volume xix, number 2 , 2002 - foundation · enhance the state of gas metal arc welding to the point...

A publication of the James F. Lincoln Arc Welding Foundation

Volume XIX, Number 2 , 2002

Technology and Change

Australia and New ZealandRaymond K. RyanPhone: 61-2-4862-3839Fax: 61-2-4862-3840

CroatiaProf. Dr. Slobodan KraljPhone: 385-1-61-68-222Fax: 385-1-61-56-940

RussiaDr. Vladimir P. YatsenkoPhone: 077-095-737-62-83Fax: 077-093-737-62-87

INTERNATIONAL SECRETARIES

That technology provides the impetus for change is nosurprise. Western civilization has made exponentialprogress through the past two millennia because of signifi-cant developments in technology. The driving force forchange though, more now than at any other time in histo-ry, is the scientific method and then subjecting the resultsfrom the use of the scientific method to applications in thereal world. If the tested technology proves to be externallyvalid – repeatable and robust away from the laboratoryenvironment, then there will be an expected increase inthe use of the newer method.

During the past one hundred years, the world of weldinghas made great technological strides: power source designevolved from power grids, motor generators, and trans-former rectifiers to today’s software driven inverter trans-formers. Such is the state of welding power sources today,that their energy-efficient transformer design and malleablesoftware based output provide unequalled arc performance.

The software-driven inverter transformer continues toenhance the state of gas metal arc welding to the pointwhere the process limits for GMAW are being challenged– and the results are nothing short of spectacular. Lowerspatter, lower weld fumes, lower hydrogen weld deposits,and improved finished weld quality are the essential fea-tures of the newer technology, and they all work togetherto provide lower welding cost. More importantly, weldingprocess development has effectively widened the windowof opportunity to apply GMAW: both thinner and thickersections of a wide range of base material types are effec-tively joined in accordance with governing welding codes.The newer technology is promoting change because thenewer methods are externally valid – they are repeatableand robust.

Central to the change process for gas metal arc welding,and any other welding process in this technocentric era,is the successful transfer of newer technology to the end-user. The precise transfer of information is fundamentalto successful implementation of technology, and thesetwo facets of change carefully intertwine for successfulimplementation.

Jean Piaget, in his theory of cognitive development, identifiesthree elements central to the learning process: assimilation,accommodation, and equilibration. The extrapolation ofPiaget’s theory meshes well with the transfer of technology,and consequently, it serves as an appropriate paradigm forfostering change.

Assimilation of the new technology:• Language should be clear and concise – no jargon.• The documentation for the new technology must support

its features, advantages, and benefits.• Examples and graphics should support the benefits

of the new technology.• Potential pitfalls ought to be clearly defined.

Accommodation of the new technology:• This is where the comparison and contrast occurs and

change begins to take place.• Testing of the newer technology takes form here and the

benefits of the technology are measured against oldermethods, ideas, or tools.

• The quantified results are either externally valid or the new technology is disgarded.

• If the technology is acceptable, then the stage is set for the next phase.

Equilibration of the new technology:• During this final phase, implementation of the new

technology occurs.• There is no conflict between the benefit of the new

technology and existing methods.• The new technology either replaces or dovetails with

the older technology.• The system is balanced.

In this issue, we are featuring new technology that hasbrought about positive changes in welding processes, weldeddesign, and fabrication materials. We think that the presenta-tions will permit you the opportunity to assimilate technologyin the form of ideas, concepts, and tools. All of them areinnovative, and each introduces creative and well reasoneddepartures from what was formerly known to be true. Weinvite you to investigate the technology presented here andmore importantly, we encourage you to begin to accommodatethe technology that we wish to transfer!

Jeff NadzamAssistant Editor

1Welding Innovation Vol. XIX, No. 2, 2002

Cover: A close-up of Amie Laird McNeel’ssculpture “Gaze” shows the importance of welding to her work. Story on page 19.

The views and opinionsexpressed in WeldingInnovation do not neces-sarily represent those ofThe James F. Lincoln ArcWelding Foundation or TheLincoln Electric Company.

The serviceability of aproduct or structure utiliz-ing the type of informationpresented herein is, andmust be, the sole responsi-bility of the builder/user.Many variables beyond thecontrol of The James F.Lincoln Arc WeldingFoundation or The LincolnElectric Company affect theresults obtained in applyingthis type of information.These variables include,but are not limited to, weld-ing procedure, plate chem-istry and temperature,weldment design, fabrica-tion methods, and servicerequirements.

Volume XIXNumber 2, 2002

EditorDuane K. Miller,

Sc.D., P.E.

Assistant EditorJeff Nadzam

The James F. Lincoln Arc Welding Foundation

Omer W. Blodgett, Sc.D., P.E.Design Consultant

Features

Departments

Visit us online at www.WeldingInnovation.com

7 Opportunities: Lincoln Electric Technical Programs

8 Design File: Mixing Welds and Bolts, Part 2

12 Technology Transfer: Tandem GMAW

The Flexibility of Pulsed Spray Transfer

15 Opportunities: Lincoln Electric Professional Programs

16 Lessons Learned in the Field: Beware the Obvious Conclusion!

2 Design & Fabrication of Aluminum AutomobilesThe importance of aluminum to automotive manufacturing is explored in an article that begins with a historical overview and includes a discussion of current state-of-the-art applications.

19 “A Critical Exchange” Welding in the Sculpture StudioAt the Cleveland Institute of Art, Chair of the Sculpture Dept. Amie McNeel leads her students to explore the potential of steel and what she calls the “magic” of welding.

THE JAMES F. LINCOLN ARC WELDING FOUNDATION

Dr. Donald N. Zwiep, ChairmanOrange City, Iowa

John Twyble, TrusteeMosman, NSW, Australia

Roy L. MorrowPresident

Duane K. Miller, Sc.D., P.E.Executive Director & Trustee

www.weldinginnovation.com

2 Welding Innovation Vol. XIX, No. 2, 2002

Design & Fabrication of Aluminum Automobiles

By Frank G. ArmaoGroup Leader, Nonferrous ApplicationsThe Welding Technology CenterThe Lincoln Electric CompanyCleveland, Ohio

HistoryAluminum was first isolated in 1888 asan element. It rapidly gained in appli-cation as people learned how to alloyit to improve its mechanical properties.By the mid-1920s, Pierce Arrow hadbegun to make at least one model ofits cars entirely from aluminum. Whileall-aluminum cars have appeared peri-odically over the years, the use of alu-minum has never become widespreadin the automotive industry. However, inthe last ten to fifteen years, the use ofaluminum in automobiles hasincreased dramatically. In fact, theaverage aluminum content of automo-biles increased 113% between 1991and 2000. Today, the average car con-tains over 250 lbs. (113 kg) of alu-minum alloys.

Over the years, some very well-knowncars have been built entirely from alu-minum. These include:• The Mercedes-Benz 300SL Gullwing

in the 1950s• The Shelby AC Cobra in the 1960s• The Jaguar D type• The Ford GT40

In the more recent past, the fabricationof high-end automobiles from alu-minum has continued with:• The Acura NSX• The Aston Martin Vanquish• The Audi A8• The BMW Z8• The Ferrari 360 Modena• The Mercedes CL coupe• The Plymouth Prowler• The Shelby Series 1• The new Ford GT40

All of these cars are made using eithera monocoque body structure (in whichthe covering absorbs a large part ofthe stresses to which the body is sub-jected) or a space frame made entirelyfrom aluminum. Therefore, it should be fairly obvious that it is possible toobtain very good structural perfor-mance from aluminum. However, all ofthe models listed above are made atrelatively low volumes (20,000 peryear maximum). Is it possible to manu-facture aluminum vehicles at highervolumes? In fact, Audi has taken alarge step by making the Audi A2 completely from aluminum alloys at avolume of 80,000 per year in Europe.

