1

vi

Bonding of Stainless Steel to Aluminum Base Alloy

1. M. Arsalan Zamir Khan05 MT 86

(Group Leader)

2. Raja Hamid Bin Amir05 MT 01

3. Riaz Jamali05 MT 46

4. Riaz Hussain Jalbani04-05 MT 90

SUPERVISED BY

Prof.Dr. Moazzam Baloch

Department of Metallurgy and Materials Engineering,

Mehran University of Engineering and Technology

Jamshoro, Pakistan.December, 2008.

Thesis submitted in partial fulfillment of requirements for the Degree of M.S Materials Engineering

DEDICATIONWe dedicate this thesis to our

PARENTS

whose guidance and love has bought us to the position we are currently at and without whom we are nothing..

Acknowledgement

All gratitude is due to Almighty Allah, the most Gracious and Merciful, Who capacitated us to complete this report. We offer our humblest thanks to Holy Prophet Hazarat Muhammad Who is the forever a model of guidance and knowledge for humanity.This project report has been a very enlightening and rewarding experience for us in an area that is of great personal interest. We would like to acknowledge my deep sense of gratitude and indebtedness to our meritorious supervisor, Prof. Dr. Moazzam Baloch with great reverence ecstasy for his encouragement, expert advice, sincere efforts and precious time. He is really a versatile genius of high order. His devoted love is worth appreciable. We couldnt find the words to express my deepest gratefulness to all our teachers at MUET, JAMSHORO, whose efforts made us what we are today, especially Prof. Dr. Abdul Hakeem Mallah, Chairman Department of Metallurgy and Materials Engineering for his guidance and wisdom throughout our academic carrier, Mr. Nisar Memon for his dedication to teaching and efforts he makes towards explaining the details in every subject he teaches, Mr. Riaz Memon for his moral advices and Mr. Ishfaque Ahmed Isani for his ever-ready help regarding anything computers.We would also like to acknowledge the help offered to us by Mr. Rehan Majid, Mr. Hamid Raza and Mr. Mehboob Shah of PMO for providing the required materials for the Project and their much valued help. M. Arsalan Zamir Khan

Raja Hamid Bin Amir

Riaz Jamali

Riaz Jalbani

Table of ContentsviiList of Figures

11Introduction

21.1Project Scope

32Literature Survey

62.1Diffusion in Metals

62.1.1Diffusion Mechanisms

62.1.2Basic One-Dimensional Equation for Diffusion Flux

72.1.2.1Fick's First Law

72.1.2.2Fick's Second Law

82.2Methods of Measuring Diffusivity

82.2.1.1Grube-Jedele Method

92.2.1.2Matano-Boltzmann method

102.3Determination of Intrinsic Diffusivities

112.4Diffusion Bonding

112.4.1Types of Diffusion Bonding

112.4.1.1Solid-State Diffusion Bonding

142.4.1.2Transient Liquid Phase (TLP) Diffusion Bonding

172.5Intermetallic Compound Formation during Diffusion Bonding

182.6Diffusion Bonding of Al Base Alloys to Stainless Steel

223Experimental

223.1Samples Preparation

233.2Formation of interlayer

233.2.1Formation of Cu Interlayer

243.2.2Formation of Zn interlayer

243.2.2.1Physical Vapor Deposition (PVD) of Zn

263.2.2.2Electrochemical Deposition of Zn

273.3Coupling the Samples and Heat Treatment

293.4Characterization of the bonded interface

314Results and Discussion

314.1Results

324.1.1SEM microscopy

344.1.2EDS spectrum of various regions in set # 3

394.1.3Microhardness test results

394.2Observations in other samples

404.3Discussion

425SEM Operating Procedures

425.1Power Up

425.2Loading Sample

435.3Operation

435.3.1Initializing

435.3.2Imaging

445.3.3More function keys

445.3.4Power down

445.3.5Image Capture

45Conclusions

45Future Recommendations

ReferencesList of Figures

10Figure 21 Boltzmann-Matano geometry for diffusion couple

12Figure 22 Stages of solid state diffusion bonding [9]

16Figure 23 Stages of liquid phase diffusion bonding [10]

18Figure 24 Al-Cu phase diagram [12]

19Figure 25 Al-Zn phase diagram [12]

20Figure 26 Al-Fe phase diagram [12]

22Figure 31 An Al alloy and SS couple aftercutting and polishing

22Figure 32 A stainless steel sample size used for diffusion bonding

23Figure 33 Cu foil used as interlayer material

23Figure 34 Cu foil placed between an SS and Al sample

24Figure 35 PVD apparatus used for Zn deposition

25Figure 36 Sample and evaporant position in the PVD apparatus

27Figure 37 (a) Steel fixturefor holding the samples (b) a diffusion couple in the fixture

28Figure 38 (a) Bovenden vacuum furnace used for heat treatment (b)furnace controllers

28Figure 39 Time vs. temperature and pressure of high vacuum furnace.

29Figure 310Muffle furnace used for heat treatment

29Figure 311 (a) Set # 3 (b) Set # 5 after diffusion bonding and mounting

31Figure 41 Al sample surface after collapse of the bond in set # 2

31Figure 42 SS sample surface after collapse of the bond in set # 2

32Figure 43 SEM micrograph of bonding interface of set # 3 in secondary electron mode

32Figure 44 SEM micrograph of bonding interface of set # 3 using back scattered electron mode

33Figure 45 Different layers formed during diffusion bonding of set # 3

33Figure 46 Al alloy microstructure after heat treatment. Dark regions are of Al+ while white regions are CuAl2

34Figure 47 EDS analysis of layer1

34Figure 48 EDS analysis of white region in layer 2

35Figure 49 EDS analysis of rod type phase in white contrast in layer 2

35Figure 410 EDS analysis of layer 3

35Figure 411 Grey contrast in Al near layer 3

36Figure 412 White contrast in Al base material ( phase)

36Figure 413 EDS analysis of black contrast within grayish grains of Al alloy

36Figure 414 EDS analysis of white contrast within grayish grains of Al alloy

38Figure 415 EDS line scan results of Al, Fe, Cr, Ni and Cu

39Figure 416 Al and SS samples with Zn interlayer (set # 4) after heat treatment

List of Tables20Table 21 Composition of SS 316 and Al alloy 5182 (wt.%)[18][19]

Table 22 Some physical properties of SS316 and Al alloy 518221Table 23 Some mechanical properties of SS316 and Al alloy 518221Table 41 Composition of various layers, regions and phases formed during diffusion bonding37Table 42 Microhardness test results39

List of Equations5

Equation 20

7

Equation 21

Equation 227

Equation 237

Equation 248

for - < x < Equation 258

Equation 269

Equation 279

Equation 289

(1/slope at CA) (area from CA1 to CA)2910

10x2/t = k Equation 210

10v = x/t = k/2x Equation 212

10v = x/2t Equation 213

Equation 21411

Equation 21511

Equation 21616

Equation 21717

List of Variables Used in EquationsEquation 2-0

s

Equation 21

J

D

C/x

Equation 2-2J

C/x

D*v

C

Equation 2-4So

Xt Equation 2-5CACA1, CA2

erf x/2t

Equation 2-8

t

CACA1

Equation 2-10

K

v

Equation 2-16

CoWoCLCmEquation 2-17

h

tHShear Strength

Solid fraction of Aluminum alloy

Flux of diffusing species

Tracer diffusion coefficient

Concentration gradient

Flux of diffusing species

Concentration gradient

Tracer diffusion coefficient

Atomic drift velocity

Concentration of diffusing species

Original concentration per unit area

Any crystallographic axisTime

Atom fraction at a distance x from weld interface

Atom Fractions of alloys A and B

Diffusivity

Error function

Time

Concentration in atomic units at a distance x measured from the Matano interfaceconcentration of one side of the diffusion couple at a point well removed from the interface where the composition is constant and not affected by the diffusion process

Constant

Velocity of some refractory markers

Solute concentration in the interlayer material

Interlayer thickness

Maximum solute concentration in the solid at the interface from phase diagram

Solute concentration in the base material

Half the width of the bonding interface

Homogenization time

AbstractFabrication of different components of a device through joining of materials by conventional welding techniques is very difficult if physical properties of the materials differ a lot. The size of the fusion zone (FZ) as well as the heat affected zone (HAZ) is large in conventional welding techniques. In case of dissimilar materials where joining components have large difference in their thermal properties the joints go through different thermal stresses and hot cracking occur in the welding regions as well as in the HAZs. Al alloys and stainless steels have different melting temperature as well as thermal coefficients. Diffusion bonding of Al alloy and stainless steel was carried out using Cu or Zn as interlayer as the Cu and Zn form eutectic with Al at 548C and 381C respectively. Samples were heat treated at different temperatures according to eutectic temperature in vacuum and microstructure were characterized using scanning electron microscope (SEM) and energy dispersive spectroscopy (EDS) techniques. Microhardness test results are correlated with intermetallic layers.

