Prediction of Fresh and Hardened State Properties of UHPC: Comparative Study of Statistical Mixture Design and an Artificial Neural Network Model

AbstractThe main objective of the research study described herein is to build two analytical models based on artificial neural networks (ANNs) and the statistical mixture design (SMD) method to predict the required performance of ultra-high-performance concrete (UHPC). Two different curing conditions—heat treatment and water storage—were applied to the specimens. To train the neural network, a total set of 53 different mixtures was designed based on the design matrix of SMD. The statistical analysis results showed the adequacy of both models to predict the required performance of UHPC; however, the ANN model could predict the compressive strength (water storage) and slump flow with higher accuracy than the SMD. The optimum combination of the cement, silica fume, and quartz flour was determined to be 24, 9, and 5% by total volume to achieve a flowable mixture with the highest compressive strength. The accuracy of the model was verified with additional experimental tests.

[1]  I-Chang Jou,et al.  Analysis of hidden nodes for multi-layer perceptron neural networks , 1994, Pattern Recognit..

[2]  James A. Anderson,et al.  Cognitive and psychological computation with neural models , 1983, IEEE Transactions on Systems, Man, and Cybernetics.

[3]  Benjamin A. Graybeal,et al.  Material Property Characterization of Ultra-High Performance Concrete , 2006 .

[4]  F ROSENBLATT,et al.  The perceptron: a probabilistic model for information storage and organization in the brain. , 1958, Psychological review.

[5]  Roger Phan-Tan-Luu,et al.  Mixture design applied to the formulation of hydrotropes for liquid detergents , 2002 .

[6]  Andreas Fischer,et al.  A Levenberg-Marquardt algorithm for unconstrained multicriteria optimization , 2008, Oper. Res. Lett..

[7]  Antoine E. Naaman,et al.  Ultra-High Performance Concrete with Compressive Strength Exceeding 150 MPa (22 ksi): A Simpler Way , 2011 .

[8]  Mark Beale,et al.  Neural Network Toolbox™ User's Guide , 2015 .

[9]  Nikola K. Kasabov,et al.  Short-term ANN load forecasting from limited data using generalization learning strategies , 2006, Neurocomputing.

[10]  Saeed Bastani,et al.  Optimization of the water-based road-marking paint by experimental design, mixture method , 2006 .

[11]  Geoffrey E. Hinton,et al.  Learning internal representations by error propagation , 1986 .

[12]  Martin T. Hagan,et al.  Neural network design , 1995 .

[13]  Jin H. Huang,et al.  Detection of cracks using neural networks and computational mechanics , 2002 .

[14]  Bahar Sennaroglu,et al.  Self-Consolidating High-Strength Concrete Optimization by Mixture Design Method , 2010 .

[15]  Erik Johansson,et al.  Mixture design—design generation, PLS analysis, and model usage , 1998 .

[16]  Kamal H. Khayat,et al.  Utility of statistical models in proportioning self-consolidating concrete , 2000 .

[17]  Eduardo Júlio,et al.  RSM-based model to predict the performance of self-compacting UHPC reinforced with hybrid steel micro-fibers , 2014 .

[18]  Eduardo Júlio,et al.  The effect of nanosilica addition on flowability, strength and transport properties of ultra high performance concrete , 2014 .

[19]  I-Cheng Yeh,et al.  Optimization of concrete mix proportioning using a flattened simplex–centroid mixture design and neural networks , 2009, Engineering with Computers.

[20]  Eduardo Júlio,et al.  Statistical mixture design approach for eco-efficient UHPC , 2015 .

[21]  P. J. Whitcomb,et al.  Find the most favorable formulations , 1998 .

[22]  Hajime Okamura,et al.  Development of self-compacting high performance concrete. , 1995 .