Laser Welding Heat Distortions on Thin Plates

INSTITUTO SUPERIOR TÉUniversidade Técnica de Lisboa

Laser Welding

P

Dissertation submitted for obtaining the degree of Master in

Presidente: Professor Doutor Rui M. S. O. Baptista

Orientador: Professora Doutora Maria Luísa Coutinho Gomes de Almeida Quintino

Co-Orientardor Professor Doutor

Vogal: Professora Doutora Rosa Maria Mendes Miranda

INSTITUTO SUPERIOR TÉCNICO niversidade Técnica de Lisboa

Laser Welding Heat Distortions on Thin Plates

Production procedures analysis

Valter Sousa Carrolo


Mechanical Engineering

Juri

Professor Doutor Rui M. S. O. Baptista

Professora Doutora Maria Luísa Coutinho Gomes de Almeida Quintino

essor Doutor Pedro Miguel dos Santos Vilaça Silva

Rosa Maria Mendes Miranda

Outubro de 2010

Distortions on Thin Plates


Professora Doutora Maria Luísa Coutinho Gomes de Almeida Quintino

Silva

I

Acknowledgements

First of all, I would like to thank to Professor Luísa for the opportunity to perform this thesis under her supervision. Her

advises, opinion and support were the beacon that guided me through the entire course of completing this work.

A special thank to Professor Pedro Vilaça, that helped me to overcome my own obstacles and exceed my own

expectations.

I would like to thank you Mr. Phill Carr for the opportunity to perform technical work in his company, Carrs Welding

Technology Lda, and for his hospitality. I also would like to thank everybody in the company for the support, specially to

Alyson, Terry and Alister.

A big thank you is in order to my colleague and dear friend Joaquim Mendes for crucial help with the Matlab’s program

and my dear aunt Carolina for the amazing at reviewing this document.

Finally I would like all my friends and family for supporting me during the whole time.

And of course thank you Sara, for everything.

III

Resumo

Nas ultimas décadas as máquinas laser têm tido grandes desenvolvimentos e são cada vez mais utilizadas nas

mais diversas industrias. A soldadura laser, por ser uma técnica com uma grande produtividade, baixa entrega térmica e

precisão era a solução para alguns dos principais problemas do processo de soldadura. Esta evolução fez com que esta

técnica começasse a ser considerada como um processo de ligação exequível e economicamente viável, em ligações que

antes eram consideradas de impossíveis, como é o caso do equipamento estudado nesta tese, que tem uma espessura

muito pequena.

Qualquer processo de soldadura introduz grandes gradientes térmicos nas peças de trabalho. Pelo que as

distorções térmicas e tensões residuais são um fenómeno sempre presente, que tem que ser encarado e deve ser

analisado e controlado, pois produz efeitos indesejados nas peças. As técnicas de controlo de distorções são sistemas

habitualmente simples mas que são indispensáveis para a produção, porque apesar de não acrescentarem valor ao

produto, asseguram a sua qualidade.

Esta tese foca-se nos processos e procedimentos envolvidos na produção de um permutador de calor,

produzido na empresa Carr’s Welding Technology, com o objectivo de a ajudar a produzir este equipamento com a

qualidade necessária a um preço competitivo.

O permutador de calor é constituído por duas chapas de AISI 304, que serão unidas por soldadura laser. Os

protótipos inicias apresentaram problemas de qualidade de soldadura devido às distorções térmicas.

Foi determinado neste trabalho as melhores condições de soldadura: focagem (+1mm), lente de focagem

(150mm) e velocidade de soldadura (v=0,008m/s), controlar os efeitos adversos das distorções térmicas, utilizando

técnicas de controlo de distorção tais como pre-flexão e aperto. Por fim, a pedido da CWT, estudou-se o padrão de

soldadura, concluí-se que o padrão actual era muito conservativo pelo que poderiam ser utilizados outros padrões mais

esparsos sem comprometer a integridade da estrutura final.

Palavras-chave

Soldadura laser, distorções térmicas, controlo de distorção, qualidade, preço

V

Abstract

Over the years, laser welding machines have developed and increased its use through the industry. Laser beam

(LBW) technology with its high production capacity, precision and low heat delivery, was the solution to many welding

problems. Such evolution only lead to more complex problems, by allowing faster welding processes and stronger welds

on thinners components. Since welding processes exposes the work piece to high temperatures, weld-induced

distortions is always present. In this work it will be investigated how distortion affects the weld quality and present

several methods to reduce its negative effects. These methods are often simple, but without them the plates could not

be welded within its conformity.

This thesis focuses on the processes and procedures involved in a water jacket production at Carr’s Welding

Technologies. The water jacket is essentially composed by two plates made of AISI 304. The preliminary work had low

weld quality, geometrical tolerance faults and aesthetic, it is in this context that this thesis appears. The final result is to

allow its production with repeatable quality and a competitive cost.

After analyzing the prototypes, it was clear that the problem in the welding process was the heat distortions.

The plates were moving defocusing the laser (warping) and were separating; consequently the welds were not reaching

the penetration necessary to resist the insufflation process.

In this work is was determined the best welding conditions: focus (+1 mm), focus lens (150mm) and welding

speed (v = 0.008 m / s), to control the adverse effects of thermal distortion, was overcome by applying distortion

control techniques such as pre-bending and clamping. Finally, as a request of CWT, it was studied the welding pattern. It

was concluded that the current pattern was too conservative therefore other patterns could be used, more sparse,

without compromising the integrity of the final structure.

.

Key-words

Laser beam weld, heat distortions, distortions control, weld quality, productivity

VII

Contents

ACKNOWLEDGEMENTS...................................................................................................................................................... I

RESUMO............................................................................................................................................................................ III

PALAVRAS – CHAVE........................................................................................................................................................... III

ABSTRACT.......................................................................................................................................................................... V

KEY- WORDS...................................................................................................................................................................... V

LIST OF FIGURES................................................................................................................................................................ X

LIST OF TABLES................................................................................................................................................................. XII

LIST OF ABBREVIATIONS................................................................................................................................................... XIII

1. INTRODUCTION ................................................................................................................................................... - 1 -

1.1. Overview ..................................................................................................................................................... - 1 -

1.2. Component Specifications .......................................................................................................................... - 1 -

1.3. Motivation and Objectives .......................................................................................................................... - 2 -

1.4. Thesis Structure .......................................................................................................................................... - 2 -

2. STATE OF ART ...................................................................................................................................................... - 4 -

2.1. Laser welding .............................................................................................................................................. - 4 -

2.1.1. Historical note ..................................................................................................................................... - 4 -

2.1.2. Laser Physics ....................................................................................................................................... - 4 -

2.1.3. Laser Process and Parameters ............................................................................................................ - 6 -

2.1.4. Advantages and limitations of laser welding ...................................................................................... - 9 -

2.1.5. Machine types ................................................................................................................................... - 11 -

2.1.6. Shielding gas ...................................................................................................................................... - 12 -

2.2. Stainless steel ............................................................................................................................................ - 13 -

2.2.1. Overview of Stainless steel ............................................................................................................... - 13 -

2.2.2. Types of Stainless Steel ..................................................................................................................... - 13 -

2.2.3. Austenitic stainless steel characterization ........................................................................................ - 14 -

2.2.4. AISI 304 ............................................................................................................................................. - 18 -

2.3. Residual stresses and distortions .............................................................................................................. - 19 -

2.3.1. Introduction ...................................................................................................................................... - 19 -

2.3.2. Mechanism ........................................................................................................................................ - 20 -

2.3.3. Residual stress consequences and parameters ................................................................................ - 21 -

2.3.4. Residual stress and distortion in the water jacket ............................................................................ - 23 -

2.3.5. Residual stress and distortion measuring techniques ....................................................................... - 25 -

2.3.6. Distortion control Techniques ........................................................................................................... - 25 -

2.4. FEM analysis .............................................................................................................................................. - 25 -

2.4.1. Introduction ...................................................................................................................................... - 25 -

2.4.2. Advantages of prediction weld distortion and residual stress .......................................................... - 26 -

3. EXPERIMENTAL METHOD .................................................................................................................................. - 27 -

3.1. Location ..................................................................................................................................................... - 27 -

3.2. Material and equipment ........................................................................................................................... - 27 -

3.2.1. Testing Plates and sample preparation ............................................................................................. - 27 -

3.2.2. Laser Machine ................................................................................................................................... - 28 -

3.2.3. Workstation ...................................................................................................................................... - 30 -

3.3. Experimental procedures .......................................................................................................................... - 31 -

4. TRIALS RESULTS PRESENTATION AND ANALYSIS ............................................................................................... - 35 -

4.1. Distortion and Weld quality analysis ........................................................................................................ - 35 -

VIII

4.1.1. Welding sequence test ...................................................................................................................... - 35 -

4.1.2. Distortion Control Techniques Test................................................................................................... - 36 -

4.1.3. Results analysis ................................................................................................................................. - 38 -

4.2. Welding procedures optimization ............................................................................................................ - 39 -

4.2.1. Lens Test............................................................................................................................................ - 39 -

4.2.2. Collimation test ................................................................................................................................. - 41 -

4.2.3. Results Analysis ................................................................................................................................. - 42 -

4.3. Tensile test ................................................................................................................................................ - 42 -


4.4. Client’s Quality Test .................................................................................................................................. - 43 -


5. FINITE ELEMENTS ANALYSIS .............................................................................................................................. - 45 -

5.1 Program used ............................................................................................................................................ - 45 -

5.1. Pattern analysis ......................................................................................................................................... - 45 -

5.1.1. Problem resume ................................................................................................................................ - 45 -

5.1.2. Objective ........................................................................................................................................... - 46 -

5.1.3. Model’s structure .............................................................................................................................. - 46 -

5.1.4. Element type ..................................................................................................................................... - 46 -

5.1.5. Geometry and mesh .......................................................................................................................... - 47 -

5.1.6. Boundary conditions ......................................................................................................................... - 47 -

5.1.7. Welds’ stress ..................................................................................................................................... - 48 -

5.1.8. Results ............................................................................................................................................... - 48 -

5.2. Welding distortion analysis ....................................................................................................................... - 51 -

5.2.1. Problem Resume ............................................................................................................................... - 51 -

5.2.2. Objective ........................................................................................................................................... - 51 -

5.2.3. Model’s structure .............................................................................................................................. - 51 -

5.2.4. Element Type .................................................................................................................................... - 52 -

5.2.5. Geometry .......................................................................................................................................... - 54 -

5.2.6. Thermal boundary conditions ........................................................................................................... - 54 -

5.2.7. Structural boundary conditions ........................................................................................................ - 55 -

5.3. Result presentation ................................................................................................................................... - 55 -

5.3.1. Thermal Results ................................................................................................................................. - 55 -

5.3.2. Structural results ............................................................................................................................... - 57 -

5.4. FEM Result analysis ................................................................................................................................... - 67 -

6. CONCLUSIONS ................................................................................................................................................... - 69 -

7. FUTURE WORK ................................................................................................................................................... - 71 -

8. REFERENCES ...................................................................................................................................................... - 73 -

9. ANNEXES .............................................................................................................................................................. - 1 -

9.1. DCT Machine ............................................................................................................................................... - 1 -

9.2. MATLAB’s Programme ................................................................................................................................ - 2 -

IX

List of figures

Figure 1- Technical drawing of the water jacket................................................................................................................ - 1 -

Figure 2- Weld geometry ................................................................................................................................................... - 2 -

Figure 3- Diagram of stimulated emission [2].................................................................................................................... - 5 -

Figure 4- Resonant cavity functioning [2] .......................................................................................................................... - 5 -

Figure 5- Diagram synthesizing the formation of a laser beam (all stages occur simultaneously [21] ............................. - 6 -

Figure 6 - Welding in thermal conduction process and in keyhole process [12] ............................................................... - 7 -

Figure 7- Laser beam geometry [12] .................................................................................................................................. - 7 -

Figure 8 - Graphic of Absorptivity of stainless steel with wavelength of the laser – Reference [7] .................................. - 8 -

Figure 9- Cr-Ni Binary Diagram ........................................................................................................................................ - 15 -

Figure 10- Grain decohesion caused by sensitzation ....................................................................................................... - 16 -

Figure 11- Carbide precipitation ...................................................................................................................................... - 16 -

Figure 12- Sensitization critical area depending on temperature and Wt% C ................................................................. - 17 -

Figure 13- Heat distribution in LBW................................................................................................................................. - 20 -

Figure 14- Diagram of residual stress mechanism [3] ..................................................................................................... - 21 -

Figure 15- Thermal stress distribution before, during and after the welding [3] ............................................................ - 22 -

Figure 16- Thermal stress after the welding [3] .............................................................................................................. - 23 -

Figure 17- Example of heat induced deformation in the work piece .............................................................................. - 23 -

Figure 18 – Residual stress distribution depending on the Radius of the circular weld [3] ............................................ - 24 -

Figure 19- Loss of stability of (left) the plate (center) inner circle (right) Outer contour [3] .......................................... - 24 -

Figure 20 – Effect of upsetting in a sheet plate ............................................................................................................... - 25 -

Figure 21- Carr’s welding technology factory .................................................................................................................. - 27 -

Figure 22 – Plates used in distortion test ........................................................................................................................ - 28 -

Figure 23 – Plates used in the welding procedures optimization .................................................................................... - 28 -

Figure 24 – Plates used in the Tensile test ...................................................................................................................... - 28 -

Figure 25- Graphic velocity vs. Penetration [17] ............................................................................................................. - 29 -

Figure 27- Example of welding setup configuration ........................................................................................................ - 30 -

Figure 26- KR16-2............................................................................................................................................................. - 30 -

Figure 28- Workstation table ........................................................................................................................................... - 31 -

Figure 29- Auxiliary television .......................................................................................................................................... - 31 -

Figure 30 – Distortion analysis test set up ....................................................................................................................... - 32 -

Figure 31 – welding procedures optimization set up ...................................................................................................... - 32 -

Figure 32- Sequence schematic on left row sequence (A) on the right columns sequence (B)....................................... - 35 -

Figure 33- Result presentation of the DCT testes ............................................................................................................ - 38 -

Figure 34 – Walls interference with the laser beam ........................................................................................................ - 42 -

Figure 35- Graphic Strength vs. Velocity .......................................................................................................................... - 43 -

Figure 36- Plates after quality test Front View ................................................................................................................ - 44 -

Figure 37- Ansys Logo ...................................................................................................................................................... - 45 -

Figure 38 – Welding pattern ............................................................................................................................................ - 46 -

X

Figure 39 – Pattern analysis algorithm ............................................................................................................................ - 46 -

Figure 40 – Element input (left) and output (right) [31] .................................................................................................. - 47 -

Figure 41- Plate geometry (left) and mesh (right) ........................................................................................................... - 47 -

Figure 42 - Boundary conditions ...................................................................................................................................... - 48 -

Figure 43 – Plate on edge of entering the plastic domain ............................................................................................... - 49 -

Figure 44 – Displacementon on the plate- Max displacement 3,2mm ............................................................................ - 49 -

Figure 45 – Top view f plate stress .................................................................................................................................. - 49 -

Figure 46 – Bottom view of plate – Maximum stress 254MPa ........................................................................................ - 50 -

Figure 47 – Graph Production time and Safety Factor Vs Welding Pattern .................................................................... - 51 -

Figure 48- welding distortion analysis algorithm ............................................................................................................. - 52 -

Figure 49- Element input ................................................................................................................................................. - 53 -

Figure 50- Element output [30] ....................................................................................................................................... - 53 -

Figure 51 – Composition of the model’s geometry ......................................................................................................... - 54 -

Figure 52 – Thermal constraints ...................................................................................................................................... - 54 -

Figure 53- Temperature distribution in the plate during the weldng process ................................................................ - 55 -

Figure 54 – Temperature in the center of the weld ........................................................................................................ - 56 -

Figure 55 – Temperature in a division of the weld bead ................................................................................................. - 56 -

Figure 56 –Displacement of the model with displacement constraints .......................................................................... - 57 -

Figure 57- Residual stress in the model ........................................................................................................................... - 58 -

Figure 58 – Displacement NO DCT Model and Model unreformed ................................................................................. - 59 -

Figure 59 – side view of the model Out-of-plane displacement ...................................................................................... - 59 -

Figure 60 – Pre-bending structural constraints ............................................................................................................... - 60 -

Figure 61 – Displacement sum in Pre-bending simulation .............................................................................................. - 60 -

Figure 62 – Out-of-plane displacement in pre-bending simulation ................................................................................. - 61 -

Figure 63 - Displacement sum in Pre-cambering simulation ........................................................................................... - 61 -

Figure 64 - Out-of-plane displacement in pre-cambering simulation ............................................................................. - 62 -

Figure 65 - Out-of-plane displacement in pre-cambering simulation (detail) ................................................................. - 62 -

Figure 66 – Displacement vector Sum in plate with pre-stress at 200MPa ..................................................................... - 63 -

Figure 67- Out-of-plane displacement in uniform pre-stress simulation at 200MPa ...................................................... - 63 -

Figure 68 - Displacement vector Sum in plate with pre-stress at 243MPa ...................................................................... - 64 -

Figure 69 - Out-of-plane displacement in uniform pre-stress simulation at 243MPa ..................................................... - 64 -

Figure 70- Displacement sum before releasing DCT (left) after releasing DCT (right) ..................................................... - 65 -

Figure 71 – Out-of-plane displacement in longitudinal pre-stress .................................................................................. - 65 -

Figure 72- Displacement sum on longitudinal stress II .................................................................................................... - 66 -

Figure 73- Out-of-plane displacement in longitudinal pre-stress II ................................................................................ - 66 -

Figure 74 – DCT’s distortion value comparison ............................................................................................................... - 67 -

XI

List of Tables

Table 1- Comparison of laser machine with competing technology [21] ........................................................................ - 10 -

Table 2- Influence of shielding gas on weld bead properties [12] ................................................................................... - 12 -

Table 3- Properties of AISI 304 [31] ................................................................................................................................. - 18 -

Table 4 – Structural and Thermal Properties Depending on the temperature [31] ........................................................ - 19 -

Table 5- miscellaneous properties [31] ........................................................................................................................... - 19 -

Table 6- Properties of the laser machine ......................................................................................................................... - 28 -

Table 7- summary of the tests performed ....................................................................................................................... - 34 -

Table 8- Results presentation of sequence test .............................................................................................................. - 36 -

Table 9- Result penetration of the lens test .................................................................................................................... - 41 -

Table 10 – Focal lens analysis resume ............................................................................................................................. - 41 -

Table 11- Result presentation of the collimation test ..................................................................................................... - 42 -

Table 12- Tensile strength of the welds ........................................................................................................................... - 43 -

Table 13 – Weld’s strength calculations data .................................................................................................................. - 49 -

Table 14 – Parametres influenced by th welding pattern ............................................................................................... - 50 -

Table 15 – Simulation result’s resume ............................................................................................................................. - 67 -

XII

List of abbreviations

LBW - Laser Beam Welding

DCT - Distortion Control Techniques

FEA - Finite Elements Analysis

FEM- Finite element model

CWT- Carr’s welding technology, Ltd

Nd:YAG - neodymium-doped yttrium aluminium garnet

WPO - Welding procedures optimization

Fc - Focal lens

1. Introduction

1.1. Overview

One of the biggest problems of welding is heat distortion. Although the l

concern with heat distortion due to the

allowed Engineers to consider smaller thickness of plates to weld.

