Volume 20 - Materials Selection and Design Part 13 potx

This article provides a brief description of various joining processes, a summary of good design practices from a joining process standpoint, and several examples of selected parts and j

Fig 9 Cross section of the initial design for a rotor housing

The boring passes are difficult to plan for the initial design There are several internal sharp corners; in addition, there is

an internal angled cut at 45° to the part axis which cannot be accessed by a standard tool that will clear other internal surfaces The grooves also have different axial widths For the initial design, therefore, two grooving tools, two standard boring tools, and additional special tools to reach the internal sharp corners would be required to make the desired cuts

Figure 10(a) shows a revised design that simplifies the machining Upon consultation, it was determined that the dimensions of the grooves could be standardized, eliminating the need for one of the grooving tools The internal sharp corners were replaced by radiused corners to permit machining with a standard boring insert Finally, the initial 45° angled cut was replaced with a 60° angled cut that could be produced with a standard 55° boring insert mounted in a standard -5° lead boring bar as shown in Fig 10(b) In the revised design, the required cuts can be made with two standard boring tools and a single grooving tool, saving at least three tool changes Because the special tools required to machine the sharp internal corners would wear rapidly, the revised design also results in increased tool life and improved part quality

Fig 10 (a) Cross section of a rotor housing redesigned to simplify machining (b) Access to internal features

using a boring bar with a standard 55° boring insert

Design for Machining

D.A Stephenson, General Motors Powertrain Group

References

1 O.W Boston, Metal Processing, 2nd ed., Wiley, 1951, p 1-8

2 J.G Bralla, Design for Excellence, McGraw Hill, 1996, p 46-47

3 C.V Starkey, Engineering Design Decisions, Edward Arnold, London, 1992, p 178-179

4 R Bakerjian, ed., Chapter 11, Tool and Manufacturing Engineer's Handbook, Vol VI, Design for Manufacturability, 4th ed., Society of Manufacturing Engineers, 1992

5 D.A Stephenson and J.S Agapiou, Chapters 2, 11, and 13, Metal Cutting Theory and Practice, Marcel

Dekker, 1996

6 Machining, Vol 16, ASM Handbook, ASM International, 1989

7 Bar Products Group, American Iron and Steel Institute, Steel Bar Product Guidelines, Iron and Steel

Society, Warrendale, PA, 1994, p 164-166

8 J.S Agapiou, An Evaluation of Advanced Drill Body and Point Geometries in Drilling Cast Iron, Trans NAMRI/SME, Vol 19, 1991, p 79-89

9 H.W Stoll, Tech Report: Design for Manufacture, Manuf Eng., Vol 100 (No 1), 1988, p 67-73

10 H.A ElMaraghy, Evolution and Perspectives of CAPP, CIRP Ann., Vol 42 (No.2), 1993, p 1-13

11 L Alting and H Zhang, Computer Aided Process Planning: The State of the Art Survey, Int J Prod Res.,

Vol 26, 1989, p 999-1014

12 S.K Gupta and D.S Nau, Systematic Approach to Analysing the Manufacturability of Machined Parts,

Comput.-Aided Des., Vol 27 (No 5), 1995, p 323-342

13 F.G Mill, J.C Naish, and C.J Salmon, Design for Machining with a Simultaneous-Engineering

Workstation, Comput.-Aided Des., Vol 26 (No 7), 1994, p 521-527

14 G Boothroyd, Product Design for Manufacture and Assembly, Comput.-Aided Des., Vol 26 (No 7), 1994,

p 505-520

Design for Joining

K Sampath, Concurrent Technologies Corporation

Introduction

JOINING is an important manufacturing activity employed in assembling parts to make components The individual parts

of a component meet at joints Joints primarily transmit or distribute forces generated during service from one part to the

other parts of an assembly A joint can be either temporary or permanent Commonly, five joint types are used in the joining of parts: butt, tee, corner, lap, and edge (Fig 1)

Fig 1 Types of joints Source: Ref 1

The selection of an appropriate design to join parts is based on a concurrent understanding of several considerations related to product and joining process Product-related considerations include codes and standards, fitness for service, aesthetics, manufacturability, repairability, reliability, inspectability, safety, and unit cost of fabrication Considerations related to joining process include material types and thicknesses, joint (part) geometry, joint location and accessibility, handling, jigging and fixturing, distortion control, productivity, and initial investment Additional considerations include whether the joint is fabricated in a shop or at a remote site, possibilities for premature failure, and containment in case of

a catastrophic failure (this is applicable, for example, to components subjected to nuclear radiation)

The term joint design emphasizes designing of a joint based on product-related considerations for meeting structural

design requirements The design or selection of appropriate joint type is determined primarily from the type of service loading For example, butt joints are preferred over tee, corner, lap, or edge joints in components subjected to fatigue loading The specific joint design aspects, such as the size, length, and relative orientation of the joint, are based on stress calculations that are derived from an evaluation of service loads, properties of materials, properties of sections, and appropriate structural design requirements An ideal joint is one that effectively transmits forces among the joint members and throughout the assembly, meets all structural design requirements, and can still be produced at minimal cost (Ref 1) Individual articles in various Sections of this Volume specifically address design of parts or components based on an understanding of several product-related considerations vis-à-vis appropriate structural design requirements

The term design for joining refers to creating a mechanism that allows the fabrication of a joint using a suitable joining

process, at minimal cost In this context, design for joining emphasizes how to design a joint or conduct a joining process

so that components can be produced most efficiently and without defects This involves selection and application of good design practices based on an understanding of process-related manufacturing aspects such as accessibility, quality, productivity, and overall manufacturing cost This article provides a brief description of various joining processes, a summary of good design practices from a joining process standpoint, and several examples of selected parts and joining processes to illustrate or highlight the advantages of a specific design practice in improving manufacturability

Acknowledgements

The following sections in this article were adapted from handbooks published by ASM International (as cited in the list of References): "Mechanical Fastening" (Ref 2), "Adhesive Bonding" (Ref 3), "Brazing" (Ref 5, 6), and "Soldering" (Ref 7)

The numbered examples were compiled from Welding, Brazing, and Soldering, Volume 6 of the 9th Edition Metals Handbook

Design for Joining

K Sampath, Concurrent Technologies Corporation

Joining Processes

Joining processes include mechanical fastening, adhesive bonding, welding, brazing, and soldering Mechanical fastening and adhesive bonding are often (but not always) used to produce temporary or semi-permanent joints, while welding, brazing, and soldering processes are used to provide permanent joints Mechanical fastening and adhesive bonding usually do not cause metallurgical reactions Consequently, these methods are preferred when joining dissimilar combinations of materials, and for joining metal-matrix, ceramic-matrix, and polymer-matrix composites that are sensitive to metallurgical phase changes or polymerization reactions

Mechanical Fastening (Ref 2) The selection and satisfactory use of a particular fastener are dictated by the design requirements and conditions under which the fastener will be used Consideration must be given to the purpose of the fastener, the type and thickness of materials to be joined, the configuration and total thickness of the joint to be fastened, the operating environment of the installed fastener, and the type of loading to which the fastener will be subjected in service

Threaded fasteners are considered to be any threaded part that, after assembly of the joint, may be removed without damage to the fastener or to the members being joined

Rivets are permanent one-piece fasteners that are installed by mechanically upsetting one end

Blind fasteners are usually multiple-piece devices that can be installed in a joint that is accessible from only one side When a blind fastener is being installed, a self-contained mechanism, an explosive, or other device forms an upset on the inaccessible side

Pin fasteners are one-piece fasteners, either solid or tubular, that are used in assemblies in which the load is primarily shear A malleable collar is sometimes swaged or formed on the pin to secure the joint

Special-purpose fasteners, many of which are proprietary, such as retaining rings, latches, slotted springs, and studs, are designed to allow easy, quick removal and replacement and show little or no deterioration with repeated use

Adhesive Bonding (Ref 3) An adhesive is a substance (usually an organic or silicone polymer) capable of holding materials together in a functional manner by surface attachment The capability of holding materials together is not an intrinsic property of a substance but, rather, depends on the context in which that substance is used Two important, basic facts about adhesive materials are that a substance called an adhesive does not perform its function independent of a context of use and that an adhesive does not exist that will bond "anything to anything" with (implied) equal utility

