWD increases current.
2. Increasing CTWD reduces actual arc voltage (Equation 5) and results in more convex
beads and reduces the tendency of porosity.
3. Momentarily increasing CTWD can be used to reduce burn-through tendency when poor fitup is encountered.
Weld Penetration
Weld Penetration (WP) is the distance into the base material when making a weld on plate
[
]
(7)
For 1 mm diameter solid carbon steel wire, the constant K = 0.0019. WS is welding speed in
cm/min.[ According to “The Science of Arc Welding” by C. E. Jackson. 1960 Welding Journal 39(4) pp
129-s thru 230-s]
Wire melting rate (WMR)
Now that CTWD and WFS were described this rate can now be introduced. WMR relates the
amount of weld material melted with the current and CTWD.
(8)
WMR is expressed as kg/hr and where "a" and "b" are constants for each wire diameter. The values
for "a" and "b" for 1mm diameter wire are a = 0.017; b = 0.00014.
The first term (a x I) is the anode voltage times current and the second term defines the energy
input due to resistance heating, these two energy sources cause the wire to melt [28].
Arc Voltage
Arc voltage is the voltage between the end of the wire and the workpiece. It should be noted that
as welding current and wire burn-off are increased, the welding voltage (arc length) must also be
increased somewhat to maintain stability. Figure 38 shows a relationship of arc voltage to welding
current for the most common shielding gases employed. The arc voltage is increased with increasing
current to provide the best operation.
30
Figure 38 - Welding voltage relation with welding amperage. [39]
In many materials, the voltage and resistance are connected by Ohm's Law:
V = IR
(9)
Summing up
If WFS, travel speed and CTWD are held constant, changing the arc voltage will have the following
effects:
1. Higher arc voltage results in a wider and flatter bead.
2. Excessive arc voltage causes porosity.
3. Low voltage causes a convex ropey bead.
4. Extremely low voltage will cause the wire to stub on the plate. That is, the wire will dive
through the molten metal and strike the joint bottom or the ceramic backing, tending to push
the gun up.
Travel Speed
The arc travel speed or welding speed is the linear rate at which the arc moves along the
workpiece. This parameter is usually expressed as centimeters per minute.
Arc travel speed has two major guidelines:
1) As the material thickness increases, the travel speed must be lowered.
2) For a given constant material thickness and joint design, as the welding current is increased,
so is the arc travel speed. The converse is also true.
Summing up
If arc voltage, WFS and CTWD are held constant, travel speed variations have the following major
effects:
1. Too high a travel speed increases the convexity of the bead and causes uneven edges.
2. Too slow a travel speed results in slag interference, slag inclusions and a rough, uneven
bead.
31
Torch angle
The first general welding technique that affects weld characteristics is torch position. This refers
to the manner in which the torch is held with respect to the weld joint. The position is usually described
from two directions – the angle relative to the length of the weld and the angle relative to the plates.
For the forehand method, often used in FCAW, the torch is angled so that the electrode wire is
fed in the same direction as arc travel, this way the filler metal is being deposited, for the most part,
directly on the workpiece.
When the welding must be performed in the vertical position there are two methods with which
this welding can be done – vertical up and vertical down. Here the torch positioning is extremely
important and welding should be performed with the arc kept on the puddle’s leading edge so as to
insure complete weld penetration.
Figure 39 - Torch angle importance.
4.3 S460M Weldability test
Moderate strength steels with yield strengths up to 350MPa are mainly produced by the
normalizing route. In general, the strength of steel is controlled by its microstructure which varies
according to its chemical composition, its thermal history and the deformation processes it undergoes
during its production schedule.
Structural steel must be readily weldable since this is the traditional fabrication route for
structures and exhibit good toughness to avoid the possibility of brittle failure, in addition to showing
high strength. Such overall requirements are often difficult to achieve because an increase in one of
these properties often leads to a decrease in the others.
The desire to increased strength to weight ratio and reduce cost leads to producing steels by
alternative processing routes such as thermomechanical controlled processing (TMCP or simply TM),
and quenching and tempering (Q & T). Metallurgical principles can also be used to satisfy the overall
mechanical property requirements for high strength structural steels, namely by reducing carbon
content to improve weldability and toughness.
After heating at temperatures of about 1100 °C the rolling takes place after that the plate cools on
calm air and the “as rolled” condition (AR) is achieved (Process A).
Normalized condition (N) (Process A + B), leads to a refined microstructure of ferrite and pearlite
the plate is reheated just above the ferrite-austenite transformation temperature (about 800 – 900 °C
and is cooled on calm air again.
32
The quenching and tempering process is quite similar to the one for normalizing (Process A + C).
After hot rolling and cooling the plate is reheated above the transformation temperature, so that
carbon can dissolve in austenite, but then cooling is not performed on calm air, but in water
(quenching) or in another medium that cools fast enough, so that there is no time for the formation of
ferrite and pearlite which needs a diffusion process. The carbon stays dissolved and at room
temperature the microstructure mainly consists of martensite, a distorted structure that has a high
strength but a low toughness.
The smaller the grain size is the higher are the tensile and toughness properties. The
thermomechanical rolling (TM or TMCP) is a method to realize such a fine grained microstructure
(Processes D to G).
The applied "rolling schedule" is individually designed, depending on the chemical composition,
the plate thickness and the required strength and toughness properties. Especially for thick plates an
accelerated cooling (ACC) after the final rolling pass is beneficial for the achievement of the most
suitable micro-structure as it forces the transformation of the elongated austenite grains before
recrystallization can happen.
Figure 40 - Steel processing routes. [Nordic steel]
For this project two structural grades of steel were considered, the widely used S355J2 and
S460M.
Commonly used in the engineering and construction industry, S355J2 offers high yield and tensile
strength and is supplied with a variety of treatments and test options making it highly usable steel for
construction.
Since S355J2 is a well-studied steel and is well documented and covered by existing codes and
standards our main attentions will be focused at S460M since the benefits expected from an increase
in the strength to weight ratio and the associated savings in the cost of materials were recognized.
The S460M is a micro-alloyed thermomechanically rolled ('M'), weldable fine grain structural steel
with higher strength and improved weldability compared to the S355J2.
Thermomechanically rolled steel generally do not show high hardness in the coarse grain heat
affected zone, which induces a better resistance to cold cracking. Considering the application in wind
turbines, S460M would represent a very large step forward compared to the currently applied grades.
33
Micro-alloyed steel can be defined as a carbon-manganese steel containing deliberately added
alloying elements totalling only 0.05 to 0.10%. Alloying elements such as small amounts of vanadium,
titanium, niobium and boron are effective in modifying steel properties thru precipitation hardening.
Vanadium addition is able to give precipitation strengthening in high carbon steels, niobium has a
particularly strong influence in reducing the recrystallization during hot rolling thus aiding grain
refinement and a small titanium addition is also very effective in refining grain size at high
temperatures in the austenite range. As result of the addition of alloying elements and controlled
rolling procedure major strength and toughness improvements are noticeable while reducing welding
problems by largely reducing the carbon equivalent value (CEV) and higher grain refinement.
The following trends are aimed at maintaining and improving the strong market position of HSLA
Steels: more closely controlled composition, reduced carbon content, increased combination of microalloy elements, lower residuals and improved cleanness, more sophisticated processing, increased
uniformity of properties, minimal heat treatment, improved shape and surface appearance, higher
strength and improved fracture properties and better weldability and toughness.
Main mechanical differences between the two grades of steels are the yield strength with
355MPa and 460MPa and slight lower value of the notch impact test in the S355. In relation to the
steel weldability both steels are very alike as it can be seen by the equal CEV values.
Table with full EN specifications can be consulted in Annex 9.6 and the certificate of the
steel used can be seen on Annex 9.7.
4.3.1 Pre Heating for HSS Steels
Both steel grades have an adequate weldability however there are some adverse aspects that
must be controlled, the most attention worthy is hydrogen cold-cracking.
When producing these types of structures the use of a variety of multipass welding process is
needed. This involves various typical levels for the diffusible hydrogen content in the weld deposit and
Q [KJ/mm]
typical heat input ranges, heavily influencing the need of preheating/interpass levels.
Risk of
HIC
Tp [ C]
Figure 41 - Qualitative welding workspace for thick high strength structural steels. [33]
34
Special attention must be taken for cold-cracking as relative high thickness is being used with special
care of:

Diffusible hydrogen content in weld metal and heat affected zone.

Brittle microstructures in heat affected zone

Tensile stress concentration in weld joint
In order to assess the effect on weldability of this alloy elements and its need of pre-heating,
Carbon Equivalent Value is used.
(10)
(Welding Handbook – 8th edition – American Welding Society)
(11)
(International Welding Institute)
The calculation of preheat temperatures was done by three different methods.
S355J2
S460M
Method
10mm
30mm
10mm
30mm
AWS D1.1
10ºC
65ºC
10ºC
65ºC
EN1011 – 2 2001
0ºC
0ºC(1)
0ºC
0ºC(2)
EN1011 – 2 2001 -B
30ºC (3)
96,5ºC (3)
0ºC (3)
51,7ºC (3)
Table 4 - Pre-heat calculation results
(1) - Value obtained for Heat input > 0.75KJ/mm
(2) - Value obtained for Heat input > 0.9KJ/mm
(3) – Values obtained for Heat input = 1KJ/mm
Pre-heating calculations can be seen on annex 9.1 and steel properties on annex 9.6
As a result, only room temperature preheating is needed for these steels for thicknesses of 10
mm (S355J2 and S460M) however for thicknesses of 30 mm preheating temperatures of 50ºC and
above are needed for both S355J2 and for S460M, and taking care of having no less than 75 to 100ºC
for interpass temperature to obtain good weldability [39].
35
140
Preaheating Temp
[ºC]
120
S460M HD 2ml/100g
100
S460M HD 4ml/100g
80
S460M HD 10ml/100g
60
S355J2 HD 2ml/100g
S355J2 HD 4ml/100g
40
S355J2 HD 10ml/100g
20
0
0,5
2,4
Heat Input [KJ/mm]
3
Figure 42 - Calculated preheating.
Net heat input diagram for the S460M and S355J2 steel grades actually employed in 30 mm thick
butt welds. Various electrode hydrogen classes are presented encompassing the probable project
fabrication scenario. At comparable conditions for heat input and diffusible hydrogen, S460M plates
are predicted to require about 30°C lower preheating temperature as calculated accordingly to EN
1011-1:2009 “Welding - Recommendations For Welding Of Metallic Materials - Part 1:General
Guidance for Arc Welding” and related with practical knowledge acquired by several trials.
Although the need for preheat has been confirmed by values calculated in Table 4 according to
EN ISO 15607:2005 "Specification and qualification of welding procedures for metallic materials General rules (ISO 15607:2003), these low temperature values can neglected if they are compensated
with additional microstructural and mechanical tests described in this standard in order to prove the
quality of welds.
4.3.2 Characterization S460M steel welded with different heat inputs.
Characterization S460M steel welded with different heat inputs was necessary in order to identify
structural changes resulting from high temperature permanence.
Parameters to analysis
• Microstructural - Identification and delineation of regions of the filler metal, base material and
heat affected zone, grain morphology, constituent phases, the presence of precipitated material
and intergranular.
• Testing of hardness (Vickers) - Membership regions hardness of the base material, weld
zone, heat affected zone so as to obtain a plot of hardness profiles and allow better identify
micro-constituents.
• Calculation of weld dilution
Welding Procedure
• Execution of 14 strands of welding (A to N) on the surface of the base material to which the
heat input values are 0.5 KJ / mm; 1KJ/mm; 1.5 KJ / mm; 2KJ/mm; 2, 5KJ/mm; 3KJ/mm and
36
3.5KJ/mm respectively, done by two different methods.
• On completion of a weld seam is necessary that the temperature of the sample to return to
room temperature before running the next string.
• Beads are performed along steel rolling direction.
Equipment used
For this procedure it was used the above mentioned data acquisition system, Kemppi welding
machine CO2 gas, and filler wire Filarc PZ6113S and Gullco tractor.
Figure 43 - From left to right: HI test layout; Welding machine; Data acquisition system; Print
screen of acquisition; HI test piece; Test weld beads
4.4 Mechanized Welding Trials
Fully mechanized welding trials were performed on smaller test pieces to check manual welding
procedures. In some cases some fine tuning was required or different approach was needed since the
mechanized system does not compensate unforeseen situations like a skilled technician.
37
Figure 44 - Fully mechanized welding trials with FCAW selected filler wire PZ6113S: welding of
a 10-mm thick plate butt joint in the vertical-up (PF) position.
A simple excel calculation tool to manage the joint filling was used, this tool provided an estimate
welding speed after inputting the joint geometry and WFS however the welding speed values
calculated must be used with caution as it is calculated for an ideal joint geometry.
The joint is not perfect it has to be taken in account that the joint suffers distortion while welding.
Gap narrowing implies changes in welding parameters in order to obtain the desired weld bead also
the final conditions of a pass may not correspond with the one calculated by the tool therefore each
pass must be assed and recalculate welding speed for the next pass.
It was decided to maintain the WFS constant therefore torch angle and alignment, stick-out, and
welding speed would be adjusted in order to compensate changes in joint geometry.
Although this process is mechanized it is no automated, this is, mechanized welding requires at all
times supervision from welding technician.
These trial samples were subjected to ultrasonic testing on site in order to establish a welding
procedure with sound weld beads, it was of the outmost importance to guarantee flawless welds
before welding the samples in the S460M and S355J2 steel for mechanical testing.
Imperfection assessment and any decision how it could be repaired, requires knowledge about
location, depth, dimensions and geometry. In this context, Non Destructive Testing (NDT) methods
play a fundamental role. The NDT techniques used in this project were Ultrasonic Testing (UT)
conventional and advanced (Phased Array) and Digital Radiography or Computed Radiography (CR)
[34].
Ultrasonic testing (UT)
Mechanical waves with wavelengths in the ultrasonic range are injected in the material by a UT
transducer propagates with a velocity that is different for different materials. These waves propagating
in an elastic medium reflect either:
 at the interface of the material with another medium, e.g. air, and the path travelled by the
wave in the material can be correlated to the material thickness;
38
 or they will be reflected by discontinuities within the material (like flaws). The reflections allow
the localization of the reflectors by measuring the sound path duration, and relating it with the
speed of sound in that material and the emission angle, making this technique useful for flaw
detection and evaluation, thickness measurement, for example.
The ultrasonic testing is a complementary solution for localizing discontinuities, more difficult to
achieve in radiography. Phased Array (PA) technique is a novel and advanced technique of
generating and receiving ultrasound based on array transducers. An array transducer is simply one
that contains a number of separate elements in a single housing, and phasing refers to how those
elements are sequentially pulsed [34,35,36].
Figure 45 - Phased array probes are made in a
variety of shapes and sizes for different
applications. [36]
Figure 46 - Inspection system during circular
welds inspection process in a wind tower. [36]
Phased array systems pulse and receive from multiple elements of an array. These elements are
pulsed in such a way as to cause multiple beam components to combine with each other and form a
single wave front travelling in the desired direction. Similarly, the receiver function combines the input
from multiple elements into a single presentation. Overall, the use of phased arrays permits optimizing
defect detection while minimizing inspection time, thereby improving productivity.
Conventional Radiography and Computed Radiography (CR)
Computed radiography uses a reusable imaging plate in place of the film. This plate employs a
coating of photostimulable storage phosphors to capture images.
When compared to conventional RT several advantages can be achieved. Efficiency is one of the
aspects to take in consideration when use CR systems since scanning time of the image plates are
less than the development of traditional film and exposure time is drastically reduced, typically 50%
less than with conventional films. The wide exposure latitude of the Digital Imaging Plates allows, in
many cases, the visualization of all information with only one exposure. In this way, the use of Digital
Imaging Plates results in a substantial reduction of the dose load [36].
39
4.5 Mechanized Welding ST Samples
Welding the ST samples, for mechanical testing, was exactly as described in the previous chapter
only differing in the parent material and acceptance criterion.
While in the previous task the objective was to consolidate the welding procedure by discovering
any imperfection and solving it, as well as any irregular bead morphology, now the objective is to
manufacture the most flawless welds possible according to the standard ISO 5817 – 2003 and then
submit then to various mechanical tests.
In order to check the quality of the welded joints both ultrasonic testing – phased array and
radiography testing was used. After the weld quality was confirmed the samples were machined into
adequate test specimens and subjected to mechanical tests such as tensile and Charpy V-notch
impact testing.
A tensile test, also known as tension test, is probably the most fundamental type of mechanical
test that can perform on material. The results from the test are commonly used to select a material for
an application, for quality control, and to predict how a material will react under other types of forces.
The major parameters that describe the stress-strain curve obtained during the tension test are
the ultimate tensile strength (UTS) or, more simply, the tensile strength, is the maximum engineering
stress level reached in a tension test, in other words is the strength of a material is its ability to
withstand external forces without breaking; yield strength or yield point (σy), characterizes a point
where beyond the material will have a plastic behavior and cannot withstand external forces without
deform permanently; elastic modulus (E); percent elongation (A%) and the reduction in area (Z%).
Toughness, Resilience, Poisson’s ratio can also be found by the use of this testing technique.
The Charpy test is most commonly used to evaluate the relative toughness or impact toughness
of materials. Impact tests are designed to measure the resistance to failure of a material to a suddenly
applied force. The test measures the impact energy, or the energy absorbed prior to fracture.
When the striker impacts the specimen, the specimen will absorb energy until it yields. At this
point, the specimen will begin to undergo plastic deformation at the notch. The test specimen
continues to absorb energy and work hardens at the plastic zone at the notch, when the specimen can
absorb no more energy, fracture occurs. Tough materials absorb a lot of energy, whilst brittle materials
tend to absorb very little energy prior to fracture.
Test specimens used measure 55x10x10mm and have a V notch machined across one of the
larger faces. V-shaped notch is 2mm deep, with 45° angle and 0.25mm radius along the base.
40
5 Results and analysis
The following results and their analysis were all obtained from welding, mechanical, metallographic
and NDT tests performed in the scope of the project SAFETOWER by ISQ.
5.1 Preliminary Specification of Welding Procedures
Preliminary specification of welding procedures was carried out manually by experienced
technicians so that a quick welding parameters setup could be reached for all three welding positions
(flat (PA), horizontal (PC) and vertical-up (PF)) and evaluate the capability of the selected wires for
future mechanization.
Figure 47 - A test piece being welded during semi-automatic welding tests.
The parameters to be studied and optimized were:
 Wire feed speed