Weld induced residual stress and distortion is among the most studied subjects for welded structures. The loca

heating and non-uniform cooling during welding results in a complex distribution of the residual stress in the joint

region, as well as the often undesirable deformation or distortion of the welded structure. As residual stress and

distortion can significantly impair the performance and relia

with during design, fabrication and in-service use of the welded structures.

All studies compiled on this subject along the years agree o

and heat distortion therefore; this thesis is

A review of the literature shows

concentrate to the ones that could be applied in specific pr

to consider their applicability due to the plates

Since the early 1990s, considerable progress has been made on welding residual stress and disto

techniques have improved significantly and, more importantly, the development and application of computational

welding mechanics methods have been remarkable due to the explosive growth in computer capability and to the

equally rapid development of numerical methods

obtained during the trials since the information that could be extracted in the trials was limited

1.2. Component Specifications

The component under analysis

of cheese. Figure 1 is the technical drawing of the product to be manufactured.

Figure

- 1 -

One of the biggest problems of welding is heat distortion. Although the laser is an exceptional technology in

the low energy delivery, new advances generates new problems because evolution

ngineers to consider smaller thickness of plates to weld.

residual stress and distortion is among the most studied subjects for welded structures. The loca

uniform cooling during welding results in a complex distribution of the residual stress in the joint


nificantly impair the performance and reliability of the welded structure;

service use of the welded structures.

subject along the years agree on one fact: it is impossible to

thesis is going to focus in controlling techniques.

shows a number of articles and patents related to theme, however the author

nes that could be applied in specific production environment at Carr’s W

the plates’ dimensions.

Since the early 1990s, considerable progress has been made on welding residual stress and disto



lopment of numerical methods therefore, in this study FEM analysis is used to corroborate the results

information that could be extracted in the trials was limited

omponent Specifications

s is a water jacket. This component is part of a tank that is

technical drawing of the product to be manufactured.

Figure 1- Technical drawing of the water jacket

aser is an exceptional technology in

new problems because evolution

residual stress and distortion is among the most studied subjects for welded structures. The localized

uniform cooling during welding results in a complex distribution of the residual stress in the joint


they must be properly dealt

s impossible to avoid residual stress

ents related to theme, however the author had to

oduction environment at Carr’s Welding Technology and had

Since the early 1990s, considerable progress has been made on welding residual stress and distortion, measurement



in this study FEM analysis is used to corroborate the results

information that could be extracted in the trials was limited.

water jacket. This component is part of a tank that is used in the production

The water jacket is composed by two stainless stee

4mm thick. Both plates have 1250mm width and 3000mm long

the plates together. As figure 2 shows, the plates

The following step is to bend the plates

phase the plates are pressurized and deform

Although the average working pressure is 4 bar the

welds are detected. The first prototype

1.3. Motivation and Objectives

Laser beam welding (LBW) is one of the best processes to do permanent joints

high welding velocity with higher energy efficiency,

Although the heat input of the laser machine is low, the distortion that it causes in thin pl

work has enabled the author to work

advanced machines.

With the knowledge accumulated from previous studies in welding distortion control and the exper

tests performed in the Company, the primary objective of this work is to help Carr's Welding Technologies Ltd to

produce water jackets, the component described in Chapter 1.2

parameters were tested, also welding sequences and distortion co

weld strength, while inducing the minimum distortion.

pattern to determine the reason why it was

could be changed to improve productivity.

1.4. Thesis Structure

This thesis is divided in six chapters.

The first chapter is the Introduction where is presented the overview, the component specifications, motivation and

objectives.

The second chapter is the state of art regarding laser welding physics and technol

general properties, performance and problems and then focus on the stainless steel AISI 304. Still in this chapter it is

- 2 -

is composed by two stainless steels plates (AISI 304), one with 1,25mm

. Both plates have 1250mm width and 3000mm long. The first stage of the manufacturing process is welding

As figure 2 shows, the plates are welded in the edges and in the center.

Figure 2- Weld geometry

The following step is to bend the plates into the position shown in figure 1. The last step is

are pressurized and deformed, specially the thinner plate, the final result is

Although the average working pressure is 4 bar the pressure applied is 25 bar, it is in this step that

he first prototypes sent to be tested did not present the needed weld quality

Motivation and Objectives

Laser beam welding (LBW) is one of the best processes to do permanent joints due to its

high welding velocity with higher energy efficiency, it allows deeper penetrations in one pass and high productivity.

Although the heat input of the laser machine is low, the distortion that it causes in thin plates

to work for a company that is a world leader in laser welding

he knowledge accumulated from previous studies in welding distortion control and the exper

he primary objective of this work is to help Carr's Welding Technologies Ltd to

ponent described in Chapter 1.2 always having in mind.

welding sequences and distortion controlling techniques with the objective of

the minimum distortion. Finally as a request from CWT it was analyzed the welding

pattern to determine the reason why it was chosen and verify if it should be changed for structural security reasons or

could be changed to improve productivity.

This thesis is divided in six chapters.


The second chapter is the state of art regarding laser welding physics and technology; it also describes stainless steels


one with 1,25mm and the other with

stage of the manufacturing process is welding

The last step is insufflation, in this

the final result is shown in figure 36.

this step that, possible defected

id not present the needed weld quality.

due to its , high seam quality,

ions in one pass and high productivity.

ates is considerably high. This

for a company that is a world leader in laser welding and work with highly

he knowledge accumulated from previous studies in welding distortion control and the experimental

he primary objective of this work is to help Carr's Welding Technologies Ltd to

always having in mind. Different laser welding

ntrolling techniques with the objective of optimize

Finally as a request from CWT it was analyzed the welding

and verify if it should be changed for structural security reasons or


ogy; it also describes stainless steels


- 3 -

given a small theoretical background that intends to help the reader to understand the experimental tests done for this

work and the conclusions extracted.

Chapter three gives a detailed description of the experimental procedures, material and parameters used and objective

of the test.

In chapter four is where the results are summarized and analyzed.

In chapter five is where the finite element analysis is summarized.

In chapter six are presented the final conclusions and benefits from this study and future research possibilities.

Chapter seven presents future developments to this work, suggestions for future work.

Chapter eight is for references used in the research and preparation for this dissertation.

- 4 -

2. State of art

This chapter introduces concepts in order to offer background knowledge that intent to help the reader to

follow the subjects discussed in this thesis.

2.1. Laser welding

2.1.1. Historical note

The first investigations that led to the invention of laser where made by Albert Einstein in 1917, the work of

Max Planck and Niels Bohr where also extremely important to understand the physics of lasers .In 1958 A. L. Schawlow

and C. H. Townes published the first article about lasers. Two years later, in 1960 Theodore Maimann built the first laser

machine.

Since industrial lasers were first introduced nearly 40 years ago, revolutionary developments in laser welding

technology have enabled new and innovative joining applications. In the early years, lasers were used primarily for

exotic applications; however advancements in the power and the beam quality of lasers made deep-penetration

keyhole welding possible and expanded the possibilities. Pulsed Nd: YAG lasers that followed in the early 1990s

provided means of welding highly reflective materials like aluminum and copper. Lamp-pumped Nd: YAG lasers reached

the multikilowatt level toward the end of the 1990s and made their entrance into high-production welding and cutting

applications. With the advancements in fiber-optic cables, manufacturers could make use of robotic automation like

never before.

Today, laser welding is a full-fledged part of the metalworking industry, routinely producing welds for common items

such as cigarette lighters, watch springs, motors, hermetic seals, battery and pacemaker cans and hybrid circuit

packages. Yet a review of the literature shows that there has not been enough investigation in this field. This aspect is

noticeable in the use of laser machines in the industrial environment as its full potential has not been explored yet , not

entirely because of the high initial costs, but mainly because most engineers still do not recognize the full potential of a

laser machine.

2.1.2. Laser Physics

Laser stands for Light Amplification by Stimulated Emission of Radiation. A laser is essentially a device which

generates a narrow beam of light sufficiently strong to drill, cut or weld. The physics principles behind the lasers can be

resumed into: excitable atoms, whose electrons are excited by an appropriate stimulus to an abnormally high level of

energy E2 (stimulated state) from which they fall spontaneously either to a normal level (base state) or intermediate

level stimulated energy lever, E1<E2; at each fall an photon is emitted in a random direction. [2]

- 5 -

There is no physical distinction between the exciting photon and the emitted photon; they all have the same phase and

wavelength.

� =��

�= ℎ� (1)

h is Planck’s constant (6.62606896(33) ×10−34

J.s)

c is light velocity (3x108 ms

-1)

�is wave length (m)

�is frequency of the radiation (s-1

)

Figure 3- Diagram of stimulated emission [2]

To make the stimulated emission effective the number of stimulated atoms must be superior to the number of

atoms on the base level. An inversion of the population is therefore sought.

Figure 4- Resonant cavity functioning [2]

Also the interaction time is artificially increased or maintained by increasing the length of the trajectory by

reflecting the beam on itself, in order cross and re-crosses the excitation tube with mirrors, in the resonant cavity. The

energy is not 100% converted into the laser beam and the excess energy has to be removed by a cooling system.

Figure 5- Diagram synthesizing

The laser beam properties are:

- Monochromatic, all photons h

- Coherent, all photons are in the same phase

- Focused, all photons have the same direction

A typical laser installation includes the following components

- Laser beam source

- Devices for guiding, shaping and focusing the laser be

- Devices to create relative movement between the laser beam and the work piece

- Fixtures to hold the piece

- Cooling system

- Control systems

-

2.1.3. Laser Process and Parameters

There are several factors that influence the weld closely related with the laser beam

know all parameters that can be controlled and to understand how they affect the welding quality.

The power output density determines if the weld is made through the thermal conduction process or, if the

density of energy is high enough (>106 W/cm²), deep welding also known as keyhole.

- 6 -

synthesizing the formation of a laser beam (all stages occur simultaneously

hotons have the same wavelength

Coherent, all photons are in the same phase

Focused, all photons have the same direction

A typical laser installation includes the following components

Devices for guiding, shaping and focusing the laser beam

Devices to create relative movement between the laser beam and the work piece

Laser Process and Parameters

There are several factors that influence the weld closely related with the laser beam


output density determines if the weld is made through the thermal conduction process or, if the

gh enough (>106 W/cm²), deep welding also known as keyhole.

simultaneously [21]

There are several factors that influence the weld closely related with the laser beam itself. It is essential to


output density determines if the weld is made through the thermal conduction process or, if the

Figure 6 - Welding in thermal conduction process and in keyhole process [12]

With thermal conduction welding the material is melted only on the surface, the depth of

tenth of millimeters.

By heating the spot of laser focus above the boiling point, a vaporize

with ionized metallic gas and becomes an effective absorber, trapping about 95 percent of the

cylindrical volume, known as a keyhole. Temperatures within this keyhole can reach as high as 25,000 °C, making the

key holing technique very efficient. Instead of heat being conducted mainly downward from the surface; it is conducted

radically outward from the keyhole, forming a molten region surrounding the vapor. As the laser beam moves along the

work-piece, the molten metal fills in behind the keyhole and solidifies to form

penetration at high welding speed. In order

high and the focal diameter very small, around the hundreds of microns.

The focus of beam is a major influence of the power density; since the smallest diamete

in the focal diameter (highest power density)

mentioned is crucial to the weld properties.

- 7 -

Welding in thermal conduction process and in keyhole process [12]

ith thermal conduction welding the material is melted only on the surface, the depth of

By heating the spot of laser focus above the boiling point, a vaporized hole is formed in the metal, then, t

with ionized metallic gas and becomes an effective absorber, trapping about 95 percent of the


technique very efficient. Instead of heat being conducted mainly downward from the surface; it is conducted

outward from the keyhole, forming a molten region surrounding the vapor. As the laser beam moves along the

piece, the molten metal fills in behind the keyhole and solidifies to form the weld. This technique allows

peed. In order have high energy density, the output power of the laser must be extremely

high and the focal diameter very small, around the hundreds of microns.

of beam is a major influence of the power density; since the smallest diamete

in the focal diameter (highest power density), the positioning of the focal diameter on the work piece as

is crucial to the weld properties.

Figure 7- Laser beam geometry [12]

Subtitle:

1-Optic fiber cable

2- collimation lens

3- Separating plate

4- O

5- Focusing lens

dk- core diameter of the fiber

dof-

fc- focal length of collimation

Subtitle:

1- Solid melt

2- Melting zone

3- Laser beam

4- Material

5- Vapor channel (keyhole)

6- Escaping metal vapor

ith thermal conduction welding the material is melted only on the surface, the depth of the weld is in the

d hole is formed in the metal, then, this is filled

with ionized metallic gas and becomes an effective absorber, trapping about 95 percent of the laser energy into a


technique very efficient. Instead of heat being conducted mainly downward from the surface; it is conducted

outward from the keyhole, forming a molten region surrounding the vapor. As the laser beam moves along the

the weld. This technique allows deep

have high energy density, the output power of the laser must be extremely

of beam is a major influence of the power density; since the smallest diameter of the beam is located

r on the work piece as previously

Subtitle:

Optic fiber cable

collimation lens

Separating plate

Observation outout

Focusing lens

core diameter of the fiber

focal diameter

focal length of collimation

Melting zone

Laser beam

Vapor channel (keyhole)

Escaping metal vapor

- 8 -

Theoretically for best beam quality the focus must be set on the surface of the work piece, but in reference

[12] it is recommended to weld with the focus a little bellow in order to produce better results.

The focal lens used has a direct relationship with the diameter of the focal diameter.

�� [��] = (�� × 10 !) ×#$�%& &'()*� $+ *�' &'(,

#$�%& &'()*� $+ �$&&-.%*-$( (2)

�� /0�ℎ � � ��/=150mm

�� =200µm

From the formula it is easily understandable that, smaller lens have smaller focal diameter and it is trough the focal

diameter that the focal lens influences the energy density therefore, more penetration.

�/��01 ��/2��1 [3/�5] = 67�� 8��8� × 9:+$�%&;

< (3)

The focal lens also influences the geometry of the beam, which in this situation is highly important. Due to the

positioning of the clamping, the beam’s interference with its walls might be high enough to influence negatively the

quality of the weld, because of the energy wasted.

The laser can be applied in two modes, pulsed mode or continuous mode. Pulse mode is better for cladding and for

regular welding. Continuous mode (CW) is better for welds where it is applied high loads, because there is less stress

concentration and fatigue problems.

In the pulsed mode we must control three out of four parameters to have the pulses with the characteristics

desired - they are frequency of the pulses, pulses length, velocity and overlap. In the CW mode we only control velocity.

In this study all the welds were made in CW mode.

Finally properties like wave length are equally important because different materials have different values of

absorptivity with different sizes of wave length. In this case we used a solid state Nd:YAG laser that emits photons with

1,64nm, this is shown in the graphic below the absorptivity of AISI 304 is around 40%.

Figure 8 - Graphic of Absorptivity of stainless steel with wavelength of the laser – Reference [7]

- 9 -

2.1.4. Advantages and limitations of laser welding

LBW had an extremely positive impact in the manufacturing sector. Materials that were once considered too

exotic, or were too complex can often be cost effectively welded by this method with excellent results. Accuracy,

repeatability, quality, high velocity and automation are already achievable with the laser welding process. Typically it

does not require a shielding gas in less demand weld. This can provide a substantial cost savings verses shielded welding

procedures. Another advantage is that there are no need to apply pressure to the materials being welded. This too can

provide a significant cost savings, in particular when dealing with laser welded parts with complex geometries that are

difficult to clamp.

It is also possible to weld different materials together because of the disk laser to narrow weld geometry.

