472 - 2 - 



The Model was filled with water and connected to a hydraulic supply, the pressure so applied 

 Being measured by a gauge connected to the Hodel at the hole marked X in Plate 1. This gauge 

 read up to 260 lbs. /square inch in 5 lbs. /square inch intervals while an additional gauge, reading 

 to 650 lbs. /square inch in intervals of 10 lbs. /square inch, was connected to the inflcw pipe at 

 a distance of about 10 feet from the Model. There was no significant diffurence between the 

 readings given by the two gauges. 



In addition to readings taken during the trial by the preceding method, measurements of 

 permanent deflection after th.> trial were taken at the points on the plate indicated in Figure 1 

 and Table 2. 



Measurements of final pull-in at the edjes were also taken as indicated in Figure 2 and 

 Table » while measurements of pertianent strain were fiiade along the centre lines ODj, LLj (Figure l) 

 by taking strain rubbings, before and after trial, of centre punch marks spaced at approximately 

 1 inch intervals. 



Description of trial . 



It was intended prior to the trial that the pressure should be increased in stages, at the 

 end of each stage the pressure being held constant while the deflection scales were read. This 

 procedure was adopted for the pressures shown in Table 1 up to and including 120 lbs. /square inch 

 but at this latter pressure a slight leak developed at the edge fixings and it became difTicult to 

 maintain the pressure at a steady value. An immediate decision was accordingly made tc continue 

 the trial by increasing the pressure slowly and to concentrate on reading the central deflection 

 scale only. In this way the central deflections shown in Table 1 for pressure above 12C lbs./ 

 square inch were recorded. It should be noted that in this second stage the central deflection 

 increased by Kss than 6 inches in a matter of minutes so that any inertia resistance offered by 

 the plat-^ was completely negligible compared with the applied pressure. 



At a pressure of 275 IDs. /square inch the trial ceased abruptly by shearing of a clamping 

 bolt in the middle of a long side and with water issuing rapidly at this point the pressure dropped 

 suddenly to zero. 



Discussion of Results. 



1. Shape of deflected plate. 



To ^void shearing of the plate rouna the edges the inner edge of the clamping frame was 

 rounded off to about -^ inch radius as shown in Figure 1, and the unsupported target plate area was 

 thus given by the dotted lines en Figure 1 ^t the conmencement of th? trial but decreased slightly 

 during the trial. To the order of accuracy of the measurements the unsupported area of plate can 

 be taken as 6 feet by 4 feel while the points 1 tc ?, at which deflections were measured during the 

 trial, are the quarter-points of this area. 



The deflections tak^n during the trial are tabulated in Table 1 while the pernanent 

 deflections after the trial are given in Table 2. The deflections quoted in Table 1 for zero 

 pressure crrrcspond to an initial asymmetric slackness in the plate. Comparison of deflections, 

 ir. both Tables 1 and 2, for points, e.g. 2 and 8, nominally symmetrically situated with respect to 

 the centre of the plate show thjt the deflection ot the plate under the uniform pressure and after 

 its r-lease was reasonably symmetrical. 



In report A, Professor G. I. Taylor developed the ttieory of a plate strained plastically 

 by uniform lateral pressure and obtained a solution for a rectangular plate on the assumption that 

 the plate bolvi'. cd as a membrane exerting uniform stress in all directions in its plane. with 

 this assumption the shape of the deflected plate is the same as that well-known for a soap film; 

 contours of which shape are shown by the dotted curves in Figure 3 in which the full curves give 

 the contours of ihe d^'-flected plate after trial as plotted from the data in Table 2. In the 

 comparison the contours have bepn drawn for equal central deflection and it is seen that the 

 theoretical and observed contours ?re in reasonable agreement. As a further check the theoretical 

 ratio of central to mean deflection is 2.06 while from Table 5 the corresponding ritiowas 2.03 after 



