Abstract: The paper investigates the field network constructed by classification scheme codes using the American Physical Society（APS） dataset, and analyzes the influence of knowledge cascading by use of raw impact per paper（RIP）and filed similarity metrics, then constructs a field cascade tree model, and then finds that there is a secondary cascading effect from root domains to subdomains. Furthermore, the cascade tree model is analyzed using structural features such as cascade depth, cascade size, and cross-domain propagation speed. The results show that the cross-field propagation speed of root domain knowledge initially increases and then decreases during the knowledge cascading process, while knowledge propagation in some field exhibits discontinuity.