The dependence of the steady-state creep rate, ε_s, on stress, σ, and temperature, T, for pure copper and a two-phase copper–4.04 wt.-% cobalt alloy at 712 K can be described as

For copper, the stress exponent, n, is ∼4.8 and the activation energy for creep, Q _c, is ∼ 115 kJ/mol. For the copper–cobalt alloy, n≃5 and Qc≃140 kJ/mol at low stresses, whereas n≃12 and Qc≃210 kJ/mol at high stresses. This variation in n and Q _c can be rationalized by measuring the ‘friction stress’, σ_o, using a technique involving consecutive small stress reductions during creep. Then, for both materials

The activation energy, Q _c *, is derived from the temperature-dependence of ε_s at the same value of (σ – σ_o) rather than at constant σ as in the determination of Q _c. For both materials, Q _c*≃110 kJ/mol, indicating that processes occurring in the matrix are rate-controlling during creep of the two-phase alloy. The greater creep-resistance of the copper-cobalt alloy is attributable to the particle dispersion decreasing A* and increasing σ_o. The results are explained in terms of a model for creep in which the rate-controlling process is the growth of the 3-dimensional dislocation network developed during creep, to produce dislocation sources allowing slip to occur.