The major function of adhesives is for mechanical fastening Because an adhesive can transmit loads from one member of

a joint to another, it allows a more uniform stress distribution than is obtained using a mechanical fastener Thus, adhesives often permit the fabrication of structures that are mechanically equivalent or superior to conventional assemblies and, furthermore, have cost and weight benefits

Although the major function of adhesives is to fasten, sometimes they are also required to seal and insulate Formulations that are good electrical and/or thermal conductors are also available Further, adhesives prevent electrochemical corrosion

in joints between dissimilar metals and resist vibration and fatigue In addition, unlike mechanical fasteners, adhesives do not generally change the contours of the parts that they join

Detailed information on adhesives and adhesive bonding is available in Adhesives and Sealants, Volume 3 of the Engineered Materials Handbook published by ASM International

Welding includes both fusion welding and solid-state welding processes

Fusion welding processes involve localized melting and solidification and are normally used when joining similar material combinations or materials belonging to the same family (e.g., joining one type of stainless steel with another type) Figure 2 illustrates the type of welds commonly used with fusion welding processes such as arc welding (Ref 1)

Fig 2 Types of welds Source: Ref 1

Fusion welding processes also include electron beam welding and laser welding These two welding processes require precise joint gap and positioning Joint designs and clearances that overwhelmingly trap the beam energy within the joint cavity are preferred for increasing process efficiency Figure 3 shows preferred and non-recommended joint designs for electron beam welding (Ref 4) When joining thick sections, the preferred joint designs allow the weld metal to freely shrink without causing cracking

Fig 3 Optimum versus least desirable weld configurations (a) Not recommended maximum confinement of

molten metal, minimum joining cross section (arrows); wastes beam energy for melting, nonfunctional metal (b) Most favorable volume of melt not confined; maximum joining cross section (arrows) (c) Not recommended maximum confinement of melt (unless gap is provided); joining cross section less than plate cross section (d) Most favorable minimum constraint and confinement of melt; minimum internal stresses; warpage can be offset by bending prior to welding; tilt can be offset by location of T-arm at less than 90° to

base prior to welding Fillet obtained by placing wire in right corner and melting it with the beam (e) Not recommended two successive welds; second weld is fully constrained by the first weld and shows strong tendency to crack Source: Ref 4

Solid-state welding processes preclude melting and solidification and therefore are suitable for joining dissimilar materials However, the process conditions may allow solid-state metallurgical reactions to occur in the weld zone When metallurgical reactions occur, they can either benefit or adversely affect the properties of the joint From a metallurgical perspective, the application of both fusion welding and solid-state welding processes must be evaluated using appropriate

weldability testing methods for their ability to either recreate or retain base metal characteristics across the joint These

weldability evaluations combine material, process, and procedure aspects to identify combinations that would provide a weld joint with an acceptable set of properties

Solid-state welding processes also have special joint design or part cross-section requirements For example, drive and inertia friction welding processes require that one of the parts exhibit a circular or near-circular cross section Diffusion bonding is another solid-state welding process that allows joining of a variety of structural materials, both metals and nonmetals However, diffusion bonding requires an extremely smooth surface finish (8 m) to provide intimate contact of parts, a high temperature, and a high pressure, first to allow intimate contact of the parts along the bond interface, followed by plastic deformation of the surface asperities (on a microscopic scale), and second to promote diffusion across the bond interface The need to apply pressure while maintaining part alignment imposes a severe limitation on joint design

continuous-Alternatively, when exceptional surface finish is difficult to achieve, a metallurgically compatible, low-melting interlayer can be inserted between the parts to produce a transient liquid phase on heating On subsequent cooling this liquid phase undergoes progressive solidification, aided by diffusion across the solid/liquid interfaces, and thereby joins the parts This process has characteristics similar to those of the brazing process

Brazing (Ref 5, 6) is a process for joining solid metals in close proximity by introducing a liquid metal that melts above

450 °C (840 °F) A sound brazed joint generally results when an appropriate filler alloy is selected, the parent metal surfaces are clean and remain clean during heating to the flow temperature of the brazing alloy, and a suitable joint design that allows capillary action is used

Strong, uniform, leakproof joints can be made rapidly, inexpensively, and even simultaneously Joints that are inaccessible and parts that may not be joinable at all by other methods often can be joined by brazing Complicated assemblies comprising thick and thin sections, odd shapes, and differing wrought and cast alloys can be turned into integral components by a single trip through a brazing furnace or a dip pot Metal as thin as 0.01 mm (0.0004 in.) and as thick as 150 mm (6 in.) can be brazed

Brazed joint strength is high The nature of the interatomic (metallic) bond is such that even a simple joint, when properly designed and made, will have strength equal to or greater than that of the as-brazed parent metal

The mere fact that brazing does not involve any substantial melting of the base metals offers several advantages over other welding processes It is generally possible to maintain closer assembly tolerances and to produce a cosmetically neater joint without costly secondary operations Even more important, however, is that brazing makes it possible to join dissimilar metals (or metals to ceramics) that, because of metallurgical incompatibilities, cannot be joined by traditional fusion welding processes (If the base metals do not have to be melted to be joined, it does not matter that they have widely different melting points Therefore, steel can be brazed to copper as easily as to another steel.)

Brazing also generally produces less thermally induced distortion, or warping, than fusion welding An entire part can be brought up to the same brazing temperature, thereby preventing the kind of localized heating that causes distortion in welding

Finally, and perhaps most important to the manufacturing engineer, brazing readily lends itself to mass production techniques It is relatively easy to automate, because the application of heat does not have to be localized, as in fusion welding, and the application of filler metal is less critical In fact, given the proper clearance conditions and heat, a brazed joint tends to "make itself" and is not dependent on operator skill, as are most fusion welding processes

Automation is also simplified by the fact that there are many means of applying heat to the joint, including torches, furnaces, induction coils, electrical resistance, and dipping Several joints in one assembly often can be produced in one multiple-braze operation during one heating cycle, further enhancing production automation

Soldering (Ref 7) is a joining process by which two substrates are bonded together using a filler metal (solder) with a liquidus temperature that does not exceed 450 °C (840 °F) The substrate materials remain solid during the bonding process The solder is usually distributed between the properly fitted surfaces of the joint by capillary action

The bond between solder and base metal is more than adhesion or mechanical attachment, although these do contribute to bond strength Rather the essential feature of the soldered joint is that a metallurgical bond is produced at the filler-metal/base-metal interface The solder reacts with the base metal surface and wets the metal by intermetallic compound formation Upon solidification, the joint is held together by the same attraction, between adjacent atoms, that holds a piece of solid metal together When the joint is completely solidified, diffusion between the base metal and soldered joint continues until the completed part is cooled to room temperature Mechanical properties of soldered joints, therefore, are generally related to, but not equivalent to, the mechanical properties of the soldering alloy

Mass soldering by wave, drag, or dip machines has been a preferred method for making high-quality, reliable connections for many decades Correctly controlled, soldering is one of the least expensive methods for fabricating electrical connections

Advantages of brazing and soldering include the following:

• The joint forms itself by the nature of the flow, wetting, and subsequent crystallization process, even when the heat and the braze or solder are not directed precisely to the places to be joined

• The process temperature is relatively low, so there is no need for the heat to be applied locally, as in welding

• Brazing and soldering allow considerable freedom in the dimensioning of joints, so that it is possible to obtain good results even if a variety of components are used on the same product

• The brazed or soldered connections can be disconnected if necessary, thus facilitating repair

• The equipment for both manual and machine brazing/soldering is relatively simple

• The processes can be easily automated, offering the possibility of in-line arrangements of brazing/soldering machines with other equipment

References cited in this section

1 O.W Blodgett, Joint Design and Preparation, Welding, Brazing, and Soldering, Vol 6, 9th ed., Metals Handbook, American Society for Metals, 1983, p 60-72

2 W.J Jensen, Failures of Mechanical Fasteners, Failure Analysis and Prevention, Vol 11, ASM Handbook (formerly 9th ed Metals Handbook), American Society for Metals, 1986, p 529-549

