WELDS STATIC AND FATIGUE STRENGTH

1.0

INTRODUCTION

The need for connections and the importance of connection designs in steel structures have been covered in an earlier chapter. It has been pointed out that steel sections are linear elements produced in certain convenient lengths due to constraints on manufacturing and transportation. Therefore connections are necessary to provide continuity, where required, as well as to create three-dimensional steel structures. One of the most efficient and possibly direct ways of providing connections in steel structures is by way of welding. Welding is the process of joining two pieces of metal by creating a strong metallurgical bond between them by heating or pressure or both. It is distinguished from other forms of mechanical connections, such as riveting or bolting, which are formed by friction or mechanical interloc ing. It is one of the oldest and reliable methods of jointing. Welding was quite an art prevalent in ancient !reece to ma e bracelets. It was probably a forging process, where metals were heated and hammered together. "odern welding has been in e#istence since World War I. It was mainly used for repairing damaged ships. $fter %&%&, the use of welding as a construction and fabrication method began to develop. 'ince then many improvements and developments have ta en place. Today there are over () different welding processes, which can be used to join various metals and their alloys. 2.0 ADVANTAGES OF WELDING

Welding offers many advantages over bolting and riveting. 'ome of the advantages are listed in the following. Welding enables direct transfer of stress between members. *ence, the weight of the joint is minimum. +esides efficiency, design details are very simple. ,ess fabrication cost compared to other methods due to handling of fewer parts and elimination of operations li e drilling, punching etc. The most stri ing advantage of welded structures is in the area of economy. Welded structures allow the elimination of a large percentage of the gusset and splice plates necessary for riveted or bolted structures. Time is saved in detailing, fabrication and field erection. In some bridge trusses it may be possible to save up to %(- of the steel weight by resorting to welding. Welding also requires considerably less labour for e#ecuting the wor . Welding offers air tight and water tight joining of plates and hence ideal for oil storage tan s, ships etc.

Welded structures usually have a neat appearance as against the cluttered surface of bolted or riveted connections. .ig. % shows a comparison of riveted plate girder and a welded plate girder. .urther, welded connections offer the designer more freedom for innovation in his design concept. It enables him to use any cross section and the best configuration to transmit forces from one member to another.

(a) Bolted girder section

(a) Welded girder section Fig.1 Comparison of appearance of riveted and welded plate girders The range of application of welding is very wide. .or e#ample, connection of a steel pipe column to other members can be made very easily by welding whereas it is virtually impossible by bolting or riveting. Welding is practicable even for complicated shapes of joints. There is no need for holes in members connected by welding e#cept possibly for erection purposes. This has direct influence in the case of tension members as the problem of determining the minimum net section is eliminated. This also results in a member with a smaller cross section.

Welded structures are more rigid compared to structures with riveted and bolted connections. The rigidity of welded structures is due to the direct connection of members by welding. In bolted or riveted structures, the connection is established through angles or plates, which deflect under loads, ma ing the structure fle#ible. It is easier to ma e design changes and to correct mista es during erection, if welding is used. It is also a simple procedure to strengthen the e#isting structures with welding. $ truly continuous structure is formed by the process of fusing the members together. This gives the appearance of a one-piece construction. !enerally welded joints are as

strong or stronger than the base metal, thereby placing no restriction on the joints. /ue to this continuity advantage, a very large number of steel frames have been constructed all over the world. 'tress concentration effect is considerably less in a welded connection. 'ome of the lesser important advantages of the welding processes are0 relative silence of the process of welding and fewer safety precautions.

'ome of the disadvantages of welding are0 Welding process requires highly s illed manpower 1#perienced manpower is needed for inspection of welded connections. $lso, nondestructive evaluation may have to be carried out to detect defects in welds Welded joints are highly prone to crac ing under fatigue loading 2ostly equipment is essential to ma e welded connections 3roper welding can not be done in the field environment ,arge residual stresses and distortion are developed in welded connections

In the earlier days, combination of bolting, riveting and welding was not practiced. 'tructures were completely welded, bolted or riveted. 3resently both are used in a structure e#cept that both connection techniques are not used in one and the same joint. The present trend is to use welding for wor shop connections or splices, and high strength bolts for field joints. 3.0 FUNDAMENTALS OF WELDING $ welded joint is obtained when two clean surfaces are brought into contact with each other and either pressure or heat, or both are applied to obtain a bond. The tendency of atoms to bond is the fundamental basis of welding. The inter-diffusion between the materials that are joined is the underlying principle in all welding processes. The diffusion may ta e place in the liquid, solid or mi#ed state. In welding the metallic materials are joined by the formation of metallic bonds. When two clean and flat surfaces are joined, the bonding ta es place between surface atoms and a perfect connection is formed 4.ig.56. Metal A

(a) Before weld

Metal B

(b) After weld

Fig.2 Two ideal surfaces brought together to form a weld

The efficiency obtained for the joint is %))- and its strength would be as much as that of the base metals. In practice however, it is very difficult to achieve a perfect joint7 for, real surfaces are never smooth. When irregular surfaces are joined by welding, contact is established only at a few points in the surface, where atomic bonding occurs. Therefore the strength attained will be only a fraction of the full strength. $lso, the irregular surface may not be very clean, being contaminated with adsorbed moisture, o#ide film, grease layer etc. as shown in .ig 84a6. In the welding of such surfaces, the contaminants have to be removed for the bonding of the surface atoms to ta e place. This can be accomplished by applying either heat or pressure. In practical welding, both heat and pressure are applied to get a good joint. Oxide layer Metal B Adsorbed layer Metal A Fig.3 (a) Surface contaminants Metal B Metal B