Usually an extreme increase of hardness in the fusion zone is expected, which can lead to cracks and compromise weld

integrity. Using a filler wire can prevent this problem , achieve a more homogeneous weld and reduce hardness within

the fusion zone. The combination of narrow weld geometry with the appropriate filler wire can eliminate the hardness

spikes and yield crack-free. Laser weld allows minimum weld metal, and minimum heat input, narrow heat affected

zone which makes the distortion caused by this method minor than the competitors. Comparing with electron beam

welding has the advantage of being performed under normal atmosphere.

The ability to uncouple components such as the pump diodes, resonator system and fiber optic cables ensure long term

cost effectiveness and minimum downtime.

On the other hand, the high cooling rate must be carefully controlled in order to avoid unacceptable materials

properties. Cracks and porosity may also occur in certain materials.

Highly reflective materials are difficult to weld because the beam energy is reflected and this is clearly shown in the

low efficiency that characterizes these machines. Because this is a very precise method, manual welding is not

recommended. In order to obtain proper weld, mechanized equipment has to be used and all operations pre-

programmed, this is why LBW is normally used in single production environments.

LBW has strict requirements for the quality of joint preparations and accurate positioning or seam tracking. Surface

coating may result in imperfections.

The biggest disadvantage of the laser machine is its high initially cost and the running cost. This is why laser machines

are usually associated with high production configurations. [21]

- 10 -

The following table, was extracted from Reference 28, it compares LBW with competing processes to weld 6mm thick of

stainless steel.

Criteria LBW TIG Plasma EBW Resistance

Power input

Weld speed

Fit-up

Precision

Special needs

1

1

1

±0,5mm

Protection

from beam

0,5

0,1

0,5

±1mm

Protection

from arc

1

0,4

0,5

±1mm

Protection

from arc

1,2

2

1,2

±0,3mm

Vacuum

chamber

Protection

from X-ray and

magnetic

fields

1,2

-

0,5

±1mm

Access to both

sides of the

plate

Frequent

change/clean

of the

electrodes

Distortion

Axial

angular

minimum

Small heat

affected zone

Parallel sides

Noticeable

V-shape weld

noticeable

noticeable

V-shape weld

noticeable

minimum

Parallel sides

minimum

Finish smooth smooth Slight surface

protrusion

Slight surface

protrusion

smooth

Initial cost

Running cost

1

1

0,1

0,8

0,1

0,9

0,7

1,2

0,8

1,4

Table 1- Comparison of laser machine with competing technology [21]

CWT’s client used to weld the plates with TIG, it is easy to verify in the table above that this is a very slow

process, therefore expensive. As consequence it was more economic to subcontract this part of the production.

Carr's Welding Technologies Ltd, proposes to weld with laser technology. A competitor in Germany uses the

resistance welding. This technology usually does not compete with laser welding, however due to the specificities of the

water jacket, it does and presents better weld quality and finishing.

- 11 -

2.1.5. Machine types

There are many different types of laser welding machines. By properly defining requirements and expectations,

along with specific operating conditions and welding parameters, we will be able to choose the correct laser welding

machine.

Based on the state of active specimen, laser machine can be divided in three basic types. These are solid state lasers,

liquid and gas lasers. For lower powered, thinner metals , pulsed solid state lasers are used. If the part to be welded is

thicker and requires a higher power input, gas lasers have preference.

Normally laser machines are divided in the following categories:

- Nd:YAG(neodymium : yttrium aluminum garnet) laser (solid state);

Probably the most popular type of solid state laser is a lower powered laser. It is possible to operate them in a

continuous mode however they are mostly used in pulse mode for welding thin materials or performing low penetration

welds. This is typically accomplished with a diode light source and crystal rods. The light is transmitted into the crystal

rod and in turn generates a laser beam with a wavelength of 1.064 microns that is pointed at the material to be welded.

Non-metal materials are invisible to this laser, it makes impossible to weld plastic. Solid state laser beam welders do

have one distinct disadvantage though, and that is the fact that they operate at wavelengths that can damage the

operators eyes. For this reason, the operator must use some type of eye protection to shield their eyes to avoid harm.

- CO2 laser(gas state)

It has been around the longest and is one of the most reliable types of laser welders. CO2 laser welding is often used

when full penetration high power welds are required. An added advantage of the CO2 laser welding machine is that it

does not produce harmful ultraviolet rays when welding .This eliminates potential harm to the operator since the wave

length of the beam is 10.64 microns. The more common gas type laser is the CO2 laser. These lasers are powered with a

high voltage and low current, and can have power outputs that are over 25 kilowatts. These laser beam welders are

normally used in continuous mode, although they are capable of pulsed mode too. The obvious advantage to the higher

power output provided by a gas laser is that it is capable of welding materials that are much thicker than a solid state

laser can weld, still the efficiency of the machine is not very good (5-8% ) and a significant part of the beam is reflected

by metals. Other benefit of a gas type laser is that it operates at a longer wavelength than solid state lasers, thus the

potential for retina damage is eliminated.

- Fiber laser

The newest type of laser welding machine utilizes what is known as a fiber laser. And as the name implies, fiber optics

are used to generate a laser beam that is used to perform the weld. The benefits of the fiber laser are that the beam

quality is vastly improved over the other methods.

- Disk laser

A new twist on Nd:YAG laser, it is what is known as a disk laser. Disk lasers combine the pumping efficiency of a

semiconductor laser with the beam quality of a solid-state laser. The difference between this and the standard Nd:YAG

laser is that the light from the diode is transferred into a disk instead of a rod. The quality of the laser beam is improved

with this method versus using a rod.

- 12 -

2.1.6. Shielding gas

The main function of the shielding gas is to reduce the welding pool’s contact with the atmosphere. The

protection is necessary because metals at high temperature are highly reactive and tend to form oxides that lead to

defects, such as porosity, and it decreases the welding seam quality. The shielding gas also interferes in the shape of the

seam and the penetration depth. As they help to create the plasma they make the welding process easier and overall

better.

There are three main shielding gas recommended by the manufacturer: Helium (He), Argon (Ar) and Nitrogen (N2).

- Nitrogen (N2) is a colorless, odorless and inert gas, well suited for welding Chrome-nickel Steel. When welding

with nitrogen it is important to remember that its use will decrease stainless protection by lowering the

Chrome-Nickel fractions.

- Argon (Ar) is a colorless, odorless and inert gas. It is inflammable and non toxic. It can be used as shielding gas

for chrome-nickel steel. Being the heaviest of them all, its use is recommended on devices where gas can

escape easily.

- Helium (He) is a colorless, odorless and inert gas. It is the lightest gas and escapes quickly, therefore, its

consumption is higher than the other gases. Its use is suited for aluminum (Al) and Al alloys. Helium can be

used as addition when welding with nitrogen.

The following table resumes the information given by the manufacturer reference [12] about the effect of shielding

gas in the welding process.

Argon (A) Helium (He) Nitrogen (N2) without shielding gas

Seam

form

B=width

T=depth

Seam

surface Very Good Good Good Bad

Splashes Good Good Neutral Bad

Blowholes Very Good Good Good Bad

Costs Bad Bad Neutral Good

Table 2- Influence of shielding gas on weld bead properties [12]

The shielding gas is not mandatory in the laser weld process, this way, the running cost is reduce, however the

weld quality maybe compromised, as shown in the table above.

Nevertheless, in more demanding welds the shielding gases is always used since it helps to achieve higher weld seam

quality.

Based on the information given by the manufacturer, in this case, the shielding gas more suitable for the given case,

given the material AISI 304 is Argon. Therefore, this was the shielding gas used in all welding trials.

- 13 -

2.2. Stainless steel

2.2.1. Overview of Stainless steel

The New York Times announced in 1915 the development of stainless steel, however the corrosion resistance of

iron-chromium alloys was first recognized in 1821 by the French metallurgist Pierre Berthier, he noticed their resistance

against attack by some acids and suggested their use in cutlery.

In the late 1890s, Hans Goldschmidt of Germany developed an aluminothermic (thermite) process for producing

carbon-free chromium. In the years 1904–1911, several researchers, particularly Leon Guillet of France, prepared alloys

that would today be considered stainless steel.

On October 17, 1912, Krupp Engineers Benno Strauss and Eduard Maurer patented austenitic stainless steel.

Nowadays stainless steel is used for many industrial, architectural and chemical applications, they are

developed to work not only under corrosive conditions but also in extreme temperature situations. From low

temperatures (cryogenic range) where they exhibit high toughness, to high temperatures where still present a good

oxidation resistance. As stainless steels are non-magnetic (due to the presence of nickel), sometimes they are used in

applications where other alloys cannot be accepted.

In metallurgy, stainless steel, is also known as inox steel and it is by definition steel alloy with a minimum of 10-

11% chromium(content by mass) and up to 26%. When nickel is added, for instance, the austenite structure of iron is

stabilized. This crystal structure makes such steels non-magnetic and less brittle at low temperatures. For greater

hardness and strength, more carbon is added. Significant quantities of manganese are also used in many stainless steel

components because it preserves an austenitic structure in steel as it does with nickel but at lower cost. Ferrite is

important in avoiding hot cracking during cooling from welding of austenitic stainless steels.

Chromium forms a passivation layer of chromium (III) oxide (Cr2O3) that is impervious to water and air,

protecting the metal beneath when exposed to oxygen atmosphere. The layer is too thin to be visible, so the metal

remains lustrous. This layer quickly reforms when the surface is scratched. This phenomenon is called passivation and is

seen in other metals, such as aluminium and titanium.

Stainless steel is a material that combines good mechanical resistance with corrosion resistance, it has low

maintenance, relatively low cost and luster that makes it an ideal base material for a host of commercial applications.

It has no transformation point, therefore it cannot be hardened on the solid phase, it is ductile and have a good

resilience even at low temperature. Their weld ability is good as it does not need pre-heating. It can be hardened by

cold work without becoming fragile. All stainless steels have a work hardening rate higher than carbon steel.

Finally on an ecological note, stainless steels are 100% recyclable. On average, stainless steel objects are composed of

about 60% recycled material of which 40% originates from end-of-life products and 60% comes from manufacturing

processes.

2.2.2. Types of Stainless Steel

Reference is often made to stainless steel in the singular sense as if it were one material but there are over 150

grades of stainless steel, of which fifteen are most common.

- 14 -

There are different grades and surface finishes of stainless steel to suit the environment to which the material will be

subjected in its lifetime. Three general classifications are used to identify stainless steels. They are: 1. Metallurgical

Structure; 2. The AlSl numbering system: namely 200, 300, and 400 Series numbers; 3. The Unified Numbering System.

When stainless steels are classified by their crystalline structure we have:

• Austenitic, or 300 series, stainless steels make up over 70% of total stainless steel production. They contain a

maximum of 0.15% carbon, a minimum of 16% chromium and sufficient nickel and/or manganese to retain an

austenitic structure at all temperatures from the cryogenic region to the melting point of the alloy.

Superaustenitic stainless steels, exhibit great resistance to chloride pitting and crevice corrosion due to high

molybdenum content (>6%) and nitrogen additions, and the higher nickel content ensures better resistance to

stress-corrosion cracking versus the 300 series. Low-carbon versions, for example 316L or 304L, are used to

avoid corrosion problem caused by welding, the "L" means that the carbon content of the alloy is below 0.03%,

this will reduce the sensitization effect (precipitation of chromium carbides at grain boundaries) caused by the

high temperature produced by welding operation.

• Ferritic stainless steels generally have better engineering properties than austenitic grades, but have reduced

corrosion resistance, due to the lower chromium and nickel content. They are less expensive. They contain

between 10.5% and 27% chromium and very little nickel, if any and some types can contain lead. Most

components include molybdenum and some, aluminium or titanium. Some alloys can be degraded by the

presence of chromium, an intermetallic phase which can precipitate upon welding.

• Martensitic stainless steels are not as corrosion-resistant as the other two classes , they are extremely strong

and tough, as well as highly machineable, and can be hardened by heat treatment. Martensitic stainless steel

contains chromium (12-14%), molybdenum (0.2-1%), nickel (0-<2%), and carbon (about 0.1-1%) (giving it

stronger hardness but making the material more brittle). It is quenched and magnetic.

• Precipitation-hardening martensitic stainless steels have corrosion resistance comparable to austenitic

varieties, but can be precipitation hardened to a even higher strengths than the other martensitic grades.

• Duplex stainless steels have a mixed microstructure of austenite and ferrite. The aim is to produce a 50/50 mix,

although in commercial alloys, the mix may be 40/60 respectively. Duplex steels have improved strength over

austenitic stainless steels and also improved resistance to localized corrosion, particularly pitting, crevice

corrosion and stress corrosion cracking. They are characterized by a higher chromium (19–28%) and

molybdenum (up to 5%) and a lower nickel contents than austenitic stainless steels. Duplex stainless steels

properties are achieved with an overall lower alloy content than similar performing super-austenitic grades ,

making their selection and use cost effective for many applications.

2.2.3. Austenitic stainless steel characterization

Carbon steel on cooling transforms from Austenite to a mixture of ferrite and cementite. In the austenitic

stainless steel, it does not happen due to the high percentage content in nickel that suppresses this transformation

(austenite stabilizers).

The transformation behavior of austenitic stainless steels can be described using the Fe-Cr-Ni pseudo binary

diagram at 70 % constant iron.

- 15 -

Figure 9- Cr-Ni Binary Diagram

Austenitic steels have a F.C.C atomic structure which provides more planes for the flow of dislocations,

combined with the low level of interstitial elements (elements that lock the dislocation chain), gives this material its

good ductility. This also explains why this material has no clearly defined yield point and why its yield stress is always

expressed as a proof stress. Austenitic steels have excellent toughness down to absolute Zero (-273°C), with no steep

ductile to brittle transition.

This material has good corrosion resistance, but quite severe corrosion can occur in certain environment. The

right choice of welding consumable and welding technique can be crucial as the weld metal can corrode more than the

parent material.

Austenitic stainless steels have high ductility, low yield stress and relatively high ultimate tensile strength,

when compare to typical carbon steel.

The thermal cycle caused by welding, have little influence on mechanical properties because a microstructure

transformation does not occur. However strength and hardness can be increased by cold working, which will also

reduce ductility. A fully annealing solution (heating to around 1045°C followed by quenching or rapid cooling) will

restore the material to its original condition, removing alloy segregation, sensitization, sigma phase and restoring

ductility after cold working. Unfortunately the rapid cooling will re-introduce residual stresses, which could be as high

as the yield point. Distortion can also occur if the object is not properly supported during the annealing process.

Austenitic steels are not susceptible to hydrogen cracking (or cold cracking), therefore pre-heating is not

required, except to reduce the risk of shrinkage stresses in thick sections. It is not required post weld heat treatment

either, as this material has a high resistance to brittle fracture; occasionally stress relief is carried out to reduce the risk

of stress corrosion cracking, however this is likely to cause sensitization unless a stabilized grade is used (limited stress

relief can be achieved with a low temperature of around 450°C ).

Possibly the biggest cause of failure in stainless steel is stress corrosion cracking. This type of corrosion forms

deep cracks in the material and is caused by the presence of chlorides in the process fluid or heating water/steam, at a

temperature above 50°C, when the material is subjected to a tensile stress (this stress includes residual stress, which ,

as mentioned before, could be up to yield point in magnitude) and significant increases in Nickel and also Molybdenum

will decrease that risk.

Stainless steel has a very thin and stable oxide film rich in chrome. This film reforms rapidly by reaction with

the atmosphere if damaged. If stainless steel is not adequately protected from the atmosphere during welding or is

- 16 -

subject to a quiet heavy grinding operations, a thick oxide layer will form. This thick oxide layer, distinguished by its

blue tint, will have a chrome depleted layer under it, which will impair corrosion resistance. Both the oxide film and

depleted layer must be removed, either mechanically (grinding with a fine grit is recommended, wire brushing and shot

blasting will have less effect), or chemically (acid pickle with a mixture of nitric and hydrofluoric acid). Once the surface

is cleaned, it can be chemically passivated to enhance corrosion resistance, (passivation reduces the anodic reaction

involved in the corrosion process).

Sensitisation is one of the corrosion mechanisms that causes widespread problems in austenitic stainless steels.

The severity of the problem can cause grain decohesion, as shown in figure below.

Figure 10- Grain decohesion caused by sensitzation

When an autenitic stainless steel is exposed to a temperature between 500ºC and 850ºC for a long period

there is a tendency to precipitate chromium-rich carbides as the alloy enters the carbide plus austenite phase field.

Precipitation of Carbides (most common M23C6 and M7C3) occurs primarily at the austenite grain surfaces which are

heterogeneous nucleation sites.

Figure 11- Carbide precipitation

In the context of welded samples leads to the phenomenon of weld decay. Regions are created in the heat-

affected zones of the welds which precipitate carbides, become sensitized and fail by localized corrosion, almost as if

the weld is unzipped in the sensitized region.

A variety of solutions exist to avoid sensitization, reducing steel carbon content, making it more difficult to

precipitate carbides is quite common. Stainless steels with an 'L' associated with their numerical designation ( 304L and

- 17 -

316L) have been manufactured with low carbon concentrations less than 0.03 wt%, which compares against the normal

grades with a typical 0.08 wt% of carbon.

Figure below shows how carbon accelerates sensitization.

Figure 12- Sensitization critical area depending on temperature and Wt% C

The use of solutes (such as Nb, Ti, V or Ta) can be an alternative which have a greater affinity for carbon than

for chromium. These are called stabilized stainless steels, for example, types 321 (Ti stabilized) and 347 (Nb stabilized)

austenitic stainless steels. Titanium cannot in general be used to make alloys deposited by arc welding because it readily

oxide. Stabilization involves more than just an addition of Nb or Ti. A heat-treatment must be performed to stimulate

the formation of TiC or NbC, by holding at 900°C for one hour and this is because during lower temperature heat

treatments, M23C6 may form faster than TiC or NbC.

