Numerical Simulation of Adaptive Control: Application to Unstable Solid Rocket Motors

Abstract : This article describes current developments in the numerical simulation of active control. The objective of this investigation is to devise software tools for the development of active control. The present approach uses a numerical simulation of the system based on the Navier-Stokes equations. It differs from the more standard simulations relying on lower order dynamical models. The main difficulties associated with the present strategy are related to the representation of the actuator in the flow simulation module and with the interfacing of this module with the adaptive control routine. These issues require careful treatment to obtain a suitable commercial model of How control. It is first shown that the actuator may be described by a distribution of sources in the field. The time stepping needed by the flow simulation module and by the control unit differ widely (the ratio between the time steps is of the order of 100 or more). This constitutes a source of perturbation and it may introduce unwanted high frequency components in the flow simulation. It is shown that this problem is alleviated by placing numerical filters at the controller input and output. A set of calculations are carried out to simulate vortex shedding instabilities of a simplified solid propellant rocket. These instabilities are then adaptively controlled. This example serves to illustrate the simulation methodology and provides insights into the operation of the flow controller.