TUNNEL TRANSFORMATION OF ANTIDUNES

Two series of experiments were run to ascertain the nature of the transformation of antidunes when suppressed by a tunnel roof. One series was performed in a recirculating flume 30cm in width and 6 meters in length using well sorted sand, 0.35 mm in mean diameter. The second series was conducted in a flume 76cm wide and 18 meters long using very well to very poorly sorted gravel and sandy gravel. In the first series, standing waves, antidunes, and breaking antidunes were generated in the open channel section of the flume; in the second series, standing waves and downstream-migrating antidunes were generated. In both series, these bedforms were replaced in the tunnel section by a plane bed, often with a rheological front. In the smaller flume, antidunes contained backset and foreset (lee-side) laminae, poorly sorted and less than about 10° in maximum dip. In the larger flume, although backsets formed, they were temporary only, and were soon eroded by the downstream migration of the antidunes. Sand was deposited on the downstream sides of bed waves, but although clearly visible in plan, it did not show up clearly as lamination in section: well sorted gravel was just as abundant as poorly sorted. The open channel data from the gravel experiments showed that in-phase waves form at about the same Froude numbers for both coarse sand and granule-pebble gravel.