3 Adhesives, Engineered Materials Handbook Desk Edition, M Gauthier, Ed., ASM International, 1995, p

Design for Joining

K Sampath, Concurrent Technologies Corporation

Basic Design Considerations

When designing a joint, one should initially consider manufacturability of the joint, whether at a shop or at a remote site For example, consider the need for a high integrity, high-performance joint between two dissimilar materials such as a low-carbon steel and an aluminum alloy If this joint has to be produced at a remote site, the available choice of joining processes is extremely limited A viable alternative would be to produce at a shop a transition piece involving the two dissimilar materials Using controlled process conditions at a shop, one could produce a high-integrity transition piece using one of the solid-state welding processes The selection of the appropriate solid-state welding process would depend

on joint (part) geometry A transition joint between a plate and a pipe is best produced using a friction welding process, while a joint between two large plate surfaces is best produced using explosive bonding Because these joining processes preclude melting and solidification, they provide high-integrity joints free from porosity or solidification-related defects Transition pieces so produced could be used at a remote site to make similar metal joints between component parts with

no undue quality assurance or quality control concerns

Design for Joining

K Sampath, Concurrent Technologies Corporation

Good Design Practices

A joint must be designed to benefit from the inherent advantages of the selected method of joining For example, braze joints perform very well when subjected to shear loading, but not when subjected to pure tensile loading When using a brazing process to join parts, it would be beneficial to employ innovative design features that would convert a joint subjected to tensile loading to shear loading For example, use of butt-lap joints instead of butt joints can provide a beneficial effect in flat parts and tubular sections

Joints must be designed to reduce stress concentration Sharp changes in part geometry near the joint tend to increase stress concentration or notch effects Smooth contours and radiused corners tend to reduce stress concentration effects Figure 4 shows a number of ways to redistribute stresses in a brazed joint (Ref 8)

Fig 4 Design of a brazed joint to redistribute stress Source: Ref 8

When determining appropriate joint designs, one should initially consider standard or recommended joint designs In practice, several standard joint designs may be suitable for producing a joint Subtle or innovative features could be added

to the recommended joint designs to improve productivity through mechanization or automation, to enhance joint performance, and to ensure safety

Orientation and Alignment. Design features that promote self-location and maintain the relative orientation and alignment of component parts save valuable time during fit-up and enhance the ability to produce a high-quality joint For example, operations involving furnace brazing or diffusion bonding with interlayers benefit from such a type of joint design, because they also require pre-placement of the brazing filler or the interlayer in the joint

The pin-socket type of temporary joints in modern electrical, telephone, and computer connectors allow temporary joining

of cables in only one way These joint designs strongly discourage any inadvertent misalignment or wrong orientation of the connectors and thereby eliminate a variety of hazards The snap-on interlocking features in twisted, threaded, or non-threaded adapter joint designs, commonly used in children's toys, often allow the snapping sound of a latch to indicate the satisfactory completion of the joint and its safety for the intended use

Jigging and fixturing can also be used to maintain relative orientation of parts When necessary, the fixturing devices should be designed for the least possible thermal mass and pin-point or knife-line contact with the parts Fixtures of low thermal mass and minimal contact with the parts reduce the overall thermal load during joining Further, arc welding processes generally allow higher deposition rates when joining is performed in the downhand position, where gravity effects tend to support a large volume of molten weld metal at the joint region When joining parts that exhibit a nonplanar joint contour, positioning equipment can be used to continuously manipulate the parts so that the welding is performed in the downhand position In such cases, the design of the joint and fixtures should be complementary to the positioning equipment used, and it should not interfere with the functioning of the positioning equipment

Joint Location and Accessibility. From a structural integrity standpoint, joint locations should be chosen such that they are not in regions subjected to maximum stress Concurrently, joints must be placed in locations that will allow operators to readily make the joints using the selected method of joining Figure 5 illustrates the effect of location of a joint on accessibility (Ref 1) Limited accessibility can reduce the overall quality of the joint, decrease productivity, or both Invariably, limited accessibility to produce joints also limits accessibility to perform nondestructive evaluation of joint quality, either during the time the joint is made or afterwards

Fig 5 Effect of joint location on accessibility Source: Ref 1

Weld joint designs employ bevel angles and root openings to enhance accessibility to the welding torch (or electrode) and provide adequate weld penetration The best bevel angles provide adequate accessibility while reducing the amount of weld metal required to complete the joint Currently, computer-based software tools are available to facilitate the selection of a weld joint for minimizing the amount of weld metal Use of such computer-based selection of joint designs increases welding productivity (joint completion rate), improves quality, and reduces overall fabrication cost, but such designs must be used only when they are consistent with structural design requirements For this reason, codes such as the ASME Boiler and Pressure Vessel Code, Section IX: Welding Qualifications, and ANSI/AWS D1.1 Structural Welding Code provide flexibility to a welding manufacturer (fabricator) to select or change weld joint design for fabrication, but they require the manufacturer to qualify the welding procedure to meet design performance requirements whenever changes are made to a previously qualified, nonstandard weld joint design In recent years, the use of narrow-gap gas-metal arc welding and submerged arc welding techniques in the place of conventional welding techniques for welding thick-section pressure vessel steels has contributed significantly to increased weld joint completion rates

Unequal Section Thickness. When constituent members of an assembly exhibit unequal section thicknesses, modifications to the recommended joint designs will be necessary for a variety of technical reasons (Ref 9), but mainly to provide a smooth flow of stress patterns through the unequal sections When making a fillet weld using an arc welding process, if thicknesses of the members are not greatly different, directing the arc toward the thicker member may produce acceptable penetration However, special designs for joining will be required when the components to be welded exhibit a large heat sink differential (difference in heat-dissipating capacities) When a thick member is joined to a thin member, the welding heat input (mainly current) needed to obtain a good penetration into the thick member is sometimes too much for the thin member and results in undercutting of the thin member and a poor weld Similarly, if the proper amount of current for the thin member is used, the heat is insufficient to provide adequate fusion in the thick member, and again a poor weld results Too little heat input can also cause underbead cracking in certain structural materials

A widely applicable method of minimizing heat sink differential is to place a copper backing block (Fig 6) against the thin member during fusion welding (Ref 9) The block serves as a chill, or heat sink, for the thin member The block can

be beveled along one edge so that it can be used when horizontal fillet welds are deposited on both sides of a thin member Copper backing bars or strips are made in a variety of shapes and sizes to dissipate heat as needed Often some experimentation and proof testing are required to obtain the optimum backing location and design Another way to obtain equalized heating and smooth transfer of stress where unequal section thicknesses are being welded is to taper one or both members to obtain an equal width or thickness at the joint Commonly, when two pipes of dissimilar internal diameter and wall thickness are to be joined, a convenient way is to introduce a "reducer" between the two pipes One end of the reducer will have the same size and wall thickness as the larger pipe, while the other end of the reducer will have the same size and wall thickness as the smaller pipe

Fig 6 Use of copper backing bar as a chill to minimize heat sink differential Source: Ref 9

Distortion Control. Design of an appropriate weld joint can also help reduce welding-related distortion Fusion welding processes employ localized melting and solidification to join component parts, which can result in excessive thermal strains These thermal strains are dependent on the type of material, the welding process, and the welding procedure Thermal strains produced by fusion welding processes can cause residual stresses and distortion, leading to

transverse and angular shrinkage Reducing the overall length of the weld or the amount of weld metal that needs to be deposited to complete a joint reduces both residual stresses and distortion For example, intermittent welding instead of continuous welding reduces the overall length of a weld Similarly, the use of a double-V groove instead of a single-V groove results in the reduction of the amount of weld metal and minimizes transverse shrinkage (Ref 10) Further, the amount of angular shrinkage is strongly influenced by the ratio of the weld metal in the top and the bottom sides of the plate To minimize the out-of-plane distortion in fillet welded joints, efforts should be directed to using the minimum size

of the welds that is consistent with strength considerations

References cited in this section

1 O.W Blodgett, Joint Design and Preparation, Welding, Brazing, and Soldering, Vol 6, 9th ed., Metals Handbook, American Society for Metals, 1983, p 60-72

8 The Brazing Book, Handy & Harmon, 1983, p 10

9 G.L Serangeli, et al., Shielded Metal Arc Welding, Welding, Brazing, and Soldering, Vol 6, 9th ed., Metals Handbook, American Society for Metals, 1983, p 91

10 K Masubuchi, Residual Stresses and Distortion, Welding, Brazing, and Soldering, Vol 6, 9th ed., Metals Handbook, American Society for Metals, 1983, p 887

Design for Joining

K Sampath, Concurrent Technologies Corporation

Case Histories

The case histories in this section illustrate the selection of design for joining, based on one or more of the good design

practices discussed above The case histories were compiled from Welding, Brazing, and Soldering, Volume 6 of the 9th Edition of Metals Handbook The examples highlight the application of industrial engineering principles to design

practices pertaining to fusion welding, diffusion welding, and brazing as they are widely used in the manufacture of components The basic principles of these design practices are also applicable to other methods of joining such as mechanical fastening, adhesive bonding, soldering, and so on

Example 1: Process Selection Obviates the Need for Joint Preparation

Bulldozer blades (Ref 11) were assembled from several low-carbon steel components that had relatively thick sections, generally 13 mm ( in.) or more (Fig 7) Most welds were 6.4 to 13 mm ( to in.) fillet welds A few were groove welds

Auxiliary gas shielded FCAW

Weld size

6.35 to 12.7 mm ( to in.)