Metal A

Fig. 3(b) !ddition of filler material

Metal A Fig.3(c) ear perfect weld When heat is applied, the adsorbed layers are driven off, o#ide films are bro en and the yield strength of the base metals are lowered7 with the application of pressure, plastic deformation ta es place and brings into contact more atoms for welding. *eat application results in the melting of the metallic members, enabling atoms to come into contact by fluid flow and form bonds. 'ometimes, a filler material, of the same type as the base material or compatible with it, is added to achieve the bond 9.ig 84b66:. When pressure is applied to form a joint it brea s the obstructing layers and sharp edges to bring the joining surfaces together for ma ing the bond. .ig. 84c6 shows a weld formed by the application of heat and pressure. $s pointed out earlier, any welding process needs some form of energy, often heat, to connect the two materials. The relative amount of heat and pressure required to join two materials may vary considerably between two e#treme cases in which either heat or pressure alone is applied. When heat alone is applied to ma e the joint, pressure is used merely to eep the joining members together. 1#amples of such a process are !as Tungsten $rc Welding 4!T$W6, 'hielded "etal $rc Welding 4'"$W6, 'ubmerged $rc Welding 4'$W6 etc., which are e#plained later. On the other hand pressure alone is used to ma e the bonding by plastic deformation, e#amples being cold welding, roll welding,

ultrasonic welding etc. There are other welding methods where both pressure and heat are employed, such as resistance welding, friction welding etc. The required heat is produced by a flame, an arc or resistance to an electric current. 1lectric arc is by far the most popular source of heat used in commercial welding practices. 4.0 BASIC WELDING PROCESSES

In general, gas and arc welding are employed7 but, almost all structural welding is arc welding. In gas welding a mi#ture of o#ygen and some suitable gas is burned at the tip of a torch held in the welder;s hand or by an automatic machine. $cetylene is the gas used in structural welding and the process is called o#yacetylene welding. The flame produced can be used both for cutting and welding of metals. !as welding is a simple and ine#pensive process. +ut, the process is slow compared to other means of welding. It is generally used for repair and maintenance wor . The most common welding processes, especially for structural steel, use electric energy as the heat source produced by the electric arc. In this process, the base metal and the welding rod are heated to the fusion temperature by an electric arc. The arc is a continuous spar formed when a large current at a low voltage is discharged between the electrode and the base metal through a thermally ionised gaseous column, called plasma. The resistance of the air or gas between the electrode and the objects being welded changes the electric energy into heat. $ temperature of 88)) ) 2 to (())) 2 is produced in the arc. The welding rod is connected to one terminal of the current source and the object to be welded to the other. In arc welding, fusion ta es place by the flow of material from the welding rod across the arc without pressure being applied. 4.1 Arc We !"#$ Pr%ce&&e& /ifferent processes of arc welding are e#plained in the following sections0 ".1.1 Shielded #etal !rc $elding (S#!$)

In 'hielded "etal $rc Welding or '"$W 4.ig. <6, heating is done by means of electric arc between a coated electrode and the material being joined. In case bare wire electrode 4without coating6 is employed, the molten metal gets e#posed to atmosphere and combines chemically with o#ygen and nitrogen forming defective welds. The electrode coating on the welding rod forms a gaseous shield that helps to e#clude o#ygen and stabilise the arc. The coated electrode also deposits a slag in the molten metal, which because of its lesser density compared to the base metal, floats on the surface of the molten metal pool, shields it from atmosphere, and slows cooling. $fter cooling, the slag can be easily removed by hammering and wire brushing.

!lectrode Arc stream "lag Weld #oating on electrode Gaseous shield Molten pool Penetration depth Base metal Fig.". Shielded The coating on the electrode thus #etal !rc $elding (S#!$) process 'hields the arc from atmosphere 2oats the molten metal pool against o#idation 'tabilises the arc 'hapes the molten metal by surface tension 3rovides alloying element to weld metal

The type of welding electrode used would decide the weld properties such as strength, ductility and corrosion resistance. The type to be used for a particular job depends upon the type of metal being welded, the amount of material to be added and the position of the wor . The two general classes of electrodes are lightly coated and heavily coated. The heavily coated electrodes are normally used in structural welding. The resulting welds are stronger, more corrosion resistant and more ductile compared to welds produced by lightly coated electrodes. =sually the '"$W process is either automatic or semiautomatic. ".1.2 Submerged !rc $elding (S!$) In this arc welding process, the arc is not visible because it is covered by a blan et of fusible powdered flu#. The bare metal electrode is deposited as a joining material. The flu#, which is a special feature of the method, protects the weld pool against the atmosphere. The arc once started is at all times covered by the flu# as shown in .ig. (. lux tube !lectrode "lag Granular fill (flux) Penetration depth Weld Base metal Fig.%. Submerged arc welding (S!$) process Molten pool

The heat of the arc melts the electrode, the object to be welded, and part of the flu#. The slag formed by the flu#, which forms a coat over the solidified weld beam, may be removed by brushing. Welds made by submerged arc welding process have high quality, good ductility, high impact strength, high density and good corrosion - resistance. Their mechanical properties are as good as the base metal. 'ince more heat is input in this process, the penetration is deeper than the '"$W process. This is normally ta en into account in the design. ".1.3 #anual #etal !rc (##!) $elding

This is a manually operated welding process and hence requires s ill to produce good quality welds. The electrode is made up of a steel core wire 48.5 > ?.) mm diameter6 and the flu# contains manganese and silicon as alloying elements. The electric arc melts the metallic object to be welded and the electrode. $s the core wire metal melts and joins the weld pool, the electrode is moved to maintain the arc length. This is important as the arc length controls the width of the weld run. The flu# also melts with steel core wire and forms the surface slag, which is removed after solidification. ,ow capital cost and freedom of movement 4up to 5) m from power supply6 are the main advantages of ""$ welding. It is well suited to structural and stainless steels. Its main disadvantage is that only a small volume of metal is deposited per electrode. This is not a problem for short welds, but for long welds this becomes a serious consideration. ".1." #etal & !ctive 'as (#!') welding This is sometimes also referred to as "etal Inert !as 4"I!6 welding. The arc and the weld pool are protected by an inert gas7 the shielding gas often used is carbon dio#ide or a mi#ture of o#ygen and carbon dio#ide. .lu# is not necessary to shield the pool7 however, occasionally a flu# - cored electrode is used to produce slag. The arc length is maintained by the power supply unit. Though "$! welding is easier, more s ill is required to establish the correct welding conditions. =sing "$! welding, production is improved, as there is no need to deslag or change electrode. It is highly suitable for fillet welded joints, such as beam to beam or stiffener to panel connections. Its disadvantage is due to restriction in movement due to equipment. This can be manual, semi-automatic or automatic process ".1.% (lectroslag $elding )rocess The method of 1lectroslag process 4.ig. ?6 used for vertical automatic welding is based on the heat produced by electrical current through molten slag. The electrode is immersed in the molten slag pool between the components to be welded and the copper moulding devices. $s the melt is heated to a high temperature by current passing between the electrodes and the base metal, the electrical conductivity is increased. The slag pool temperature must e#ceed the melting points of the base and filler metals. Then the slag melts the faces of the connecting wor and the electrode is immersed in the molten slag.

