The application of electron beam welding for the ... - Science Direct

$ouraal~

ELSEVIER

Journal of Materials Processing Technology 59 (1996) 257 -267

Materials Processing Technology

The application of electron beam welding for the joining of dissimilar metals: an overview Z. Sun

R. Karppi b

a GINTIC Institute of Manufacturing Technology, Nalo,ang Avemw, Singapore 638075, S#~gapore b VTT Mamtfacturing Technology, Technical Research Centre of Finland (VTT), E~poo, Finland Received 2 February 1995; accepted 10 August 1995

Industrial summary

Electron beam welding (EBW) has been developed for many years and is being increasingly implemented in various industrial applications. Joining dissimilar metals using EBW has also been a subject of interest in recent years. Due to special features of EBW, e.g., high energy density and accurately controllable beam size and location, in many cases it has proven to be an efficient way of joining dissimilar metals. Numerous successful results have been achieved, and some of them have already been exploited in production. EBW continues to be the subject of investigations and further development, and improvements in the joining of dissimilar metals remains one of the aims. This paper reviews the state-of-the-art EBW of dissimilar metals, with special emphasis on showing the potential of the process for achieving high-quality dissimilar-metal joints. Since EBW is a fusion-welding process, metallurgical phenomena associated with fusion still exist and cause difficulties. However, these are often minor as cc,mpared to those in conventional arc welding. Problems encountered and possible solutions are discussed. The survey indicates that although many studies have been performed, there is still a considerable need to further examine existing and new combinations. Wherefore. future R&D trends are highlighted. Keywords: Electron beam welding; Dissimilar metals

1. Introduction

Dissimilar-metal joints are used widely in various industrial applications due to both technical and economic reasons. The adoption of dissimilar-metal combinations provides possibilities for the flexible design of the product by using each material efficiently, i.e., benefiting from the specific properties of each material in a functional way. Fusion welding is one of the most widely used methods for the joining of metals. Therefore, continuous efforts are made to apply these methods to the joining of dissimilar-metal combinations also, despite the many difficulties encountered. These difficulties include problems associated with metallurgical incompatibility, e.g., the formation of brittle phases, the segregation of high- and low-melting phases due to chemical mismatch, and possibly large residual stresses from the physical mismatch. There are several choices amongst the fusion welding processes, such as common conventional shielded metal arc, gas tungsten arc, gas metal arc, and submerged arc welding. They also include processes characterized by * Corresponding author. 0924-0136/96/$15.00 © 1996 Elsevier Science S.A. All rights reserved S S D I 0924-0136(95)02150-K

high energy density, such as plasma arc, electron beam, and laser beam welding. In addition to fusion welding, several other types of joining techniques are also available, and may often be associated with less difficulties for producing dissimilar-metal joints. These are pressure welding, e.g., friction, resistance and diffusion welding, as well as brazing and soldering, adhesive bonding, and mechanical joining. Most of these techniques can eliminate the fusion problems because the base metals remain in the solid state during joining. Therefore, they are better than fusion welding in this respect. However, the service conditions may make particular processes unsuitable, e.g., for high-temperature applications, soldering and adhesive bonding cannot be candidates, and for leak-tight joints, mechanical joining is not acceptable. Furthermore, the required joint geometry can make, e.g., friction welding difficult to apply. Diffusion welding often provides superior technical benefits for joining small dissimilar-metal parts, but the process is rather time consuming. Therefore, solutions relying on high energy density processes, e.g., electron beam welding (EBW) and laser beam welding (LBW), are still of great industrial interest. This was revealed in a recent survey on processes used for producing dissimilar-metal joints, Fig. 1. It is shown clearly that EBW has been

Z. Sun, R. Karppi /Journal of Materials Process#~g Technology 59 (1996)257-267

258

amongst the most frequently investigated processes in the joining of dissimilar metals during the last decade. This trend may be attributed to several factors: (i) the rapid development of EBW equipment; (ii) the high quality requirements of the products which EBW can realize; (iii) the special features of the process which offer solutions that other alternative methods cannot provide; and (iv)the economic benefits obtained by using the process in mass production. The obj~tive of the present paper is to review and discuss the EBW of dissimilar metals with the aim of showing the current potential, and highlighting the future needs of R&D efforts.

Etec'l:ron

beam gun

±

glos etec±rode

Focus;n9 colt

PM~

2. Characteristics of EBW

EBW, as an industrial welding process, was started in the late fifties. The process v.as used primarily in the nuclear, aircraft and aerospace industries, initially due to the requirements of high quality and reliability of the joints [1]. EBW has, subsequently, also been used successfully in various applications, such as medium-thickness welding in workshops, and the welding of high-precision parts. More recently, the welding of heavy engineering parts was realized with high power levels. In all of these phases, EBW of dissimilar metals was frequently appended to investigations due to the progress in equipment, and its ability to perform various welding tasks. The principle of an EBW operation is to use the kinetic energy of electrons as the heating source to melt the metals to be joined. These electrons are generated by heating a negatively-charged filament (cathode) to its thermionie emission temperature range, upon which electrons are emitted, Such electrons, being accelerated by the electric field between the negatively-charged bias 200

Metadex Database 1985-1994 160

'/, !'/, /t

//

I / !// 100

I/

tt

//I// 2

so

// /t

Fig. 1. Distribution of papers describing dissimilar-metaljoining using various processes.

