Argon gas has been widely injected into submerged entry nozzles (SENs) to prevent nozzle clogging during the slab continuous casting (CC) process. Meanwhile, argon bubbles also reduce the negative pressure and air aspiration, [1,2] promoting the mixing and flotation of inclusions from the molten steel in the CC mold. [3][4][5][6] Abnormal larger bubbles float up close to the outer wall of the SEN and escape from the top surface quickly, which causes slag entrainment into the molten steel. [7,8] Smaller bubbles are easily captured by the solidified shell, causing segregation, [9] pencil blister, and sliver defects. [10][11][12] Therefore, it is necessary to better understand and optimize the bubble size and distribution in the CC mold to avoid bubble-related defects.The gas injection method, casting parameters, [13,14] nozzle design, [15] physical properties of the gas and liquid phases, [16] and temperature have a great effect on the bubble size and distribution inside the SEN and the CC mold. Water model measurements have shown that bubble diameters decrease with a higher casting speed and submergence depth of the SEN, as well as with a lower gas flow rate. [14,[17][18][19][20] However, the effect of temperature on the bubble size has been rarely studied in the water model of a continuous steel caster. The gas injection method in the water model experiment included the gas channel of the stopper rod, [21][22][23][24][25] small holes or porous refractory in the upper tundish nozzle (UTN), [16,[26][27][28] and the slide gate of the SEN. [11,29] Bai and Thomas [16] investigated the effects of the liquid velocity, gas-injection flow rate, injection hole diameter, and gas composition on the initial bubble-formation behavior. The gas was injected through a square tube below the tundish tank outlet. Kasai and Iguchi [29] obtained the critical value of the water flow rate to carry the gas injected through the slide gate into the mold. The variation of the bubble size inside the SEN and the mold was also studied. [30] Lee et al. [31] investigated the effect of refractory properties on the initial stages of bubble formation. Air was injected through the porous refractory in the center of the UTN wall. For gas injection from the stopper rod tip, single [21,23,31] and multichannel [32,33] blowing methods have been studied in recent years. Gas injection from a single channel created abnormal large bubbles or gas pockets, and produced slug or annular flows inside the SEN, resulting in unstable flow and violent level fluctuations, whereas gas injection from multiple channels produced a more stable surface level in the mold and retarded the clogging on the stopper rod tip. [24] However, studies on the effect of the hole diameter of the gas channel of the stopper rod on the bubble size distribution are little reported.In this study, the bubble size and distribution in the CC mold were investigated and measured using a 0.5-scale water model. First, the effect of the water temperature on the bubble size and