Key points are not available for this paper at this time.
Abstract The effect of side-levers and two-handed turning on accuracy of constant speed courses. A handwheel with a side-lover roughly simulating the turning-efforts of tank traverse gear almost doubled the mean following error, as compared with turning a similar handwheel with no side lover and against no load. The maximum time for which an average man could turn the wheel with load and side-lever without undue discomfort was about 7 min at 20 r.p.m. and 2 min at 200 r.p.m. Turning one wheel with two hands on its rim was worse than with one hand, in continuous turning at I, 3 and 10 r.p.m., owing to the difficulty of changing hand-grip smoothly. Positional following of slowly accelerating courses shows considerable learning over the first four courses, of two minutes each. At ail acceleration of 0.1 turn per sec2 of the handwheel, the following errors are increased two or five fold, as compared with laying on constant speed courses of the same instantaneous velocity. A simple technique has been developed for recording the corrective movements of the layer in response to large misalignments (e.g. in getting on to the target). This has been used with positional controls, but should be applicable also to velocity control and other systems. After some practice the layer seems to make a ballistic corrective movement roughly proportional to the misalignment about 0.25 sec curlier, and subject to a percentage error in amplitude of 5 to 15 percent on successive occasions. Further investigations of this kind should provide some basis for mathematical investigations of the best types of control. Additional informationNotes on contributorsK. J. W. CRAIK with the assistance of MARGARET A. VINCE With special reference to ground-tank and A.A. tank gunnery
CRAIK et al. (Tue,) studied this question.