1 Introduction

Stainless steel has the properties of high strength, high resistance to etching and wear. Aluminum alloy has the properties of high specific strength and perfect thermal and electrical conductivity. The joint between Al based alloy and stainless steel is required in wide application foreground in many fields, such as automobile, aviation, railway, family electrical equipment, cooker, chemical industries, and so on.

Because of the great difference in physical and chemical properties between stainless steel and aluminum alloy, reasonable bonding methods are scarce. The joining methods can be divided into two kinds by the state of materials, solid solid bonding and solidliquid bonding. For solidsolid bonding, the bonding ways of stainless steel and aluminum are mechanical and partial physical bonding, the shear strength is only about 40 MPa, so heat treatment must be adopted to improve the bonding quality. For solid liquid bonding, brittle chemical compounds layer can be easily formed at the interface because of the higher bonding temperature, and the shear strength is also lower and is only about 70 MPa[1].Due to the fact that mechanical properties and deformation and fracture behavior of bi-material plate, are mainly dependent on their interfacial bonding of the two materials, the selection of processing variables affecting interfacial structure, diffusion distance and compound formation and morphology are critical [2].

In recent years, transient liquid phase diffusion bonding technology is widely applied. But as a method of different metal joining it is relatively rare, because there exists micro-segregation in the electromagnetic stirred aluminum alloy cast slab, the parts of rich alloy elements that have low melting point will melt first at the effect of surface tension of the liquid phase. The parts with higher melting point become smoother and the semi-solid in structure with solid phase suspended in the liquid phase formed. With this feature, during bonding period of stainless steel and semi-solid aluminum alloy, a new type of interfacial structure is constructed. Solid phase and liquid phase are bonded with stainless steel by turns along the bonding interface in the process. The layer structure of FeAl metallic compounds in the interface is destroyed, so the mechanical properties of the material are improved. But this technique requires the heating of the Al alloy to the semisolid state which requires more energy. Also, this technique may lead to the formation of NiAl3.1.1 Project Scope

This present work is about lowering the bonding temperature and shortening the bonding time as much as possible. To achieve these goals some interlayer is incorporated between the two very materials. The interlayer material forms a eutectic with Al at a temperature well below the semisolid state of Al alloy. The eutectic allows an easy diffusion of the major constituting elements of the two adjoing materials into each other.

Intermetallic formation can be avoided by optimizing the soaking time. Intermetallics are formed only when constituting elements are present in stoichiometric quantities. Building of stoichiometric quantity at large distance is prevented by restricting the excessive diffusion. As diffusion is a time dependent phenomenon so diffusion time is optimized in such a way that diffusing atoms form a continuous lattice across the two material but not penetrate deep into the opposite side to form brittle intermetallics.

Potential interlayer materials for joining of Stainless steel and Al alloys are Cu and Zn. Cu forms a eutectic with Al at 548C while Zn has a eutectic with Al at 381C. In both cases joining temperature can be selected well below the temperatures involved in previous joining method of SS and Al alloys.

The prime requirement for this joining technique is that diffusion bonding should be carried out in a fine vacuum. Oxidation of the two adjoining surfaces hinders diffusion and it may also cause a dispersion of oxide at the interface which lowers the strength of the joint. Application of high strength Al alloy and stainless steel joint are in the following fields of engineering: High strength Al alloy and SS joint may be required in the cladding material of nuclear reactor core. In high temperature turbo engine parts, joining of Al alloys with SS is required. In cryopumps stainless steel cladding is used while some Al alloys are used as cryoconductor in temperature sensing transducers, so a joint may be required between SS and Al alloy. In pulsed power applications, SS structural component are required to join with Al Conductors.2 Literature SurveyDifferent researchers have worked on various aspects of diffusion bonding of dissimilar materials. Most related and helpful work regarding this project is explained below with the corresponding author name:

He and Liu [3] worked on mechanism of forming interfacial intermetallic compounds at interface for solid state diffusion bonding of dissimilar materials. They concluded that: I. The kinetic driving force of the phase at the interface is determined by specific properties of the components in the diffusion couple, and the composition ratio in the phase formed is in agreement with the diffusion flux ratio. II. The principle of formation of intermetallic compounds at the interface for the multi-composition diffusion couple is the flux-energy principle, and the phase with the largest thermodynamic driving force forms first based on the diffusion flux ratio at the interface. Deqing et al. [2] worked on making cladding of stainless steel on aluminum and carbon steel by interlayer diffusion bonding. They concluded that:

I. An AlCuMg interlayer alloy has been successfully used in preparing the 304L/L2Y2 (an Al-Fe-Si alloy) and the 304L/Q235A (a low carbon steel) clad sheets. II. The thickness of the diffusion layers in the Q235A and 304L steels is increased with increasing cladding temperature and time, and the Q235A steel acquires much thicker diffusion layer than 304L does at the same cladding conditions. The Interfacial cohesion of the substrates and the clad stainless steel is closely related to the thickness of the diffusion layers in the substrates and the clad, and controlled by the diffusion layer thickness in 304L side. A thickness of about 30 and 90 m diffusion layers in the 304L clad material and Q235A steel substrate respectively is needed to achieve sound bonding of the clad samples. Juan et al. [4,5] studied the diffusion bonding of Fe-28Al alloy with austenitic stainless steel and low carbon steel in vacuum and diffusivity of Al and Fe near the diffusion bonding interface of Fe3Al with low carbon steel and 18Ni-8Cr stainless steel and concluded the following:I. The calculated values of concentration of Al, Fe near the Fe3Al/Q235 (low carbon steel) interface conform to the experimental values obtained by EPMA. With increasing heating temperature T and holding time t, the diffusion distance of the elements in the diffusion transition zone for Fe3Al/Q235 dissimilar materials increases.II. Diffusion coefficient of Fe and Al in the Fe3Al/Q235 interface transition zone is larger than that in Fe3Al and Q235 base materials at the same temperature. The microstructure near the Fe3Al/Q235 interface is favorable for diffusion.III. There exists a latent time, t0, for the formation of diffusion transition zone near the Fe3Al/Q235 interface. The relation between the width of transition zone, x, and holding time, t, follows a parabolic growth law.IV. Activation energy for diffusion of Fe and Al has been estimated.V. The reacted phases are produced in the diffusion reaction layer and they are determined to be FeAl, Fe3Al, Ni3Aland Fe(Al) solid solution. These phases may decrease diffusion activation energy of Al, Fe, Cr and Ni to promote element diffusion and even accelerate the formation of diffusion reaction layer.VI. The growth of the diffusion reaction layer with bonding time obeys parabolic law.VII. The diffusion of Cr and Ni at the interface is favourable to improve the strength of the interface between Fe3Al and steel. The width of Fe3Al/Cr18Ni8 interface transition zone is 35 mm. It is larger than that of Fe3Al/Q235 interface transition zone (29 mm) under the condition of 1333 K 60 min and a pressure of 15 MPa. The second phase rich in Cr and Ni distributes dispersively in Fe3Al formed at the Fe3Al/Cr18Ni8 interface to cause dislocation couples with different spacing and even dislocation network. This is favourable to improve the joint strength at the Fe3Al/Cr18Ni8 interface. Shirzadi et al. [3] studied the interface evolution and bond strength when diffusion bonding materials with stable oxide films. They concluded that:I. Bond strength of diffusion bonded metals or alloys like Al increases with increasing surface roughness.