The weld pool that forms when the molten base and filler metal collect at the bottom of the slag pool solidifies and forms the weld joining the faces of the members. 1lectroslag welding is useful for joining thic sections in a vertical position. 'ingle - pass welds can be made in any reasonable thic ness of steel. Welding usually starts at the bottom of the joint and progresses towards the top of the vertical connection.

Fig. * (lectroslag welding process ".1.* Stud $elding The shear connectors that are used e#tensively in composite construction have to be attached to the steel beam by welding. 'tud welding is an accurate and best method of attaching shear connectors with minimum distortion. The figure and details of a stud welding machine are given in 2hapter <% on .abrication and 1rection of 'tructural 'teel Wor . 'tud welding is another form of arc welding and is mechanised for the welding of studs to plane surfaces. The stud may be a plain bar with an upset head or threaded. The stud itself forms the electrode and is held in a chuc , which is connected to the power source. Initially, the stud is touched on the steel plate or section surface and the current is switched on. 'oon the stud is moved away to establish the arc. $ weld pool is formed and the stud end melts. Then the stud is forced into the steel plate automatically and the current supply is switched off. The molten metal is collected in an enclosure around the stud formed by a ceramic collar or ferrule. Thus the molten metal is formed into a fillet and is also protected from atmosphere by the ferrule. 4.2 C'%"ce %( Pr%ce&& The choice of a particular process is made based on a number of parameters listed below0 The location of the welding operation0 In a protected place li e a fabrication shop, '$W and "$! are best suited. .or field conditions ""$ is easier.

 ).0 ).1

$ccuracy of setting up0 '$W and "$! require good and accurate set-up. 3enetration of the weld. @olume of weld to be deposited $ccess to joint0 The welding plant and the welding torches have to be properly positioned during the welding operation. In easily accessible joints '$W or "$! is used, whereas in cramped locations ""$ is preferred. 3osition of welding0 '$W and "$! are not suitable for overhead positions. ""$ is the best for overhead wor s. 'teel composition0 '$W and "$! do not generally develop *$A 4*eat $ffected Aone6 crac ing. This offsets the disadvantage of "$! for site wor s. 2omparative cost0 cost of welding is calculated for unit length considering the duty cycles. WELDING PROCEDURE Ge#er*

The term Bwelding procedure; encompasses the complete operation of ma ing a weld. Thus, it includes choice of electrode, edge preparation, preheat, welding parameters such as voltage, current, welding position, number of weld run to fill the groove and post weld treatments 4e.g. grinding, heat treatment etc.6. 1stablishment of such procedures helps to minimise the cost, achieve good impact properties, eliminates defects and controls distortion. 'ome of the important elements of weld procedure are elaborated below. (nvironment+ Weld procedure must account for actual site conditions. In cold regions, it may be necessary to heat steel up to 5)-. The humid weather or condensation might help formation of porosity. 1lectrodes must be ept in dry condition. In moist C humid environments the electrodes may be ept in a warm container to avoid moisture entrapment in the flu# coating. $elding position+ @ertical welding is slower compared to welding in the flat position. Overhead welding causes weld splutter and require special s ills. It is better avoided. Current+ The current controls the heat input. $ minimum current is required for fusing the plate and to eep the arc stable. !enerally a high current is used to obtain quic er welding so as to reduce cost. It may not be possible to use ma#imum current always, a specific e#ample being welding in the overhead position. The current limit for overhead use is %?)$. =sually high current results in low impact properties. .urther very high value of current may cause crac s in the *eat-$ffected Aone 4*$A6. Shrin,age+ While cooling after the welding operation, the hot metal in the welded region contracts causing the joint to shrin . +ut this contraction is prevented by the adjacent colder metal. This causes stress, sometimes even beyond yield stress, and causes plastic deformation. This also might cause distortion of the member. +y following proper edge preparation and weld procedure, this can be minimised. $fter the plastic deformation a residual stress pattern is formed in the joint. Tensile stresses are formed in the weld metal and *$A Dones, whereas compression in the adjacent steel.

)re heating+ *ydrogen induced crac ing 4cold crac ing, delayed crac ing6 is a serious problem affecting weldability. The degree of crac ing occur due to the combined effects of four factors+ %. 5. 8. <. +rittle microstructure 3resence of hydrogen in weld metal Tensile stresses in the weld area Temperature range 4-%))o 2 to 5))o 26

3re heating of the weld area is the most effective and widely used method to prevent hydrogen induced crac ing. Welding involves a cycle of sudden heating and cooling. +y preheating the parent metal, the difference in temperature between the preheated temperature and the final temperature is reduced. This, in the cooling cycle, also helps to obtain a lower thermal gradient. $s e#plained in the first chapter on B*istorical /evelopment and 2haracters of 'tructural 'teels;, sudden cooling of steel results in a hard and less ductile material called martensite. The main function of preheating is to reduce the weld metal cooling rate so that transformation to martensite is reduced below a certain critical level. The slower cooling gives more time for hydrogen to diffuse out of the weld area and delays the development of ma#imum residual stresses. !as torches, heat-treating furnaces or electric-resistance heaters are used in preheating the weld area. ).2 We !*," "-. %( S-ee & The term weldability is defined as the ability to obtain economic welds, which are good, crac - free and would meet all the requirements. Of great importance are the chemistry and the structure of the base metal and the weld metal. The effects of heating and cooling associated with fusion welding are e#perienced by the weld metal and the *eat $ffected Aone 4*$A6 of the base metal. The *$A i.e. base metal surrounding the weld metal and the weld itself will have unduly varying hardness distribution across a weld. The hardness in steel depends upon the rate at which steel is cooled near the fusion Done7 the hardness is ma#imum due to the higher temperature at that location. .urther, these locations also have the ma#imum rate of cooling. *igher value of hardness leads to crac s in *$A or in the weld. 2rac s might be formed during or after the welding process. !ood design and standard welding procedure will minimise the crac ing problem. 'everal features that affect weld crac ing during the welding processes are Eoint restraint that builds up high stress in the weld +ead shape 4conve# or concave6 2arbon and alloy content of the base metal 2ooling rate *ydrogen and nitrogen absorption

