Abstract: Abstract: In order to study the influence of inlet velocity and structural characteristics on heat transfer performance of herringbone plate heat exchanger, the realizable k-epsilon turbulence model was used to simulate the two-dimensional sinusoidal flow channels in herringbone plate heat exchangers. The internal velocity contours, pressure contours and temperature contours for different Reynolds numbers were described. The flow velocity distribution in the flow channel of the herringbone heat exchanger was not uniform due to the existence of the vortex flow. The velocity increased and then gradually decreased along the vertical line of the flow channel, and the areas with higher velocity were mainly concentrated in the middle and lower parts. The pressure increased and then decreased along flow direction, and there was an area with greater pressure at the apex. The area with lower temperature gradually became smaller along flow direction, and the temperature was higher at the concave wall due to the existence of the dead flow zone, which affected the heat transfer. But the increase of inlet velocity made the fluid distribution in the flow channel more uniform, and enhanced vortex mixing. Therefore, the average Nusselt number was greater for higher inlet velocity, that was higher heat transfer performance. However, an excessively high flow velocity caused a large negative pressure area near the outlet of the flow channel, resulting in a large pressure drop and energy loss in the flow channel. Through comparison and analysis, the optimal flow rate 0.4-0.5m/s was obtained. Furthermore, the effects of corrugated geometric parameters (corrugated spacing and corrugated depth) on heat transfer performance by average wall Nusselt number were investigated. For a given inlet velocity, decreasing corrugated spacing lead to a higher average wall Nusselt number, and thus the heat transfer efficiency was better. This was attributed to that the smaller corrugated spacing increased the number of contact points in the flow channel, which enhanced the disturbance between fluids, and thus strengthened heat transfer performance. However, further decreasing corrugated spacing caused a greater pressure drop between the heat exchanger plates and energy loss, so the optimal corrugated spacing was 12-16 mm. Additionally, the increase of corrugated depth led to a higher average wall Nusselt number. There were two main reasons, 1) the large corrugated depth promoted sufficient mixing of the large vortices and enhanced the heat transfer; 2) the deep ripple increased the heat exchange area, and thus improved the heat transfer performance. But higher corrugated depth could cause scaling problems in practical applications. Therefore, the optimal corrugated depth was 4-5mm. In comparison, the effect of the corrugated depth on the average wall Nusselt number in the flow channel was greater than that on the corrugated spacing. In conclusion, the law of fluid flow and heat transfer dead zone distribution in the flow channel is better presented, and the optimal flow rate is 0.4-0.5m/s, corrugated spacing is 12-16 mm, corrugated depth is 4-5 mm. The simulation results are of great significance for improving heat transfer efficiency of the herringbone plate heat exchanger and saving engineering energy consumption.