This paper describes both the construction and creep test of a 20-m span ice dome carried out in Asahikawa during the winter of 1985.
The test dome was constructed at a riverbed as follows. 1) Inflating a membrane bag covered with ropes anchored to the snow-ice foundation circular ring. 2) Covering the membrane with thin snow-ice sherbet by blowing the milled snow with a rotary snowplow and spraying water pumped up from the Ishikari river. 3) Allowing the snow-ice sherbet to be solidified by cooling by cold outside air. 4) Repeating 2) and 3) up to the desired shell thickness. 5) Removing the bag and ropes for reuse.
Because too much milled snow was blown onto the membrane during one blowing operation, geometrical and material imperfections of the constructed dome were observed in some places. Blowing snow in thin layers is not so difficult, and the technique could be improved gradually with experience.
After construction, a creep test was carried out under dead and snow load, and its structuralbehaviour up to the collapse was examined. In spite of the above-mentioned imperfections, it was very ductile and took about one month to collapse.
On the basis of the results of this test, if blowing by a snowplow is done carefully, it is concluded that the proposed construction technique satisfies the requirement for rapid, easy and economical construction of a large ice shell, and the production of 20-m span ice domes could be practicable in principle.

View full abstract

When the hydraulic conveying technique is applied to remove snow from an urban area, a tree configuration pipeline system is considered to be effective. In this work, energy loss of granulated-snow/water mixture flow combining at a T-junction of pipes with circular cross section, which is the simplest geometry of connecting elements used in a pipe line, was investigated experimentally.
A branch pipe met a straight main pipe at an angle 90°, both diameters being 77 mm. Granulated-snow/water mixture flows with equal snow fractions were combined at the junction. The flow rates in the main pipe upstream the junction and in the branch pipe, Q₁ and Q₂, were measured by electro-magnetic flow meters. The pressure distribution along the pipes and the volume snow fraction in the discharge were measured for variable flow rate Q₂/Q₃ under the condition of constant Q₃, the flow rate in the main pipe downstream the junction. Beads of ABS resin were used as the solid sample for comparison.
Power losses due to the confluence were obtained by subtracting frictional losses in straight pipes from the total losses in the junction. The confluence loss coefficient ζc was defined by assuming the mixture to be a homogeneous fluid. While ζc for the beads/water mixture increased with an increasing solid fraction, the relation between ζc and Q₂/Q₃ for the snow/water mixture was nearly equal to that for water irrespective of the snow fraction. In conclusion, the power loss due to the confluence for the graduated-snow/water mixture is well estimated from ζc data for water.

View full abstract