The crac s in *$A are mainly caused by high carbon content, hydrogen enbrittlement and rate of cooling. .or most steels, weld crac s become a problem as the thic ness of the plates increases.

/.0 T0PES OF 1OINTS AND WELDS +y means of welding, it is possible to ma e continuous, load bearing joints between the members of a structure. $ variety of joints is used in structural steel wor and they can be classified into four basic configurations as shown in .ig. F. They are0 %. 5. 8. <. ,ap joint Tee joint +utt joint, and 2orner joint

(a) Butt $oint

(b) %ap $oint

(c) &ee $oint

(d) #orner $oint Fig. - T.pes of /oints .or lap joints the ends of two members are overlapped, and for butt joints the two members are placed end to end. The T- joints form a Tee and in 2orner joints, the ends are joined li e the letter ,. The common types of welds are shown in .ig. G. "ost common joints are made up of fillet weld and the groove weld. 3lug and slot welds are not generally used in structural steel wor . .illet welds are suitable for lap joints and Tee joints and groove welds for butt and corner joints. !roove welds can be of complete penetration or incomplete penetration depending upon whether the penetration is complete through the thic ness or partial. !enerally a description of welded joints requires an indication of the type of both the joint and the weld. Though fillet welds are wea er than groove welds, about G)- of the connections are made with fillet welds. The reason for the wider use of fillet welds is that in the case of fillet welds, when members are lapped over each other, large tolerances are allowed in erection. .or groove welds, the members to be connected have to fit perfectly when they are lined up for welding. .urther groove welding requires the shaping of the surfaces to be joined as shown in .ig. &. To ensure full penetration and a sound weld, a bac up plate is temporarily provided as shown in .ig. &.

Welds are also classified according to their position into flat, horiDontal, vertical and overhead 4.ig. %)6. .lat welds are the most economical to ma e while overhead welds are the most difficult and e#pensive.

(a) Groo(e welds A A

(b) illet welds

!nds shall be semi circular A

"ection A'A "ection A'A (c) "lot weld (d) Plug weld Fig.0 Common t.pes of welds

Bac,up plate Fig. 2 Shaping of surface and bac,up plate

(a) lat (b) )ori*ontal (c) +ertical

(d) O(erhead

Fig. 11 Classification based on position

2.0 GROOVE WELDS The main use of groove welds is to connect structural members, which are in the same plane. $ few of the many different groove welds are shown in .ig. %%. There are many variations of groove welds and each is classified according to its particular shape. 1ach type of groove weld requires a specific edge preparation and is named accordingly. The proper selection of a particular type depends upon 'iDe of the plate to be joined. Welding is by hand or automatic. Type of welding equipment. Whether both sides are accessible. 3osition of the weld. "ingle be(el groo(e weld

"-uare groo(e weld (a) #orner $oint

(b) Butt $oint

.ouble + groo(e weld

"-uare groo(e weld

(c) !dge /oint Fig.11 T.pical connections with groove weld The aim is to achieve the most economical weld of the requisite efficiency and strength. The butt weld whether of full penetration or partial penetration should attain the required strength of the joined parts. The siDe of the butt weld is defined by the thic ness i.e. the thic ness of the connected plate for complete penetration welds or the total depth of penetration for partial penetration welds. !roove welds have high strength, high resistance to impact and cyclic stress. They are most direct joints and introduce least eccentricity in the joint. +ut their major disadvantages are0 high residual stresses, necessity of edge preparation and proper aligning of the members in the field. Therefore, field butt joints are rarely used.

To minimise weld distortions and residual stresses, the heat input is minimised and hence the welding volume is minimised. This reduction in the volume of weld also reduces cost. *ence for thic er plates, double groove welds and = welds are generally used. 2.1 E!$e Pre3*r*-"%# (%r B4-- We ! Typical edge preparations are shown in .ig.%5 0

0oot opening 0 (a) Be(el with feathered edge (b) Be(el with bac,up plate

%and (root face) (d) .ouble be(el with a spacer

(c) Be(el with a land

Fig.12 T.pical edge preparation for butt weld .or a butt weld, the root opening, 0, is the separation of the pieces being joined and is provided for the electrode to access the base of a joint. The smaller the root opening the greater the angle of the bevel. The depth by which the arc melts into the plate is called the depth of penetration 9.ig.%8 4a6:. Houghly, the penetration is about % mm per %))$ and in manual welding the current is usually %() > 5)) $. Therefore, the mating edges of the !lectrode
Arc

1ncluded angle 0oot face

0oot gap

.epth of penetration

(a) .epth of penetration #apping run

illing run

(b) 0oot gap

0oot run (c) 0oot run Fig.13 'roove weld details

plates must be cut bac if through-thic ness continuity is to be established. This groove is filled with the molten metal from the electrode. The first run that is deposited in the bottom of a groove is termed as the root run 9.ig.%84 c6:. .or good penetration, the root faces must be melted. 'imultaneously, the weld pool also must be controlled, preferably, by using a bac ing strip. The choice of edge preparation depends on %. 5. 8. <. (. ?. Type of process 3osition of welding $ccess for arc and electrode @olume of deposited weld metal 2ost of preparing edges 'hrin age and distortion.

