There is a recurring problem in the field of concrete petrography: What caused the cracks in a piece of deteriorated concrete? This is a difficult question to answer from just a field evaluation and often is a difficult task in petrographic investigations. A typical example is cracked concrete with crystals growing in the cracks. Did the crystal growth cause the cracks, or did the crystals grow in the open space provided by the cracks? Knowledge of the materials and environmental history can help guide the petrographer's approach to the problem. For example, if the temperature of the concrete was always above freezing, then frost attack can be excluded as a possible mechanism, or if there was no source of external sulfate ions anywhere near the concrete during its lifetime, then external sulfate attack can be ruled out. Even with environmental information available, however, the crack-causing mechanism(s) usually cannot be uniquely determined.
Most causes of deterioration involve mechanics, in that some phase grows/shrinks because of chemical or physical attack and induces stresses that produce cracking of various patterns. Examples include frost attack, sulfate attack, and alkali-silica reaction. Using video imaging, stereo microscopy, and back-scattered scanning electron microscopy, we can take images of damaged areas, and identify the different phases, including the cracks. For the SEM, if gray scale alone is not enough to distinguish among phases, then x-ray bitmaps can be produced that give chemical information that can further distinguish phases [1, 2]. This article describes a hybrid imaging-finite element modelling technique that can be used with these kinds of images to help determine the mechanism of damage. This is accomplished by qualitatively linking mechanics to the microstructural image via simulations of selected deterioration mechanisms. The details of the technique, as well as its limitations and research needs, will be described in this paper.