Current

Arc voltage

Electrode extension - Stickout

Travel speed

Torch angle

Weaving frequency and trajectory
5.1.1 Manual Welding tests
The tests begun with the self-shielded wires with and without ceramic backing, the initial welding
parameters used, optimum technique and special wire care were provided by manufacturers technical
sheets from Lincoln Electric Innershield Electrodes Welding Guide [37] and ESAB Cored Wire
Handbook [38].
41
Weaving Technique
Flat position
Gapped root passes are made with a small, back-and-forth weave pattern. For fill and cover passes
the same weave, with an adjustment for the desired width, is used with adequate pause at the
sidewalls to obtain the necessary fill in these areas.
Vertical position
For a beveled, multipass joint a ”U” pattern is used for the root. The fill and cover passes are
made using a side-to-side weave with a backstep at the walls. The length of the backstep is on the
order of a wire diameter.
Figure 48 - Test piece geometry
Figure 49 - Weaving technique for PF
position. [39]
In order to select the adequate wire and optimize welding parameters three characteristics are of
utmost importance. The welded joint must have proper bead height and width, adequate penetration
and the highest possible deposition rate.
Welding trials
Welding trials were initiated using manufacturers parameters as a start point for the optimization,
after satisfactory visual assessment of weld beads the optimization task begun. The optimization goal
was to maintain or improve weld quality while increasing parameters in order to obtain a higher deposit
rate and consequently higher productivity.
This process suffered some setbacks as the physical limits of the molten metal and filler wire
were reached. Some examples are shown below.
Figure 50 - Example of trial and error with NR233.
42
Figure 51 - NR233 Root morphology
Figure 52 - NR233 Penetration on root pass
This type of wires has a deep penetration as it can be seen on the left image. On the right image
the standard morphology of an open back root weld.
5.1.2 Welding defects encountered while optimizing and solutions
While manual assessing self-shielded wire and optimizing parameter some visual imperfections
were encountered and such as:
Convex Bead
A convex or “ropy” bead indicates that the settings being used are too cold for the thickness of
the material being welded, in other words, there is insufficient heat in the weld to enable it to penetrate
into the base metal.
Solution

Increase voltage (within wire specifications) – this allows a more disperse electrical arc
improving the dispersion of molten metal instead of concentrating it on the middle.

Decrease CTWD (Contact Tip to Work Distance) – decreasing CTWD by Joule effect will
increase the current which will also improve the dispersion as higher temperatures will be
reached.
Figure 53 - Ropey convex bead
43
Concave bead
Solution

Decrease CTWD

Increase WFS – by increasing WFS an increase in current is also happens.

Decrease voltage

Decrease travel speed – decreasing travel speed allows a better filling of the joint.

