The primary requirement for the flexure control system, if it is found to be necessary, is that it limit image motion to 0.1 pixels rms over a one hour exposure.
The two initial options for flexure control use either the collimator or the detector array as the active element in the control system. Translation of the detector array will require precision translation stages, either inside or external to the evacuated CCD enclosure. Alternately, tip/tilt motion of the collimator barrel, or a subset of the collimator optics, could provide the flexure correction.
The flexure control system may operate in an open-loop mode, with corrections based on a look-up table of flexure versus instrument rotation angle. Alternately, a closed-loop control system, based on error signals from additional flexure control CCDs in the focal plane, could be used. In either case, separate control systems would be required, depending on which camera was in use.