Austenitic stainless steel have can be separated according to their carbon content:

- 304 L grade Low Carbon, typically 0.03% Max

- 304 grade Medium Carbon, typically 0.08% Max

- 304H grade High Carbon, typically Up to 0.1%

The higher the carbon content the greater the yield strength, hence the advantage in using stabilised grades.

Typical Alloy Content

- 304 (18-20Cr, 8-12Ni)

- 316 (16-18Cr, 10-14Ni + 2-3Mo) 304 + Molybdenum

- 316 Ti (316 with Titanium Added) 304+Moly+Titanium

- 320 same as 316 Ti -

- 321 (17-19Cr, 9-12Ni + Titanium) 304 + Niobium

- 347 (17-19Cr, 9-13Ni + Niobium) 304 + Extra 2%Cr

- 308 (19-22Cr, 9-11Ni) 304 + Extra 4%Cr + 4% Ni

- 309 (22-24Cr, 12-15Ni) 304 + Extra 4%Cr + 4% Ni

All the above stainless steel grades are, basically, variations of a 304. They are all ready to weld and all have

matching consumables, exception for a 304 which is welded with a 308 or 316 and 321 is welded with a 347 (Titanium

is not easily transferred across the arc) and a 316Ti is normally welded with a 318.

- 18 -

Molybdenum has the same effect on the microstructure as chrome except that it gives better resistance to

pitting corrosion. Therefore a 316 needs less chrome than a 304.

2.2.4. AISI 304

The 300 series designation tells one that the grade is composed basically of 18% chromium and 8% nickel.

Type 304 serves a wide range of applications. It withstands ordinary rusting in architecture, it is resistant to food

processing environments (except possibly for high-temperature conditions involving high acid and chloride contents), it

resists organic chemicals, dyestuffs and a wide variety of inorganic chemicals. Type 304 L (low carbon) resists nitric acid

well and sulfuric acids at moderate temperature and concentrations. It is used extensively for storage of liquefied gases,

equipment for use at cryogenic temperatures (304N), domestic appliances, kitchen equipment, hospital equipment,

transportation, wastewater treatment and other consumer products.

Type 304 (sometimes referred to as 18-8 stainless) is the most widely used alloy of the austenitic group. It has a nominal

composition of 18% chromium and 8% nickel.

By far, the most popular grade of stainless steel is 304. It is non-magnetic and cannot be hardened by heat treatment,

but it can with cold work and without becoming fragile.

The following tables (3,4 and 5) present the chemical, mechanical and thermal properties of AISI 304.

Table 3- Properties of AISI 304 [31]

- 19 -

Table 4 – Structural and Thermal Properties Depending on the temperature [31]

Table 5- miscellaneous properties [31]

All the data is according to standards EN 10088-3: 2005.

2.3. Residual stresses and distortions

2.3.1. Introduction

The residual stresses in a component or structure are stresses caused by incompatible internal permanent

strains. They may be generated or modified at every stage in the component life cycle, from original material

production to final disposal. Welding is one of the most significant causes of residual stresses. It typically produces large

tensile stresses whose maximum value is approximately equal to the yield strength of the materials being joined, and

balanced by lower compressive residual stresses elsewhere in the component.

The distortion causes the degradation of the product performance and the increase of the manufacturing cost

due to the poor fit-up, so that it need be eliminated or minimized below a critical level.

- 20 -

Tensile residual stresses may reduce the performance and ultimately causing failure of manufactured products.

They may increase the rate of damage by fatigue due to the increase of the average stress applied, creep or

environmental degradation. They may reduce the load capacity by contributing to failure by brittle fracture, or cause

other forms of damage such as shape change. Compressive residual stresses are generally beneficial, but cause a

decrease in the buckling load.

The purpose of this chapter is to present a method that simulates such processes involving phase transformation

when considering the effect of the coupling just mentioned.

2.3.2. Mechanism

Four types of distortion induced by welding were discovered. The first two are longitudinal shrinkage and

transverse shrinkage that occur in plane. The other two are angular distortion and longitudinal distortion (bowing),

which appear out of plane. The angular distortion is mainly caused by the non-uniform extension and contraction

through thickness direction due to the temperature gradient. The longitudinal distortion (also called buckling distortion)

is generated by the longitudinal tensile residual stress. [11]

Heat distortions have origin in fast temperature variation that generates non-uniform dilation and

contractions. The localized heating and non-uniform cooling during welding results in a complex distribution of the

residual stress in the joint region, as well as the often undesirable deformation or distortion of the welded structure. A

number of factors influence the residual stress and distortion of a welded structure. They are related to the

solidification shrinkage of the weld metal, non-uniform thermal expansion and contraction of the parent metal, the

internal constraints of the structure being welded, and the external structural restraints of fixtures used in a welding

operation. For many engineering materials, a transient welding thermal cycle also results in micro structural changes in

the joint region, which can further complicate the formation of the residual stress field.[11]

Figure 13- Heat distribution in LBW

∆ = >?(@5 − @B) (4)

l - elongation (m)

? - thermal explasion coeffcient m(m x K)

T – Temperature (K)

- 21 -

The effect of coupling between metallic structures, including the molten state, temperature, and stress or/and

strain occurring in processes accompanied by phase transformation, sometimes play an important role in industrial

processes like welding. Figure 14 represents the schematic features of the effect of metallurgy, thermal and mechanical

coupling in heat distortions phenomena. When the temperature distribution in a material varies, thermal stress is

caused in the body, and the induced phase transformation affects the structural distribution, which is known as melting

or solidification in the solid–liquid transition and pearlite or martensite transformation from austenite in the solid

phase, this is for general cases because as said before austenitic stainless steel does not have phase transformation

since it presents an austenitic microstructure at 25ºC. Still local dilatation due high temperature creates stress and

interrupts the stress or strain field in the body.

In contrast with these phenomena, which are well known in ordinal analysis, arrows in the opposite direction

indicate coupling in the following manner. Part of the mechanical work done by the existing stress in the material is

converted into heat , which may be predominant in the case of inelastic deformation, thus disturbing the temperature

distribution. The acceleration of phase transformation by stress or strain, which is called stress- or strain-induced

transformation ➄, has been discussed by metallurgists as one of leading parameters of transformation kinetics. The

arrow ➅ corresponds to the latent heat due to phase transformation, which is essential to determine the

temperature.[3]

Figure 14- Diagram of residual stress mechanism [3]

2.3.3. Residual stress consequences and parameters

Warping is a common problem experienced in the welding fabrication of thin plate structures. There are

several factors that influence distortion control strategy, they may be categorized into:

- Design-related and process-related variables, that includes weld joint details, plate thickness and thickness

transition if the joint consists of plates of different thickness, stiffener spacing and number of attachments,

corrugated construction, mechanical restraint conditions, assembly sequence and overall construction

planning.

- Welding process, there are important variables

method, travel speed and welding sequence.

The implementation of distortion mitigation techniques can be applied before, during and after the welding and

their objective is to counteract the effects of shrinkage during cooling, which distorts the fabricated structure. These

mitigation techniques include controlled pre

bending, fillet joints, presetting butt joints and using appropriate

mitigation techniques is to balance weld shrinkage

is used to control the cool down velocity and have a more adequate heat flux; heat sinking balances welding heat about

the neutral axis of the joint.

Once the parameters that affect the h

due to the better knowledge of mechanism

general all measures that can be taken to prevent

usage of more material, or more energy

One common problem associated with

is the finished products’ dimensional tolerance an

The pictures below synthesize the typical temperature variation and

welding process.

Figure 15- Thermal stress distribution before, during and after the welding

- 22 -

mportant variables related to the technology used such as

speed and welding sequence.



nclude controlled pre-heating, mechanical tensioning, clamping, thermal tensioning, pre

bending, fillet joints, presetting butt joints and using appropriate heat sinking arrangements. The

o balance weld shrinkage forces as they prevent the typical component distortion. Pre


affect the heat distortion are identified, theoretically they can be

mechanism, less conservative decisions regarding weld

general all measures that can be taken to prevent distortion, add cost to the process. This cost can

more energy consumption.

One common problem associated with welding, which has been made aware and documented for many years,

’ dimensional tolerance and stability.

e typical temperature variation and stresses induced and material

Thermal stress distribution before, during and after the welding

such as heat input, heat delivery



heating, mechanical tensioning, clamping, thermal tensioning, pre-

heat sinking arrangements. The purpose of these

they prevent the typical component distortion. Pre-heating


can be controlled. However,

are considered, because in

cost to the process. This cost can be either by the

and documented for many years,

induced and material behavior in a

Thermal stress distribution before, during and after the welding [3]

Fig

Allowing for residual stresses in the assessment of service performance varies according to the failure

mechanism. Normally, it is not necessary to tak

ductile materials. Design procedures for fatigue or buckling of welded structures usually make appropriate allowances

for weld-induced residual stresses, and hence it is not necessary to include them explicitly.

major effect on fracture in the brittle and transitional regimes, and hence the stress intensity, K, or energy release rate,

J, due to residual stresses must be calculated and included in the fracture assessment. K or J may be obt

distribution function, crack size and geometry by various methods, including handbook solutions, weight functions,

and/or finite element analysis.

2.3.4. Residual stress and distortion in the

Manufacturing of sheet-metal-formed plat

fusion welding is applied. It has been

because of their lower critical compre

distortions in thin-walled structural elements

component in analysis the heat distortions

separation of the plates, therefore, the normal production

distortion removal tasks.

Figure 17-

- 23 -

Figure 16- Thermal stress after the welding [3]

Allowing for residual stresses in the assessment of service performance varies according to the failure

necessary to take into account residual stresses in calculations

. Design procedures for fatigue or buckling of welded structures usually make appropriate allowances

induced residual stresses, and hence it is not necessary to include them explicitly.


J, due to residual stresses must be calculated and included in the fracture assessment. K or J may be obt

, crack size and geometry by various methods, including handbook solutions, weight functions,

and distortion in the water jacket

formed plates, panels and shells are always accompanied by distortions where

It has been identified the problem of buckling distortions in plates with less than 4mm,

of their lower critical compressive stress. Buckling distortions are more marked than other forms of welding

walled structural elements and they are the main problem in sheet metal fabrication

heat distortions are critical because they affect the welding

the normal production procedures have to be broken for time

- Example of heat induced deformation in the work piece

Allowing for residual stresses in the assessment of service performance varies according to the failure the

calculations of the static strength of

. Design procedures for fatigue or buckling of welded structures usually make appropriate allowances

induced residual stresses, and hence it is not necessary to include them explicitly. Residual stresses have a


J, due to residual stresses must be calculated and included in the fracture assessment. K or J may be obtained as a stress

, crack size and geometry by various methods, including handbook solutions, weight functions,

accompanied by distortions where

the problem of buckling distortions in plates with less than 4mm,

re more marked than other forms of welding

in sheet metal fabrication. In the

welding quality by inducing the

procedures have to be broken for time-consuming and costly

- 24 -

Welding-induced buckling differs from bending distortion by its much greater out-of-plane deflections and several

stable patterns. Buckling patterns depend much more on the element’s geometry, and types of weld joint especially

dependent on the thickness of sheet materials under certain conditions, it also depends on rigidity of elements to be

welded and welding heat inputs.

In order to use the best DCT in the work piece it is important to analyze the particularities of it. The plates are

welded together using circular welds with 7mm radius in a pattern with distance of 70mm in width and 80mm in length.

Buckling distortions caused by circular welds in the plates are mainly determined by transverse shrinkage of welds in the

radial direction, whereby compressive stresses are produced in the tangential direction.

The straight- and curve path welding are open-loop paths so that the induced bending distortion is relatively

simple however in the work piece the welds are circular, close-loop path welding, the non-uniformity of bending

distortion is more serious than the former cases. A saddle shape with negative Gaussian curvature is observed.

On circular welds the bending direction changes along the path, due to the adjacent bending effect which is induced by

moving heat source, and also due to the direction change of the major inherent mechanical constraint. [3]

Figure 18 – Residual stress distribution depending on the Radius of the circular weld [3]

Figure 19- Loss of stability of (left) the plate (center) inner circle (right) Outer contour [3]

It is also important to highlight the upsetting phenomenon that occurs on the plates and that plays a major role on the

welds ‘quality that are made later on the edges of the plate. Before it is heated the metal sheets are square and flat. As

the welds on the center are done, the uneven heating of the plate, the restraint offered by the weld already made and

the cooler areas cause dimensional change called upsetting. The upsetting configuration is shown on figure 20.

- 25 -

Figure 20 – Effect of upsetting in a sheet plate

2.3.5. Residual stress and distortion measuring techniques

To have the ability to accurately measure the size and shape of the structure during the fabrication process

provides information about the amount of distortion generated during the fabrication process and how the distortion

varies along the production line, this kind of insight is very helpful when one is trying to select he best DCT and adjust

them to the workpiece.

Residual stresses may be measured by non-destructive techniques; or by locally destructive techniques and by

sectioning methods. The selection of the measurement technique should take account of volumetric resolution,

material, geometry, accessibility and if the component needs to be used after the testing or not. [13]

Non-destructive

• x-ray or neutron diffraction

• ultrasonic methods

• magnetic methods

• Electronic Speckle Pattern Interferometry

• Photogrammetric measurement Vision

• Direct method with pachymeter

Detsructive

• Core Hole drilling and strain gage technique

• Hole drilling and strain gage technique

• Block removal, splitting, slicing

• Contour method

2.3.6. Distortion control Techniques

Distortion control methods increase manufactures costs due to requirements for more energy, more time

consumed in non-adding value activities, increase of labor and potentially high-cost capital equipment. Some methods

may not be suitable for automated LBW due to interruption from fixtures or stiffener arrangements. Depending on

manufacturer’s circumstances, environment and type of structures, different distortion control methods may provide

more adequate solutions to certain problems than others. Understanding their capability and limitations of all these

distortion control methods is critical to a successful welding fabrication project.

Methods for removal, mitigation and prevention of distortions can be applied before, during or after welding.

Recent progress in eliminating distortions ,has resulted in trends from the adoption of passive technological

measures to the creation of active in-process control of inherent (incompatible residual plastic) strains during welding

without having to undertake costly reworking operations after welding.

- 23 -

Control methods listed below are some of the more popular weld manufacturing methods to control residual

stresses and distortions [3],[16], [31].

- Weld technique

Each situation must be analyzed carefully; the wrong technique will lead to a more expensive and harder job. In

extreme situations it may even be impossible at all. Fusion welds often lead to the largest distortions while laser,

electron beam welding and friction stir welding result in lower distortions.

However, friction stir welding can impart large plastic strains to the structure even though the residual stresses may be

low. These large strains, which locally strain harden the material, can influence the fracture response of the structure.

- Weld parameter optimization

Weld technique parameters applied must be tuned to obtain better results. The most important parameters

are travel speed, power input, weld groove geometry and weld size.

- Weld sequencing

Weld sequence simply means the order in which the welds are deposited. Sequencing is more important for

distortion control although it can affect weld residual stresses as well. For some fabrications, weld sequencing is not

sufficient for distortion control and it is used in conjunction with some of the other.

- Fixture design

Fixtures control residual stresses and displacements by forcing the displacements and rotations of some

portions of the welded component to be zero. The ‘zero points’ should be carefully designed to achieve the distortion

control goals. Highly stiffened (restrained) structures offer better resistance to distortion during welding. This in turn

leads to higher residual stresses and requires the weld zone to yield during the heating and cooling of a weld cycle,

placing a higher demand on weld metal and HAZ properties.

- Pre-cambering

Pre-cambering consists of elastically (or plastically) bending some of the components (usually in a specially

designed fixture) in a predefined manner and then welding. After welding, the pre-camber is released and the

fabricated structure ‘springs back’ to a minimally distorted shape. The pre-camber pattern must be carefully designed.

Pre-camber also affects weld residual stresses.

- Pre-bending

Pre-bending consists of plastically bending some of the components before welding and possibly before placing

them in a fixture. The welding is performed with or without a fixture.

- Pre-tensioning

Pre-tensile loading is applied, before welding, as an active in-process control method to decrease welding

residual stress and avoid buckling distortions.

- Thermal tensioning

Thermal tensioning consists of strategically moving a heat source ahead , beside, behind (or combinations of

these), the moving weld torch. It is most convenient to design the heat source locations and power through the use of a

- 24 -

computational model. This method can control distortions and residual stresses during welding by controlling the

heating and cooling rates of the weld.

- Heat sink welding

Heat sink welding is similar to thermal tensioning except a cooling source is strategically moved (or kept

stationary) with the weld torch.

- Control of weld consumables

Welding consumables have recently been developed which result in a particular weld bead shape and can be

used to control residual stresses. The only consumable used was shielding gas. In Chapter 2.1.7 there is detailed

information about the influence of shielding gas in the welding final quality.

Finally, post-weld corrective methods are used to correct a distorted component or reduce residual stresses in a

component. Corrections made after the weld are often the most expensive and time consuming to implement.

- Post-weld heat treatment

Post-weld heat treatment consists of heating parts (or all) of the welded fabrication to a high temperatures

(depending on the material) and holding it for a period of time, while the stresses are relieved. Often the stresses

cannot be fully relieved. Again, this is an expensive method and it is often used to prevent a service fracture problem

such as corrosion, fatigue, creep or combinations.