Fig. 2. A simplified representation of the EBW process [2].

electrode locat:.d slightly below the cathode and the anode, pass through the hole in the anode and form a beam, see Fig. 2 [2]. The electron beam can be focused under vacuum, and strikes the metal surface at velocities of up to 70% of the speed light. About 95% of the electrons' kinetic energy is converted into heat. The electron beam can be focused on diameters in the range of 0.3-0.8 mm, and the resulting power density can be as high as 10~°W m -2. One of the key features of EBW is its ability to perform deep penetration welding with the 'key-hole' mechanism. The EB weldments can, therefore, exhibit a high depth-to-width ratio. This feature allows for not only single-pass welding of thick plates, but also for the welding of relatively thin plates with a high travel speed. The basic EBW equipment consists of a welding chamber, an electron beam gun, a power supply system, a vacuum pumping system, and a control system. During the initial period of commercial application, the process was strictly limited to operations in a high vacuum chamber. However, the later development of the system permitted the option of welding in either a medium-vacuum chamber or a non-vacuum environment, which widened the use of EBW over a broad range of industrial applications [3]. Therefore, three modes of EBW are available at present: the high-vacuum mode (< 10-3 Torr); the medium-vacuum mode (10-2-25 Torr); and the non-vacuum mode (atmospheric). It should be noted that the use of non-vacuum EBW still remains very limited. The selection of EBW modes depends on the materials to be joined and the penetration requirement. Reactive and refractory metals have to be welded in a high vacuum to avoid atmospheric contamination, for instance, this applies to joints involving Ti or Ti alloys. Furthermore, the pene-

Z. Sun, R. Karppi/Journal of Materials Processing Techs'ohJgy 59 (1996) 257-267

259

Table 1 Process characteristics of EBW related to dissimilar metal joining Advantages - Accurately controllable energy density nnd the small beam size makes it possible to control dilution, to weld with high precision, and thus, to weld both very thin and very thick metals, e.g., from 0.025300 mm. - Possible accurate beam alignment at any position allows the two base metals to melt selectively to better satisfy the metallurgical compatibilities. - Low total heat input per unit length of weld produces a narrow weld bead and HAZ, and results in low residual stresses and minimum distortion, which can cause serious problems for conventional fusion welding processes. - It is possible to solve problems associated with metallurgical incompatibility more accurately with EBW when using a suitable filler material, although this can also be a possible solution for arc welding. High purity environment (vacuum) for welding minimizes surface contamination of the metal by oxygen, nitrogen and hydrogen, which is particularly beneficial lbr reactive and refractory metals. - Dissimilar-metal combination involving high thermal conductivity metals such as copper can be welded withol~t preheating.

Limitations Problems related to melting and mixing of dissimilar metals during fusion welding still exist. Possible beam deflection by electrostatic and magnetic fields due to dissimilar metals. Vacuum environment normally necessary, impossible with metals vaporising easily. High accuracy requirement in groove preparation. Rapid solidification may result in brittleness of the weld, and defects, e.g., porosity. Use of vacuum chamber may reduce product size and limit the product design. Although beam oscillation can minimize the groove preparation requirement, it may cause problems for dissimilar-metal joining due to the possible uncontrollable fusion ratio of the two metals. High equipment and run'ring cost.

-

tration capability decreases with lower levels of vacuum. As a rule, the highest weld quality is achieved with a high vacuum, although the two other modes can be applied satisfactorily in selected cases. The primary processing parameters in the EBW are the beam current (current), the beam accelerating voltage (voltage), the focusing lens current (focal beam spot size), the welding speed, and the vacuum level. The quality of the joints depends on the selection of these parameters, although some secondary variables can, to some extent, refine the parameters, and thus produce welds with the desired properties. Some of the parameters particularly affect the results of dissimilar-metal joining. Current, voltage, and welding speed are the main factors that determine the heat input of the process, although it is also affected by a number of other parameters, such as focus position, focal length and beam oscillation width. Therefore, selection of these parameters can influence the residual stresses, the

distortion, the heat-affected zone (HAZ) size, etc.. The selection of a suitable vacuum level is important for dissimilar-metal joining, particularly for joints with members involving refractory metals, such as niobium or highly-reactive metals like titanium and zirconium or their alloys. Other important parameters, such as the beam alignment relative to the joint centreline, and the focal beam spot size, could be domi,~ating factors in dissimilar-metal joir, ing due to their ability to control the desirable fusion ratio of the two metals. When using EBW, it is possible to weld particular dissimilar-metal combinations which "were previously considered to be non-weldable with conventional fusion-welding processes. For instance, joints between Cu and Ti can be EB welded by locating the beam to the Cu side to minimize the melting of the titanium and increase the heat dissipation of the copper [4]. The basic characteristics of the EBW process, related to dissimilar-metal joining, are summarized in Table 1.

