Turning off a high electric current, especially at high voltages, needs special arc extinguishing devices, because the current makes the insulating gas between the separating contacts conductive, resulting in an electric arc. Between the splitter plates of such an arch chute, the motion, and finally the breaking of the arc depends on the balance of the attracting magnetic forces from the ferromagnetic plates, on the repelling forces between two arc columns at the two sides of a plate and on the viscous forces from the insulating gas. In order to understand the behavior of the arc chute and optimize the splitter plate arrangement, Lucy Electric engineers needed to simulate this motion during the development of a medium voltage switchgear.
A high speed video recording of a breaking test can be seen in the first part of the animation. The electric arc in the SF6 gas at the current level of switchgear formed a contracted channel. This fact led engineers to the idea of modeling the short arc column burning between the splitter plates of the arc chute as a thin cylinder, and of taking the viscous forces as a drag force into account. The equations of motion of one arc column, including an electromagnetic (EM) acceleration and a braking term due to drag, are first and second order nonlinear differential equations for the velocity and the position. It was assumed that the variation in the relation between the position and the EM force was negligible within a tiny distance (step size). The previous step gave the initial velocity, and an ANSYS Mechanicalanalysis provided the EM force at the beginning of the step. The EM force itself varied within one step as the current changed. Also, the drag force changed proportionally to the velocity squared. An APDL script solved the equations of motion numerically, determining the displacement and velocity of the arc columns at the end of the step. In the next step, it repositioned the arc columns, regenerated the geometry and the mesh, and ran the Mechanical analysis.
This calculation method made it possible to compare several splitter plate arrangements within a reasonable time. Engineers tested some of those that performed better during the simulations, and the test and simulation results were in a very good agreement (at the end of the animation, you can see that the calculated motion paths were very close to the traces the arc columns left over the splitter plate surfaces). The simulations helped to reduce the number of prototypes and costly test shifts during the development process.