Abstract: The self-impact seal represents an innovative sealing technology, distinguished by its numerous internal flow channels and complex structural design. Within these flow channels, the main and tributary flows exhibit significant impact effects at their intersection points. The research findings indicate that the fluid inside the self-impact seal does not always follow the typical flow direction of decreasing pressure (i.e., forward flow). Instead, under certain operating conditions, the tributary channels can experience reverse flow. Investigating the underlying causes of this phenomenon, as well as studying the evolution of the flow field under different operational conditions, is crucial for the design and optimization of the self-impact seal structure. This study delved into the flow field and leakage characteristics of the self-impact seal across a range of structural and operating parameters. The research demonstrated that, under specific conditions, reverse flow could occur in the tributary channels, and the evolution of flow direction in the flow field was influenced by multiple variables, resulting in a complex behavior. By analyzing the leakage trends, the study established a logical relationship between the evolution of the flow field and throttling efficiency. Building on this analysis, the study proceeded to optimize the cascading-type self-impact seal structure. The results showed that reverse flow within the tributary channels significantly contributed to enhanced leakage suppression efficiency. The sealing gap was identified as the primary factor that governed the forward and reverse flow behavior within the channels. When the gap was relatively small, both the main and tributary flows followed a forward direction. However, when the gap exceeded 200 μm, reverse flow began to emerge in the tributary channels. On the other hand, changes in pressure had a relatively minor impact on the flow direction. In contrast, an increase in rotational speed leaded to a shift in the fluid flow within the tributary channels from reverse to forward flow, thereby reducing the proportion of reverse flow. For certain key parameters, such as the shunt angle (α), flow distance (l), and number of sealing stages (Z), which could not be optimized solely based on leakage performance, the study proposed a quantitative approach to optimize these parameters based on the variation of the tributary reversal rate. The optimal values for the shunt angle (α), flow distance (l), and sealing stages (Z) were determined to be 50°, 7 mm, and 20, respectively. These values provided valuable technical support for designing self-impact seals under real-world operational conditions. Additionally, in consideration of manufacturing feasibility, the study introduced a regular-type self-impact seal structure as an alternative to the cascading-type design. The results showed that, compared to the cascading-type structure, the leakage of the regular-type structure increased by less than 10%. However, the regular-type structure demonstrated superior throttling efficiency at smaller sealing gaps. Moreover, the regular-type structure was capable of reducing manufacturing costs by over 60%. The proposed regular-type self-impact seal structure combined high efficiency in leakage suppression, excellent manufacturability, and significant cost savings. The findings of this study provided strong technical support for the rapid industrialization of this innovative sealing technology, offering substantial practical and industrial value for its widespread application in various industries.