Techno Press
You logged in as Techno Press

Computers and Concrete
  Volume 19, Number 3, March 2017 , pages 233-241

Compressive strength prediction of CFRP confined concrete using data mining techniques
Aires Camões and Francisco F. Martins

    During the last two decades, CFRP have been extensively used for repair and rehabilitation of existing structures as well as in new construction applications. For rehabilitation purposes CFRP are currently used to increase the load and the energy absorption capacities and also the shear strength of concrete columns. Thus, the effect of CFRP confinement on the strength and deformation capacity of concrete columns has been extensively studied. However, the majority of such studies consider empirical relationships based on correlation analysis due to the fact that until today there is no general law describing such a hugely complex phenomenon. Moreover, these studies have been focused on the performance of circular cross section columns and the data available for square or rectangular cross sections are still scarce. Therefore, the existing relationships may not be sufficiently accurate to provide satisfactory results. That is why intelligent models with the ability to learn from examples can and must be tested, trying to evaluate their accuracy for composite compressive strength prediction. In this study the forecasting of wrapped CFRP confined concrete strength was carried out using different Data Mining techniques to predict CFRP confined concrete compressive strength taking into account the specimens\' cross section: circular or rectangular. Based on the results obtained, CFRP confined concrete compressive strength can be accurately predicted for circular cross sections using SVM with five and six input parameters without spending too much time. The results for rectangular sections were not as good as those obtained for circular sections. It seems that the prediction can only be obtained with reasonable accuracy for certain values of the lateral confinement coefficient due to less efficiency of lateral confinement for rectangular cross sections.
Key Words
    CFRP confined concrete; data mining; artificial neural networks; support vector machines
Aires Camões: CTAC, Department of Civil Engineering, University of Minho, Guimarães, Portugal
Francisco F. Martins: ISISE, Department of Civil Engineering, University of Minho, Guimarães, Portugal

Techno-Press: Publishers of international journals and conference proceedings.       Copyright © 2024 Techno Press
P.O. Box 33, Yuseong, Daejeon 305-600 Korea, Tel: +82-42-828-7996, Fax : +82-42-828-7997, Email: