Home Archived January 13, 2017
(i)

Potential Drift Accumulation at Bridges

Width of Accumulations and Blocked Spans

The length of the longest pieces of drift determines the maximum width of the common types of drift accumulation. Long logs hold together large accumulations and support them against lateral forces. The width of the channel influences the length of drift delivered to the bridge, and thereby helps to determine accumulation potential and characteristics.

Width of observed single-pier accumulations

An accumulation resting against a single pier typically contains one or more logs extending the full width of the accumulation perpendicular to the approaching flow. These key structural logs convey lateral hydraulic forces to the pier and prevent the accumulation from breaking apart and passing downstream on both sides of the pier. Sometimes these logs are visible (figure 15). Single-pier accumulations without visible full-width logs generally contain smaller logs arranged in a pattern similar to the smaller logs in accumulations with visible, full-width logs. This common pattern suggests that full-width logs are present, but either submerged or concealed beneath smaller drift.

137K GIF version of this photograph

Figure 15. Large log supporting a single-pier drift accumulation.

The upstream ends of some large accumulations that form across spans and on island heads have the same structural pattern as single-pier accumulations (figure 16). Island-head accumulations typically terminate in a raft one log thick at its upstream edge, with individual logs extending its full width. The raft typically has a curved upstream edge when viewed from above, and the center of its downstream side rests across thicker parts of the accumulation that support the raft against lateral hydraulic forces.

128K GIF version of this photograph

Figure 16. Drift accumulation at the upstream end of an island.

The submerged drift accumulation surveyed with scanning sonar at the FM 2004 bridge over the Brazos River near Lake Jackson, Texas, was about 23 m (about 75 feet) wide, or about as wide as the length of the largest logs in the Brazos River. It had an irregular, convex upstream face. The floating raft of drift that formed over this accumulation during a flood had about the same width and length, and a curved upstream edge (figure 17).

74K GIF version of this photograph

Figure 17. Raft of floating drift at the FM 2004 bridge over the Brazos River near Lake Jackson, Texas.

Data from scour-potential studies are consistent with the structural pattern observed in single-pier accumulations. Plots of drift width versus upstream channel width (for drift accumulations presumed not to span the gap between two piers) show that few accumulations on a single pier are more than 15 m (50 ft) wide (figures 18 and 19). In narrow channels, such accumulations tend to be narrower than the channel.

31K GIF version of this graph

Figure 18. Width of inferred single-pier drift accumulations at scour-potential sites in Indiana.

33K GIF version of this graph

Figure 19. Width of inferred single-pier drift accumulations at scour-potential sites in Tennessee.

Single-pier and island-head accumulations wider than the length of a single log were observed at three sites along the White River of Indiana. The White River has a wide-bend point-bar channel with a high rate of lateral migration (Brice and others, 1978a). All of these unusually wide accumulations were at sites with potential for bar aggradation and island development. At Paragon, Indiana, island growth clearly contributed to the large size of the accumulation. The island and the upstream part of the accumulation formed part of a scroll bar, one of the longitudinal sand bars that characteristically form parallel to the inside bank of channel bends.

Width of observed span blockages

Drift accumulations between piers typically occur where the length of drifting logs exceeds the effective width of openings between piers, and the logs can come to rest against two piers (figure 20). The effective width between piers is the distance between lines parallel to the approaching flow that pass through the nose of each pier (figure 21). Any fixed object that divides the flow can provide a place for one end of a log to rest. Such objects include island heads, trees and utility poles on the flood plain, and isolated piers and pilings remaining from previous bridges at the same site.

118K GIF version of this photograph

Figure 20. Logs lodged from pier to pier and from pier to bank.

8K GIF version of this diagram/I>

Figure 21. Definition sketch of the effective width of horizontal gaps.

Where the structure of logs bridging the gap between piers could be observed, individual logs typically bridged the gap between piers. Some of these key logs rested directly on the piers; others rested against drift accumulated on the pier noses. In all confirmed cases of drift extending from pier to pier without interaction with additional stationary objects, the effective span was short enough to be bridged by a single log (figure 22).

28K GIF version of this graph

Figure 22. Effective width of drift-blocked spans outside the Pacific Northwest.

At most sites with narrow channels, the main span must bridge the entire channel to achieve low potential for span blockage. Many box culverts and timber trestles have an effective span length much shorter than the channel width upstream or the typical length of the longest drift in the stream. At such sites, span blockage is the dominant mode for large drift accumulations, and considerations of pier placement and single-pier accumulations are of secondary importance.