Aside from the all-aluminum car, thereis increasing use of aluminum in outerbody panels (i.e., fenders, hoods,decklids) in virtually every manufactur-er’s model lines. Most bumper beamstoday are made from aluminum alloys.Perhaps even more noteworthy from a

structural standpoint, there areincreasing volumes of aluminumengine cradles (the Chevrolet Impalaand Malibu and the 2002 NissanAltima) and rear suspension cradles(the Chrysler Concorde, the DodgeIntrepid, and the BMW 5 Series). Thefact that these are being made at vol-umes as high as 700,000 per yeargoes a long way toward proving theviability of high volume, all-aluminumautomobiles. As we will see below,welding is a major contributor to mak-ing this possible.

Why Aluminum?

Aluminum has a number of propertiesthat make it attractive for application inautomobiles. However, it has one char-acteristic that overrides all others: itslight weight. Aluminum automotivealloys are one third as dense assteels, while many of them have ten-sile and yield strengths almost equal

Figure 1. Aluminum front frame railcrushed in crash test showing uniformcrushing and energy absorption.

Welding Innovation Vol. XIX, No. 2, 2002 3

to those of construction grade steels.Does this mean that we can make alu-minum parts that weigh one third ofsteel parts? In general, no. Most partsof a car are not strength-limited, butare stiffness-limited. (There are excep-tions to this – the areas around theshock towers are usually strength-limited). Because stiffness is a functionof Young’s modulus, which is 10 x10bpsi (68,950 MPa) for aluminumalloys and 30 x 10bpsi (206,850 MPa)for steels, weight reductions of 2/3 arenot usually possible. Weight reductionsof 40%–45% are more typical.

The U.S. Federal Government publish-es Corporate Average Fuel Economy(CAFE) standards. These standardsdictate the fuel economy levels thatevery auto maker must meet. Failureto meet them can result in penalties.Car manufacturers are under a greatdeal of pressure to increase fuel econ-omy across the board. One of the easiest ways to do this is to reducethe weight of the automobile. Reducingthe total weight of the car by 10% normally results in an 8%–10%improvement in fuel economy. Evensomething as simple a substitution ofan aluminum hood for a steel one hasa significant effect on average fueleconomy.

Aluminum has another advantage oversteel. It can be easily extruded, whilesteel can’t. This allows the designer tocreate complex shapes of varying wallthickness using extruded sections.Internal stiffening ribs can be integrallyextruded, so that cross sections con-sisting of multi-hollows are routinely

used. The only closed section tubingavailable in steels is simple shapessuch as rounds, squares, ovals, etc.This has allowed designers of alu-minum auto structures to venture intoautomotive space frames and hybridstructures, instead of using only themonocoque sheet construction used in steel automobiles.

But what happens in a crash? Won’tan aluminum car just crumple into aball of aluminum foil? The answer isan emphatic “No!” For a detailed dis-cussion of the behavior of aluminumautomotive structures in crash tests,the interested reader is referred to“Automotive Aluminum Crash EnergyManagement Manual,” publication AT5,published by the Aluminum Associationin Washington D.C. For our purposes,

it is sufficient to say that it is not diffi-cult at all to make aluminum automo-biles that meet or exceed the NHTSAcrash test requirements set out inFMVSS 208, which is the same criteri-on steel cars must meet. Innovationsin alloys and processing have resultedin materials that crush uniformly andabsorb energy better than steels.Figure 1 shows an actual front crashrail from a production car. It has begunto crush and has buckled in a con-trolled, uniform manner, absorbingcrash energy and ending up about halfas long as it started. Good designsand improved materials are the keysto superior crash performance.

Space Frames versus Sheet Cars

Until approximately thirty years ago,cars were made as an assembly of asheet metal body and a heavier, sepa-rate chassis. The body provided little, ifany, structural strength and wasassembled by resistance spot welding(RSW) and bolting. The frame, madefrom thicker members, was assembledprimarily by arc welding, rivetting, andbolting.

Then, in the late 1960s and early1970s, automotive design changed.The so-called “unibody” was born. Inthis construction method, the entirebody, except for the hang-on panels, ispart of the car’s structure and con-tributes to the car’s stiffness andstrength. There is no separate frame,although small front or rear subframesmay be used to hold the engine, sus-pension, etc. These cars are madealmost exclusively of steel sheet ofvarious thicknesses which is stampedand joined together by RSW. In 2002,the car makers have seventy plusyears of experience in RSW and arevery good at it. All of the infrastructureto support RSW is in place.

Why not just make aluminum cars byup-gauging the material thicknessfrom steel to aluminum and assemblethem by RSW? Indeed, that’s possibleand one of the major aluminum com-panies supports this strategy, using acombination of RSW and adhesivebonding. However, this approach oftenresults in extra costs.

Aluminum can be easily extruded, while steel can’t


For production volumes under 100,000per year, it has been shown that eithera pure space frame or a hybrid spaceframe/sheet approach is more costeffective. This results mostly from thefact that extrusion dies are relativelyinexpensive, while stamping dies aremuch more costly.

Figure 2 shows a photograph of anAudi A8 space frame. This method ofconstruction employs “nodes” whichare made from castings, formed sheetor extrusions. Each node serves as a joining point for several structuralmembers. The nodes are designed sothat there is at least one slip plane foreach joint. This serves to minimizegaps between the node and the struc-tural member coming into it. While it ispossible to use other joining methods,such as adhesive bonding, most of theexisting aluminum space frame cars,including the Audi A8 and A2, theFerrari 360, the Ford GT40, the BMWZ8, and the Shelby Series 1, are arcwelded. The Audi A8 space frame contains 70 m (approximately 230 ft.)of gas metal arc welding. Only theAston Martin Vanquish is adhesivelybonded and riveted.

Why Gas Metal Arc Welding?

If automakers go to aluminum vehi-cles, why not just spot weld them

together as they do now on steel vehi-cles? There are a number of reasons.

RSW of aluminum presents someunique challenges. The aluminumreadily alloys with the copper spotwelding tips, so electrode life can bevery short. The electrical conductivityof aluminum is much higher than thatof steel, so not as much resistanceheating takes place at the interface of the two pieces to be joined.

Consequently, currents required forRSW are often three times what theyare for steel, so the equipment usedfor steel seldom can be used for aluminum.

Because of these issues, manyautomakers have moved away fromRSW for aluminum. For joining alu-minum sheet parts, many have gone to self-piercing riveting, often in combi-nations with adhesives. However, forjoining extrusions and/or castings,these processes have some limitations:• Only lap joints are possible. Tee or

butt joints cannot be made.• Physical access to both sides of the

joint is required.

• When joining castings or extrusions,it is usually necessary to add aflange in order to make the joint.This adds back some of the weightthat has been saved.

GMAW is not without limitations either.When joining thin sheet, welding dis-tortion is sometimes excessive. Theheat of the welding arc softens theHAZ of the joint, reducing mechanicalproperties. However, GMAW has anumber of advantages that have madeit the preferred method for joining ofcastings, extrusions, and thicker sheet(thicker than 0.070 in. or 1.8 mm), asfollows• It is usable for all types of joints –

lap, tee, and butt.• It is easily automated using robotics• Access to only one side of the joint

is required.• It is fairly tolerant of part misalign-

ment and joint gaps.• Capital equipment costs are low.• It is a well-established, widely used

process.

