Abstract:
Full-scale impact tests made on a Sunderland at the Marine Aircraft Experimental Establishment were in disagreement both with basic model tests and existing theories. The existing information on theoretical, model land full-scale work has therefore been examined to clarify as far as possible the position at the time of writing. Detailed considerations have had to be limited to a few selected numerical examples of impact but the treatment has been kept as general as possible. None of the general theoretical solutions convincingly fit both model and full-scale experimental evidence. Fair agreement is obtained between model data and theoretical prediction but not between the Sunderland full-scale data and theory as so far taken. There is generally a much longer time to peak accelerations full scale, although the peak-acceleration values are of the right order. Somewhat better agreement is obtained with some few American results but the basic theoretical errors are still in the same sense. Use of a more empirical theory does, however, give good agreement in one case-the Martin Model 270. These theories are all based on an assumption of transfer of momentum to an associated mass of water and the selection of the value of this is open to a whole range of interpretations. Assuming this approach to be valid, the value depends very much on the geometry of the wetted hull surface and how this varies with time. There is evidence that the consequent errors in time could be very large for small aspect ratios, which are normally much less than 1 in seaplane impacts. This error is also much increased in the same sense by the effects of rotation, flexibility, possible flow sticking to the afterbody and above the chines. It is also likely that a quasi-static assumption of the values of speed draught and incidence is not permissible when deducing associated mass values.
