A study on analysis and control aspects of a doubly excited, double translator, linear switched reluctance machine

Switched reluctance machines are increasingly being considered for various applications in recent times. As with conventional machines, a linear switched reluctance machine configuration is also feasible. In addition to retaining the advantages of the rotary switched reluctance motor, the linear version offers the flexibility of some unique configurations. In particular, a linear machine typically has to deal with attractive forces between the stationary and translating parts. This is particularly true for switched reluctance machines. This dissertation presents results of analysis of a linear switched reluctance machine through analytical and finite element techniques. In addition, a control algorithm is developed for operation as an actuator. Experimental data from a proof-of-principle machine yield good correlation to predictions.; Use of the doubly excited configuration alleviates if not eliminates this problem. Additionally, use of a double translator section yields a "force multiplying" effect, which permits an increase in output, thrust capability of the motor.; The analysis of this machine was based on equivalent magnetic circuits using the permeances of quasi-flux tubes in the air regions. Factors were developed to help account for saturation effects. Predictions of force were found to be in good agreement with measured data.; Further detailed analysis was carried out using finite element methods. To avoid the cost and complications of three dimensional finite element analysis, a novel scheme is presented that helps account for third dimensional effects in the more economical and easier to use two dimensional finite element analysis. Again, the predictions were found to be in excellent agreement with measurement.; A scheme is presented for constant force operation as an actuator. No assumptions of magnetic linearity are made. The algorithm was tested from a static point with measurements and simulated using a new software to study dynamic operation. The results are presented that show the validity and applicability of this approach.