II. Bond strength of the metals with oxides which are soluble in the bulk of the metal or decompose at bonding temperature like Steels, copper, titanium, tantalum, columbium and zirconium show good bonding as the surface finish is improved. Guo et al. [1] investigated the interfacial strength and structure of stainless steelsemi-solid aluminum alloy clad metal. They concluded that:

I. Using semi-solid joining technique, the bonding of stainless steel and semi-solid aluminum alloy is successfully realized. The best solid fraction of aluminum alloy is 30%, and the maximum interfacial shear strength is 106.17 MPa. The relationship between interfacial shear strength and solid fraction of aluminum alloy is as follows:

Equation 20II. Solid phase and liquid phase are bonded with stainless steel by turns along the bonding interface, because there are different diffusion ability of solid and liquid phase of aluminum alloy. A new type of nonequilibrium diffusion interfacial structure is constructed at the interface of stainless steel and aluminum alloy.III. Compound mechanism of plastic and brittle fracture interface is formed in the shear fracture interface. That also means that the bonding property is not the same along the interface. Khan et al. [12] studied the effect of transient liquid-phase bonding variables on the properties of a micro-duplex stainless steel by using a Cu and an Fe-B-Si alloy interlayer. They concluded that:

I. Bonds made using the Cu interlayer show microstructural homogeneity with grain structure continuous with that of the parent alloy with an increase in bonding time.II. The joint properties of bonds made using the Cu interlayer are similar to that of the parent alloy due to good compositional homogeneity, better / phase ratio and less grain growth within the joint region compared to bonds made using the FeBSi interlayer.The pitting corrosion resistance of joints made with the Cu interlayer shows the least change in percentage weight loss when compared with bonds made with the FeBSi interlayer2.1 Diffusion in MetalsDiffusion is a basic metallurgical process which results from more or less random motion of individual atoms. Because of thermal energy, the atoms in a metallic crystal are in constant motion around their equilibrium lattice sites. Occasionally as a result of this motion, an atom will jump to a neighboring site. At room temperature, the frequency with which any atom makes a move to a neighboring site is usually small. However as the temperature increases, the atom jump frequencies increase, with a net rate of atomic migration eventually becoming large enough to provide readily observable effects, including transport of atoms over considerable distances and appreciable changes in chemical composition.2.1.1 Diffusion MechanismsDiffusion is a basic metallurgical process which results from more or less random motion of individual atoms. Because of thermal energy, the atoms in a metallic crystal are in constant motion around their equilibrium lattice sites. Occasionally as a result of this motion, an atom will jump to a neighboring site. At room temperature, the frequency with which any atom makes a move to a neighboring site is usually small. However as the temperature increases, the atom jump frequencies increase, with a net rate of atomic migration eventually becoming large enough to provide readily observable effects, including transport of atoms over considerable distances and appreciable changes in chemical composition

2.1.2 Basic One-Dimensional Equation for Diffusion Flux

Theories of diffusion based on the simple idea of atom jumps by diffusion mechanisms are very successful in relating a wide variety of diffusion effects to each other and to atomic parameters. The diffusion equations are obtained by the random walk ideas.2.1.2.1 Fick's First Law

Adolf Fick in 1855 theoretically proposed an equation for diffusion in solutions which was famed later as Fick's first law, given by [7]:

Equation 21Where J is the flux of diffusing species (number of atoms crossing unit area normal to the diffusion direction per unit time), D is the tracer diffusion coefficient (usually expressed in cm2/sec) and C/x is the concentration gradient.

For the flux of diffusing species in the direction x, Jx (number of atoms crossing unit area normal to the diffusion direction per unit time) with the tracer diffusion coefficient D* (usually expressed in cm2/sec), the concentration C of the diffusing species (atoms per unit volume) and v the atomic drift velocity in the x direction from driving forces acting on the atoms, the expression is given by [7]:

Equation 22D* and v depend on the jump frequencies and atom jump distance.

In all cubic crystals, diffusion is isotropic, as can be demonstrated from a consideration of crystal symmetry. Therefore equation 2-1 is valid for planner diffusion in every direction in these crystals.

2.1.2.2 Fick's Second Law

Adolf Fick in 1855 theoretically proposed an equation for diffusion in solutions which was famed later as Fick's first law, given by [7]:

Basic equation for planar diffusion is given by equation 2-1. For the flux of diffusing species in the direction x, Jx (number of atoms crossing unit area normal to the diffusion direction per unit time) with the tracer diffusion coefficient D* (usually expressed in cm2/sec), the concentration C of the diffusing species (atoms per unit volume) and v the atomic drift velocity in the x direction from driving forces acting on the atoms, the expression is given by:

Equation 23D* and v depend on the jump frequencies and atom jump distance.

In all cubic crystals, diffusion is isotropic, as can be demonstrated from a consideration of crystal symmetry. Therefore equation 2-1 is valid for planner diffusion in every direction in these crystals.

For planar diffusion [8] along any crystallographic axis x in a cubic system, with v = 0, other quantities independent of position, and the diffusant originally confined to a very thin layer at x = 0, the concentration C of diffusant after time t is given by:

Equation 24Where S0 is the original concentration per unit area.2.2 Methods of Measuring Diffusivity

There are two common methods of measuring diffusivity in substitutional solid solutions.

a) Grube-Jedele method

b) Matano-Boltzmann method

2.2.1.1 Grube-Jedele Method

Solution to the equation 2-4 for the case of a diffusion couple consisting originally of two alloys of the elements A and B, one having the composition CA1 (atom fraction) and the other having the composition CA2 at the start of the diffusion process, is [7]:

for - < x < Equation 25Where CA is the composition or atom fraction at a distance x from the weld interface, t is the time and is the diffusivity. The symbol represents the error function, or the probability integral with the argument. This function is defined by the equation:

Equation 26If a diffusion couple has been maintained at some fixed temperature for a given period of time t so that diffusion can occur, then a single determination of the composition at an arbitrary distance x from the weld permits the determination of the diffusivity.

2.2.1.2 Matano-Boltzmann method

In Matano-Boltzmann method [7], the diffusivity is assumed to be a function of concentration, which requires the solution of Fick's second law in the form:

Equation 27This method uses a graphical integration. The first step, after the diffusion anneal and chemical analysis, is to plot a curve of concentration versus distance along the diffusion bonded samples measured from a suitable point of reference, say from one end of the couple. The second step is to determine that cross-section of the bar through which there have been an equal total fluxes of the two atomic forms (A and B). This cross-section is known as Matano interface and lies at the position where areas A1+A2+A3=A5 in figure 2-1. the position of the Matano interface is determined graphically, but, in general, it has also been experimentally determined that, in the absence of porosity, the Matano interface lies at the position of the original weld. Once the Matano interface is located, it serves as the origin of x coordinate.

The Boltzmann solution of Fick's equation is

Equation 28Where t is the time of diffusion, CA is the concentration in atomic units at a distance x measured from the Matano interface, and CA1 is the concentration of one side of the diffusion couple at a point well removed from the interface where the composition is constant and not affected by the diffusion process. Diffusivity at a particular concentration CA, two quantities should be evaluated first from the figure 2-1. The first is the derivative x/CA, the reciprocal of the slope of the penetration curve at CA. thye other quantity is the integral, the integration limits of which are CA1 and CA. this integral corresponds to the area under the curve between CA1 and CA. Evaluation of this area by a graphical method like Simpson's rule yields the value of the integral. The diffusivity can be evaluated accordingly,

(1/slope at CA) (area from CA1 to CA)29

Figure 21 Boltzmann-Matano geometry for diffusion couple2.3 Determination of Intrinsic DiffusivitiesDetermination of intrinsic diffusivities can be done by using the data of composition versus distance from Matano interface [7]. First the velocity v of some refractory markers placed at the interface between the two materials in the diffusion couple is to be determined in terms of marker displacement and the time of diffusion t. experimentally, it is found that the markers move in such a manner that the ratio of their displacement squared to the time of diffusion is a constant. Thus

x2/t = k

Equation 210where k is a constant. The marker velocity is, accordingly,

v = x/t = k/2x

Equation 212but k equals x2/t, so that

v = x/2t

Equation 213Now from Darken's equations, we know that

Equation 214 and

Equation 215By inserting the values of, CA, CB, v and CA / x, i.e., the slope of the curve, in the equations 3-13 and 3-14, the solution of this pair of equation yields the value of DA and DB. 2.4 Diffusion BondingDiffusion bonding, as a subdivision of both solid-state welding and liquid-phase welding, is a joining process wherein the principal mechanism is interdiffusion of atoms across the interface. Diffusion bonding of most metals is conducted in vacuum or in an inert atmosphere (normally dry nitrogen, argon or helium) in order to reduce detrimental oxidation of the faying surfaces. Bonding of a few metals which have oxide films that are thermodynamically unstable at the bonding temperature (e.g. silver) may be achieved in air. The International Institute of Welding (IIW) has adopted a modified definition of solid-state diffusion bonding, proposed by Kazakov [9].