Z. Sun, R, Ka~7~pi/Journal of Materials Processing Technology 59 (1996)257-267

260

Table 2 Some physical properties of commonly used metals [5] Metal

Fe A/ Cu Ni Ti Zn Mo W Zr Nb

Melting temperature (K)

Density (kg m-3)

1808 933 1356 1728 1933 693 2890 3683 3125 2740

7870 2700 8930 8900 4510 7130 10 220 19 250 6500 8600

3. Dissimilar-metal joints Dissimilar-metal joints are characterized particularly by compositional gradients and microstructural changes, which yield large variations in chemical, physical, and mechanical properties across the joint. The joining of dissimilar metals is, therefore, normally far more complex than the joining of similar metals. The difficulties encountered when joining dissimilar metals include the problems experienced when joining each base metal individually, and the problems specific to the range of compositions, as well as to properties and their incompatibility possible in phases and/or compounds of the two metals in various proportions. Further complexity arises from the addition of filler or insert materials, which is a common practice in the joining of dissimilar metals. Table 2 shows the physical properties of some commonly-used metal.~ r51. It is seen, readily, that large property differences ex~,s, tween various metals. The large variation of these and other properties provide an attractive source for the choice of materials for special service conditions and environments. However, the large differences in physical properties also yield difficulties in joining the different metals together. Both physical and chemical mismatches easily result in incompatibility. For example, a large difference in the melting temperature between two metals makes the conventional fusic~n-welding processes inapplicable due to, e.g., the segregation of low-melting eutectics, which can cause hot cracking. Large differences in thermal expansion will, in turn, lead to the formation of large residual stresses, and thus reduce the joint strength and cause fatigue problems. A difference in the thermal conductivity of the two metals easily causes uneven heat dissipation. Furthermore, chemical mismatches in dissimilar-metal joints can result in the formation of brittle phases and the diffusion of particular elements. These will adversely affect the service properties of the joints, and are the main reasons why dissimilar-metal

Thermal conductivity (W m - i k - l ) (0-100 °C)

Thermal expansion coefficient (k -1 x 10 6)

80 238 397 88.5 22 120 137 174 22.6 54.1

12.1 23.5 17.0 13.0 8.9 3.1 5.1 4.5 5.9 7.2

(o- lOO°c)

joining is normally more complex than similar-metal joining in many respects. Although EBW is a fusion-welding process and may still encounter these problems, it may, however, offer chances to reduce or overcome the problems to a certain extent, and hence, to produce satisfactory joints. For instance, EBW can solve the problem of the large difference in melting temperature more easily than, e.g., arc welding, due to the high-energy density resulting in a high heating and cooling rate. A low total-heat input per unit length of weld of EBW can also reduce the residual stresses substantially, as compared to arc welding. The thermal conductivity problem can be overcome by directing the beam correctly to the required location. The small weld bead size of EB welds minimizes the mixing of dissimilar metals, and thus, limits the brittle zones arising from the chemical mismatch, to some extent. Therefore, these advantages make EBW a viable technique for solving the problems of some specific metal combinations. According to metallurgical compatibility and other factors, the EB-weldability of different metal combinations have been ranked as shown in Table 3, the data originating partly from phase-diagram information, and partly from practical experience [6]. The joinability is classified on five levels. The first class represents the best joinability. The metals are both liquid and solid soluble in all combinations, therefore, they are readily weldable together without difficulties. The second class represents combinations probably acceptable, provided that suitable measures are taken. These measures can take different forms, such as the use of a filler material and/or preheating, suitable beam alignment, etc.. For the other three classes, the joinability gradually decreases as far as can be judged from the analysis of phase diagrams due to, e.g., the formation of brittle phases. How much more difficult these combinations are to be welded in practice remains an open question, because the solutions, especially in these categories, have not been experimentally studied in detail yet.

Z. Sun, R. Karppi / Journal o,/"MateriaLs" Processing Technology 59 (1996)257-267

261

Table 3 Weldability of dissimilar metal combinations using electron beam welding [6] AI atu ,Be Co Cu Fe Mg Mo Nb Ni Pt Re Sn Ta Ti W Zr

2 1 5 3 2 3 5 3 4 2 2 3 2 5 2 3 5 Ag

5 2 5 2 5 2 5 5 5 5 4 2 5 5 5 5 A!

5 2 ! 2 5 2 4 I I 4 5 4 5 4 5 Au

5 5 5 5 5 5 5 5 5 3 5 5 5 5 Be

2 2 5 5 5 1 1 1 5 5 5 5 5 Co

2 5 3 2 1 1 3 2 3 5 3 5 Cu

3 2 5 2 1 5 5 5 5 5 5 Fe

3 4 5 5 4 5 4 3 3 3 Mg

1 5 2 5 3 I 1 I 5 Mo

5 5 5 5 ! 1 ! I Nb

i 3 5 5 5 5 5 Ni

2 5 5 5

3 5 5

5 5

1

1

5

3

1

2

5 Pt

5 Re

5 Sn

2 Ta

!

5

Ti

W

1. Very desirable (solid solubility in all combinations), 2. Probably acceptable (complex structures may exist), 3. Use with caution (insufficient data for proper evaluation), 4. Use with extreme caution (no data available), and 5. Undesirable combinations (intermediate compounds formed).