GMAW Technology Development

On the surface, gas metal arc welding(GMAW) might appear to be an older,low tech process. It is anything but.Even ten years ago, it would havebeen very difficult, if not impossible, toGMAW aluminum members as thin as0.040 in. (1 mm) thick. Today, weldingthin aluminum is fairly easy and the

There are 17 pulsing variablesthat can be programmed

Figure 2. The Audi A8 spaceframe.


Figure 3. Software screen used for programming Lincoln pulsing power supplies.

development of GMAW has becomean enabling technology for the use ofaluminum in automotive fabrication.

GMAW of thin aluminum was compli-cated by the fact that short circuitingarc transfer (short arc) is not recom-mended for GMAW of aluminum alloys.When gas metal arc welding steels toweld thin material, the welder uses afiner welding wire and keeps goinglower in current and deeper into shortarc transfer. However, if this approachis used on aluminum, incompletefusion defects occur. Short circuitingarc transfer is never recommended for aluminum because of this.

Spray transfer is always recommendedfor welding aluminum. In years past, itwas impossible to weld thin aluminum,say, of 1/16 in. thickness (1.6 mm),because even with the smallest diameteraluminum wire available for GMAW,0.030 in. (0.8mm), the welding currenthad to be above 85 amperes to getspray transfer. This was just too muchcurrent to weld thin materials, soGMAW of thin aluminum simply wasnot performed in production.

However, electronics technology devel-oped and made it possible to control

the welding process much more pre-cisely and to change the welding cur-rent very quickly. Pulsed GMAW wasdeveloped. In fact, it was developedover twenty years ago. However, it isvery different today than it was then.

Pulsed GMAW has proved to be espe-cially applicable to welding of thin alu-minum. Fundamentally, the weldingcurrent is pulsed between a high peakcurrent where spray transfer isobtained and a low background cur-rent where no metal is transferredacross the arc. This means that wehave spray transfer, but the averagewelding current is much lower. So nowwe can weld aluminum as thin as0.020 in. (0.5 mm) and we can havespray transfer at average currents aslow as 30 amperes or so, even withlarger diameter 0.047 in. (1.2 mm)wires.

Early pulsed GMAW power supplieswere transformer controlled and limit-ed to 60 or 120 Hertz pulsing frequen-cies. Today’s power supplies areinverter based, software controlled,and programmable. Control frequen-cies are often 20 KHz. This flexibilityhas allowed a tremendous amount ofGMAW process development.

Figure 3 shows a computer screen ofproprietary software used for pro-gramming pulsed GMAW in a con-temporary power supply. There areseventeen pulsing variables that canbe programmed. The programmerchooses a wire feed speed (WFS) anddevelops the optimum pulsing vari-ables for that WFS and saves them.This process is repeated over therange of wire feed speeds and thedata is saved as a program.

This whole process is invisible to theuser, who merely picks a WFS. Thepower supply then automatically setsall of the pulsing variables. The onlyother control is a “Trim” control thatgives the welder control over arclength. All the welder has to do is picka program number and a WFS to haveaccess to a program that is optimizedfor pulsing for the specific filler alloyand wire diameter being used.Furthermore, if the specific applicationis so unique that the standard programis inadequate, it can easily be repro-grammed by the manufacturer or, insome cases, by the user.

Figure 4 shows a photo of one suchpower supply. This power supply canbe combined with a push-pull weldingtorch. Using such a welding system,

Figure 4. Multi–process programmablewelding system.


Figure 5. Schematic of Pulse OnPulse waveform.

Figure 6. A weld in 3 mm aluminum made using Pulse On Pulse welding.

aluminum wires as fine as 0.030 in.(0.8 mm) can be fed as far as 50 ft.(15 m).

However, GMAW technology develop-ment still has not stopped, or evenslowed down. The tremendous capa-bilities available today to control and

switch the welding process are stimu-lating continuing development. Forinstance, a recent development is acontrol pulsing logic for thin aluminumcalled “Pulse On Pulse.” This wave-shape is shown schematically inFigure 5. In this process, a number ofrelatively high energy pulses are alter-nated with the same number of lowenergy pulses, causing a weld rippleto be formed each time the low energypulses fire, and resulting in a very uni-form weld bead. An example of PulseOn Pulse welding is shown in Figure6. This type of pulsing has shown itselfto be very applicable to automotivefabrication and is in use already insuch applications.

The FutureAs in all areas of life, the future is hardto predict. Is there an all-aluminum carin your future? This depends on a lotof factors. If the Federal governmentincreases CAFE requirements, it willdrive automakers to reduce vehicleweight further. If aluminum ingot pricesstay low, additional aluminum use ismore likely. However, ingot prices havebeen volatile in the past, and thatscares auto manufacturers. Whateverthe future, though, there is likely to begreater use of aluminum in cars. Thatmeans that some of us will continue totry to improve gas metal arc weldingtechnology.


Lincoln Electric Technical Programs

Opportunities

High Productivity Welding

Submerged Arc Welding, October 7-8, 2002Gas Metal Arc Welding, October 9-10, 2002Each program will provide the basic theoretical concepts that support the process, advanced topics,and new developments. Both programs will empha-size welding process optimization and welding cost reduction opportunities. These seminars aredesigned to benefit welding engineers, technicians,supervisors, instructors, quality assurance personnel,and manufacturing engineers.Fee: $375 each, or $595 for both.

Welding of Aluminum Alloys,Theory and PracticeOctober 15-18, 2002Designed for engineers, technologists, techniciansand welders who are already familiar with basicwelding processes, this technical training programprovides equal amounts of classroom time andhands-on welding.Fee: $595.

Welding & Fabricating Nickel AlloysOctober 21-22, 2002A joint effort of the Nickel Development Institute andthe Welding Technology Center of Lincoln Electric,this program provides welding, metallurgical, andconstruction practice information about joining nickelbased alloys. It will specifically cover thin-sheetmetallic lining methods for FGD installations andnickel clad base material used for chimney con-struction, but others interested in the fabrication of nickel alloys are also encouraged to attend.Fee: $595.

Space is limited, so register early to avoid disappointment. For full details, see

www.lincolnelectric.com/knowledge/training/seminars/

Or call 216/383-2240, or write to Registrar, Professional Programs, The Lincoln Electric Company,22801 St. Clair Avenue, Cleveland, OH 44117-1199.



Mixing Welds and Bolts, Part 2Practical Ideas for the Design Professional by Duane K. Miller, Sc.D., P.E.

Design File

In a previous edition of Welding Innovation (Volume XVIII,Number 2, 2001), Part 1 of “Mixing Welds and Bolts” waspublished. That column dealt with snug-tightened and pre-tensioned mechanical fasteners, including rivets, combinedwith welds, as well as existing specification requirementsfor such combinations. Part 1 can be obtained by down-loading a PDF file from the Welding Innovation web site atwww.weldinginnovation.com. Part 2 will address combiningwelds with slip-critical, high-strength bolted connections,and will also examine existing specification provisions forvarious combinations of welds and bolts in light of recentresearch.

Review of Part 1

In Part 1, general information was provided on bolted con-nections. Snug-tightened, pretensioned, and slip-criticalbolted connections were defined. ASTM A325 and A490bolts were identified, and the capacity of rivets identified astypically about half of the strength of A325 bolts. Slip-criti-cal joints have bolts that have been installed in a mannerso that the bolts are under significant tensile load with theplates under compressive load. They have faying surfacesthat have been prepared to provide a calculable resistanceagainst slippage. Slip-critical joints work by friction: the pretension forces create clamping forces and the frictionbetween the faying surfaces work together to resist slip-page of the joint. The basic design philosophy relies on friction to resist nominal service loads. The provisions fordesign of slip-critical connections are intended to provide90–95% reliability against slip at service load levels. In itsstrength limit state, slip can occur and the bolts will go intobearing. This should not be the case for service loads.

