The Science and Practice of Welding

7/16/2019 The Science and Practice of Welding

1/362

The Science and Practice of Welding, now in its tenth edition andpublished in two volumes, is an introduction to the theory and practice ofwelding processes and their applications.Volume 1, Welding Science and Technology, explains the basic principlesof physics, chemistry and metallurgy as applied to welding. The section onelectrical principles includes a simple description of the silicon diode andresistor, the production and use of square wave, and one-knob steplesscontrol of welding current. There is a comprehensive section on non-destructive testing (NDT) and destructive testing of welds and crack tipopening displacement testing. The text has been brought completely up todate and now includes a new chapter devoted to the inverter power unit.Duplex stainless steel has been included in the list of materials described.Volume 2, The Practice of Welding, is a comprehensive survey of thewelding methods in use today, and gives up-to-date information on alltypes of welding methods and tools. Processes described include: m anualmetal arc welding (MMA or SMAW); gas shielded arc welding (MIG,MAG or GMAW); tungsten inert gas welding (TIG or GTAW) andplasma arc welding (PA) and cutting. Resistance, flash butt and oxy-acetylene welding are also included. Cutting processes are given a separatechapter. This new edition has been brought right up to date with a newchapter on the welding of plastics, and new sections on the welding ofduplex stainless steel and air plasma cutting. The text is illustrated byup-to-date photographs of plant and equipment. As in previous editions,the appendices bring together a wealth of essential information, includingBritish and American welding symbols, tables of conversion, informationon proprietary welding gases and mixtures, testing practices, safetyfeatures and tables of brazing alloys and fluxes.Both volumes contain numerous questions of the type set at craftsmanand technician grade of the City and Guilds of London Instituteexaminations.


2/362


3/362

The science and practice of weldingVolume 1Welding science and technology


4/362


5/362

The science and practiceof weldingVOLUME 1Welding science andtechnologyA.C.DAVIESB.Sc (London Hons. and Liverpool), C.Eng., M.I.E.E., Fellow of the Welding Institute

TENTH EDITION

C A M B R I D G EUNIVERSITY PRESS


6/362

Published by the Press Syndicate of the University of CambridgeThe Pitt Building, Trumpington Street, Cambridge CB2 1RP40 West 20th Street, New York, NY 10011-4211, USA10 Stamford Ro ad, O akleigh, Victoria 3166, Australia Cambridge University Press 1963, 1972, 1977, 1984, 1989, 1992First published 1941Second edition 1943Third edition 1945Reprinted 1947, 1950Fourth edition 1955Reprinted 1959Fifth edition 1963Reprinted 1966, 1969, 1971Sixth edition 1972Reprinted 1975Seventh edition 1977Reprinted with revisions 1981Eighth edition 1984Reprinted with new Appendix 3 1986Ninth edition 1989Reprinted 1990Tenth edition 1992A catalogu e record for this book is available from the British LibraryLibrary of Congress cataloguing in publication data availableISBN 0 521 43403 3 hardbackISBN 0 521 43565 X paperback

Transferred to digital printing 2003

UP


7/362

Contents

Preface ixTh e me tric system and the use of SI units xiWelding science 1Heat 1Behaviour of metals under loads 22Chem istry applied to welding 33Fluxes 55Metallurgy 65Pro duc tion and prope rties of iron and steel 65Effect of add ition of car bon to pu re iron 80Alloy steels 85Effect of heat on the stru ctu re of steel 94Effect of welding on the stru ctu re of steel 103Effect of defo rm ation on the pro per ties of me tals 115Non -ferrous metals 120Stress and disto rtion in welding 138Metallic alloys and equilibrium diagrams 152Me tallic alloys 152Equ ilibrium diag ram s and their uses 154Basic electrical principles 163Electrical technology 163Rectifiers 199Welding gene rators 206Alterna ting current welding 218Earthing 238The inverter 241Transformer and inductor 242Inverter design and ope ration 246Co ntro l systems in inverters 247


8/362

viii ContentsIro n loss 249Summary 251

6 Inspection and testing of welds 253Non -destructive tests 255De structive tests 272Some notes on Crack Tip Open ing Displacement 296

7 Engineering drawing and welding symbols 302Eng ineering drawing 302W elding symb ols 305Appendixes 3101 Tab les of elem ents and conversions 3102 Selection of British Sta nd ard s relating to welding 3173 No tes on Guidance on some methods for

the derivation of acceptance levels fordefects infusion welded joints PD6493 322

City and Guilds of London Institute examinationquestions: Welding science, metallurgy andtechnology 324Index 340


9/362

Preface

The Science and Practice of Welding was divided into two volumes for theeighth edition: Volume 1, Welding Science and Technology; Volume 2,The Practice of Welding. Volum e 1 covers all the basic material on thephysics and chemistry of welding, including metallurgy, equilibriumdiagrams, testing of welds, drawing and welding symbols and an appendixwith tables and conversions. Volume 2 gives a comprehensive survey ofthe welding methods in use today and the tools involved.This tenth edition has been brought completely up to date throughout.Volum e 1 has a new chap ter on the inverter, which has become po pul aras a power unit because of its reduced weight and size compared with aconv entional unit. There is also an up-to -date section on the classificationof stainless steels. Volume 2 has a new chapter on welding plastics andnew sections on welding duplex stainless steel and air plasma cutting.There are two new appendices (one illustrating the latest plant andequipment, and one on refraction). The appendix on proprietary weldinggases has been completely revised.

My thanks are due to each of the following firms who have renderedevery technical assistance and have supplied information and photographsas indicated.

Air Products Ltd: NDT of welding fabrications and all radiographs.Alcan W ire Ltd : classification and welding of alum inium and its alloys.Alpha Electronics Ltd: instrument technology.Aluminium Federation: classification and properties of aluminium andits alloys.Amersham International: X-ray and gamma-ray testing of welds, and

gamma-ray sources.


10/362

Andrex Products (NDT) Ltd: sources of X-rays and the testing ofwelds, with illustrations.Baugh and Weedon L td: ultrasonic testing of welds, with photographs.BOC Ltd : technology of stainless steel; and BOC Ltd (Gases Division):gas technology.Copper Development Association: classification of copper and itsalloys.INCO Alloys International Ltd, Hereford; details of nickel-copper,nickel-chromium and nickel-chromium-iron alloys.Magnesium Elektron Ltd: classification and properties of magnesiumand its alloys.

Megger Instruments Ltd: instrument technology.Murex Welding Products: thyristor control of welding sources;chemistry of submerged arc fluxes.Radiographic Supplies Ltd: diagrams and information on magneticparticle, ultrasonic and dye penetrant testing of welds.RGB Stainless Steels Ltd (Smethwick): classification of stainless steel.Salford Electrical Instruments Ltd: link testing ammeter.Testrade Ltd: information and photographs of gamma-ray testing ofwelds.The Welding Institute : information on crack tip opening displacement;

information and photograph on alloy steel electrodes.G. J. Wogan Ltd: information and testing of welds, with photographs.My thanks are also due to Messrs D. G. J. Brunt, A. Ellis, C. Owen,M. S. Wilson, R. A. Wilson and Dr S. J. Garwood for technical helpreceived in the preparation of the ninth edition, to Messrs R. P. Ham pson,R. Moulding, J. B. Stokes and P. V. C. Watkins for the tenth edition, andto the City and Guilds of London Institute for permission to include some

up-to-date questions in welding science and technology to add to thequestions at the end of the book.Extracts of up-to-date British Standards are included by permission ofthe British Standards Institution. Copies of the standards can be obtainedfrom The Sales Department, British Standards Institution, LinfordWood, Milton Keynes MK14 6LE.

Oswestry A. C. Davies1992


11/362

The metric system and the use of SI units

The m etric system was first used in Fra nc e after the Fren ch Re volution andhas since been adopted for general measurements by all countries of theworld except the United States. For scientific measurements it is generallyused universally.

It is a decimal system, based on m ultiples of ten, the following mu ltiplesand sub-multiples being added, as required, as a prefix to the basic unit.

Prefixes for SI unitsPrefix Symbol Factoratto a 10"18femto f 10"15pico p 10"12nano n 10 ~9micro // 10 ~6milli m 10"3centi c 10 "2deci d 10 ~ ldeca da 10lhecto h 102kilo k 103mega M 106giga G 109tera T 1012peta P 1015exa E 1018

Examples of the use of these multiples of the basic unit are: hectobar,milliampere, meganewton, kilowatt.

In past years, the C GS system, using the centimetre, gram an d second asthe basic units, has been used for scientific measurements. It was latermodified to the MKS system, with the metre, kilogram and second as the

xi


12/362

xii The metric system and the use of SI unitsbasic units, giving many advantages, for example in the field of electricaltechnology.

Note on the use of indicesA velocity measured in metres per second may be written m/s,

indicating that the second is the den om inato r, thus : .or . Since second s = a~n, the velocity can also be expressed as metre second" 1 o r m s ' 1 .This method of expression is often used in scientific and engineeringarticles. Oth er exam ples are; pressure and stress: newton per squa re m etreor pascal (N/m 2 or Nmf 2); density: kilograms per cubic metre (kg/m 3 orkg m" 3 ) .

