A Multivariate Computational Framework to Characterize and Rate Virtual Portland Cements. (26th September 2018)

• Record Type: Journal Article
• Authors: Tao, Chengcheng; Watts, Benjamin; Ferraro, Christopher C.; Masters, Forrest J.
• Is Part Of: Computer-aided civil and infrastructure engineering. Volume 34:Number 3(2019:Mar.)
• Page Start: 266
• Page End: 278
• Publication Date: 2018-09-26
• Subjects: Civil engineering -- Data processing -- Periodicals; Computer-aided engineering -- Periodicals; 624.0285
• DOI: 10.1111/mice.12413
• Languages: English
• ISSNs: 1093-9687

Abstract

Abstract: This article presents a novel application of metaheuristic optimization and rating techniques to virtual test results for cement and mortar, and presents objective computational methods for the evaluation and selection of cementitious materials based on simulated material testing. A scalable approach based on particle swarm optimization of the National Institute of Standards and Technology Virtual Cement and Concrete Testing Laboratory (VCCTL) is successfully demonstrated using ∼150, 000 combinations of cement phase distributions and water‐to‐cement ratios, with as few as 10% of the VCCTL runs required to obtain the optimal solutions. The application of Pareto front analysis reveals an inherent trade‐off between the modulus of elasticity, time of set, and kiln temperature (using alite:belite as a proxy) at the performance limits. This study also provides a means to objectively characterize cements in these contexts using a maximum likelihood estimator to calculate a score from the cumulative distribution function, using a random subset of the 150, 000 simulated cement pastes. The proposed approach provides a new pathway to optimally proportion raw materials (and eventually waste by‐products) to reduce production costs, extend the life of a quarry, or reduce the carbon footprint of cement plants.