UNSTEADY MULTI-BODY SEPARATION TRAJECTORY PREDICTION TECHNOLOGY

In diverse multi-body separation scenarios—including internal weapon deployment, stage separation of two-stage-to-orbit vehicles, submunition dispersion, fairing/cockpit canopy separation, and ejection seat operation—the motion trajectories of separated components are frequently subject to unsteady interference flow fields, resulting in trajectory dispersion. This dispersive behavior may lead to deviations of the actual separation trajectory from the predefined path, thereby inducing critical safety incidents during the separation process. Consequently, accurate prediction of the dispersion range of unsteady separation trajectories is of profound engineering significance. The precision of trajectory dispersion range prediction is highly dependent on an adequate number of unsteady trajectory samples; generally, a larger sample size corresponds to more reliable prediction results. However, in current domestic research practices, whether relying on numerical simulations or wind tunnel tests, only a single unsteady separation trajectory can be acquired per simulation run or test campaign. This poses substantial challenges in meeting the massive data requirements for trajectory range prediction, both in terms of data acquisition efficiency and associated costs. To address this bottleneck, this study proposes a simulation-based prediction method for separation trajectories using the Monte Carlo approach. The core of this method involves constructing an unsteady instantaneous aerodynamic model for the payload and coupling it with the six-degree-of-freedom (6-DOF) motion equations of rigid bodies. This coupling enables the rapid generation of unsteady separation trajectories with a specified sample size, thereby facilitating the determination of the payload’s separation trajectory range. To validate the effectiveness of the proposed method, a typical unsteady multi-body separation scenario—“weapon launch from a backward-facing step internal bay”—was selected as the research object. Trajectory range prediction was conducted using the proposed method, and dynamic similarity deployment tests were simultaneously performed under identical initial separation conditions. The results demonstrate that the predicted separation trajectory range fully encompasses all test-acquired separation trajectories, which comprehensively confirms the reliability of the proposed method.