Optical method for measuring slow crack growth in cementitious materials

Abstract

The double-torsion (DT) test is commonly used to calculate slow or subcritical crack velocities in (quasi-)brittle engineering materials directly from the measured load relaxation of notched DT-specimens. In cementitious materials a significant part of the recorded load relaxation in the DT-test may be due to specimen creep deformation, and this would then lead to overestimated crack velocities. In this paper we describe a method to optically measure slow crack growth in cementitious materials by carrying out DT-tests under the optical microscope or inside the environmental SEM. Crack tip detection is facilitated by digital image correlation of the time-lapse microscope recordings. DT-tests at 10 % relative humidity in hardened cement paste (with w/c-ratio of 0.4, 0.5 and 0.6) showed that optically measured crack velocities were significantly lower than those calculated from the DT-specimen relaxation. In many experiments the subcritical crack growth rapidly stopped, while an ongoing specimen load relaxation was recorded. At 90 % relative humidity, load-relaxation in the DT-test was much stronger than at 10 % relative humidity, because subcritical cracking and creep-induced relaxation both increase with moisture content.