Estimating Young Modulus of Elasticity of Terminalia catappa: A Machine Learning Approach

Abstract

The purpose of this research was to evaluate the potential of Magnetic Resonance Spectroscopy (MRS) in estimating Young’s modulus of elasticity of wood species. To do so, Terminalia catappa, a wood species of common occurrence was chosen and its mechanical properties such as bending strength, compression parallel to the grain, and shear parallel to the grain properties were determined using testing methods for small and clear specimens of wood with the British (BS 373, 1957) and American Society of Testing Materials’ specifications (ASTM D143, 1983s. The results showed that at 18% moisture content the wood has a density of 520 kg/m³ with a mean modulus of rupture of 86.04 Mpa, compressive strength parallel to the grain of 42.02 Mpa, modulus of elasticity of 10,500 Mpa, and shear strength parallel to the grain of 16.42 N/mm². This dataset was used on machine learning approaches such as decision tree and random forest.  The estimated value of Young’s modulus using the machine learning models varies between 1000 to 13000 MPa. The obtained results indicated that the use of Magnetic Resonance Spectroscopy (MRS) is an efficient tool for predicting Wood-Young’s modulus. This research paves the way for further investigations on the application of MRS and machine learning for predicting a wider range of wood properties. By employing machine learning techniques such as decision trees and random forests, researchers can develop robust models for estimating Young's modulus in other wood species. This approach allows for leveraging large datasets that encompass various influencing factors, ultimately leading to more accurate predictions compared to traditional methods.