Joseph J. Biernacki

Tennessee Tech University

Computationally efficient algorithms for developing cement hydration microstructures

Dario Cruz

X-ray micro-tomography was used to visualize hydrating alite in real-time [1].Alite is the most abundant component in type I portland cement. Tomography of this sort produces large sets of data consisting of grey-scale images, or tomographs, of the subject material, alite in this case. Such data might provide a critical link for the development of highly efficient hydration models that include both solution phase chemistry [2] and microstructure generation. Current models for predicting microstructure development are slow and extremely computationally intensive [3].  And, while microstructures generated form such models have been shown to correlate well for some experimental properties [4], one questions if there could be a much more efficient way of generating realistic microstructures. My strategy is to utilize microstructural information from 3-D images in combination with machine learning (ML) technology to develop algorithms that might be conjoined with fast continuum codes to generate both hydration trajectories and microstructure using a highly efficient hybrid strategy.   Extracting useful information from such large datasets could provide a robust link that bridges continuum models with microstructure, however, the exact methodology is yet to be identified and is the subject of this work. Strategies for mining information form the large 3-D tomography datasets include image processing and ML toolboxes.

[1] Y. J He, “Characterization of microstructure evolution of cement paste by micro computed tomography”, Journal of Central South University, vol. 20(4),  pp 1115-1121, 2013.

[2] M. Gottapu, “Investigation of recent  c3s  hydration inferences using  a multi-constrained multi-ionic single particle modeling strategy” Ph.D. dissertation, Dept Chem. Eng, Tennessee Technological University., Cookeville, TN, 2013.

[3] J. W. Bullard, “A determination of hydration mechanisms for tricalcium silicate using a kinetic cellular automaton model,” Journal of American Ceramic Society, vol. 97(7), pp.2088-2097, 2008.

[4] D. P. Bentz, “Three dimensional computer simulation of cement hydration and microstructural development,” Journal of American Ceramic Society, vol. 80(1), pp. 3-21, 1997.

Leave a Reply



XHTML: You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>