In the present study, copper was used as an intermediate material in diffusion bonding between titanium and stainless steel. The process was carried out in the temperature range of 850 [degrees]C to 1000 [degrees]C for 60 minutes and at 900 [degrees]C for 30 to 150 minutes under the compressive stress of 3 MPa in a vacuum. The effects of temperature and time on the microstructure of Ti | Cu | stainless steel diffusion-bonded joints were studied. The interface microstructures of the bonded assemblies were observed in optical and scanning electron microscopes and an electron probe microanalyzer. Formation of various types of reaction products near the interface was detected using the X-ray diffraction technique. The maximum tensile and shear strength of ~101 pct of that of Ti and of ~86 pct of that of Ti, respectively, along with ~8.5 pct elongation, were obtained at 900 [degrees]C for 60 minutes. Observation of fracture surfaces demonstrates that failure takes place through the SS-Cu interface. DOI: 10.1007/s11661-008-9562-x

(c) The Minerals, Metals & Materials Society and ASM International 2008

I. INTRODUCTION

IN recent years, nuclear and chemical processing industries strongly demanded transition joints between titanium/titanium alloy and stainless steel for application in service.[1,2] The diffusion bonding technique provides a near-net-shape forming process for similar and dissimilar materials without gross macroscopic distortion and with minimum dimensional tolerances.[3] These two materials, when joined by conventional fusion welding, result in segregation of chemical species, stress concentration sites, and formation of intermetallic phases near the interface of the joint due to the very low solubility of Ti and Fe at room temperature.[4]

The literature reports that joints produced by direct bonding between Ti and stainless steel showed the formation of Cr, Fe, and Ti base reaction products.[5-8] Aleman et al.[6] reports that the TEM studies on titanium-306 stainless steel diffusion-bonded joints indicated the formation of s, Fe^sub 2^Ti, FeTi, ?, Fe^sub 2^Ti^sub 4^O, and TiC in the diffusion interface. Ghosh et al.[7] report that solid-state diffusion bonding was carried out between titanium and stainless steel and achieved a maximum tensile strength of 222.1 +- 9 MPa when processed at 850 [degrees]C for 90 minutes and also report that bond strength drops with a rise in bonding temperature due to the increase in the width of intermetallic phases at the diffusion zone. Copper can also be considered as a useful intermediate material to reduce the bonding parameters (i.e., temperature and time) due to its low melting point with respect to Fe, Cr, Ni, and Ti. An increase in the transportation of atoms of these elements at higher temperature (>0.5 T^sub m^, with T^sub m^, melting point in K) will encourage a good contact between the mating surfaces. Moreover, copper does not form any intermetallic phases with iron. The Cu-Ti binary phase diagram shows the occurrence of Cu^sub 2^Ti, Cu3Ti2, CuTi, and CuTi^sub 2^ with increasing Cu content.[8] Eroglu et al. also report that the Cu-Ti base intermetallic phases have higher plasticity than the Fe-Ti base intermetallic phase.[9] In the previous attempt, nickel was used as an intermediate material for the same diffusion couple and the bond tensile and shear strengths of 311 and 236 MPa, respectively, were obtained due to the absence of Fe-Ti and Fe-Cr-Ti base intermetallic compounds.[10]

The present investigation reports diffusion bonding of commercially pure titanium and 304 stainless steel using copper as an interlayer at various temperatures and times in vacuum. The effect of bonding temperature and time on microstructure and mechanical properties has been investigated.

II. EXPERIMENTAL PROCEDURE

The chemical compositions and room-temperature mechanical properties of commercially pure titanium and 304 stainless steel are given in Tables I and II, respectively. Both of them were received in the form of rod having 20-mm diameter. From the base materials, cylinders of 15-mm diameter x 30-mm length were machined.

The mating surfaces of the cylinders were prepared by conventional grinding and polishing techniques by final polishing with l-[mu]m diamond paste. The copper foil of 300-[mu]m thickness and 99.95 pct purity was used as an intermediate material, and both the surfaces of the interlayer were polished in the same fashion. The mating surfaces were cleaned in acetone and dried in air. The Ti | Cu | SS assembly was kept in contact in a fixture and was inserted in a vacuum chamber. The diffusion bonding was carried out in the temperature range of 850 [degrees]C to 1000 [degrees]C in steps of 50 [degrees]C for 60 minutes and 900 [degrees]C for 30 to 150 minutes in steps of 30 minutes in (6 to 8) x 10^sup -4^ Pa vacuum. The compressive stress of 3 MPa was applied along the longitudinal direction of the sample, and the stress was measured at room temperature. Heating was done at the rate of 0.24 [degrees]C s^sup - 1^ at the time of processing, and after the operation, the samples were allowed to cool in vacuum at a cooling rate of 0.1 [degrees]C s^sup -1^ up to 300 [degrees]C.

The diffusion bonded joints thus formed were cut longitudinally and prepared by conventional techniques for metallographic observations. The titanium side was etched in an aqueous solution of 88 mL H2O, 4 mL HF, and 8 mL HNO^sub 3^. The stainless steel side was etched by a mixture of 10mL HNO^sub 3^, 40mL HCl, and 50mL glycerol. A solution containing 5 g FeCl^sub 3^, 2 mL HCl, and 96 mL ethanol was used for etching pure copper. The structural change owing to diffusion was observed in a light microscope (Correct SDME TR5, Seiwa Optical Co. Ltd, Makaro-Ku, Tokyo, Japan). Polished samples were also examined in a scanning electron microscope (JEOL* JXA 840A) using backscattered mode (SEM-BSE) to reveal the reaction layers near the diffusion-bonded interface. The compositions of the reaction layers were determined in atomic percent using an electron probe microanalyzer (Cameca Sx 100). The presence of intermetallic phases in the reaction zone was confirmed by X-ray diffraction study (PHILIPS** PW 1840) on the fracture surfaces of the couples using a copper target. The scanning span of 20 to 80 deg with a step size of 0.02 deg ( = 20) was used during this investigation. Tensile properties of the transition joints were evaluated in a tensile testing machine (Instron 4204) at a crosshead speed of 8.33 ? 10-4 mm s at room temperature. Cylindrical tensile specimens were machined per ASTM specification (vol. 03.01 E8M-96) with a gage diameter and length of 4 and 20 mm, respectively. The interlayer was at the center of the gage length. The shear strength of the bonded joints was evaluated at room temperature using a screw tensile testing machine set at a crosshead speed of 8.3 ? IO-3 mm s~ '.'91 The shear test specimens were machined to a diameter of 10 mm. Four samples were tested at each processing parameter. The microhardness along the cross section of the diffusion-bonded joints was determined by a microhardness tester. A dwelling time of 20 seconds and a 10-g load were used for the measurement. Fracture surfaces of the samples were observed in secondary electron mode in SEM (Leica S440, Cambridge, UK) using energy-dispersive spectroscopy (Oxford 5431, High Wycombe, UK) to reveal the nature and location of failure under loading.

* JEOL is a trademark of Japan Electron Optics Ltd., Tokyo.