The square groove joint is used to connect thin material up to about G mm thic 7 for thic er material, single-vee groove and the double-vee groove welds have to be used. 5.0 FILLET WELDS $ typical fillet weld is shown in .ig.%< 4a6. Weld and leg si*e ace of weld
s

&heoretical throat (t234535s)

0oot of weld Fig. 1" (a) T.pical fillet weld Owing to their economy, ease of fabrication and adaptability, fillet welds are widely used. They require less precision in the fitting up because the plates being joined can be moved about more than the groove welds. $nother advantage of fillet welds is that special preparation of edges, as required by groove welds, is not required. In a fillet weld the stress condition in the weld is quite different from that of the connected parts.

Fig.1" (b) T.pical fillet weld connections

The siDe of a fillet weld is defined by the length of the two sides of the largest right triangle, which can be inscribed within the weld cross section. $ major share of welds of this type has equal legs i.e. they form right isosceles triangle shown in .ig. %<4 c6. The typical fillet weld connections are shown in .ig. %<4 b6. The critical dimension of a fillet weld is its throat, the shortest distance from the root to the hypotenuse of the defining triangle shown in .ig. %<4c6. "i*e 2 s8 t &hroat
s t s s

"i*e 2 s &hroat 234535s

0oot (6) 7ne-ual legs

0oot (16) !-ual legs

Fig.1" (c) 3ne4ual and (4ual legs fillet welds The root of the weld is the point where the faces of the metallic members meet. The theoretical throat of a weld is the shortest distance from the root to the hypotenuse of the triangle. The throat area equals the theoretical throat distance times the length of the weld. Though a fillet weld is specified by defining the two sides of the inscribed triangle, its actual cross section will be quite comple#. $ fillet weld must penetrate the base metal and the interface of the weld is either concave or conve# 9.ig .%(4a6I4b6:.

(a) #onca(e

(b) #on(ex Fig.1% Cross section of fillet weld

(c) 7nder cutting

The concave shape of free surface provides a smoother transition between the connected parts and hence causes less stress concentration than a conve# surface. +ut it is more vulnerable to shrin age and crac ing than the conve# surface and has a much reduced throat area to transfer stresses. On the other hand, conve# shapes provide e#tra weld metal or reinforcement for the throat. +ut while ma ing a conve# surface there is always the possibility of causing undercut at the edges, which undermines the strength of the joint 9.ig. %(4c6:. The stress concentration is higher in conve# welds than in concave

welds. It is generally recommended that for statically loaded structures, a slightly conve# shape is preferable, while for fatigue > prone structures, concave surface is desirable. ,arge welds are invariably made up of a number of layers or passes. .or reasons of economy, it is desirable to choose weld siDes that can be made in a single pass. ,arge welds scan be made in a single pass by an automatic machine, though manually, G mm fillet is the largest single-pass layer. 6.0 FACTORS AFFECTING THE 7UALIT0 OF WELDED CONNECTIONS

$ good weld is obtained from a combination of many factors, from the design of the weld to the welding operation. 1ven a well-designed weld may not give a strong connection if it is not properly made. Therefore, a structural engineer must be aware of the various factors that affect the quality of the weld. 'ome of those factors are e#plained in the following. 6.1 Pr%3er E ec-r%!e&8 We !"#$ A33*r*-4& *#! Pr%ce!4re&

/epending on the grade of steel and its thic ness, appropriate electrode with suitable diameter has to be selected. The siDe of the electrode is chosen based on the siDe of the weld to be made and also on the output of the welding apparatus. It is important that the welding apparatus is capable of delivering enough current for the siDe of the electrode. 'ince the output of welding apparatus can be controlled within limits, an electrode of small siDe may also be used. In metal>arc welding, the metal is deposited by electromagnetic shield and not by gravity. Therefore the welder is not limited to horiDontal or flat welding positions. It is better to avoid overhead welding as the controlling of the process is very difficult and requires a highly s illed welder. In the field, it may not be possible to avoid overhead welding fully7 so adequate care must be ta en in specifying and ma ing such a weld. 6.2 We !"#$ Se94e#ce 'equence of welding plays a ey role in obtaining a satisfactory welded fabrication. The smallest weld siDe that can fulfil the requirements is the hallmar of a good weld designer. It is always advisable to weld away from a point of restraint7 welding of a joint should start from the centreline and proceed towards the free end. The principle of doubling>up method is employed for a single run fillet weld on either side of the vertical member 4.ig. %?6.
69 6: 5 ; 6 < = 65 6=

66

63

>

6?

6;

6<

Fig.1* Se4uence of welding fillet welds on either side of a vertical member (doubling5up method)

This eliminates transverse angular distortion of a Tee joint. The planned wandering method of welding, shown in .ig.%F is used for butt welded joints with two operators.
:

Fig.1- )lanned $andering #ethod for $elding 6ong 7utt 8oint with Two 9perators $nother alternative is the stepbac method 4.ig. %G6 in which the welding is carried out in an evenly balanced manner about the centerline of a joint. Operator (:)
6 :

Operator (6)
9 ? ;

Fig.10 Step 7ac, #ethod for $elding 6ong 8oint $ith Two 9perators 10.0 RESIDUAL STRESSES AND WELD DISTORTION 10.1 Re&"!4* S-re&&e& In any fusion welding process, the first step is to heat the surfaces to be joined. This heating process is fast and restricted to a narrow portion, where the joint is to be made. The region, which is heated, e#pands, whereas the rest of the metal resists the e#pansion. $s a consequence compressive stresses are induced near the joint during the heating. $fter the welding operation is over, the region of intense heating starts cooling. The contraction, which should result, is prevented because the two separate pieces, which were originally free to e#pand, have now been joined and are not free to contract. .urther, the temperature distribution of the weldment would be vastly different from that of the two separate pieces. The net result is that tensile stresses are developed near the weld and compressive stresses away from it. This is shown in .ig. %&. Thus in any welding process involving rapid heating and cooling, residual stresses would always be present. The amount of the residual stresses would vary depending up on the joint restraint, geometry, thermal properties of the material, welding process employed etc. If the parts to be joined are free to e#pand and contract then the residual stresses would be minimum. *owever the presence of e#ternal restraints would prevent free e#pansion and contraction resulting in large levels of stresses. In general, residual stresses are increased when increased number of passes is made. The thermal e#pansion coefficient and thermal conductivity influence the residual stress level. 3oor thermal conductivity and high

thermal e#pansion coefficient would significantly increase the residual stresses in a weldment. Jield strength of material is the upper limit for the residual stress. If the stresses e#ceed the yield strength, which is normal in welding, the material yields leading to permanent deformation or change of shape. This permanent deformation caused by stresses resulting from thermal cycles is called distortion. This is e#plained in the ne#t section. &ension #ompressio n

