A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Statistical Comparisons of Creep and Shrinkage Prediction Models Using RILEM and NU-ITI Databases. Paper by Akthem Al-Manaseer and Armando Prado/AUTHORS' CLOSURE
In complex problems with many random variables of widely different scatter magnitudes, diverse overlapping trends, and an imperfect database that had to be drawn from separate studies conducted without a coordinated scientific sampling scheme, global statistical inferences may be misleading or even support erroneous conclusions. This is what plagues the authors' paper, and also previous similar statistical studies. Furthermore, the authors use an obsolete database missing data on modern concretes with autogenous shrinkage and data on admixture and aggregate type effects. These are all included in the new NU database, which contains approximately 4,000 test curves-more than twice as many as the RILEM and NU-ITI databases. The purpose of the paper being discussed is to provide designers and code committees with insight into models' prediction accuracy. Overall, the ranking of the models can change according to the criteria described in this discussion. Therefore, ranking of the models is a progressing topic of research influenced by factors such as the selected statistical methods, database, and data elimination criteria.
Statistical Comparisons of Creep and Shrinkage Prediction Models Using RILEM and NU-ITI Databases. Paper by Akthem Al-Manaseer and Armando Prado/AUTHORS' CLOSURE
In complex problems with many random variables of widely different scatter magnitudes, diverse overlapping trends, and an imperfect database that had to be drawn from separate studies conducted without a coordinated scientific sampling scheme, global statistical inferences may be misleading or even support erroneous conclusions. This is what plagues the authors' paper, and also previous similar statistical studies. Furthermore, the authors use an obsolete database missing data on modern concretes with autogenous shrinkage and data on admixture and aggregate type effects. These are all included in the new NU database, which contains approximately 4,000 test curves-more than twice as many as the RILEM and NU-ITI databases. The purpose of the paper being discussed is to provide designers and code committees with insight into models' prediction accuracy. Overall, the ranking of the models can change according to the criteria described in this discussion. Therefore, ranking of the models is a progressing topic of research influenced by factors such as the selected statistical methods, database, and data elimination criteria.
Statistical Comparisons of Creep and Shrinkage Prediction Models Using RILEM and NU-ITI Databases. Paper by Akthem Al-Manaseer and Armando Prado/AUTHORS' CLOSURE
Zdenek P Bazant (author) / Roman Wendner / Giannis Boumakis / Mija Hubler
ACI materials journal ; 112
2015
Article (Journal)
English
Statistical Comparisons of Creep and Shrinkage Prediction Models Using RILEM and NU-ITI Databases
| Online Contents | 2015
Features - ACI-RILEM WORKSHOP ON CREEP AND SHRINKAGE
Online Contents | 1999
Unbiased Statistical Comparison of Creep and Shrinkage Prediction Models
| Online Contents | 2008
Unbiased Statistical Comparison of Creep and Shrinkage Prediction Models
| British Library Online Contents | 2008
4th RILEM International Symposium on Creep and Shrinkage of Concrete Mathematical Modeling
| Online Contents | 1986