Heavy oil burner is the key component of asphalt mixing equipment. However, heavy oil is not easy to atomize due to its high viscosity and surface tension, which results in low combustion efficiency and high pollutant discharge in heavy oil burners. Y-type atomizing nozzle is widely used in heavy oil atomization because of its simple structure and wide adjustment range. A large number of experiments and simulations were carried out on the atomization characteristics of heavy oil. However, the existing research only analyzed the atomization characteristics from the aspects of structure or atomization parameters, ignoring the gas-liquid two-phase flow characteristics in the nozzle. Therefore, the research on the gas-liquid two-phase flow and atomization characteristics of Y-type atomizing nozzle has an important guiding role in the design of heavy oil burner. In order to improve the quality of heavy oil atomization, computational fluid dynamics software were used to explore the influence of nozzle structure parameters (mixing chamber length, inlet diameter ratio, angle between air and fuel inlet) and atomization parameters (mass flow rate of oil, air inlet pressure, temperature of oil) on the gas-liquid two-phase flow and heavy oil atomization characteristics of Y-type nozzle. The results showed that the structure parameters and atomization parameters directly affected the oil film thickness and gas-liquid two-phase velocity difference in the nozzle. Under different parameters, the change rule of gas-liquid rate and droplet Sauter mean diameter was opposite. Considering the two atomization performance indexes, the optimal length of mixing chamber was 15~20 mm, the reasonable range of inlet angle was 60°~75° and the optimal inlet diameter ratio was 1.0~1.1. In order to ensure the better atomization effect of heavy oil, the air inlet pressure should be greater than 0.5 MPa, the above parameters obtained by simulation provide guidance for the structure optimization design and operation parameter setting of Y-type nozzle.