OCR Text |
Show It can be seen that extremely heavy loads are required to achieve considerable compaction. For example, to compress snow of density 0.40, average density for a newly formed layer of windblown snow, a load is needed equal to the weight of an 8 m layer of snow of density 0.50. The layered structure of a snow sample is brought out clearly by compression. The column of compressed snow taken from a test cylinder nearly always shows the stratification not so evident before loading. We compressed the surface layers of loose, natural snow. Compression was observed in the upper layers only; the lower layers did not respond to weight pressure. This shows that when the snowcover is of considerable depth, only the upper layers compress under load. Recrystallized snow is responsive to lateral load ( pressure) and is more easily destroyed by this means than by a force perpendicular to the layer. The binding capacity of freshly fallen snow is greatly increased by compression; but the further recrystallization has proceeded, the less the increase in density. Since loose snow, coarse granular snow, and granular snow are not sufficiently bound to evaluate hardness ( strength), resistance to penetration by cone was determined. The cone was positioned with its apex touching the snow surface and its axis perpendicular to it; the tread attached to the cone base was left free ( to move vertically) and the cone was allowed to penetrate the snow under its own weight. Depth of penetration was determined both by a scale on the cone surface and by the imprint ( perforation) diameter. In view of the fact that snow strength ( hardness) varies between wide limits, an assortment of cones of varying weight ( 10, 60, and 300 g) and conical angle ( 30° and 60°), or a " Swedish cone," were used. The strength of the snow ( resistance to penetration) was calculated as the ratio of cone weight to the volume of snow dislodged. Expecially interesting results were obtained with loose snow. In some cases of snow in which recrystallization had already set in, penetration strength was two times less than for fresh snow. No other physicomechanical property of snow is so variable. Comparable results can be obtained only with the same cone. The hardness ( strength) of bound snow, which becomes less of a factor as conical angle decreases, varies hardly at all with cone weight differences. The reverse is true for loose snow. For example, in snow of density 0.12, tested immediately following its deposition by a windstorm, the ratio of penetration strength, measured with a 60° cone, to cone weights of 60 and 300 g was 1: 12; but on transformation to loose granular snow, the ratio was 1: 0.07. The different results obtained with different cones characterize the type of snow and are closely related to other physicomechanical properties such as intermittent shearing, crushing and compression strength. Sharpness of the cone apex is of special significance in types of snow where crushing strength exceeds shearing strength; cone weight is the chief factor where crushing strength is less than shearing strength. The loss of data makes illustration with examples impossible. We are convinced, however, that the sensitivity and simplicity of the above test make it an invaluable aid in field work, especially so because, with a sufficient number of results, it can fully replace more complicated methods of determining the mechanical properties of snow. It should be pointed out that the data obtained by penetration with cones of varying angle and weight not only characterize the physicomechanical properties of snow, but define its structure and density as - 29- |