Armature Motion Characteristics of Electromagnetic Propulsion Devices Considering Dynamic Armature Conductivity

In electromagnetic propulsion devices (EPD), the equivalent armature conductivity varies dynamically with bulk conductivity and contact resistance, which directly affects the motion characteristics of devices. However, under extreme electromagnetic, thermal, and force shock conditions, real-time in-situ measurement of armature conductivity remains challenging. To improve the computational accuracy of numerical models, this paper proposes a method for analyzing the motion characteristics of EPD based on dynamic armature conductivity inversion. First, real-time armature velocity and spatial magnetic field information at measurement points are obtained through electromagnetic propulsion experiments. Subsequently, a dynamic armature conductivity inversion model is developed by integrating deep neural networks (DNN) with an improved gradient descent method (DGD). This model utilizes the measured magnetic flux density to correct the computed values of the parameterized model, thereby obtaining spatiotemporal characteristics of armature conductivity. Next, a transient electromagnetic, thermal, and force coupling finite element model of the armature-rail is developed based on the inversion results, enabling the analysis of the impact of dynamic armature conductivity on the device’s motion characteristics. Finally, experimental validation shows that the method considering dynamic armature conductivity significantly improves the computational accuracy of the finite element model. This work provides theoretical support for further reliability prediction and structural optimization design of EPD. It also offers new solutions for real-time in-situ measurement of material properties under extreme conditions.