Smooth support vector regression (SSVR) modelling of self-compacting concrete properties / Yoyok Setyo Hadiwidodo
Material type:
TextPublication details: Kuantan, Pahang : UMP, 2013Description: xviii, 155 p. : ill. ; 30 cm. + 1 CD-ROMISBN: - THE0002501(Local)
| Item type | Current library | Call number | Copy number | Status | Date due | Barcode | |
|---|---|---|---|---|---|---|---|
Thesis
|
UMPLIB GAMBANG | TA439 .H33 2013 rs Thesis (Browse shelf(Opens below)) | 1 | Not for loan | 0000075946 | ||
|
Thesis
|
UMPLIB GAMBANG | CD 7243 | TA439 .H33 2013 rs Thesis (Browse shelf(Opens below)) | 1 | Not for loan | 0000075947 |
Thesis (PhD in Civil Engineering) -- Universiti Malaysia Pahang – 2013
Bibliography : p. 133-142
Self-compacting concrete (SCC) is a type of concrete that can flow under its own weight without vibration, filling small interstices of formwork, passing through complicated geometrical configurations, be pumped through long distances and resist segregation. SCC is a complex material, which makes modelling its behaviour a very difficult task. SCC constituent materials and mix proportions which must be properly selected to achieve these flow properties required. The effects of any changes in materials or mix proportions on fresh and hardened concrete performance must be considered in evaluating SCC. It is crucial to use a systematic approach for identifying optimal mixes and investigates the most effective factors on SCC properties under a set of constraints. Due to this reason Taguchi method with the L18 (36) orthogonal array is used in this study to investigate the properties of SCC. Taguchi method is a promising approach for optimizing mix proportions of SCC to meet several fresh concrete properties. Taguchi method can simplify the test procedure required to optimize mix proportion of SCC by reducing the number of trial mixes. This study has shown that it is possible to model SCC which fulfilling its criteria. The application of the Taguchi method gave the optimal mix design proportions for fresh properties and hardened properties as well. This study has also demonstrated the capability of regression analysis and Smooth Support Vector Regression (SSVR) modelling to predict the properties of SCC. The performance of the proposed method is evaluated using a coefficient of determination (R2) and mean square error (MSE). Results have shown this model is accurate in prediction of the properties of SCC because it has maximum R2 and minimum MSE. The performance of the proposed method is also verified by comparing the predicted levels with actual values. It can be concluded that SSVR method can predict properties of self-compacting concrete with higher estimation accuracy.