Decrease drag angle – with a smaller drag angle smaller penetration will be achieved while
using higher currents and avoiding burn-through moreover is easier to fill the joint.
Figure 54 - Concave bead
Hot Cracking
Hot cracks are those that occur while the weld bead is between the liquidus (melting) and solidus
(solidifying) temperatures.
Any combination of the joint design, welding conditions and welding techniques that results in a
weld bead with an excessively concave surface can promote cracking. The major reason for this
defect is the incorrect technique for ending the weld. To properly end a weld, the crater should be
filled. This is done by reversing the arc travel direction before breaking the arc, if the welding control is
designed to supply gas for a short time after the arc is broken, the crater should be shielded until it is
completely solidified.
Hot cracking has been identified both in PA and PF welding positions mostly in the root or fill
passes, no special attention was required as the following pass will melt a significant portion of the
root pass and eliminate that imperfection. If it was located in the cap pass a small welding speed delay
in the end of the weld bead, as suggested above, should be enough to avoid the crater or the crack.
Figure 55 - Hot Cracking
44
Worm tracking or gas tracking
Are marks on the surface of the weld bead that are caused from the gas that is created by the flux
in the core of the wire. Stiffer slag create wile welding may inhibit outgassing and promote worm
tracking. The gas causes worm tracking when there is excessive voltage, for a given wire feed
setting/amperage, creating a longer arc which essentially needs more shielding and slower cooling
rates [39,40,41,42].
Figure 56 - Standard Front Weld Bead
Figure 57 - Worm tracking
Weld bead unsatisfactory geometry
Two main characteristics of the weld bead are the bead height and width, these characteristics
are important to assure that the weld joint is properly filled, with a minimum of defects, particularly in
multi-pass weldments.
Welding current and travel speed are the welding parameters primarily used to control weld bead
size. If the bead height is too great, it becomes very difficult to make subsequent weld passes that will
have good fusion, poor lateral fusion may occur with more peaked and narrow weld beads.
Figure 58 - Weld bead dimensions. [39]
Arc voltage and weaving pattern/frequency is used to control the shape of the weld bead, as the
arc voltage (arc length) increases, the bead height decreases and bead width increases. This increase
implies a flatter bead height the weld metal is said to ”wet” the base materials more efficiently and
fusion to the base plate is improved.
Insufficient penetration or burnthrough
Weld penetration is the distance that the fusion line extends below the surface of the material
being welded.
The primary factor that has an effect on penetration is the welding current, an increase (or
decrease) of the current will have same effect on penetration as seen by Equation 7. Penetration can
be controlled not only by varying welding current via wire feed speed but also through the variation of
the tip-to-work distance, however the effect of tip-to-work distance on weld penetration is opposite in
nature to that of welding current. By increasing the tip-to-work distance, welding current and
penetration will decrease and of course, the converse is also true.
45
Preventing burnthrough when there are discontinuities in material thicknesses or joint gap is an
advantage in controlling the tip-to-work distance while welding. The remaining factors have
comparatively little effect on penetration and do not provide a good means of control.
5.1.3 Manual Welding Pre-Qualification and deposit rates.
After all possible imperfections were solved a robust welding procedure was achieved. The
following figures demonstrate the two best welds achieved in the optimization process of the ten
chosen wires for test.
Visual assessment of all weld trial samples was executed and evaluated according to ISO 5817.
Figure 59 - Example of a semi-automatic
welding joint using a self-shielded wire (NR203Ni1): root pass performed on open root,
with no backing. Left: front side of welded
joint; right: back side of welded joint.
Figure 60 - Example of a semi-automatic welding
joint using a gas-shielded wire (PZ6113S): root
pass on performed on ceramic backing. Left: front
side of welded joint; right: back side of welded
joint.
Figure 61 - General view of test pieces produced throughout the course of manual semiautomatic welding tests of each wire chosen.
46
Deposition rate
The deposition rate describes how much usable weld metal will be deposited in one hour of
actual arc-on time
The current to achieve a given deposition rate can also be varied by changing the tip-to-work
distance. As Figure 62 shows, the wire feed speed can be increased by increasing tip-to-work distance
to maintain a constant welding current, this results in a higher deposition rate than usually associated
with a given current level.
Figure 62 - Effect of Tip-To-Work distance, WFS and deposition rate on welding current. [39]
The most reliable way to calculate the deposition rate of a specific process and wire is to weigh
the body specimen, set the wire feed speed ( and correspondent current intensity) and weld for a
determined amount of time (simple weld deposition).
After welding the body specimen is weighted again and by knowing the increase of weight and
welding time the deposition rate can be calculated.
Executing several of these trials, recording the welding parameters above mentioned and
weighing one meter of wire from each consumable a mathematical approach can be reached do
calculate the deposition rate [39,43].
Knowing the weight per meter of wire, the wire efficiency and the wire feed speed it is easy to
calculate the deposition rate thru the formula below.
(12)
Where Deposition Rate (Kg/h), WFS= Wire Feed Speed (m/min)), WPM = weight per meter of
wire (g/m), and WE = Wire deposition efficiency (for these wires it is around 0.86), the value of 0.06 is
a conversion constant from gr/min to kg/hour.
Using the Equation 12 and by weighing all the tested wires the following results were obtained.
Figure 63 - Wires used in weighing
47
Figure 64 - Deposition rates for one-side butt welding, 12-mm thick plate, vertical-up (PF)
position after welding process optimization.
Welding joints, sequences and parameters
Wire feed
speed
Voltage
[m/min]
1
2, 3
Pass no.
Current
(typical)
Weld
speed
Heat
input
Deposition
rate
[V]
[A]
[cm/min]
[kJ/mm]
[kg/h]
6.0
23
160
11
2.0
2.1
8.0
25
190
18
1.7
2.8
Wire feed
speed
Voltage
Current
(typical)
Weld
speed
Heat
input
Deposition
rate
[m/min]
[V]
[A]
[cm/min]
[kJ/mm]
[kg/h]
1
6.0
23
160
10
2.2
2.1
2-8
9.0
27
210
17
2.0
3.2
Pass no.
Figure 65 - Welding joints, sequences and parameters for vertical-up (PF) position.
48
Wire feed
speed
Voltage
[m/min]
1
2, 3
Pass no.
Current
(typical)
Weld
speed
Heat
input
Deposition
rate
[V]
[A]
[cm/min]
[kJ/mm]
[kg/h]
8.0
25
190
7
1.6
2.8
12.0
29
260
26
1.6
4.3
Wire feed
speed
Voltage
Current
(typical)
Weld
speed
Heat
input
Deposition
rate
[m/min]
[V]
[A]
[cm/min]
[kJ/mm]
[kg/h]
1
8.0
25
190
7
1.5
2.8
2-8
13.0
30
275
22
2.0
4.6
Pass no.
Figure 66 - Welding joints, sequences and parameters for flat (PA) position.
Wire feed
speed
Voltage
[m/min]
1
2-4
Pass no.
Current
(typical)
Weld
speed
Heat
input
Deposition
rate
[V]
[A]
[cm/min]
[kJ/mm]
[kg/h]
8.0
25
190
16
1.8
2.8
8.0
25
190
22
1.3
2.8
Wire feed
speed
Voltage
Current
(typical)
Weld
speed
Heat
input
Deposition
rate
[m/min]
[V]
[A]
[cm/min]
[kJ/mm]
[kg/h]
1
8.0
25
190
20
1.4
2.8
2-16
8.0
25
190
27
1.1
2.8
Pass no.
Figure 67 - Welding joints, sequences and parameters for horizontal (PC) position.
For the PC position it was found that in order to obtain a sound root with good bead morphology
the ceramic backing must be uncentered with the joint. The lower edge of the CB should be coincident
with the lower joint edge as seen in Figure 67.
49
5.1.4 Analysis of results
Wires tested
Self-shielded wires seemed to have good potential for application in the scope of this study.
These wires, from standard self-shielded filler wire ranges, were valid choices if chemical composition
as well as mechanical properties of the weld metal are to be considered, these consumables are allposition self-shielded wire for structural welding of mild and some alloy steels but also suitable for
single and multi-pass welding and with an expected good performance in both semiautomatic and
mechanized applications.
From all self-shielded wires tested ESAB Coreshield 8 allowed an easier welding technique with
improved control over deposit and penetration, a much quieter welding arc, producing lower spatter
levels and a better bead appearance.
NR-Offshore is possibly among the very few self-shielded wires capable of matching properties of
S460M steel. It was initially expected that this wire could fully match and surpass all requirements
resulting from the possible choice of S460M as the structural steel for the tower however this wire
wasn’t even near the results achieved by Coreshield 8.
All gas-shielded filler wires that have been tested are 1.2mm diameter all-position rutile wires
which differ among them mainly on nickel content and corresponding toughness properties of the weld
metal.
The gas-protected rutile flux-cored wires tested, PZ6113 (with Ar/CO2), PZ6113S, PZ6114S,
PZ6116S and PZ6138, all match entirely the properties of S460M steel. Rutile types yield somewhat
higher values, generally between 3 and 4 ml / 100 g, thus still well within class H5.
Self-shielded FCAW wires did not match mechanical properties of S460M steel, and have poor
behavior on ceramic backings, even with quite low process settings, which means that it would be very
difficult to mechanize one-side welding root passes in butt joints during production welding.
Surprisingly when welding open root passes with self-shielded wires was when the best deposition
rates were obtained.
On fill and cap passes self-shielded wires behavior improved but still revealed very harsh and
difficult to control welding arc which demanded quick responsive actions of a highly skilled manual
welder therefore not being compatible with all-position mechanized welding.
As for gas-shielded wires the performance and welding parameters achieved were very similar
amongst all of them. Unlike self-shielded, these wires have a remarkable all-position performance and
behavior on ceramic backings.
It should be noted that all root passes with gas-shielded wires were performed on ceramic
backings since the welding of open roots is not recommended by manufacturers for this type of wires.
Fill and cap passes with gas-shielded wires were much more “civilized” being the perfect candidate for
mechanization.
Not only gas-shielded wires are capable of a much better overall weldability but also have
deposition rates notoriously higher than their self-shielded counterparts. An additional fact is related to
the operational limits of these wires, self-shielded were pushed to their limits while gas-shielded could
withstand higher process settings.
50
It is known that as the cross sectional area of a conductor decreases, the resistance to current
flow increases as can be seen in Equation 1.
This resistance to current flow will cause considerable heating of the conductor, this
phenomenon is show by Equation 3, if the current is relatively high and the conductor is small in cross
sectional area. In other words, at a given current in amperes, the current density within the conductor
will increase as the diameter of the conductor is reduced.
The current density is considerably higher in the small diameter flux cored wire and therefore the
deposition rate will also be somewhat higher.
It is this high current density that makes flux cored wires the success they are. The high
resistance heating of the wire is confined to a small area, and the electrode reaches its melting point
very quickly, producing a concentrated deep penetrating arc, the efficiency and deposition rate are
also very high.
This resistance factor is the key to understand the performance differences between self-shielded
and gas-shielded wires, the fabrication process and the type and amount of flux required for a selfshielded implies a slightly larger diameter that makes all the difference.
For example the smallest self-shielded (SS) wire tested was 1.6mm and gas-shielded (GS) was
1.2mm, the resistive area of SS is 77% bigger than GS and even increasing the stickout length (L) to
its limit it was proven impossible to compensate the resistance difference, as it can be seen in Figure
64 were the best SS wire has a deposition rate of about 75% of any GS.
To match GS wires performance SS wires must increase its welding current however the only
parameter which increases the current besides CTWD (which is limited) is WFS that also has a limit.
More current implies hotter molten metal and more wire fed, in these conditions it is impossible to the
molten metal solidify quickly enough to sustain the following metal deposited by the weaving motion as
described in Figure 49.
A simple math exercise proves the limitation of the stickout length relative to the diameter of the
wire: Welding machines work with constant power, using Equation 6 any change in the resistance will
affect directly the welding current, and only the wire diameter and stickout can influence the
resistance.
Using
,
, P= 1500W, Dconst =1.5mm and Lconst =22mm the ratio between increments
in the wire diameter or stickout length can be calculate since:
√
(ρ can be discarded taking into account the objective of the exercise).
51
Stickout
Variation L
Current (I)
Variation I
1
0
51
0%
1,1
+10%
49
-5%
1,2
+10%
47
-4%
1,3
+10%
45
-4%
1,4
+10%
44
-4%
1,5
+10%
42
-3%
Diameter
Variation D
Current (I)
Variation I
1
0
7
0%
1,1
+10%
8
10%
1,2
+10%
9
9%
1,3
+10%
10
8%
1,4
+10%
10
8%
1,5
+10%
11
7%
Table 5 - Current, stickout and diameter
variations.
12%
Stickout Var.
10%
Diameter Var.
8%
6%
4%
2%
0%
1,1
1,2A or L variation
1,3
1,4
1,5
Figure 68 - Variation ratio between stickout and
wire diameter.
As stated previously the ability of the stickout modifying the current values is about half the
variation obtained with diameter variations.
Parameter optimization
Weld bead appearance is mainly controlled by welding current and travel speed, with a decrease
of current the wire melting rate (Equation 8) will also decrease as it is governed by the current and in a
smaller importance CTWD. With this reduction in current and in the amount of melted material if travel
speed is maintained the weld bead will be smaller. The opposite is also true.
The other parameter which can be used to control the weld bead is the travel speed, in opposition
to the current, a decrease in travel speed will increase the deposited metal. Since the melting rate is
independent of travel speed the amount of filler metal deposited in a linear meter of weld is increased
producing a fuller bead.
The CTWD, in a limited extent, can also affect weld bead characteristics. Returning to Equation 8
it can be seen that while CTWD has no exponent the current is raised to the 2nd power giving a
greater role in geometry control. When long extensions are used to increase deposition rates, bead
height will increase to a greater extent than bead width.
In the case of the convex bead the solution is to disperse the molten material by the joint, this is
achieved by increasing the welding voltage and decreasing CTWD due to Joule effect as explained in
chapter 4.2.1. For the case of concave bead the problem isn’t a less fluid and undispersed molten
material but an insufficient and badly distributed filling of the joint. The solution is to decreasing CTWD
or increase WFS to maintain the molten material fluid and decrease voltage, which will reduce the
lateral flow cause by the excessive electric arc. Travel speed and drag angle should be adequate to
enable a more complete filling of the joint as more material is deposited.
The effect on weld penetration of arc travel speed is similar to that of welding voltage –
penetration is a maximum at a certain optimal value and decreases as the arc travel speed is varied.
52
With lower welding speeds, too much metal is deposited in an area and the molten metal tends to roll
in front of the arc and ”cushions” the base plate, this prevents further penetration. On the other hand at
high welding speeds, the heat generated by the arc hasn’t sufficient time to substantially melt the area
of base material.
The welding current can also be used to control penetration thru CTWD. The resistance heating
2
of the wire (the 2nd Amp term in the Equation 8) is a very efficient heating process. Therefore the
current needed to finish melting the wire as it enters the arc, becomes less as the wire is hotter with
longer stickout resulting in less penetration.
It is very important to keep the torch stickout constant, small wire length changes affect the
welding current which is raised to the 4th power and has great weight in weld penetration as seen on
Equation 7. Also the shorter the distance from tip to work for a fixed wire feed speed the greater the
penetration since current also increases.
Welding voltage or arc travel speed have a slightly smaller effect than torch position does on
welding penetration. Changing the longitudinal torch angle, and using the pulling or pushing method
can help control the penetration. In the pull technique the torch is dragged backward across the weld
joint and gives a bit more penetration and a narrower bead for deeper weld joints as the wire is directly
aimed at the gap. On the other side the push technique pushes the torch forward into the weld, this
technique will give a bit less penetration for weld joints that are shallow and produces a wider bead.
● – No considerable effect
 - Little effect
 - Increase
 - Decrease
Figure 69 - Parameters optimization
summary.
Shielding gas
Pure carbon dioxide is not an inert gas because the heat of the arc breaks down the CO2 into
carbon monoxide and free oxygen. Under the heat of a welding arc, these active gases react with
alloys in the molten weld metal, such as manganese and silicon, leading to the loss of these elements
in the solidified weld.
Because of these types of reactions, failure to follow the welding wire manufacturer's guidelines
regarding what shielding gas can be used with each formulation can cause unexpected results with
chemical analysis, tensile strengths, impact strengths, and crack resistance. The use of CO2 alone as
53
welding gas may be slightly detrimental for yield and tensile strength of weld metal, but sound welds
can be consistently and easily achieved which are free of porosity and defects.
The use of pure CO2 as welding gas, instead of 20% CO2 argon-based mixtures, results normally
in lower values of hydrogen as CO2 react at these high welding temperatures producing a hotter
puddle than truly inert atmospheres. The thermal conductivity of the gas at arc temperatures
influences the arc voltage as well as the thermal energy delivered to the weld. As thermal conductivity
increases, greater welding voltage is necessary to sustain the arc. For example, the thermal
conductivity of helium and CO2 is much higher than that of argon, because of this they deliver more
heat to the weld. Therefore, helium and CO2 require more welding voltage and power to maintain a
stable arc.
As result the use of CO2 improves the molten puddle flow characteristics and hydrogen diffusion,
since the elevated solubility of hydrogen allows hydrogen to diffuse out of the metal while this is at
elevated temperatures.
The advantage of CO2 is fast welding speeds, deep penetration, common availability and quality
weld performance as well as its low cost and simple installation.
The major drawbacks of CO2, are a harsh globular transfer and high weld spatter levels, also the
weld surface resulting from pure CO2 shielding is usually heavily oxidized. A welding wire having
higher amounts of deoxidizing elements is sometimes needed to compensate for the reactive nature of
the gas.
5.1.5 Selected Filler Wire
In spite of the need to provide protection from wind in the immediate vicinity of the welding area,
gas-shielded wires are strongly recommended for mechanized FCAW applications within the scope of
this study.
For S460M steel, all gas-shielded wires but PZ6113 could be used. The final choice of FCAW
filler material fell on PZ6113S due to lower cost related to welding (gas type and wire price) and very
similar properties with the remaining PZ wires.
In this chapter several parameter combinations were tried in order to achieve the optimum
welding parameters for each wire. Highest deposition rate and therefore higher productivity with less
or none welding imperfections and adequate bead morphology were taken in account when choosing
the most suitable wire for this work. In the next chapter, microstructural changes due to high
temperature permanence of the S460M steel will be studied.
5.2 S460M Weldability Test
Fourteen specimens were created for this trial using two ways to obtain the desired HI. The first
approach was combining the WFS (using the synergic welding mode implies the machine will select
54
the correct voltage and amperage based on the wire feed speed set by the operator) and welding
speed, this method was used for test specimens from A to G. The second approach consisted on
maintaining the WFS value constant and varying only the WS. Specimens H to N were executed with
this method.
Specimens identified as A, B, C, D, E, G, H, I, M and N were selected to be tested.
Figure 70 - Test specimens obtained
Heat Input calculation
To calculate the heat input for arc welding procedures, the following formula was used:
(13)
Where Q = heat input (KJ/mm), V = voltage (V), I = current (A), and S = welding speed (mm/min)
The efficiency is dependent on the welding process used, for FCAW is 0.86.
Spec.
A
B
C
D
E
F
G
H
I
J
K
L
M
N
HI(KJ/mm)
HI real
(KJ/mm)
0,5
1
1,5
2
2,5
3
3,5
0,5
1
1,5
2
2,5
3
3,5
0,51
1,05
1,5
2,17
2,58
3,04
3,11
0,49
1,03
1,5
2
2,52
3
3,82
Table 6 - Heat input test table
Detailed Heat Input data can be seen on annex 9.2
Microstructure characterization after HI test.
All specimens presented acicular ferrite and polygonal ferrite in the weld material however in the
low HI specimens, 0.5KJ/mm and 1KJ/mm, some formations of martensite, bainite were also detected.
In the unaffected parent material all specimens presented ferritic and perlitic structures.
Microstructural differences were best seen in the heat affected zone as it was expected.
Specimens with lower HI (A, B, H, I) have martensitic and bainite structures in the grain growth
zone and in the refined grain as well. Traces of perlite and some aggregated carbides are present in
the subcritical region. In the specimens with higher HI than 2.5KJ the grain growth zone has ferrite
with aligned M-A-C and ferrite carbide aggregates while in the subcritical region ferrite and
spheroidized pearlite are present.
Micrographs can be seen on annex 9.3
55
Zone
Weld Material
HAZ
Parent Material
0.5
KJ/mm
Ferrite + Martensite
Martensite + Bainite
Ferrite + Perlite
1.0
KJ/mm
Acicular and Polygonal Ferrite
Martensite + Bainite
Ferrite + Perlite
Acicular and Polygonal Ferrite
Ferrite Aligned M-A-C
Ferrite + Perlite
2.5
KJ/mm
Table 7 - Microstructures.
Weld Dilution
Total area of
Penetrations of weld
weld metal
material (mm²)
(mm ²)
A
18,803
5,73
B
45,909
11,258
C
51,002
10,984
D
69,237
10,923
E
78,27
12,567
G
93,435
17,239
H
23,392
9,787
I
38,375
11,733
M
82,395
7,508
N
97,269
11,259
Table 8 - Weld Dilution
Weld
dilution
(%)
30,47
24,52
21,54
15,78
16,06
18,45
41,84
30,57
9,11
11,58
Spec.
Figure 71 - Weld diluition areas
Hardness Test
400
HV10
Teste Piece A
Teste Piece B
Teste Piece C
Test Piece D
Test Piece E
Test Piece G
Test Piece H
Test Piece I
Test Piece M
Test Piece N
350
300
250
200
150
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17Indentation
18
Figure 72 - Hardness Test Results
56
Indentations number 1,2,3,4 and 5 were taken in the Weld Metal Zone
Indentations number 6,7,8,9 and 10 were taken in the Heat-affected Zone (HAZ)
Indentations number 11,12,13,14,15,16,17 and 18 were taken in the Parent Material
For hardness detailed data please seen annex 9.4
HV10
400
350
300
250
200
0,49
0,51
1,03
1,05
1,5
2,17
2,58
3
3,11
3,82
Heat Input KJ/mm
Figure 73 - Maximum Hardness for each heat input value tested
Welding Cooling Rates
The most widely used and the best known analytical solutions to predict weld thermal history and
cooling rate are those of Rosenthal. To enable solving Equation 14 the following simplifications are
assumed:

