The present study aims to develop a mathematical model based on previous research work for the prediction of the compressive strength of plastic pavement blocks. The present study consists of the development of a mathematical model using Machine Learning (ML) and experimental validation. The background of the present study provides a brief overview of the existing research, along with the data collected and utilized for developing the ML model. The literature review indicated a lack of prior research on the use of machine learning models to predict the physical and mechanical properties of paving blocks incorporating LDPE and HDPE. The novelty of this study lies in utilizing machine learning to predict the properties of paving blocks made with recycled HDPE pellets obtained from Plasbin, a recycling facility based in Oman. According to the developed model, the mix proportions of plastic and fine aggregate is 1:3 corresponding to the required compressive strength of 20 MPa since C20 grade concrete is commonly used for the pavement construction. For the experimental study, plastic paving blocks are prepared using melted high/ low-density polyethylene and fine aggregate. The experimental study involved measuring the internal temperature, thermal conductivity, and compressive strength of LDPE and HDPE paving blocks. The results show that both materials have similar internal temperature characteristics. A comparison of the thermal conductivity of LDPE and HDPE paving blocks reveals that LDPE has 29.41% higher thermal conductivity than HDPE. However, the comparison of average compressive strengths between HDPE and LDPE pavement blocks indicates that the compressive strength of the HDPE block is 33.05% greater than that of the LDPE block. The Gaussian Process Regression (GPR) model with a Matern 5/2 kernel and constant basis function achieved the best performance, with the lowest RMSE, MSE, and MAE, though 1393 s for training. The percentage difference between the expected compressive strength predicted by the ML model and the experimental results ranges from 13 to 16%, indicating that the developed ML model performs effectively. One limitation of the present study is the release of harmful or toxic gases during the plastic melting process. For future research, the use of acetone is recommended as a potential solvent for melting plastics, as it may help reduce the emission of these gases.
Every day a huge quantity of glass fiber reinforced polymer (GFRP) and glass reinforced epoxy (GRE) waste powder is coming from the composite pipe industries in Oman. This powder is harmful to human beings and aquatic animals. Composite pipe manufacturing industries are struggling a lot with the disposal of waste GRE dust. At present, these GRE waste materials are buried under the ground without affecting the groundwater. Impervious layers are constructed to avoid seepage and protect the groundwater. Composite pipes manufacturing industries are spending huge amounts to save the environment from GRE waste disposal. In this experimental project, GRE waste is added to the concrete to partially replace cement with proportions such as 5%, 15% and 25%. Mechanical properties of concrete with and without GRE waste are investigated like compressive, split tensile strength and bending strength tests experimentally. Experimental results depict that the spilt tensile value of concrete with 5% GRE dust is 6.45% more when compared to the control specimen. Also, compressive and bending strengths of concrete with 5% GRE dust are not much affected and it is almost equal to the control specimen.