Diffusion bonding of materials in the solid state is a process for making a monolithic joint through the formation of bonds at atomic level, as a result of closure of the mating surfaces due to the local plastic deformation at elevated temperature which aids interdiffusion at the surface layers of the materials being joined.2.4.1 Types of Diffusion Bonding1) Solid state diffusion bonding

2) Transient liquid phase(TLP) diffusion bonding

2.4.1.1 Solid-State Diffusion Bonding

Solid-state diffusion bonding is a process by which two nominally flat interfaces can be joined at an elevated temperature (about 50%-90% of the absolute melting point of the parent material) using an applied pressure for a time ranging from a few minutes to a few hours.

2.4.1.1.1 Mechanism of Solid-State Diffusion Bonding

The mechanism of solid-state diffusion bonding can be classified into two main stages [9]. During the first stage, the asperities on each of the faying surfaces deform plastically as the pressure is applied. These asperities arise from the grinding or polishing marks that have been produced in the surface finishing stage. The micro plastic deformation proceeds until the localized effective stress at the contact area becomes less than the yield strength of the material at the bonding temperature. In fact, initial contact occurs between the oxide layers that cover the faying surfaces as shown in figure 4-1.

Figure 22 Stages of solid state diffusion bonding [9]As the deformation of asperities proceeds, more metal-to-metal contact is established because of local disruption of the relatively brittle oxide films which generally fracture readily. At the end of the first stage, the bonded area is less than 10% and a large volume of voids and oxide remains between localized bonded regions. In the second stage of bonding, thermally activated mechanisms (creep and diffusion) lead to void shrinkage and it increases further the bonded areas.

2.4.1.1.2 Modeling of Solid-State Diffusion Bonding

All attempts at modeling diffusion bonding have two main aims. The first is to optimize the process variables e.g. surface finish, bonding temperature, pressure and time so that the proper bonding conditions for a particular material can be identified. Secondly, a model attempts to obtain a reasonable and profound understanding of the mechanisms involved and their relative contributions not only for different bonding conditions but also for different materials being joined. 2.4.1.1.3 Advantages of Solid-State Diffusion Bonding1. The process has the ability to produce high quality joints so that neither metallurgical discontinuities nor porosity exist across the interface.

2. With properly controlled process variables, the joint would have strength and ductility equivalent to those of the parent material.

3. Joining of dissimilar materials with different thermo-physical characteristics, which is not possible by other processes, may be achieved by diffusion bonding. Metals, alloys, ceramics and powder metallurgy products have been joined by diffusion bonding.4. High precision components with intricate shapes or cross sections can be manufactured without subsequent machining. This means that good dimensional tolerances for the products can be attained. 5. Apart from the initial investment, the consumable costs of diffusion bonding are relatively low as no expensive solder, electrodes, or flux are required (although the capital costs and the costs associated with heating for relatively long times may be high). 6. Diffusion bonding is free from ultraviolet radiation and gas emission so there is no direct detrimental effect on the environment, and health and safety standards are maintained.

2.4.1.1.4 Problems with Solid-State Diffusion Bonding

The aim in diffusion bonding is to bring the surfaces of the two pieces being joined sufficiently close that interdiffusion can result in bond formation. There are two major obstacles that need to be overcome in order to achieve satisfactory diffusion bonds.

Even highly polished surfaces come into contact only at their asperities and hence the ratio of contacting area to faying area is very low.

In most metals, the presence of oxide layers at the faying surfaces will affect the ease of diffusion bonding. For some metals and alloys, their oxide films either dissolve in the bulk of the metal or decompose at the bonding temperature (e.g. those of many steels, copper, titanium, tantalum, columbium and zirconium), and so metal-to-metal contact can be readily established at the interface. The joining of these materials is relatively straightforward and is not included in this review. However, if the oxide film is chemically stable, as for aluminum-based alloys, then achieving a metallic bond can be difficult.

In practice, because of inevitable surface roughness and also the presence of oxide layers on most faying surfaces, it is neither feasible to bring together the surfaces of two pieces within interatomic distances nor to establish complete metal-to-metal contact by simply putting two pieces together. Shirzadi et al. [3] have discussed various aspects of the effects of surface oxides on interface morphology and bond strength, and a summary of the existing approaches used to overcome the oxide problem. The bonding technique which is adopted for bonding of SS to Al alloy is TLP diffusion bonding, so further discussion is about TLP diffusion bonding.2.4.1.2 Transient Liquid Phase (TLP) Diffusion Bonding

Liquid-state diffusion bonding relies on the formation of a liquid phase at the bond line during an isothermal bonding cycle. This liquid phase then infuses the base material and eventually solidifies as a consequence of continued diffusion of the solute in to the bulk material at constant temperature. Therefore, this process is called Transient Liquid Phase (TLP) diffusion bonding. Despite the presence of a liquid phase, this process is not a subdivision of brazing or fusion welding as the formation and annihilation of the liquid phase occurs at a constant temperature and below the melting point of the base material. The liquid phase in TLP diffusion bonding generally is formed by inserting an interlayer which forms a low melting point phase, e.g. eutectic or peritectic, after preliminary interdiffusion of the interlayer and the base metal at a temperature above the eutectic temperature. Note that the liquid phase could, alternatively, be formed by inserting an interlayer with an appropriate initial composition e.g. eutectic composition which melts at the bonding temperature.

The diffusion rate in the liquid phase enhances dissolution and/or disruption of the oxide layer and so promotes intimate contact between the faying surfaces. Therefore, the presence of a liquid phase reduces the pressure required for TLP diffusion bonding in comparison with solid-state diffusion bonding and may overcome the problem associated with solid-state diffusion bonding of the materials with a stable oxide layer [3].

Achieving high integrity joints with minimal detrimental effects on the parent material in the bond region and also the possibility of joining metal matrix composites (MMC) and dissimilar materials are the most promising features of TLP diffusion bonding.2.4.1.2.1 Theoretical Aspects of Transient Liquid Phase (TLP) Diffusion Bonding

Figure 4-2 shows a simple eutectic phase diagram where A represents a pure parent material and B is the diffusing solute (i.e. the originally solid interlayer) with limited solubility in A. Basically, TLP diffusion bonding consists of two major stages as follows: Dissolution of the base metal;

Isothermal solidificationThe dissolution stage can be divided into two hierarchical sub-stages in which filler metal melting is followed by widening of the liquid zone. However, if the melting process occurs as a result of interdiffusion of A and B, then melting of the interlayer and widening of the liquid phase may occur simultaneously. According to the phase diagram, equilibrium in the liquid can be established by dissolution of A atoms into the supersaturated B-rich liquid to decrease its concentration to CL as shown in figure 2-3. During this stage, homogenization of the liquid phase continues and the width of the liquid zone increases until the composition profile in the liquid phase levels out, i.e. diffusion in the liquid ceases.The rate of this homogenization is controlled mainly by the diffusion coefficient in the liquid phase and, therefore, this stage takes a short time to be completed. In the next stage, isothermal solidification occurs as B atoms start to diffuse into the solid phase, and the liquid zone shrinks in order to maintain the equilibrium compositions of CL and CL at the solid/liquid moving boundaries. The interdiffusion coefficient in the solid phase () controls the rate of solidification and, because diffusivity in the solid is low, the annihilation of the liquid phase is very slow compared to the initial rapid dissolution stage for which diffusivity in the liquid controls the rate of the reaction.

Figure 23 Stages of liquid phase diffusion bonding [10]2.4.1.2.2 Isothermal Solidification Time

If the amount of solute diffused into the base metal during the heating and dissolution stages is ignored (i.e. when rapid heating to the bonding temperature is employed), at the end of the isothermal solidification stage all the solute in the interlayer controlling the solidification process has diffused into the base material, then isothermal solidification time can be estimated by following expression [6]:

Equation 216Where C0 is the solute concentration in the interlayer material, W0 is the interlayer thickness, CL is the maximum solute concentration in the solid at the interface from phase diagram and Cm is the solute concentration in the base material.

2.4.1.2.3 Homogenization Time

At the end of the homogenization stage, the maximum solute concentration is attained in the centre line (x = 0), thus at homogenization time (tH), the solute concentration is given by [6]:

Equation 217Where Cm, CL and D has the same meaning as previously. h is half the width of the bonding interface and tH is the homogenization time.2.5 Intermetallic Compound Formation during Diffusion Bonding

Sometimes new phases different from base metals are found in the bonded zone when diffusion bonding between dissimilar materials takes places [11]. The brittle intermetallic compounds formed could weaken the bonding performance. Therefore, it is important to study the formation and growth mechanism of intermetallic compounds at the interfaces in order to control the process during the diffusion bonding of dissimilar materials.