The heat source is concentrated in one point and has infinite temperature

The physical properties of the metal are independent of temperature

No heat exchange occurs between the plate and environment

The plate is flat and has large dimensions

Latent heat of fusion is ignored
(
(14)
The solutions, for the 2D and 3D heat flow, give the temperature variation during cooling as a
function of time for a given location, the peak temperature (TP) as a function of distance from the heat
source, and the weld time constant (Δt8–5), which is the cooling time from 800° to 500°C.
2D
(
(
(
3D
)
)
(
(
(
(
)
)
(
(
(
(
(
(
)
(
(
(
)
(
(23)
57
Symbol Definition and unit
Value
T
Temperature, ºC
Tp
Peak Temperature, ºC
t
Time, s
Δt8-5
Cooling time from 800º to 500ºC, s
r
Radial/lateral distance from weld, m
T0
Initial temperature, ºC
22
λ
Thermal conductivity, Js-1m-1ºC-1
41
a
Thermal diffusivity, m2/s
9.1x10^6
ρc
Specific heat per unit volumeJm-3ºC-1 4.5x10^6
d
Plate thickness, m
0.003
HI
Heat input, J/m
Table 9 - Cooling rate variables and values. [Properties of Structural steels and effects of
steelmaking and fabrication , McGraw Hill]
There is also an equation to determine a critical thickness for a given heat input at which the 2-D
condition changes to 3-D. Equation 24 calculates the critical plate thickness, dc, which the crossover
between the 2-D and 3-D conditions of heat flow takes place, moreover cooling rate is considered
independent of the distance from the heat source, at least in the HAZ.
HI (KJ/mm)
Dc value (mm)
0,50
6,69
1,00
9,47
1,50
11,60
2,00
13,39
2,50
14,97
3,00
16,40
Figure 74 - 2D and 3D flow.[39]
3,80
18,46
Table 10 - Dc value for each HI.
[
(
)]
(24)
Real welds are more likely to lie between the two limiting solutions, a situation classified by some
researchers as 2.5-D for which there is no simple solution [43].
Based on the assumption that the actual situation lies between the two limiting solutions, the
actual HAZ width can be related to the values below [39].
2D
3D
(
(
) (
(
(
(
) (
(
(
(
) (
(
(
)
(
The following equations can be used to find the critical temperatures A1 and Tm (melting point,
which is an upper-bound value for Ts), in case this information is not readily available. The alloying
additions are in wt-% and the temperatures are calculated in K [44].
(29)
(30)
58
Temperature
(ºC)
1000
0.5KJ/m
m
1KJ/mm
800
600
1.5KJ/m
m
400
200
1
7
13
19
25
31
37
43
49
55
61
67
73
79
85
91
97
103
109
115
121
127
133
139
145
0
Time (s)
Figure 75 - Steel welding temperature over time for different HI @ r=3mm.
HI
(KJ/mm)
HAZ Width (mm)
Measured
3D
2D
(≈)
0,5
1,60
1,88
1,68
1
2,26
2,83
2,34
1,5
2,76
5,62
3,25
2
3,19
7,50
4,42
2,5
3,57
9,60
5,35
3
3,91
11,28
5,97
HAZ Width
(mm)
16,00
14,00
12,00
10,00
8,00
6,00
4,00
2,00
0,00
Thick Plate
Thin Plate
Measured
0,5
2,5
3
3,8
Heat Input (KJ/mm)
Figure 76 - HAZ width of each HI.
3,8
4,40
14,29
6,78
Table 10 - HAZ Width calculations
and measurements for different HI.
HI
(KJ/mm)
Cooling
time (s)
Cooling
rate
(ºC/s)
Peak
Temperature
(r=3mm)ºC
0,50
1,00
1,50
2,00
2,50
3,00
3,80
1,57
3,13
4,70
6,26
7,83
9,39
11,90
191,60
95,80
63,87
47,90
38,32
31,93
25,21
2913
5805
8696
11588
14480
17371
21998
Table 11 - Cooling times, rates and peak
temperature @ r=3mm for different HI.
1
1,5
2
Max.
Hardness
400
350
y = 364,35e-0,062x
R² = 0,9715
300
250
200
1,57
3,13
4,70 6,26 7,83 9,39 11,90
Cooling times T8-5
Figure 77 - Max. hardness relation with cooling
times.
5.2.1 Analysis of results
Microstructures
All the samples analyzed had the same parent material microstructural constitution, ferrite and
perlite.
59
In the weld metal the predominant structures are polygonal ferrite and acicular ferrite. Acicular
ferrite is comprised of intragranularly nucleated Widmanstatten ferrite which consists of small laths of
ferrite with low aspect ratio which occur in several distinct orientations giving the appearance of a fine
grain size interlocking microstructure, this microstructure is usually associated with excellent
toughness as it provides maximum resistance to crack propagation by cleavage. Acicular ferrite is also
characterized by high angle boundaries between the ferrite grains. This further reduces the chance of
cleavage, because these boundaries impede crack propagation. Composition control of weld metal is
often performed to maximize the volume fraction of acicular ferrite due to the toughness it imparts
hence welding consumables employ sophisticated alloying techniques, incorporating the optimum
balance of deoxidizing elements (aluminum, silicon and manganese) to produce a high density of
small non-metallic inclusions which are known to act as intragranular nucleation sites for acicular
ferrite.
Polygonal ferrite can nucleate both at inside austenite grain boundaries and in intragranular
regions, its formation is therefore favored in high heat input welds and decreases with the increase in
carbon and chromium content. Large amounts of grain boundary polygonal ferrite are not generally
considered beneficial for toughness especially in higher strength steels. Lower strength and an
increase in ductility can be obtained with slower cooling rates which lead to larger volume fraction of
less dislocated polygonal ferrite structure.
The HAZ zone is subdivided in two main regions, the grain growth and the subcritical region, the
grain growth presented with an interesting change in its microstructure with evidence of bainite and
martensite which is a metastable structure consisting of a supersaturated solid solution of carbon and
alpha ferrite created when C-Mn steel is cooled rapidly from the austenite phase into a new solid
phase. Martensite can be found in common C-Mn steels welded at low heat input and when cooling is
very quick, the toughness is generally very poor, the strength very high and can give rise to HAZ
hydrogen cracking.
Not surprisingly in the subcritical region (subcritical is between the intercritical and unaffected
base metal) the only change from the parent material was the existence of spheroidized pearlite
together with ferrite and ferrite with aggregated carbides. In pearlitic structures, lamellas of soft ferrite
alternate with lamellas of hard cementite. Under the influence of shearing stresses, plastic deformation
occurs essentially only in the soft ferrite lamellas where dislocations can move relatively easily. The
thinner the ferrite lamellas are the more restricted, by the rigid cementite lamellas, is the mobility of the
dislocations. This explains why the ductility of pearlite increases and its hardness correspondingly
decreases with the thickness of its ferrite lamellas. Spheroidite is the most ductile variety of pearlite
because, in its coherent ferrite matrix, dislocations can move more freely, restricted only to a lesser
degree by the spherical cementite inclusions. A tempering heat treatment of the base metal occurs in
this region, however, the pearlite does not completely spheroidize since the weld thermal cycle is too
short for this to happen.
All samples tested with HI higher than 2.5KL/mm have ferrite with aligned M-A-C (martensite,
retained austenite, or carbide) and ferrite carbide aggregates in the HAZ grain growth region. This is
generally the predominant microstructural constituent in C-Mn steels, occurring over a wide range of
60
heat inputs moreover ferrite with M-A-C can have an aligned AC or non-aligned FN appearance, but
this variation is most likely due to a sectioning effect.
Retained austenite does not always have a beneficial effect, martensite-austenite constituents
have a strong embrittling effect. However thin films of interlath retained austenite in martensite
improve the fracture toughness and even resistance to hydrogen assisted cracking probably due to
the instability of this austenite upon deformation.
The existence of carbides can be controlled by increasing the cooling rate which minimizes the
diffusion of carbon and hence carbide segregation. The nature and the size of the carbides have also
an influence on irradiation embrittlement since the growth of carbides increases the embrittlement.
Cooling rates
The microstructural changes between different heat input samples can be explained by the
cooling time. A high heat input will result in a slow cooling rate. The temperature-time cycles during
welding have a significant effect on the mechanical properties of a welded joint and the cooling time
from 800°C to 500°C (t8/5) is a good choice to characterize the cooling conditions of an individual
weld pass for the weld metal and the corresponding heat affected zone (HAZ).
As can be seen from the diagram below for a carbon steel, a long cooling time over the
temperature range of 800 to 500º C, results in a predominantly ferrite and pearlite microstructure.
A low heat input will result in a high cooling rate and fast cooling time over the temperature range
of 800 to 500º C, results in a predominantly martensite and equal quantities of bainite [32,46,47].
Martensite is rare in weldments with 0.1 to 0.25% C and 1.0 to 2.0% Mn carbon steel when
welded with suitable fillers, a cooling time of less than 1 second over the temperature range of 800 to
500º C would be necessary to form martensite. For low carbon steels, martensite formation would
have to be artificially enhanced by the use of a very low heat input [ 48,49,50].
Figure 78 - C-Mn Steel Weld CCT Diagram of S460M. [51]
The application of Rosenthal analytical solutions for this kind of experimental work is always a
rough approximation to real welding conditions, however with the necessary precautions when
analyzing the data obtained some conclusions can be associated with the real case.
Assuming heat transfer behavior is 2-D above the critical value of Table 9 is not reasonable, for
the 2-D situation to prevail, the weld metal zone should span over the whole width of a thin plate, so
61
that the heat transfer occurs only within the plane of the plate, which is not the case even for the
lowest heat input used here.
In Figure 75 the temperature evolution was calculated for the complete range of HI done at 3mm
of the heat source, main differences are noticeable between 0 and 86 seconds where, as expected,
lower values of HI will have a most pronounce slope than higher HI since less energy is delivered to
the piece and its dissipation is quicker.
The evidence of the approximation characteristic of this type of analytical solution can be seen
on Table 10 and Figure 76 although the actual measurement of the HAZ is very difficult to perform the
values obtained by the equations have a relatively large deviation. This gives greater weight to the
idea of 2.5D situation as mentioned above and becomes one more reason to be careful when
interpreting results obtained.
When the deviation of real versus calculated measurements was perceived HAZ width
measurements using 2D equations took place and plotted along real and 3D results, this allowed
understanding de dimension of the deviation. As can be seen the real values are much closer to the
3D thick plate results than to the 2D thin plate and although disparities exist the 3D approach is still
more accurate.
The last results obtained with this analytical analysis were the cooling times and rates and the
possibility to measure up with the microstructures obtained.
The predicted range of cooling times for a 0.5 kJ/mm weld sample was found to be around 1.6s,
this translates to mean cooling rates of approximately 192ºC/s. On the opposite for a heat input of
3.8KJ/mm the cooling time increased to 12s and the cooling rate was about 25°C/s.
In Figure 77 the maximum hardness versus cooling times were plotted and a trend lines was
added using an exponential equation, this trend line has a R2=0.9715 which is a fairly good accuracy
to the values plotted.
(
(
(31)
This equation can be used to back track from a hardness value desired up to the HI necessary to
achieve it.
With cooling times less than 4.7s there is a high probability to find martensite on the weld metal
and also on the HAZ, these hard microstructures are above de 300HV since the supersaturated solid
solution of carbon and alpha ferrite are not maintained between 800 and 500ºC have enough time to
promote the grain growth and carbon diffusion.
Test pieces C, D and M had cooling times between 5 and 9s, this higher cooling time translated in
the complete microstructural transformation of the weld metal in acicular and polygonal ferrite and long
enough cooling time to transform austenite in bainite, avoiding martensite, in the HAZ.
For the remaining test pieces, G and N, with cooling time, above 9s, allowed the attainment of
ferrite and spheroidized perlite in the subcritical zone very similar with the parent material whilst in the
grain growth zone was identified ferrite with aligned M-A-C.
62
Hardness Test
Hardness test has as main goal measuring the resistance to indentation of a material, this test
allows relating mechanical properties with metallurgical properties. All three most important areas
where tested, Parent Material (PM), Heat Affect Zone (HAZ) and Weld Metal (WM), with a load of
10kg.
Consulting the standard ISO TR 15608 – 2005 the steel S460M belongs to the group 2.1 as it is a
thermomechanically treated fine-grain steel with a specified minimum yield strength 360 /mm2 < ReH
2
≤460 /mm and consulting ISO 15614-1 2004 section 7.4.6 Table 2 the maximum permitted hardness
