.png){kind=icon}
.png){kind=icon}
Return
Integration of automated experiment and domain adaptation for precise evaluation of cement dispersant performance
.png){kind=icon}
PRE
AI
I
J
DOI：10.1016/j.cemconres.2026.108342.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Evaluating the rheological performance of modern concrete requires large datasets due to the diversity of mixture compositions and the inherent variability of cementitious materials. Our previous study introduced automated experimentation for high-throughput rheological measurement of cement-based materials. In this study, the system is extended to enable reliable estimation of Bingham parameters by integrating domain adaptation. A total of 100 mortar samples with systematically varied mix proportions were tested using the proposed framework. The resulting large, high-quality dataset, combined with machine learning, enables accurate estimation of Bingham parameters that are consistent with the measured torque response. Shapley additive explanations further confirm the proposed approach provides physically reasonable Bingham parameter estimates. The proposed system demonstrates its applicability to PCE performance evaluation, detection of subtle variations in rheological behavior, and quantitative assessment toward rheological standardization. Through this work, an automated experimentation platform for rheological characterization of cement-based materials has been completed, establishing a methodological transition toward data-centric concrete rheology.
Journal
IF:
13.1
Papers:
6.9K
Citations:
7.5W
Organization
No organization information available
Cited Papers