SI units (Systeme Internationale d'Unites)To rationalize and simplify the metric system the Systeme

Internationale d 'Unites was adopted by the ISO (International Organiz-ation for Sta nda rdiza tion). In this system there are six prima ry units, thus :

Quantity Basic SI unit Symbollengthmasstimeelectric currenttemperatureluminous intensity

metrekilogramsecondamperekelvincandela

mkgsAKcd

In addition there are derived and supplementary units, thus:Quantity Unit Symbolplane angle radian radarea square metre m2volume* cubic metre m3velocity metre per second m/sangular velocity radian per second rad/sacceleration metre per secondsquared m/s2frequency hertz Hzdensity kilogram per cubicmetre kg/m3force newton Nmoment of force newton per metre N/mpressure, stress newton per square N/m 2metre (or pascal, Pa)


13/362

The metric system and the use of SI units xin

Quantitysurface tensionwork, energy, quantity ofheatpower, rate of heat flowimpact strengthtemperaturethermal coefficient of linearexpansionthermal conductivitycoefficient of heat transferheat capacityspecific heat capacityspecific latent heatquantity of electricityelectric tension, potentialdifference, electromotiveforceelectric resistanceelectric capacitancemagnetic fluxinductancemagnetic flux densitymagnetic field strengthmagnetomotive forceluminous fluxluminanceillumination

Unitnewton per metrejoulewat tjoule per squaremetredegree Celsiusreciprocal degreeCelsius or kelvinwatt per metredegree Cwatt per squaremetre degree Cjoule per degree Cjoule per kilogramdegree Cjoule per kilogramcoulomb

voltohmfaradweberhenryteslaampere per metreamperelumencandela per squaremetrelux

SymbolN /m

J (N/m)W (J /s)

J /m 2C C \ K " 1

W/m CW /m 2 CJ/CJ/kg C

J/kgC ( A s )

V (W/A)Q (V/A)FWbHT (Wb/m 2 )A/mAlm

cd/m 2lx* Note. N m 3 is the sam e as m 3 at norma l tempe rature and pressure, i .e. 0C and 760mm H g (NTP or STP).

Th e litre is used instead of the cubic decim etre (1 litre = 1 dm 3) and is used in the welding industry toexpress the volume of a gas.

Pressure and stress may also be expressed in bar (b) or hectobar (hbar) instead of newton per squaremetre.

Conversion factors from British units to SI units are given in the appendix.1 metric ton ne = 1000 kg.


14/362


15/362

1Welding science

HeatSolids, liquids and gases: atomic structureSubstances such as copper, iron, oxygen and argon which cannotbe broken down into any simpler substances are called elements; there areat the present time over 100 known elements. A substance which can bebroken down into two or more elements is known as a compound.An atom is the smallest particle of an element which can take part in achemical reaction. It consists of a number of negatively charged particlestermed electrons surrounding a massive positively charged centre termed

the nucleus. Since like electric charges repel and unlike charges attract, theelectrons experience an a ttraction due to the positive charge on the nucleus.Chemical compounds are composed of atoms, the nature of the compounddepending upon the number, nature and arrangement of the atoms.A molecule is the smallest part of a substance which can exist in the freestate and yet exhibit all the properties of the substance. Molecules ofelements such as copper, iron and aluminium contain only one atom andare monatomic. Molecules of oxygen, nitrogen and hydrogen contain two

atoms and are diatomic. A molecule of a compound such as carbon dioxidecontains three atoms and complicated compounds contain many atoms.An atom is made up of three elementary particles: (1) protons, (2)electrons, (3) neutrons.The proton is a positively charged particle and its charge is equal andopposite to the charge on an electron. It is a constituent of the nucleus ofall atoms and the simplest nucleus is that of the hydrogen atom, whichcontains one proton.The electron is 1/1836 of the mass of a proton and has a negative chargeequal and opposite to the charge on the proton. The electrons form a cloudaround the nucleus moving within the electric field of the positive chargeand around which they are arranged in shells.

1


16/362

2 Welding scienceThe neutron is a particle which carries no electric charge but has a mass

equal to that of the proton and is a constituent of the nuclei of all atomsexcept hydrogen. T he atom ic numb er of an element indicates the num ber ofprotons in its nucleus and because an atom in its normal state exhibits noexternal charge, it is the same as the number of electrons in the shells.

Isotopes are forms of an element which differ in their atomic mass butno t in some of their chemical properties. Th e atom ic weight of an isotope isknown as its mass number. F or example, an atom of carbon has 6 proton sand 6 neu tron s in its nucleus so that its atom ic num ber is 6. Othe r carbo natoms exist, however, which have 7 neu tron s and 8 neu tron s in the nucleus.These are termed isotopes and their mass numbers are 13 and 14respectively, compared with 12 for the no rm al carbo n atom . On e isotope ofhydrogen, called heavy hydrogen or deuterium, has a mass number 2 sothat it has one proton and one neutron in its nucleus.Electron shells. The classical laws of mechanics as expounded by Newtondo not apply to the extremely minute world of the atom and the density,energy and position of the electrons in the shells are evaluated by qu an tumor wave mecha nics. Since an at om in its no rm al state is electrically n eutra l,if it loses one or m ore electrons it is left positively charge d and is kn ow n as apositive ion; if the atom gains one or more electrons it becomes a negativeion. It is the electrons which are displaced from their shells, the nucleus isunaffected, and if the electrons drift from shell to shell in an organ ized wayin a completed circuit this constitutes an electric current.

In the periodic classification, the elements are arranged in order of theirmass numbers, horizontal rows ending in the inert gases and verticalcolumns having families of related elements.

The lightest element, hydrogen, has one electron in an inner shell and thefollowing element in the table , helium, has tw o electron s in the inner shell.This shell is now complete so that for lithium, which has three electrons,two occupy the inner shell and one is in the next outer shell. Withsucceeding elements th is shell is filled with e lectrons un til it is comp lete withthe inert gas neon, w hich has two electron s in the inner shell and eight in theouter shell, ten electrons in all. Sodium has eleven electrons, two in theinner, eight in the second an d o ne in a further ou ter shell. Electron s now fillthis shell with succeeding elements until with argon it is temporarily filledwith eight electrons so that argon has eighteen electrons in all. This isillustrated in Fig. 1.1 and this brief study will suffice to indicate how atomsof the elements differ from each other. Succeeding elements in the tablehave increasing n um be rs of electrons w hich fill mo re shells un til the table is,at the present time, comp lete with ju st over 100 elements.


17/362

Heathydrogen

lithium21sodium281

helium2beryllium~)

magnesium

The shells22222222

82

boron23aluminium283

are then filled8.8888888

8.18.18 8.18 18.18 1818 3218 32

carbon24silicon284

up thus:

8.18.18 \

nitrogen25phosphorus85

i.

oxygen26sulphur

86

fluorine27chlorine287

neon-)z8argor

88

T he electron s in their shells possess a level of energy and with any chan gein this energy light is given out or absorbed. The elements with completedor tem porarily comp leted shells are the inactive or inert gases helium, neon,argon, xenon and rado n, whereas when a shell is nearly comp lete (oxygen,fluorine) or has only one or two electrons in a shell (sodium, magnesium),the element is very reactive, so that the characteristics of an element aregreatly influenced by its electron structure. When a metal filament such astung sten is hea ted in a vac uu m it emits electrons, an d if a positively ch argedplate (anode) with an aperture in it is put in front near the filament, theelectrons stream through the aperture attracted by the positive charge andform an electron beam . Th is beam can be focused a nd guided and is used inthe television tube, while a beam of higher energy can be used for weldingby the electron beam process (see Volume 2).

Fig. 1.1

LITHIUM2 1

BERYLLIUM2 2


18/362

4 Welding scienceIf the atoms in a substance are not grouped in any definite pattern the

substance is said to be amorphous, while if the pattern is definite thesubstance is crystalline. Solids owe their rigidity to the fact that the atomsare closely packed in geometrical patterns called space lattices which, inmetals, are usually a simple pattern such as a cube. The positions whichatoms occupy to make up a lattice can be observed by X-rays.

Ato ms vib rate ab ou t their mean position in the lattice, and when a solidis heated the heat energy supplied increases the energy of vibration of theatoms until their mutual attraction can no longer hold them in position inthe lattice so tha t the lattice collapses, the solid melts and tur ns in to a liquidwhich is amorphous. If we continue heating the liquid, the energy of theatoms increases until those having the greatest energy and thus velocity,and lying near the surface, escape from the attraction of neighbouringatoms and become a vapo ur or gas. Eventually when the va po ur pressure ofthe liquid equals atm osph eric pressure (or the pressure abo ve the liquid) theatoms escape wholesale throughout the mass of the liquid which changesinto a gaseous state and the liquid boils.

Suppose we now enclose the gas in a closed vessel and continue heating.The atoms are receiving more energy and their velocity continues toincrease so that they will bombard the walls of the vessel, causing thepressure in the vessel to increase.

Ato m s are grouped into molecules, which may be defined as the smallestparticles which can exist freely and yet exhibit the chemical properties ofthe original substance. If an atom of sulphur, tw o atom s of hydrog en, andfour a tom s of oxygen com bine, they form a molecule of sulphu ric acid. Themolecule is the smallest particle of the acid w hich can exist, since if we splitit up we are back to the original atoms which combined to form it.

Fr om the foregoing, it can be seen tha t the three states of ma tt e r- solids,liquids and gases - are very closely related, and that by giving or takingaway heat we can change from one state to the other. Ice, water and steamgive an everyday example of this change of state.

Metals require considerable heat to liquefy or melt them, as for example,the large furnaces necessary to melt iron and steel.

We see examples of metals in the gaseous state when certain metals areheated in the flame. The flame becomes coloured by the gas of the metal,giving it a characteristic colour, and this colour indicates what metal isbeing heated. Fo r example, sodium gives a yellow coloratio n and copper agreen coloration.

This change of state is of great importance to the welder, since he isconcerned with the joining together of metals in the liquid state (termed


19/362

Heatfusion welding) and he has to supply the heat to cause the solid metal to beconverted into the liquid state to obtain correct fusion.

Temperature :* thermometers and pyrometersTh e tem pera ture of a body determines whether it will give heat to ,

or receive heat from, its surroundings.Ou r sense of determ ining hotness by tou ch is extremely ina ccurate, since

iron will always feel colder than wood, for example, even when actually atthe same temperature.