For 6.35 to 7.94 mm ( and in.), flat position 1

For 2.03 mm ( in.), horizontal position 2

For 0.44 and 12.7 mm ( and in.), flat position

Fig 7 Bulldozer blade and comparison of joint penetration (and actual throat depth) of fillet welds made by

shielded metal arc welding and by auxiliary gas shielded flux cored arc welding (FCAW) Low-carbon base metal; low-carbon steel filler metal Source: Ref 11

To produce fillet welds involving thicker plates, flux cored arc welding (FCAW) was selected in preference to shielded metal arc welding (SMAW), for three reasons: deeper joint penetration, permitting the use of smaller fillets without decreasing the strength of the joint; higher deposition rate; and greater visibility of the arc to the welder, resulting in a better weld The difference in joint penetration for the two processes is shown in Fig 7 Use of SMAW to achieve the same level of penetration as FCAW would have required beveling the edge of the vertical member, or an additional number of weld passes Details for FCAW are given in the table accompanying Fig 7

Example 2: Two-Tier Welding

Electron beam welding provided a unique solution to the problem of designing and fabricating a component of an aircraft gas turbine engine (Ref 12) As indicated in Fig 8, the component consisted of a cylinder with an external flange on one end, an internal flange on the other, and a tubular annulus between The components were assembled by welding the trough shape ends of the two subcomponent cylinders by a single two-tier circumferential weld

Joint type Circumferential, two-tier butt

Machine capacity 150 kV at 40 mA

Maximum vacuum 1.3 ×10-3 Pa (10-5 torr)

Fixtures Bolted end plates; rotating positioner

Welding power 125 kV at 9.3 mA

Welding vacuum 1.3 × 10-2 Pa (10-4 torr) minimum

Working distance 31 cm (12 in.)

Beam focal point Midway between tiers(a)

Welding speed (at 0.42 rpm):

Upper tier 76 cm/min (30 in./min)

Lower tier 73 cm/min (28.7 in./min)

Beam oscillation None

Number of passes One, plus 30° downslope

Postweld heat treatment Aged 16 h at 760 °C (1400 °F)

(a) Indicated machine setting See text of example for

explanation of beam focal point

Fig 8 Tiered welds made simultaneously using the electron beam welding process Source: Ref 12

The components were made of René 41, for service at elevated temperature The chief welding objectives were to obtain sound welds and to avoid distortion of the part, especially the alignment of the 288 holes located in the annulus less than

13 mm ( in.) from the joints The holes had to be drilled before welding because they could not be deburred if drilled after welding

Arc welding was rejected as a joining method because it would have been necessary to use internal chills with gas backing in the annulus to minimize distortion and avoid atmospheric contamination Electron beam welding not only met the basic requirements but also made both joints simultaneously, even though the two joints were separated by approximately 13 mm ( in.), as shown in detail A of Fig 8

Fixturing was relatively simple The joints were accurately machined square and the components were assembled between two aluminum plates fitted over the flanges The plates were connected and forced together by bolts located inside the inner flange This fixture was then mounted on the faceplate of a welding positioner in the vacuum chamber so that the part would rotate with its axis horizontal The electron beam gun was in a fixed, overhead position

Success of the two-tier welding operation depended on careful control of part alignment, beam alignment, beam focal point, power adjustment, and travel speed The joints for the upper-tier and the lower-tier welds had to rotate in the same vertical plane, although by direct viewing they could not be observed simultaneously In addition, beam impingement on the joint of the lower-tier weld could be verified only by emergence of the weld bead from the underside, or by sectioning

of the test pieces

Alignment of the part for true horizontal-axis rotation was done with the aid of a precision level (sensitivity of 0.0004 mm per centimeter, or 0.0005 in per foot) and precision spacer blocks placed on the face of the flange (62 cm, or 24.5 in OD) Beam alignment was done by centering the beam spot on the reticle of the scope and moving the joint to this position Beam focal point, beam power, and travel speed were adjusted by trial and error on test components until a satisfactory welding procedure was established The final settings are shown in the table with Fig 8 By adjusting the beam focus for an indicated setting midway between the two tiers, the weld shapes of Fig 8, details B and C, were obtained

The mushroom-head shape of the upper-tier weld was caused by the defocused condition of the beam at that point, while the somewhat oversize root reinforcement resulted from the excess of power needed to penetrate to and through the lower tier The relatively narrow face of the lower-tier weld, as well as the narrowing of the weld in its progress through the joints, was explained as an effect of a charged plasma that surrounded and refocused the beam on its passage through the material The plasma, having a net negative charge, repelled the beam electrons, causing the beam to constrict and to change its focal point The net result on the lower tier was to produce a weld closely approaching the contour of a normal

single-thickness weld made with a tight, surface-focused beam Thus, the indicated focal-point setting was more virtual than real

Both welds were satisfactory as to soundness and shape Because of lack of access to the interior of the joint, weld spatter and undercutting were of concern, and the spatter associated with penetration of the upper tier was a problem Most of the particles were loosened with pipe cleaners and were flushed out with solvent at high pressure The few small particles that remained were judged to be acceptable after radiographic examination Undercutting was not a problem The René 41 material was capable of withstanding considerable excess beam power, which was especially important in making the upper-tier weld

Components produced using the two-tier welding procedure met all test requirements

Example 3: Design for Diffusion Bonding

Design for diffusion bonding must facilitate intimate contact of parts and local (microscopic) plastic deformation to promote the formation of a joint The following example illustrates the use of an innovative joint design that exploits the differences in thermal expansion between parts and tooling (the tooling material exhibits a higher strength than the part material at the bonding temperature) to promote intimate contact, incrementally increase pressure at the joint interface, cause localized plastic deformation, and thereby produce a diffusion bond

Figure 9 shows diffusion bonding of a titanium part using tooling blocks and spacers of 22-4-9 stainless steel (Ref 13) Initially, the parts and the tooling are fitted into a welded retort made of 1.6 mm (0.063 gage) muffler steel and conforming to the shape of the part The retort contains an end rail of 22-4-9 spanning the entire width This end rail contains machined grooves that allow the air to escape when a vacuum pump is turned on Similar 7.6 cm (3 in.) thick, 22-4-9 plates line the bottom, walls, and opposite end of the retort, and one covers the filled retort before a lid is welded

to the retort to seal the container and make it leakproof To prevent sticking to titanium, all 22-4-9 tooling is surface oxidized by oven baking at 760 °C (1400 °F) for 4 h During the loading of the retort, titanium slip shims are inserted to separate the tooling blocks slightly from the titanium parts Later, these shims are removed to create a vacuum path for the air to escape during evacuation of the retort A steel tube is used to connect the retort with a vacuum pump

Fig 9 Processing sequence during diffusion bonding of a titanium part using stainless steel tooling Source: Ref

13

Initially, the container is evacuated to 0.13 Pa (10 torr) vacuum and checked for leaks A higher vacuum, below 1.3 × 10

4

Pa (10-6 torr), is then obtained The leakproof container is placed within reusable ceramic blocks, which in turn are covered by steel plates containing thermocouples The entire assembly is placed in a bonding press The ceramic blocks that press against the top and bottom of the retort contain heating coils that bring the entire assembly to about 927 °C (1700 °F) The ceramic blocks on the sides and ends of the retort transmit heat and pressure to the assembly during bonding A press is used to apply about 13.8 MPa (2000 psi) pressure on the retort in all directions, and the retort is held

at the bonding temperature and pressure for about 2 to 12 h Thermal expansion of the titanium part against the relatively rigid stainless steel tooling allows intimate contact of the titanium parts across the joint line, and it facilitates localized plastic deformation and the formation of a diffusion bond Following the bonding cycle, the entire assembly is cooled slowly and dismantled, and then the retort is cut open to retrieve the diffusion-bonded titanium part Generally, the tooling

is reused, while the retort (made from cheap muffler steel) is scrapped

Example 4: Revision of Joint Design to Reduce Cost

The longitudinal butt joints in 6.1 m (20 ft) long sections of SA-106, grade B carbon steel pipe (Ref 14) used for boiler headers were originally designed as shown at lower left in Fig 10 With this design, the root pass and the second pass were made by SMAW using a backing bar, and then the weld was completed by submerged arc welding (SAW)

power-Welding conditions for improved design

Filler metal (low-carbon steel)

