ings end-on to the direction of the blast were damaged less severely than those side on. Buildings which were closed tightly received more damage than those which were left open. The damage to two heavily reinforced concrete shelters on Able and Charlie Islands was also documented by Project 3.5 (Figure 4.3). The damage inflicted upon these two massive instrument shelters, which were in the high-pressure region of approximately 130-psi peak overpressure (estimated 170-psi peak dynamic pressure), is significant background material for the design of maximum-protection shelters for either personnel or equipment. These shelters maintained their structural integrity, but failed functionally because of detail failure. Failure of the reinforced concrete, by either shear or tension, was predominantly around walls supporting doors and special windows and other structural discontinuities. The value of earth cover over structures, where practicable, was also indicated by the reduced damage to one of the two massive concrete structures, which was exposed to approximately the same 130-psi peak overpressure. Primary failures in the latter shelter were in ripping of portions of the concrete parapet and retaining walls at the rear of the shelter structure, which were torn off by the blast. A study of these failures may suggest corrective design improvement. Some of these improvements are appropriate for inclusion in future test-operation instrument shelters and other utilitarian structures. 4.2 CRATER SURVEY The immediate objective of Project 3.2 was to determine the dimensions of the apparent craters formed by Shots 1, 3, and 4 (Figure 4.4). The long-range objective of the work was to obtain data to assist in the prediction of the crater produced by any high-yield nuclear weapon. Two situations were of particular interest in this regard in Operation Castle: surface burst on land and surface burst in relatively shallow water. The major military interest in craters stems from the observation that the limiting distance of important damage to well-constructed underground fortifications lies only a relatively short distance outside the crater. For the prediction of such damage, the shape of the crater near the rim is more important than its shape or depth near the center. Although of somewhat less military interest, the crater produced by the surface shot in shallow water -— determining the limiting distance of damage to tunnels and the pos- sibility of damming a harbor by the formation of a crater with a shallow or above-water lip——was also of some concern. In planning for Castle, it was found that previous crater studies utilizing full-scalenuclear, high-explosive, and theoretical data had reached the point where additional full- scale-nuclear data was required. The interest was actually not in water oratoll detonations, but there was no prospec: of obtaining full-scale test data for surface or underground shots in continental tests. As a result, the participation in Castle represented a compromise measure. A second compromise was necessary: one between what was desired (measurement of true craters) and what was operationally and financially feasible (measurement of ap~ parent craters only). This compromise was also based on the lack of detailed information of the geologic structure at the detonation sites. Deep drilling and coring operations at Eniwetok Atoll in connection with Ivy indicated the presence of extensive sand lenses and other geologic nonhomogeneities, which made it uncertain that the demarcation line between the true and apparent craters could be readily ascertained by any means. In addition, the time interval between Shots 1 and 4 and the ready date for the shots follow- 61