Concrete's tendency to "creep," or deform progressively, under mechanical stress is well known. However, the reasons behind this behavior have generally remained poorly understood and sometimes scientifically contested.

As a result, engineers estimate creep using empirical models, which often poorly predict creep behavior, says Gaurav Sant, associate professor at the University of California-Los Angeles Department of Civil and Environmental Engineering.

Empirical models often poorly predict concrete creep behavior. Image credit: Pixabay.Sant and colleagues say they believe they have determined the root cause of concrete deformation. “By careful unifications of experimental and computational data, we clarified that creep originates from a dissolution-precipitation process that acts at nanoscale contact regions of [calcium-silicate-hydrates] grains,” he says.

Sant and his colleagues describe their work in The Journal of Chemical Physics. They found that calcium-silicate-hydrates, the binding phase that holds cement paste together, tend to dissolve at high-stress regions and re-precipitate at low-stress regions. The authors say this corresponds with Le Chatelier’s principle, also known as “the equilibrium law.”

“As a result of such dissolution-precipitation behavior, a macroscopic, time-dependent ‘creep’ deformation manifests," says Sant. The idea of a dissolution-precipitation process is familiar to geologists—its effects can lead to deformation in the earth's crust. However, Sant says that this is one of the first times it has been shown to be relevant to concrete.

“Such behavior shows a dependence on the chemical composition of the calcium-silicate-hydrate, a result [that] permits identification of ‘isostatic’ calcium-silicate-hydrate compositions, which feature a minimum in creep and dissolution rates. This data reveals a previously unknown ‘compositional route’ to minimize creep of concrete,” the researchers say.

Future work will involve putting together a comprehensive description of concrete creep from the atomic to macroscopic scale. This will help them develop mechanistic models for predicting creep behavior and identifying cementation agents with reduced sensitivity to creep.