GTAW (argon shielding) ER70S-G consumable insert

Interpass temperature 260 °C (500 °F) max

Postheat Stress relieve at 621±25 °C (1150±25 °F)(c)

With the original joint design, the root and second passes were made by SMAW using a backing bar Weld was then completed by SAW With the improved joint design, the root pass was made by GTAW using a consumable insert (no backing), the second pass by SMAW, and the remainder by SAW

(b) Weld head on boom-type manipulator, workpiece supported on power and idler rolls for turning

Fig 10 Revision of joint design Use of a consumable insert permitted change to a lower-cost method of

welding boiler-header pipes Carbon steel (SA-106, grade B; 0.30 max C) base metal; low-carbon steel filler metal GTAW, gas-tungsten arc welding; SMAW, shielded metal arc welding; SAW, submerged arc welding Source: Ref 14

To reduce the cost, the joint design was revised to that shown at lower right in Fig 10 This permitted making the root pass by gas-tungsten arc welding (GTAW), using a consumable insert, instead of by SMAW with a backing bar Then, as with the original joint design, the second pass was made by SMAW, and the weld was completed by SAW The SMAW process was used for the second pass to provide a deposit thick enough to ensure against melt-through by SAW The improved joint design and change in welding procedure resulted in a 25% saving in cost (material, labor, and overhead) per foot of seam welded

Example 5: Submerged Arc Welding of a Large Piston

The large hydraulic-jack piston shown in Fig 11 was assembled by welding three low-carbon steel castings (head, piston body, and seat) at girth joints (Ref 15) When similar smaller pistons with wall thicknesses of 7.6 to 12.7 cm (3 to 5 in.) had been assembled by SMAW, about one welded joint in eight was found to be defective and had to be reworked

Conditions for SAW

Joint type Circumferential butt

Weld type Single-U-groove, integral backing

Joint preparation Machining

Power supply 1000 A transformer

Wire feed Fully automatic, constant speed

Welding head Machine held, air cooled

Fixture 50 ton variable-speed roll

Auxiliary equipment Exhaust fan, vacuum flux remover, positioning arm

Current and voltage:

Passes 1 through 3 700 A, ac; 38 V

Remaining passes 750 A, ac; 40 V

Preheat 204 °C (400 °F) (by torch)

Postheat (stress relief) 7 h at 600 °C (1115 °F), furnace cool to 315 °C (600 °F)

Welding speed

24 cm (9 in.) per min (a) Electrode and flux yielded a weld deposit containing 0.12 C,

0.84 Mn, 0.72 Si, 0.018 S

Fig 11 Large piston assembled by submerged arc welding (SAW) Low-carbon steel base metal; low-carbon

steel filler metal (EL12) Source: Ref 15

Because of the experience with the pistons with 7.6 to 12.7 cm (3 to 5 in.) wall, it was decided to use SAW to assemble four large pistons, in which a 19.2 cm (8 in.) wall was to be joined to a 16.8 cm (6 in.) wall, using the joint design shown in detail A The outside surfaces of the three castings to be welded were rough machined and the joints were prepared by machining The joints were of the interlocking type (see Fig 11, detail A) and provided support for the unwelded components during positioning on variable-speed welding rolls Joint areas were preheated to 204 °C (400 °F) with the gas torches as the piston was rotated The welds were made in 380 passes and were produced oversize and machined to size after magnetic-particle inspection and stress-relief The welded pistons were stress relieved at 600 °C (1115 °F) for 7 h and furnace cooled to 315 °C (600 °F)

Each welded joint was ultrasonically inspected for a distance of 7.6 cm (3 in.) on each side of the weld After inspection, the pistons were hydrostatically tested at 2 MPa (300 psi) There were no rejections Production time for welding the large piston was 101 h, which was a considerable improvement over the production time of 212 h for the smaller pistons assembled by SMAW

Example 6: Use of an Offset to Eliminate Backing Rings

A component of a heat-exchanger shell assembly (Ref 15) was initially made by SAW: a medium-carbon steel pipe cap with a wall thickness of 6.4 mm ( in.) was attached to a low-carbon steel pipe of the same wall thickness by means of a circumferential butt joint, supported and aligned by a backing ring, as shown in the "original design" in Fig 12

Joint type Joggled lap

Weld type, original design Square-groove, with backing ring

Weld type, improved design Modified single-V-groove, with integral backing

Joint preparation:

Original design Backing ring machined

Improved design Cap end machined, pipe end reduced

Welding current 350 to 410 A (DCEN)

Welding speed 46 to 51 mm (18 to 20 in.) per min

Number of passes, original design Three

Number of passes, improved design Two

Power supply 40 V, 600 A transformer-rectifier (constant-voltage)

Fixturing Chuck-type turning rolls; alignment clamps for tack welding

Fig 12 Cap-to-pipe weldment Low-carbon steel welded to medium-carbon steel; low-carbon steel filler metal

(EL12) Source: Ref 15

When it became apparent that the wall thickness of the pipe cap could be less than that of the pipe without adversely affecting service performance, the joint was rede-signed as a joggled lap joint (see "improved design," Fig 12) The offset incorporated in the pipe for the redesigned joint took the place of the backing ring previously used and furnished a locating surface for the cap The redesigned joint was made by SAW under the same conditions as those for the original joint, except that only two passes were required, rather than three

Cost reduction was realized from eliminating the backing ring, from the savings in material resulting from the use of thinner pipe caps, and from eliminating one circumferential welding pass The change in joint design led to a savings of approximately 35% in total factory cost All joints were inspected visually and radiographically to check for full penetration and absence of slag inclusions The rejection rate was less than 1%

Example 7: Elimination of Backing Bars

A 3.7 m (12 ft) long header assembly for a large high-pressure heat exchanger (Ref 15) was manufactured to Section VIII, Division I, of the ASME Boiler and Pressure Vessel Code (Fig 13)

Electrode

2.4 mm ( in.) diam flux cored wire 2.4 mm ( in.) diam solid wire

Root-pass welding conditions

Welding speed, cm/min (in./min) 15 (6) 20 (8)

Filler-pass welding conditions

Deposition rate, kg/h (lb/h) 2.7 (6) 8.2 (18)

Welding speed, cm/min (in./min) 25 (10) 56 (22)(b)

(a) Power supply for welding of both designs was an 80 V (open-circuit)

transformer-rectifier

(b) Welding speed for the first filler pass was 71 cm/min (28 in./min)

Fig 13 Submerged arc welding (SAW) setup for heat-exchanger header Carbon steel, 0.35% max C (ASTM A

515, grade 70) base metal; carbon steel filler metals FCAW, flux cored arc welding Source: Ref 15

As originally designed (see upper left in Fig 13), for manual FCAW from the outside, the assembly consisted of four steel components and was welded at corner joints that incorporated backing bars, as shown in Fig 13, section A-A It was difficult to ensure a uniformly tight fit of the backing in the joint Under radiographic inspection, slag was revealed that ran between the backing bars and the adjacent 38 mm (1 in.) thick components

The problem was eliminated by redesigning the header assembly for automatic SAW without backing bars The redesigned assembly, shown in the upper right of Fig 13, consisted of two 38 mm (1 in.) thick channels formed in a press brake The two components were welded at two longitudinal butt joints of the double-V-groove design (Fig 13, section B-B)

For this improved design, the welding was done by the use of a boom-mounted automatic welding head The formed channels were held stationary while the welding head was advanced along the joints First, root passes were made along the inside grooves of the two joints, then filler passes were made along the outside grooves After the root passes, the outside grooves were machined-out to sound metal before the filler passes were begun

A major benefit of the change to two-piece design was that only about one-third as many filler passes were required for the entire weldment (10 passes, as compared with 28 to 32 passes for the original four-piece design)