To study the mechanism of formation of intermetallic compounds during the diffusion bonding of dissimilar materials, understanding the law of early interdiffusion at the interface is necessary. Here early means the moment when intermetallic compounds appear. In contrast to liquid-state phase changes, homogeneous nucleation seldom happens during solid state phase changes, but heterogeneous nucleation at impurities, grain boundaries and dislocations, etc. occurs. Many investigations indicate the probability of precipitation of new phase depends on kinetic factor, state of diffusing atom, thermodynamic driving force, reaction temperature, contact region and so forth. But by far, no theory can predict and interpret accurately the formation and growth of intermetallic compounds during diffusion bonding. Researchers gradually agree on that primary stage of the formation and growth of intermetallic compounds during diffusion bonding mainly contains the following stages: both mating halves interdiffuse at different rates, then supersaturated solid solution is formed. The crystal nuclei of new phases are formed at defects where concentration of diffusing element is high. Crystal nuclei of intermetallic compound grow along the interface; the regions of grown intermetallic compound connect and grow longitudinally as usual. After that crystal nuclei of a second intermetallic compound are formed at the interface and grow up 2.6 Diffusion Bonding of Al Base Alloys to Stainless SteelThe joining of aluminum and aluminum base alloys to themselves and to other metals has long created problems because of the tenacious layer of surface oxide which is always present. The difficulties become more acute when melting of the components to be joined is not an option. Some advanced materials cannot be welded by conventional techniques because the high temperatures involved would destroy their properties. For such materials, diffusion bonding is an attractive solution because it is a solid state joining technique, which is normally carried out at a temperature much lower than the melting point of the material.For joining aluminum alloys, insertion of a thin copper or zinc interlayer will allow a low melting point eutectic phase to be formed at a temperature about 100 to 250C lower than the melting point as shown in Figure 2-4 and 2-5. It is anticipated that this technique can be used for joining dissimilar metal combinations, metal matrix composites and possibly nickel based materials. Figure 24 Al-Cu phase diagram [12]

Figure 25 Al-Zn phase diagram [12]Bimetallic sheets which consist of dissimilar metal components have been widely used in many industrial fields due to their excellent mechanical and functional properties. Several methods such as explosion, rolling, weld overlay, inversion casting and laser cladding have been developed for producing bimetallic sheets. Generally, the commonly used bilayer sheets are stainless steel and aluminum or carbon steel, and they are produced by cladding two metallic materials directly, which has some drawbacks either in engineering or economics. Due to the fact that mechanical properties and deformation and fracture behavior of bi-material plate are mainly dependent on their interfacial bonding of the cladding materials, the selection of processing variables affecting interfacial structure, diffusion distance and compound formation and morphology are critical. In order to improve the bonding of stainless steel on aluminum and carbon steel, a Cu or Zn interlayer material will be incorporated. The advantage of using the interlayer is that the Al-Cu and Al-Zn eutectic in semisolid state at processing temperature can promote metallurgical bonding between the two cladding materials. This project will cover the effects of cladding temperature and time on the interfacial bonding of the stainless steel/aluminum through the Cu and Zn interlayer. Optimum conditions for a sound joint between Al alloy and austenitic stainless steel without producing intermetallics will be searched in this project. Phase diagram of Al-Fe system is shown in Figure 2-6. Chemical composition, some important physical and mechanical properties of austenitic stainless steel 316 and Al-Mg alloy are given in table 2-1, 2-2 and 2-3 respectively.

Figure 26 Al-Fe phase diagram [12]Table 21 Composition of SS 316 and Al alloy 5182 (wt.%)[18][19]Alloy typeCrNiMoCMnSPMgAlSiFe

Al alloy 5182----0.35--4.5Bal.0.1-

SS 31617.5122.50.0320.030.04--0.75Bal.

Table 22 Some physical properties of SS316 and Al alloy 5182

Alloy typeMelting point (C)Density (g/c.c)Elastic modulus (GPa)Mean CTE at 538C (m/m)Thermal conductivity at 500 C (W/m.K)Sp. Heat 0-100 C (J/Kg.K)

SS 3161390-1440[13]8.027 [13,14]193 [13,14]17.5 [13,14]21.5 [13,14]500 [13,14]

Al-Mg alloy 5182630-640 [15]2.7 [16]75 [16]23.1 [17]123[16]903 [17]

Table 23 Some mechanical properties of SS316 and Al alloy 5182

MaterialYield strength(MPa)Tensile strength (MPa)% elongationHardness vicker

VHN

SS 316L290 [13,14]558 [13,14]40 [13,14]180

Al-Mg alloy 5182195 [16]420 [16]22 [17]60

3 Experimental3.1 Samples Preparation A commercial Al-Mg alloy 5182 as cast ingot and an austenitic stainless steel (AISI 316) hot rolled bar was taken as starting material. The samples of the stainless steel were cut by milling machine to the dimensions of 20(10(3 mm3(Figure 3-1) and 15(12(4 mm3(Figure 3-2). Al alloy samples were cut by thin abrasive wheel cutter to the dimensions as that of stainless steel counterpart sample. One large face of each samples was polished to give a uniform finish with 80, 320, 500, 800, 1200 grade emery paper. Prepared samples were stored in alcohol to avoid excessive oxidation.

Figure 31 An Al alloy and SS couple aftercutting and polishing

Figure 32 A stainless steel sample size used for diffusion bonding3.2 Formation of interlayer

Interlayer between the two materials was formed by either Cu or Zn. Both the interlayer materials were placed by different techniques discussed below.

3.2.1 Formation of Cu Interlayer

Three set of samples of the dimensions 20(10(3 mm3 were incorporated by Cu interlayer. To place the Cu interlayer, a 99.99% pure Cu foil of the thickness 200m was cut to the dimensions as that of the polished surface of the base materials samples. One set or couple was prepared with the Cu interlayer of 200m thickness. The foils for other two sets were then carefully ground on both sides by 800 grit size emery paper to an average thickness of 100m (Figure 3-3). One such prepared foil was placed between each set of the two base materials samples as shown in Figure 3-4.

Figure 33 Cu foil used as interlayer material

Figure 34 Cu foil placed between an SS and Al sample3.2.2 Formation of Zn interlayer

Two sets of samples were bonded with Zn as interlayer. As Zn is a brittle material and cannot be shaped to thin foil, so a Zn layer was formed by following two methods:

a) Physical vapor deposition

b) Electrochemical deposition

3.2.2.1 Physical Vapor Deposition (PVD) of Zn

PVD of Zn was done on one surface of each type of sample in one couple. PVD of Zn was carried out in thermal evaporation vacuum chamber. The apparatus is shown in figure 99% pure Zn chips were placed on a tungsten boat. The boat was be heated by electric resistance.

Each sample in the couple was dried by hot air drier and weighed by a highly sensitive microbalance. Then all the faces of both samples were masked carefully by scotch tape except the prepared face. Both samples were clamped in a fixture just above the tungsten boat with their prepared surfaces normal downward. This configuration was thought to give a maximum flux faced by the sample surface and a maximum uniformity of the deposit that could be achieved by the apparatus used.

Figure 35 PVD apparatus used for Zn deposition

Figure 36 Sample and evaporant position in the PVD apparatusAfter placing the target material and the substrate at their proper places, the chamber was closed and sealed by silicon grease. Then rotory pump was switched on to create a rough vacuum (10-2 torr). After achieving rough vacuum, diffusion pump was managed to take charge for creating high vacuum. When a vacuum level of 5(10-4 torr was achieved, the boat current was turned on. Operating voltage was 20V. The current was slowly increased to raise the temperature of the boat. At a current of 3A, Zn started to evaporate. While maintaining the current, all the Zn in the boat was slowly evaporated. Boat current was switched off and the boat was allowed to cool. Then vacuum pumps were isolated from the vacuum chamber and switched off. Chamber vacuum was released and samples were taken out of the chamber. The masking tape was removed carefully without any damage to the Zn coating formed. Alcohol swabbing of unmasked surfaces was done by a cotton wad to remove any adhesive remained after removing masking tape. Then samples were dried by a hot air drier. Samples were again weighed by microbalance. Gain in weight was the weight of Zn deposited. Assuming uniform deposition, the thickness of the Zn layer on each sample was calculated as follows:

Initial weight of SS sample = SW1 = 7.35670 g

Final weight of SS sample = SW2 = 7.38560 g

Weight gain = SW1- SW2 =SW = 0.0289g

Density of pure Zn = = 7.14g/cm3Surface area coated = A=19.7(10mm2 = 197mm2 =1.97cm2Thickness of layer on SS sample = t1

Density = mass/volume

= SW/ (A(t)

and t1 = SW/(A()

t1 = 0.0289/ (1.97(7.14)

t1 = 0.00205cm = 20.5m

Initial weight of Al sample = AW1 = 1.58990g

Final weight of Al sample = AW2 = 1.698250g

Weight gain = AW1- AW2 =AW = 0.02835g

Thickness of layer on Al sample = t2

t2 = AW/ (A()

t2 = 0.02835/ (1.97(7.14)

t2 = 0.00201cm = 20.1m

Now the total thickness t of interlayer is the sum of the thickness of layers formed on both sample. So

t = t1 + t 2t = 40m approx.