- Hammer peening

Hammer peening is used to introduce compressive residual stresses at the weld. It counteracts the shrinkage

forces of a weld bead as it cools. Peening consists in slightly reshaping the weld bead, its stretches and makes it thinner,

thus relieving (by plastic deformation) the stresses induced by contraction as the metal cools. But this method must be

used with care. For example, a root bead should never be peened, because of the danger of either concealing a crack or

causing one. Generally, peening is not permitted on the final pass, because of the possibility of covering a crack and

interfering with inspection, and because of the undesirable work-hardening effect. Thus, the utility of the technique is

limited, even though there have been instances where between-pass peening proved to be the only solution for a

distortion or cracking problem. This method can be helpful but on the skill of the welder.

- Press straightening

The distorted component is placed within a press and mechanically straightened, is time consuming. Moreover,

the press can be an expensive shop floor component, especially for large welded components. Other machines can be

used to do the straightening such as bender. The time lost in the straightening process is an expense as well. Moreover,

the redistribution of the residual stresses after straightening is not usually considered.

There is not a technique superior to all, each case has to be analyzed carefully, parameters such as the

geometry of the component and the weld are a important when choosing DCT, furthermore the geometrical tolerance

(precision) and cost, also influences the choice of the distortion control techniques. DCTs can also be combined, to

achieve better results.

- 25 -

This is not an exhaustive list, more techniques exist like using pre-stress and preheat during assembly. With the

evolution of FEM analysis, the continuous improving of the computational weld models, predicting the distortions

behavior will be more accurate and naturally more effective techniques will emerge.

2.4. FEM analysis

2.4.1. Introduction

Residual stresses in welded structures are unavoidable. High-tensile stress exists in the weld areas. It changes

to compression in the areas away from the weld. To investigate residual stress distribution and magnitude for stiffened

plates, numerical analysis is still the best and cheaper option if time is not an important factor and it is available a huge

computational power.

For many years, researchers have studied the predictive methods for welding-induced distortions using the

finite element method (FEM). Models with various complexities were developed. Many complex models contributed to

the knowledge of distortion, but might be impractical for industrial applications due to the required computational

intensity.

Welding simulation of a large-size and/or complicated component impose tremendous computational

demands. Welding simulation normally requires transient analysis to capture the component response during the

welding, such as the welding sequence, welding direction, cooling and fixture. Welding simulation is nonlinear and often

difficult to converge because of the material behavior at the elevated temperature. Moreover, welding simulation

requires quality finite-element meshing with sufficient mesh density along the welds and the heat-affected zone. This

requires both time and expertise in finite-element meshing and usually results in a large number of degrees of freedom

in the model. The computational cost quadruples with increase in the model size. Despite the fast-growing computer

technology, the complexity of an industrial problem could easily make the simulation model infeasible to complete in an

acceptable turnaround time. In many cases, decisions have to be made to balance between the model size hat one

could afford and the level of the detail that one would want to capture.

One significant conclusion from these studies is that the weld residual stresses and distortion are not

influenced much by the weld heating cycle but, instead, it occurs as a result of shrinkage in the weld metal. It is adjacent

base metal during cooling when the yield strength and modulus of the material elasticity are restored to their higher

values at lower temperatures. Therefore, analysis of the shrinkage welds phenomena alone may be sufficiently accurate

to predict the state of weld residual stresses and distortion. This conclusion has led to the development of a modeling

scheme referred to as the ‘inherent shrinkage model by some researchers. The root cause for welding-induced residual

stresses and distortion may be described using a plasticity-based hypothesis.

Prediction of residual stresses by numerical modeling of welding and other manufacturing processes has

increased rapidly in recent years. Modeling of welding is technically and computationally demanding, and simplification

and idealization of the material behavior, process parameters and geometry is inevitable, apart from the computational

usage intensity, the biggest problem is to create a valid physic model of the component.

All simplifications, limitations, constraints, all elements used in the creation of the model, will be, presented

and justified in Chapter 5.

- 26 -

Numerical modeling is a powerful tool for the behavior of the materials, but validation with reference to

experimental results is still essential.

2.4.2. Advantages of prediction weld distortion and residual stress

Two tremendous advantages are obtained by developing fabrication solutions via the computer. First,

designing the fabrication to minimize or control distortions can significantly reduce fabrication costs. Second, controlling

the fabrication-induced residual stress state can significantly enhance the structure service life.

For distortion control, fabrication design via modeling can achieve the following [19];

• It can eliminate the need for expensive distortion corrections.

• It can reduce machining requirements.

• it can minimize capital equipment costs.

• It can improve quality.

• It can permit pre-machining concepts to be used.

Residual stress control via modeling has the following results;

• It can reduce weight.

• It can maximize fatigue performance.

• It can lead to quality enhancements.

• It can minimize costly service problems.

• It can improve damage resistance during attack (e.g. naval structures).

It is important to note that fabrication modeling tools can be used to develop new control methods since the

new methods can be first attempted on the computer. Some of these methods will be considered in the examples

discussed later.

- 27 -

3. Experimental method

This chapter describes the materials and equipment used in the experiment trials. It is also described the

procedures of each trial and his objective.

3.1. Location

The tests performed during this report were done essentially in three locations:

- Carr’s Welding Technology Ltd, in Kettering, England. This company is specialized in laser welding for moulds, tools

repair and production of precision components.

At this location were performed all the welds for testing, trials pieces.

Figure 21- Carr’s welding technology factory

- NDT services limited, in Derby, England. This certified laboratory performed the tensile tests needed to verify the

strength of the welds.

- IST, in Lisbon, Portugal. It was where it was analyzed weld bead quality and penetration.

3.2. Material and equipment

3.2.1. Testing Plates and sample preparation

In the trials all plates were made of stainless steel AISI 304 since it is the material used in the real components. The

material was ordered from the usual certified company supplier. The material’s characteristics are described in detail in

section 2.2.5.

It was ordered Plates with 3 different sizes:

- 1000x350x1.5, these plates were used in the distortion test. The size of these metal sheets is around one third

of the real size. It was important to use big sheets since the residual stress and distortions have a cumulative

effect. It was engraved small cycles (1mm diameter) in a attempt to measure local strains. It turned out to be

impossible to measure strains with this method.

- 28 -

Figure 22 – Plates used in distortion test

- 2x (150x120x1.5) were used to test the welding conditions in order to optimize parameters such as velocity,

collimation and lens. The analysis consisting in welding in a straight line two small plates. After it was done an

macrographic analysis.

- 100x30x1,5 – 100x30x4 were welded together with the best parameters determined previously after were sent

to NDT where it was performed a tensile test.

- 500x500-1.25 – 500x500x3.75 were welded together using the best parameter determined previously and

were sent to the client, so that he could perform a quality/approval check test.

-

3.2.2. Laser Machine

During the trials the laser machine used HL 1006D from Trumpf. The following table resumes the pertinent information

regarding the HL 1006D.

Laser Unit Max output

power [W] Laser power [W]

Beam quality [mm-

rad]

Power consumption

[KW]

HL 1006 D 1400 1000 25 32KW

Table 6- Properties of the laser machine

Figure 23 – Plates used in the welding procedures optimization

Figure 24 – Plates used in the Tensile test

The machine uses Neodymium doped yttrium

energy source comes from a flash light for pulsed mode and electrical arc lamps for CW, the wave length is

1,06micrometers the efficiency is around 40% in stainless steel. The overall energy efficiency of the machine is around

10%.

The parameters controlled in the machine are:

- Focal lenses that controlled the diameter of the focal point

- Length of the pulses

- Power output

- Frequency of pulse

- Overlap of pulse

- Velocity

- Collimation

The velocity plays a major role in the penetration of the weld, in the user’s manual of HL 1006D,

graphic where penetration varies with the velocity. It was with this graphic’s help that the first approxim

optimum velocity was made.

The machines are guided by an industrial arm robot KR16

maximum load capacity of 16Kg, max reach of 1610mm and repeatability of < ±0, 05 mm.

- 29 -

The machine uses Neodymium doped yttrium-aluminum garnet single crystal, Nd is the laser active component, the


around 40% in stainless steel. The overall energy efficiency of the machine is around

The parameters controlled in the machine are:

Focal lenses that controlled the diameter of the focal point

The velocity plays a major role in the penetration of the weld, in the user’s manual of HL 1006D,

graphic where penetration varies with the velocity. It was with this graphic’s help that the first approxim

Figure 25- Graphic velocity vs. Penetration [17]

The machines are guided by an industrial arm robot KR16-2 produced by Kuka. This is a 6

max reach of 1610mm and repeatability of < ±0, 05 mm.

aluminum garnet single crystal, Nd is the laser active component, the


around 40% in stainless steel. The overall energy efficiency of the machine is around

The velocity plays a major role in the penetration of the weld, in the user’s manual of HL 1006D, Trumpf present a

graphic where penetration varies with the velocity. It was with this graphic’s help that the first approximation to the

2 produced by Kuka. This is a 6 axes robot with a

- 30 -

To make the welds it was used a Trumpf laser machine HL 1006 D with a maximum output power of 1400 W

and a maximum laser power of 1000W, with a wavelength of 1.064 μm. The detailed of the laser machine

characteristics are described in section 2.1.6.

The laser is operated by an industrial robotic arm from KUKA.

3.2.3. Workstation

The workstation at CWT where the trials were performed is composed by:

- Laser machine

- Robotic arm

- Auxiliary television

- Shielding gas nozzle

- Controlling block

- Work table and accessories

Due to the laser high power and the hazards that it may cause to health, the laser is in separated room in the shop.

Figure 27- Example of welding setup configuration

Figure 26- KR16-2

- 31 -

The laser machine is extremely powerful and danger, the radiation emitted can instantly blind someone looking

directly at the light without the proper protection. The auxiliary television main function is to allow the user to see the

weld being made without looking at it. The television is also used to focus the laser, basically the laser has the same

focus depth. Therefore, when the camera is out of focus the laser is also out focus and as we alter the laser and

camera position (they move as one) the image becomes sharper. When the camera is focus the image is clear and sharp

and the laser is focus right on the surface and is in the correct position to weld, exception being made when collimation

was needed/used.

The television also has another very useful function; the stripes drawn in the screen represent the limits of the

conic beam. This can easily be verified if there is any wall interfering with the beam and consequently reducing power

delivery in the welding area.

The collimation is a system to change the laser focal position, without moving the laser’s focal point position. It

can be placed above, on the surface or bellow.

3.3. Experimental procedures

The trials were composed by four sections. Distortion analysis were it analyzed the weld sequence and the

DCT. Welding procedures optimization where it was analyzed the lens, the collimation and speed. In third it was the

tensile test, that was used to verify the results obtained previously and quantify the strength of the welds. The last was

the quality test to ensure the quality of the welds.

• Distortion analysis

Objective: Test the influence of heat distortions in weld quality and analyze the available techniques in

reducing thermal distortion.

Procedure:

1- Were performed rows with four circles along the plate’s width with the pattern and radius

determined by the client (70x80). In each plate different distortions controlling techniques were

applied.

2- The distortion induced by welding were measured, different DCT were applied, in some cases none.

Measure out-of-plane displacement and plate separation with pachymeter.

Figure 28- Workstation table Figure 29- Auxiliary television

- 32 -

The velocity used was calculated with help of a graphic Velocity vs. Penetration provided by Trumpf.

The target penetration was 3mm. The literature recommends a 0 mm collimation, so by default that was the

value used. The focal used was 150mm following the recommendations in the user’s guide of the machine.

Figure 30 – Distortion analysis test set up

� Welding parameters optimization (WPO)

Objective: Determine parameters that produced the strongest weld possible, without undesirable distortions

and minimize the residual stress

Procedure:

1- Were performed straight linel welds across the plates as shown in Figure 31 with different

combinations of collimation, Focal lens and speed.

2- After was analyzed weld quality, trough visual inspection method and macro graphic analysis:

Aspects analyzed:

- Weld defects such as, porosity or cracks

- Width and penetration of the weld

Figure 31 – welding procedures optimization set up

- 33 -

� Tensile Test

Objective: Verify quality and determinate strength of the welds of the previus tests best welds.

Procedure:

1- The best welds conditions that were discovered were used in straight line welds with 30mm long.

2- Sent test pieces to NDT’s lab to perform tensile test.

� Quality control test

Objective: component approval

Procedure:

1- Test pieces were sent to the client, to be tested.

The pieces were done using the best control distortion technique and best weld conditions.

2- It is applied a pressure of 60Bar between the joint sheets to verify if the welds can hold them together

The table below resumes the tests performed.

Test Run Material conditions objective Knowledge acquirement

Distortion analysis #1

Plate 350x350

bottom

2xPlate 350x125

Thk=1,5/1,5

No DCT applied

Reference

Worst case

scenario

- verify hammering, clamping and

bending efficiency in reducing

distortions

- which configuration (sequence and distortion controlling techniques)

produces better results


Plate 350x350

Thk=1,5/1,5

Sequence A

Vs.

Sequence B

Compare

distortions


Plate 1000x350

Thk=1,5/1,5

Clamping

Hammering

Sequence B

Compare

distortions


Plate 1000x350 Thk=1,5/1,5

Clamping Sequence A

Compare distortions


Plate 1000x350

Thk=1,5/1,5

Clamping

Pre-bending

Sequence A

Compare

Distortions


Plate 1000x350

Thk=1,5/1,5

Pre-bending

Sequence A

Compare

Distortions

WPO #1 2x Plate 150x120

Thk=1.25/3mm

Power =1Kw

Fc= 100mm

WV=0,006m/s 0,008m/s

0,01m/s

Collimation: -1, o ,

+1

Check weld quality and

penetration - which combination of speed,

power, lens and collimation,

produces the desirable results, Stronger welds and low distortion

WPO #2 2x Plate

150x120 Thk=1.25/3mm

Power=1Kw

Coalimation:

-1,0,+1

Fc= 150mm WV=0,006m/s

0,008m/s

0,01m/min

Check weld

quality and penetration

- 34 -

WPO #3 2x Plate

150x120 Thk=1.25/3mm

Power 1Kw Coalimation:

-1,o,+1

Fc= 200mm WV=0,006m/s

0,008m/s

0,01m/s

Check weld

quality and penetration

WPO #4

Plate 10x20 Thk= 5mm

C=-2, -1.5, 1, -0.5,0, +0,5, +1, +1,5,

+2

Check welding penetration

Tensile Tests Plate 100x30

Thk=1,25 & 4mm

C=0 V=0,006; 0,008;

0,01 m/s C=1 V=0,006; 0,008;

0,01m/s

Max stress allowable

- Discover welds parameters that give more strength to the weld chord

Final Simulation

Plate 1000x350

Thk=1,25/1,5

& Plate 1000x350

Thk=4mm

Best sequence

Best laser parameters

Be approves in

client’s quality test

- Verify if optimized weld can sustain all the stress

Table 7- summary of the tests performed

- 35 -

4. Trials Results Presentation and Analysis

In this chapter it will be presented and discussed the results obtained on the trials described in the previous

chapter.

4.1. Distortion and Weld quality analysis

This section is divided in two tests, welding sequence test and DCT test.

The objective was to analyze heat distortion influence on the weld quality and efficiency of DCT to control

thermal distortion.

After welding the plates, using different DCT and welding sequences, it was measured the out-of-plane

distortion, the distance from the table’s surface to the top of the top plate and plate separation. A basic visual

inspection was also performed on the welds to determine its quality and penetration. The weld parameters were always

the same: V=0,008m/s, c=0mm, Fc=150mm, Power=1Kw.

4.1.1. Welding sequence test

This test’s objective is to determine which welding sequence should be applied. No DCT was applied.

Two pairs of plates were welded with different sequences. One pair was welded in a row sequence the other was

in a column sequence. Since the welds are circular the typical solutions, like back step welding, balancing welds, or weld

near the neutral axis can not be used.

Figure 32- Sequence schematic on left row sequence (A) on the right columns sequence (B)

A B

- 36 -

Sequence test Observations

No distortion reduction techniques were applied.

Welds were performed in sequence A(row sequence)

and B (columns sequence).

(See Figure 32)

the lack of penetration can easily be seen in the

sequence A in the center of the plate.

Maximum plate separation occurred in row

sequence A. The value is 6,31mm. In collumn

sequence B the maximum value is 4,5. In both plates

the maximum plate separation occurred on the edges

due to penetration problems.

Maximum plate warping occurred in row sequence A.

Value is 14mm. This value occurs because of a local

combination of plate separation and upsetting. In

collumn sequence B the maximum out-of-plane

displacemet is 13,08.

Table 8- Results presentation of sequence test

4.1.2. Distortion Control Techniques Test

The efficiency of different distortion control techniques and its effects in weld quality was analyzed. First it was

studied the clamping technique, with different pressures. Then it was analyzed pre-bending, also with different

bending values. Hammering was also used a few times. Finally it was analyzed the combination of clamping with

pre-bending.

DCT test observation

Plate after the weld, different clamping

pressures were tested and some welding

conditions.

Best results C=0 and C+=1mm

V=0,008m/s Focal lens 150mm

- 37 -

Side view of the plate. The clear gap in

between the plates shows lack of

penetration.

Warping is still considerable. The upsetting

is still noticeable.

Maximum warping value 10.79mm.

Using clamping only.

Maximum plate separation2.82mm. Using

clamping only.

In this plate it was tested the effect that

clamping pressure had when combined with

pre-bending. The positive effect of the pre-

bending on the weld quality was clearly

noticeable.

- 38 -

Best results were +1 collimation without

clamping. This result was the first indication

that collimation +1mm could be better than

0mm or -1mm. This test also revealed that

clamping is less important than pre-bending

in weld quality. However it helps controlling

the out-of-plane displacement the pre-

bending induces.

