In seeking to analyse the stress distribution in a complex structure, our object is really twofold. In the first place, if the stress in each part is accurately known, the strength of each part separately can be assessed on the basis of tests on simple units and the necessity for testing the complete structure may be avoided. In the second place, if the stress distribution can be expressed as a function of the geometry of the structure, the effect of varying that geometry, particularly with the object of realising Holmes’ one-hoss-shay, can be examined theoretically and arbitrary-features of the design can be eliminated. If our knowledge were complete, any structure designed to a maximal or minimal condition would not permit any variation from one ideal form, the design would be uniquely determined by the stated requirements.

Tests were made in the N.A.C.A. gust tunnel to determine the effectiveness of a long-period dynamically overbalanced flap in reducing aeroplane accelerations due to atmospheric gusts. For two gust shapes, one gust velocity, one forward velocity, and one wing loading, a series of flights was made with the flap locked and was then repeated with the flap free to operate. The records obtained were evaluated by routine methods.

The results indicate that the flap reduced the maximum acceleration increment 39 per cent, for a severe gust with a representative gust gradient distance of 8 chord lengths and that, for an extreme gust shape (a sharp-edge gust), the reduction was only 3 per cent. The results also indicate that the flap tended to reduce the longitudinal stability of the aeroplane. Computations made of the effectiveness and the action of the flap were in good agreement with the experimental results.