U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590

Skip to content
Facebook iconYouTube iconTwitter iconFlickr iconLinkedInInstagram

Federal Highway Administration Research and Technology
Coordinating, Developing, and Delivering Highway Transportation Innovations

This report is an archived publication and may contain dated technical, contact, and link information
Publication Number: FHWA-HRT-07-026
Date: February 2007

Bottomless Culvert Scour Study: Phase II Laboratory Report

Chapter 5: Conclusions

Phase II improved on the Phase I study results by providing additional research data, including the following.

  • Additional riprap tests improved the riprap analysis. More data were developed, including data from experiments with wingwalls and under submerged conditions.

  • Fixed-bed experiments accurately measured initial flow distributions and flow redistribution in the culvert. One of the problems encountered with the movable-bed experiments was that conditions change as soon as the experiments begin. The information from the fixed-bed experiments was used to validate three approximations of the representative velocity.

  • Different outlet wingwall shapes were used to analyze outlet scour. Results from the observed outlet scour experiments are presented in spatial maps in appendix A.

  • Many different theoretical approaches were used to help the practitioner calculate the maximum scour under unsubmerged flow conditions. However, the results for submerged bottomless culverts are only preliminary.

Equations are presented to estimate the maximum expected scour depths at the upstream corners of bottomless culverts under clear-water conditions. New equations are also presented to estimate the riprap sizes needed to protect bottomless culvert footings from scour.

All experiments outlined in the test matrix in table 1 were completed in Phase II, but there were some limitations in the experimental setup. The experimental results were based on laboratory flume experiments with a flat approach cross section with uniform flow conveyance, which is not typical of field conditions. The experiments were also conducted under clear-water approach flow conditions with no sediment being transported into the culvert. The authors attempted to present the results in terms of overbank flow rather than geometric variables; presenting the results is this fashion allows accounting for the reduced conveyance that is typical of overbank flow for natural streams. These results have not been tested for field conditions; however, they are offered as initial guidance for field applications. An anticipated next step is that MDSHA will adopt the results as preliminary design guidelines and test them for field sites using engineering judgment to decide if the applications are reasonable.

The abutment scour concept of using the flow distribution at the culvert entrance to compute the primary scour depth component and adjusting that with an empirical factor based on laboratory data appears to be valid for bottomless culverts. Three different equations for the initial representative velocity and two different equations for the critical incipient motion velocity were tested to compute the flow distribution scour. The Froude numbers in the experiments did not cover the full range that is expected in the field, and the negative slopes presented in table 2 are probably not realistic. In fact, other experiments performed by GKY and Associates, Inc., show that the correlation of ks with the Froude number is positive.(3) For this reason, we recommend changing the Froude number multiplier to zero for equations in table 2 with negative slopes. This change is equivalent to changing the ks equations with a Froude number in them so that ks equals only the intercept. Nevertheless, the laboratory data suggest that calculations of ks as a function of either VRA, VCL, and F1 or VRM, VCN, and Qblocked are the two best functions for calculating scour in an unsubmerged bottomless culvert. The kp results, however, are still too preliminary to suggest the best predictors of scour in submerged bottomless culverts.

The culvert entrance flow conditions were a significant influence on the scour. The flow through various inlet and outlet configurations was investigated as both submerged (pressure flow) and unsubmerged to determine the overall effects of the flow conditions on scour hole formation. The results show that submerged flow conditions induce greater inlet scour depths, while unsubmerged flow conditions induce greater outlet scour depths. The results also show that 45-degree inlet wingwalls are effective at reducing inlet scour, whereas 8-degree inlet wingwalls are not effective.

The outlet scour experimental results showed the effects of using different wingwall configurations at the outlet. Changing the angle of the wingwalls reduces the turbulent shear stress, and thus reduces the scour depth created. The outlet experiments clearly demonstrate that outlet scour can be substantially reduced by using outlet wingwalls with a streamlined shape. The elongated streamlined bevel wingwall was best at reducing scour. Experimental results indicate that turbulence is reduced and “vortex shedding” caused by abrupt changes in pressure is almost eliminated by using this shape. In other words, the streamlined wall eliminates flow separation and decreases turbulence.(10) Hence, with the streamlined bevel, vortices do not propagate downstream and the resulting turbulence is more evenly distributed—not concentrated in a single location. Conversely, the abrupt change in pressure that results from a square exit shape (as found in culverts without wingwalls at the outlet) induces vortex shedding and increased scour depths.

Eight-degree outlet wingwalls were also tested because streamlined wingwalls may not be practical in the field. These results revealed reduced turbulence and scour depth at the outlet. This is an encouraging finding because wingwalls with an 8-degree flare are easy to construct or can be ordered prefabricated, which may make this design more cost-effective than the streamlined design.

Equation 27 is useful for sizing riprap to reduce scour. Chang’s pile dissipators dissipated some of the energy at the outlet and thus reduced the scour depth. The MDSHA Standard Plan for countermeasures did not significantly reduce the scour depth, but it is considered a good practice because the riprap that was employed in this plan moved and fell into the scour holes, after which the riprap stabilized. However, since these results are still preliminary, this report does not make any recommendations about sizing or placing riprap for this design. Cross vanes are not recommended at the inlet because the results show that they contribute to rather than hinder scour due to a spiral current effect.

Additional research could extend and improve upon the Phase I and Phase II study results. This research could include:

  • Conceptual sediment balance relationships to extend the analysis to live-bed conditions. The authors propose that Laursen’s “sediment-in equals sediment-out” logic (that the amount of sediment entering a stream segment must equal the amount of sediment exiting) should apply with reasonable assumptions about flow distributions. An inherent assumption is that the empirical adjustment factors from the clear-water experiments can be applied to live-bed conditions. Live-bed flume experiments with sediment transport in the main channel and clear water (no sediment) in overbank flow are needed to test these assumptions.

  • Derivation of a safety factor to envelop the experimental riprap data. Engineers often find that they use the same class of riprap for a wide range of requirements. A safety factor provides a level of confidence in applying engineering judgment in these situations.

Previous | Contents | Next

Federal Highway Administration | 1200 New Jersey Avenue, SE | Washington, DC 20590 | 202-366-4000
Turner-Fairbank Highway Research Center | 6300 Georgetown Pike | McLean, VA | 22101