Maximum plate separation 0.38mm. This

extraordinary difference points out the pre-

bending efficiency in preventing plate

separation. It was observed that plate

separation plays a major role in weld quality.

Maximum warping 10,27mm. As the weld

are made correctly the upsetting effect is

small.

Figure 33- Result presentation of the DCT testes

4.1.3. Results analysis

Distortion was measured with a pachymeter. This method provides accurate data but is limited. It is not possible

to measure angular distortions and in the center of the plate.

- Sequence B (columns) presented better results and has less Non-added value time per plate. However due

to the size of the water-jacket, clamping is impossible and the bending is harder to be applied. Although

welding sequence is very important to obtain better welding results, DCT will prevail over it. However due

to the size of the water-jacket, clamping is impossible and the bending is harder to be applied, with

sequence B.

- Hammering did not produced satisfactory results towards the distortion problem but it might be helpful in

relieving some residual stress as mentioned on the literature, however the author did not have means to

verify that.

- Clamping alone is not effective enough because of not being possible to clamp on all the areas at the same

time, so in reverse, clamping helps propagate the distortion to areas further way from the welding zone

- 39 -

that are not clamped. This is particularly evident on the edges, where the clamping seems to increase the

upsetting effect. If more pressure was applied it would cause more distortions on the area near the clamp.

If less pressure was applied the welds would not have been done properly.

- Pre-bending revealed to be a very effective method to prevent upsetting and plate separation. If the

bending force is increased, the plate separation and upsetting decreases, however the out-of-plane

displacement will increase. Although the bending moment applied is nowhere near the plastic bending

moment, combined with the welding process (high temperature) it has always bent the plates a little.

- Clamping and pre-bending combined produced the best results. When combined with pre-bending,

clamping seems to have little effect on the weld quality, even with no pressure the welds were done

properly, but it helps decreasing the out-of-plane displacement that pre-bending tends to induce on the

plates, keeping this value within the acceptable limits.

- An important factor to prevent plate separation and upsetting is to have a good penetration in every weld

because each weld acts like a clamp that keeps the plates together. This phenomenon is particularly

intense on the edges, as the biggest values registered were there as consequence of lack of penetration.

- Weld penetration is not constant along the weld.

4.2. Welding procedures optimization

In this section of the trials it was made the first approach to determine the best laser parameters. The difficulty

of optimization is that all variables are connected. As an example, when the heat delivery increases, so does the

penetration and heat distortion. With the Velocity is the opposite. So the key-word is compromise. The objective is to

determine the parameters that will create strong welds but that keep the heat distortions to the minimum.

After the plates were welded, the test pieces were analyzed in the microscope and the penetration was

measured.

This section is divided in two tests, the lens test and the collimation test.

4.2.1. Lens Test

The first test, lens test, primary objective was to determine which focal lens is the best. The literature review

indicated advantages in all lenses, only the specifications of this job would determine which one is the most suitable.

Since different velocities were tested it would be also the first approach to determine the optimum velocity.

- 40 -

Lens Test Observations

All the test pieces that were welded

successfully.

On left with 100mm lens; in the

middle 150mm lens; 200mm lens on

the right.

The test consisted in welding together

two plates with 1,25mm and 3mm

thick with different speeds and

collimation.

Back of the 200mm lens samples.

A few samples were welded together

successfully. The weld chord is thicker

but the penetration is smaller.

It is evident from the color, that the

welds performed are not very deep.


With this lens it was possible to weld

with a wider range of parameters. The

purple color on the back of the

samples indicates almost full

penetration.

- 41 -


This was the lens that managed to

weld better outside the optimum

zone and produced the best results.

Table 9- Result penetration of the lens test

weld properties

width penetration energy loss

Fc=100mm 1mm Full-penetration 40%

Fc=150mm 2mm Full-penetration 5%

Fc=200mm 2,5/3mm insuficient 0%

Table 10 – Focal lens analysis resume

4.2.2. Collimation test

The second test, collimation test, was performed to determine where the focal point should be positioned. The

literature review shows that it would be better on the surface or bellow.

From the previous test it was determined that the best lens is the 150mm and the optimum velocity although

not completely proved the results indicate V=0,008m/s. These were the fixed parameters and on a 5mm plate a group

of welds were performed. The only difference between the previously mentioned welds is the collimation value.

The following table resumes the results obtained.

- 42 -

Collimation Test Observations

Velocity [m/s] 0,008

Lens [m] 150

Collimation [mm] -2 -1,5 -1 0,5 0 0,5 1 1,5 2

Penetration [mm] 1,8 2,1 2,5 2,7 2.9 3.2 3.3 3.2 3

Table 11- Result presentation of the collimation test

4.2.3. Results Analysis

- Focal lens 200mm allows wider welds, therefore stronger; however the penetration rate is low, being most

times insufficient therefore, it was discarded.

- Focal Lens 100 has wide cone and the welding perimeter is very close to the clamp’s walls. Consequently the

energy loss due to wall interference is almost 40%. Considering all penetration was still good, however, the

welds are thinner than the ones produced with the 150mm lens, bringing no advantage in using 100mm lens

over the 150mm lens.

Figure 34 – Walls interference with the laser beam

- The focal lens 150mm is balanced with good penetration and the welds are wider than the ones made with 100mm

lens. The beam cone is thinner so the energy loss due to wall interference is less than 5%. This lens was the one

that performed better outside the optimum zone; this is a very desirable characteristic, since heat distortion is

controllable but not unavoidable, it is important that even with a little distortion the welds are being made in good

conditions.

4.3. Tensile test

This test allowed to the author to obtain precise values of the welds’ strength and to determine beyond any

doubt which parameters are the best. Due to the high price of these tests, the order was to the absolute minimum

quantity that would validate the data collected previously. Since the results met immediately all the expectations there

was no need to order more tensile tests because it was not expected to extract further useful data.

The following table resumes the information gathers from NDT’s tensile test.

- 43 -

Tensile test number Parameters Force applied to rupture [KN]

#1 V=0.06 C=0 25,21

#2 V=0.08 C=0 24,41

#3 V=0.1 C=0 8,43

#4 V=0.06 C=1 28,07

#5 V=0.08 C=1 26,64

#6 V=0.1 C=1 16,52

Table 12- Tensile strength of the welds

Figure 35- Graphic Strength vs. Velocity


- The graphic (Figure 25- Velocity vs. Penetration) provided by Trumpf revealed to be very accurate since the initial

velocity calculated with the graphic turned out to be the optimum velocity.

- Table 12 clearly shows that if the laser goes slower (0.06m/s) the weld strength does not increases

significantly (4%), however heat input will rise (more thermal distortions) and the production rate will

decreasse 15%. If the laser goes faster (0.01m/s) the strength of the weld decreases significantly (35%) which is

not compatible with the goal.

- Based in the literature review positive collimation should present worse results than negative or 0 collimation

but the tensile tested proved the opposite. This result was already expected since this tendency was seen in

the macrographic analysis on the weld beads

4.4. Client’s Quality Test

This was the most important test as it represents the client’s product aproval. The critical phase of the

production is the insufflation process. In this step the plates already welded are blown apart with a pressure of 2,5MPa.

Initial attempts were not successful. The analysis of the rupture revealed a problem in welding process, because of the

- 44 -

heat distortions some welds were not having enough penetration and this way, they did not have the standard strength.

Consequently the welds were not capable of withstanding the pressure.

The plates sent to the client were welded with the following parameters: Fc=150mm V=0,08m/s C=+1mm

Power=1Kw DCT: Pre-Bending and Clamping

Figure 36- Plates after quality test Front View


- There were no record of problems during insufflation.

- The plates were approved by the client.

The goal was reached, the plates must be welded with Fc=150mm V=0,08m/s C=+1mm Power=1Kw. These

conditions create a weld that has a 440MPa of tensile strength. To control the distortions, pre-bending and

clamping should be used.

The pre-bending prevents plate separation, which is the main responsible for bad weld quality and upsetting.

Also if the upsetting effect is minor, the welds made after on the edges are easily made. The clamping

decreases the out-of-plane displacement caused by the pre-bending, resulting in no after work needed to

flatten the plates back to normal.

- 45 -

5. Finite Elements Analysis

In this chapter infomation will be given regarding tha finite element program used, advantages, limitations and

the algorithm used in each situation analyzed. Finnaly the outcome is shown and connected with the results obtained

in the trials. Two situations were analyzed through the finite element method: the pattern of the welds and welding

induced distortion.

5.1 Program used

The ANSYS CAE (Computer-Aided Engineering) software program was used in conjunction with 3-D CAD

(Computer-Aided Design) solid geometry to simulate the mechanical behavior of a two coupled plates under thermal

and structural loading conditions. ANSYS automated FEA (Finite Element Analysis) technologies from ANSYS, Inc. to

generated the results listed in this report.

Figure 37- Ansys Logo

To be able to accurately predict how the plates will behave in real-world environments where multiple types of

coupled physics interact. Multiphysics simulation software from ANSYS allows users to create virtual prototypes of their

designs operating under real-world multiphysics conditions.

The analysis is separated in two parts: the Thermal analysis where the temperature and heat distribution are studied

and the Structural analysis where stress and strains are analyzed.

Each scenario presented represents one complete engineering simulation. The definition of a simulation

includes known factors about a design such as material properties, contact behavior between the two plates, types and

magnitudes of loading conditions (thermal and structural). The results of a simulation provides insight into how the

bodies may perform and how the design might be improved. Multiple scenarios allow comparison of results given

different loading conditions, materials or geometric configurations.

5.1. Pattern analysis

5.1.1. Problem resume

The welds’ pattern is defined by the gap between the center of circle welds, and the distance from the plate’s

edge to the center of the circle welds. It determines the force applied per length of weld and the stress in the plates, it

also influences the heat distortion and the productvity. By making a wider net of welds, the force that each weld will

hold will increase, what may compromise the component integrity. If the distance is smaller more welds are done, the

productivity decreases and the heat input increases, consequently weld induced distortion becomes harder to control.

The optimum pattern is a result from compromise of productivity and strutural resistance.

- 46 -

Figure 38 – Welding pattern

5.1.2. Objective

The client recommends CWT to use the folowing pattern: A=70mm B=80mm C=70mm D=80.

The objective of this analysis is to discover why this pattern has been choosen and to verify if it is possible to use

a wider pattern in order to improve productivity, within the standard level of safety.

5.1.3. Model’s structure

To analyze this situation a static analysis was performed. The real event that is being simulated is the insufflation

process that is done after the plates are welded together. Since the insufflation process is not performed at CWT the

only way to test the pattern influence of the stress transmission is trough a numerical analysis.

First the geometry and element type were defined, then material properties, boundary conditions and finally the

loads were applied. When the solution is calculated it is possible to view the plate’s behaviour and the stress in the

plate.

Figure 39 – Pattern analysis algorithm

The stress in the welds is calculated in a excell calculus sheet. The details are described in section 5.1.7.

5.1.4. Element type

The element used is SHELL63. This is a four node element that has both bending and membrane capabilities. Both in-

plane and normal loads are permitted. The element has six degrees of freedom at each node: translations in the nodal x,

y, and z directions and rotations about the nodal x, y, and z-axes. Stress stiffening and large deflection capabilities are

included.

- 47 -

The element is defined by four nodes, four thicknesses, an elastic foundation stiffness, and the orthotropic material

properties. Pressures may be input as surface loads on the element faces, positive pressures act into the element. [31]

Figure 40 – Element input (left) and output (right) [31]

5.1.5. Geometry and mesh

The plate has 1.25mm thickness.The length and width is controlled by the welding pattern. Instead of changing the

number of welds each time a new pattern was tested, it was the size of the plate that is changed. The number of welds

remains the same through the tests. Each model has 27 welds. A full numerical model was not made in order to speed

up the calculations. This model is valid because if enlarged, only the stable part in the center of the weds would be

added, so no aditional information would be extracted.

Figure 41- Plate geometry (left) and mesh (right)

5.1.6. Boundary conditions

• Fixtures: The 1.25mm plate is going to be welded to one with with 4mm, so it is assumed that it will be

the 1.25 thick plate that will deform and will fail, therefore it was not considered to be necessary to

simulate the second plate.

The circle welds and the side welds do not move, as they are attached to the 4mm plate that is considered to be a rigid

body. The displacement constragiments in the welds are signalled with green arrows in figure 42.

• Pressure applied: On the bottom of the plate it is applied a pressure of 2.5Mpa (25 Bar), signaled with

red arrows in figure 42.

- 48 -

Figure 42 - Boundary conditions

5.1.7. Welds’ stress

Ansys does not calculate the stress in the welds, only in the plate. So the force applied in the welds were

calculated separately in an Excel calculus sheet.

The chosen weld was made using the following parameters: C=1, focal lens=150mm, velocity=0.08m/s Power

input=1KW; this weld (#5), is capable of withstanding a forces up to 888N/mm of weld.

AreaGHIJK = 1250x30000 [mm2] (5)

P��Q'&R = S8�� 7� �2 ×B<;9

< [mm

2] (6)

P��TU',,VU' = AreaGHIJK − AreaWXKY [mm2] (7)

@�� = 6��228�� × P�� @�� = 2,5 × 10[ × P�� @�� [N] (8)

\�/0�ℎQ'&R = S8�� 7� �2 ×B<9

5 [mm] (9)

��]'&R =^$*%& +$U�'

_'()*� $+ ]'&R,≤ 888S/�� (10)

5.1.8. Results

• 70x80 70x80

The client sugested a circle pattern of 70mm between centers of the weld in a line and 80mm in a row. Nothing

was said about the distance from the center of the weld to the side.

Pattern variables: A=70mm B=80mm C=70mm D=80mm

- Inicially the author had doubts about the high pressure value applied on the plate so after a few simulation it

concluded that it was necessary a pressure of 1,85Mpa to create plastic deformation on the plate.

- 49 -

Figure 43 – Plate on edge of entering the plastic domain

- In the insuflation process it is applied a pressure of 2,5MPa to deform the plates, into the plastic domain, in

order to separate the plate as shown in figures 38-41. This translate in the following data.

Area Pressured [mm2] 3,67x106

Total Force [N] 9,17x106

Total Length of weld [mm] 31,6 x103

Force Weld [N/mm] 296

Table 13 – Weld’s strength calculations data

Using Ansys, the process was simulated, with the process described above. It produced the following results:

Figure 44 – Displacementon on the plate- Max displacement 3,2mm

Figure 45 – Top view f plate stress

- 50 -

Figure 46 – Bottom view of plate – Maximum stress 254MPa

- With a pressure applied of 2,5Mpa the maximum displacement occurs on the spaces between the welds. The

maximum displacement value is 3,216mm. The maximum stress occurs in the weld area and its value is

254Mpa. It is clear from figure 39 and 40, the diferent stress distribution between the top and bottom of the

plate.

The safety factor used by client is 2,03.

• Pattern optimization

From the results shown previously the author verified that the client used a relavitively big safety factor for the

insuflation, considering that the working pressure is not superior to 0,5Mpa.

With this information and having in mind that the bigger the number of welds the longer it will take to produce

each workpiece. It would be interesting to analyze the evolution of stress and displacement with different

patterns.

AxB CxD Number of

welds

Force per mm of weld [N/mm]

Maximum displacement

[mm]

Maximum plate’s stress

[MPa]

Safety Factor

75x75 70x70 532 287 3,1 242 2,13

70x80 70x80 525 290 3,2 254 2,03

75x80 70x70 490 305 3,4 262 1,97

80x80 70x70 455 322 3,6 265 1,95

80x85 70x70 429 336 3,8 271 1,90

90x90 80x80 372 371 4,4 305 1,69

100x100 100x100 280 445 5,2 327 1,58

Table 14 – Parametres influenced by th welding pattern

- From the results above, it is clear that there is no reason to why it could not used a wider pattern; if there is no

other constraint to determine the pattern than the capability to resist the pressure applied in the insuflation

- 51 -

process. There are wider pattern with good safety factors, that have significantly lower the production time, as

shown in the graphic bellow.

Figure 47 – Graph Production time and Safety Factor Vs Welding Pattern

5.2. Welding distortion analysis

5.2.1. Problem Resume

CWT was interested that was developed a finite element model; in order validate the results obtained in trials

and test future procedures.

5.2.2. Objective

Test the efficiency of the distortion control techniques used in order to validete the results obtained in the

trials.

5.2.3. Model’s structure

To analyze such a complex situation it is necessary a sequentially coupled physics analysis. A combination of analyses

from different engineering disciplines will interact in a single model to solve a global engineering simulation.

This analysis will simulate a core hole drilling and strain gage technique. Before the welding process is performed, an

electric strain gauge is glued in the surface, in the zone to be analyzed. After the weld it will register some deformation.

Next, a piece is cut around the gauge, the material will contract, elastically, so what the deformation that gauge is still

registering is residual pos-weld strain. With deformation values and based on elastic-plastic models stress is calculated.

Ansys has some limitations, not all physics phenomenon can be replicated successfully, some approximations have to be

introduced and others over-looked. Limitations introduce errors in the results that are hard to quantify. Since in the

experimental process in CWT was not possible to measure the residual stress or distortion with accuracy, the important

information to extract is the comparison between different methods.

61 60 5753 50

4335

2,13 2,03 1,97 1,95 1,9 1,69 1,58

0

10

20

30

40

50

60

70

75x75

70x70

70x80

70x80

75x80

70x70

80x80

70x70

80x85

70x70

90x90

80x80

100x100

100x100

Productivity Vs. Pattern

Production

time [min]

Safety

Factor

- 52 -

The process requires the creation of all the necessary environments, which are basically the preprocessing portions

for each environment, and write them to memory. Then in the solution phase they can be combined to solve the

coupled analysis.

