A reflected kinetics model for nuclear space reactor kinetics and control scoping calculations

Typescript (photocopy).Renewed interest in space nuclear applications over the past several years has motivated the study of a specialized reactor kinetics model. Long term civilian and dedicated military missions warrant the consideration of a kinetics model favorable for study of the feasibility of automatic control of these devices. The need to bridge this gap between reactor kinetics and automatic control in conjunction with the control drum design characteristic of next generation ''paper" space reactors inspired the development of a new Reflected Kinetics (RK) model. An extension of the conventional point kinetics (PK) model was done in order to explicitly correlate reactivity and the reflector/absorber control drums characteristic of space nuclear reactor designs. Open loop computations and numerical comparisons to analytic PK equations indicated that the RK model is a functional alternative to "equivalent bare point kinetics" in the analysis of moderate transients. It was shown that variations in the RK reflector-to-core transfer probabilities and coolant flow rate do indeed drive the transient differently than the lumped insertion of equivalent reactivity amounts in the core. These computations illustrated the potential importance of the utilization of variable coolant flow rate to aid control in space reactor systems limited by minimal drum reactivity worth. Additionally the Doppler reactivity shutdown mechanism was concluded to be the primary reliable means of safety shutdown in such systems. The structure of the RK equations proved to be advantagous for integration of automatic control. Unity feedback and two existing control techniques, state feedback and m odal, were applied in the solution of step and ramp servo mechanism problems. Advantages and disadvantages of state feedback control and the recently developed modal control as applied to the RK model were identified. Numerical computations of the resulting closed loop RK equations lead to the conclusion that the drawbacks of a large closed loop system and blind eigenvalue assignment in the state feedback case outweigh the advantantage of definite system stability. Even though stability cannot be guaranteed in the modal case, it was found that the structure of the RK equations was well suited for the success of this output control technique..