Fig.12 6ongitudinal residual stress due to weld 10.2 D"&-%r-"%# "# We !:e#-& In a welding operation, due to heating and cooling of the weld metal and base metal regions near the weld, thermal strains are induced in them. The strains thus produced may be accompanied by plasticity. The stresses due to these strains combine and react to produce internal forces causing bending, buc ling and rotation. The displacement arising out of these forces are called distortion. Three basic dimensional changes that occur during the welding process cause distortion in fabricated structures0 %. Transverse shrin age perpendicular to the weld line. 5. ,ongitudinal shrin age parallel to the weld line. 8. $ngular distortion 4rotation around the weld line6. These dimensional changes are classified according to their appearance into the following 4a6 Transverse shrin age0 the shrin age is perpendicular to the weld line 4b6 $ngular change0 Kon-uniform thermal distortion in the thic er direction causes the angular change close to the weld line, otherwise nown as transverse distortion. 4c6 Hotational distortion0 angular distortion in the plane of the plate due to thermal e#pansion. 4d6 ,ongitudinal shrin age0 shrin age in the direction of the weld line.

4e6 ,ongitudinal bending distortion0 distortion in plane through the weld line and perpendicular to the plate. 4f6 +uc ling distortion0 thermal compressive stresses cause instability in thin plates. 'ome of the above are illustrated in .ig. 5).

(b) Angular change &rans(erse shrin,age

(c) 0otational distortion

(d) %ongitudinal shrin,age

(e) %ongitudinal bending distortion Fig. 21 $eld distortion

(f) Buc,ling distortion

The three main approaches to overcome the problems of distortion are0 4%6 "inimising the distortion using carefully controlled welding procedures. 456 /eveloping standards for acceptable distortion limits. 486 =sing appropriate techniques to remove distortion. $ judicious combination of the above approaches would result in ma ing distortion free welds0 The first approach is to use a welding process that would produce no shrin age or distortion. *owever there is no process that would completely eliminate distortion. 2lose attention should be paid to factors such as welding sequence, degree of restraint, welding condition, joint details, the preheat etc. that contribute to weld distortion. $s distortion is unavoidable, it would be prudent to set up rational standards for acceptable distortion, considering the reliability, economic value and fabrication cost of the structure. If distortion e#ceeds the acceptable limits, then the distortion has to be removed with minimum damage to the structure. /istortion may occur even during the service due to overloading or impact. 'ome of the techniques used for removal of distortion are0 4a6 flame heating at selected spots and cooling with water, and 4b6 hammering while it is being heated or by applying a local force to cause counter distortions.

11.0 WELD S0MBOLS Welding will become a powerful engineering tool when the information required for welding is provided by the designers to the operators. The information concerning type, siDe, position, welding process etc. of the welds in welded joints is conveyed by standard symbols in drawings. =sage of standard symbols by all designers and fabricators would help avoid confusion and misunderstanding. The symbolic representation gives clearly all necessary indications regarding the specific weld to be achieved. The symbolic representation includes elementary symbols along with a6 supplementary symbol, b6 a means of showing dimensions, or c6 some complementary indications. I'0 G%8-%&G?, L'cheme Of 'ymbols for WeldingM gives all the details of weld representation in drawings. 1lementary symbols represent the various categories of the weld and loo similar to the shape of the weld to be made. 2ombination of elementary symbols may also be used, when required. 1lementary symbols are shown in Table %. 'upplementary symbols characterise the e#ternal surface of the weld and they complete the elementary symbols. 'upplementary symbols are shown in Table 5. 2ombinations of elementary and supplementary symbols are given in Table 8. The weld locations are defined by specifying, a6 position of the arrow line, b6 position of the reference line, and c6 the position of the symbol. "ore details of weld representation may be obtained from I' G%8-%&G?. Table 1. (lementar. S.mbols I 4&-r*-"%#;F"$.< S.:,% De&cr"3-"%# +utt weld between plates with raised edgesN4the raised edges being melted down completely6

'quare butt weld

'ingle-@ butt weld

'ingle-bevel butt weld

'ingle > @ butt weld with broad root face

'ingle > bevel butt weld with broad root face

'ingle > = butt weld 4parallel or sloping sides6

'ingle > E butt joint

+ac ing run7 bac or bac ing weld

.illet weld

3lug weld7 plug or slot weld

'pot weld

'eam weld

Table 2. Supplementar. S.mbols .lat 4flush6 single > @ butt weld

2onve# double > @ butt weld

2oncave fillet weld .lat 4flush6 single > @ butt with flat 4flush6 bac ing run

Table 3 Combination of (lementar. and Supplementar. S.mbols S'*3e O( We ! S4r(*ce 4a6 flat 4usually finished flush6 4b6 conve# 4c6 concave 11.1 P%&"-"%# %( &.:,% & "# !r*="#$& $part from the symbols as covered earlier, the method of representation 4.ig. 5%6 also include the following0 &.:,%

:a 6

/oint

:b

62arrow line :a2reference line(continuous line) :b2identification line(dashed line) 92welding symbol

Fig. 21 #ethod of :epresentation @Other sideA @Arrow sideA @Arrow sideA @Other sideA

!rrow line

Arrow line

(a)Weld on the arrow side Fig. 22 T58oint with one fillet weld

(b)Weld on the other side

$n arrow line for each joint $ dual reference line, consisting of two parallel lines, one continuous and the other dashed. $ certain number of dimensions and conventional signs The location of welds is classified on the drawings by specifying0 3osition of the arrow line 3osition of the reference line and The position of the symbol

The relation between arrow line and the joint, shown in .ig. 55 4a, b6 e#plain the meaning of the terms Barrow side; of the joint and B other side; of the joint. The position of arrow line with respect to the weld has no special significance. The arrow line joins one end of the continuous reference line such that it forms an angle with it and shall be completed by an arrowhead or a dot. The reference line is a straight line drawn parallel to the bottom edge of the drawing.

