Relationships between agents for modelling the intermediate stages of cement hydration

Abstract

Agent-Based modelling is, in many cases, most natural for describing and simulating a system composed of “behavioral” entities. Whether one is attempting to describe the complex process of cement hydration, an agent-based approach makes the model seem closer to reality. Therefore, it is more natural to describe how the calcium silicate phases individually behave during the hydration Chemistry than to come up with the equations that govern the dynamics of the anhydrous constituents. Because the dynamic equations result from the behavior of the calcium silicate phases, the agent-based model will also enable to study aggregate properties. A model based on agents for the cement hydration, also makes it possible to analyse the formation and growing of the C-S-H gel: the main binding agent in hardened cement. Knowing the initial concentration of the anhydrous constituents makes it possible to create virtual agents which represent and behave according with their reactivity. This work describes an algorithm which cores an agentbased approach for modelling the intermediate stages of cement hydration, focusing on the microstructure development and C-S-H gel formation. Theories relating to the mechanism of cement hydration are examined and these are discussed in terms of the agents behavioral. Monte Carlo method is applied to obtain and analyse the algorithm results