Instruments to measure temperature are termed thermometers andpyrometers. Thermometers measure comparatively low temperatures,while pyrometers are used for measuring the high temperatures as, forexample, in the melting of metals.

In the thermo me ter, use is ma de of the fact th at som e liquids expand by agreat am ou nt when heated. Me rcury and alcoh ol are the usu al liquids used.Mercury boils at 375C and thus can be used for measuring temperaturesup to about 330C.

M ercury is contained in a glass bulb which conn ects into a very fine bo reglass tub e called a capillary tu be an d up w hich the liquid expa nds (Fig . 1.2).

The whole is exhausted of air and sealed off. The fixed points on athermometer are taken as the melting point of ice and the steam from purewater at boiling point at standard pressure (760 mm mercury).

In the Celsius or C entigrad e therm om eter the freezing po int is ma rked 0and boiling point 100; thus there are 100 divisions, called degrees andshown thus . The Kelvin scale (K) has its zero at the absolute zero oftem pera ture, which is 273.16C To convert approximately from C to Kadd 273 to the Celsius figure.

Fig. 1.2. Celsius or Centigrade graduations. STEAM FROMBOILING WATERAT PRESSUREOF 760 mm Hg -

Kelvin scale of temperature. This scale of absolute temperature is a thermodynamictemperature scale based on the efficiency of a reversible heat engine. If the reversiblecycles are performed at various points on the scale, the amounts of work done on eachcycle are equal. The scale is compared with the gas scale of temperature by choosing thedegree size and defining the absolu te zero K elvin (K) as 273.16 C.


20/362

6 Welding scienceTo measure temperatures higher than those measurable with an ordinarythermometer we can employ:

(1) Temperature cones.(2) Temperature-indicating paints or crayons.(3) Pyrometers:(a ) Electrical resistance.(b ) Thermo-electric.(c) Radiation.(d ) Optical.(1) Tem perature cones (Seger cones) are triangular pyramids made of amixture of china clay, lime, quartz, iron oxide, magnesia, and boric acid invarying proportions so that they melt at different temperatures and can beused to measure temperatures between 600 C and 2000 C. They arenumbered according to their melting points and are generally used inthrees, numbered consecutively, of approximately the temperature re-quired. When the tem perature reaches that of the lowest melting point coneit bends over until its apex touches the floor. The next cone bends slightly

out of the vertical while the third cone remains unaffected. The temperatureof the furnace is that of the cone which has melted over.(2) Temperature-indicating paints and crayons either melt or changecolour or appearance at definite temperatures. Temperature indicators areavailable as crayons (sticks), pellets or in liquid form and operate on themelting principle and not colour change. They are available in a range from30 C to 1650 C and each crayon has a calibrated melting point. To use thecrayon, one of the temperature range required is stroked on the work as thetemperature rises and leaves a dry opaque mark until at the calibratedtemperature it leaves a liquid smear which on cooling solidifies to atranslucent or transparent appearance. Up to 700 C a mark can be madeon the work piece before heating and liquefies at the temperature of thestick. Similarly a pellet of the required temperature is placed on the workand melts at the appropriate temperature while the liquid is sprayed on tothe surface such as polished metal (or glass) which is difficult to mark with acrayon, and dries to a dull opaque appearance. It liquefies sharply at itscalibrated temperature and remains glossy and transparent upon cooling.(3) Pyrometers, (a) Electrical resistance pyrometers. Pure metals increasein resistance fairly uniformly as the temperature increases. A platinum wireis wound on a mica former and is placed in a refractory sheath, and the unitplaced in the furnace. The resistance of the platinum wire is measured (in aWheatstone's bridge network) by passing a current through it. As thetemperature of the furnace increases the resistance of the platinum


21/362

Heat 1increases and this increase is measured and the temperature read from achart.

(b) Thermo-electric (thermo-couple) pyrometers. When two dissimilarmetals are connected together at each end and one pair (or junction) of endsis heated while the other pair is kept cold, an electromotive force (e.m.f.) orvoltage is set up in the circuit (Peltier Effect). The magnitude of this e.m.f.depends upon (a) the metals used and (b ) the difference in temperaturebetween the hot and cold junctions. In practice the hot junction is placed ina refractory sheath while the other ends (the cold junction) are connectedusually by means of compensating leads to a millivoltmeter which measuresthe e.m.f. produced in the circuit and which is calibrated to read thetemperature directly on its scale. The temperature of the cold junction mustbe kept steady and since this is difficult, compensating leads are used. Theseare made of wires having the same thermo-electric characteristics as thoseof the thermo-couple but are much cheaper and they get rid of the thermo-electric effect of the junction between the thermo-couple wires and the leadsto the millivoltmeter, when the temperature of the cold junction varies. Thecouples generally used are copper/constantan (60 % Cu, 40 % Ni) used up to300 C; chromel (90 % Ni, 10 % Cr)/alumel (95 % Ni, 3 % Mn, 2 % Al) up to1200 C; and platinum /platinum -rhodium (10% Rh) up to 1500 C (Fig.1.3a).

Fig. 1.3(a )

^ - ^ COMPENSATING LEADS THERMO-COUPLE WIRESAJUI I i A S / J / _ > HOT. L>-'- *\ ~r ~f >AJUNCTIONMLUVOLTMETER*

REFRACTORYTHERMO-COUPLE PYROMETER $HEL L

(MILLI-\1LTMETER

^THERMO COUPLE \L * CONCAVE MIRROR

SHIELD PROTECTINGCOUPLE FROMDIRECT RADIATION

VIEWING POINT

RADIATION PYROMETER

LENS FORMING IMAGE OF(c ) RED GLASS FILTER FILAMENT HOT BODY ON FILAMENT

\ETEROPTICAL PYROMETER


22/362

8 Welding science(c) Radiation pyrometers. These pyrometers measure the radiation

emitted from a hot body. A 'black body' surface is one that absorbs allradiation falling upon it and reflects none, and conversely will emit allradiations. For a body of this kind, E, the heat energy radiated, isprop ortion al to the fourth power of the absolute tempe rature, i .e. E oc T 4(Stefan-Boltzmann Law) so that E = kT4. If a body is however radiatingheat in the open, the ratio of the heat which it radiates to the heat that ablack body w ould rad iate at the same temp eratu re is termed the emissivity,e, and this varies with the nature, colour and temperature of the body.Knowing the emissivity of a substance we can calculate the true tempera-ture of it when radiating heat in the open from the equation:

(Apparent temperature)4(True temperature)4 = - emissivity(temperatures are on the absolute scale).In an actual radiation pyrom eter the radiated h eat from the hot source is

focussed on to a thermo -cou ple by me ans of a mirr or (the focussing can beeither fixed or adjustable) and the image of the hot body must cover thewhole of the thermo-couple. The e.m.f. generated in the thermo-couplecircuit is measured as previously described on a millivoltmeter (Fig. 1.3b).

(d ) Optical pyrom eters. The disap pearin g filament type is an exam ple ofthis class of pyrometer. A filament contained in an evacuated bulb like anelectric light bulb is viewed against the hot body as a background. Bymeans of a control resistor the colour of the filament can be varied byvarying the current passing through it until the filament can no longer beseen, hot body and filament then being at the same temperature. Anammeter measures the current taken by the filament and can be calibratedto read the temperature of the filament directly.

Th e judg ing of tempe ratures by colour is usually very inacc urate. If steelis heated, it undergoes a colour change varying from dull red to brilliantwhite. After considerable experience it is possible to estimate roughly thetemperature by this means, but no reliance can be placed on it (Fig. 1.3 c).

Temperature gradient and heat affected zoneThe gradient is the rate of change of a quantity with distance so

that the temperature gradient along a metal bar is the rate of change oftemperature along the bar.This can be illustrated by Fig. 1.4a, which shows the graph of

tem peratu re plotted ag ainst distance for a bar heated at one end, the otherend being cold. The hot portion loses heat by conduction, convection and


23/362

Heatradiation. The conduction of heat along the bar will be greater the betterthe thermal conductivity of the metal of the bar so that heat travels morequickly along a copper ba r than along a steel bar, and the graph shows thatthe hot end is losing heat to its surroundings much more quickly than thecolder parts of the bar. T he greater the difference of tempe rature betweentwo points, the more rapidly is the heat lost.

If two steel plates are welded together, Fig. 1.46 shows the temperaturegradient from the molten po ol to the cold pare nt plate on each side of theweld. The gradient on one side only is shown in the figure. That portion of

Fig. 1.4

LENGTH OF BAR, MMHOTC Z^COLDTEMPERA TURE GRA DIENT A L ONG BA R HE A TED A T ONE EN D

(a)TEMP. OFWELD A APOOL

AMBIENTTEMPERATUREWeld^ DISTANCE, MM

HAZ \ y HAZ -cold-- plate-

Weld-

TEMPERA TURE GRADIENT AND HAZ FOR A BUTT WELDBETWEEN TWO PLA TES (ONE SIDE ONL Y SHOWN)(b)

-Weld

HEA T D/SSIPA TION B Y CONDUCTION(c)


24/362

10 Welding sciencethe plate on either side of the weld, affected by the heat and in which theme tal suffers therm al disturban ce, is kno wn as the Heat-Affected Z on e(H AZ ) and the areas nearest the weld in which this disturbanc e is greatestund ergo a change in structure which may include recrystallization, refiningand grain growth (q.v.). The larger and thicker the plate the more quicklywill the molten pool lose heat (or freeze) by conduction and for this reasonan ar c should nev er be struck briefly o n a thick section cold pla te especiallyin certain steels as the sudden quenching effect may lead to cracking.

