In order to enhance the heat transfer intensity and uniformity in an air impinging freezer, an impinging freezing experimental table was designed as the research object, and a new nozzle structure—circular funnel nozzle was proposed. The numerical simulation technology was used to simulate the flow fields in the impinging freezing experimental table, which was testified by experiments. The flow medium was air, and the simulation process assumes： （1） The wall of static pressure chamber was adiabatic. （2） Air was an incompressible, homogeneous viscous fluid. （3） During the normal operation, the internal flow field of the model was regarded as steady state. The three-dimensional continuity equation, the momentum equation, the energy equation, the kinetic energy k equation and the turbulent dissipation ε equation were used. The cooling air inlet and outlet pressure were 250 Pa（P_in） and 0 Pa（P_out） respectively. For the frozen area, the cooling air inlet and outlet temperature was 230 K and 235 K. The mass flow rate at the cooling air inlet is 0.064 4 kg/s. The thermal conductivity of steel strip was 16.3 W/（m·℃）. Using the Plackett-Burman design, the significant factors for the average Nusselt number on the steel strip surface were obtained from the factors of outlet diameter D_E, funnel width L₃, funnel height L₁, jet height L₂, nozzle number N, nozzle spacing S and nozzle-to-surface distance H. These significant factors were funnel width L₃, jet height L₂ and nozzle number N. The others had little effect on the average Nusselt number. So in the next study, these factors adopted the median values. Then, using the Box-Behnken design, a mathematical model between those three significant factors and the two response values which were average Nusselt number Nu_ave and the heat transfer uniformity index η on the steel strip surface was established to determine the optimal structural parameters. The F value in the regression equation can be used to determine the influence of the factors on the response value. Therefore, the order of the factors affecting the average Nusselt number and the heat transfer uniformity index on the steel strip surface was nozzle number>funnel width>jet height. The interaction between the funnel width and the number of nozzle rows, the jet height and the number of nozzle rows were the most significant in the response surface analysis figures, which were consistent with the results of the variance analysis. The results showed that the optimal structural parameters were funnel width L₃=17 mm, jet height L₂=50 mm, nozzle number N=3. Substituting these parameters into the second-order polynomial equation, the average Nusselt number on the steel strip surface Nu_ave=448.68, heat transfer uniformity index η on the steel strip surface=0.232 9. The numerical simulation value was generally consistent with the predicted value, so it was reasonable to optimize the average Nusselt number and the heat transfer uniformity index on the steel strip surface by using the Box-Behnken design.