Some possibilities may exist, but further investigations are required in these classes. Though similar judgement from the phase diagrams can be made for conventional fusion-welding processes, particular measures suitable for EBW cannot readily be realized with these processes. A good example is the accurate beam alignment used for producing a desirable fusion ratio for dissimilar metals. For instance, the focused beam diameter can be in the range of 0.3-0.8 ram, which can be located to _0.1 mm accuracy. The EB-solution can, furthermore, be associated with very precise control of filler-material addition typical of feeding units used in EBW, e.g., a CNC-controlled filler-wire feeding system has been successfully u~ed in production [7]. Therefore, EBW does possess its own characteristics for dissimilar-metal joinrag. It should be emphasized that the data given here provide the user only with broad guidelines, because specific weldability may depend upon a number of factors, such as variation in alloy compositions, welding parameters, differences in component design, service requirements, etc.. A recent survey about dissimilaralloy combinations successfully EB-welded is shown in Fig. 3 [81.

detail to show how the advantages of EBW are realized, and how the problems are soh, ed. 4.1. EB W with filler-metal addition The following example is an austenitic/ferritic transition joint. This material combination is widely used in many industrial applications, such as power-generation systems. When using conventional fusion-welding processes, nickel-based fillers have proven to be satisfactory. The theoretical background originates from the following two main factors. Firstly, the low diffusioncoefficient of carbon in nickel-based weld metal minimizes carbon diffusion across the ferritic steel/weld metal interface, and thereby reduces the tendency to form a soft decarburized zone in the ferritic steel. Secondly, the use of nickel-based fillers produces a weld C- and lowalloyed steels

High.alloyed steels with high C-content

Co-alovs

High-alloyed steels with low C-content

Ni and Ni-alloys ~

"

....... - 7.'~. ~

4. EBW of dissimilar-metal combinations A variety of metal combinations welded with the EB process has been investigated, many of them already being used in industrial applications. Table 4 lists some combinations found in a recent literature survey. Potential application fields, research items, or solutions, are mentioned in the table when possible. In the following, a few examples from Table 4 are introduced in greater

Cu a l l o y s ~ ~ _

~

Cu

O Ti

-'0

AI-alloys

good weldability weldable with special precautions Fig. 3. EB-weldability of dissimilar metals and alloys [8].

262

Z. Sun, R. Karppi/Journal of Materials Process#zg Technology 59 (1996) 257-267

Table 4 Recent investigations concerning the EBW of dissimilar metals Material combination

Potential applications

Remarks

2 I/4Cr-IMo/C-Mn steel

Steam turbine diaphragms Heavy equipment Power generation systems Saw blades

Beam deflection should be overcome

SA E ! 136/1010 ~teel 2 I/4Cr- 1 Mo~AlSl 405 High speed/carbon or low alloy steels High speed/mild steels 20NiMoCr6E- sintered powder/mild steel lnconel 713c/A!Sl 1045 13Cr steel/mild steel

Cutting tools Car industry Turbocharger impeller Steam turbine nozzle diaphragms

Pearlitic steel 20/321 stainless steel 20Khl2VNMF20KhN3MFA 321 stainless steel/ADI A 1 2 I/4Cr-! Mo/AISI 316 2 l/4Cr- 1 Mo, A204 and SS41/AISI 405 2 l/4Cr-I Mo/AISI 304 and 321 Cr-Ni alloy/EP517

Power generation systems Power generation systems Power generation systems Electromagnetic devices

EP288 austenitic-martensitic

type steel/VZhL-14 Ni base alloy Vn-2AE Nb base alloy/321 stainless steel AISI 422/Inconel 600/Stellite alloy Cu/Ag

Steam turbine blades Low temperature applications

CI0700/CI7510 Cu alloy Beryllium copper/copper

Crimp-type connector contacts

Silver/copper lnvar/stainless steel

Switch contacts Thermostatic bimetallic components Lead frames Nickel clad limiter Aerospace equipment

Steel/copper C u - C r - Z r ailoy/NiCrl 5Fe Nb/Cu Nb/Mo and porous W/Mo

Ta or Nb/321 stainless steel

Gold/Phosphorus bronze C-103Nb/Ti-6A! -4V

Emitting electrode, Ba-W cathode of microwave tubes Heating elements of high temperature vacuum furnace High reliability miniature switches

Ta/Mo or W

Ni-base superalloy/stainless steel EN 36/FV 520

Marine engine valves Marine pump ~haft

Ref. 9 10 11,12

Creep-fatigue property prediction was developed Saving of high speed steels can be achieved Optimization of parameters

Laser cutting and EBW

13,14 15 16 13 17

Effect of oscillations of beam on microstructure was studied PWHT is beneficial Multi-pass welding was used Fatigue-property prediction was developed Effect of PWHT, fracture toughness was investigated Using nickel-base filler wire

23-25

Optimization of parameters

26

Horizontal beam welding produce defect-free joints

27

Corrosion resistance was investigated Optimization of parameters

28

Contact resistance was studied Mechanical properites were investigated The joints combine good spring properties {Be-Cu) with good fonnability (Cu) Finely defined weld zone with minimal HAZ

Beam was aligned to the copper side Optimization of parameters

Microstructural study

18

19 20 21 22

29 30 31 32

32 32 32 33 34 35

36

32 Evaluation of various properties of the joints High temperature annealing after welding is beneficial for bend strength Good wear and corrosion properties

