The nanoparticles aggregation aspects on the chemically reactive unsteady flow of alumina-water based nanofluid: A Keller box approach with applications of wavelet physics inspired neural networks

The present study explores the unsteady flow of a nanoliquid past a stretching cylinder with the consequence of heat source/sink and chemical reaction. Additionally, the effect of nanoparticle aggregation, convective boundary conditions, and magnetic field on the liquid flow is taken into consideration. Utilizing similarity variables, the modeled equations are transformed into dimensionless ordinary differential equations (ODEs). Further, the obtained ODEs are numerically solved by using the Keller box method. Moreover, the physics-informed neural network (PINN) is applied to analyze the flow, heat, and mass transport features. Graphical illustrations are used to display the influence of various parameters on the velocity, concentration, and temperature profiles for aggregation and without aggregation cases. As the value of the magnetic parameter increases, the temperature and concentration profile upsurge, but the reverse trend can be seen in the velocity profile. The concentration and temperature profiles rise as the unsteadiness parameter increases, but the velocity profile declines. The concentration, velocity, and temperature profiles are strengthened by an improvement in the curvature parameter value. The intensification in the values of the chemical reaction parameter declines the concentration.