Fig. 1.4c shows that a butt weld loses heat by conduction in twodirectio ns while in a fillet jo int the he at h as three dire ction s of travel. Bothjoints also lose heat by convection and radiation.

Expansion and contractionWhen a solid is heated, the atoms of which it is composed vibrate

ab ou t their mean position in the lattice more and more . This causes them totake up more room and thus the solid expands.

Most substances expand when heated and contract again when cooled,as the atoms settle back into their normal state of vibration.

Metals expand by a much greater amount than other solid substances,and there are many practical examples of this expansion in everyday life.Gaps are left between lengths of railway lines, since they expand andcontract with atmospheric temperature changes. Fig. 1.5a shows theexpansion joint used by British Rail. With modern methods of trackconstruction only the last 100 m of rail is allowed to expand or contractlongitudinally irrespective of the total con tinuo us length of welded rail, andthis movement is well within the capacity of the expansion joint oradjustment switch.

Iro n tyres are mad e smaller than the wheel they are to fit. They are heatedand expand to the size of the wheel and are fitted when hot. On beingquickly cooled, they contract and grip the wheel firmly.

Large bridges are mounted on rollers fitted on the supporting pillars toallow the bridge to expand.

In welding, this expansion and c ontra ction is of the greatest im portan ce.Sup pose we have two pieces of steel bar ab ou t 1 m long. If these are settogeth er at an an gle of 90 , as shown, an d the n welded an d allowed to cool,we find that they have curled or bent up in the direction of the weld (Fig.1.56 and c).The hot weld metal, on contracting, has caused the bar to bend up asshown, and it is evident that considerable force has been exerted to do this.

A well-known example of the use to which these forces, exerted duringexpansion and contraction, are put is the use of iron bars to pull in orstrengthen defective walls of buildings.


25/362

Heat 11Plates or S pieces are placed on the threaded ends of the bar, whichprojects through the walls which need pulling in. The bar is heated toredness and nuts on each end are drawn up tight against the plates on thewalls. As the bar cools, gradually the walls are pulled in.Different metals expand by different amounts. This may be shown byriveting together a bar of copper and a bar of iron about 0.5 m long and 25mm wide. If this straight composite bar is heated it will become bent, withthe copper on the outside of the bend, showing that the copper expandsmore than the iron (Fig. 1.6). This composite bar is known as a bi-metalstrip and is used in engineering for automatic control of temperature.

Fig. 1.5

* DIRECTION OFTRAFFIC ADJUSTMENT SWITCH

PLATES ARESTRAIGHT BEFORE WELDING^(c)

SHAPE ON COOL/NGDUE TO CONTRACT/ONFig. 1.6COPPER STRIP

AIRON STRIP


26/362

12 Welding scienceCoefficient of linear expansionThe fraction of its length which a bar will expand when heated

through one degree rise in temperature is termed its coefficient of linearexpansion. (This also applies to contraction when the bar is cooled.) Thisfraction is very small; for example, for iron it is

121000000

T ha t is, a bar of iron length / wou ld exp and by / x for every1000 000degree rise in temperature. Hence, if the rise was f, the expansion

^ 2Th e fraction is usually deno ted by the letter a. Thus the1UUU 00Uincrease in length of a bar of original length /, made of material whosecoefficient of linear expansion is a, when heated through f is lat.

Thus, the final length of a bar when heated equals its original length plusits expansion, that is:

L = I + la tFina l length = original length + expansion.

This can also be written: L= / (1 + at).I la t

Length after beingheated through tC

ExampleGiven tha t the coefficient of linear expansion of copper is or 0.000017

1 UUO UUUper degree C,find he final ength of a bar of copper whose original length was 75 mm,when heated through 50 CFinal length = original length + expansion, i.e.:F inal length - 75 + ( 75 x ^ x 50

63 750 63757 5 +


27/362

Heat 13The above is equally true for calculating the contraction of a bar when

cooled.Table of coefficients of linear expansion of metals per degree CMetalLeadTinAluminiumCopperBrass

60 % copper,40 % zinc

a0.0000270.0000210.0000250.0000170.000020

MetalZincCast ironNickelWrought ironMild steel

a0.0000260.0000100.0000130.0000120.000012

Invar, a nickel-steel alloy containing 36% nickel, has a coefficient oflinear exp ansio n of only 0.000 000 9, tha t is, only -^ of tha t of mild steel,and thus we can say that invar has practically no expansion when heated.

The expansion and contraction of metal is of great importance to thewelder, because, as we have previously shown, large forces or stresses arecalled in to play w hen it takes place . If the m etal th at is being welded is fairlyelastic, it will stretch, or give, to these forces, and this is a great help,although stresses may be set up as a result in the welded metal. Somemetals, however, like cast iron, are very brittle and will snap rather thangive or show any elasticity when any force is applied. As a result, thegreatest care has to be taken in applying heat to cast iron and in welding itlest we introduce into the metal, when expanding and contracting, anyforces which will cause it to break. This will be again discussed at a laterstage.

Coefficient of cubical expansionIf we imag ine a solid being hea ted, it is evident tha t its volu me will

increase, because each side undergoes linear expansion.A cube, for example, has three dimensions, and each will expand

according to the previous rule for linear expansion. Suppose each face ofthe cube was originally length / and finally length L after being heatedthrough fC Let the coefficient of linear expansion be a per degree C.

The original volume was / x / x / = P.Each edge will have expanded, and for each edge we have:Final length L = I (1 + at) as before (Fig. 1.7).Thus the new volume = /(I + at) x /(I + at) x 1(1 + at)

= /3(1 + 3at) approximately.


28/362

14 Welding scienceThus, the final volume = original volume (1 + 3at).That is, the coefficient of cubical expansion may be taken as being three

times the coefficient of linear expansion.ExampleA brass cube has a volume of 0.006 m 3 (6 x 106 mm3) and is heatedthrough a 65 C rise in temperature. Find itsfinalvolume, given that thecoefficient of linear expansion of brass = 0.000 02 per degree C.

Final volume = original volume (1 + 3at)V= 0.006(1 + 3 x 0.00002 x 65)= 0.006(1.0039)= 0.006023 4m3.The joule and the newtonHe at is a form of energy a nd the un it of energy is the jou le (J). A

joule may be defined as the energy expended when a force of 1 newton (N)moves thro ugh a distance of 1 me tre (m). (N ote : a new ton is tha t forcewhich, acting o n a mass of 1 kilog ram (kg), gives it an acceleration of 1metre per second per second (1 m /s2). Th e gravitation al force on a mass of 1kg equals 9.81 N so that for practical purp oses, to con vert from kilogram sforce to newtons, multiply by 10.)

Specific heat capacityThe specific heat capacity is defined as the quantity of heat

required to raise unit mass of a substance through 1 rise in tem peratu re.Th e SI unit is in kilojoules per kilogram K (kJ /kg K , sym bol c). N ot e t ha tsince K adds 273 to C, a rise in temperature will be the same in K or C.

ExampleFind the heat gained by a mass of 20 kg of cast iron which is raisedthrough a temperature of 30 C, given that the specific heat capacity ofcast iron is 0.55 kJ/kg K.

Fig. 1.7


29/362

Heat 15Heat gained = mass x c x rise in temperature

= 20 x 0.55 x 30 kilojoules= 330 kJ.

Specific heat Specific heatSubstance capacity Substance capacityWaterAluminiumTinLeadCopperBrass

4.2 x0.91 x0.24 x0.13 x0.39 x0.38 x

103103103103103103

Mild steelWrought ironZincCast ironNickel

0.45 x0.47 x0.4 x0.55 x0.46 x

103103103103103

The above values are approximately 4.2 x 10 3 as great as the values when specific heat capacities wereexpressed in calories per gram degree C.

Melting pointThe melting point of a substance is the temperature at which the

change of state from solid to liquid occurs, and this is usually the sametemperature at which the liquid will change back to solid form or freeze.

Substances which expand on solidifying have their freezing pointlowered by increase of pressure while others which contract on freezinghave their freezing point raised by pressure increase.

The melting point of a solid with a fairly low melting point can bedetermined by attaching a small glass tube, with open end con taining someof the solid, to the bulb of a therm om eter. T he therm om eter is then placedin a container holding a liquid, whose boiling point is above the meltingpo int of the solid, and fitted with a cover, as shown in Fig. 1.8, an d a stirreris also included. The container is heated and the temperature at which thesolid melts is observed. The apparatus is now allowed to cool and thetemperature at which the substance solidifies is noted. The mean of thesetwo readings gives the melting point of the solid. By using m ercury, whichboils at 357C, as the liquid in the container, the melting point of solidswhich melt between 100 and 300 C could be obtaine d.

Determination of the melting point by method of coolingThe solid, of which the melting point is required, is placed in a

suitable container, fitted with a cork or stopper through which atherm om eter is inserted (Fig. 1.9). A hole in the stopp er preve nts pres surerise. Th e con tainer is heated until the solid m elts, and heating is continueduntil the temperature is raised well above this point. The liquid is nowallowed to cool and solidify and th e tempe ratur e is taken every quarte r or


30/362

16 Welding scienceFig. 1.8

STIRRER THERMOMETER

SOLID

Fig. 1.9

VENT HOLE


31/362

Heat 17half minute. This temperature is plotted on a graph against the time, andthe shape of the gra ph shou ld be as show n in Fig. 1.10.

If the melting po int of a me tal is requ ired, the m etal is placed in a fireclayor graphite crucible and heated by me ans of a furnace, and the tem pera tureis meas ured, at the same intervals, by a pyrome ter. Th e metal, on co oling,begins to solidify and form crystals in exactly the same way as any othersolid. The portion A shows the fall in temperature of the liquid or moltenmetal. Th e portion B indicates the steady tem pe ratu re while solidification istaking place, and portion C shows the further fall in temperature as thesolid loses heat. The temperature f of the portion B of the curve is themelting point of the solid.

