Date of Award
Master of Science In Civil Engineering Degree
Civil And Environmental Engineering
Owusu-danquah, Josiah Sam
Dr. Jacqueline M Jenkins
Dr. Christopher Robert Huhnke
Old and demolished structures profusely exist in landfills because they are not being recycled frequently nor being employed correctly. This leads to an increase of construction and demolished wastes (C&D). These demolished structures and blocks can be broken down into smaller components to serve as aggregates (which are called recycled aggregates). Recycled aggregates are not being used regularly because they sometimes have detrimental influence on the compressive strength of concrete. Recycled concrete aggregate (RCA) reduces compressive strength of the concrete samples due to absorption issues related to the type, and age of the old concrete. Increase in water absorption levels leads to reduction in the compressive strength. If this issue is resolved, consumption of natural resources would decrease, and the use of recycled aggregate would increase which has beneficial reflection on the economy and the environment. The objective of this research was to develop a model to predict the compressive strength of concrete containing different percentages of RCA. This research studied the physical properties that reduce compressive strength, and even included the parameters that are aligned to the concrete mixture and treated them as input parameters in a prediction model. The model was created using artificial intelligence neural network. The built model included a specific prediction algorithm which was Bayesian Regularization Backpropagation which can deal with many types of data, even those of the random type. Although, the data was considered as non-linear, the Bayesian v probability algorithm was able to determine the pattern between the data and reduce the error by using the error function which was Mean Squared Error. The experimental data was collected from previous published research works in literature. The collection of the data and the evaluation of the model were both built upon specific criteria. The training results showed the success of the model. The model can be used as a tool by engineers to calculate compressive strength when recycled aggregates are added by entering the physical properties of the mixture. The work done here can be extended in the future to cover optimization of mechanical properties of concrete containing RCA.
Nijem, Abdelaziz, "Artificial Intelligence Neural Network: Compressive Strength Prediction of Recycled Aggregate Concrete Samples" (2020). ETD Archive. 1274.