Investigating features and output correlation coefficient of natural fiber-reinforced poly(lactic acid) biocomposites

Polylactic acid (PLA) material has the potential to be applied in various industrial fields, but this material has shortcomings in terms of mechanical properties, especially mechanical strength, due to brittleness nature of PLA. The manufacture of PLA composite material with the addition of natural fibers as a reinforcing phase is one of the methods to increase the impact strength and maintain the biodegradable properties of the material. However, in theory, there are many factors that affect the mechanical properties of composite materials, thus making the mechanical properties of composites more complex than monolithic materials. The mechanical properties of these composite materials can be predicted using deep learning by paying attention to the relationship between factors, and between factors and their mechanical properties. This relationship has an important role in creating a predictive model with good accuracy. Therefore, correlation analysis is an important thing to do. Correlation analysis was applied using Python programming language to determine the relationship between the impact strength of natural fiber-reinforced PLA biocomposites with its feature information: chemical composition, density, dimensions, surface chemical treatment of natural fibers, matrix-reinforcement volume fraction, and the type of processing used to manufacture the material.