In practice we may find tha t the tem pera ture falls below the dotted line,as shown, that is, below the solidifying temperature. This is due to thedifficulty which the liquid m ay experience in com me ncing to form crystals,and is called 'sup er-c oo ling '. It then rises again to th e true solidifying poin tand cooling then takes place as before (Fig. 1.11).

This method of determination of the melting point is much used infinding the melting point of alloys and in observing the behaviour of theconstituents of the alloys when melting and solidifying.

The melting point of a metal is of great importance in welding, since,Fig. 1.10

TIME.

Fig. 1.11

-EFFECT OFSUPER-COOUNC

TIME


32/362

18 Welding sciencetogether with the capacity for heat of the metal, it determines how muchheat is necessary for fusion. Th e addition of other substances o r metals toa given metal (thus forming an alloy) will affect its melting point.

Specific latent heatIf a block of ice is placed in a vessel with a therm om ete r and hea t is

applied, the tem pera ture rem ains steady a t 0 C (273 K) until the whole ofthe ice has been melted and then the temperature begins to rise. The heatgiven to the ice has no t caused any rise in tem per atur e b ut a chang e of statefrom solid to liquid and is called the specific latent he at of fusion. W hen thechange of state is from liquid to gas it is termed the specific latent heat ofvap oriza tion (o r evap oratio n) and is expressed in joule s or kilojoules (kJ )per kilogram (J/kg or kJ/kg).

Specific latent hea t of fusion in kJ/kgAluminiumCopperIron

393180205

NickelT inIc e

27358333

Specific latent heat of fusion is more important in welding than specificlatent heat of vapo rizatio n, because a com pariso n of these figures gives anindication of the relative am ou nts of heat required to chan ge the solid m etalinto the liquid state before fusion.

Since the hea t m ust be given to a solid to conv ert it to a liquid, it followsthat heat will be given out by a liquid when solidifying. This has alreadybeen demonstrated when determining the melting point of a liquid by themethod of cooling. When the change of state from liquid to solid takesplace (B on the curve in Fig. 1.10) heat is given out and the temperaturerem ains steady u ntil solidification is com plete, wh en it again begins to fall.

Transfer of heatHeat can be transferred in three ways: conduction, convection,

radiation.Conduction. If the end of a short piece of metal rod is heated in a flame, itrapidly gets too hot to hold (Fig. 1.12). Heat has been transferred byconduction from atom to atom through the metal from the flame to thehan d. If a rod of copper and one of steel are placed in the flame, the co pperrod gets hotter more quickly than the steel one, showing that the heat has


33/362

Heat 19been conducted by the copper more quickly than the steel. If the rods areheld in a cork and the cork gripped in the hand, they can now be heldcomfortably. The cork is a bad conductor of heat. All metals are goodconductors but some are better than others, and the rate at which heatis condu cted is termed the thermal conductivity and is measu red in watts permetre degree (W /m C).

Th e conductivity depen ds on the purity of the metal, its structure a nd thetemperature.

As the temperature rises the conductivity decreases and impurities in ametal greatly reduce the conductivity.

The thermal conductivity is closely allied to the electrical conductivity,that is, the ease with which an electric current is carried by a metal. It isinteresting to compare the second and third columns in the table. Fromthese we see that in general the better a metal conducts heat, the better itconducts electricity.

Table of comparative conductivities (taking copper as 100)

SilverCopperAluminiumZincNickelIronSteelTinLead

Thermalconductivity1061006229251713-1715

8

Electricalconductivity1081005629151713-1717

9

Fig. 1.12


34/362

20 Welding scienceTh e effect of cond uctivity of hea t on w elding practice ca n clearly be seen

from the calculations in Fig. 1.13, where a block of copper and one of steelof equal mass are to be welded. It is seen that if the two blocks were to beeach brought up throughout their mass to melting point, the steel wouldtake a much greater quantity of heat than the copper would.

When the heat is applied at one spot, copper being such a goodcond uctor, he at is rapidly transferred from this spot throu gh ou t its mass,and we find that the spot where the heat is applied will not melt until thewhole mass of the copper h as been raised to a very high tem pera ture indeed.

With the mild steel block, on the othe r han d, the heat con ductivity is onlyabout \ (from the table) that of the copper, that is, the heat is conductedaway at only \ the rate. Hen ce we find th at th e spot where the heat is appliedwill be raised to melting point long before the rest of the block has becomevery hot.

Because of this high conductivity of copper, it is usual to employ greaterheat than when welding the same thickness of steel or iron.

For this reason also, when welding copper, whether by arc or oxy-acetylene, it is always advisable to heat the w ork u p to a high tem pera tureover a large area aro un d the area t o be welded. In this way the heat will notbe con duc ted to colder regions so rapidly an d b etter fusion in the weld itselfcan be obtained.

Cast iron is a comparatively poor conductor of heat compared with

Fig. 1.13THIS POINT CANNOT BEMELTED UNTIL THE WHOLEMASS OF COPPER IS ATA HIGH TEMPERATUREDUE TO THE HIGHCONDUCTIVITY OFCOPPER

THIS POINT CAN BE MEL TEDBEFORE THE MASS OF THESTEEL IS AT A HIGHTEMPERATURE BECAUSESTEEL HAS ONLYTH THE CONDUCTIVITYOF COPPERLINE OF WELD

Mass 1 kgMelting point of copper 1083 CSpecific heat capacity 0.39 X1 0 3 J/kgCHeat required to raise copper

to melting point= 1 X 1083 X 0.39 X10 3 J= 422 370 J= 0.422 X10 6 J.

Mass 1 kgMelting point of steel 1400 CSpecific heat capacity 0 .45 X1 0 3 J/kgCHeat required to raise steel to

melting point= 1 X 1400 X 0.45 X1 0 3 J= 630 000 J= O.63X1O 6 J.


35/362

Heat 21copper. If we heat a casting in one spot, therefore, heat will only betransferred away slowly. The part being heated thus expands more quicklythan the surrounding parts and, since expansion is irregular, great forces,as before explained, are set up and, since cast iron is brittle and has verysmall elasticity, the casting fractures. The welding of cast iron is thus astudy of expansion and contraction and conduction of heat and, to weldcast iron successfully, care must be taken that the tem perature of the wholecasting is raised and lowered equally throughout its mass. This will bediscussed at a later stage.

Convection. When heat is transferred from one place to another by themotion of heated particles, this is termed convection. For example, in thehot water system of a house, heat from the fire heats the water and hotwater, being less dense than colder water, rises in the pipes, formingconvection currents and transferring heat to the storage tank.In the heat treatment of steel it is often necessary to cool the steel slightlymore quickly than if it cooled naturally, in order to harden it. It is cooled,therefore, in an air blast, the heat being transferred thus by convection.

Radiation. Heat is transferred by radiation as pulses of energy, termedquanta, through the intervening space. We sit in front of a fire and it feelswarm. There is no physical contact between our bodies and the fire. Theheat is being transferred by radiation. Heat transferred in this mannertravels according to the laws of light and is reflected and bent in the sameway.The sun's heat is transmitted by radiation to our planet but the methodby which the heat travels through space is not fully understood. Metal, ifallowed to cool in a still atmosphere, loses its heat by radiation and anyother bodies in the neighbourhood will become warmed.It is evident that the outside of the hot metal will lose heat more quicklythan the interior, and wefind,for example, that the surface of cast iron ismuch harder than below the surface, because it has lost heat m ore quickly.Chills are strips or blocks of metal placed adjacent to the line of weldduring the welding operation in order to dissipate heat and reduce the areaaffected by the input of heat, the heat-affected zone (HAZ). Heat isremoved by conduction, convection and radiation, and copper is often usedbecause of its good heat conducting properties. Heat control can be effectedby moving the chills nearer or further from the weld.


36/362

22 Welding scienceBehaviour of metals under loadsStress, strain and elasticityWhen a force, or load, is applied to a solid body it tends to alter theshape of the body, or deform it.

The atoms of the body, owing to their great attraction for each other,resist, up to a certain point, the attempt to alter their position and there isonly a slight distortion of the crystal lattice.If the applied force is removed before this point is reached, the body willregain its original shape.This property, which most substances possess, of regaining their originalshape upon removal of the applied load is termed elasticity.Should the applied load be large enough, however, the resistance of theatoms will be overcome and they will move and take up new positions in thelattice. If the load is now removed, the body will no longer return to itsformer shape. It has become permanently distorted (Fig. 1.14).

The point a t which a body ceases to be elastic and becomes permanentlydistorted or set is termed the yield point, and the load which is applied tocause this is the yield-point load. The body is then said to have undergoneplastic deformation or flow.Whenever a change of dimensions of a body occurs, from whatevercause, a state of strain is set up in that body. Strain is usually measured (forcalculation purposes) by the ratio or fraction:

change of dimensions in direction of applied loadoriginal dimensions in that direction

Fig. 1.14 (a) Original length of specimen, (b) Extension produced = lv Elasticlimit not reached. Fx = applied force, (c) Force removed, specimen recovers itsoriginal dimensions, (d) Extension produced = l2. Elastic limit exceeded byapplication of force. Specimen now remains permanently distorted or set, anddoes not recover its original dimensions when force is removed. F2 = appliedforce.

7? NOFIRMLY HELDt'/.y/'A

(a) ib) (0 id)


37/362

Behaviour of metals under loads 23ExampleA bar is 100 mm long and is stretched \ mm by an applied load along itslength. Find the strain.. _ change in length _ \ _ 1n ~ original length ~ TOO ~ 400'

The magnitude of the force or load on unit area of cross-section of thebody producing the strain is termed the stress.

