A new predictor–corrector approach for the numerical integration of coupled electromechanical equations

Abstract

Summary In this paper, a new approach for the numerical solution of coupled electromechanical problems is presented. The structure of the considered problem consists of the low‐frequency integral formulation of the Maxwell equations coupled with Newton–Euler rigid‐body dynamic equations. Two different integration schemes based on the predictor–corrector approach are presented and discussed. In the first method, the electrical equation is integrated with an implicit single‐step time marching algorithm, while the mechanical dynamics is studied by a predictor–corrector scheme. The predictor uses the forward Euler method, while the corrector is based on the trapezoidal rule. The second method is based on the use of two interleaved predictor–corrector schemes: one for the electrical equations and the other for the mechanical ones. Both the presented methods have been validated by comparison with experimental data (when available) and with results obtained by other numerical formulations; in problems characterized by low speeds, both schemes produce accurate results, with similar computation times. When high speeds are involved, the first scheme needs shorter time steps (i.e., longer computation times) in order to achieve the same accuracy of the second one. A brief discussion on extending the algorithm for simulating deformable bodies is also presented. An example of application to a two‐degree‐of‐freedom levitating device based on permanent magnets is finally reported. Copyright © 2015 John Wiley & Sons, Ltd.

Null-Flux Coils in Permanent Magnets Bearings

“…It is based on a 3D integral formulation it can simulate complex electromagnetic devices, transforming the solution of the field equations in the solution of an equivalent electric network [19], [20]. The code allows to obtain a set of equations that can be seen as the equilibrium equations set of an equivalent electric network.…”

“…x represents the center of mass of the rotor, F is the resultant of the forces, G I is the tensor of inertia and G T is the resultant of the torques; both are referred to the center of mass of the rotor. Since the differential equations for the electrical and dynamic equilibrium are coupled, the resulting system is time varying; the equations are integrated by using a recently proposed predictor-corrector [19]. The solution of the model allows to know all the electromagnetic and mechanical quantities which govern the operation of a given 6 DOF device.…”

Optimization of a Novel Magneto-Rheological Device With Permanent Magnets

“…Once the values of the elements of the equivalent magnetic network have been calculated, the Nodal Analysis Method for the electrical network is used to solve the problem [37], [38]. As an example of the set of equations obtained by the model, in the right side of Figure 3 …”

Analysis of the Static and Dynamic Behavior of a Non Hysteretic Superconductive Passive Magnetic Linear Bearing by Using an Electromagnetic Integral Formulation

“…Once the currents are known, the force f kj exerted on the kth elementary volume due to the jth one can be calculated as follows [27]. In correspondence to a given configuration, the dynamic interactions between couples of elementary volumes (one on the fixed body and one on the moving one) were firstly evaluated in the hypothesis of unit current on them.…”

“…However, the electrical equations are integrated by using a conventional integration scheme (Single Step Time Marching) [27]. The solution of the model allows to know all the electromagnetic and mechanical quantities which govern the operation of a given 6 DOF device.…”