Although this study indicates that the SSPS would not be disturbed about the Z axis from its nominal orientation, it is realized that in a practical sense this will not be the case. Therefore, it is suggested that future structure and control interaction studies be expanded to include the effect of cross-coupling and torsional bending, as well as a system of non-linearities such as control system dead zones and saturation limits. Identification of Areas Requiring Further Flight Control Performance Analysis, — The results of the structural analysis [Reference 43] indicate the predominance of torsional vibrational modes, as well as cross-coupling between all modes. However, the math model presently employed in the controllability study cannot accept torsional modes and is restricted to the analysis of only uncoupled modes. While this model has proven adequate for this feasibility study, it is suggested that for future studies the math model be expanded from a planar model to a three-dimensional model which can accept all the bending modes identified in Reference 43. Furthermore, it is felt that an analysis of this expanded system could only be accomplished through the use of a non-real time simulation. A digital computer program has been developed at Grumman under contract #NAS 10 10973 which would fulfill the above requirements. It is felt that this program would provide a good interim capability for predicting the three-dimensional dynamic behavior of an SSPS. This time-history program can treat rotating or non-rotating satellites of any shape or mass distribution. The satellite may be idealized by using up to 100 masses. Since the rotatory inertia of each mass is considered, up to 600 physical coordinates may be employed. A maximum of 20 elastic modes are then used to represent the elastic motions of the vehicle and to reduce the number of coordinates being integrated. Control systems may be added to this program in subroutine form. The program has already been used to analyze the behavior of a realistically modeled space station containing five different types of control systems. Before using this program, the gravity-gradient and orbital centrifugal loads would be computed as a function of time by assuming that the vehicle is rigid and in perfect control. These loads would then be used as the time-dependent forcing function for the program and the dynamic response of the vehicle would be obtained. Since the gravity-gradient and orbital centrifugal loads are a function of the vehicle attitude, the true loads might differ from the computed loads. Thus, it may be necessary to recompute the loads based on the program's output and repeat the time-history run using the new loads. One or more iterations of this procedure may be required. Since gravity-gradient and orbital-centrifugal loads are primary disturbances, for the final analysis another program should be developed which is similar to the above program, but has the capability to include these effects more directly and accurately. The loads would be determined including the vehicle's orientation (response), and an iterative approach would no longer be required. The baseline attitude control system presently used for the SSPS is linearly modeled and uses proportional thrust actuators. It is recommended that, in the expanded study, the control system be modified to include such non-linearities as dead zone characteristic and control-thrust saturation limits. Although pointing accuracy has been found to be well with the ±1 deg pointing requirements, the actual attitude control system could conceivably have an attitude dead zone of ± 1 deg.
RkJQdWJsaXNoZXIy MTU5NjU0Mg==