Stress = force or load per unit area.Hooke's Law states that for an elastic body strain is proportional to

stress.Th e mass of a body is the quan tity of ma tter w hich it con tains, so that it is

depen dent u po n the num ber of atom s in its structure. M ass is measu red inkilogram s (kg) and 1000 gra m s (g) equ al 1 kg. No te 1 lb = 0.4536 kg and 1kg = 2.2 lb. Ne wto n's U niversal L aw of Gr av itatio n states tha t everyparticle of matter attracts every other particle of matter with a force (F )which is pro po rtio na l to the pro du ct of the masses (m 1 and m 2) of the tw oparticles and inversely proportional to the square of the distance (d )between them , Foe m1m2/d2. Th e weight of a bod y is the force by w hich it isattracted to the earth (the force of gravity), but because the earth is aflattened sphere, this force and hence the weight of the body vary som ewh ataccording to its position on the earth 's surface. On th e surface of the m oon ,which has abo ut one-sixth of the mass of the earth, a mass of one kilogramwould weigh about one-sixth of a kilogram. To distinguish a mass of onekilogram from a force of one kilogram, w hich is the force of attraction du eto the gravitational pull of the earth, the letter/is added thus, kgf.

Th e unit of force termed the new ton av oids the distinction between massan d weight and is defined as ' th at force w hich will give an a cceleratio n of 1metre per second per second to a mass of 1 kilogram'.

Units of stress or pressureThe following multiples of units are used:tera- (T) = on e million million 1012giga- (G ) = one thou sand million 109mega- (M ) = one million 106kilo- (k) = one thou san d 103hecto- (h) = one hund red 102.

The SI unit of stress or pressure is the newton per square metre (N/m 2 )which is also know n as the pascal (Pa), and 1 N /m 2 = 1 Pa. T his is a sm allunit and when using it to express tensile strengths of materials large


38/362

24 Welding sciencenumbers are involved with the use of the meganewton per square metre( M N / m 2 ) or megapascal (MPa).

If, h owever, the new ton per squ are millimetre (N /m m 2) is used, a s in thisbook, large figures are avoided and the change to the SI unit is easily madesince 1 N/ m m 2 = 1 MN /m 2 or 1 MP a.

Th e bar (b) and its multiple the hectob ar (h ba r) are also used as units ofpressure and stress. 1 bar is equ al to the pressure of a vertical column ofmercury 750 mm high and for conversion purp oses it can be take n to e qual15 lbf* per squa re inch. It should be noted tha t 1 ba r = 105 N /m 2 or 105 Pa,and 1 hba r = 10 N /m m 2 .

Gauges for cylinders of compressed gases can be calibrated in bar, acylinder pressure of 2500 lbf/in2 being 172 bar.Tensile strength can be expressed in hbar. A specimen of aluminium mayhave a tensile strength of 12 hbar which is equal to 120 N/mm 2 .

If stress is stated in tonf/in 2 or kgf/mm 2 the following con versions can b eused. (A full list of conversion factors is given in the appendix.)

Tonf/in 2 to MN/m 2 or N/mm 2 , multiply by 15.5; M N /m 2 or N/m m 2 totonf/in 2, multiply by 0.0647; kgf/m 2 to N/m 2 , multiply by 9.8; andapproximately 1 hba r = 1 kgf/mm 2.

If a stress is applied to a body and it changes its shape within its elasticlimits, the ratio stress/strain is termed the mo du lus of elasticity or Y ou ng 'smodulus (E ) of the material. Th e unit is N /m 2 or Pa, and a typical value fora specimen of aluminium is 69 x 103 M N / m 2 or MPa or N/mm 2 .

There are three kinds of simple stress: (1) tensile, (2) compression, (3)shear.

Tensile stressIf one end of a me tal rod is fixed firmly and a force is applied to the

othe r end to pull the rod, it stretches. A tensile force ha s been applied to therod and when it is measured on unit cross-sectional area it is termed atensile stress.

ExampleA force of 0.5 M N is applied so as to stretch a bar of cross-sectional area400 mm 2. Find the tensile stress.

load 500000 , XT / ,Tensile stress = : = . . . = 1250 N/m m 2 .area of cross-section 400

* 1 ba r = 14.508 lbf/in 2.1 lbf/in2 = 0.0689 bar.


39/362

Behaviour of m etals under loads 25A machine known as a tensile strength testing machine, which will be

described later (Chapter 5), is used for determining the tensile strength o fmaterials and welded joints.

The specimen under test is clamped between two sets of jaws, one fixedand one moving, and the force can be increased until the specimen breaks.

Suppose a piece of m ild steel is placed in the machine. A s the tensile stressis increased, the bar becomes only very slightly longer for each increase offorce. Then a point is reached when, for a very small increase of force, thebar becomes much longer. This is the yield point and the bar has beenstretched beyond its elastic limit, and is now deforming plastically.

If the applied load had been reduced before this point was reached, thebar would have recovered its normal size, but will not do so when the yieldpoint has been passed.

As the load is increased beyond the yield point the elongation of the barfor the same increase of loading becomes much greater, until a point isreached when the bar begins to get reduced in cross-sectional area andforms a waist, as shown. Less load is now required to extend the bar, sincethe load is now applied on a smaller area, the waist becomes smaller and thebar breaks. The accom panying diagram (Fig . 1.15) will make this clear.

When the stress is first applied, the extension of the bar is very small andneeds accurate measurement, but it is proportional to the load so that thegraph of stress/strain is a straight line. At one point X th e graph deviates alittle from the straight line OX, so that after X the strain is no longerproportional to the stress. The point X is the limit o f proportionality andHooke's Law is no longer obeyed. At Y the extension suddenly becomes

Fig. 1.15. Stress-strain diagram, mild steel.

' PLASTICI DEFORMATION

ULTIMATE ORMAXIMUM STRESS

WBREAKING POINT

LIMIT OF PROPOR TIONALITY

' ELASTIC DEFORM ATIONWAISTuFRACTURE

STRAIN


40/362

26 Welding sciencemu ch greater tha n before for an equal increase in load and Y is termed theyield point, the stress at this point being the yield-point stress.

Increase of load produces progressive increase of length to the point Z.At this point the waist forms; Z is the maximum load. Breakage occurs atS u n d e r a smaller load than at Z. A substance which has a fair elongationdu ring t he plastic stage is called du ctile, while if the elong ation is very sm allit is said to be brittle.

Given a table of tensile strengths of vario us m etals, we can calculate themaximum force or stress that any given section will stand.Table of tensile strengthsThe tensile strength of a metal depen ds upon its condition, whether cast, ann ealed,work-hardened, heat-treated, etc.

LeadZincTi nAluminiumCopper

N/mm 212-2230-453060-90220-300

hbar1.2-2.23-4.536- 920-30

tons f/in20.8-1.42-324-614-20

BrassCast ironWrought ironMild steelHigh tensilesteel

N /m m 2220-340220-300250-300380-450600-800

hbar22-3422-3025-3038-4560-80

tonsf/in214-2214-2016-2025-30

ExampleA certain grade of steel has a tensile strength of 450 N / m m 2 . What tensileforce in newtons will be required to break a specimen of this steel cross-section 25 mm x 20 mm?

Area of cross-section = 500 mm 2Forc e required = tensile strength x area of cross-section

F = 450 x 500 newtons= 225 000 new tons= 225 x 103 newtons = 0.225 MN.

Th e tensile strength of a metal dep end s largely upon the way it has beenworked (hammered, rolled, drawn, etc.) during manufacture, its actualcomposition and the presence of impurities (see Fig. 1.16).

From the tensile test we can obtain:/ ix *r- u yie ld s t ress ,^TI . t l ,(1) Yield point = ^ : (N/mm 2 or hbar).original area of cross-section(2) Ultimate tensile stress (UTS)

_ max imum stressoriginal area of cross section (N /m m 2 or hbar).


41/362

Behaviour of metals under loads(3) Percentage elongation on length between gauge marks

extension

27

original length between gauge marks x 100.Th e distance between the gauge mar ks can be 50 mm or 5 x diameter of thespecimen. Standard areas of cross-section can be 75mm 2 or 150 mm 2 .

(4) Percentage reduction of area (R of A)reduction of area at the fracture

original area x 100.A typical example for one particular grad e of weld metal is: Co mp osition:0.07% C, 0.4% Si, 0.68% Mn, remainder Fe. Yield stress 479 N/mm 2 ,ultimate tensile stress 556 N/mm 2 , elongation on gauge length of 5 x Z),26%; reduction of area, 58%.

The elongation will depend on the gauge length. The shorter this is thegreater the percentage elongation, since the greatest elongation occurs inthe short length where 'waisting' or 'necking' has occurred. Reduction inarea and elongation are an indication of the ductility of a metal.As tem pera ture rises there is usually a decrease in tensile strength and anincrease in elongation, and the limit of pro po rtion ality is reduced so that atred heat application of stress produces plastic deformation. A fall intemperature usually produces the opposite effect. Internal stresses which

Fig. 1.16. Stres s-strain diag ram , for a steel in (1) anne aled , (2) cold dra wncondition.

/

si1/ /\ \

ANNEALED

EXTENSION {STRAIN)


42/362

28 Welding sciencehave been left in a welded structure can be relieved by heating the membersand lowering the limit of pro po rtion ality. Th e stresses then pr odu ce plasticdeformation, and are relieved. This stress relief, however, may causedistortion.

Proof stressNon-ferrous metals, such as aluminium and copper, etc., and also

very har d steels, do n ot sh ow a definite yield poin t, as ju st exp lained, andload-extension curves are shown in Fig. 1.17. A force which will produce adefinite perm ane nt extension of 0.1% or 0.2% of the gauge length is know nas the proof stress (Fig. 1.18) and is measured in N/mm 2 or hbar.