37 38

39 40

z. Sun, R. Karppi / Jou,nal of Materials Processing Technology 59 (1996) 257-267

having a thermal expansion coefficient very similar to that of ferritic steel. Therefore, the magnitude of ferritic steel/we~d metal interracial stresses can be reduced during the thermal cycling using nickel-based filler welds, as compared to autogenous welds or to welds made with austenitic stainless-steel fillers. Ruge et al. [23] investigated this material combination with EBW. Type 304 austenitic stainless steel was EBwelded to 2-~Cr-1Mo ferritic steel with different procedures: (i) autogenous welding; (ii) welding with E308 austenitic stainless steel filler wire; and (iii) welding with Inconel 82 nickel-based filler wire. The best metallurgical quality of the dissimilar-metal joints, in terms of microstructure, was obtained, just as in conventional arc welding, when using nickel-based filler wire. It was found that autogenous welding, and welding with austenitic stainless-steel filler wire produced an inferior weld-metal microstructure, i.e., they yielded the formation of martensite with high hardness values. Although conventional fusion welding with nickel-based filler wire can produce satisfactory joints, several further advantages of using EBW can be realized. Firstly, thick-section joints can be welded with less passes. For instance, a single-pass 20 mm-thick austenitic/ferritic joint was produced successfully, 60 mm thick joints being welded with 4 passes [24]. Secondly, due to its high energy density (low total heat-input per unit length of weld), EBW can yield a narrow HAZ and low residual stresses and distortions, which are important properties for thi~ type of transition joint when they are subjected to thermal cycling. Therefore, EBW with nickel-based filler wire has great potential for such applications. It can be seen from the above example, that EBW with a suitable filler metal can be a good solution when tackling metallurgical problems for dissimilar-metal joints, although autogenous welding is often regarded as an obvious advantage of the EB process, especially in the joining of similar metals. More generally, problems of metallurgical incompatibility and physical mismatches can be solved, also, in many other metal combinations by the use of particular filler or transition materials. Table 5 gives a few examples.

263

Fig. 4 A bimetallic saw blade produced by continuous EBW of a narrow strip, made of high-speed steel, to a spring-steelbacking [13]. 4.2. Precision welding

A typical example, taking full advantage of the high level of accuracy of EBW in dissimilar-metal joining, is the production of bimetallic saw blades. These saw blades usually contain joints between high-speed steel and either a plain-carbon or low-alloy steel due to both technical and economical reasons. High-speed steel is used only for the cutting teeth and has to be welded to a backing strip of another type of steel, with accurate control and a high welding speed of up to l0 m minI13]. An example of the product is shown in Fig. 4 [13]. One typical arrangement of equipment for realizing the continuous production of bimetallic saw blades is illustrated in Fig. 5 [43]. A recent study also showed that EB-welded saw blades exhibit the cutting performance of bulk high-speed steel blades, simultaneously saving 82% of high-speed steel expenditure [14]. This result was obtained with partial-vacuum EB welding. A more complicated geometry of joints is necessary for hobbing cutters which have also been successfully EB welded, see Fig. 6 [15]. Furthermore, EBW has been used for welding dissimilar-metal continuous strips in production lines for joints as thin as 0.125 mm [32].

Table 5 Examples of filler metals or transition metals ill between the two surfaces for electron beam welding[6,8,41,42] Metal A

Metal B

Filler or transition material

Tough pitch copper Hastelloy X

Mild steel SAE 8620 steel

Nickel or Monel 321 stainless steel

304 stainless steel

Monel

Hastelloy B

304 stainless steel Ta-8W-2Hf TZM (mainly Mo)

AI W-25Re-3Mo Haynesalloy 230 (ca. NiCr22 WI4)

Ni Pure Mo Nb-Zr

Fig. 5 Schematic illustration of an EBW-equipment arrangement used for the continuous production of bimetallic saw blades [43].

264

Z, Sun, R, Karppi/Jourmd of Materials Processing Technology 59 (1996) 257-267

4.3. Fitness-for-purpose applications

Fig. 6 A hobbing steels [! 5].

high-speed and carbon

Further examples of multi-material products realized ',ith EB-welded joints are the sensor components of instruments used for geological surveys [44]. A typical component, shown schematically in Fig. 7, is made of glass, Kovar and copper. The joints between the glass and Kovar were made by vacuum brazing. The major requirements for the EB-welded joints between the Kovar and the copper tubes are that they are leak tight, and have minimized distortion. Although the welding of Kovar to copper could be realized easily with fusionwelding processes, the requirement of accurate joint size and location, and the cleanness of the product, make EBW a much more suitable method than any of the other conventional welding processes. The utilization of EBW to join dissimilar metals does not only originate from the avoidance of metallurgical problems faced by conventional fusion-welding processes, but also, in many cases, from special features of the process which meet the requirements of the product. For instance, the distortion caused by welding is a very common problem that is emphasized in the welding of dissimilar metals. Strictly speaking, there is no welding process that can completely eliminate distortion. However, a process that minimizes distortion is increasingly preferred in many applications. EBW with a high energy density produces much smaller weld distortion than other conventional fusion-welding processes. Therefore, the process can be a real benefit, and is often the best choice for joining finished machined parts to a component, provided, of course, that EB,equipment is available. Fig. 8 shows two examples of products consisting of dissimilar alloys [8]. They are both readily available in a serviceable condition, without any postweld treatments, after EB-welding. All these examples demonstrate that EBW can provide practical solutions for the precision welding of dissimilar metals.

An example of the fitness-for-use of dissimilar-metal EB-weldments is the Cu/Ag joints used in low-temperature applications [30]. This type of a joint requires proper thermal contact between the dissimilar metals at low temperature due to the working mechanism of the component. EB-welded Cu/Ag joints were found to have a much better contact resistance than the joints made by other processes, such as diffusion-welded joints, screw-fastened joints and soldered joints [30]. Although the Cu/Ag combination can also be easily welded with many fusion welding processes, a weld metal with poor thermal conductivity remains the main problem. For instance, TIG welding cannot meet this technical requirement of the products, because the larger newly formed fusion zone of the TIG weld will change the thermal conductivity of the contact area substantially. When using EBW, a very thin layer between the dissimilar metals is formed, thus ensuring a small contact resistance. This example again demonstrates that EBW can offer viable solutions for those joints where conventional processes are unfit, not because of weldability problems, but due to the requirement of the product.