Example 8: Use of Modified Butt Joint to Save Tooling and Labor Costs

An offset lap joint (middle view in Fig 14) is frequently used in welding the components of a variety of pressure cylinders and spheres However, the use of this lap joint for welding the spherical refrigerant container shown in Fig 14 would have resulted in high labor and tooling costs The lip of the offset hemisphere would have caused interference in assembly, and additional tooling would have been needed to offset the lip

Joint type Circumferential modified butt Weld type Single-flare V-groove

Power supply 300-A transformer-rectifier

Electrode wire (a) 0.162 mm (0.030 in.) diam ER70S-3

Welding gun Mechanized, fixed, water cooled

Wire feed Push-type motor, on welding gun

Welding speed 118 cm (46.6 in.) per min

Weld time per container 42 s

(a) Selection of wire wound to a large diameter

eliminated need for wire straighteners and reduced leakage rate

(b) Argon of 99.999% purity from bulk-liquid holder

Fig 14 Girth welded refrigerant container Labor and tooling costs were reduced by use of a modified butt joint

instead of an offset lap joint Low-carbon steel (ASTM A 620) 0.045 in base metal; low-carbon steel filler metal (ER70S-3) Source: Ref 16

The modified butt joint shown at the bottom in Fig 14 allowed use of identical forming tools for both halves of the sphere and was the best compromise among weldability, tooling costs, and labor costs (Ref 16) To reduce labor costs still further, each welder operated two girth welding machines and thus was not able to observe the welding operation; therefore, automatic seam tracking was necessary

The automatic seam tracking system consisted of two recirculating ball-screw cross-slides mounted at right angles and driven by reversible alternating current motors A probe, which was mounted to move with the welding gun (see view at upper left in Fig 14), sensed the location of the joint in relation to the gun A movement of the probe tip caused the appropriate slide to bring the probe and gun back to the neutral position The probe was mounted on a small screw-adjusted slide to provide quick and accurate adjustment of both the horizontal position of the welding gun and the distance between it and the work At the end of the welding cycle, the probe and the welding gun were raised by the vertical slide After the next assembly was in place, the probe and gun were lowered by the same means and welding was started automatically

The hemispheres were held in a special welding machine (lathe) consisting of one fixture rotated by a continuously variable drive and a second fixture mounted on the tailstock Both fixtures were mounted on air-operated slides Thus, the parts were held together and rotated under the welding gun When the gun was retracted after completing the weld, the air cylinders separated, releasing the welded workpiece

The operator loaded the hemispheres in the machine and pushed a button to close the fixtures The operator then had an option of using automatic start, whereby the weld started as soon as the welding gun was in position, or manual start, whereby the position of the gun could be observed, corrections could be made (if required), and the weld could be started

by pushing a button While one container was being welded, the operator loaded and started a second machine After each container had been welded, the operator checked it for visible defects The containers requiring repair welding were set aside, and those with no visible defects were transferred on a conveyor to a testing area

The workpiece was a disposable refrigerant container with a water capacity of 11.6 kg (25.5 lb), produced under a special permit that specified two types of pressure tests Each welded container was tested by subjecting it to 2.1 MPa (300 psi) internal air pressure while within a heavy steel safety chamber The pressure in the container was then reduced to 0.7 MPa (100 psi), the chamber was opened, and the sphere was forced under water to check for leaks If repairs were required, the spheres were resettled after repair A destructive test was required on one container out of each lot of 1000, with a minimum of one per day (although, in practice, at least one container was tested from each machine during each shift) This test consisted of filling the container with water, connecting it to a high-pressure pump, and increasing the pressure until the container burst The minimum bursting strength was 5.5 MPa (800 psi) Fewer than 5% of the spheres required weld repairs

The guidance system caused some problems, primarily because of the maintenance required Repair and adjustment of the probe switch were difficult, and improperly adjusted probes sometimes caused misplaced welds Spare systems were available for replacement of defective probe units

Although the guidance system added to the machine cost and caused maintenance problems, these disadvantages were soon canceled out by decreased welding costs Satisfactory welds were difficult to produce manually, because the horizontal variance of the welding gun position had to be held to 0.8 mm ( in.) to prevent melt-through In addition, it would have been necessary for the manual operator to correct for differences in the heights of the weld seams, limiting the welder to operating one machine and thus doubling the labor cost

The hemispheres were press formed and vapor degreased No edge preparation or postweld finishing was done Welding conditions are given in the table with Fig 14

Example 9: Joining Sections of Unequal Thickness

An application involving components of unequal section thicknesses (Ref 9) is the welding of heat-exchanger tubes having 2.4 mm (0.093 in.) wall thickness to a tube sheet as thick as 25.4 cm (10 in.) The usual method of avoiding difficulty is to cut a circular groove, 6.4 mm ( in.) deep, in the upper surface of the tube (Fig 15) By restricting heat transfer, this groove minimizes heat sink differential between the thin tube wall and thick tube sheet

Fig 15 Thick tube sheet with machined groove Minimizes heat sink differential during welding of thin-walled

heat-exchanger tube to the tube sheet Low-carbon steel base metal; low-carbon steel filler metal Source: Ref

9

Example 10: Redesign of a Joint to Improve Dimensional Control

The corner-welded channel sections shown in Fig 16 were parts of rectangular frames for data-processing machines (Ref 17) Originally, 45° miter joints were used (Fig 16a), but dimensions after welding were unsatisfactory because of joint location and weld restraint

Joint type Corner

Weld type Fillet and V

Power supply 200 A, constant-voltage rectifier

Electrode wire (a) 0.76 mm (0.030 in.) diam ER70S-2

Wire feed Constant feed

Current 80-85 A (DCEN)

Voltage 26 V

Shielding gas 75% argon-25% carbon dioxide, at 1.1 m3/h (40 ft3/h)

Wire-feed rate 76-254 cm (30-100 in.) per min

Welding speed 25 cm (10 in.) per min

(a) 0.04% C; triple deoxidized, with flash copper plating

Fig 16 Corner section of a rectangular frame A cope joint was substituted for a miter joint to improve

dimensional control Low-carbon steel base metal; low-carbon steel filler metal (ER70S-2) Source: Ref 17

To provide a more positive joint location with less weld restraint, cope joints (Fig 16b) were substituted for the miter joints Tolerances of ±0.25 mm (±0.010 in.) on length and width, and ±0.813 mm (±0.032 in.) on squareness, were met on channel sections welded with the improved joint design

The channel sections were contour roll formed from 3.05 mm (0.120 in.) thick low-carbon steel strip Pieces were cut to length by a cutoff die in a press The cut lengths were also coped by a die in a press, and all parts were inspected Tolerances on individual pieces were held to ±0.127 mm (±0.005 in.)

Example 11: Change in Joint Design to Reduce Distortion and Cost

Figure 17 shows a 305 cm (120 in.) long steam-drum shell course, roll formed with a welded longitudinal seam (Ref 15) Originally, the butt joint for this seam was of single-V-groove design and was welded with the use of a backing strip (see

"original design" in section A-A in Fig 17) Fit-up and removal of the backing strips were time-consuming operations, and welding from one side distorted the weldment

Welding conditions for both joint designs

Weight of electrode and flux deposited per hour 7.1 kg (15.6 lb)

Weight of electrode and flux deposited per foot of weld:

Original design (single-V-groove) 11.8 kg (26 lb)

Improved design (double-V-groove) 6.4 kg (14 lb)

Length of weld (including runoff tabs at ends)

3.2 m (10 ft)

Time for installing and removing backing strip (original design) 12 h

Weld types Single-V-groove (original); double-V-groove (improved)

Postheat 621±25 °C (1150±25 °F) (furnace), 1 h/25 mm (1 in.) of section

Electrode

4.8 mm ( in.) E7018

Intermediate passes (single-electrode SAW):

Electrode wire

5.6 mm ( in.) diam 0.5% Mo steel(b)

Electrode wire

5.6 mm ( in.) diam 0.5% Mo steel(b)

Electrode wire

2.4 mm ( in.) diam 0.5% Mo steel(b)

(a) Workpiece supported on one power roll and one idler roll

(b) Electrode wire contained 0.11% C, 0.50% Mo, 0.85% Mn, and was used at a 1-to-1 ratio of wire to

flux

(c) Tandem welding head was mounted on a boom-type manipulator

Fig 17 Submerged arc welding setup for steam-drum shell course Based metal: carbon steel, 0.35% max C