Fig. 1.17

EXTENSION {STRAIN)

Fig. 1.18. Load-extension curve of hard steels and non-ferrous metalsillustrating proof stress.

PROOFSTRESS

PERMANENT EXTENSION'.OR SET BEGINS

* * EXTENSIONDEFINITE PERCE NTAGEOF PERMANENT SET


43/362

Behaviour of metals under loads 29Compressave stressIf the forces app lied in the previo us exp eriments on tensile streng th

are reversed, the body is placed under compression.Compressive tests are usually performed on specimens having a short

length compared with their diameter to prevent buckling when the load isapplied. Ductile metals increase in diameter to a barre l shape and crackingrou nd the periphe ry is some indication of the ductility of the specimen. Fo rpractical purposes E, the Yo ung 's modu lus, can be assumed to be the samefor compression and tension.

^ . compressive load (N)Compressive stress = -z : ; N/m m 2 .area of cross section (mm

2)A goo d exam ple of compressive stress is found in building and structu ral

work. All foundations, concrete, brick and steel columns are undercompressive stress, and in the making or fabrication of welded columns andsupports, the strength of welded joints in compression is of greatimportance.

Shearing stressIf a cub e has its face fixed t o the ta ble on which it stand s and a forceis applied parallel to the table on one of the upp er edges, this force per un it

area is termed a shearing stress and it will deform the cube, as indicated bythe dotte d line (Fig. 1.19). The angle 6 thro ug h w hich the cube is deformedis a measure of the shearing strain, while the shearing stress will be inN / m m 2 .

This is a very common type of stress in welded construction. Forexample, if two plates are lapped over each other and welded, then a loadapplied to the plates as shown pu ts the welds und er a shearing stress. If theload is kno wn and also the shearing strength of the metal of the weld, thensufficient metal can be deposited to withstand the load.

A welded structure should be designed to ensure that there is sufficientarea of weld metal in the join t to w ithstand safely the load required.

Fig. 1.19

\6


44/362

30 Welding scienceMechanical properties of metals and the effect of heat on thesepropertiesPlasticity may be defined as the ease with which a metal may bebent or moulded into a given shape. At ordinary temperatures, lead is one

of the most plastic metals. The plasticity usually increases as temperaturerises. Iron and steel are difficult to bend and shape when cold, but itbecomes easy to do this when heated above red heat. Wrought iron,however, because of impu rities in it, sometimes break s when we attem pt tobend it when ho t (called ho t shortness), and thus increase of tem pera tures isnot always accompanied by an increase in plasticity.Brittleness is the opposite of plasticity and denotes lack of elasticity. Abrittle metal will break w hen a force is applied. Cast iron and high carb onsteel are examples of brittle metals. The wrought iron in the aboveparagraph has become britt le through heating. Copper becomes britt lenear its melting point, but most metals become less brittle when heat isapplied. Ca rbo n steel is an exam ple; when cold it is extremely brittle, bu t caneasily be bent and worked when hot. Brittle metals require care whenwelding them, due to the lack of elasticity.Malleability is the prope rty possessed by a metal of becoming p erm ane ntlyflattened or stretched by hamm ering or rolling. Th e more m alleable a metalis, the thinner the sheets into which it can be hammered. Gold is the mostmalleable metal (the gold in a sovereign can be ham me red into 4 m 2 of goldleaf, less than 0.0025 mm thick).

Co ppe r is very malleab le, except nea r its me lting poin t, while zinc is onlymallea ble between 140 and 160 C . Metals such as iron and steel becomemuch more malleable as the temperature rises and are readily hammeredand forged.

The presence of any impurities greatly reduces the malleability, as wefind that the metal cracks when it stretches.Order of malleability when cold( l) G o ld (3) Aluminium (5) Tin (7) Zinc(2) Silver (4) Co ppe r (6) Lead (8) Iro n.Ductility is the prop erty possessed by a substan ce of being draw n out into awire and it is a property possessed in the greatest degree by certain metals.Like malleability this property enables a metal to be deformed mechani-cally. Metals are usually more ductile when cold, and thus wire drawingand tube drawing are often done cold, but not always.


45/362

Behaviour of metals under loads 31In the wire-drawing operation, wire is drawn through a succession of

tapered holes called dies, each operation reducing the diameter andincreasing the deformation of the lattice structure. The brittleness thusincreases and the wire must be softened again by a process termedannealing.Order of ductility( l ) G o l d(2) Silver (3 )(4 )

IronCopper (5 )(6 )

AluminiumZinc

(7 )(8 )

TinLead.

Tenacity is another name for tensile strength. The addition of varioussubstances to a metal m ay increase or decrease its tensile strength. Sulphurreduces th e tenacity of steel while car bo n increases it (see section o n T ensileStrength).Hardness is the property possessed by a metal resisting scratching orindentation. It is measured on various scales, the most common of whichare: (1) Brinell, (2) Rockwell, (3) Vickers.Table of comparative hardnessMaterialLeadTinAluminiumpureannealedZincCopper, castcold worked

Brinell614

194540-4580-100

Vickers615

204842-4885-108

MaterialBrass 70/30,annealedrolledCast ironMild steelStainless steel

Brinell

60150150-250100-120150-165

Vickers

64162160-265108-130160-180

Hardness decreases with rise in temperature. The addition of carbon tosteel greatly increases its hard ness after h eat treatm ent, and the ope rationsof rolling, drawing, pressing and hammering greatly affect it.

It will be noted that there is considerable latitude in the higher figures.Copper, for example, varies from 40 to 100 according to the way it isprepared. Copper is hardened by cold working, that is drawing, pressingand hammering, and this also decreases its ductility.

Th e tensile strength of steels can be approxim ately d etermined in N/ m m 2by multiplying their Brinell hardness figure by 3.25 for hard steels and by3.56 for those in the soft or annealed condition.


46/362

32 Welding scienceCreepThis is the term ap plied to the grad ual ch ange in dimen sions w hich

occurs when a load (tensile, compressive, bending, etc.) is applied to aspecimen for a long period of time. Creep generally refers to the extensionwhich occurs in a specimen to which a steady tensile load is applied over aperiod of weeks and months. In these tests it is generally found that thespecimen shows greater extension for a given load over a long period thanfor a short period a nd may fracture at a load mu ch less tha n its usu al tensileload. The effect of creep is greater at elevated temperatures and isimportant, as for example, in pipes carrying high-pressure steam at highsuperheat temperatures. In creep testing, the specimen is surrounded by aheating coil fitted with a pyrometer. The specimen is heated to a giventemperature, the load is applied and readings taken of the extension thatoccurs over a period of weeks, a gr ap h of the results being m ade . Th e test isrepeated for various loads and at various temperatures.

Special electrodes usually containing molybdenum are supplied forwelding 'creep-resisting' steels, that is, steels which have a high resistanceto elongation when stresses are applied for long periods of time at eitherordinary or elevated temperatures.

FatigueFatigue is the tendency which a metal has to fail under a rapidly

alter na ting lo ad, th at is a load which acts first in one direction, decre ases tozero and then rises to a maximum in the opposite direction, this cycle ofreversals being repeated a very great n um ber of times. If the stress is plottedagain st the num be r of stress reversals, the curve first falls steadily an d thenruns alm ost para llel to the stress reversal axis. T he stress at which th e curvebecomes horizontal is the fatigue limit (fig. 1.20). The load causing failureis generally m uch less than wou ld cau se failure if it was applied as a steadyload. Many factors, such as the frequency of the applied stress, tempera-

Fig. 1.20. Number of cycles of stress reversal-stress curve.

NUMBER OF CYCLES OF STRESS REVERSAL


47/362

Chemistry applied to welding 33ture, internal stresses, variation in section and sharp corners leading tostress concentration, affect the fatigue limit. Methods of fatigue testing aregiven in Chapter 5.

Chemistry applied to weldingElements, compounds and mixturesAll substances can be divided into two classes: (1) elements, (2)

compounds .An element is a simple substance which can no t be split up in to anything

simpler. F or example, aluminium (Al), coppe r (Cu), iro n( Fe ), tin(Sn ), zinc(Zn), sulphur (S), silicon (Si), hydrogen (H), oxygen (O) are all elements.A table of the elements is given in the appendix, together with their

chemical symbols.A compound is formed by the chemical combination of two or more

elements, and the property of the compound differs in all respects from theelements of which it is composed.

We have already mentioned the occurrence of matter in the form ofmolecules, and now it will be well to consider how these molecules arearranged among themselves and how they are made up.

If a m ixtu re of iron filings and sand is m ade , we can see the grain s of sandamong the filings with the naked eye. This mixture can easily be separatedby mean s of a magn et, wh ich will att rac t the iron filings an d leave the sand.Similarly, a mixture of sand and salt can be separated by using the factthat salt will dissolve in water, leaving the sand. In the case of mixtures,we can always separate the components by such simple means as this(called mechanical means).

Similarly, a mixture of iron filings and powdered sulphur can beseparated, either by using a magnet or by dissolving the sulphur in a liquidsuch as carbon disulphide, in which it dissolves readily.

Now suppose we heat this mixture. We find that it first becomes blackand then, even after removing the flame, it glows like a coal fire and muchheat is given off. After cooling, we find that the magnet will no longerattract the black substance which is left, neither will the liquid carbondisulp hide dissolve it. Th e black su bstan ce is, therefore, totally different incharac ter from the iron filings or the sulphur. I t can be shown by chemicalmeans that the iron and sulphur are still there, contained in the blacksubstance. This substance is termed a chemical com pou nd and is called ironsulphide. It has properties quite different from those of iron and sulphur.

Previously it has been stated that molecules can be sub-divided.


48/362

34 Welding scienceMolecules are themselves composed of atoms, and the number of atomscontained in each molecule depends upon the substance.