4.4. Welding involving refi'actory metals EBW is an excellent process for joining refractory metals and their dissimilar combinations because of the high energy density, and hence, the minimum total heat input per unit length of weld. This is a- especially important factor with welding molybdenum and tungsten, since fusion and recrystallization raise the ductileto-brittle transition temperatures (DBTT) of these two metals to above room temperature. Thus, the HAZ produced by conventional welding processes becomes prone to brittle fracture. The short time at high temperature in EBW can, however, minimize the grain growth and other phenomena that raise the DBTT, improving, in this way, the HAZ properties. A study of the EBW of niobium to molybdenum thin sheets (0.5 or I mm thick) for the emitting electrodes of thermionic energy converters reports satisfactory joints [35]. The successful joints were obtained by exploiting the advantages of EBW, such as a highly focused beam and a low total heat input per unit length of weld, which reduce the fusion zone and the HAZ. Furthermore, the beam alignment to the niobium side was found to minimize adverse high-temperature influences on the molybdenum. In the same report, joints between porous tungsten and molybdenum for cathode assemblies were also EB-welded successfully. Although these combinations would be considered fusion weldable from the judgement of the phase diagram, difficulties do exist for conventional fusion-welding processes due to the high

Z. Sun, R. Karppi /Journal of Materials Processing Technology 59 (1996) 257-267

265

Eb -welded joint

kovar tube Copper tube

Fig. 7 Schematic diagram of a sensor component used in geological survey instruments. The length of the copper tube is about 200 mm.

melting temperature of the base metals, and the accurate dimensional requirement of the joint. In this aspect, EBW does show its advantages over conventional fusion-welding processes. A further example is a recent study on the EBW of niobium/copper joints. The Nb/Cu system is under consideration for a variety of applications, such as electrode plating drums, superconducting accelerating cavity components, microcomposite wires and sheets, and in the US for National Aerospace Plane heat exchanger panels [34]. The binary phase diagram indicates very limited solubility between Nb and Cu. Nb has its largest solubility, around 1%, in molten copper at 1090 °C. Cu reaches its maximum solubility, 1.5%, in Nb at 1090 °C. Despite these figures not offering expec-

EB-welds

Wear-resistantsteel

,

i

/~

Constructional

IIIIIIIIIIIl te '

/

_.L.

i

tations of success in fusion welding, an electron beam weld was made by directing the beam to the copper side. Surprisingly, a satisfactory weld bead quality and joint strength were obtained. Fig, 9 shows a sample of the EB-welded Nb/Cu joint [34]. This study indicates the potential of the EBW of a Nb/Cu combination, regardless of the reasons behind this success still not being fully understood and needing further investigation. The author suggests that the dispersion of small Nb particles in the copper weld metal are responsible for the good strength of the weld metal. In any event, it is clear that conventional fusion-welding processes are very liable to fail in joining the Nb/Cu combination due to the substantial difference in both melting *emperature and thermal conductivity. It should, therefore, be concluded that EBW is a vital method for welding dissimilar-metal combinations involving metals having high thermal conductivities. It has been reported that 100 mm thick copper can be EB-welded with a single pass, due to its high energy density [10]. This feature of EBW can, certainly, be exploited to produce many other combinations with metals having high thermal conductivities in a wide range of plate thicknesses.

i J_

/

(

,

Quenched-tempered steel

Carburizingsteel

(a,~ EBEB

EBEB

EB

EB

T

~ F/Z/Z/Z/Z. ,x Xll/r . ~ ' ~ \

IM3 I

CO

crF--0

x.,,/# (b) Fig. 8 (a) Gear produced with EB welding; and (b) an EB-welded section of a gas turbine rotor (M! =X2CrNi i8-i0, M2=St 35, M3 = Ck 15) [8].

Fig. 9 Electron-beamweldedcircumferentialmock-up oi"copper and niobium cylinders [34].

266

Z. Sun, R. Karppi/Journal of Ma: 'rials Processing Technology 59 (1996) 257-267

5. Future developments Many advantages can be gained in joining dissimilar metals using EBW. However, there are also a number of metallurgical factors that must be carefully considered when applying EBW. To fully utilize these advanf u t u r e applications, the following aspects should be considered, (i) High energy density and deep-penetration capability are the most important features of EBW, particularly for the welding of thick materials. In this respect, EBW can be regarded as a more effective method compared to laser-beam welding. Therefore, thick-section dissimilar-metal joining by EBW can be one potential R&D field. Indeed, welding technology for several combinations has already been developed, but many others can still be exploited. (ii) Although many combinations have been proven to have a good weldability, R&D for many difficult joint types are still required. Therefore, further R&D efforts should be directed, especially, to those combinations with which an alternative joining process cannot provide a technically and/or economically satisfactory result. To avoid major difficulties, the potential solutions can be judged from, e.g., the pl~ase diagram prediction. (iii) EBW, with filler-metal addition, should be emphasized in these trials, because recent studies have demonstrated that this method can be applied to solve many problems associated with metallurgical imcompatibility, although autogenous welding is often regarded as one of the advantages of the EB-process. (iv) As EBW is a high-precision joining process, it can be augmented to many new applications with highaccuracy requirements. These can vary from micro.joining to the high-precision assembly of thick components, with no need for post-weld treatments. (v) EBW possesses the feature of low total heat input per unit length of weld which can be beneficial for achieving a joint with low distortion and residual stresses. This characteristic can be applied to weld components where tolerances of distortion are limited. Consequently, these parts can be used without postweld machining. This, in turn, provides the designer with increased flexibility for the product and even manufacturing-route development to gain substantial economic benefits.

