| Abstrakt: |
Artificial intelligence is the current trend in the world, which has taken the opportunity to advance in all its fields, particularly in scientific research. In materials engineering, the results obtained from classic methods such as experimentation, homogenization methods, or finite element methods have become input and validation elements for intelligent models to obtain more effective results in an optimal time frame. In this article, we discuss the use of artificial neural networks to determine the mechanical properties of biocomposites, which are the subject of much research due to the advantages they represent. The properties of these complex materials depend on various parameters, such as the behavior of the constituent materials, the percentage of the mixture, and the manufacturing process. In this work, our goal is to predict how polypropylene behaves elastically when reinforced with 15% various natural fillers. and we will study the impact of bamboo on polypropylene to test and validate our model. By exploiting the results of the Mori-Tanaka model, we were able to generate our dataset, with which we feed our feedforward backpropagation neural network and demonstrate that our biocomposite gained in terms of stiffness, marked by an increase in Young's modulus to 550.3 MPa, with better performance validation and a very good regression coefficient. [ABSTRACT FROM AUTHOR] |