The focus of this Design File series is not upon bolted connections, but rather upon connections that are composedof both welds and bolts. For the snug-tightened and preten-sioned bolted connections, it was shown that welds cannotbe assumed to be capable of sharing loads with the mechan-ical fasteners. AWS D1.1 Structural Welding Code-Steel and

AISC LRFD Steel Specification require that the welds bedesigned to carry the entire load under these conditions. TheCanadian standard CAN/CSA-S16.1-01 provides a morerational criterion by permitting load sharing between weldsand bolts for service loads, providing the higher of the twocapacities can carry all factored loads alone.

Part 2 focuses on slip-critical joints, combined with welds.As mentioned in Part 1, this topic is the subject of ongoingresearch and consideration by the various technical com-mittees. Much of this work has been done by Drs. G. Kulakand G. Grondin and their co-workers of the University ofAlberta, Canada, and definitive conclusions have not yetbeen reached as to how these findings should be incorpo-rated into US standards, such as AWS D1.1 and AISCLRFD. However, at least some parts of current standardsare likely to be determined to be unconservative, and prac-ticing engineers should review these data and determinehow specific projects should be addressed in light of thesefindings. The same research has drawn into question someof the current specification requirements for snug-tightenedconnections when welds are added, and these findings willbe reviewed.

Code Provisions for Slip-Critical Connections with Welds

The issue of mixing mechanical fasteners and welds isaddressed in AWS D1.1: 2002 Structural WeldingCode–Steel. Provision 2.6.7 states:

“Connections that are welded to one member and boltedor riveted to the other shall be allowed. However, rivetsand bolts used in bearing connections shall not be con-sidered as sharing the load in combination with welds ina common faying surface. Welds in such connectionsshall be adequate to carry the entire load in the connec-tion. High-strength bolts installed to the requirements forslip-critical connections prior to welding may be consid-ered as sharing the stress in the welds. (See:

www.weldinginnovation.com/papers/default.asp


A

ADISPLACEMENT

LOAD

TRANSVERSE FILLET WELD

LONGITUDINALFILLET WELD

BOLTEDCONNECTION

B

B

Figure 1.

Specifications for Structural Joints Using ASTM A325 orA490 Bolts of the Research Council on StructuralConnections.)”

Note: Part 1 cited the 2000 version of D1.1, in whichthese provisions were contained in 2.6.3. The latest version is largely unchanged in concept, although theunderlined words in 2.6.7 are new for the 2002 edition.

The fourth sentence deals with slip-critical connections.Notice that, in order for sharing to be considered, this pro-vision requires that the high-strength bolts be installed“prior to welding.” More will be said on this issue later.AISC LRFD – 1999, Provision J 1.9, expresses the samegeneral philosophy when it states:

“In slip-critical connections, high-strength bolts are permitted to be considered as sharing the load with the welds.”

The commentary to this provision provides some additionalunderstanding of both the AISC and AWS provisions:

“For high-strength bolts in slip-critical connections toshare the load with welds it is advisable to fully tensionthe bolts before the weld is made. If the weld is placed

first, angular distortion from the heat of the weld mightprevent the faying action required for the developmentof the slip-critical force. When bolts are fully tensionedbefore the weld is made, the slip-critical bolts and theweld may be assumed to share the load on a common-shear plane. The heat of welding near bolts will not alterthe mechanical properties of the bolts.”

The straightforward reading of these provisions, andindeed the intent of them, is to permit the direct combina-tion of the capacity of the slip-critical connection and theweld. However, recent research indicates that this is notthe case, and such an assumption may be unconservative.

The commentary that addresses the angular distortionexplains the apparent justification for requiring that thebolts be installed before welding. The basis for such arequirement is suspect, however. Kulak and Grondin pointout that “slip resistance of the bolted joint is independent ofthe amount of area between faying surfaces. As long asthere is some area, which is a physical necessity for properpreloading of the bolts…, then the slip resistance will bedeveloped.” (Kulak and Grondin, from the minutes of theAISC TC6 Connections Task Committee, June 12-13,2002.) Thus, the apparent justification for the sequentialrequirement may be suspect.

Different Deformation Capabilities

In Part 1, the differences in the deformation capabilitiesbetween welded connections and those joined with bolts in either a snug-tightened or pretensioned manner wasidentified as the factor that precluded the simple arithmeticaddition of the capacities of the two systems. The weldswere identified as being “stiff,” whereas the snug-tightenedor pretensioned bolted connection could slip to distributethe applied loads on the mechanically fastened joint.

The concept presented in codes with respect to slip-criticalconnections was presumably based upon the lack of slip in the connection (that is, their “stiffness”), justifying theassumption that the capacities of the two types of joiningsystems (welds and bolts) can be joined. Ultimately, a slip-critical bolted connection will slip, but if a weld is added,such a connection cannot slip. Thus, the capacities of thetwo elements cannot be combined in terms of the ultimatestrength capacity.

Figure 1 contains a conceptual plot of the load/displace-ment relationships for welds and bolts. Note that theload/deformation relationships are different for each of thethree elements. It should be noted that the two types ofwelds shown are not equally “stiff.” The actual curve for thebolted connection is illustrative only; in fact, there would bevarious curves for the different types of bolted connections.


Additionally, while in this illustration the three curves are allshown having the same strength, under most conditions,the capacity of each element will be different. The differ-ences in stiffness preclude simple mathematical additionsof the various capacities.

Figure 2 illustrates six possible connection details: a) boltsonly, b) longitudinal fillets only, c) transverse fillets only, d)bolts and transverse fillets, e) bolts and longitudinal fillets,and f) bolts with both longitudinal and transverse fillets. Inthis illustration, it is assumed that the strength of the con-nections in Figure 2a-2c is equal, as is illustrated in Figure1. The bolts, for example, offer the same load resistance,as do the transverse fillet welds. All the bolts shown inFigure 2 are assumed to be slip-critical.

If the code provisions cited above were correct, that is, if the capacities of welds and slip-critical bolts could bemathematically combined, then the connections with boltsand welds in Figure 2d and 2e would both be twice thevalue of the connections in Figure 2a-2c. Further, if theseprovisions were correct, the capacity of Figure 2f would bethree times that of Figure 2a-2c. Loads, however, are notevenly split between the various elements in the mixedconnection, because of the differences in the load/defor-mation curves.

Referring again to Figure 1, the bolted connection in Figure2a would have a load/deformation curve like the bolt curve.For a unit strength of 1, the deformation experienced wouldalso be a unit of 1. For the longitudinal fillet in Figure 2b,the strength is also normalized to a value of 1, but thedeformation capacity is estimated to be 1/6 of the boltedconnection. The transverse fillet of Figure 2c also hasstrength of 1, but with a deformation capacity of about onesixth of the longitudinal fillet weld.

To analyze the combination of welds and bolts and theirultimate load capability, constant displacements for eachelement must be considered, and the resistances to defor-mation for each element added to determine the totalcapacity of the combination. Consider the combination oflongitudinal fillet and bolts (Figure 2e). Line A in Figure 1illustrates a likely deformation level that would contribute tothe total connection strength of a level 1. However, ratherthan a 50-50 split, the weld contributes about 60% of thestrength, with 40% coming from the bolts. At line B wherethe weld is capable of delivering 100% of its strength, thebolts can contribute only about 80% of theirs, and the com-bination is not 200%, but rather about 180%, or 10% less.Of course, the code provisions would suggest 200%, thedirect addition of both members.

