that the fallout particles originated as calcium carbonate (coral) which was converted to calcium oxide by the heat of the detonation. This rapidly changed to calcium hydroxide with the formation of a very thin layer of calcium carbonate on the outer surfaces. Particles were influenced by a sea water environment which caused the formation of an insoluble shell of magnesium hydroxide, surrounded by calcium carbonate. On the interior of the shell, weliedeveloped hydrated calcium sulfate crystals (gypsum) were formed. This leaching effect, by causing partial solution and reprecipitation of the soluble calcium compounds, was felt to account. for the adherence of the particles, As indicated above, experimental evidence supports the view that ion exchange is primarily responsible for the adherence of the contamination at CASTLE, However, the role of ion =xchange in the wet contamination-decontamination behavior of materials will require further exploration before its implications are fully understood, Appreciable differences in initial contamination levels existed among the various panel surfaces with no evident correlation to surface properties. Vertical surfaces facing upwind became equally or more highly contaminated than pitched or horizontal surfaces, probably due to the combined action of wind currents and tenacious contaminant. A possible explanation for this phenomenon is that deposition of wet contaminant was irfluenced by wind currents which tended to - impact the contaminant onto surfaces normal to the wind, Although this effect was most pronounced on shipboard contaminated panels following Shots 2 and wnere ship speed into the wind and ship structural georetries may increase the impact, the contamination of panels on the stationary barge following Shot 6 exhibited similar relationships. The effect of slope panels used during Shot 6 were all of the same material and had similar contamination characteristics, The construction material panels were placed at the same mounting slopes as practiced in building construction. No direct comparison of identical surfaces at different slopes can be made with these panels, but it is important to ccmpare the contamination of materials at the slopes as they are encountered in actual buildings. The experinental decontamination work was done in a sequence of operations and only qualitative comparisons between decontamination methods can be made. If a less efficient method were applied after | a more efficient one, it is believed that only negligible removal would result. The selection of the sequence of the different decontamination metheds was predicated on laboratory tests as suggested in reference’. The basic physical parameters which appear to affect resistance to wet or slurry forms of contamination and/or ease of decontamination are impermeability to moisture, non-absorptivity, and hardness, Results illustrating the effects of these parameters are presented as part of Table 4.6. Smoothness of surface did not appear to be as important for wet and slurry forms of contaminant as it was for the dry form found at JANGLE. Scrubbing followed by flushing seems to have been the most effective and economical of the reclamation techniques employed. Furthermore, the addition of a detergent increased the effectiveness 39 CONFIDENTIAL— RESTRICTED DATA

Select target paragraph3