value (HV10) after welded for this type of joint and steel group cannot exceed 380HV10. [45,46]
Even with the lowest heat input the maximum HV value obtained was 363 therefore according to
the standards any of the welding parameters used are acceptable, however lower values of hardness
ensure a less brittle structure and HI higher than 1.5KJ/mm should be used to safeguard a more
ductile microstructure.
As expected test pieces with lowest HI such as A, B, E,I with microstructures rich in bainite and
martensite which has limited slip possibilities and a high yield strength, these test pieces also show
higher hardness and brittleness in WM up to refined grain growth than all the others test pieces. The
other test pieces with higher HI and slower cooling rates have small amount of polygonal ferrite and
acicular ferrite, characterized by needle shaped crystallites with chaotic ordering, in the WM which
confers a more ductile microstructure than martensite. Ferrite with aligned M-A-C was also found and
although this kind of microstructures is not beneficial due to the presence of martensite and carbides
which contribute to the embrittlement, hardness values were very alike to test pieces without this type
of microstructures which leads to think that there is only a small amount of M-A-C.
In Figure 72 the 7th point on specimen G is not coherent with the rest of the data, a reason may
be due to chemical inhomogeneity and segregation in the HAZ located on the point where the
indentation took place.
It was also found that for same values of heat input obtained by higher wire feed speed result in
slight lower values of hardness. This can be seen in Figure 73 where hardness values of 0.49 KJ/mm
and 1.03KJ/mm, respectively test pieces H and I welded at 9cm/min, were 20HV10 lower than the two
test pieces, A and B, welded at 6cm/min. One reason for this to happen is the cushion effect, at a
slower welding speed the arc force is damped by the extra weld metal deposited this translate to a
lower heat transmission to the piece as heat is partly dispersed by the weld bead itself, although the
heat input is equal. As seen before lower temperatures are accompanied by higher cooling rates which
lead to harder structures.
Microstructural changes due to high temperature permanence of the S460M steel and
corresponding hardness were identified for a wide range of heat inputs, the need for pre-heating was
also taken in account. The data gathered until now allows us to initiate the next chapter with the
knowledge of optimum welding parameters and steel weldability boundaries allowing full attention to
the task of welding mechanization and its issues.
63
5.3 Mechanized Welding Trials
With the production of several trial samples it become noticeable that even using a robust jig,
which grants precision and repeatability, constant monitoring of welding speed, stick-out and torch
alignment with the joint was required due to transverse weld shrinkage of the gap. Angular distortion
were only noticeable in 30mm thickness when welding the first side, after welding completion on both
sides the trial sample had no distortion.
One parameter of major importance is the angle of the torch in all passes but especially on root
and cap passes.
By controlling the torch angle a good penetration or fill can be obtained, bead morphology is also
an important characteristic of a sound welding bead. For root passes it was found that a downward
angle about 15º with horizontal plane provided the correct amount of penetration and good joint filling.
For cap passes the decision of what the most appropriate angle was difficult because not only the
welding speed and torch angle influenced the cap but also weaving frequency and amplitude.
Several tryouts combining torch angle, welding speed and weaving were made in order to
achieve the intended result.
Figure 79 - Cap parameters tryout upward and downward angle
Figure 80 - Cap parameters tryout 90º angle
64
The preferred angle for all pass besides root pass was an angle of 90º between torch and the
joint, with a 0.2 seconds dwelling right and left with a frequency of 1Hz, as for the width of the weaving
it was normally 2mm shorter than the gap of the joint in the actual pass.
Figure 81 - Mechanized root pass welded on ceramic backing. Left: front side of welded joint;
right: back side of welded joint immediately after removal of backing.
Figure 82 - Completed welding joint. Left: front side of welded joint; right: back side of welded
joint.
Figure 83 - On site UT testing
65
Figure 84 - Macrographs of two cross sections of perfect welded joint.
5.3.1 Imperfections detected in mechanized welding by NDT
Incomplete penetration
There are three ways in which incomplete penetration can occur, the first is when the weld bead
does not penetrate the entire thickness of the base plate, the second occurs when two opposing weld
beads do not interpenetrate and the third and final is when the weld bead does not penetrate the toe
of a fillet weld but only bridges across it.
Principal factor to affect penetration is low welding current, also too slow travel speed or incorrect
torch angle may promote incomplete penetration since molten weld metal to roll in front of the arc,
acting as a cushion to prevent penetration. The arc must be kept on the leading edge of the weld
puddle to ensure proper penetration.
Lack of fusion
Lack of fusion, occurs when there is no fusion between the weld metal and the surfaces of the
base plate. A common cause of lack of fusion is either the weld puddle is too large (travel speed too
slow) and/or the weld metal has been permitted to roll in front of the arc. To avoid this imperfection,
like incomplete penetration, the arc must be kept on the leading edge of the puddle.
Another cause is when the molten weld metal will only flow and cast against the side walls of the
base plate without melting them. The heat of the arc must be used to melt the base plate. This is
accomplished by making the joint narrower or by directing the arc towards the side wall of the base
plate. When multipass welding thick material, a split bead technique should be used whenever
possible after the root passes.
Undercutting
Undercutting is a defect that appears as a groove in the parent metal directly along the edges of
the weld. One of the causes can be a travel speed too high which will give a very peaked weld bead
due to its extremely fast solidification. The fast solidification makes the forces of surface tension draw
the molten metal along the edges of the weld bead and pile it up along the center.
The undercut groove is where melted base material has been drawn into the weld and not
allowed to wet back properly. Welding speed should be decreased leading to a gradual reduction of
the size of the undercut and eventually eliminate it.
66
Torch angle is also a corrective action producing a flatter weld bead and improve wetting.
Raising the arc voltage and maintaining the arc length short, not only will avoid undercutting but
will also increase penetration and weld soundness. However excessive welding currents can cause
undercutting since the arc force, arc heat and penetration are so great the base plate under the arc is
actually ”blown” away.
Porosity
Porosity are gas pores found in the solidified weld bead they can be found either under or on the
weld surface.
The common causes of porosity are excessively oxidized work piece surfaces, inadequate
deoxidizing alloys in the wire and the presence of foreign matter such as excessive lubricant on the
welding wire and atmosphere contamination.
Oxygen and entrapped moisture are products of excessive oxidation of the work pieces which
can lead to formation of pores while welding however this was not the cause of porosity detected since
the work pieces had a factory primer which was removed immediately before welding.
Bad choice of parameter such as extremely high travel speeds and low welding current levels
were avoided, since solidification rates are extremely rapid, trapping any gas that would normally
escape, puddle turbulence was avoided to the maximum by maintaining as constant as possible the
arc characteristics. This turbulence will tend to break up the shielding gas envelope and cause the
molten weld metal to be contaminated by the atmosphere.
Atmospheric contamination can be caused by inadequate shielding gas flow both in excess as in
shortage, severely clogged gas nozzle or damaged gas supply system (leaking hoses, fittings, etc.)
and an excessive wind in the welding area which can blow away the gas shield.
The welding took place indoor and there were no evidence of wind draughts strong enough to
disrupt gas protection more over the welding gun nozzle was cleaned after each pass.
The only causes remaining to explain the porosities are related to gas-shielding inadequate
shielding gas flow both in excess as in shortage or gas supply system leaking.
The gas supply system was checked and no leaks were found, for example the remaining CO2,
after a welding, which was in the supply line between the gas cylinder, thru hose, welding machine
and the gun nozzle would lose very little pressure during a weekend. This fact testifies the gas supply
system was well sealed.
All possibilities to explain porosity were ruled out except for inadequate shielding gas flow.
A 16mm diameter nozzle was used in these trials and gas-shielding rate was set to 26l/min some
porosity was detected in early trial pieces. Wires datasheet were consulted however there was no gas
flow rate data specified or data available did not include the used nozzle diameter.
It is generally advised when welding in still air may require gas flow rates of 14 to 19 L/min and in
moving air may require flow rates up to 27 L/min.
By consulting gas flow efficiency information in “ Wilkinson, M. E., Direct Gas Shield Analysis to
Determine Shielding Efficiency. Report of The Welding Institute (TWI), Cambridge, England;” and
applying, the Reynolds number (Re) from fluid mechanics, a flow line was plotted which facilitated the
67
establishment of a new and more efficient flow rate however the exact transition will be dependent on
the diameter of torch used, this study provided some perspective and cautions on the use of too high a
flow rate.
Flow regions
Flow Rate (l/min)
30
25
Turbulent
20
17,5
15
26
22,7
20
26,9
1st used
flow
2nd used
flow
Flow Line
Laminar
10
6,6
5
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Nozzle Size (mm)
Figure 85 - Shielding gas flow regions.
Shielding gas flow rate was decreased to 20L/min, a conservative value relatively to de data in
Figure 85, the appearance of porosities did not disappeared however its frequency of occurrence
became quite low.
5.3.2 Analysis of Results
As foreseen mechanized welding of this type of joints is not “fire and forget”.
Torch angle and alignment, stick-out and welding speed must be controlled during the entire
welding time to compensate changes in joint geometry or other unforeseen events. The parameters
obtained from the manual testing were very accurate working as guide lines needing just some
corrections case by case as it is impossible to guarantee identical joint geometry of every test pieces.
The most delicate and troublesome pass was the cap, not only because it was the surface one
but also because the weld bead has the transitions which starts within the joint and ends on the
surface of the test piece.
In PA position with 10mm thickness was found to be impossible to make the root pass with
welding speed higher than 7cm/min as the electric arc overtakes the molten bath and begins to
unsettle. Same scenario was found in PF position with the root welding speed limit being 11cm/min.
Shielding gas flow
A shielding gas leak means air is leaking back, this can be very wasteful and allow moistureladen air to enter the shielding gas lines.
By applying the law of partial pressures to the gas-shielding system it can be observed that the
total pressure of a gas mixture is the sum of the partial pressure of each gas. P
total
= P1 + P2 + P3 +
...Pn, the partial pressure is defined as the pressure of a single gas in the mixture as if that gas alone
occupied the container.
Since there is no Nitrogen in a shielding gas hose the partial pressure of Nitrogen in the hose is
zero. Since there is 78% Nitrogen in the surrounding air, the partial pressure of Nitrogen is 14.7
68
psia(absolute pressure) x .78 =11.5 psia. Therefore, there is a driving force for the Nitrogen to reach
equilibrium of 11.5 psia into the hose leak.
It will not move as fast as the CO2 coming out, which may have a driving force of say 25 psig +
14.7 psi = 39.7 psia versus 0.009 x 14.7= 0.13 in the air.
The driving force of 39.6 psia is 3.4 times as much however very little Nitrogen in the gas stream
is needed to cause problems. If welding a material needing a low hydrogen deposit is being used and
leaks are present in gas lines, hoses or fittings hydrogen from moisture laden air is entering back
though those leaks.
Shielding gas flow rate is often a neglected factor in FCAW, the assumption that more gas flow
results in better protection is misleading.
High gas flow surge at weld start causes turbulence in the shielding gas stream. This turbulence
causes air to be mixed into the shielding gas stream until the flow rate stabilizes to the preset level.
Figure 86 - Laminar and turbulent flow.
This entrained air causes, in addition to wasting shielding gas, excess weld spatter and can