The same exercise could be performed with bolts andtransverse welds. The reduced deformation capacity of the transverse fillet makes the differences even more

Figure 2.

(a)

(b)

(c)

(d)

(e)

(f)


pronounced. Thus, the significance of these differences in displacement is more pronounced for the connections composed of bolts and transverse welds.

A Proposed Model

Kulak and Grondin propose a model whereby the ultimateload resistance of the joint can be computed from the fol-lowing relationship:

Rutl joint = Rfriction + Rbolts + Rtrans + Rlong

Where Rfriction is the frictional resistanceRbolts is the bolt shear resistanceRtrans is the transverse weld shear resistance Rlong is the longitudinal weld shear resistance

Rfriction is estimated to be 0.25 times the slip resistance ofthe slip-critical bolted joint. For slip-critical connections inconjunction with welds, this factor is always present, but isaccounted for differently, depending on the orientation ofthe weld (longitudinal versus transverse). This factor, ofcourse, would be zero for bearing-type bolted connections.

Rbolts depends on the type of weld (transverse or longitudinal)and the condition of bearing, whether already in bearing(positive) or unknown (indeterminate).

For transverse welds along with slip-critical bolted joints,the ultimate joint resistance is the strength of the trans-verse weld plus the frictional resistance, or the bolt shear,whichever is greater.

For longitudinal welds along with slip-critical bolted joints,the ultimate joint resistance is a percentage of the boltshear plus the shear resistance of the longitudinal weld plus the frictional resistance, or the bolt shear, whichever is greater. Under positive bearing conditions, 75% of theultimate bolt shear strength is used, and for indeterminatebearing conditions, 50% is used.

The work of Kulak and Grondin indicates that for slip-criti-cal connections, the code provisions are unconservative.For example, the capacity of a combined longitudinal weldand bolts is equal to the weld capacity plus 50% of the boltcapacity. For this condition AWS and AISC would indicatethe weld capacity plus 100% of the bolts capacity, thusoverestimating the capacity of the connections.

Recall from Part 1 that in the general case, AWS and AISCrequire that combinations of welds and bolts of the bearingtype be designed such that the entire load is transferred

through the weld. Thus, regardless of the capacity of thebolts, any small weld addition effectively eliminates thecapacity of the bolts. The preceding model could be used toaddress these snug-tightened connections. In such cases,Rfriction is zero. The greater capacity of the bolts and weldscould then be used, as is the case in the Canadian stan-dard CAN/CSA-S16.1-01. The Kulak work would indicatethat this approach is correct and conservative.

The Kulak work has also revealed new information regard-ing the combination of longitudinal and transverse filletwelds, which are subject to some of the same deformationcapacity differences. This will be addressed in a future edi-tion of Design File.

Conclusion

The responsible technical committees are evaluating theseresearch findings, and changes to specifications will nodoubt result. Currently, the data suggest that while a por-tion of the slip-critical bolt capacities can be directly addedto the capacities of longitudinal welds, the same is not thecase for slip-critical bolts and transverse welds. In the caseof the latter, the greater capacity of the two elements is aconservative assumption.

Acknowledgements

A sincere thanks to all of my professional colleagues whogenerously gave of their time to review and critique thecontent of this article.

ReferencesAISC (1999). Load and Resistance Factor Design Specification for Structural

Steel Buildings. American Institute of Steel Construction (AISC), Chicago,Illinois.

AWS (2002). AWS D1.1: 2002 Structural Welding Code—Steel. AmericanWelding Society (AWS), Miami, Florida.

Jarosch, K.H., and Bowman, M.D., (1985). “Behavior of Tension Butt JointsUsing Bolts and Welds in Combination.” Struct. Engrg. Rep. CE-STR-85-20, Purdue University, Lafayette, Indiana.

Kulak, G.L., Fisher, J.W., and Struik, J.H., (1987). Guide to Design Criteria forBolted and Riveted Joints, 2nd Ed., John Wiley and Sons Inc., New York,New York.

Kulak, G.L., and Grondin, G.Y., (to be published). “Strength of Joints thatCombine Bolts and Welds,” from the minutes of the AISC TC6Connections Task Committee, June 12-13, 2002.

Manuel, Thomas J. and Kulak, Geoffrey L., (2000). “Strength of Joints ThatCombine Bolts and Welds,” J. of Struct. Engineering, ASCE, 126(3), 279-287.


Suitable advances in technology fosterchange, and the changes that we arewitnessing in the welding industrytoday are due to the forces of newpower source technology. Moreover, itis no accident that much recentresearch and development work focus-es on gas metal arc welding (GMAW).Historically, the bridge and structuralfabrication industries have tended toview GMAW with derision, and withsome cause. Issues regarding incom-plete fusion set a precedent that hasseriously limited the implementation ofGMAW in code quality applications.However, recent advances in GMAWpower source technology allow anopportunity to challenge conservativethinking regarding the use of gasmetal arc welding.

GMAW Process Review

Gas metal arc welding, by definition, isa process that produces coalescenceof metals by heating them with an arcbetween a continuously fed filler metalelectrode and the work. The processuses shielding from an externally sup-plied gas to protect the molten weldpool. The application of GMAWrequires DC+ (reverse) polarity to theelectrode.

In gas metal arc welding there are fourtraditional modes of metal transfer:

• Axial spray transfer is the higherenergy mode and it provides thebenefit of excellent fusion, but it isrestricted to the flat and horizontalpositions. Spatter is absent and thepenetration profile is uniform.

• Pulsed spray metal transfer (GMAW-P) is a higher technology mode ofmetal transfer that provides the ben-efits of axial spray, but at a loweraverage current. The pulse exceedsthe transition to spray transfer for a

very short period, then it is reducedto a lower energy level often associ-ated with short circuit transfer. Metaltransfer only occurs during the pulsepeak. This mode of metal transfer

was developed because of its abilityto lower spatter levels, overcomeincomplete fusion defects, and beimplemented in out-of-position weld-ing applications.

• Globular metal transfer is historicallyassociated with the use of 100% car-bon dioxide shielding. The moltenmetal transfer occurs non-axially indroplets that are one and a half tothree times larger than the diameterof the electrode. Globular transferusually results in higher spatter levels,and it is prone to incomplete fusion.

• Short-circuit transfer is the low heatinput mode of GMAW, in whichmetal transfer occurs as a result ofphysical contact between the elec-

Technology TransferContributed by Jeff Nadzam, Group Leader, GMAW and GTAWThe Welding Technology Center, The Lincoln Electric CompanyCleveland, Ohio

Tandem GMAW:

The Flexibility of Pulsed Spray Transfer

Figure 1. Welding ASTM 516 grade 70 base material test plates with tandemGMAW.

Pulsed spray metal transfer provides the benefits of axial spray at a

lower average current


trode and the weld pool. This modeof metal transfer historically excelsin the joining of sheet metal thick-ness material or open root pipewelding. When applied to a basematerial thicker than 3/16 in. (5 mm),short-circuit transfer can result inincomplete fusion.

New Developments

The transformation of GMAW, led bythe development of software-driveninverters, has earned a second chancefor critical code quality applications.Leading the way is the pulsed spraymode of metal transfer employed withTandem GMAW (Figure 1). The newerfeatures of the process invite severalopportunities to reduce the manufac-turing costs associated with both lightand heavy steel fabrication. Considerthe following:

• The arc can be tailored for penetra-tion profile, heat input, and mechani-cal properties.

• The use of tandem GMAW, whichincorporates two independentpulsed arcs in the same weld pool,leads to higher deposition rates andhigher electrode efficiencies thanwas previously possible.

