Prediction of Stress Increase at Ultimate in Unbonded Tendons Using Sparse Principal Component Analysis
| Autor: | Minwoo Chang, Eric McKinney, Marc Maguire, Yan Sun |
|---|---|
| Přispěvatelé: | Springer Singapore |
| Rok vydání: | 2019 |
| Předmět: |
Principal Component Analysis
Structural material Mathematical analysis 0211 other engineering and technologies Linear model 020101 civil engineering Ocean Engineering 02 engineering and technology LASSO strand stress increase 0201 civil engineering Stress (mechanics) Sparse Principal Component Analysis Lasso (statistics) lcsh:Systems of building construction. Including fireproof construction concrete construction 021105 building & construction Principal component analysis Principal component regression lcsh:TH1000-1725 Linear equation Mathematics Civil and Structural Engineering Test data unbonded tendons |
| Zdroj: | Mathematics and Statistics Faculty Publications International Journal of Concrete Structures and Materials, Vol 13, Iss 1, Pp 1-18 (2019) |
| Popis: | While internal and external unbonded tendons are widely utilized in concrete structures, an analytical solution for the increase in unbonded tendon stress at ultimate strength, $$\Delta f_{ps}$$ , is challenging due to the lack of bond between strand and concrete. Moreover, most analysis methods do not provide high correlation due to the limited available test data. The aim of this paper is to use advanced statistical techniques to develop a solution to the unbonded strand stress increase problem, which phenomenological models by themselves have done poorly. In this paper, Principal Component Analysis (PCA), and Sparse Principal Component Analysis (SPCA) are employed on different sets of candidate variables, amongst the material and sectional properties from a database of Continuous unbonded tendon reinforced members in the literature. Predictions of $$\Delta f_{ps}$$ are made via Principal Component Regression models, and the method proposed, linear models using SPCA, are shown to improve over current models (best case $$R^{2}$$ of 0.27, measured-to-predicted ratio [λ] of 1.34) with linear equations. These models produced an $$R^{2}$$ of 0.54, 0.70 and λ of 1.03, and 0.99 for the internal and external datasets respectively. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|