cause internal weld porosity.
The gas flow required to efficiently protect the molten puddle cam be directly related to the
welding gun nozzle size and distance to work piece
First flow rate value used, which lead to some porosity, is marked as a red triangle in Figure 85.
As can be seen it was not an excessive value however for the gas type and nozzle size being used it
was too much. After plotting a flow line that separates approximately the turbulent regime from the
laminar a new and conservative flow rate was tested, marked in the graph with a green square, with
this new value the incidence of porosities cause by inadequate gas flow rate dropped to virtually zero.
All issues found in mechanized welding were quickly pinpointed and solved, this was due to the
work done in previous chapters which narrowed the origins of defects. NDT and visual inspection
ensured imperfection free weld and confirmed the welding procedure. With this task successfully
completed the production of test pieces for mechanical testing can now be initiated.
69
5.4 Mechanized Welding ST Samples
ST samples required a more precise and complete welding record as it is very important to know
how the joint was welded to relate with the future mechanical test results. The next example shows the
detailed welding information gathered from ST018.
Pass
1
2
3
Oscillation data
Dwell
WFS
WS
Width
Dwell left
Frequency
right
-1
[m/min] [cm/min]
[mm]
[s]
[s]
[min ]
6,0
9,0
4,5
0,2
0,2
60
8,0
22,0
7,5
0,2
0,2
60
8,0
17,0
10,0
0,1
0,1
60
Table 12 - Welding speed and oscillation of ST018
Kemppi Welding machine
data
Acquisition data
Heat
input
[m/min]
[V]
[A]
[m/min]
[V]
[A]
[kJ/mm]
6,0
22,8
164
6,0
22,9
160
2,5
8,0
25,8
201
8,0
26,0
198
1,3
8,0
25,9
198
8,1
25,9
193
1,7
Table 13 - Welding parameters, HI and Deposit rate calculation
WFS
Voltage
Current
WFS
Voltage
Current
Dep. rate
[kg/h]
2,1
2,8
2,8
ISO 5817
quality
level
back bead width
14,5mm
—
back bead height
2mm
B
front bead width
18mm
—
front bead height
2,5mm
B
Table 14 - Weld bead measurements and corresponding ISO 5817 level
Actual
measurements
ST samples from ST001 to ST011 plus ST016, ST017, ST021, ST022 and ST023 are 10mm
thickness and the material is S460M steel, from ST012 to ST015 plus ST018, ST019 and ST020 are
also 10mm thickness and the material is S355J2 steel.
ST samples with 30mm thickness where only made with the S460M steel corresponding to
ST101 up to ST106.
5.4.1 NDT imperfections detected in ST test pieces
Twenty nine ST test pieces were welded, in S460M and in S355J2, with thicknesses of 10mm
and 30mm in PF and PA positions. All twenty nine test pieces were controlled by UT-PA and RT and
evaluated by standard ISO 5817.
Even with all the trials done before and welding procedure well established some imperfections
were detected in some samples for mechanical testing.
In ST003 intermittent undercut was found, gas pores were present in ST007 and ST020 had gas
pores and slight slag inclusion.
The majority of imperfections detected were intermittent undercut and some gas pores, these
imperfections were only detected on four test pieces (3 of them were in the PA position). Three test
70
pieces, despite the imperfections found, are classified by ISO 5817 as class B (highest quality level),
only ST005 had imperfection not permitted by the standard and required repair.
ST005 Imperfection
The most interesting and dimensionally large imperfection, present in almost two thirds of the
joint, was found in PA test piece ST005.
Figure 87 - Radiography of ST005, 0 to 165mm
Figure 88 - Radiography of ST005, 170mm to 335mm
Figure 89 - Radiography of ST005, 335 to 500mm
A section with defect was extracted from ST005 and a macrograph was made in order to identify
which kind of imperfection was detected by NDT.
Figure 90 - Macrograph of ST005 weld bead showing a slag channel
71
Figure 91 - ST005 removed defect sample and air carbon arc thinning
It was found that the imperfection was a slag inclusion.
After evaluating the imperfection the most probable cause found for this to happen was due to the
method of removing the root slag.
The root slag had a concave geometry which translated into a very difficult slag removal
operation, the preferred removal method using a chisel and a hammer did not proved to be an efficient
method as the slag did not detach from the weld bead.
The method used to achieve a good and easy slag detach consisted in making a cut on the
center of the slag with a grinder without or by removing very little of the weld bead (an estimate for the
groove made by grinding is less than 0.5mm), after this cut was made the remaining slag detach very
easily with chisel and hammer.
There are two hypotheses to explain the slag inclusion:
1-
The cut made by the grinder entrapped existing slag in the small groove of the cut and
the second pass did not melt adequately the root pass leaving slag embedded in the bead.
2-
The cut originated a small groove that was enough to disturbing the weld flow and
entrap the slag formed by the following pass.
The first hypothesis appeared to be more unlikely to cause the imperfection however not being
sure which one was the cause of imperfection, slag removal procedure and welding procedures were
altered in order to eliminate all possible causes.
Slag cut was made, with grinder, on left and on right side instead of being centered, the cut did
not touched the weld bead and welding speeds were increased in order to avoid a cushion effect
(”cushioning” of the arc force by the extra weld metal deposited when welding at slower speed) and
provide a better penetration by the arc travelling on the edge of the molten material.
Figure 92 - Difference between initial preparation (left) and correct preparation (right)
72
Two new test pieces were produced to replace ST005 and they had no evidence of imperfections
thus the procedure changes were correctly done and were repeated in the following test pieces in the
PA position with no indication of imperfections.
5.4.2 Mechanical Testing Results
Width
Test
diameter
Reference
mm
Initial
section
mm2
Yield
Tensile Elongation
Reduction
Strength Strength
after
of Area %
Mpa
MPa
fracture %
ST004_T1
5,0
19,7
531
614
25,3
67,6
ST004_T2
5,0
19,7
540
650
27,1
67,9
ST004_T3
5,0
19,8
610
695
27,7
67,8
ST006_T1
5,0
19,8
476
596
29,0
71,7
ST006_T2
5,0
19,6
524
644
26,2
70,1
ST006_T3
5,0
19,7
496
599
24,8
73,7
ST013_T1
5,0
19,7
552
680
27,5
67,9
ST013_T2
5,0
19,8
559
699
27,4
74,0
ST013_T3
5,0
19,8
558
715
29,9
67,1
ST020_T1
5,0
19,7
495
625
28,1
72,9
ST020_T2
5,0
19,8
516
632
28,8
72,6
ST020_T3
5,0
19,8
536
664
27,6
71,1
Table 15 - ST samples tensile testing
ST008
Temp.
Weld
Fusion
(ºC)
Metal
Line
22
100J
158J
-20
76J
57J
-50
44J
36J
Table 16 - Average values of
Charpy Test of ST008 in S460M
steel(welded transversely to
rolling direction)
200
WM
Energy (J)
150
FL
100
50
0
-20
0
20
Temperature (ºC)
Figure 93 - ST008 Charpy test.
-60
ST004 & ST007
FL
FL
Temp Weld
Fusion
+2m
+5m
Line
. (ºC)
Metal
m
m
22
102J
129J
206J
189J
-20
63J
95J
123J
180J
-50
45J
46J
98J
164J
Table 17 - Average values of Charpy Test
of ST004 & ST007 in S460M steel (welded
parallel to rolling direction)
-40
40
Energy (J)
250
WM
200
FL
150
FL+2
100
FL+5
50
-60
0
-20
0
20
Temperature (ºC)
Figure 94 - ST004&ST007 Charpy test.
-40
40
73
ST013 & ST014
FL
FL
Temp Weld Fusion
+2m
+5m
Line
. (ºC) Metal
m
m
22
91J
67J
48J
56J
-20
69J
60J
28J
40J
-50
47J
20J
19J
17J
Table 18 - Average values of Charpy Test
of ST013 & ST014 in S355J2 steel(welded
transverse to rolling direction)
100 Energy (J)
WM
FL
80
FL+2
60
40
FL+5
20
0
-20
0
20
Temperature (ºC)
Figure 95 - ST013&ST014 Charpy test.
-60
ST012 & ST015
Weld
FL
FL
Temp
Fusio
Meta
+2m
+5m
. (ºC)
n Line
l
m
m
22
93J
123J
104J
110J
-20
64J
54J
61J
76J
-50
39J
25J
25J
29J
Table 19 - Average values of Charpy
Test of ST012 & ST015 in S355J2 steel
(welded parallel to rolling direction)
-40
40
Energy (J)
140
WM
FL
120
FL+2
100
FL+5
80
60
40
20
-60
0
-20
0
20
Temperature (ºC)
Figure 96 - ST012&ST015 Charpy test.
-40
40
Full Charpy tests are in table 12 up to 15 and can be consulted in Annex 9.8
5.4.3 Analysis of Results
Although some imperfections were detected in the ST samples, according to ISO 5817, no repair
was mandatory, except for ST005. The results obtained by NDT confirmed not only the quality of
welded joints but also the ability to detect any kind of imperfection.
Methodology and welding parameters were very alike with the ones used in the mechanized trials
and in twenty nine test pieces only one failed the standard requirements, this shows that the welding
procedure is robust and easily repeatable with high quality welds.
The major imperfection found on ST005 reveals that the welding position PA is more prone to
develop this kind of imperfection than the PF position, this is related to the less efficient convection in
the molten pool when welding in PA making any failure in the joint preparation a possible triggering
defect. The key to eliminate the imperfection was the homogeneous preparation of the joint to receive
the second pass as it is nearly impossible to control weld flow conditions.
The disturbance of the weld flow can be explained by two types of flow generated.
The surface temperature of the weld pool will usually be maximum in the center of the weld pool
and decreases with increasing distance from the center creating a temperature gradient which in turn
will generate a surface tension gradient. The negative gradient of the surface tension generates
outward directed flow called Marangoni flow, on the opposite side is the Lorentz flow which generates
electromagnetic forces in the weld pool due to divergency of the electric current causing pressure
differences and resulting in downward directed flow.
74
As result of these reversals, undesirable variations in penetration depth can occur and the
existence of the small groove can be enough to produce the imperfection above described [54].
Figure 98 - Lorentz flow.
Figure 97 - Marangoni flow.
Mechanical testing plays an important role in evaluating fundamental properties of engineering
materials. Tension test is widely used to provide basic design information on the strength of materials.
Tensile Testing
The most relevant data obtained by tensile test for the present work are yield and tensile strength,
area reduction and elongation after fracture.
The values obtained for the yield and tensile strength confirm the mechanical properties expected
when welding with the chosen wire and these types of steel since the lowest value of yield strength
registered was 476MPa and tensile strength was 599MPa above minimum value in the wire certificate
present in Annex 9.5 as well in the steel certificate present in Annex 9.6.
The values for these properties vary even between specimens of the same test sample, the
maximum variation is about 81 MPa and this can be explained by some changes in the welding
parameters to compensate inhomogeneous joint zones. These changes modify the heat input and
joint filling, as explained before, this can originate different microstructures consequently different
mechanical properties. Values obtained are within the range stipulated by Tensile Testing of Metallic
Materials (ISO 6892-1:2009) where a statistical study made to develop criteria for load and resistance
factor design showed that the mean yield points can exceeded the specified minimum yield point Fy
(specimen located in web) as indicated below and with the indicated coefficient of variation (COV)
Plates: 1.10Fy, COV = 0.11.
The ductility of a material is a measure of the extent to which a material will plastically deform
before fracture. Ductility is also used a quality control measure to assess the level of impurities and
proper processing of a material. Ductility can be assessed by the other two parameter, area reduction
and elongation after fracture. A material that experiences very little or no plastic deformation upon
fracture is termed brittle and the converse is also true.
Ductility measurements may be specified to assess material quality (indicator of changes in
impurity level or processing conditions) even though no direct relationship exists between the ductility
measurement and performance in service. Ductility can be expressed either in terms of percent
elongation (A) or percent reduction in area (Z):
[
(
[
(
]
(32)
]
(33)
75
Elongation is the change in axial length divided by the original length of the specimen or portion
of the specimen. It is expressed as a percentage. Reduction of area is the change in cross-sectional
area divided by the original cross-sectional area. This change is measured in the necked down region
of the specimen. Like elongation, it is usually expressed as a percentage.
Values of elongation are above the ones of wire certificate and steel. The minimum elongation
belongs to S460M steel with a value of 17% from tensile test done the lowest of all specimens was
24.8%, regarding the values of reduction of area and using once again ISO6892-1 this type of fracture
is considered ductile.
Impact Testing
Identical analysis can be made for the Charpy impact tests which determine the amount of energy
absorbed by a material during fracture. This absorbed energy is a measure of a given materials notch
toughness and acts as a tool to study temperature-dependent ductile-brittle transition. All results
obtained while testing confirm the predicted behavior, supplied by manufacturers, at all temperatures