For example, a molecule of the black iron sulphide has been formed bythe combination of one atom of iron and one atom of sulphur joinedtogether in a chemical bond. This may be written:

Th e molecules of some elements contain m ore tha n o ne atom . A moleculeof hydrogen contains two atoms, so this is written:

O. Similarly, a molecule of oxygen contains two atoms,O. O

l l 2 " v Hthus: O, =

HO

A molecule of copper contain s only one atom , thu s: Cu = _CiiThe atmosphereLet us now study the com position of the atm osph ere, since it is of

primary importance in welding.Suppose we float a lighted candle, fastened on a cork, in a bowl of water

and then invert a glass ja r o ver the ca ndle, as sh own in Fig. 1.21. We findthat ther water will gradually rise in the jar, until eventually the candle goesout. By measurem ent, we find that the w ater has risen up the ja r j of theway, that is ,} of the air has been used up by the b urnin g of the candle, whilethe remaining f of the air still in the jar will not enable the candle tocon tinue bu rning. T he gas remaining in the jar is nitrogen (Fig. 1.22). It hasno smell, no taste, will not burn and does not support burning. The gaswhich has been used up by the burning candle is oxygen.

Fig. 1.21


49/362

Chem istry applied to welding 35Evidently, then, air consists of four parts by volume of nitrogen to one

pa rt of oxygen. Th at oxygen is necessary for b urnin g is very evident. Sandthrown on to a fire excludes the air, and thus the oxygen, and the fire isextinguished. If a per son 's clothes catch fire, rolling him in a blanke t or m atwill exclude the oxygen and put out the flames. In addition to oxygen andnitrogen the atmosp here con tains a small percen tage of carb on dioxide andalso small percentages of the inert gases first discovered and isolated byRayleigh and Ramsay. These gases are argon, neon, krypton and xenon.An inert gas is colourless, odourless, and tasteless, it is not combustibleneither does it support combustion and it does not enter into chemicalcombination with other elements. Argon, which is present in greaterpro po rtio n than the other inert gases, is used as the gaseous shield in ' inertgas welding' because it forms a protective shield around the arc andpreven ts the mo lten metal from com bining with the oxygen and nitrogen ofthe atm osph ere. Helium, w hich is the lightest of the inert gases, occurs onlyabout 1 part in 200000 in the atmosphere but occurs in association withother natural gases in large quantities especially in the United States,where it is often used instead of argon. The various gases of theatmosphere are extracted by fractionation of liquid air.

The percentage composition by volume, of dry air of the Earth'satmosphere is: nitrogen (N 2) 78.1, oxygen, (O 2) 20.9, argon (Ar) 0.93,carbon dioxide (CO 2) 0.03, neon (Ne) 0.0018, krypton (Kr) 0.00014,xenon (Xe) 0.0000086, helium (He) 0.00052, hydrogen (H 2) 0.00005,methane (CH 4) 0.00015, oxides of nitrogen (nitrous oxide N 2 O, etc.)traces of the order 0.00005, ozone (O 3) variable traces of the order0.00004, ozone layer in polar regions of the orde r x 10"5 to x 10"6.Note 0 " 1 ^ - 10~ 2 io-5 = (p.p.m.)

Fig. 1.22

NITROGEN LEFTTOTAL VOLUMEOF THE JAR


50/362

36 Welding scienceNitrogenNitrogen is a colourless, odourless, tasteless gas, boiling point

195.8 C, which does not b urn or sup por t co mb ustion. It is diatom icwith an atomic weight of 14, and dissociates in the heat of the arc to formiron nitride, which reduces the ductility of a steel weld. For this reason it isnot used as a shielding gas to any extent. It also forms nitrogen dioxideN O 2 and nitric oxide NO, which are toxic. It is widely dispersed incom pou nd form in nitrates, am mo nia and amm onium salts. It is producedby the liquefaction of air, and the considerable volumes produced by unitssuch as those supplying tonnage oxygen to steel plants can be used as thetop pressu re gas in blast furnaces and for the displacement of air in tank s,pipelines, etc.Nitrogen is supplied in compressed form in steel cylinders of 1.2, 3.1,4.6, 6.2 and 7.77 m 3 capacity at pressures of 137 and 175 bar at 15 C, andin liquid form by bulk tankers to an evaporator which in turn feeds gasinto a pipeline. (See liquid oxygen.)

ArgonThis mon atomic gas, chemical symbo l Ar and atom ic weight 18, is

presen t in the atm osph ere to the extent of abou t 1% and is obtaine d byfractional distillation from liquid air. It has no taste, no smell, is non-toxic,colourless and neither burns nor supports combustion. It does not formchemical com po un ds and has special electrical prop erties. It is extensivelyused in welding, either on its own or mixed with carbon dioxide orhydrogen, in the welding of aluminium, magnesium, titanium, copper,stainless steel and nickel by the TIG and MIG processes and in plasmawelding of stainless steel, nickel and titanium, etc. Argon is used for theinert gas filling of electric lamps and valves, with nitrogen, and in metalrefining and heat treatment, for inert atmospheres. It is supplied incompressed form in steel cylinders of 1.72, 2.00, 8.48 and 9.66 m 3 capacityat pressu res of 175 an d 200 ba r, an d in liquid form delivered by ro adtankers which pump it directly into vacuum-insulated storage vessels asfor liquid oxygen (q.v.).

HeliumHelium is an inert gas only present in the atmo sph ere to an extent

of 0.000052%. It is obtained from underground sources in the USA and isvery mu ch mo re expensive in this cou ntry th an argon . It is m on ato m ic withan atom ic weight 4, is lighter tha n a rgo n a nd is the lightest of the rare gases.Like the other inert gases it is colourless, odourless and tasteless, doesno t burn or supp ort com bustion , is non -toxic and does not formchemical com pou nd s. B ecause of its lightness, a flow rate of 2 to 2\ times


51/362

Chemistry applied to welding 37that of argon is required to provide an efficient gas shield in inert gaswelding processes. Mixed with argon it gives a range of proprietary gasesfor TIG and MIG welding processes contained in steel cylinders of 8.5 to9 m 3 capacity at a pressure of 200 ba r a t 15 C.

Carbon dioxide CO 2Ca rbo n d ioxide is no w extensively used as a shielding gas in the gasshielded metal arc w elding process. It is a non-flam mable gas of m olecularweight of 44.01, with a slightly pu nge nt smell and is ab ou t 1 j times as heavyas air (specific gravity relative to air is 1.53). It is soluble in water, givingcarbonic acid H 2 C O 3 , and it can be readily liquefied, the liquid beingcolourless; the critical temp eratu re (tha t is the tem per ature abov e which itis impossible to liquefy a gas by increasing the pressure) is 31.02C.Because its hea t of formation is high it is a stable com po und , en abling it tobe used as a protective shield around the arc to protect the molten metalfrom con tam inatio n by the atmo sphe re, and it can be mixed with argon forthe same purpose. During the CO 2 shielded metal arc process some of themolecules will be broken down or dissociated to form small quantities ofcarbon monoxide and oxygen. The carbon monoxide recombines withoxygen from the atmosphere to form CO 2 again and only very smallquantities (the generally accepted threshold is 50 p.p.m.) escape into theatmosphere and the oxygen is removed by powerful deoxidizers in thewelding wire. The ga s is very much cheaper than argon; i t is no t an inert gas.

Ca rbo n diox ide is formed when limestone is heated strongly in the limekiln and also by the action of hydrochloric acid on limestone. It may beobtained as a by-prod uct in the production of nitrogen and hydrogen in thesynthesis of am mo nia an d also as a by-p rod uct in the fermentation processwhen yeast acts on sugar or starch to produce alcohol and carbon dioxide.

Large supplies for indus trial use may be obtain ed by burnin g oil, coke orcoal in a boiler. Th e steam gen erated can be used for driving prim e m oversfor electricity gen eratio n and the flue gases, consisting of C O 2 , nitrogen andother im purities, are passed in to a washer where the impurities are removedand then into an absorb er where the CO 2 is absorbed and the nitrogen thusseparated. The absorber co ntaining the CO 2 passes into a stripping co lumnwhere the CO 2 is remove d and w ater vap ou r set free, and is removed by acondenser. The CO 2 is then stored in a gas-holder from which it passesthrough a further purifying process, is then compressed in a compressor,passed through a drier and a condenser and stored in the liquid state at apressure of 2 N / m m 2 at a temperature of 18 C, the storage tank beingwell insulated. The liquid CO 2 is then pumped into the cylinders used forwelding purpose s or into bulk supply tan ks, or it may be further convertedinto the solid state (Cardice) which has a surface te m per atur e of 78.4 C.


52/362

38 Welding scienceThe use of CO 2 as a shielding gas in the M I G - M A G processes is fullydiscussed in the chapter on these processes and, in addition to this, the

following are the main uses of the gas at the present time: in nuclearpower stations, where it can be used for transference of heat from thereactor to the electricity generating unit; for the CO 2-silicate process inthe found ry for core and mou ld m ak ing ; for the soft drink trad e where thegas is dissolved under pressure in the water of the mineral water or beerand gives a sparkle to the drink when the pressure is released; and in thesolid state for refrigerated transport, the perishable foodstuffs beingpacked in heavily insulated containers with the solid CO 2 , whichevaporates to the gaseous state and leaves no residue.

OxygenIn view of the im po rtan ce of oxygen to th e welder^ it will be useful

to prepare some oxygen and investigate some of its properties.Place a small quantity of potassium chlorate in a hard glass tube (test

tube) an d hea t by means of a gas flame. The su bstance melts, acc omp aniedby crackling noises. Now place a glowing splinter in the mou th of the tub e.T

Date post:	30-Oct-2015
Category:	Documents
Upload:	orlandompsilva
View:	95 times
Download:	2 times

The Science and Practice of Welding

Documents