Purpose
The purpose of this study is to describe the combustion of a magnesium particle falling into a hot oxidizer medium.
Design/methodology/approach
The governing equations, including mass, momentum and energy conservation equations, are numerically solved. Afterward, the influences of effective parameters on the temperature distribution and burning time are investigated. Artificial neural network (ANN) is applied to approximate the particle temperature as a function of time, diameter and porosity factor. To obtain the best arrangement of the ANN structure, an optimization process is conducted.
Findings
The results show that by considering variations of the particle size, the maximum temperature increases compared to the case in which the particle diameter is constant. Also, the ignition and burning times and the maximum temperature of the moving particle are lower than those of the motionless particle. Optimum network has the best values of regression coefficient and mean relative error whose values are found to be 0.99991 and 1.58 per cent, respectively.
Originality/value
In this study, particle size varies over the combustion process that leads to calculation of particle burning time. In addition, the effects of the motion and porosity of the particle are examined.