Predicting early-age stress evolution in restrained concrete by thermo-chemo-mechanical model and active ensemble learning

Predicting early-age stress evolution in restrained concrete by thermo-chemo-mechanical model and active ensemble learning

Liang, M. (TU Delft Materials and Environment)
Chang, Z. (TU Delft Materials and Environment)
He, S. (TU Delft Materials and Environment)
Chen, Y. (TU Delft Materials and Environment)
Gan, Y. (TU Delft Materials and Environment)
Schlangen, E. (TU Delft Materials and Environment)
Šavija, B. (TU Delft Materials and Environment)

2022

Early-age stress (EAS) is an important index for evaluating the early-age cracking risk of concrete. This paper encompasses a thermo-chemo-mechanical (TCM) model and active ensemble learning (AEL) for predicting the EAS evolution. The TCM model provides the data for the AEL model. First, based on Fourier's law, Arrhenius’ equation, and rate-type creep law, a TCM model is built to simulate the heat transfer, cement hydration, and viscoelasticity, which together determine the EAS evolution. Then, a material model composed of an eXtreme Gradient Boosting model and adjusted Model Code 2010 is built to allow for parametric study and database construction. Finally, an AEL framework is built, which incorporates principal component analysis (PCA), Gaussian process, and light gradient boosting machine (LGBM). This study resulted in the following findings: (1) The dimensionality of the 672-by-1 EAS vector can be effectively reduced by PCA, and the first principal component (PC) is a global index representing the magnitude of the EAS; (2) the mechanical field of the TCM model is validated by testing data. Correlation analysis on the first PC quantifies the influence of various input parameters of the TCM model, which is in accordance with common understandings of the EAS evolution process. (3) The AEL and one-shot ensemble learning (OSEL) both achieve high prediction performance in the testing set, whose R² reaches 0.961 and 0.948, respectively. Thanks to the uncertainty-based query procedure, comparing with OSEL, AEL shows advantages in prediction performance over the whole training history. (4) AEL can significantly reduce the number of samples required for training, which can be a major improvement in efficiency considering the computational cost of the TCM model.

http://resolver.tudelft.nl/uuid:402bd8b6-3b0d-4d05-b0a2-24abae01efa1

https://doi.org/10.1111/mice.12915

1093-9687

Computer-Aided Civil and Infrastructure Engineering, 37 (14), 1809-1833

Institutional Repository

journal article

© 2022 M. Liang, Z. Chang, S. He, Y. Chen, Y. Gan, E. Schlangen, B. Šavija