including at -50ºC which is an extreme temperature for both steel and filler wire. All test pieces
absorbed amounts of energy above the minimum value in the wire certificate present in Annex 9.5 as
well in the steel certificate present in Annex 9.6 for all temperatures.
The transition from ductile to brittle can be seen when impact energy is plotted as a function of
temperature, the curve will show a rapid dropping off, very sharply, of impact energy as the
temperature decreases. The transition temperature is often a good indicator of the minimum
recommended service temperature.
Some values are not consistent, as seen in Figure 94 the value of FL+2 at 20ºC is above de FL+5
at the same temperature, this fact might raise suspicions however it is necessary to remember that
this type of test relies on an created notch as mean of crack propagation. The location of the notch is
very important, if the notch is created in a zone with hard and brittle microstructures such as
martensite or carbides pockets the energy absorbed will be much lower. The separation of ferrite and
perlite in strips, results from microsegregation of manganese in austenite and when cooling
accelerates the formation of pro-eutectoid ferrite enriching the adjacent area with carbon and
formation of pearlite can lead to inconsistence values.
One other aspect of Charpy results is the energy difference between longitudinal and transverse
specimen. Generally in the longitudinal specimen the energy absorbed increases along the distance
from the weld metal in other words FL+5 will absorb the most amount of energy of all the locations
followed by FL+2 then FL and finally WM, on the other and the transverse specimen behaves in the
opposite way and has generally lower values than the longitudinal one.
One explanation is that longitudinal test metal across the grain of steel and have higher notch
toughness than transverse. Thermomechanically rolled material possess long grain boundaries in
rolling direction and another reason for the anisotropy is the elongation of non-metallic inclusions like
manganese sulphide. At high temperatures the sulphide inclusions are harder than the matrix material,
but in the temperature regime of the thermomechanical treatment the inclusions are softer. During the
rolling process they stretch in the rolling direction and elongate resulting in anisotropic toughness
76
behavior as a consequent, the Charpy impact values in the transverse direction are usually inferior to
those in the rolling direction.
However, modern steelmaking techniques using calcium-silicon desulphurization can reduce the
effect of inclusions in the steel. The addition of elements (titanium, zirconium, rare earth metals) that
influence the hardness of the sulphide also have a beneficial effect on the toughness [54].
Figure 99 - Anisotropic mechanical properties.[54]
77
6 Conclusion
The main conclusions we can gather from the work done for this thesis are:
Although self-shielded FCAW wires have evolved greatly they are still a step behind the gasshielded wires, self-shielded wires seemed to have good potential for application in the scope of this
project as mobility being the best property however self-shielded FCAW wires do not match
mechanical properties of S460M steel. Surprisingly when welding open root passes with self-shielded
wires was when the best deposition rates and behavior were obtained. Self-shielded wires presented a
very harsh and difficult arc behavior translating into an additional difficulty when trying to mechanize
the procedure
Gas-shielded surpassed in every way their self-shielded counterparts, mechanical properties
similar to S460M, better deposition rates, steady behavior on ceramic backing provided a good start
point to mechanize. Gas-shielded wires have a low hydrogen class (H5) and although the spatter is
reduced with when using 20% CO2 argon-based mixtures the use of pure CO2 as welding gas results
normally in lower values of hydrogen and it is more economical.
Not only the selection of gas type is important but also the relation of shielding gas flow rate and
gun nozzle. If a too high flow rate is set it may result in a turbulent flow which causes air to be mixed
into the shielding gas stream, on the other side a low flow rate will provide insufficient protection.
Selecting the right wire for this project was greatly simplified by testing manually each one. If a
poor wire behavior was detected while welding manually it would be almost impossible to mechanize
them thus saving time from experimenting the wires in a mechanized setting where any change would
translate in a rearrangement of the mechanized structure. Parameters found manually needed very
few corrections when mechanized nevertheless constant oversight of the welding procedure is
mandatory and the parameters obtained in chapter 5.1, Figure 65, 66 and 67 are guidelines for the
ideal joint, each joint needs to be assed in order to choose the correct parameters, also a good joint
preparation minimizes the corrections during welding and reduces the probability of imperfections.
S460M weldability test gave an idea of the steel behavior when subjected to certain heat inputs,
results show some detrimental structures like martensite and carbides but in small quantities and the
hardness values show that even with very low HI, according to ISO standard, all welds are acceptable.
However the use of HI equal or higher than 1.5KJ/mm will result in more favorable microstructures and
so it is advisable.
The Rosenthal analytical solution applied to welding proved to be a fairly accurate tool to predict
heat behavior in steel however have some attention must be taken when working with to intermediate
thicknesses.
The NDT chosen to verify the weld quality will also be important for the future of the project,
especially UT as it is safe to operate, no radiation emission, fairly easy to mechanize which are two
very important properties since it is supposed to be operated on site. Weld quality was verify not only
by NDT but also by mechanical testing such as tensile test and impact tests, that ensured structural
78
integrity of welded joints. The values obtained by these tests were well within the acceptable range of
materials in question.
The tensile and Charpy tests performed proved the mechanical qualities expected of this type of
material and process and within the range recommended by ISO standards.
From this work it can be concluded that use of FCAW for onsite build and erection of WEC is a
good substitute for the current method of fabrication and erection of WEC using bolted flanges. SAW
has better deposition rates than FCAW, mechanically and metallurgically weld beads are very similar
with SAW and FCAW and from the NDT results of the ST test pieces it can be seen that this welding
procedure is very robust and easily repeatable, when done with proper precautions.
79
7 Future work
The results obtained in this thesis are just the first step to evaluate the capability of replacing the
SAW done in factory and tower bolted connections by using FCAW on site for WEC. The results here
presented can only testify the capability of using FCAW in this type of joints with none or very few
imperfections.
Mechanical testing is of the upmost importance for this project. Mechanical testing such as impact
tests (Charpy), CTOD, high cycle fatigue, tensile tests and bending tests will determinate if tower
modular connections done with FCAW are appropriate under typical loads of a on service WEC.
A higher number of mechanical tests should take place allowing the creation of a much more
complete test data base and it is never too much to overemphasize the importance of fatigue tests for
this kind of structures.
Although there are no data from the behavior of these welded joints in high cycle fatigue it would
be beneficial to study the effects of surface treatment of the welded joints by shot peening or ultrasonic
impact treatment, local compression or thermal and vibratory stress relief.
Finally one other important aspect is the Cost -Benefit analysis. With all the welding data now
available it is possible to calculate the costs associated with this alternative way for constructing a
WEC compared with the current.
80
8 References
[1] UNFCC, 2007. Kyoto Protocol. [online] Available at:
<http://unfccc.int/kyoto_protocol/items/2830.php> [Accessed 14 July 2011]
[2] http://www.epa.gov/air/airpollutants.html
[3] Manwell, J.F., Rogers, A.L, McGowan, J.G., 2009. Wind Energy Explained: Theory, Design and
Application. Second Edition. Amherst. Wiley
[4] Jain, Pramod, 2011, Wind Energy Engineering, McGraw Hill
[5] Pelamis wave power. Available from:
http://www.pelamiswave.com/gallery?category=images&page=1 [Accessed 14 July 2011]
[6] DTI, 2005. What is the potential for commercially viable renewable generation technologies? Oxera
for the Department of Trade and Industry,2004.
[7] http://www.a-nossa-energia.edp.pt/centros_produtores/complementariedade_hidroeolica_pe.php
[Accessed 14 July 2011]
[8] Eric, FRANCISCO; Energy buffering for large wind farms, MSc Energy Systems and the
Environment – University of Strathclyde 2006
[9] D'Ambrosio, Marco & Medaglia, Marco; Vertical Axis Wind Turbines: History, Technology and
Applications; University essay from Högskolan i Halmstad/Högskolan i Halmstad/Sektionen för
ekonomi och teknik (SET)Energiteknik; 2010
[10] Åke Larsson; The Power Qualities of Wind Turbines; Doctoral thesis - University dissertation
from Chalmers University of Technology [2000]
[11] H. H. Hubbard, K. P. Shepherd. Wind Turbine Acoustics. In D. A. Spera (ed.). Wind Turbine
Technology. ASME Press 1994. P. 393.
[12] http://www.lorc.dk/Knowledge/Wind/Towers [Accessed 26 July 2011]
[13] Velkovic, Milan and Husson Wylliam. High-strength wind turbine steel towers. Elforsk rapport
09:11.
[14] Rodeja, Josep Pigem , Sustainability assessment of towers for wind turbines; University essay
from Luleå/Department of Civil, Environmental and Natural Resources Engineering 2012
[15] Mukund R. Patel, Wind and Solar Power Systems: Design, Analysis and Operation, second
edition. Published in 2006 by CRC press, Taylor & Francis Group.
[16] Global Wind Energy Council GWEC website. Available from:
http://www.gwec.net/index.php?id=181 [Accessed 03 September 2011]
81
[17] Jacobson, Mark Z. & Archer, Cristina L. ; Saturation wind power potential and its implications for
wind energy; Proceedings of the National Academy of Sciences, 1208993109v1-6; 2012
[18] Kate Marvel, Ben Kravitz, Ken Caldeira. Geophysical limits to global wind power. Nature Climate
Change, 2012; DOI: 10.1038/nclimate1683 2012
[19] Staffan Engström, Tomas Lyrner, Manouchehr Hassanzadeh, Thomas Stalin and John
Johansson, Tall towers for large wind turbines- Report from Vindforsk project V-342 Höga – July2010
[20] Schumacher, A (2003), “Fatigue behaviour of welded circular hollow section joint in bridges”,
doctoral thesis, Ecole Polytechnique Federale de Lausanne, Lausanne, Switzerland;
[21] EN 1993-1-9, (2005), “Design of steel strucutures – Fatigue”, European Committee for
Standardization, Brussels, Belgium
[22] Germanischer Lloyd Windenergie GmbH, (2005), Rules and guidelines IV Industrial Services –
Guideline s for the Certification of Offshore Wind Turbines, Hamburg, Germany;
[23] Hau, E., 2006. Wind Turbines: Fundamentals, Technologies, Application, Economics. Second
Edition. Krailling. Springer.
[24] M. Hassanzadeh, T. Stalin and J. Johansson. High towers for wind power onshore – information
dissemination. Vattenfall Reasearch and Development AB. 2008-12-05. P.8.
[25] FATHOMS final report, (2008) , contract number RFSR-CT-2005-00042
[26] Jeppesen, Allan Hammer; An alternative connection in steel towers for wind turbines - friction
connection – HISTWIN University essay from Luleå/Department of Civil, Mining and
Environmental Engineering [2008]
[27] Oliveira, Vanessa Reich de; The Use of Wind Energy for Electricity Generation in Brazil; Master
in Science In Energy Systems and the Environment, University of Strathclyde 2002
[28] AWS, Welding Processes Part 2, 9th Edition, Volume 3
[29] James F. Lincoln Arc Welding Foundation, 2000, The Procedure Handbook of Arc Welding, 14th
Edition, Lincoln Electric
[30] Quintino, L.; Santos, J.; “Processos de Soldadura”, Instituto de Soldadura e Qualidade, 2.ª
Edição, 1998
[31] http://www.iiwindia.com/pdf/ASME_&_ISO_EN_Welding_Position_Comparison.pdf [Accessed 30
November 2011]
[32] http://www4.hcmut.edu.vn/~dantn/Weld%20pool/ [Accessed 30 November 2011]
82
[33] Fernandes, Paulo Eduardo Alves; Evaluation of fracture toughness of the heat affected zone
(HAZ) of API 5L X80 steel welded SMAW and FCAW. São Paulo,2011
[34] http://www.ndt.net/article/wcndt00/papers/idn151/idn151.htm [Accessed 05 May 2012]
[35] Anmol S. Birring, "Ultrasonic Testing in Electric Power Plants," NDT Handbook, Ultrasonic
Testing, 3rd Edition, ASNT, 2007
[36] http://www.ndt-validation.com/technologies/pr_2.jsp?menu_pos=0 [Accessed 05 May 2012]
[37] http://www.lincolnelectric.com/assets/en_US/Products/Consumable_Flux-CoredWires-SelfShielded-Innershield-InnershieldNR-233/c32400.pdf [Accessed 30 November 2011]
[38] http://www.esabna.com/us/en/support/downloads/litDownloads/CoredWireBooklet/index.html#/24/
[Accessed 30 November 2011]
[39] http://www.esabna.com/EUWeb/MIG_handbook/592mig9_8.htm [Accessed 30 November 2011]
[40] Mostafa ; M.N. Khajavi Optimization of welding parameters for weld penetration in FCAW 2006