3.2.2.2 Electrochemical Deposition of Zn

Electroplating of Zn was also carried out on both the samples. Composition of the solution used is shown below:

ZnCl2 20g/100ml

H3BO3 4g/100ml

Rough polishing of the surfaces to be coated with Zn of both the samples was done (800 grit size) to ensure some sticking. Fine polishing causes a continuous stable oxide layer on both the materials (Cr2O3 on SS and Al2O3 on Al) which hinders the adhesion of deposit with the base material. Then the surface of SS was cleaned by acetone and activated by 10% H2SO4 solution. Al surface was cleaned by 2% HF solution.

Electroplating was done at a current density of 50mA/cm2 for 20 minutes while maintaining the bath temperature at 45oC. Thickness of the interlayer formed was calculated by the same method as that for PVD interlayer. Total interlayer thickness was calculated to be approximately 20m in this case.3.3 Coupling the Samples and Heat Treatment

After surface preparation, samples of both the materials were coupled with a different method to form a set or couple. After coupling each couple was given different heat treatment conditions. Interlayer type and thickness, cold welding or coupling method and pressure (if measurement was available), heat treatment temperature, furnace atmosphere and holding time for each sample are given in Table 3-1.

Set #1 was wrapped with Kanthal wire and pressed in a bench vice for 24 hours. Then it was heat treated in a lower vacuum (410-3 torr) quartz tube for 40 minutes and then air cooled. Set # 2 and 3 (Table 3-1) were pressed in a hydraulic press under a maximum pressure of 50 MPa for 4 hours and then fastened in a screw fixture. The steel fixture is shown in Figure3-7. Figure 37 (a) Steel fixturefor holding the samples (b) a diffusion couple in the fixtureThe fixture holding the samples couple was placed in a high vacuum Ruhsttrat D-3406 Bovenden tube furnace (Figure3-8 and 3-9) for heat treatment. Heating was done at a rate of 10 C/min. After the specified holding time, samples were air cooled to room temperature. Time and temperature of heat treatment was different for both the sets. Time-temperature and time-furnace atmospheric pressure profile during the whole heating and cooling cycle of set# 2 and 3 is shown in Figure 3-10.

Figure 38 (a) Bovenden vacuum furnace used for heat treatment (b)furnace controllers

Figure 39 Time vs. temperature and pressure of high vacuum furnace.Set # 4 with PVD coated Zn interlayer was wrapped with Kanthal wire tightly and then fastened in a bench vice for 24 hours to form some cold weld. Then it was heat treated in a quartz tube, low vacuum furnace. After specific holding time, sample was air cooled to room temperature. Set # 5 having electroplated Zn interlayer was pressed in a hydraulic press under a maximum pressure of 50 MPa for 4 hours. Then it was placed in a screw fixture to keep some pressure on the samples. Heat treatment was done in a chamber furnace (Figure 3.11) with ordinary atmosphere. After heating for specific time, sample was air cooled to room temperature. After cooling, samples were removed from the fixture and saved for characterization of the bonds.

Figure 310Muffle furnace used for heat treatment3.4 Characterization of the bonded interface

Bonded samples were mounted in cold setting resin and then ground and polished with emery paper # 320, 500,800,1000,1200,2400 and then finally with diamond paste. Etching was done with Kellers etchant (95 ml H2O, 2.5 ml HNO3, 1.5 ml HCl and 1 ml HF). Microstructural investigations were done using scanning electron microscopy (SEM) and the analysis of phases was carried out with an energy dispersive system (EDS). Hardness of each phase or layer in the microstructure was determined by Vicker microhardness tester. Figure 311 (a) Set # 3 (b) Set # 5 after diffusion bonding and mountingTable 31 Interlayer type, thickness and heat treatment conditions for various sets of samples

Set#Interlayer material with thickness in mMax. temperature (C)Furnace atmosphereHolding time (min)Cold welding method

1Cu

200560Vacuum

410-3 torr40By bench vice

2Cu

200555Vacuum

5010-6 torr120By hydraulic press 50MPa applied

3Cu

100550Vacuum

5010-6 torr60By hydraulic press 50MPa applied

4PVD Zn

40420Vacuum

410-3 torr30By bench vice

5Electroplated

Zn

20400Room atmosphere60By hydraulic press 50Mpa applied

4 Results and Discussion4.1 ResultsNo bonding was seen in set # 1. Cu interlayer remained as such and bonded only to Al side. SS surface was oxidized slightly.

During heat treatment of set # 2, vacuum conditions were improved than that for set # 1 and a known pressure of 50MPa (much more than a table vice) was applied for cold welding. A weak bond was observed. Bond was collapsed during metallography. Bonded surface of both the samples after opening of the bond is shown in figure

Figure 41 Al sample surface after collapse of the bond in set # 2

Figure 42 SS sample surface after collapse of the bond in set # 2In set # 3, good bonding was observed. Bonding interface was characterized by SEM and phases were identified by EDS analysis.4.1.1 SEM microscopy

SEM micrographs of the bonding interface are shown in Figure 4-3 and 4-4. Three distinct layers were observed in the bond region as shown in Figure 4-5. In the microstructure of Al alloy after bonding, two types of phases can be observed. White regions are of CuAl2 ( phase) while dark regions are of Al+ as shown in Figure 4-6.

Figure 43 SEM micrograph of bonding interface of set # 3 in secondary electron mode

Figure 44 SEM micrograph of bonding interface of set # 3 using back scattered electron mode

Figure 45 Different layers formed during diffusion bonding of set # 3

Figure 46 Al alloy microstructure after heat treatment. Dark regions are of Al+ while white regions are CuAl24.1.2 EDS spectrum of various regions in set # 3

EDS spectrums of various regions and phases in the microstructure of set # 3 are shown in Figure 4-7 to 4-14. In the layer 1(Figure 4-7), Fe, Cr and Ni contents were maximum while small amount of Al can also be seen. Oxygen peak was due to Al2O3 formed on the surface after polishing.

Figure 47 EDS analysis of layer1EDS spectrum of white region in layer 2 shows dominant peaks of Fe, Cr and Al (Figure 4-8).

Figure 48 EDS analysis of white region in layer 2In the spectrum of rod type phase in layer 2, peaks of Fe, Cr, Cu, Ni and Al were observed (Figure 4-9).

Figure 49 EDS analysis of rod type phase in white contrast in layer 2Figure 4-10 shows the presence of large amount of Al and Fe in layer 3. Small amount of Ni, Cu and Cr was also observed.

Figure 410 EDS analysis of layer 3Just close to layer 3, there was a gray region on Al side. EDS analysis (Figure 4-11) shows large amount of Al, Cu, and Mn. Mg, Fe, Ni and Cr were also found. Interesting thing in this analysis is that Mn (7.24 wt.%) was only found in this region throughout the whole sample. Mg contents were also found maximum (5.5 wt.%) in this region.

Figure 411 Grey contrast in Al near layer 3Figure 4-12 shows presence of Al and Cu in white region at the grain boundaries of dark grains in Al base material.

Figure 412 White contrast in Al base material ( phase)

Figure 4-13 and 4-14 show presence of Al and Cu in black and white areas in the large dark grains of Al alloy. White regions are richer in Cu than blacker ones.

