diffractive phase) correlated reasonably well with the early drag phase of loading (out tn about 50 msec). Actually, for the Castle Structure 3.1 in which the target width was twice the length, the ARF net-loading prediction was not quite as good an approximation to the experimental data curve as was the AFSWP-226 prediction. However, the ARF method of computing the net blast load on a closed, diffractive-type structure stipulates that the target length must be “. . . greater than the height or half width, whichever is smaller.” For this reason, the net~loading ccmparison may not have presented the ARF method in its best light. On the busis ef the record p: svided by eleven pairs of gages on Structure 3.1, the reproducibility of the Wianko gage measurements was good. The probable error from the mean of the impulse ratios of each gage pair was only about 9 percent, while the probahie error of the arithmetic mean itself was only about 3 percent. In view of the failure of Project 3.1 to meet its original specific objectives, the question arises as to whether even a modest structure program should be included in avy future developmental test series at the EPG. A comparison of the planned shot schedule {estimaied yield and intended shot sites) with the actual shot schedule reveals that there was no feasible location either at Bikini or Eniwetok Atoll at which the test structure could have been placed to be in the desired 15-psi overpressure zone. Certainly, these facts emphasize that the inclusion of a structures program in an EPG developmental! test series must be considered in the light of yield uncertainties, possible changes in detonation sites, and the restrictions imposed by small land areas. In addition, possible water-~wave damage and the radiation hazard imposed upon the existing land masses by prior detonations in a series as well as the shot in which participation is desired, must be carefully considered in planning. The documentation made by Project 3.5 (see Appendix) was not planned, but rather ar opportunity initiated because Shot 1 gave a higher yield than originaiy predicted. The objective of this project was to determine the effects of air blast froma high-yield device on miscellaneous structures. The unexpected high yield of Shot 1 (approximately 15 Mt instead of 5 Mt) caused damage to certain structures at ranges where no damage had been expected. It was considered highly desirable to obtain all the data possible about this unexpected blast damage, since such knowledge could assist in establishing design criteria for blast protection. That part of Project 3.5 which documented damage to a campand facilities on Tare (Figure 4.2) and Peter Islands, some 14 to 16 miles from Shot 1, presented a picture of conditions to be expected in the fringe zone between no damage and light damage for met- ropolitan targets. Analytical prediction of such damage on the basis of overpressures and positive-phase duration would be difficult if not impossible. Therefore, documenta- tion of such damage was probably of just as great valuc as data obtained from a project specifically designed to obtain such damage data. At the location of the camp installations on Tare and Peter Islands, the estimated peak overpressure was about 1.4 pai, with a positive-phase duration of about 13.4 seconds. Damageto light wood-frame structures varied from light to severe damage. For a given design, the larger structures received greater damage than the smaller structures. Light knee bracing or truss work was effective in preventing collapse of rafters and walls of smail buildings. The atructures oriented parallel to the direction of the blast suffered less damage than those oriented normal to the direction of burst. Generally, the sides of the buildings facing toward ground zero were caved in, usually by bending fractures of the studs. Also, the roof rafters on the burst side were usually broken. The damage to the side and roof away from the burst direction varied widely: some were compietely blown out, others partially damaged, and some received no visible damage. The build-