Derek N. Dyck

Automated design of magnetic devices by optimizing material distribution

This paper presents a method for automatic design of magnetic devices. The process can be used to determine both the topology and shape of the required device, as well as the materials to be used. Sensitivity information derived directly from Maxwells equations is used to direct the optimization process. The method is applied to the design of a jumping ring (magnetic bearing).

A method of computing the sensitivity of electromagnetic quantities to changes in materials and sources

This paper presents a method of computing the sensitivity of electromagnetic quantities (e.g. induced voltage, force, field uniformity) to perturbations in sources and material parameters. The method applies to systems with linear materials and time harmonic source currents; static problems are included as a special case. Sensitivity is equivalent to the derivative with respect to the distributed system parameters. Therefore, with the expressions for the sensitivity presented in this paper, first order optimization methods can be applied to optimization problems in electromagnetics. Applications are in the area of inverse problems, including electromagnetic device design, device optimization, and imaging. >

Composite microstructure of permeable material for the optimized material distribution method of automated design

In the automated design of magnetic devices using the optimized material distribution method, the material properties are allowed to vary continuously in each cell of the discretized design region. This leads to the appearance of intermediate materials whose properties do not correspond to an actual solid material. However, a correspondence can be made to composite materials, which are constructed from of a number of different solid materials. This approach permits accurate evaluation and stable optimization of the performance of a device. This paper derives the general composite microstructure of permeable intermediate materials.

Response surface models of electromagnetic devices and their application to design

The use of a response surface in the design of an electromagnetic system is described. The surface is generated using a minimum number of points based on the design of experiments. The final surface is integrated in SPICE to provide dynamic performance modeling.

The logical control of an elevator

This paper presents a detailed example of the design of a logical feedback controller for finite state machines. In this approach, the control objectives and associated control actions are formulated as a set of axioms each of the form X implies Y, where X asserts that (i) the current state satisfies a set of conditions and (ii) the control action y will steer the current state towards a given target state; Y asserts that the next control input will take the value y. An automatic theorem prover establishes which of the assertions X is true, and then the corresponding control y is applied. The main advantages of this system are its flexibility (changing the control law is accomplished through changing only the axioms) and the fact that, by the design of the system, control actions will provably achieve the control objectives. The illustrative design problem presented in this paper is that of the logical specification and logical feedback control of an elevator. >

A density driven mesh generator guided by a neural network

A neural network guided mesh generator is described. The mesh generator uses density information provided by the neural network to determine the size and placement of elements. This system is coupled with an adaptive meshing and solving process, and is shown to have major computational benefits compared with adaptation alone. The novel mesh generator is faster than conventional meshing approaches and allows the solver to take fewer adaptive steps in generating an optimum solution. It provides a method of hiding the meshing process from the user in a totally automatic system. >

An NDT pulse shape study with TEAM problem 27

This paper performs a study of the effect of pulse shape on signal amplitude for TEAM Problem 27-Eddy Current NDT and Deep Flaws. A procedure is described to reduce the discretization errors in the finite element mesh. The results are compared with experimental waveforms. It is found that beyond a certain point the signal amplitude is insensitive to further decreases in the pulse turn-off time.

A Performance Model of an Induction Motor for Transient Simulation With a PWM Drive

The goal of this paper is a method for the simulation of an induction motor driven with a pulse width modulation (PWM) three-phase bridge. The approach taken here decouples the circuit simulation from the electromagnetic field simulation by first constructing a performance model of the machine using 2D finite element (FE) analysis. The model consists of the complete matrix of the self- and mutual inductances of the stator windings and each of the rotor bars. Skew is taken into account by a correction applied to the winding-bar inductance entries in the matrix.

Internally Consistent Nonlinear Behavioral Model of a PM Synchronous Machine for Hardware-in-the-Loop Simulation

Hardware-in-the-loop simulation allows hardware implementations of some components of a system to be tested in conjunction with other components that are only emulated based on behavioral models. For example, a physical prototype of an engine control unit for a hybrid vehicle can be tested with virtual prototypes of the motor and vehicle. This paper presents a new method to obtain realistic behavioral models of PM synchronous machines from finite element analysis. The new method guarantees that the model will be internally consistent; for example, it will not violate the principle of conservation of energy.

Transient analysis of an electromagnetic shaker using circuit simulation with response surface models

A numerical model of an electromagnetic shaker (a linear actuator used for active vibration control) is constructed based on magnetostatic analyses and response surface models. This model is incorporated into a transient circuit simulator to obtain the frequency response of the shaker and its behavior at resonance. The simulation results are compared to experimental measurements.