Journal of Achievements in Materials and Manufacturing Engineering
[41] STARLING, Cícero Murta Diniz; MODENESI, Paulo J. and BORBA, Tadeu Messias Donizete.
Bead characterization on FCAW welding of a rutilic tubular wire. Soldag. insp. (Impr.) [online].
2009
[42] V. Kumar and N. Murugan, "Effect of FCAW Process Parameters on Weld Bead Geometry,"
Journal of Minerals and Materials Characterization and Engineering, Vol. 10 No. 9, 2011, pp. 827842. doi: 10.4236/jmmce.2011.109064.
[43] Myers, P. S., Uyehara, O. A., and Borman, G. L. 1967. Fundamentals of heat flow in welding.
Weld. Res. Council Bull. 123: 1–46]
[44] Ion, J. C., Easterling, K. E., and Ashby, M. F. 1984. A second report on diagrams of
microstructure and hardness for heat-affected zones in welds. Acta Metall. 2(11): 1949–1962
[45] http://weldingdesign.com/processes/news/wdf_10760/ [Accessed 30 November 2011]
[46] https://canteach.candu.org/Content%20Library/20053428.pdf [Accessed 30 November 2011]
[47] Krauss, G., 1992, “Heat Treatment and Processing Principles Materials Park” American Society
for Metals.
[48] Doherty, R.D., Martin J.W. Cantor, B. 1997, “Stability of Microstructure in Metallic Systems”, 2. ed.
Cambridge, England: Cambridge University Press, 1997
83
[49] Smith, W. F., Princípios de Ciência e Engenharia de Materiais, 3ª Edição McGraw-Hill de
Portugal, 1998
[50] Lopes Dias, E. M.; Miranda R. M.; “Metalurgia da Soldadura”; Instituto de Soldadura e Qualidade,
1993
[51] http://www.staff.ncl.ac.uk/s.j.bull/mmm211/PHASE/index.htm [Accessed 6 December 2011]
[52] SADEK A (Cmrdi, Cairo, Egy) IBRAHAM R N (Monash Univ., Melbourne, Aus) PRICE J W H
(Monash Univ., Melbourne, Aus) SHEHATA T (Monash Univ., Melbourne, Aus) USHIO M (Osaka
Univ., Osaka, Jpn) ; Effect of Welding Parameters of FCAW Process and Shielding Gas Type on
Weld Bead Geometry and Hardness Distribution. 2001
[53] Murray, Amanda. Examination of SAW and FCAW high strength steel weld metals for offshore
structural applications Cranfield University Current Institution 1997
[54] W. Haumann, F. Koch and W. Recknagel, “Anwendung von fannenmetallurgischen Verfahren bei
der Herstellung von Vormaterial für geschweißte Hochdruckgasleitungsrohre”, Stahl und Eisen,
104, 25-26 (1984), 1357 – 1360.]
9 Annexes
9.1 Pre-heat calculation methods
1º Method – AWS D1.1
Consulting Table 3.1 and 3.2
For A913 Gr.65 (S460M) with thickness between 3mm and 20mm and using low hydrogen electrodes MPIT =
10ºC
For S460M with thickness superior to 3mm and using electrodes with maximum diffusible hydrogen of
8ml/100g (H8) MIPT = 0ºC (In case base metal is below 0ºC is mandatory to preheat to 20ºC).
For S355J2 with thickness of 30mm MIPT = 65ºC and with 20mm thickness MIPT=10ºC
2º Method - EN1011 – 2 2001
Select hydrogen scale
Hydrogen scale C
84
Calculate combined thickness
Max Combined Thickness = 60mm
Calculate heat input in accordance to EN1011-1 1998
Identify preheat from charts
CE = 0.36 therefore in Fig. C.2.a Preheat temp = 0ºC (20mm) & 0ºC (30mm) for heat input
>0.75KJ/mm (S355J2)
CE = 0.41 therefore in Fig. C.2.b Preheat temp = 0ºC (20mm) & 0ºC (30mm) for heat input >0.9KJ/mm
(S460M)
3º Method - En1011- 2 (optional)
Calculate CET
CET = 0.256 (S460M)
CET = 0.315 (S460M)
Calculation by plate thickness
Tpd= 0ºC (10mm) & 1.17º (30mm)
Calculation by heat input
TpQ = (53xCET-32)xQ-53xCET+32 (ºC)
With 1.0KJ/mm Tpq= 0ºC
Calculate preheat temperature
TPCET= 750xCET(%)-150 (ºc)
TpCET = 41.2 (S460M)
TpCET = 86.6 (S460M)
Calculation by Hydrogen content
HD value in ISO3690
TpHD=62 x HD^0.35 -100 (ºC)
TpHC = 8.9º C
Preheat Temperature
Tp=TpCET+TpD+TpHD+TpQ (ºC)
Tp= 51.2ºC (30mm) & <0ºC (10mm) – (S460M)
Tp= 96.5ºC (30mm) & 30ºC (10mm) – (S355J2)
85
9.2 Heat Input Trial Data
Spec.
ET(KJ/mm)
WS(cm/min)
Ia
(A)
Im
(A)
Va
(v)
Vm
(v)
WFSa
(m/min)
WFSm(m/min)
ET real
(KJ/mm)
A
0,5
35
164
160
22,8
22,9
6,1
6
0,51
B
1
18
174
171
22,8
22,9
6,1
6
1,05
C
1,5
18
212
209
26,8
26,5
9,2
9
1,50
D
2
13
224
217
26,5
26,8
9,2
9
2,17
E
2,5
11
226
220
26,4
26,7
9,2
9
2,58
F
3
13
281
278
29,3
29,7
12,3
12
3,04
G
3,5
12
267
259
29,3
29,8
12,3
12
3,11
H
0,5
54
214
202
26,6
26,9
9
9
0,49
I
1
27
222
210
26,6
26,9
9
9
1,03
J
1,5
18
216
204
26,6
26,9
9
9
1,50
K
2
13,5
215
206
26,5
26,9
9
9
2,00
L
2,5
11
223
210
26,5
26,9
9
9
2,52
M
3
9
215
205
26,6
26,9
9
9
3,00
N
3,5
7,6
9
3,82
233 220 26,5 26,9
9
Table 20 - Heat Input Trial Data
ET – Heat input
WS – Welding speed
Ia – Welding Current registered on data acquisition
Im - Welding Current registered welding machine
Va – Welding Voltage registered on data acquisition
Vm - Welding Voltage registered welding machine
WFSa – Wire Feed Speed registered on data acquisition
WFSm - Wire Feed Speed registered welding machine
86
9.3 Macrographs and Micrographs of Heat Input Trial Samples
A
C
B
D
E
G
H
I
M
N
87
All magnifications: 500X
Sample A –
Weld Material
Heat Affected Zone-grain growth region
Heat affected zone- grain refined region
Heat affected zone - intercritical region
Heat affected zone – subcritical region
Unaffected parent material
Sample B
Weld Material
Heat affected zone-grain growth region
Heat affected zone – intercritical region
Heat affected zone - subcritical region
Unaffected parent material
88
Sample C
Weld Material
Heat Affected zone-grain growth region
Heat affected zone-grain refined region
Heat affected zone-intercritical region
Heat affected zone-subcritical region
Unaffected parent material
Sample D
Weld Material
Heat affected zone –grain growth region
Heat affected zone –grain refined region
Heat affected zone – intercritical region
Heat affected zone subcritical region
Unaffected parent material
89
Sample E
Weld material
Heat affected zone –grain growth region
Heat affected zone –grain refined region
Heat affected zone-intercritical region
Heat affected zone-subcritical region
Unaffected parent material
Sample G
Weld material
Heat affected zone – grain growth region
Heat affected zone – intercritical region
Heat affected zone subcritical region
Unaffected parent material
90
Sample H
Weld material
Heat affected zone-grain growth region
Heat affected zone (top: grain growth region;
bottom: grain refined region)
Heat affected zone-intercritical region
Heat affected zone-subcritical region
Unaffected parent material: ferrite and pearlite
Sample I
Weld material
Heat affected zone-grain growth region
Heat affected zone-intercritical region
Heat affected zone-subcritical region
Unaffected parent material
91
Sample N
Weld material
Heat affected zone – grain growth region
Heat affected zone – grain refined region
Heat affected zone – intercritical region
Heat affected zone – subcritical region
Unaffected parent material
92
9.4 Hardness Values
Sample A
Local
WM
HAZ
PM
Sample B
Indentation
HV10
Value
1
Indentation
HV10
Value
Indentation
HV10
Value
277
1
237
1
228
2
291
2
238
2
231
3
270
3
232
3
229
4
-
4
-
4
228
5
-
5
-
5
-
6
363
6
331
6
302
7
235
7
226
7
267
8
-
8
-
8
207
9
-
9
-
9
-
10
-
10
-
10
-
11
196
11
202
11
193
12
181
12
182
12
184
13
170
13
179
13
177
14
163
14
174
14
173
15
172
16
170
17
168
18
168
15
163
16
158
Local
WM
HAZ
PM
15
Sample D
Local
WM
HAZ
Sample C
168
Local
WM
HAZ
PM
Sample E
Indentation
HV10
Value
1
Sample G
Indentation
HV10
Value
Indentation
HV10
Value
221
1
214
1
215
2
225
2
215
2
210
3
222
3
220
3
209
4
222
4
212
4
211
5
-
5
-
5
-
6
289
6
274
6
248
7
264
7
274
7
203
8
208
8
208
8
237
9
195
9
194
9
188
10
-
10
-
10
-
11
187
11
182
11
175
12
174
12
175
12
178
PM
Local
WM
HAZ
Local
WM
HAZ
PM
PM
13
174
13
178
13
179
14
170
14
172
14
177
93
15
175
15
175
15
173
16
172
16
171
16
173
17
164
17
165
17
168
Sample H
Local
WM
HAZ
PM
Sample I
Indentation
HV10
Value
1
Sample M
Indentation
HV10
Value
Indentation
HV10
Value
281
1
238
1
212
2
275
2
236
2
214
3
281
3
239
3
214
4
-
4
234
4
210
5
-
5
313
5
-
6
345
6
-
6
256
7
217
7
-
7
253
8
-
8
224
8
206
9
-
9
201
9
192
10
-
10
-
10
-
11
185
11
180
11
179
12
176
12
176
12
179
13
172
13
174
14
171
14
171
15
175
16
173
17
172
13
166
14
160
15
157
Local
WM
HAZ
PM
Local
WM
HAZ
PM
Sample N
Local
WM
HAZ
Indentation
HV10
Value
1
197
2
203
3
211
4
207
5
201
6
234
7
231
8
210
9
187
10
181
11
177
12
175
13
170
14
169
PM
94
15
171
9.5 Filler Wire Specification
Filler wire
NR-233
Manufacturer Lincoln Electric
NR-203Ni1
NR-Offshore
Coreshield 8
Coreshield 8Ni1 H5
Lincoln
Electric
Lincoln
Electric
ESAB
ESAB
Shielding
type
Self-Shielded
Self-Shielded
Self-Shielded
Self-Shielded
Self-Shielded
Classification
E71T-8 (*)
E71T8-Ni1 (§)
NA
E71T-8 (*)
E71T8-Ni1-J
(§)
Chosen
diameter
[mm]
1.6
2.0
2.0
1.6
1.6
Weld metal
Nickel [%]
Not Specified
0.90
1.10
≤ 0.25
1.00-1.10
―
420
NA
420
―
AWS
Tensile
strength
[MPa]
Min/range
EN/ISO
400
400
NA
400
400
―
500-640
NA
500-640
―
AWS
Elongation
[%]
Min EN/ISO
483-655
483-621
NA
483-655
483-621
―
20
NA
20
―
Min AWS
22
20
NA
22
20
―
47/ -30
NA
47 / -20
―
27 /-29
27 /-29
NA
27 /-29
27 /-29
Min Yield
strength
[MPa]
EN/ISO
Charpy Vnotch impact
energy [J]
EN/ISO
Min/@ [ºC]
AWS
Min/@ [ºC]
Table 21 - Filler wire specification- SS.
95
Filler wire
Filarc PZ6113
Filarc
PZ6113S
Filarc
PZ6114S
Filarc
PZ6116S
Filarc PZ6138
Manufacturer
ESAB
ESAB
ESAB
ESAB
ESAB
Gas - CO2
Gas - CO2
Gas - CO2
Gas - Ar/CO2
Shielding
type
Classification
Gas CO2
Gas Ar/CO2
E71TE71T-1M1C-H4 (*)
H8 (*)
E71T-9CH4 (*)
E71T-1C-JH4 E81T1-K2C- E81T1-Ni1M(*)
JH4 (§)
JH4 (§)
Chosen
diameter
[mm]
1.2
1.2
1.2
1.2
1.2
1.2
Weld metal
Nickel [%]
≤ 0.50
≤ 0.50
≤ 0.50
0.25 – 0.5
1.30 – 1.70
0.80 - 1.10
420
460
460
460
460
500
400
400
400
400
469
469
500-640
530-680
530-680
530-680
530-680
560-720
AWS
Elongation
[%]
Min EN/ISO
483-655
483-655
483-655
483-655
552-689
552-689
20
20
20
20
20
18
Min AWS
22
22
22
22
19
19
47 / -20
47 / -20
47 / -30
47 / -40
47 / -60
47 / -60
27 /-18
27 /-18
27 /-29
27 /-40
27 /-39
27 /-39
Min Yield
strength
[MPa]
EN/ISO
AWS
Tensile
strength
[MPa]
Min/range
EN/ISO
Charpy Vnotch impact
energy [J]
EN/ISO
Min/@ [ºC]
AWS
Min/@ [ºC]
Table 22 - Filler wire specification- GS.
96
9.6 S355J2 and S460M Steel properties
Steel designation
Standard
ReH - Minimum yield
strength [MPa]
Nominal
Thickness[mm]
Rm - Tensile strength
[MPa]
Nominal
Thickness[mm]
Elongation [%]
Nominal
Thickness[mm]
Notch impact test.
Charpy - Min energy
[J]@[ºC]
Nominal
Thickness[mm]
Chemical composition
[max %]
C
Si
Mn
Ni
P
Si
Cr
Mo
V
Ni
Nb
Ti
Al
Cu
t ≤ 16
16 ≤ t ≤ 40
40 ≤ t ≤ 63
3 ≤ t ≤ 100
t ≤ 40
40 ≤ t ≤ 63
3 ≤ t ≤ 40
40 ≤ t ≤ 63
S355J2+N (1.0577)
EN 10025-2
355
345
335
470 - 630
20
19
27 @ -20
t ≤ 150
t ≤ 30
Maximum carbon
equivalent CEV [%]
Nominal
Thickness[mm]
0,23
0,6
1,7
0,035
0,45
0,45
30 ≤ t ≤ 150
t ≤ 16
S460M (1.8827)
EN 10025-4
460
440
430
540 - 720
530 - 710
17
17
55 @ +20
47 @ 0
43 @ -10
40 @ -20 (*)
0,16
0,6
1,7
0,8
0,03
0,025
0,3
0,2
0,12
0,025
0,05
0,05
0,02
0,55
0,47
0,45
16 ≤ t ≤ 40
0,46
40 ≤ t ≤ 63
0,47
Table 23 - HSS properties
(*) - 40 J @ ‒20 ºC º 27 J @ ‒30 ºC (according to Eurocode 3)
97
9.7 Steels certificate
The mechanical properties according to the mill certificate are summarized in the next tables.
plate
thickness
ReH
Rm
A
CVN L -20°C
5972218-1 /
5972218-2
10
521
575
31.8
149 / 164 / 156 (7.5mm)
5972219-1 /
5972220-1
30
459
541
27.2
236 / 233 / 244
Table 24 - Mechanical properties of S460M plates
C
Mn
Si
S*
P*
Al
N
Cu
Ni
Cr
Mo
Nb
V
Ti
Ceq
.11
1.42
.241
10
140
.029
62
.025
.021
.03
.003
.047
.035
.014
.36
Table 25 - Nominal composition of S460M plate
plate
Heat
thickness
ReH
Rm
A
CVN L -20°C
4047233-1
195582
30
426
575
22.6
35 / 34 / 30
4045434-1
195148
30
403
548
24.3
143 / 115 / 125
4045624-2
195296
10
452
563
26.7
65/84/98
4045628-2
195297
10
384
508
36.1
84/77/78
Table 26 - Mechanical properties of S355J2 plates
Heat
number
Th
195582
30 .16
195148
C
Mn
Si
S*
P*
Al
N
Cu
Ni
Cr
Mo
Nb
V
Ti
Ceq
1.34 0.427 150 210 .023
31
.034
.036
.03
.002
.032
.004
.018
.39
30 .15
1.38
.344
80 150 .035
33
.021
.016
.024
.002
.035
.002
.021
.39
195296
10 .16
1.47
.346
80 220 .035
34
.018
.015
.025
.001
.001
.001
.001
.41
195297
10 .16
1.44
.334 150 200 .032
29
.016
.013
.022
.001
.001
.001
.001
.41
Table 27 - Nominal composition of S355J2 plates
9.8 Charpy test results
Test Piece
ST008
Room
Temperature
Location
Weld
Metal
Energy
(J)
99
104
Average
(J)
100
98
97
147
Fusion
155
Line
172
79
Weld
68
Metal
80
-20ºC
32
Fusion
37
Line
101
43
Weld
40
Metal
48
-50ºC
27
Fusion
44
Line
36
Table 28 - Charpy results for ST008.
Test
Piece
158
76
57
44
36
Location
Weld
Metal
ST004
Room
Temperature
Fusion
Line
LF+2mm
ST007
LF+5mm
Weld
Metal
Fusion
Line
-20
ST007
LF+2mm
LF+5mm
Weld
Metal
-50
ST007
Fusion
Line
LF+2mm
Energy
(J)
105
100
101
125
121
140
208
200
209
188
192
187
56
65
67
81
88
115
157
132
79
184
181
176
43
47
47
55
44
39
96
104
93
Average
(J)
102
129
206
189
63
95
123
180
45
46
98
99
148
160
164
184
Table 29 - Charpy results for ST004&ST007.
LF+5mm
Test
Piece
Energy
Average
Test
(J)
(J)
Piece
93
Weld
89
91
Metal
91
ST012
65
Fusion
ST013
63
67
Line
73
Room
Temperature
48
LF+2mm
48
48
49
ST015
57
ST014
LF+5mm
53
56
56
77
Weld
69
68
Metal
63
ST012
48
Fusion
59
60
Line
75
-20
ST014
27
LF+2mm
27
28
31
ST015
45
LF+5mm
35
40
41
55
Weld
43
47
Metal
44
ST012
21
Fusion
23
20
Line
16
-50
ST014
19
LF+2mm
17
19
20
ST015
17
LF+5mm
13
17
20
Table 30 - Charpy results for ST013&ST014 and ST012&ST015.
Location
Location
Weld
Metal
Fusion
Line
LF+2mm
LF+5mm
Weld
Metal
Fusion
Line
LF+2mm
LF+5mm
Weld
Metal
Fusion
Line
LF+2mm
LF+5mm
Energy
(J)
89
93
96
137
105
125
108
95
108
107
120
105
65
61
67
40
55
67
55
65
63
70
82
76
39
39
40
21
29
25
24
20
32
31
27
29
Average
(J)
93
123
104
111
64
54
61
76
39
25
25
29
100