SBIR-STTR Award

Reaction wheel disturbances are some of the largest sources of noise on sensitive telescopes. Such wheel-induced mechanical noises are not well characterized. Disturbances can be amplified by wheel and other structural dynamics (for example, isolator modes), that are coupled to gyroscopic effects and therefore are wheel speed dependent. Tonal disturbances and wheel structural modes thus sweep across a frequency band. When one or more tones crosses the frequency of an observatory mode, a large jitter response will result. These higher harmonic effects have not been very significant in the past, for larger spacecraft having looser pointing requirements. However, many current and planned missions have much tighter pointing requirements than past missions. The higher harmonic wheel disturbances are being found to interact with structural modes to cause jitter exceedances. The lack of knowledge of higher harmonic wheel disturbance behavior forces engineers to carry more conservatism in the observatory design, resulting in potentially higher costs. The proposed innovation is a modeling tool that will create a hybrid physical/empirical wheel disturbance model from reaction wheel Induced Vibration data, suitable for high-confidence on-orbit jitter prediction.

Reaction wheel mechanical noise is one of the largest sources of disturbance forcing on space-based observatories. Such noise arises from mass imbalance, bearing imperfections, and other sources. It takes the form of a number of discrete harmonics of the wheel speed, often also with a broadband noise component. Jitter problems can arise when harmonics sweep across observatory modes, and can be exacerbated by gyroscopically coupled spin-rate-dependent wheel structural modes that dynamically amplify the tonal and broadband disturbances. For a well-balanced wheel, higher harmonic forces can be on the same order as the fundamental, therefore when there is a jitter problem it can occur at very low wheel speed. These higher harmonics are generally less well-characterized than the fundamental. The proposed Reaction Wheel Disturbance Model Extraction Software (RWDMES) is a tool for fitting a hybrid physical/empirical model to wheel induced-vibration data. The physical model captures the wheel structure including gyroscopic effects, while the empirical model captures the harmonic forcing and broadband noise. The Phase I effort demonstrated the ability to fit a highly accurate harmonic/broadband/structural model, including 43 harmonics up to 14.63 times the fundamental, to measured wheel disturbance data in a point-and-click environment in about 2 hours. The benefits of the technology include reduced program effort to produce wheel disturbance models, leading to more accurate jitter prediction earlier in a mission. This in turn allows jitter problems to be mitigated at the design stage when changes are relatively inexpensive.