Each physic enviroment (thermal and structural) was contructed separatly, but only one geometry exists, only one set

of nodes and elements type is used for the entire analysis.

First the geometry, and element type were defined, then material properties, boundary conditions and load steps of

each enviroment. The Thermal Environment is where the laser heat is applied. On the Structural Environment, different

types of constrains were applied and then displacement and stress were calculated based on the previous thermal

analysis.

Figure 51 shows what input data is received by Ansys and what information can be extracted.

Figure 48- welding distortion analysis algorithm

5.2.4. Element Type

The direct coupled-field elements allow users to solve a coupled-physics problem by employing a single finite

element model with the appropriate coupled-physics options set within the element itself. A direct coupled-field

solution simplifies the modeling of multiphysics problems by allowing users to create, solve and post-process a single

- 53 -

analysis model. From the coupled-field elements avaiable in Ansys, few had the capabilitie to be used in a Thermal-

Structural analysis, excluded the 2D elements, SOLID5 that is a tetrahedral elements and reavelled difficulties in

adjusting to the mesh since a circular mesh was used in this model, also tetrahedral elements tend to suffer from

overstiffness (locking) which does not happen in the pyramidal elements, such as SOLID98. For these reasons SOLID98

was the element choosen to do the numerical model.

Bellow is resumed the information of element SOLID98.

Figure 49- Element input

Figure 50- Element output [30]

SOLID98 is a 10-node tetrahedral element, with up to six degrees of freedom at each node, it has a quadratic

displacement behavior and is well suited to model irregular meshes. When used in structural,has large deflection and

stress stiffening capabilities. Since it is a quadratic element is one of the reasons for the enormous weight of the model.

Various combinations of nodal loading are available for this element such as displacement, temperature,

voltage and scalar magnetic potencial.

Element loads are pressure, convection or heat flux (but not both). Positive pressures act into the element.

These forces are applied in solution as structural loads

The body loads; temperature, and heat generation rate may be input based on their value at the element's

nodes or as a single element value. When the temperature degree of freedom is active, applied body force

temperatures, are ignored. In general, unspecified nodal values of temperatures and heat generation rate default to the

uniform value specified. the temperature degree of freedom is present, the calculated temperatures override any input

nodal temperatures.[31]

- 54 -

5.2.5. Geometry

The geometry is composed by one plate with 4mm and one with 1.25. In order to obtain better a better

meshing net, concentric rings are made in the center of eahc plate plates. To simulate the key-hole heat zone a thin

ring is created with 7mm radius 0.2mm thickness and 3mm deepth. This hollow cylinder is represented has may collour

because is divided in smaller volumes to simulate the heat source.

Figure 51 – Composition of the model’s geometry

5.2.6. Thermal boundary conditions

- Themal constraints

It was considered that the limit walls had an imposed temperature of 25ºC.

The initial and reference temperature was 25ºC

The following picture shows were the the thermal constrains

Figure 52 – Thermal constraints

- 55 -

- Heat source

Ansys does not have the ability to simualte moving loads. To overcome this obstacale a routine in Matlab was

developed. It divides the weld bead in smaller elements, from which it is generated heat, one at a time, in order to

simulate the moving laser heat input (see figure 51). The more divisions, the more closer to reality the model would be.

The limit of divisions was less than 150. At this stage the size of each weld bead element is so smal that is within the

tolerance range of Ansys. In the analisys was used models with 50 and 100 divisions.

5.2.7. Structural boundary conditions

- Structural constraints

Ansys needs to have a reference point, otherwise the displacement results may have no meaning because if

the loads applied are unbalanced the component moves. The solution is to fix one node during the whole analysis, this

node will work as a reference.

The outer walls are in the solid state during the hole analysis, meaning that they have no displacement.

As decribed in section 5.2.3.1 bending can not be applied, however it can be applied the straigth plates theory,

so the bending moment is substituted by traction and compressive pressures.

Pre-stress DCT was also tested, in two modes. The first one was the uniform pre-stress, it was a applied presure

in all outer walls. The second one was longitudinal pre-stress, where the pressure was only applied in two opposite

walls.

5.3. Result presentation

5.3.1. Thermal Results

The figure bellow shows the thermal distribution on a segment of the while during the welding process.

Figure 53- Temperature distribution in the plate during the weldng process

- 56 -

The maximum temperature, as expected, is reached in the in the welding bead. The four temperatures signaled

in the welding bead demonstrate that the cooling is quite rapid since the difference between them is less then one

second. As expected the temperature near the outer walls remains almost the same during the welding process

therefore, close to the thermal boundaries pre-determined. This outcome does make sense because the austenitic

stainless steel is characterized by a low thermal conductivity.

Figure 54 shows how the temperature in the center of segment evolves during welding and the cool down

phase.

Figure 54 – Temperature in the center of the weld

As previously mentioned, the thermal conductivity of ASI 304 is low, in comparison with average of steels.

Although the welding process stops in second 6, the temperature continues to rise, as shown in figure 54, but at a lower

rate. In spite of the temperature not being stabilized, it is not expected to reach values that would justify a longer

simulation, that would show the maximum temperature reached in the center.

The most relevant information to extract from the graphic above is that the initial prevision that the lateral walls

remained at the surrounding temperature during the whole process, is correct.

Figure 55 – Temperature in a division of the weld bead

- 57 -

On figure 55, is shown the evolution of the temperature on a segment in the weld bead. This area will be heated

directly by the laser; this is why such high temperature is reached (maximum temperature 1650°C).

The most important information to extract from this graphic is that it confirms, that the parameters used in welding

process are well defined since the maximum temperature, is only 230°C above the melting temperature. If the spike

temperature were higher, it would aggravate the distortion problems (excessive heat input), if the spike temperature

were lower the risk of appearing non-melted areas would increase.

This hypothesis is sustained by Figure 38. The graphic clearly shows that the benefits of increasing heat input

(decrease welding speed) is not reflected in the increase of weld strength. On the other hand, a small decrease of heat

input (increase of speed) leads to a significant decrease of the strength of the weld.

At the moment of the structural analysis the plate’s temperature at any point is bellow 600°C.

5.3.2. Structural results

The thermal analysis, calculates the temperature in every points of the plate’s segment for the duration of the

analysis, 5 seconds of welding and 4 seconds of cooling down. The structural analysis runs on top of the thermal

analysis. It is in this phase that the structural constraints are introduced and the thermal stress and heat distortions are

calculated.

5.3.2.1 Without Distortion control technique

The thermal analysis revealed that, as expected, the outer walls remain at a low temperature (>30°C), so the

mechanical properties remain the same during the analysis. In terms of computational structural constraints it means

that the outer walls do not move since they are in the solid state during the whole simulation.

As the welds are performed, the material in center (around the weld bead) expands and presses the material

surrounding it deforming it permanently. As expected the displacements in this phase are very smal (>0,005mm).

Figure 56 –Displacement of the model with displacement constraints

- 58 -

The literature review indicates that the stress induced by thermal expansion can be bigger than the yield

strength (205MPa). This is coherent with the fact that these stresses induce plastic deformation. With the structural

analysis it was possible to measure that phenomenon, since in the trials the author had no means to measure the weld

induced stress. Discovering the residual stress value is very important since it is a reference that will help to optimize

the parameters of the DCT.

The maximum values are registered near the outer walls., as it is where the displacement constraints are

applied. In the analysis the maximum stress registered is 243MPa.

Figure 57- Residual stress in the model

When the displacements constraints are withdrawn the plate shrinks elastically. Therefore the residual stress is

the value registered, with constraints, less the yield stress and the maximum residual stress is 38MPa.

The value of the residual stress is of the most importance, because if it was nullified theoretically the weld

induced distortion would be null. This value will be a reference point that will guide the DCT parameters. For this reason

in DCT simulations it was always made one simulation in the elastic regime, where the stress introduced was 200MPa

and another in the plastic regime, where the stress introduced was 243MPa.

- 59 -

Figure 58 – Displacement NO DCT Model and Model unreformed

Decreasing size of the model is easily noticeable in figure 56. The maximum displacement sum is 0,157mm and

it occurs on the corner opposite to the fixed point (origin - inferior left corner).

The out-of-plane displacement maximum value is also located in the opposite corner to the reference point.

The maximum Z-axis movement is 0,0252mm.

The angular distortion that can be seen in figure 56 and 57 occurs as a consequence of the shape of the weld.

The bending moment vector changes direction as the circle weld is made, resulting that the radial stress while the

welding is made is not balanced.

Figure 59 – side view of the model Out-of-plane displacement

- 60 -

5.3.2.2 Pre-Bending

In the model it is not possible to apply bending moment. However the plates fulfill the pre-requisites to be

analyzed in elastic bending regime. So computationally, the bending is represented by two faces of the model that have

a pressure applied in such way that replaces the bending and the other two are free. The stress applied in the walls

induces a maximum stress of 200MPa in the plates. These stresses are introduced with the objective of counteract the

weld induced tress.

Figure 60 – Pre-bending structural constraints

Figure 61 – Displacement sum in Pre-bending simulation

The maximum displacement occurs as expected in the opposite corner to the reference point. Its value is

0,103mm.

- 61 -

Figure 62 – Out-of-plane displacement in pre-bending simulation

The maximum out-of-plane displacement occurs in the same position as the sum displacement its value is

0,0181mm.

5.3.2.3 Pre-cambering

In one of the trials, by accident it was applied a plastic bending moment (Pre-cambering) instead of an elastic

bending moment (pre-bending). The results regarding plate separation and upsetting were very good. Unfortunately the

plate was too much bent. It needed to be worked on, re-bended back to its flat state therefore, pre-cambering was

discarded.

The maximum sum displacements are in the usual location, the corner further away from the reference point.

The maximum value registered was 0,0742mm.

Figure 63 - Displacement sum in Pre-cambering simulation

- 62 -

This technique present very good results with the out-of-plane distortion. The maximum value is 0,0155mm.

Figure 64 - Out-of-plane displacement in pre-cambering simulation

Figure 65 - Out-of-plane displacement in pre-cambering simulation (detail)

5.3.2.4 Uniform pre-stress at 200MPa

Pre-stress was not available during the trials however the literature review indicates pre-stress as one of the

best distortion control techniques. As stated before one of the main advantages of FEA is that it can be used to develop

and test different control methods via FEM instead of laboratory experiments, which is more costly.

Pre-stress is an interesting method not only for the efficiency in controlling distortions but also because the

plates can be welded in one pass. Different weld sequences would have to be analyzed; whichever the case production

time will always be lower than the ones presented with Pre-bending or clamping techniques.

Based on the arguments presented above, uniform pre-stress was simulated to verify if this technique could

produce better results than the tested methods. A stress of 200MPa was applied on the four outer walls.

The maximum displacement is 0,12mm and as usual located in the corners further away from the fixed point

- 63 -

Figure 66 – Displacement vector Sum in plate with pre-stress at 200MPa

The maximum out-of- plane displacement value is 0,0209mm.

Figure 67- Out-of-plane displacement in uniform pre-stress simulation at 200MPa

5.3.2.5 Uniform Pre-stress at 243Mpa

Considering the residual stress value calculated previously and its influence in the distortions, it was used in

this test a stress value that would eliminate the residual stress. Traction stress was applied a in the four outer walls with

a value of 243Mpa.

- 64 -

The same phenomenon that occurred with pre-bending and pre-cambering happened with the pre-stress. The

results obtained at 243Mpa were better than the ones obtained at 200MPa.

The maximum displacement sum is 0,0511mm and it is located in the corner further away from the reference

point.

Figure 68 - Displacement vector Sum in plate with pre-stress at 243MPa

The maximum out-of plane displacement registered was 0,0145mm and it is located in the same place.

Figure 69 - Out-of-plane displacement in uniform pre-stress simulation at 243MPa

- 65 -

5.3.2.6 Longitudinal Pre-stress at 200MPa

Uniform pre-stress is harder to apply in such big plates, however if the stress is applied in only one direction it

might be easier. In this case 200MPa traction stress was applied in two opposite walls.

Figure 70- Displacement sum before releasing DCT (left) after releasing DCT (right)

The maximum sum displacement value is 0,237mm and the maximum out-of-plate distortion is 0,0303mm. In

both situations the maximum is located, as usual, in the opposite corner to the reference point.

Figure 71 – Out-of-plane displacement in longitudinal pre-stress

- 66 -

5.3.2.7 Longitudinal Pre-stress at 243

This technique did not present good results in the elastic regime (200MPa) but it would be interesting to verify

if it would get better results if the stress was increased, into the plastic domain (240MPa) since the same phenomenon

occurred with pre-bending and pre-cambering. The maximum displacement occurs in position as in previous tets and

the value is 0,185mm.

Figure 72- Displacement sum on longitudinal stress II

The maximum out-of-plan displacement is located in the opposite corner to the fixed point and its value is 0,0246mm.

Figure 73- Out-of-plane displacement in longitudinal pre-stress II

- 67 -

5.4. FEM Result analysis

The initial consideration that the lateral walls did not move was correct as the thermal analysis revealed that

the temperature never surpass the 30°C. Therefore their mechanical properties remain the same during the whole

welding simulation. However the FEA shows that the material around the weld bead does move, due to thermal

expansion caused by the welding heat input. The dilatation induced stress up to 243MPa and the residual stress is

38MPa.

Following the thermal analysis, the maximum temperature on the weld spot was 1650°C. The number

confirms, that if the used speed of v=0,008m/s is not the optimum it is very close; because If the speed increases the

heat input and the maximum temperature will decrease, resulting in a rapidly decrease of weld quality. In the welding

process the conditions changes along the time, so in a extreme case the melting temperature might not be reached.

This effect is clearly shown in Figure 34. On the other hand if speed decreases, as the weld already has the desirable

penetration, the strength will not increase proportionally, but the heat distortions will. Following this the productivity

will also be lower.

The following table resumes the structural results obtain from ANSYS.

No DCT Pre-

bending Pre-

cambering

Uniform Pre-stress 200MPa

Uniform Pre- stress

243MPa

Pre-stress longitudinal

200MPa

Pre-stress longitudinal

243MPa

Disp. Sum 0,157 0,103 0,0742 0,12 0,0511 0,237 0,185

Z Disp 0,0252 0,0181 0,0154 0,0209 0,0145 0,0303 0,0264

Table 15 – Simulation result’s resume

Figure 74 – DCT’s distortion value comparison

The structural results show that the DCT in the plastic domain presents better results than those in the elastic

domain. This indicates that tuning the DCT based on the residual stress value is a good strategy to avoid welding

induced distortions as long as the residual stress value is accurate, otherwise extra unnecessary stress might be

introduced.

0

0,05

0,1

0,15

0,2

0,25

Disp. Sum

Z Disp

- 68 -

The Ansys simulations results are coherent with the trials results. Pre-bending technique showed effectively

better results, however this model does not account with the out-plane displacement that is induced when many welds

are done in a row.

The technique that produced better results in the simulations was pre-stress. However at the time of the trials

it was not available, but with this simulation results it might be interesting to try to do a laboratory test to verify if pre-

stress can indeed produce better results than were tested. This technique is interesting since no clamping is needed and

if used it would allow the plate to be welded all at once. This would decrease significantly the production time.

The results show that pre-bending is a good option from the ones available. It is not the best DCT but it does not need

any re-work of the plates after the weld and the results are satisfactory.

- 69 -

6. Conclusions

This chapter presents the conclusions of this work.

• The best focal lens is 150mm. This choice was expected based on the literature review [11] as it was the more

balanced lens. Table 9 and 10 resumes the results that confirm that tendency that the welds performed with

this lens has good penetration and generates a wide weld bead. As the laser cone is not too wide the energy

loss due to interference with the clamp is residual.

• The best collimation is +1mm. On the contrary, the tendency of the literature [11] , [21]. that suggest that the

best collimation is either 0 or negative, the results on Table 11 clearly demonstrates the opposite. Positive

collimation produces better results. Since there is no mechanic or physic explanation for this fact, the logical

conclusion that the laser machine used on the trials was not focused properly, being the offset positive by

1mm-2mm, although this hypothesis has not been confirmed by CWT.

• The optimum speed is 0,008m/s. This was the initial estimated speed calculated with Figure 25, a graphic

provided by Trumpf that revealed to be very accurate. This statement is supported by the results shown on

Table 12 and Figure 35 and Figure 55. On Figure 55 it shows that the spike temperature on the weld bead is

only 230°C above the melting temperature which demonstrates that the energy delivery is only the necessary

to perform a weld plus a 15% safety factor which is needed to assure the weld quality since the welding

conditions changes during the welding process, mainly due to loss of focus caused by distortion.

• The conditions with the following parameters C=+1mm V=0,008m/s Fc=150mm Power=1KW; generate a weld

bead with a tensile strength of 440MPa. These calculations were based on the strength determined on the

tensile test perform by NDT as seen on Table 12.

• Column sequence B was produced less distortion than in row sequence A and had less non-added-value time

per plate. However due to the size of the plates (1250mm x 3000mm), the DCT’s avaiable could not be applied

with sequence B. Since the DCT has a much bigger weight in reducing distortion than the welding sequence,

row sequence A was the chosen one.

• Hammering is ineffective in removing heat distortion. It might be useful in relieving some of the residual stress

but that fact could not be test or proved.