Figure 413 EDS analysis of black contrast within grayish grains of Al alloy

Figure 414 EDS analysis of white contrast within grayish grains of Al alloyTo have an overview of composition of different layers and phases, Table 4-1is quite helpful.Table 41 Composition of various layers, regions and phases formed during diffusion bonding

RegionFeCrNiAlCuMgMnSi

Layer 1(point1)49.6418.9520.469.081.87---

Layer 1 (point2)57.6719.763.278.1311.18---

Layer 2 rod type in white contrast32.8815.6332.4915.513.49---

Layer 2 white region33.1117.043.3539.826.68---

Layer 3 (point1)24.117.414.7161.322.45---

Layer 3 (point2)20.286.287.0665.201.18---

Gray contrast in Al near layer 35.204.211.2320.9655.645.517.24-

White contrast in Al---45.5154.49---

White contrast within grayish grains in Al---93.066.250.69--

Black contrast within grayish grains in Al---95.723.950.33--

EDS line scan results of Al, Fe, Cr, Ni and Cu for an SEM image at a magnification of 200X are shown in Figure 6-15.

Figure 415 EDS line scan results of Al, Fe, Cr, Ni and Cu4.1.3 Microhardness test results

Microhardness test results for various layers, regions and phases are shown in the Table 4-2. Maximum hardness was observed for white areas at the grain boundaries of dark grain in Al alloy. Minimum hardness was also observed in the Al alloy which was of dark grain. Hardness of stainless steel was in the range of 145-180VHN at various points. Table 42 Microhardness test resultsRegionAverage Vicker hardness (VHN)

SS base material148

SS just close to layer1167

Layer 1246

White areas in layer 2315

Dark areas in layer 2382

Layer 3417

Dark areas in Al134

White areas in Al466

4.2 Observations in other samples

Heat treatment of set # 4 was carried out in rough vacuum. No bonding was seen. Zn on SS surface was found oxidized during examination after heat treatment. Zn adjacent to Al surface had formed a eutectic with Al which did not wet the SS surface but solidified on the Al surface.

Figure 416 Al and SS samples with Zn interlayer (set # 4) after heat treatment

Heat treatment of set # 5 was done in ordinary room atmosphere because available high vacuum furnace had a very slow cooling rate. A good bond was observed by naked eye. SEM micrograph was done to characterize the microstructure of the bonding interface and EDS analysis was done to identify the phases.

4.3 Discussion

Set # 1 showed no bonding. Holding time was insufficient to dissolve the 200m thick Cu interlayer. So interlayer thickness was to be reduced and holding time was required to optimize. Small oxidation on SS surface was observed. Therefore, vacuum was improved in the further experiments with Cu interlayer.

Set # 2 was annealed in a high vacuum to eliminate oxidation problems seen during bonding of set # 1. Cu interlayer thickness was reduced from 200m to 100m. It can be thought that holding time was longer in this experiment because a weak bond was observed. Longer holding time caused diffusing atoms to travel a larger distance and acquired a stoichiometry according to chemical formulae of different intermetallics. Formation of intermetallics was excessive. But evidence of formation of Al-Cu eutectic was also found as there was a small amount of flown out liquid material which had solidified at the mating line. Diffusion flux from the interface to Al side was more than that towards it. This causes some small voids at the interface. Formation of intermetallic compounds in large amount and voids causes a drastic decrease in the bond strength.

Set # 3 had shown a good compact bond between the two materials. Bonding interface was 30m wide. Three distinct layers were identified in the bonding region during SEM microscopy. The layers were found to form only on Al side. EDS analysis showed that total amount of Fe, Cr, Ni, was nearly 89 wt.% while Al and Cu were approximately 11wt.%. This composition is of (Fe, Cr, Ni)3(Al, Cu) complex intermetallic compound. At some points in layer 1, Al and Cu contents were totally 19wt.% approximately. This composition assured the presence of (Fe,Cr,Ni)(Al,Cu) intermetallic. High hardness of layer 1 (VHN246) as compare to the adjacent region of SS was due to presence of Fe3Al. In layer 2 there were two types of phases, one was black rod like phase while second was white phase. The black phase was found to have a composition of (Fe,Cr,Ni)(Al,Cu). White regions in layer 2 had a composition of (Fe,Cr,Ni)(Al,Cu)2. It can be seen that hardness is still increasing as one travels from SS to Al side. Increase in hardness was due to hard intermetallics of Fe and Al. Layer 3 composed of FeAl3 with small amount of free Al (from Al-Fe phase diagram). Harness was observed to be VHN417 in layer 1.

Above results show a typical layer by layer formation of Al-Fe intermetallics which are richer in Fe on SS side and then amount of Al increases as Al base material is approached, i.e. Fe3Al, FeAl, FeAl2 and FeAl3. From the table, it is seen that Mn was found only in the grayish region in Al just close to layer 3. Small amounts of Fe, Cr, Ni and Mg was also found. Al6Mn, Al2Cu, Al3Mg2 and Cu2Mg intermetallics are likely to be formed in a highly dispersed form in -Cu because of large amount of Cu present at this region. White phase in Al base material was found to have Al and Cu in such an amount which is typical of Al2Cu ( phase in Al Cu phase diagram). Maximum hardness (VHN466) was observed in this region.

White and dark regions in the large grayish grains in Al side are rich in Al but each have a different amount of Cu. Microstructure is of -Al+Al2Cu. Darker areas contain less Cu as compare to white regions. Hardness of this region was lowest among all the phases (VHN134) because of presence Al but it is much more than that of Al alloy 5182 before heat treatment which was 80VHN.

Layer 1 was formed first which act as a barrier layer against the Al and Cu because no evidence of Al or Cu presence were found in SS side. Layer 2 which is a mix of FeAl and FeAl2 was formed later as a result of a diffusion of small amount of Fe into the Al-Cu liquid eutectic. This diffusion occurred prior to the formation of layer 1. Cr and Ni also came with Fe. Layer 3 was formed after layer 2. This layer act as a barrier layer for Fe, Cr, Ni because very small traces of only Fe were found beyond this layer in Al. Reason for small amount of Fe beyond layer 3 was high Fe flux when there was no barrier layer. So Fe had diffused to a larger distance but its large amount was consumed in the formation of layer 1, 2 and 3.

Small cracks which are perpendicular to the bonding interface were also formed in the layer 1 and 3. These very small cracks were due to volume change during formation of these layers. Moreover, these layers were in contact with a different type of materials (layer 1 with SS and layer 3 with Al) during cooling. But one thing which is fruitful is that no layer or phase was totally cracked. It means that good buffering between coefficients of thermal expansion of Al and SS was done by these layers.

Surface preparation plays an important role in bond strength. Set # 2 was finely polished (2400grit size) which results in a weak bond. On the other hand, set # 5 was grinded finally with 800 grit size.

5 SEM Operating Procedures

5.1 Power Up

1. Log In

2. Power on switch on front panel of beam

3. Record in log book: vacuum pressure of IP1, IP2, IP3

Filament chamber = top cylindrical column; holds the filament, column, etc. Sample chamber (SC) = large chamber under filament chamber where substrate is subjected to electron scanning Sample exchange chamber (SEC) = small antechamber used for loading wafers. 5.2 Loading Sample

1. Check that the stage is in home position, as listed on the stickers on the stage drivers.

2. Check all chambers are closed:

a. Check SC Air Lock valve is closed

b. Manual gate between SC and SEC is closed

c. Check SC/SEC switch is on SEC

3. AIR the SEC chamber

4. When SEC door releases, open it and load sample, then close SEC door.

5. EVAC the SEC chamber; wait until high vacuum is indicated on the panel

6. Open portal between SC and SEC using the manual gate

7. Carefully insert sample fixture snugly in the stage, unload sample, and remove probe from SC.

8. Close the manual gate

9. Make sure high vacuum is achieved in SC.

10. Open SC to expose filament to sample with the toggle switch on the panel.

11. Again, make sure high vacuum is achieved in SC 5.3 Operation 5.3.1 Initializing

o Flashing the filament vaporizes organic material that may have deposited on the filament, which will ultimately improve stability and imaging quality. However, emission current will fluctuate for approximately 30 minutes after flashing. To flash:

On the control panel, press the Flash standby button, it should blink

Press the Adjust button to its left

Observe maximum applied current and record in log book as the flash current.

o Press HV On switch

5.3.2 Imaging

o Selection of secondary electron detector PF2

As a guideline: Upper detector for WD < 18; Lower detector for longer WD or tilted specimen

o ABC = Automatic brightness/contract adjustment

Manually control brightness/contrast using image or MONIT

MONIT waveform view of brightness/contrast; containment of line around center is best; press this button again to exit this mode

o Initial focus using the focus pad

SEARCH FOCUS buttons

Adjust focus manually

o Adjust applied voltage: PF1

Total range is from 0.5 30 kV

For resist under thin metal, use 0.8-5.0 kV

o Adjust column after any change in applied voltage: PF3 o Adjust brightness, focus, and stigmation

5.3.3 More function keys

o PF3 - Column Adjust

Column adjustment requires practice and training for proper beam alignment.

o PF7 Auto data display = 20-30 range

o PF12 measure data use cursor knobs

5.3.4 Power down

Press HV Off to power down applied voltage

Lock SC chamber

Zero x, y, R, and tilt with values indicated on Exchange Position tag

Ensure high vacuum in SEC. Open manual SEC portal and carefully retrieve sample; close portal.

