Numerical investigations of six-fold dendritic icing process in subcooled water subject to natural and forced convective environments

Abstract

Six-fold symmetric ice dendrite growth in subcooled water is investigated based on a GPU-accelerated hybrid lattice Boltzmann (LB) method and the cellular automation (CA) method. A reduction factor is proposed to reduce the discretization induced anisotropy in the hybrid method. The accuracy and feasibility of this hybrid method is validated by comparing simulated tip velocity of ice dendrite with existing experimental data under microgravity conditions, as well as comparing with the LM-K theory under small degrees of subcooling. Effects of natural convection and forced convection environments on dendrite morphology and tip velocity are further studied based on this improved method. It is shown that natural convection promotes ice dendrite to grow upward, and the arm growth velocity can be enhanced or weakened by natural convection, depending on the relative angle between intrinsic growth orientation and gravity direction. Under large natural convection effects, a growth competition phenomenon between different arms on a single ice dendrite is observed which had not been predicted by previous theoretical models. On the other hand, forced convection is shown to have strong enhancing effects on upstream arms and inhibit the growth of downstream arms. The arm growth is always enhanced with increasing inlet Re to various degrees for different relative angles between intrinsic growth orientation and gravity direction.