Prediction of compressive strength of concrete containing fly ash using data mi...

• Author(s): Martins, Francisco F. ; Camões, Aires
• Date: 2013
• Persistent ID: http://hdl.handle.net/1822/23707
• Origin: RepositóriUM - Universidade do Minho
• Subject(s): Concrete strength; Fly ash; Data mining; Artificial neural networks; Support vector machines
• Document Type: Article
• Language: English

Description

The concrete compressive strength is the most used mechanical property in the design of concrete structures. Therefore, the use of rational models to its prediction, to simulate the effects of its different constituents and its properties can play an important role in the achievement of the safety-economy required. Models to forecast the concrete compressive strength have already been presented before by some researchers. However, the comparison of different rational models and the application of models to predict the importance of the different constituents in the concrete behaviour have not yet been approached. Therefore, developing these models will be necessary namely to take into account the quality, i.e. the activity, of the most used mineral addition in concrete: fly ash. This study compared different Data Mining techniques to predict the compressive strength of fly ash concrete along time. The presented models are able to learn the complex relationships between several variables like the uniaxial compressive strength, the different concrete compounds and its mix design, the different properties of the fly ash used and the relative influence of its.