Figure 2. Software employed to manipulate the pulsedspray mode of metal transfer. Note the nine pulsed spraytransfer waveform components: Ramp Up Rate, RampOvershoot, Peak Current, Peak Time, Tailout Time, TailoutSpeed, Step-off Current, Background Current, BackgroundTime, and Frequency.

• GMAW produces lower levels ofhydrogen and the process lendsitself to a cost-effective alternativefor joining high performance andhigh strength low alloy steels.

• Weld bead appearance is excellent.

• GMAW produces low levels of weldingfumes, which can be an advantage forshop or other indoor fabrication.

The Tailorable Arc

Essential to the development or themanipulation of a particular weldingwaveform is an understanding of therelationship of the electrode type, itsdiameter, and the shielding gasemployed. Shielding gas is central to determining the outcome of the fin-ished weld and it has a profound effectupon penetration profile, bead shape,toe wetting, and mechanical proper-ties. For example, the arc characteris-tics of a 95% argon + 5 % oxygenblend may be suitable for high travelspeed welding on sheet metal thick-ness, but the penetration profile maybe undesirable for plate thicknessmaterial. The electrode diameter,whether it is solid or metal cored, hasan associated maximum current carry-ing capability. A 0.045 in. diameter (1.1

mm) electrode reaches its maximumcurrent at approximately 420 Amps,and the maximum for a 0.035 in. diam-eter (0.9 mm) is approximately 220Amperes.

The deposition rate is associated with a specific useable current range and,quite naturally, the higher the currentthe more energy there is to apply to the weld joint. For example, a solid0.045 in. (1.1 mm) diameter electrodecan carry more welding current than a0.035 in. diameter (0.9 mm) electrode.Given similar wire feed speeds and arctravel speeds, we would expect that thepenetration into the base material

would be greater for the larger diameterthan the smaller. Similarly, it would betrue that the deposition rate potentialwould be higher for the larger diameterelectrode. Selecting the electrode diam-eter and the shielding gas thenbecomes a choice based upon theneeds of the joint, the base materialtype and thickness, the requiredmechanical properties, and through-put.

Advancing technology has produced astounding

improvements in GMAW-P


It is true, then, that each welding appli-cation carries specific user-definedrequirements for the finished weld.Because the new technology permits

modification to the components of the GMAW-P waveform, these weldrequirements may be more readilyachieved.

Nine components are associated withthe pulsed spray waveform (see Figure2). Each component in combinationwith the others provides specific attrib-utes to the finished weld. The energylevel and appearance of the arcchange with the manipulation of eachof the components. For example, highpulsed frequency produces a narrowarc that may lend itself to use on flare-bevel type weld joints. Increasing thepeak current results in an increase inenergy associated with deeper weldpenetration, and the opposite is alsotrue. The use of higher front ramp

rates stiffens the arc and thus increas-es the immunity of the arc to arc blowconditions.

Pulsed GMAW waveforms for the tan-dem GMAW process include a widerange of electrode diameters, from0.030 in. to 1/16 in. (0.08 mm to 1.6mm) and material types. Shielding gasselections include argon and carbondioxide or argon and oxygen combina-tions. Ternary, three-part shielding gasblends are rarely employed. TandemGMAW typically uses electrodes of thesame diameter, but in some instancesthe diameter of the trail electrode maybe smaller. The judgement about whento employ differing diameters generallyfollows a case-by-case evaluation.

Tandem GMAW

Tandem GMAW is a welding processthat uses two DC electrode positive(DC+) arcs in the same weld pool, andthey are identified as the ‘Lead’ andthe ‘Trail’ arcs (Figure 3). Each arcuses its own power source and wiredrive, and the possible combinationsof modes of metal transfer employedin tandem GMAW are as follows:

• Axial Spray Transfer Lead Arc +Axial Spray Transfer Trail Arc

• Axial Spray Transfer Lead Arc +Pulsed Spray Transfer Trail Arc

• Pulsed Spray Transfer Lead Arc +Pulsed Spray Transfer Trail Arc

Figure 3. The tandem torch permitsthe delivery of two electrodes: thelead arc and the trail arc. The nozzleprovides uniform shielding gas cover-age for the molten weld pool.

Table 1. Results of mechanical testing for each of 2 ASTM A516 grade 70 testplates.

The lead arc, in all cases, determinesthe penetration level of the weld, andthe trail arc provides the final beadshape and weld bead reinforcement.The programs that are created for tan-dem GMAW, particularly in the case ofthe Pulse + Pulse, are developed withDC+ arc compatibility in mind.

The development of tandem GMAWhad higher arc travel speed as its coreobjective, and it is generally used withmaterial the thickness of sheet metal.The travel speed on sheet metal appli-cations is usually 1.5 to 1.9 times thetravel speed for a single arc, and it isnot uncommon to find travel speeds of more than 100 ipm (2.5 m/min).

For thicker sections of material wheremultiple pass welding is necessary,the deposition rate for tandem GMAWcan vary from 20–45 lbs/hr (9–21kg/hr). The arc travel speeds rangefrom 25–45 ipm (0.6–1.2 m/min). Toachieve anticipated mechanical prop-erties, tandem GMAW may require theuse of special welding techniques.

Testing Tandem GMAW

Two ASTM A516 grade 70 test plateswere assembled, welded, and testedto the requirements of ANSI/AASH-TO/AWS D1.5-96 “Bridge WeldingCode,” Section A. The welding proce-dure involved the use of two 0.052 in.(1.4 mm) diameter ER70S-6 elec-trodes in tandem, and the shieldinggas employed was a 90% argon + 10

Both bridge and structural fabricators could take advantage

of the tandem GMAW process

Mechanical Properties

Tensile Strength

Yield Strength

Avg. Ft.-Lbs. @ -20° F

Avg. Joules @ -29° C

Test Plate #1

90,000 psi (620 Mpa)

73,300 psi (505 Mpa)

76

103

Test Plate #2

89,600 psi (617 Mpa)

73,300 psi (505 Mpa)

78

106


Figure 4. This multiple pass TandemGMAW weld has a finished appearancethat resembles a Submerged ArcWeld.

Lincoln Electric Professional Programs

Opportunities

Blodgett’s Design of WeldedStructuresSeptember 24-26, 2002Blodgett’s Design of Welded Structuresis an intensive 3-day program whichaddresses methods of reducing costs,improving appearance and function,and conserving material through theefficient use of welded steel in a broadrange of structural applications.Seminar leaders: Omer W. Blodgett andDuane K. Miller. 2.0 CEUs. Fee: $595.

Blodgett’s Design of WeldmentsOctober 29-31, 2002Blodgett’s Design of Weldments is an intensive 3-day program for those concerned with manufacturing machinetools, construction, transportation,material handling, and agriculturalequipment, as well as manufacturedmetal products of all types. Seminarleaders: Omer W. Blodgett and DuaneK. Miller. 2.0 CEUs. Fee: $595.

Fracture & Fatigue Control in Structures: Applications of Fracture MechanicsOctober 15-17, 2002Fracture mechanics has become theprimary approach to analyzing andcontrolling brittle fractures and fatiguefailures in structures. This course willfocus on engineering applicationsusing actual case studies. Guest seminar leaders: Dr. John Barsomand Dr. Stan Rolfe. 2.0 CEUs.Fee: $595.

Space is limited, so register early toavoid disappointment. For full details, see

http://www.lincolnelectric.com/knowledge/training/seminars/

Or call 216/383-2240, or write toRegistrar, Professional Programs, TheLincoln Electric Company, 22801 St. ClairAvenue, Cleveland, OH 44117-1199.

% carbon dioxide. Both the lead arcand the trail arc employed the pulsedspray mode of metal transfer. Theeffective deposition rate for the testwas 32 lbs/hr (14.5 kg/hr). The resultsof the mechanical testing, for each ofthe two plates, were as shown in Table 1.

Discussion of Results

The results indicate that tandemGMAW using pulsed spray transferwith 0.052 in. (1.4 mm) diameter elec-trodes is viable for complete penetra-tion weld joints in structural fabrication.In addition, it would be fair to assumethat both bridge and structural fabrica-tors could take advantage of the tan-dem GMAW process not only forcomplete penetration groove welds,but also for completing fillet welding on stiffeners. Moreover, it would

appear that the use of robotic or specialized hard automation would inmany cases provide both the motionand process control necessary toimplement tandem GMAW.

Conclusion

Advancing technology permits the useof GMAW-P on a wide range of weld-ing applications, including thicker sec-tion base materials. The use of tandemGMAW for depositing high quality weldmetal with excellent fusion representsthe culmination of several years ofapplication research and development.The optimized arc condition permitsthe use of two DC+ pulsed arcs in thesame molten puddle. Moreover, it rep-resents a reasonable alternative forwelding components of bridges andother structures.



The most dramatic example of a lesson I learned in the field occurredabout five years ago. At the time, withtwo decades of welding industry salesand technical support experienceunder my belt, I thought I had prettymuch “seen it all.” As it turned out, Iwas about to get my come-uppance,and learn a lasting lesson in theprocess.

The Initial Trouble Call

My responsibilities in the WeldingTechnology Center of The LincolnElectric Company include direct cus-tomer support. One day in the earlydays of the use of high performance

(HP) steel, I received a trouble callfrom a bridge fabrication shop thatwas having cracking problems whilewelding a test girder for the FederalHighway Turner Fairbanks ResearchLab. This HP steel had a 70 ksi (485MPa) minimum yield strength with

low sulfur and carbon content, ascompared to A852. I asked the usualquestions and learned that the weld,specifically, a complete joint penetra-tion groove weld, had passed the fabri-cator’s procedure qualification tests,but for some reason, it was failing onthe job. The transverse and longitudi-nal cracking was delayed, taking threedays to show up. The fabricator wasusing LA -100 electrode and 960 flux,and welding in accordance with theAWS D1.5 Bridge Welding Code.

Together with the customer, I carefullyreviewed the possible causes. TheCJP groove weld was on a 1-5/16 in.(33 mm) thick plate (Figure 1), so thepotential for high residual stressesexisted. I also thought we might bepicking up some alloy out of the basematerial. This scenario led me to rec-

ommend slowing down the coolingrate through increasing preheat andinterpass temperature, trying to mini-mize the influence of the base metalon the weld deposit.

The Second Failure

The fabricator made another attempt,following all of my recommendations.To our consternation, the crackingproblem persisted. At this point, I washaving a hard time figuring out whatwe could have missed. Since the testgirder was being welded for theFederal Highway Administration, theirpersonnel became involved. The prob-lem was starting to mushroom.Eventually, there were a lot of peoplegiving a lot of opinions about what wasreally going on.

Beware the Obvious Conclusion!

Lessons Learned in the Field Contributed by Lon Yost, Group Leader, FCAW & SAWThe Welding Technology Center, The Lincoln Electric CompanyCleveland, Ohio

Figure 1. Test plate with flange splice.

To our consternation, the cracking problem

persisted


The Third AttemptFor the third attempt, two representa-tives and one consulting welding engi-neer from the Federal HighwayAdministration, a representative fromthe steel supplier, and yours truly allconvened at the fabricator’s facility.When we got there, we went to theoffice to review as much as we could,including drawings and joint designs.Once we were done reviewing thepaper side of the job, we went into theshop to investigate details related tothe handling of consumables, whatthey were doing to determine preheat,and so forth.

I was still convinced that the problemwas somehow related to the steel. Itjust seemed to me that if we had the

right preheat and interpass tempera-ture, that would solve the problem. Atour of the shop seemed to lend somecredence to my theory. I observed:

• Low winter time ambient tempera-ture in the unheated shop

• Flux was being heated in a holdingoven (Figure 2), rather than in a trueflux drying oven

• The welding set-up would not permitaccess to the bottom of the plate forheating

We dealt with the lack of access to thebottom of the plate by increasing thepreheat substantially beyond levelsrequired by AWS D1.5. We stayedthere while the CJP groove weld wasagain attempted on the bridge girder.In this case, the third time was acharm—the weld did not crack thatday, nor in the subsequent days. Butthe mystery of the first two failuresremained.

Lab Testing of the Failed Welds

Sections of the failed welds were dis-tributed to the Federal HighwayAdministration, Lincoln Electric, andthe steel supplier for analysis. Afterreturning to our Welding TechnologyCenter, I subjected my sample to a fullarray of metallographic tests in our lab.Upon analyzing the results, MarieQuintana, Lincoln’s Manager of NewProducts, Consumables, felt that wewere dealing with a case of hydrogenassisted cracking in the LA100 fillermetal (Figure 3). This was something Ihadn’t considered. My twenty years of

experience had included many exam-ples of steels experiencing hydrogenassisted cracking, but never had weldmetal been more sensitive to this typeof cracking than the base metal. Newdevelopments in base metals, such asthis HP steel, resulted in a new possi-bility. I had assumed that when you’re

having a cracking problem, the steel is to blame. But this time, my assump-tions had misled me. The problem wasactually due to the susceptible chem-istry of the weld.

It turned out that on the third attemptat the weld, our team effort to supplymore than sufficient preheat, to main-tain interpass temperature and to

250 F

FLUX

500 F

Figure 2. A flux holding oven.Figure 3. Weld sample with hydrogenassisted crack.

My assumptions had misled me


meticulously condition the flux hadactually overcome this tendency tohydrogen assisted cracking. Even afterseveral days, no cracks appeared. Atthat point, the customer and theFederal Highway Administration weresatisfied.

What Took Me So Long?

I was just mystified at how I couldhave spent two decades in this indus-try, and never before run into a dif-fusible hydrogen cracking problemattributable to Lincoln’s LA100 elec-trode. So I tried to remember all thetimes in the past when I had recom-mended the use of LA100. It becameapparent to me that most of the appli-cations for which I had suggestedLA100 had been on projects which

entailed small, single pass welds—sit-uations that were very different fromthis multi-pass CJP weld on a relative-ly thick bridge girder such as the oneshown in Figure 4.

It was a humbling experience to admitto my professional associates on thisjob that the actual problem was not inthe HP steel, but actually in my weldmetal. And it certainly taught me tolook beyond the obvious whenattempting to diagnose and solve aweld cracking problem.

The Consequences

As an outgrowth of this cracking prob-lem and some other similar cases, theuse of controlled hydrogen fluxes hasbroadened. Controlled hydrogen sub-

merged arc fluxes have been specifi-cally designed to resist moisture pick-up and aid in the diffusion of hydrogenout of the weld deposit as the weld isbeing made. Different welding con-sumables have been tested to deter-mine the best consumables forwelding HP steel. Information aboutthe recommended consumables isavailable from the AmericanAssociation of State Highway andTransportation Officials (AASHTO) inthe Guide Specifications for HighwayBridge Fabrication with HPS70WSteel, which can be accessed onlineat www.aashto.org.

