Proceedings of the 8th International conference on Research in Engineering, Science and Technology
Developing New Software for Determination of Concrete Compressive Strength of the Existing Reinforced Concrete Buildings
Gamze Dogan, Musa Hakan Arslan, Ahmet Ozkis And Murat Ceylan
Ability to determine the compressive strength of the concrete used at existing reinforced concrete buildings according to the traditional destructive methods is a quite time– consuming, costly and difficult process. As for the non–destructive methods, they are in general not preferred due to their high error shares. Therefore, it is extremely important to develop a practical method and the software related hereto. It is aimed that compressive strength determination–oriented software study to be developed to be the first in the literature, reinforced concrete used in existing buildings using Image Processing Technical (IPT) and Artificial Neural Networks (ANN) together. For experimental study, 192 concrete core samples were produced from 32 different sets of concrete block. Concrete surficial images and compressive strength of the core samples are matching. The 192 concrete core samples images were digitized with IPT and were used ANN for training and test set. 27 images of concrete samples for validation process (validation set) selected randomly from reinforced concrete building. Even though results from samples produced in laboratory is quite satisfactory, results from the case samples obtained from real existing RC buildings is not reached the desired level.
Keywords: Reinforced Concrete Building, Concrete, Core, Non-destructive Methods,Image Processing, Artificial Neural Network.