Study of bridges along the Harpeth and Wolf Rivers in Tennessee confirmed the importance of span length in determining which bridges underwent span blockage. On the Harpeth River, bridges with spans shorter than the length of large logs (15 to 18 m, or about 50 to 60 ft) typically had one or more spans blocked at least once, whereas longer spans were not blocked. Most of the oldest and newest bridges along the Harpeth River have long spans. Current design practice seems to favor spans in the range of 20 to 30 m (about 70 to 100 ft) and single-column piers with rounded noses. Such bridges did not undergo span blockage during the study period. Most bridges along the lower Wolf River have spans longer than 18 m (60 ft). No blocked spans were observed on these bridges. During a period of high water on the Wolf River, abundant drift formed several single-pier accumulations, but did not bridge any spans.

Wide spans can be bridged by drift accumulations where the size of the drift exceeds the effective span width. An extreme example is the accumulation that occurred in June 1994, in the Brazos River at U.S. Highway 59, near Richmond, Texas (David Dunn, USGS, oral commun., 1994; James Fisher, USGS, oral commun., 1994; David Mueller, USGS, oral commun., 1995). At that location, flow approached two parallel bridges at an angle of about 45 degrees to the highway centerline. This situation produced an effective span length of about 24 to 27 m (70 to 80 ft), although the span length along the centerline of the bridges is 56 m (185 ft). A large accumulation grew directly upstream from a pier on the downstream bridge into the middle of a span of the upstream bridge. Additional drift accumulated between the head of this accumulation and the adjacent piers of the upstream bridge. In Steamboat Slough at Interstate 5 near Everett, Washington, 37-m (120-ft) spans were blocked (Brice and others, 1978b; Martin Fisher, Washington DOT, written commun., 1994). Sawlogs that were cabled together into rafts wider than the spans broke free from a storage area and became lodged against the spans. On the White River near Paragon, Indiana, the blockage of a 27-m (90-ft) span was related to the growth of an island upstream from the span (figure 23).

116K GIF version of this photograph

Figure 23. Blockage of a 27-meter span over the White River at Paragon, Indiana, by drift accumulation and island formation, September 25, 1992.

On the Harpeth River at Tennessee State Route 250 near Frog Pond, Tennessee, a span was bridged by drift extending from a single-pier accumulation to the side of a span blockage. This drift was washed away in a subsequent period of high water, leaving the single-pier accumulation and span blockage intact. This was the only observed instance in which the narrow gap between adjacent drift accumulations trapped drift in a manner analogous to a gap between piers. Such a structural pattern would seem to require the development of significant shear strength between the logs in the mass of drift. A theoretical basis exists for the development of shear strength capable of supporting lateral forces over a greater width perpendicular to flow than the length of single logs (Kennedy, 1962). However, wide span blockages that involve the development of shear strength in an accumulation of drift seem to be rare.

Where several adjacent spans are short enough to be bridged by drift, the accumulation can extend across all spans to which drift is delivered, sometimes blocking nearly all of the channel and producing significant backwater. On the Harpeth River at State Route 46, for example, two of the 14-m (46-ft) spans and most of the third span were blocked, causing about 1 m (3 ft) of backwater and supercritical flow through the remaining chute. Drift can also accumulate on immersed superstructures, and can extend across all areas to which drift is delivered.

Accumulations between a pier and an abutment or bank are similar in structure to accumulations between piers. Large, sturdy, fixed objects on the bank, such as boulders and trees, can support one end of a log. Logs can become lodged perpendicular to flow against a pier and a tree or boulder just as they would across two piers. Woody vegetation and riprap on river banks at bridges may make such accumulations possible at most sites. Based on this possibility, the effective width between a pier and a bank or abutment is the distance perpendicular to the approaching flow between a line projected upstream from the pier nose and the nearest point on the bank or abutment (figure 21).

More typically, such accumulations extend diagonally upstream from the pier to the bank. Banks without objects projecting into the flow apparently trap one end of a log only when it is forced into the bank. At the other end of the log, a corresponding force away from the bank is exerted on the pier. Logs trapped diagonally must be somewhat longer than the effective span. Measures that keep the bank smooth (such as clearing the right-of-way of woody vegetation) may reduce the potential for drift to accumulate between the banks and nearby piers (figure 20).


previous || table of contents || next


U.S. Department of the Interior, U.S. Geological Survey
Maintainer:webmaster@tn.water.usgs.gov
Last update: Tuesday, 08-Jan-2013 14:46:14 EST
URL: http://tn.water.usgs.gov /publications/FHWA-RD-97-028/accumwid.htm
USGS Privacy Policy || Disclaimer