Abstract

The power-law creep equation, ε̇ ∞ σⁿ exp(−Q/RT), is commonly used to relate strain rate, ε̇, stress, σ, and temperature, T, for thermally activated dislocation creep in rocks. When triaxial deformation experiments on marble and limestone samples are performed at temperatures of 400–1050°C, to strains <0.2, and with strain rates between 10⁻³ and 10⁻⁷s⁻¹, the variations in strength among different rocks at nominally identical conditions are much larger than the experimental uncertainty. During dislocation creep, the strengths of various limestones and marbles decrease with increasing grain size, similar to the Hall-Petch effect in metals. The stress sensitivity of strain rate, n′ = ∂ln ε̇/∂lnσ, and the temperature sensitivity of strain rate, Q′ = −R∂lnε̇/∂(1/T), differ greatly for the various calcite aggregates. There is a systematic dependence of n′ and Q′ on stress, grain size, and perhaps, temperature, and there is no interval in stress where n′ is constant. Thus, the steady-state power-law equation is an inadequate description of dislocation creep in calcite rocks. To improve the constitutive law, it may be necessary to include at least one additional state variable that scales with grain size.