The application of ultrasound to the casting process can improve the quality of the casting through the effects of grain refinement, degassing, wetting, deagglomeration and dispersion. To predict an effective energy range of liquid metal, computational fluid dynamics (CFD) study has recently developed an analysis tool capable of modeling cavitation and acoustic streaming in liquid metal. In general, the generation of a cavitation bubble is a very small region directly under the sonotrode in the fluid, and acoustic streaming is the main factor for the scale-up of the ultrasound process because it is a non-linear physical effect which can assist in effective dispersion of the cavitation bubble with the propagation of ultrasonic wave in the liquid. Therefore, this study is focused on the evaluation of the effect of ultrasonic injection conditions on acoustic streaming by particle image velocimetry and CFD modeling.