(ASTM A 515, grade 70), normalized Filler metals: low-carbon steel (E7018) for root passes (shielded metal arc welding); 0.5% Mo steel for remaining passes (submerged arc welding) Source: Ref 15

The joint was changed to a double-V-groove design (shown as "improved design" in section A-A of Fig 17) This change resulted in the need for much less weld metal; the need for a backing strip was eliminated; and distortion was reduced by sequential deposition of weld beads on the inside and outside of the joint The amount of back gouging needed was less than that required to remove the backing strip from the single-V-groove weld As a result of these improvements, electrode, flux, and labor costs were reduced by 46%, and the total cost of welding was reduced by 62% Welding procedures and post-weld operations for the two designs are described below For both designs, the shell courses were hot roll formed into a cylinder and descaled, and the joint grooves were flame cut

Originally, the single-V-groove joint was preheated to 79 °C (175 °F) with a propane torch, the backing strip was installed, and the temperature of the joint was raised to 121 °C (250 °F) At least two root passes were made, using SMAW This operation was followed by depositing six single-pass layers, each 3.2 mm ( in.) thick, by single-electrode SAW Tandem SAW was used to complete the weld, single-pass layers 3.2 mm ( in.) thick being deposited to a weld level of 38 mm (1 in.), followed by two-pass (split) layers 3.2 mm ( in.) thick Then the backing strip was removed by air carbon arc gouging and grinding, and back welding was done, as required, to provide a flush joint

In the improved design, the double-V-groove joint was also preheated in two stages (79 and 121 °C, or 175 and 250 °F) with a propane torch, except that instead of a backing strip being installed between stages, a spacer rod of 6.4 mm ( in.) diameter 0.5% Mo steel electrode material was tacked in place and seal welded by SMAW Shielded metal arc welding was used also for root passes The first increment of single-electrode submerged arc welds consisted of eight 3.2 mm (in.) thick single-pass welds on the outside of the weldment The workpiece was rotated 180°, and the joint was back gouged and ground to a radius of 6.4 to 9.5 mm ( to in.) The first increment of welding on the inside of the joint consisted of 3.2 mm ( in.) thick single-pass welds to a 38 mm (1 in.) level, using single-electrode SAW Then the workpiece was again rotated 180°, and the remainder of the outside welding was completed using tandem SAW to deposit two-pass (split) layers of 3.2 mm ( in.) thickness After a final 180° rotation of the workpiece, the inside welding was completed using the same sequence of two-pass tandem SAW of 3.2 mm ( in.) thickness

References cited in this section

9 G.L Serangeli, et al., Shielded Metal Arc Welding, Welding, Brazing, and Soldering, Vol 6, 9th ed., Metals Handbook, American Society for Metals, 1983, p 91

11 G.L Serangeli, et al., Flux Cored Arc Welding, Welding, Brazing, and Soldering, Vol 6, 9th ed., Metals Handbook, American Society for Metals, 1983, p 108

12 E.A Metzbower, et al., Electron Beam Welding, Welding, Brazing, and Soldering, Vol 6, 9th ed., Metals Handbook, American Society for Metals, 1983, p 609-646

13 S Bangs, Diffusion Bonding: No Longer a Mysterious Process, Source Book on Innovative Welding Processes, American Society for Metals, 1981, p 259-262

14 D Hauser, et al., Gas Tungsten Arc Welding, Welding, Brazing, and Soldering, Vol 6, 9th ed., Metals Handbook, American Society for Metals, 1983, p 202-203

15 D.L Olson, et al., Submerged Arc Welding, Welding, Brazing, and Soldering, Vol 6, 9th ed., Metals Handbook, American Society for Metals, 1983, p 114-152

16 D Hauser, et al., Gas Metal Arc Welding (MIG Welding), Welding, Brazing, and Soldering, Vol 6, 9th ed., Metals Handbook, American Society for Metals, 1983, p 165

17 D Hauser, et al., Gas Metal Arc Welding (MIG Welding), Welding, Brazing, and Soldering, Vol 6, 9th ed., Metals Handbook, American Society for Metals, 1983, p 166

Design for Joining

K Sampath, Concurrent Technologies Corporation

Future Directions

The foregoing examples illustrate that value engineering, methods study, and time study principles can be applied to select the best design for joining of parts Future efforts could be directed toward developing computer-based simulations with graphic user interfaces that would integrate appropriate part design and manufacturing databases Such efforts would

allow one to effectively consolidate existing knowledge on basic design practices, design criteria for joining, and appropriate case examples involving parts and processes These computer-based simulations can serve as powerful learning tools, and their effective use can be expected to eliminate or minimize trial-and-error methods of design for joining, and thereby facilitate agile manufacturing at minimal cost

Design for Joining

K Sampath, Concurrent Technologies Corporation

8 The Brazing Book, Handy & Harmon, 1983, p 10

9 G.L Serangeli, et al., Shielded Metal Arc Welding, Welding, Brazing, and Soldering, Vol 6, 9th ed., Metals Handbook, American Society for Metals, 1983, p 91

10 K Masubuchi, Residual Stresses and Distortion, Welding, Brazing, and Soldering, Vol 6, 9th ed., Metals Handbook, American Society for Metals, 1983, p 887

11 G.L Serangeli, et al., Flux Cored Arc Welding, Welding, Brazing, and Soldering, Vol 6, 9th ed., Metals Handbook, American Society for Metals, 1983, p 108

12 E.A Metzbower, et al., Electron Beam Welding, Welding, Brazing, and Soldering, Vol 6, 9th ed., Metals Handbook, American Society for Metals, 1983, p 609-646

13 S Bangs, Diffusion Bonding: No Longer a Mysterious Process, Source Book on Innovative Welding Processes, American Society for Metals, 1981, p 259-262

14 D Hauser, et al., Gas Tungsten Arc Welding, Welding, Brazing, and Soldering, Vol 6, 9th ed., Metals Handbook, American Society for Metals, 1983, p 202-203

15 D.L Olson, et al., Submerged Arc Welding, Welding, Brazing, and Soldering, Vol 6, 9th ed., Metals Handbook, American Society for Metals, 1983, p 114-152

16 D Hauser, et al., Gas Metal Arc Welding (MIG Welding), Welding, Brazing, and Soldering, Vol 6, 9th ed., Metals Handbook, American Society for Metals, 1983, p 165

17 D Hauser, et al., Gas Metal Arc Welding (MIG Welding), Welding, Brazing, and Soldering, Vol 6, 9th ed., Metals Handbook, American Society for Metals, 1983, p 166

Design for Joining

K Sampath, Concurrent Technologies Corporation

Selected References

• Brazing Handbook, 4th ed., American Welding Society, 1991

• P.W Marshall, Design of Welded Tubular Connections: Basis and Use of AWS Code Provisions,

Elsevier, 1992

• R.C Juvinall and K.M Marshek, Rivets, Welding, and Bonding, Chapter 11, Fundamentals of

Machine Component Design, 2nd ed., John Wiley & Sons, 1991

• R.O Parmley, Ed., Standard Handbook of Fastening & Joining, 3rd ed., McGraw-Hill, 1997

• D Radaj, Design and Analysis of Fatigue Resistant Welded Structures, Halsted Press/Woodhead

Publishing, 1990

• M.M Schwartz, Brazing, ASM International, 1987

• M.M Schwartz, Ceramic Joining, ASM International, 1990

• M.M Schwartz, Joining of Composite Matrix Materials, ASM International, 1994

• J.E Shigley and C.R Mischke, Welded, Brazed, and Bonded Joints, Chapter 9, Mechanical

Engineering Design, 5th ed., McGraw-Hill, 1989

• Weld Integrity and Performance, ASM International, 1997

Design for Heat Treatment

William E Dowling, Jr and Nagendra Palle, Ford Motor Company

Introduction

THE SELECTION OF MATERIALS and manufacturing processes for a component design is a complex process and often involves iterative decision making The component is designed to provide a specific mechanical function, and its design is often limited by space and cost considerations The component must be able to survive extreme external loading conditions from thermal and/or applied mechanical forces Therefore, a high level of performance needs to be achieved at

a minimum cost Based on the design and loading conditions, a material and manufacturing process are selected to cost effectively provide adequate properties for the operating environment Very often a low-cost material and processing combination requires heat treatment after component shaping to enable the part to meet its design criteria The relationship between performance at minimum cost, design, material, and the manufacturing process is analogous to a three-legged stool with all legs having equal importance In order for a component to fulfill its cost and performance criteria, its design should accommodate all the loading conditions of the component, its material properties should meet the expectations of the design, and its manufacturing processes should produce the component at a minimum cost