6. Conclusions Attempts have recently been made to weld dissimilar metals using EBW with high energy density. These studies have already demonstrated many advantages over conventional fusion-welding processes. EBW has the potential of not only being able to weld material

combinations, which often cause problems with conventional welding techniques, but also of showing functional advantages for some combinations which also can be produced with conventional welding processes. However, the use of the EBW process to join dissimilar metals is still at the development stage. Much more of the potential will be exploited in future via extensive R&D efforts. The limited amount of information available concerning the joinability of different dissimilar combinations, and the technical practice for potential weldable combinations, will stimulate many studies. The results of these stadics should provide industries with competitive solutions and consequent benefits.

Acknowledgements The authors wish to thank M. Vilpas and P. Vartia for their helpful discussions.

References [1] D.E. Powers, Electron beam welding--an overview, in E.A. Metzbower and D. Hauser (eds.), Proc. Con.f~ Power Beam Processing, ASM International, Ohio, 1988, pp. 25-33. [2] Y. Arata, Development of ultra high energy density heat source and its application to heat processing, Okada Memorial Jpn. Soc., Tokyo, 1985. [3] D.E. Powers and B.W. Schumacher, Using the electron beam in air to weld conventionally produced sheet metal parts, Wehl. J., 68(2) ( !989) 48- 53. [4] M.M. Schwartz, Modern Metal Joining Techniques, Wiley, New York, 1969. [5] E.A. Brandes (ed.), Smithwells Metals Reference Book, Butterworth, London, 6th edn., 1983. [6] Welding Handbook, American Welding Society, Miami, Florida, 8th edn., 1991, Vol. 2. [7] 1. Meuronen, private communication with Valmet Lentokoneteollisuus Oy, Linnavuori, Finland, 1994. [8] O. Braun, J.J. Chene and G. Schmelzer, Towards a full exploitation of the electron beam welding process, Proc. bn. ConJ~ Joining~Welding 2000, Tire Hag:re, The Netherlands, !-2 July, 1991, Pergamon, Oxford, 1991, pp. 107-118. [9] Y. Akutsu, Application of electron beam welding to steam turbine diaphragms, Beam technoi, Lectures of the Int. Beam Technol. Conf, Essen, Germany, 7-8 May 1980, pp. 63-69. [10] G. Sayegh, Application of electron beams in welding metals, Cot~ 6th General Assembly CENIM (6th Asamblea General del CENIM) I1, Madrid, Spain, 8-11 Oct. 1985. Paper No. 43, CENIM, Avda. Madrid, Spain, 1985, p. 10. [11] M. Okazaki, Y. Mutoh and M. Yamaguchi, Creep-fatigue fracture of dissimilar metal electron beam welded joints at elevated temperature, J. Eng. Mater. Technol., !10(3)(1988) 212-218. [12] M. Okazaki, Y. Mutoh and Y. Itoh, Low cycle fatigue strength and its prediction for dissimilar-metal electron-beam-welded joints at high temperature. High Temperature Creep-Fatigue, Elsevier Applied Science, Essex, UK, 1988, pp. 183-202. [13] S. EUiott, New fusion joining processes for dissimilar metals, Welding Dissimilar Metals, The Welding Institute, Cambridge, UK, 1986, pp. 16-19.