The symbol is placed either above or beneath the reference line. The symbol is placed on the continuous side of the reference line if the weld is on the other side of the joint7 the symbol is placed on the dashed line side 12.0 DEFECTS IN WELDS

If good welding methods and procedures are not followed a number of defects may be developed causing discontinuities within the weld. 'ome of the important defects are described in the following. 12.1 I#c%:3 e-e (4&"%# 2omplete fusion may not ta e place, if the mating surfaces are not properly cleaned of all coatings such as mill scales, slag, o#ides etc. This defect may also be caused by insufficient current, because of which the base metal does not melt properly. Hapid rate of welding also leads to improper fusion. The different types of incomplete fusion are shown in .ig.58. %ac, of sidewall fusion

%ac, of interpass fusion

%ac, of root fusion Fig.23 6ac, of fusion (or) incomplete fusion 12.2 P%r%&"-. 3orosity is formed when a number of gas poc ets or voids are trapped during the cooling process. =se of e#cessively high current and longer arc length are the reasons for this type of defect. 3orosity may occur in two ways0 1ither dispersed through the weld or as a large poc et at the root near to the bac up plate in a groove weld. Improper welding procedures and careless use of bac up plates result in porosity in groove welds. 12.3 I#*!e94*-e 3e#e-r*-"%# In certain instances, partial penetration may be adequate. *owever when the weld penetration is less than that specified, it is termed as inadequate penetration. This type of defect, primarily occurring in groove welds, is due to insufficient groove angles, very large electrodes, inadequate weld current, larger welding rates, or insufficient gaps at the

root of welds. The defect can be avoided by means of bac up plates. Inadequate penetration is shown in .ig.5<.

1ncorrect se-uence

Omission of bac,gouging or inade-uate bac,gouging with too small a root gap

Fig.2" 6ac, of penetration 12.4 U#!er c4--"#$ This type of defect is formed due to the use of e#cessive current or an e#cessively long arc. $ portion of the metal is burnt away reducing the thic ness of the joint at the edge of the weld. The defect is detected easily by visual inspection and repaired easily by depositing additional weld material. 1#amples of under cutting are shown in .ig.5(. 7ndercuts 7ndercuts

7ndercuts

(a)

(b)

Fig. 2% T.pical e;amples of undercut defect. (a) $ide and curved< (b) arrow and crac, li,e 12.) S *$ "#c 4&"%# 'lag is formed in the welding process due to the chemical reaction of the melted electrode coating. It normally consists of metal o#ides and other compounds. 'ince it has less density than the molten weld metal the slag usually floats on the surface. On cooling, this is removed by the welder. +ut, if the cooling is rapid, the slag may get trapped before it can rise to the surface. When several passes of weld are made to achieve the desired weld siDe, the slag that forms between each process must be removed completely. The

main reason for slag inclusion is due to the failure to remove the slag fully between runs. Overhead welds are also susceptible to slag inclusion and hence adequate care should be ta en. 'lag inclusion is shown in .ig. 5?. Bot cleaned or incorrect electrode

"lag trapped in undercut Barrow root gap Fig.2* Slag inclusion 12./ Cr*c>& 2rac s are by far the most severe form of weld defects. 2rac s occur in the form of brea s in the weld metal. They are the result of internal stresses and form either longitudinally or transversely to the line of weld. 2rac s may e#tend from the welded metal into the base metal. They may also be completely in the base metal very near to the weld in *$A. 2rac s may occur either in the hot or cold form. *ot crac s are formed as the weld begins to solidify. =niform heating and slower cooling will prevent hot crac s. 2old crac s, which occur at room temperature, run parallel to but under the weld in the base metal. =sing certain special electrodes and proper preheating and post heating, cold crac ing can be reduced. 12.2 L*:e *r Te*r"#$ ,amellar tearing is a type of crac ing that occurs in the base metal beneath the weld. It is caused by the combined effects of high, localised stresses from weld contraction and poor through - thic ness ductility in the steel. The tearing is started by the separation of the interface between inclusions and metal 4also nown as delamination6 or by fracture of an inclusion itself. The crac s grown by the joining of the delamination in the same plane or by the shear steps, which join the crac s in different planes. This results in a characteristic step-li e appearance for lamellar tearing. ,amellar tear is shown in .ig. 5F.

Fig. 2- 6amellar tear The influencing factors in lamellar tearing are0 $mount of Kon-metallic inclusion and their orientation. "agnitude of induced normal stresses normal to the plate surface.

The presence of inclusions reduces the ductility of steel in the through - thic ness direction because the bond between the inclusion and steel is much wea er than steel itself. The magnitude of stresses induced depends on the joint design, the imposed degree of restraint on the joint, plate thic ness, siDe of the weld and orientation of the weld. Hestrained corner or T-joints are most susceptible to lamellar tearing, as the through -thic ness contraction stresses are high. +utt welds rarely e#perience lamellar tear. Thic plate, high restraint 4rigid clamping6 and large weld beads all contribute to residual stresses and the chances of tearing. *ydrogen also increases the vulnerability to lamellar tearing. *ence preheat is beneficial to reduce the tendency to lamellar tearing. It is also beneficial to use low hydrogen consumables. The following precautions may be ta en to reduce lamellar tearing. =sing plate material with improved through > thic ness properties. /esigning the joint with minimum through-thic ness stresses. =sing lower strength welding consumables. In special cases, the plate may be ground to a level below where lamellar tearing is e#pected. The area can be provided with weld metal and the attachment weld can be made. 13.0 13.1 WELD DEFECT ACCEPTANCE LEVELS Ge#er*

