The fault propagation process of rotating machinery is affected by multi-physical field coupling and multi-level causality, and traditional methods have limitations in data fusion and causal inference, which make it difficult to effectively analyze the dynamic mechanism of fault chain propagation. In this paper, we propose a fault propagation analysis method for rotating machinery based on the fusion of multidimensional data and causal chain explanatory structural model, aiming to realize the dynamic modeling of fault paths and causal explanatory analysis. Through CCISM, a multi-level causal chain network model is constructed to analyze the explicit causal logic and implicit correlation paths of fault propagation, and the research results provide a theoretical basis for optimizing the reliability of rotating machinery systems and formulating active maintenance strategies.