Z. Sun, R. Karpt,i / Journal ~" Materials Processing Technology 59 (1996) 257-267 [14] Y. Liu, C. Sun and X. Chen, Investigation on the equipment and technology of partial vacuum electron beam welding in the continuous manufacture of bimetallic saw blade, Trans. China Weld. Inst., 8(1)(1987) 31-37. [15] A.M. Kosecek, J. Bobcik and J. Andrlik, Electron beam welding of hobbing cutters, Co~!Ii Electron and Laser Beam Welding, Tokyo, Japan, 12-19 July 1986, Pergamon, Oxford, UK, 1986, pp. 101 - 107. [16] G. Sayegh and R. Cazes, Where does laser beam compete with electron beam in industrial welding applications? Conf. the Laser versus the Electron Beam in Welding, Cutting and Surface Treatment: State of the Art, Reno, Nevada, USA, 6-8 Nov. 1985, Vol. I, Bakish Materials Corp., Englewood, New Jersey, USA, 1985, pp. 273-288. [17] S. Matsui, H. Matsummura and K. Yasuda, The application of laser cutting and electron beam welding to the manufacturing of turbine nozzle diaphragms, Weld. World, 25 (1-2) (1987) 16-25. [18] A.P. Lopatko and N.I. Nikitin, Effect of beam oscillations in electron beam welding on the structure of the fusion zone of pearlitic and austenitic steels, Weld. Prod. (USSR), 32(11) (1985) 37-39. [19] V.M. Nesterenkov, D.Yu. Novikov and I.P. Kirpach, Structure of electron beam welded joints in dissimilar steels, Paton Welding J., 3(!!) (1991) 841-845. [20] G.M. Grigorenko and A.A. Gordonnaya, Effect of the number of passes on the structure, phase composition and properties of electron-beam welded joints between joints between alitised 12KhI8NIOT steel and AD1 aluminium, Paton Weld. J., 3(11) (1991) 861-864. [21] M. Okazaki, Y. Mutoh and Y. lkeda, Creep-fatigue strength and its prediction of dissimilar metal electron beam welded joint, J. Soc. Mater. Sci. Jpn., 38(425) (1989) 168-174. [22] Y. Itoh, K. Tomod~ and Y. Yuasa, Fracture properties of electron beam welded dissimilar metal joints with incomplete penetration, J. Soc. Mater. Sci. Jpn., 36(405) (1987) 549-555. [23] J. Ruge, C. Oestmann, i. Decker and K. ~versen, Welding of dissimilar metals by using the method of electron beam welding with filler wire, Proc. Cot!['. Eh,ctron and Laser Beam lYeMing, Tokyo, Japan, 12-19 Juh' 1986, Pergamon, Oxford, UK, 1986 pp. 193-203. [24] J.L. Ruge, C.M. Oestmann and I.A. Decker, Joint quality of EB welds using filler wire (EBW-FW), Conf. Power Beam Technol., Brighton, UK, 10-12 Sept. 1986, The Welding Institute, Cambridge, UK, 1987 pp. 351-359. [25] I. Decker, C. Oestmann and J. Ruge, Metallurgical properties of electron beam welded dissimilar metal joints, Conf. Welding ha Nuclear Eng#leer#tg, Nuremberg, Germany, 26-28 Nov. 1986, Deutschen Verband fur Schweisstechnik e.V., Dusseldorf, Germany, 1986, pp. 147-152.

267

[26] V.K. Dragunov, T.P. Murav'eva and Yu.P. Rodionov, Electron beam welding dissimilar alloys used in electromagnetic devices, WeM. bit., 5(9) (1991) 723--725. [27] N.P. Voronov, Electron beam welding EP-288 steel and VZhL14 alloy, Paton Weld. g, 3(4) ( 1991 ) 317- 3 i 8. [28] G.D. Nikiforov, E.N. Sivov, N.P. Krutogolovov and N.E. Sibova, Corrosion resistance of weld joints of niobium with steel in boiling nitric acid, Svar. Proizvod, (1) (1986) 22-24. [29] F. Franchini, Electron beam welding of heterogeneous materials for steam turbine blades, Weld. Int., 7(3) (1993) 206-210. [30] S. Yin and P. Hakonen, Electron-beam welded copper-to-silver joints for thermal contact at low temperatures, Rev. Sci. histrum., 62(5) (1991) 1370-1371. [31] T.A. Siewert and D.P. Vigliotti, Mechanical properties of electron beam welds in thick copper, Cot~ Advances #~ Cryogenic Enghleering: Materials, Los Angeles, California, USA, 24-28 July 1989, Plenum, New York, 1990, Vol. 36B, pp. 1185-1192. [32] R.J. Russell, Electron beam welding for dual metal strip, Weld. J. 61(1) (1980) 21-24. [33] G. Israel, Experiences with transition welds between a copperchromium-zirconium alloy and NiCrl5Fe on nickel limiters, Schweissen Schneiden, 38(10) (1986) 495- 501. [34] R.J. Sinko in S.A. David and J.M. Vitek, (eds.) Niobium to copper electron beam weldability study, Proc. Int. Conf on Trends 01 Welding Research, Gatl#lburg~ TN, June 1992, ASM Int., 1993, pp. 787-791. [35] S. Lin, B. Fan and Z. Guan, Electron beam welding of dissimilar metals (niobium-molybdenum, porous tungsten-molybdenum), Conf. Electron and Laser Beam Welding, Tokyo, Japan, 12-19 July 1986, Pergamon, Oxford, UK, 1986, pp. 227-235. [36] A. Detka, Electron beam welding of tantalum and niobium, Przegl. Spawalnictwa, 41(2) (1989) 14-16. [37] F. Franchini and P. Pierantozzi, Electron beam welding of different materials: Niobium-base alloy C-103 with titanium-base alloy Ti-6AI-4V ELI, WeM. Int. 6(I0) (1992) 792-797. [38] Y. Hiraoka and R. Fukuda, Electron-beam welding of dissimilar refractory metals, J. High Temperature Soc .lpn., 18(!)(1992) 45-55. [39] N.N. Electron beam process improves weld quality, Shiprepair. (June 1993) 45-55. [40] F.J. Becket, Electron beam welding in product design and manufacture, Col!L Power Beam Tecimol.. Brighton, UK, I0-12 S~Tt. 1986, The Welding Institute, Cambridge, UK, 1987, pp. 97-!12. [41] G. Tardiff, Fusion welding of a tungsten alloy to a tantalum alloy, Weld. J., 51(1)(1972) 36s-40s. [42] G. Metzger and R. Lison, Electron beam welding of dissimilar metals, Weld. J., 55(8)(1976) 230s-240s. [43] Leybold-Heraeus GMBH Brochure, 1986. [44] P. Vartia, private communication with the Technical Research Centre of Finland, Espoo, Finland, 1993.