Weld defects acceptance levels are closely related to the available methods of Kon -/estructive 1#amination 4K/16. 3reviously, radiography was the best technique available for inspection of welds. Then, acceptance conditions were in terms of ma#imum slag inclusion and porosity levels. 3resence of crac s was not acceptable and repairs were absolutely essential. The use of ultrasonic technique has made it possible to detect smaller crac s, when compared to radiography. Thus, welds, which passed radiographic inspection, required e#tensive

repair by the new method of inspection. The method of fracture mechanics has made it possible to assess the potential of crac s to cause serious damage and thus to come up with tolerable defect siDes. 'lag inclusion and porosity may not be particularly deleterious defects unless fatigue type of loading is anticipated. +y fracture mechanics approach, it has been established that the crac s detected by the ultrasonic methods are far smaller than those that affect the safety of the structure. =ses of fracture mechanics method has proved that tolerable defect siDes are large. This would result in cost saving in terms of repair and reduce the level of K/1 inspection. 13.2 Acce3-e! Cr"-er"* (%r We !e! 1%"#-& In general the following weld defects detected during inspection are acceptable for structures. .or joints welded from both the sides, incomplete penetration with thic ness up to (of the parent metal thic ness, but not e#ceeding 5 mm and the length more than ()) mm can be accepted. The aggregate length of flaw shall not be more than 5)) mm per meter length of the joint. Incomplete penetration and crac s are not allowed at or near the end or beginning of a joint. .or joints welded from one side without bac ing strip, incomplete penetration with thic ness up to %(- of parent metal thic ness but not e#ceeding 8 mm at the root is allowed. 'lag inclusion located along the weld as a chain or unbro en line is allowed if their aggregate length does not e#ceed 5)) mm per meter of weld length. 'iDe of the slag may also be considered. Total of isolated gas pores and slag inclusion shall not e#ceed ( in number per square centimetre of the weld. Total of incomplete penetration, slag inclusion on pores located separately or as a chain shall not e#ceed %)- of metal thic ness but not greater than 5 mm when welding is done from both the sides and %(- of metal thic ness, but not greater than 8 mm when welding is done from one side. .or metal thic ness up to %) mm, undercuts shall not be more than ).( mm. .or metal thic ness more than %) mm, undercuts shall not be greater than % mm.

Incomplete weld, molten metal flow, pits and crac s shall not be allowed. 14.0 WELDING INSPECTION There are essentially three steps to be followed to ensure good welding7 they are0 % % 5 1stablishing good welding procedures =se of pre-qualified welders $vailability of competent inspectors in shop and field

It is essential that welded joints are thoroughly e#amined and defects are detected so that any possible distress could be averted. There are several non-destructive testing methods to chec the quality of welds. They are e#plained in the following.

14.1 V"&4* "#&3ec-"%# @isual inspection by a competent person will give a good indication of the quality of welds7 but may not be able to gauge the sub surface condition of the welds. $n e#perienced welder, by visual inspection, would be able to now whether satisfactory fusion and penetration are obtained. *e will be able to recognise good welds by their shape, siDe and general appearance. In a good weld, the metal should be nearly its original colour after it has cooled. In case of over heating, it will give a rusty appearance. There are several scales and gauges to chec the siDe and shape of welds. "ethods of determining the internal soundness of a weld are described in the following section. 14.2 L"94"! Pe#e-r*#-& In this method, a type of dye is spread over the weld surface. This dye penetrates into the surface crac s of the weld. $fter the penetration of the dye, any e#cess material is removed and a powdery developer is sprayed to draw the dye out of the crac s. Then, the outline of the crac s can be seen with na ed eye. In some cases, fluorescent dyes are used for improved visibility of the crac s. 14.3 M*$#e-"c P*r-"c e& The weld that is inspected is Bmagnetised; electrically. 2rac s, which are present at or near the surface, would cause Korth and 'outh poles to form on each side of the crac s. /ry iron filings are then ept on the weld. They form patterns when they cling to crac s. .rom the patterns, the location of crac s, their siDe and shapes are established. 14.4 U -r*&%#"c Te&-"#$ +y means of the ultrasonic equipment, sound waves are sent through one side of the material and they are reflected from the opposite side. These reflections are indicated in a cathode ray tube. $ny defect in the weld will alter the time of the sound transmission. +y the help of the picture in the tube, flaws can be detected and their severity can be judged. 14.) R*!"%$r*3'. This is an e#pensive method and can be used to chec the welds in important structures. 3ortable O-ray machines along with radium or radioactive cobalt would give e#cellent pictures. This method is reliable for butt welds, but is not satisfactory for fillet welds due to difficulty in interpreting pictures. $nother drawbac of the method is the radioactive danger. "uch care has to be ta en while carrying out this inspection to protect the wor ers on the job. $ properly welded connection is usually much stronger 4%.( to 5 times6 than the strength of the members being connected. The reasons for the e#tra strength are0 electrode wire is

made up of premium steel, the metal is melted electrically and the cooling rate is rapid. /ue to these factors, the weld strength is always higher than required by the design. 1).0 CONCLUSION In this chapter, fundamentals of welding, details of the various welding processes, types of welds, common weld defects and weld inspection have been covered. $dvantages of welding over other forms of connection such as bolting and riveting are e#plained in detail. /esign of various types of welded connections in steel structures is e#plained in ne#t chapter. 1/.0 REFERENCES %. 5. 8. <. "orrow *.W., L1lements of 'teel /esignM M, 3rentice *all, Inc. C 1nglewood 2liffs, Kew Eersey-%&GF. /owling 3.E., Pnowles 3.H. and Owens !.W., 'tructural 'teel /esign, The 'teel 2onstruction Institute, +utterworths, ,ondon, %&GG. "c2ormac E.2., L'tructural 'teel /esignM4Third 1dition6., *arper and How, 3ublishers, Kew Jor , %&G%. I'0 G%8 > %&G?. L'cheme Of 'ymbols .or WeldingM, +ureau of Indian 'tandards Kew /elhi.