Figure 4. Girders for highway bridges such as this one require multi-pass CJP groove welds.

www.aashto.org


“It’s magic,” declares Amie LairdMcNeel when she is asked to talkabout welding and its application in the sculpture studio. McNeel, anaccomplished sculptor and Chair of the Sculpture Department at theCleveland Institute of Art, is dedicatedto sharing that magic with her students.She came to the profession in aroundabout way, starting as a marinebiologist documenting the migrationpatterns of humpback whales. But itseemed to her that marine biologistsspent a great deal of time communi-cating their findings to other marine

biologists, and she now says, “I didn’tthink I could make a difference in thatfield.” Instead, she turned her skills ofobservation and thoughtful analysis tothe creation of art, and she fell in lovewith the medium of steel.

It is clear from a single conversationwith McNeel that she is equally pas-sionate about creating art and teach-ing it. “Teaching, for me,” she says, “isa way of not only sharing what I havelearned over the years, but of explor-ing new ideas and possibilities everyday.” She regards making sculpture as “a way of maintaining a criticalexchange between hands, materialsand tools.”

Two Student Sculptors

McNeel referred Welding Innovation totwo of her sculpture students whosework is represented here. Jeff Guhde,class of 2004, describes MIG weldingas being “just like hot glue.” To createthe six segments that compose his“Stretch and Reach” (Figure 1), heexplains, “I forged these centers out ofwhat started as a cube, and then all ofthe linear extensions were rods that Iforged just a little bit to get a ham-mered texture, and then they werewelded on to each of the centers. So

welding is not a predominant designelement, but it was used just as ameans of connecting the pieces.”Guhde was drawn to working in metalinitially because of the structural quali-ty of steel, and what he refers to asthe “immediacy” of welding. On a morephilosophical level, he wants toexplore the way that sculptural formsorganize the interior space in whichthey are installed. He says, “It’s some-thing that my teacher, Amie, got mestarted on—seeing space as a physi-cal material.”

“A Critical Exchange”

Welding in the Sculpture Studio

By Carla RautenbergWelding Innovation Contributing WriterThe James F. Lincoln Arc Welding FoundationCleveland, Ohio

Figure 1. The close-up on the left shows one of the welded connections Jeff Guhdeused to create his sculpture "Stretch and Reach," shown on the right.

Seeing space as a physical material

20 Welding Innovation Vol. XIX, No. 2 2002

Having grown up on a farm, McNeellearned early to work with her hands,and welding came fairly naturally toher. She had this in common withanother of her students, Trevor Korns,who received his Bachelor of Fine Artsfrom the Institute this year. He creditshis fascination with creating mythicalcreatures to having grown up aroundanimals on his parents’ farm. Hissculptures are made from scrap steeland occasionally cut-up pieces of oldmachinery, and fabricated using a MIGwelder. Korns describes his “WingedLion” (Figures 2 and 3) as follows:“The self-destructive nature of humani-ty is represented by the death of thewinged lion. I used the winged lion torepresent humanity because of thepredatory status of the lion andbecause it has the ability to fly.”

“Thoughts Existing in Space”

Amie McNeel refers to her sculpturalwork as “thoughts existing in space.”Some of her pieces, for example, “Fins I”(Figure 4) and “Fins III” (back cover),seem to evoke the machinery of thefarm and a lost agricultural Eden.“Fins II” (Figure 5) projects more of an industrial aspect. However, McNeelinsists that “It is not nostalgia that

motivates this work, but a need toreevaluate such ideas as obsoles-cence, necessity, progress, ingenuity,and physical labor.” The 10 ft. (3m)diameter work entitled “Gaze” (Figure 6)is a visual exploration of movementwithout any actual moving parts.Figure 7 depicts McNeel installing“Gaze” for an exhibition.

Teaching the Welding Part

McNeel describes her teachingprocess with great animation. Askedhow she begins, she replies, “I learnedfrom one of my instructors that you letpeople experience the material first,and try to do it on their own, as longas it’s safe. Teach them safety. Andteach them how to use the tool. Ibreak down the functions of thetool…so they can problem-solve ifthere’s a variable that keeps themfrom achieving a good weld.”

Since steel is expensive and requirespractice to manipulate, McNeel hasher students do their initial designwork using materials such as paperand cardboard. As she puts it, “Theydesign out of a material that has thesame geometric qualities but lessdemands. And then they move tosteel.”

The beginning students learn threemethods of joining: oxy-acetylenewelding, oxy-acetylene brazing, andgas metal arc welding. “They feel lessintimidated by the wire feed welder, soI teach them harder ones first,” saysMcNeel.

In the second semester, studentslearn to cast bronze and aluminumand, in due course, are introduced togas tungsten arc welding of aluminum,and in some cases, steel.

McNeel laughs as she explains,“When the students have to buy theirfirst plate of steel, they’re shocked. It’s80 bucks for a sheet of 1/4 in. (6 mm)steel. Then, they start looking at all thescraps in a very different way. So itbecomes precious once they put amonetary value on it. With aluminum

and brass and different alloys ofsteels, then they start to really appre-ciate their designs. They becomemuch more focused on practicingbecause they don’t want to screw itup. We balance mechanical joinery,drilling, bending, designs of connec-

tions, we practice mechanical joinery,stitching, sewing, pinning, using steeldowels. They practice design alongwith just welding. They learn how tochoose what’s appropriate design—when to weld, when to pin and bolt,when to adjust.”

A Commitment

She goes on, “Then, we start gettingphilosophical about welding. Becauseit’s magic. Then they realize how giv-ing and flexible and malleable steel is.But it takes a few years before theirdesigns and their ability and practiceall combine to the point where theycan say, isn’t this magic? Where youcan take a metal and you can estab-lish a molten pool of it, and only in thatmolten pool, when you have that arcright, and that gas shield right, will itaccept this material, and it’ll agree tobecome molten, and it’ll agree tomerge with this base material, andyou’ve made a permanent understand-ing, you’ve established a weld. Thatpiece is now one piece. Now what kindof a commitment is that? You’ve decid-ed this form is worthy of being a per-manent thing. You kind of push them to realize that their decisions are notcasual.”

McNeel sums up her experience asteacher and artist with a wry grin, say-ing, “Students are kind of dumbfoundedwith the patience you have to havewith yourself and with the process inorder to learn it when you deal directlywith materials…it takes practice.People don’t like to practice. Just theact of doing it sometimes is the mostimportant thing.”

Figure 2. Recent CIA graduate TrevorKorns created “Winged Lion” usingscrap steel and MIG welding.

They realize how giving and flexible steel is


Figure 7.Amie L. McNeel,Chair of theSculpture Dept. atthe ClevelandInstitute of Art,installing her steelwelded sculpture,“Gaze.”

Figure 3. A close-up of the Trevor Korns’mythical “Winged Lion.”

Figure 4. “Fins I” (steel, wood,rubber, cables, 10 ft. [3m] long,4 ft. [1.2 m] in diameter) byAmie Laird McNeel.

Figure 5.“Fins II” (steel,wood, plastic,cables, 9 ft.[2.7 m] long, 5 ft. [1.5 m] indiameter) byAmie LairdMcNeel.

Figure 6. “Gaze” (steel, cables,10 ft. [3 m] in diameter) byAmie Laird McNeel.

▲

▲

▲

▲▲

Photograph by MJ Toles

NON-PROFIT ORG.U.S. POSTAGE

PAIDJAMES F. LINCOLN

ARC WELDING FND.

“Fins III” (steel, wood, rubber, cables, 14 ft. [4.3] long, 6 ft. [1.8 m] in diameter) is just one example ofAmie Laird McNeel’s welded sculpture. See story on page 19.

P.O. Box 17188Cleveland, OH 44117-1199

TheJames F. LincolnArc WeldingFoundation

volume xix, number 2 , 2002 - foundation · enhance the state of gas metal arc welding to the point...

Documents