• Clamping is one of the most DCT used in the welding industr,y however in this study case due to the size of the

plates this method was not effective by itself. All the distortions were prevented in the clamped area but it

seems to help propagate the distortions to areas further away that are not clamped, regardless of the pressure

applied.

- 70 -

• Pre-bending is effective in reducing weld induced distortions. This happens because, from the point of view of

distortion resistance, the moment of inertia of the cross-section of the entire panel structure increases as the

structure is bent. Since the weld shrinkage force remains unchanged, the magnitude of global bending of the

panel structure is reduced by the increasing cross-sectional rigidity. This method was particularly good in

avoiding plate separation since the process of bending the plates pushes them against each other, and once

the welds are done correctly they will act as clamp keeping them together.

• By far the best results were obtained using Pre-bending and clamping together and it is this combination and

DCT the recommended to be used in this work. The results showed that the clamp pressure does not need to

be high since its main function is to the out-of-plane distortion that the bending induces.

• Pre-cambering was used by accident in one test and it produced good results, however the value of the out-of-

plane distortion induced in the plates was high (the plate was bent). The plate would have to be re-worked to

be brought back to the straight position, which is another time consuming activity that does not add value to

the product therefore, it was discarded.

• The results obtained in Ansys simulations are coherent with the trials’ results. Pre-cambered continued to

exhibit good results in controlling weld induced distortion. However in the finite element analysis, pre-stress

stood out, this technique should be now studied in the laboratory in order to verify further good results.

• The FEA also showed that DCT’s applied in the plastic regime presented better results than the ones in elastic

regime (see Table 14). This result was expected as they were tuned to counteract the residual stress that the

welding process induces.

• One of the standard safety factor used throughout the industry is two. During the insufflation process the

stress value with the client’s welding pattern (70x80 70x80) is 254MPa. The material has an tensile strength of

515MPa showing that the safety factor in this production step is two. Although this might be a coincidence, it

also may be the reason why this pattern was chosen.

• Further analysis on the weld pattern revealed that if a lower safety factor was accepted, there would be other

patterns that would allow faster production rates and safety would not be compromised.

• The initial difficulties that this work presented, with low weld quality in the preliminary attempts, indicated

that this was not the ideal technology to be used in this job. However this study shows that using the

conditions and the DCT suggested this work can be done with the desired quality. Further to this study, with

laser beam welding inherent capacities, water jackets can the produced at a competitive cost. Finally useful

data is provided that allow smaller production time.

- 71 -

7. Future work

Althought Ansys simulations revealled to be trustworthy, there are many flaws that in the future could be solved

in order to obtain a more reliable and accurate results.

The most important is that this model only has a plate’s segment with only one weld, it would be very

interesting to develop a model that would have more, specially because residual stress and weld induced distortios are

cumulative. As this model is only composed by only one weld the upsetting effect cannot be studied, plate separation

could not be simulated either. Finnaly for some unknown reason th clamping echnique could not be simulated, as the

needed surfaces could not be selected.

Although the author had conscience of the limitations that would be introduced by this decision, the increase of

computational weight and programming difficulty was so high that would jeopardy the time table of this work. A lighter,

smaller mode had to be considered. However the processing time of this lightweight model still is high (8h for the

thermal analysis and 4h for the structural analysis).

This model can still be used to test and optimize parameters such as speed and parameters of mechanical DCT

as previously shown but also thermal DCT like heat sink or pre-heat. This FEM can also be rapidly adapted to test other

situations like diferent plates thickness, materials, power, weld radius and penetration.

Many of the so called flaws, come from the finite element program limitations that was used, so maybe some

could be solved simply by using a different program.

Ansys gave good references of the pre-tension distortion control technique,so it might be interesting to CWT

further study this method and if the results justified, as they are expected, to own a equipment that can apply these

stress. Such equipment would also help to decrease significantly the production time.

Finnaly in section 9.1 is presented an apparatus developed by the author, that applies the advised DCT’s and that

facilitates the welding process therefore, decreasing the production time.

- 72 -

- 73 -

8. References

[1] K. Masubuchi – “Analysis of welded structures” – Ed. Pergamon Press, Oxford, UK, 1980

[2] Steen W. M. – “Laser Material Processing” – Springer – Verlag, 3ª Ed, London, England , 2003, Cap. 1, 2, 4, 5.

[3] Feng, Zihli – “Processes and mechanism of welding residual stress and distortion” – Woodhead Publishing

limited, 1st Ed, Cambridge, England, 2005, Chap. 1-3, 5-9.

[4] C. L. TSAI, S. C. PARK AND W. T. CHENG – “Welding Distortion of a Thin-Plate Panel Structure” – People’s

Republic of China, May, 1999

[5] Tiago Carlos Pereira e Rosa – “Modelação Térmica e de Tensões Residuais de Soldadura de Metais Duros” –

Dissertação de mestrado em Engenharia Mecânica, Faculdade de Ciências e Tecnlogia da Universidade Nova de

Lisboa, 2008

[6] Hugo Marques – “Modelação das propriedades mecânicas de juntas soldadas por fricção linear” –Dissertação

de mestrado em Engenharia Mecânica, Faculdade de Ciências e Tecnologia da Universidade Nova de Lisboa

[7] Maria Alexandre S. R. P.da Costa – “Soldadura Laser de Pastilhas de Carboneto de Tungsténio a Aço” -

Dissertação de mestrado em Engenharia Mecânica, Instituto Superior técnico da Universidade Técnica de

Lisboa.

[8] T. Stolarski, Y. Nakasone, S. Yoshimoto – “Engineering Analysis with Ansys Software” Elsivier Butterworth-

Heinemann, 1st

Ed, Oxford, England 2006, Chap. 6.

[9] O. A. Vanli, P. Michaleris – “Distortion analysis of weld stiffners” Pennysilvania State University, Universit Park,

June, 2001

[10] DT.D Huang, P.E,P. Keene and L. Kvidahl – “Distortion Mitigation Technique for Lightweight Structure

Fabrication”, University of New Orleans

[11] Michaleris, P. and DeBiccari, A., 1997, “Prediction of Welding Distortion” Welding Journal, Welding Research

Supplement, 76 (4), p 172s-181s

[12] Trumpf User’s manual guide

[13] P. J. Withers and H. K. D. H. Bhadeshia - “Residual stress Part 1 – Measurement techniques” University of

Manchester, University of Cambridge, England, March 2000

[14] Dong, P. – “Residual stresses and distortions in welded structures: a perspective for engineering applications”,

Science and Technology of Welding & Joining, Volume 10, Number 4, July 2005 , pp. 389-398

[15] Jingyong Li, Yaming Lu, Zhipeng Liu and Mingming Zhang – “residual stress and distortions in Aluminm Alloy

Sheet After welding under Pre-tension” Advanced welding TechnologuProvincial Key laboratory, Jiangsu

University of Science, Republic of China, 2009

[16] F. W. Brust, Paul Scott – “Weld Distortion Control Methods and Applications of Weld Modeling” - SMiRT 19,

Toronto, August 2007

[17] Wentao Cheng, B.S., M.S. – “ In-plane shrinkage strains and their effects on welding distortions in thin-walled

structures” Ohio State University

[18] You Chul Kim, Kyong Ho Chang, Kohsuke Horijawa – “Characteristics of out-of-plane deformation and residual

stress generatedby fillet welding”, JWRI, Vol 27, 1998, No. 1.

- 74 -

[19] Peng Cheng, Andrew J. Birnbaum, Y. Lawrence Yao – “Correction of Butt-Welding induced Distortion by Laser

Forming”, Department of Mechanical Engineering, Columbia University, New York, NY, 10027, USA

[20] J. Cornu. “Advanced Welding Systems - V. III. TIG and Related Processes,” Springer Verlag, 1988, Chap. 5

[21] F. W. Brust, Paul Scott - “Weld Distortion Control Methods and Applications of Weld Modeling”, Transactions,

SMiRT 19, Toronto, August 2007

[22] Dhingra, A. K.; Murphy, C. L. - “Numerical simulation of welding-induced distortion in thin-walled structures”,

Science and Technology of Welding & Joining, Volume 10, Number 5, September 2005 , pp. 528-536

[23] Erdogan Madenci, Ibrahim Guven – “The Finite Element Method and Applications in Engineering using ANSYS®”

– Springer – Verlag, 3ª Ed, London, England , 2006

[24] T. Stolarski, Y. Nakasone, S. Yoshimoto – “Engineering Analysis with Ansys Software” Elsivier Butterworth-

Heinemann, 1st

Ed, Oxford, England 2006, Chap. 6.

[25] Ansys INC. Theory Reference, Element Library

[26] University of Alberta ANSYS Tutorials – “Simple Conduction”

[27] University of Alberta ANSYS Tutorials – “Mixed boundary”

[28] University of Alberta ANSYS Tutorials – “Transient ConductionSaeed Moaveni – “Finite Elemente Analysis –

Theory and Applicationswith Ansys” Prentice Hall, Upper Saddle River, New Jersey, 1999

[29] University of Alberta ANSYS Tutorials – “Coupled and Structural analysis”

[30] British Stainless Steel Association – AISI 304 Thermo-Mechanical Properties

[31] Defence Industry Group - “Distortion control in shipbuilding”

- 1 -

9. Annexes

9.1. DCT Machine

As previously mentioned, DCT are indispensable to weld the two plates of the water jacket, however they are non-

added value activities that are time consuming, which makes the water jackets more expensive.

During the course of this work a device was developed to minimize the set up time of the DCT suggested (Pre-bending

and clamping).

By marking the first weld of each set in the plates, the work could quickly be done only by sliding the plates through the

machine.

- 2 -

9.2. MATLAB’s Programme

N = input('Número de divisões da fonte de Calor:');

T = N + 15;

%### GEOMETRIA ############################################################

cd('C:\Documents and Settings\Valter Carrolo\Ambiente de trabalho');

fid = fopen(sprintf('Geometria_N=%0.d.txt',N),'w'); % Criação da Geometria

if fid == -1,error('Error opening the file'),end

fprintf(fid,'/PREP7\n\n');

fprintf(fid,'/NOPR \n\n');

fprintf(fid,'/PMETH,OFF,0\n');

fprintf(fid,'KEYW,PR_SET,1\n');

fprintf(fid,'KEYW,PR_STRUC,1\n');

fprintf(fid,'KEYW,PR_THERM,1\n');

fprintf(fid,'KEYW,PR_FLUID,0\n');

fprintf(fid,'KEYW,PR_ELMAG,0\n');

fprintf(fid,'KEYW,MAGNOD,0\n');

fprintf(fid,'KEYW,MAGEDG,0\n');

fprintf(fid,'KEYW,MAGHFE,0\n');

fprintf(fid,'KEYW,MAGELC,0\n');

fprintf(fid,'KEYW,PR_MULTI,1\n');

fprintf(fid,'KEYW,PR_CFD,0\n');

fprintf(fid,'/GO\n\n');

fprintf(fid,'ET,1,SOLID98\n\n');

fprintf(fid,'KEYOPT,1,1,0\n');

fprintf(fid,'KEYOPT,1,3,0\n');

fprintf(fid,'KEYOPT,1,5,2\n\n');

fprintf(fid,'TOFFST,273\n\n');

fprintf(fid,'MPTEMP,,,,,,,,\n');

fprintf(fid,'MPTEMP,1,20\n');



- 3 -





fprintf(fid,'MPDATA,KXX,1,,15\n');

fprintf(fid,'MPDATA,KXX,1,,16.3\n');





fprintf(fid,'MPDATA,KXX,1,,25.1\n\n');







fprintf(fid,'MPDATA,C,1,,500\n');




fprintf(fid,'MPDATA,C,1,,630\n\n');



fprintf(fid,'MPDATA,DENS,1,,7930\n\n');

fprintf(fid,'BLC4,,,0.04,0.035,0.00125\n');

fprintf(fid,'BLC4,,,0.04,0.035,-0.004\n');

fprintf(fid,'CYL4,0.02,0.0175,0.014,,,,0.00125\n');

fprintf(fid,'CYL4,0.02,0.0175,0.014,,,,-0.004\n\n');

fprintf(fid,'VSBV,1,3\n');


fprintf(fid,'numcmp,KP\n');

fprintf(fid,'numcmp,line\n');

fprintf(fid,'numcmp,volu\n\n');

- 4 -

fprintf(fid,'CYL4,0.02, 0.0175, 0.0069, 0, 0.0071, 360, -0.004\n');

fprintf(fid,'CYL4,0.02, 0.0175, 0.0069, 0, 0.0071, 360, -0.00175\n');

fprintf(fid,'numcmp, volu\n\n');




fprintf(fid,'numcmp,volu\n\n');

for i = 4:1:N+4

x = i - 4;

fprintf(fid,'CYL4,0.02,0.0175,0.0069, %0.d*(360/%0.d), 0.0071, (360/%0.d) + %0.d*(360/%0.d), -

0.00175\n',x,N,N,x,N);

end

end

fprintf(fid,'\n');

for i = 4:1:N+4

x = i - 4;

fprintf(fid,'CYL4,0.02,0.0175,0.0069, %0.d*(360/%0.d), 0.0071, (360/%0.d) + %0.d*(360/%0.d),

0.00125\n',x,N,N,x,N);

end

end

fprintf(fid,'\n');



fprintf(fid,'numcmp,volu\n\n'); % (os volumes do cordão são 4 a 2xN+3)

for y = 4:1:N+3

fprintf(fid,'FLST,2,2,6,ORDE,2\n');

fprintf(fid,'FITEM,2,%0.d\n',y);

fprintf(fid,'FITEM,2,%0.d\n',y+N);

fprintf(fid,'VADD,P51X\n\n');

end

fprintf(fid,'CYL4,0.02,0.0175,0.014, 0, 0.012, 360, -0.004\n\n');

- 5 -





fprintf(fid,'CYL4,0.02,0.0175,0.002, 0, 0, 360, -0.004\n\n');

fprintf(fid,'CYL4,0.02,0.0175,0.014, 0, 0.012, 360, 0.00125\n\n');





fprintf(fid,'CYL4,0.02,0.0175,0.002, 0, 0, 360, 0.00125\n\n');



fprintf(fid,'numcmp,volu\n\n'); % (os volumes do cordão são 4 a N+3)

fprintf(fid,'FLST,2,%0.d,6,ORDE,2\n',T);

fprintf(fid,'FITEM,2,1\n');

fprintf(fid,'FITEM,2,-%0.d\n',T);

fprintf(fid,'VGLUE,P51X\n\n');



fprintf(fid,'numcmp,volu\n\n'); % (os volumes do cordão são 4 a N+3)

fprintf(fid,'LESIZE,ALL,0.001,,,,1,,,1,\n');

fprintf(fid,'MSHKEY,0\n');

fprintf(fid,'MSHAPE,1,3d\n');

fprintf(fid,'FLST,5,%0.d,6,ORDE,2\n',T);


fprintf(fid,'FITEM,5,-%0.d\n',T);

fprintf(fid,'CM,_Y,VOLU\n');

fprintf(fid,'VSEL,,,,P51X\n');

fprintf(fid,'CM,_Y1,VOLU\n');

fprintf(fid,'CHKMSH,''VOLU''\n');

fprintf(fid,'CMSEL,S,_Y\n\n');

fprintf(fid,'VMESH,_Y1\n\n');

- 6 -

fprintf(fid,'CMDELE,_Y\n');

fprintf(fid,'CMDELE,_Y1');

fclose(fid);

%### GEOMETRIA ############################################################

%### CONDIÇÕES FRONTEIRA ##################################################

fid = fopen(sprintf('Condicoes_Fronteira_N=%0.d.txt',N),'w'); % Condições Fronteira

if fid == -1,error('Error opening the file'),end

fprintf(fid,'/SOL\n\n');

fprintf(fid,'ANTYPE,4\n');

fprintf(fid,'TRNOPT,FULL,4\n');

fprintf(fid,'LUMPM,0\n');

fprintf(fid,'TREF,25,\n\n');



fprintf(fid,'FITEM,2,-101056\n');

fprintf(fid,'IC,P51X,ALL,25, ,\n\n');



fprintf(fid,'/GO\n');

fprintf(fid,'BFV,P51X,HGEN,8e11*(%0.d/50)\n',N);

fprintf(fid,'TIME,1+5/%0.d\n',N);

fprintf(fid,'AUTOTS,1\n');

fprintf(fid,'DELTIM,1,1,1+5/%0.d,1\n',N);

fprintf(fid,'KBC,1\n');

fprintf(fid,'TSRES,ERASE\n');

fprintf(fid,'LSWRITE,1,\n\n');

for z = 5:1:N+3

fprintf(fid,'FLST,2,N+15,6,ORDE,2\n');


fprintf(fid,'FITEM,2,-N+15\n');

fprintf(fid,'BFVDELE,P51X,HGEN\n\n');

- 7 -


fprintf(fid,'FITEM,2,%0.d\n',z);



fprintf(fid,'TIME,1 + (%0.d-3)*(5/%0.d)\n',z,N);


fprintf(fid,'DELTIM,1 + (%0.d-4)*(5/%0.d),1 + (%0.d-4)*(5/%0.d),1 + (%0.d-3)*(5/%0.d),1\n',z,N,z,N,z,N);



fprintf(fid,'LSWRITE,%0.d,\n\n',z-3);

end





fprintf(fid,'TIME,10\n');


fprintf(fid,'DELTIM,6,6,10\n');+



fprintf(fid,'LSWRITE,1,\n\n');

fclose(fid);

%### CONDIÇÕES FRONTEIRA ##################################################

Laser Welding Heat Distortions on Thin Plates

Documents

Transcript of Laser Welding Heat Distortions on Thin Plates