VENT SEC chamber

Open outer SEC door and remove sample from handle; close SEC door

Evacuate SEC wait until high vacuum

Power off front panel

Log Off 5.3.5 Image Capture

To capture an image:

1. Open the Quartz PCI software

2. Press single image on the toolbar or Acquire Single Image

3. Press a slow scanning mode (3 or 4) on the SEM panel

4. Press direct on upper right of SEM panel

5. After image is acquired, press stop acquisition on the toolbar

If you want to adjust contrast and brightness of the captured image:

1. Go to Process Histogram

2. On the graph, drag the center x-axis marker to the green portion of the histogram. To learn more about this feature, press Help located at the bottom right of this window.

To save the image you can use the Database. It is set up and ready for images. Or you can bypass the database by Exporting the image. Conclusions

From above results, it is concluded that:

Diffusion bonding of stainless steels and Al alloys is feasible if process parameters are tightly controlled. Heating and cooling rates are highly important to achieve a sound joint. Too slow heating rate increases the time span for which the material remain at high temperature. Same is true if cooling rate is slow. Slow cooling rate causes excessive formation of Fe and Al intermetallics.

Using a Cu interlayer facilitates a liquid phase formation which allows easy diffusion of Fe to Al side but Cu interlayer thickness should be optimum according to the Al part size to be joined with SS.

Fe and Al form intermetallics in layered form which are rich in Fe on stainless steel side and then gradual decrease in Fe contents and increase in Al contents towards the Al side.

Coefficient of thermal expansion gap between Al alloy and stainless steel can be successfully buffered by using Cu interlayer.

A rougher surface of both the materials gives a better joint than that by using finely polished surface.

Future RecommendationsDiffusion bonding by applying temperature gradient is a new technique for joining dissimilar materials. An experimental set up which can raise the temperature on one side rapidly while the other side can be maintained at a lower temperature will be highly beneficial. Moreover, a measured pressure should be applied to estimate the minimum deformation required to break the surface oxide layer. Vacuum conditions are not required to be high. An intermediate vacuum is sufficient if Al is to be joined with SS but the samples should be heated under such a pressure that assures the breakage of surface oxide layer. Bonding by using a Zn interlayer is required to be investigated throughly. Samples can be prepared for mechanical testing so that the mechanical properties of the bond between Al and stainless steel can be determined i.e. shear test, tensile test, fatigue test, interface toughness test, etc.

References

[1] H.W. Liu, C. Guo, Y. Cheng, X.F. Liu and G.J. Shao, "Interfacial strength and structure of stainless steel-semi-solid aluminum alloy clad metal," Mater. Lett., vol. 60, pp. 180, August. 2005.

[2] Wang Deqing, Shi Ziyuan and Qi Ruobin, "Cladding of stainless steel on aluminum and carbon steel by interlayer diffusion bonding," Scripta Materialia vol.56, pp. 369372, December. 2006.

[3] A.A. Shirzadi, H.Assadi and E.R. Wallach, "Interface evolution and bond strength when diffusion bonding materials with stable oxide films," Surface and Interface Analysis vol.31, pp. 609-618, 2001.[4] Wang Juan, Li Yajiang and Ma Haijun, "Study of diffusion bonding of Fe-28Al with austenitic stainless steel in vacuum," Vacuum vol.79, pp.4551, December.2004.[5] Wang Juan, Li Yajiang and Yin Yansheng, "Interface characteristics in diffusion bonding of Fe3Al with Cr18-Ni8 stainless steel," Journal of Colloidal and Interface Science vol.285, pp.201-205, January.2005.[6] T.I. Khan, M.J. Kabir and T. Padron, "Modelling the transient liquid phase diffusion bonding behavior of a duplex stainless steel using copper interlayers," Materials Science and Engineering A vol.385 pp.220228, June.2004

[7] Robert E. Reed-Hill, Reza Abbaschian, "Physical Metallurgy Principles", 3rd ed., PWS, pp. 369-380, 1994.

[8] H. I. Aarinson, "Diffusion," ASM, Metals Park, Ohio, pp.5, October.1972.[9] URL: http:// www.msm.cam.ac.uk/phase-trans/2005/Amir/bond.html

[10]

HYPERLINK "URL:%20http://%20www.twi.co.uk/j32k/getfile/c685a.html"

URL: http:// www.twi.co.uk/j32k/getfile/c685a.html

[11] P. He and D. Liu, "Mechanism of forming interfacial intermetallic compounds at interface for solid state diffusion bonding of dissimilar materials," Materials Science and Engineering A vol.437, pp.430, August.2006.[12] ASM Metals Handbook, Vol.3 Alloy Phase Diagrams 2.44, 2.56, 1994.[13] URL: http:// www.fergusonmetals.com/stainlesssteels[14] URL: http:// www.lenntech.com/stainless/properties

[15] W.F. Smith, Structure and Properties of Engineering Alloys, Mc-Graw Hill pp.188-313, 1993.[16] URL:http://www.efunda.com/materials/alloys/alloys_home/show_alloys_found.cfm

[17] URL: http://www.webelements.com/webelements/elements/text/Al/heat.html[18] http://www.huysindustries.com/articles/HuysArticle17.pdf[19] http://www.calfinewire.com/metals/tds/ss316.htm

Zn Pieces

Heating Boat

Samples to be coated with Zn

Al-alloy

Cu foil

SS

(b)

(a)

Al alloy

Bonding interface

Al alloy

Al alloy

Bond region

PAGE 5

_1248131445.unknown

_1248203441.unknown

_1248212445.unknown

_1248378932.unknown

_1248381997.unknown

_1251843004.xlsChart3

3280

11580

220100

312400

411500

511450

555450

555450

555400

555300

555250

555250

555200

555150

555150

555150

555150

533100

508100

490100

46480

42870

40960

40050

38840

37640

35340

34535

33835

32825

31725

31125

30325

29625

28225

27025

25925

24925

23925

19525

18625

17525

16425

15525

14320

13520

12520

11920

11220

10620

10025

9525

9125

8625

8325

7925

7525

7225

6925

6725

651890

631950

612010

592070

572130

552190

532250

522310

512370

492430

482490

4725550

462610

452730

442790

Temperature

Pressure

Time (minutes)

Temperature (OC) and Pressure (X10-7 torr)

Time Vs Temperature and Pressure

Sheet1

032880

10115880

2022010100

3031240400

4041150500

5051145450

5455545450

6055545450

7055540400

8555530300

10055525250

11555525250

13055520200

14555515150

16055515150

17555515150

18055515150

19053310100

20550810100

22049010100

235464880

265428770

280409660

295400550

310388440

325376440

355353440

3703453.535

3853383.535

4003282.525

4153172.525

4303112.525

4453032.525

4602962.525

4902822.525

5202702.525

5502592.525

5802492.525

6102392.525

7001952.525

7301862.525

7901752.525

8501642.525

9101552.525

970143220

1030135220

1090125220

1150119220

1210112220

1270106220

13301002.525

1390952.525

1450912.525

1510862.525

1570832.525

1630792.525

1690752.525

1750722.525

1810692.525

1890672.525

195065

201063

207061

213059

219057

225055

231053

237052

243051

249049

2555048

261047

273046

279045

285044

Sheet1

Temperature

Pressure

Time (minutes)

Temperature (OC) and Pressure (X10-7 torr)

Time Vs Temperature and Pressure

Sheet2

Sheet3

_1248212579.unknown

_1248209820.unknown

_1248211732.unknown

_1248205610.unknown

_1248131827.unknown

_1248132077.unknown

_1248131586.unknown

_1248126143.unknown

_1248131198.unknown

_1248126070.unknown