High Strength Copper-Titanium Alloys: Part One

The age hardening of copper–titanium (Cu–Ti) alloys containing approximately 1–5 wt.% Ti (1–6 at.% Ti) has been known since the 1930s. The mechanical and physical properties were found to be comparable to the widely used copper–beryllium (Cu–Be) alloys with better high temperature strength and superior stress relaxation behavior. The electrical conductivity of the aged alloys, however, falls somewhat below that of the Cu–Be alloys.

Cu–Ti alloys are now receiving a great deal of attention as ultra-high strength conductive materials for applications such as conductive springs, interconnections etc. essentially displacing the conventional Cu–Be alloy series. This move away from Cu–Be has been catalyzed to a large extent by the full recognition of serious health hazards associated with the Be-based metallurgy in production. It is likely that the precipitation hardened Cu–Ti alloys will rise to prominence as a commercial alternative in a myriad of applications over the next decade.

A lot of electrically conductive springs of thin plate type are made of an age-hardened copper titanium alloy because of its excellent mechanical strength and electrical conductivity. Such spring sheets are usually produced by a process which includes preparing a copper titanium melt, casting it, hot working the cast copper titanium, subjecting the hot-worked copper titanium to alternate annealing and cold working to final shape, subjecting the copper titanium to solution heat treatment, and age-hardening it after cold working, if required. Its solution heat-treated structure has, however, an average crystal grain size of at least 40 microns and even up to 100 microns.

The age-hardened copper titanium alloy has been developed as an inexpensive substitute for a well-known age-hardened copper beryllium alloy which is disclosed in, for example, U.S. Pat. No. 4,425,168 issued to Goldstein et al. on January 10 1984. The conventional age-hardened copper titanium alloy is, however, not satisfactory, and required to be improved, in view of mechanical properties such as formability, fatigue strength, elongation and yield strength. Moreover, it has the drawback of having different properties in the rolling direction and in a direction perpendicular thereto.

The early phase diagrams of the Cu–Ti system showed the terminal FCC Cu-rich solid solution (α) to be in equilibrium with a Cu₃Ti phase characterized by a space group Pnmm. However, a Au4Zr-type structure was reported and this orthorhombic phase Cu₄Ti exhibits the space group Pnma.

Work in the early 1970s by Hakkarainen and Laughlin and Cahn reported that a metastable coherent precipitate phase with the tetragonal D1_a structure (Ni₄Mo-type; I 4/m) having the composition Cu₄Ti forms during aging below approximately 600–700°C. This important finding was further verified in a comprehensive electron microscopy study of the microstructural behavior in a series of alloys containing 1–4 wt.%Ti by Datta and Soffa. The resultant metastable, finely dispersed two-phase mixtures of α'+D1_a which arise during aging produce highly aligned and quasi-periodic microstructures often called modulated structures.

During prolonged aging at low or moderate aging temperatures a coarse lamellar microconstituent composed of the equilibrium phase and terminal solid solution (α) nucleates and grows consuming the fine-scale dispersion of coherent/semicoherent D1_a particles. Fig. 1a–b shows recent versions of the equilibrium diagram as well as the measured metastable solvus of the D1_a precipitate phase. The portion of the Cu–Ti equilibrium diagram in Fig. 1b indicates a polymorphic transformation of the Cu₄Ti phase; it suggests that the tetragonal D1_a phase is the stable phase below ≈ 500°C and that the orthorhombic Au₄Zr-type structure is the equilibrium high temperature phase.

The existence of the D1_a structure as the stable low temperature phase is not surprising since the D1a structure is a ground state for FCC-based alloys. The orthorhombic Pnma structure (Au₄Zr-type) is presumably stabilized at higher temperatures by entropic effects.

A fundamental interest was directed at the Cu–Ti age hardening alloys in the 1960s and 1970s when these alloys were recognized to be prototypical binary ‘‘sideband alloys’’ perhaps exhibiting spinodal decomposition. The strong age hardening response also received new attention since these precipitation hardening alloys were found to show extraordinary plastic flow behavior including profuse deformation twinning.

The occurrence of spinodal decomposition during the early stages of precipitation resulting in the formation of an ordered phase provoked new thinking regarding the interplay of clustering and ordering tendencies in solid solutions similar to those articulated by Allen and Cahn and Ino. During the anticipated expanded commercialization of age hardened Cu–Ti alloys the nature of the cellular reaction or discontinuous precipitation mode will undoubtedly be a major focal point since the appearance of the cellular microconstituent leads to deleterious effects on mechanical properties. Also, the influence of plastic deformation on the aging response will be of paramount importance in the thermomechanical processing of these materials in production.

Numerous experimental methods have been applied to elucidate the basic mechanisms controlling the formation of the characteristic periodic and aligned two-phase microstructures which arise during aging, including X-ray and neutron diffraction, electron microscopy and diffraction and atom probe field-ion microscopy (APFIM). The early stages of decomposition and the origin of the ‘‘sideband state’’ will be discussed along with the evolution of the microstructures during prolonged aging.

The decomposition of supersaturated Cu–Ti alloys containing 1–6 atomic % titanium (Ti) embodies a complex interplay of clustering and ordering effects in solid solutions as well as a subtle synergy between short range order (SRO) and long range order (LRO). In addition, the supersaturated Cu–Ti alloys exhibit well-known ‘‘sideband’’ phenomena in the diffraction patterns during the early stages of decomposition.

As pointed out above, the precipitate phase of particular interest in the age hardening response of these high-strength alloys exhibits the D1_a superstructure (Ni₄Mo-type; I 4/m) and therefore is the focus of this discussion. A generalized Bragg–Williams description of the thermodynamics of the LRO Cu₄Ti/ D1_a structure requires fourth-nearest neighbors to stabilize the structure as a non-degenerated ground state for FCC-based superstructures but third nearest-neighbor interactions will lead to the emergence of SRO and LRO structures from a disordered solid solution at low temperatures.

The A₄B superstructure is favored in systems where the first nearest-neighbor interchange potential V⁽¹⁾ and second nearest-neighbor interchange potential V⁽²⁾ are of the same sign when considering just the first and second nearest-neighbors in the structural energies of the various FCC-based superstructures.

Figure 1: Recent versions of the Cu–Ti phase diagram: (a) Cu₄Ti (orthorhombic; Pnma) shown as the equilibrium phase above 500°C and (b) detailed portion of diagram showing polymorphic transformation in Cu₄Ti phase. X axis denotes atomic percentage of Cu.