Abstract

The need for sustainable development requires environmentally friendly solutions; this fact leads engineers from all research areas to seek and employ methodologies that integrate new construction materials. One of these areas of study is the slope stability in rocks and soils. Where traditional materials such as concrete have been substituted for biomaterials, which are few known in terms of mechanical behavior and modeling. The present work trains a model of Artificial Neural Network (ANN) for modeling the tensile strength of the roots of Vetiver plants, considering the diameter of the root, its age and the soil type. A total of 40 ANN models were trained using 100 laboratory tests, selecting the best model according to their mean square error and generalization capacity. The model shows a better fit than the potential type models proposed in the literature; it explaining the variability of the data for diameters smaller than one millimeter. it was found that the resistances for these diameters are inversely proportional to the age of the plant.