The mechanical design process has evolved from an experience-based process using design factors, such as concentration factors, and design rules (based on experimental data) to the current reliance on analytical processes The ability to analytically design components combined with the constant desire for system cost reduction while achieving greater performance has led to decreased design times and increased component complexity The designer, when provided with accurate component loading information, can accurately assess part life with reliable input of experimentally determined material properties These material property databases are continually growing, both in the open and in proprietary databases (see the article "Sources of Materials Properties Data and Information" in this Volume) In addition

stress-to these databases, the mechanical properties resulting from a broad range of heat treatment processes, for many classes of materials, are well documented The ability to select materials and process parameters to achieve the desired property goals is becoming more automated as evidenced by the success of Jominy hardness and carburizing prediction programs

(Ref 1) The foundation of the currently used design practices requires a close relationship among analytical design procedures, material property databases, and the ability of the heat treat process to achieve desired mechanical properties

In addition to providing appropriate physical and mechanical properties to meet design requirements, heat treatment also produces dimensional changes and residual stress patterns that in some cases can lead to component cracking The dimensional changes and residual stresses produced by heat treatment are very sensitive to geometric and processing specifics Currently, the relationship between design, dimensional changes, residual stresses, and cracking is determined

by experience and results in the development of general rules for design Prototype component designs must be experimentally evaluated and iterated to bring the component within acceptable tolerances for dimensions and residual stresses A clear need for reducing development costs has motivated significant corporate and academic research in the area of analytical prediction of the response of a component to heat treatment

This article presents an overview of techniques that are currently in use to design for heat treatment The primary design criteria addressed in this article are the minimization of distortion and undesirable residual stresses The article presents both theoretical and empirical guidelines to understand sources of common heat treat defects and how they can be controlled A simple example is presented to demonstrate how thermal and phase-transformation-induced strains cause dimensional changes and residual stresses This example also serves as a representation of a typical "process model." The final sections of the article describe the state-of-the-art in heat treatment process modeling technology

Reference

1 M.A.H Howes, Factors Affecting Distortion in Hardened Steel Components, Quenching and Distortion Control: Proceedings of the First International Conference on Quenching and Control of Distortion, ASM

International, 1992, p 251-258

Design for Heat Treatment

William E Dowling, Jr and Nagendra Palle, Ford Motor Company

Overview of Component Heat Treatment

Component heat treatment is often the most cost-effective method for a manufacturing process to produce the desired

material properties (Detailed information about heat treating processes is provided in Heat Treating, Volume 4 of ASM Handbook, Ref 2.) However, in addition to material strength, heat treatment can result in the development of residual

stresses (both compressive and tensile), dimensional changes (with respect to size and shape), and, in an extreme situation, component cracking, often referred to as quench cracking These factors (residual stresses and dimensional changes) have the greatest influence on the design process of a component Often, the inability to produce components with acceptable dimensions and residual stress patterns will cause changes in design, materials, and process selection leading to additional cost and lower material strength

A typical component manufacturing process includes the following five steps, all of which can influence dimensional changes and residual stress patterns in heat-treated components:

1 Metalworking, machining, or other forming operations

2 Component heat-up

3 Hold at temperature for through-heating, solutionizing, or thermal chemical treatments such as carburizing or nitriding

4 Quenching from elevated temperature

5 Postquench tempering or aging treatment

Within these five process steps there are seven major factors that lead to size and shape changes and the development of residual stresses in heat treated components (Ref 3, 4):

• Variation in structure and material composition throughout the component, leading to anisotropy in properties and transformation behavior

• Movement due to relief of residual stresses from prior machining and forming operations

• Creep of the part at elevated temperature under its own weight or as a result of fixturing

• Large differences in section size and asymmetric distribution of material causing differential heating and cooling during quenching

• Volume changes caused by phase transformation

• Nonuniform heat extraction from the part during quenching

• Thermal expansion

All of these factors, except relief of prior residual stresses (second item) and creep at elevated temperature (third item), can be directly related to thermal and transformation-induced strains in the component Residual stresses from forming operations can be reduced by stress relief prior to final shaping operations Creep at elevated temperature can be addressed by appropriate component loading in the furnace Neither of these two factors are directly associated with the design of the component The other five factors are directly related to component design, and more precisely, strains introduced by transformations and nonuniform temperature distributions in the component The relationship among design, material properties (yield strength at temperature, coefficient of expansion, thermal conductivity, etc.), and heating and cooling processes determine the distortion and residual stress patterns in heat treated components

References cited in this section

2 Heat Treating, Vol 4, ASM Handbook, ASM International, 1991

3 J.S Kirkaldi, Quantitative Prediction of Transformation Hardening in Steels, Heat Treating, Vol 4, ASM Handbook, ASM International, 1991, p 20-34

4 H Walton, Dimensional Changes During Hardening and Tempering of Through-Hardened Bearing Steels,

Quenching and Distortion Control: Proceedings of the First International Conference on Quenching and Control of Distortion, ASM International, 1992, p 265-273

Design for Heat Treatment

William E Dowling, Jr and Nagendra Palle, Ford Motor Company

Thermal and Transformation-Induced Strains in Heat Treated Components

Thermal strain is developed in a component when differential thermal expansion (or contraction) occurs The magnitude

of strain is directly proportional to the thermal expansion coefficient of the material ( ) and the temperature difference between two points ( T) This strain translates directly to a thermal stress th (Ref 5):

where E is the elastic modulus of the material However, if the thermally induced stress is greater than the flow strength

of either the cooler or hotter material, permanent (plastic) deformation occurs This plastic flow causes permanent shape change (distortion) and impacts the magnitude and distribution of residual stresses Without plastic deformation, the component would return to its original dimensions once the part has thermally equilibrated In addition to thermal strains, many materials systems undergo phase transformations as a function of temperature Often, the new phase(s) that form have different volumes and different coefficients of expansion as well as different mechanical behavior(s) than the parent phase(s) These differences increase the complexity of understanding the effect of thermal gradients on strains produced and the resulting plastic deformation

Example: Thermal and Transformation-Induced Strains in a Large, Thick Plate

This simplified example demonstrates how thermal and transformation-induced strains can result in substantial plastic deformation and residual stresses A large, thick plate (Fig 1a) has a very thin layer uniformly heated from room temperature to 850 °C and then cooled It is assumed that the bulk of the material is unheated and does not strain because

of its much greater thickness The heating and cooling processes can be broken down into 6 segments (shown schematically as A to F in Fig 2; key process points are labeled 1 to 7) During the heating and cooling cycle, the thin layer of heated material will undergo a phase transformation from phase 1 to phase 2 upon heating and from phase 2 to phase 3 upon cooling The thermal profiles through the sample for process point 4 is shown in Fig 1(b) to illustrate that the heated layer is considered to be at the same temperature all the way through, with a thermal step change from the heated layer to the unheated bulk

Fig 1 Schematic (a) of a large, thick plate and assumed temperature distribution (b) at process point 4 (Fig 2)

Fig 2 Strain history during heating and cooling for a large, thick plate See text for discussion of the labeled

process segments

The strain produced by thermal ( T = th) and transformation ( tr) strain is calculated for each segment identified in Fig 2 The total induced strain must be accommodated in the thin layer through either elastic ( el) or plastic ( pl) strain, which sums to the total strain:

The induced strain and the accommodation strain must equalize in the thin layer because the assumption has been made that the bulk material will not accommodate any strain The values for (thermal expansion coefficient) for phases 1, 2, and 3 are chosen to be 15, 21, and 13 × 10-6/°C, respectively, and are assumed to be independent of temperature for this example The transformation strains chosen are tr 1-2 = 0.0025 for the transformation from phase 1 to phase 2 at 725 °C and tr 2-3 = 0.0075 for the transformation from phase 2 to phase 3 over the temperature range 300 to 150 °C

In order to determine the accommodation strain values, Young's modulus (E), and the yield strength ( ys) are required as

a function of phase and temperature; the data for this example are shown in Fig 3 For simplicity, the flow behavior is assumed to be elastic-plastic in nature with no work hardening Thus, the yield strength is the flow stress at any plastic strain level