Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
66 CHAPTER 7. RECOMMENDATIONS FOR DESIGN ESTIMATION OF ABUTMENT AND CONTRACTION SCOUR DEPTH From the foregoing discussion, it is apparent that none of the abutment scour or contraction scour formulas listed in Tables A-1 and B-1 satisfy fully the criteria specified in Section 5.4. Furthermore, it is obvious that no simple abutment scour formula can be recommended that will apply to all of the complexities found in the field. Finally, the idealized contraction scour formulas currently in use are problematic because of their basis in assumptions that are not satisfied for bridge contractions which are inherently short contractions, and for which the flow is nonuniform. Given this state of our current understanding of abutment/contraction scour, or lack thereof, it is very difficult to develop design-specific recommendations at this time. Some important strides forward have been made in the past decade, but there much remains to be learned before we have arrived at the more settled and defined state of knowledge that currently exists with respect to pier scour. In spite of these caveats, some general recommendations can be made and then discussion of a possible path forward to create a unifying model of abutment scour is presented in the following. 7.1 GENERAL RECOMMENDATIONS ⢠The definition of abutment scour and the formulas by which it is estimated pertain to a combination of local scour due to large-scale turbulence generated by flow separation and constriction scour due to flow acceleration caused by the flow contraction itself. Under these circumstances it would appear that contraction scour should not be computed separately from abutment scour estimates. Therefore, it is recommended that a combined abutment/contraction scour formula be developed. ⢠Given that many abutment failures due to scour are the result of collapse of an erodible embankment, it is recommended that geotechnical estimates of stability should accompany hydraulic scour estimates as suggested by the NCHRP 24-20 Final Report (Ettema et al. 2010). The precise approach to formulating such estimates requires further work, however. ⢠It is recommended that abutment toe protection and/or guide banks should be considered for all new installations of abutments. Furthermore, for setback abutments, the setback distance should be large enough to avoid failure of the main channel bank in the event of an embankment failure. This distance depends on the flow distribution in the bridge opening as well as the abutment characteristics. ⢠A small group of abutment scour formulas using the flow distribution in the bridge section or a similar independent variable are best for estimating abutment scour in compound channels. The idealized long contraction scour depth is a useful reference scour depth for these formulas. ⢠It is recommended that a clear distinction be made between abutment scour depth estimates by formulas developed for solid abutments and those for erodible embankments and abutments. These formulas should be applied only to the case for which they were developed. The solid abutment scour formulas will predict the maximum potential scour
67 depth in comparison to erodible abutment/embankment scour formulas that consider the flow relief associated with embankment failure. Unifying these formulas with an adjustment factor for erosion strength of the embankment would be a useful goal. ⢠It is recommended that a renewed effort be undertaken to educate hydraulic engineers with respect to the complexities of abutment scour and the new numerical and physical modeling tools available to resolve difficult cases. 7.2 SPECIFIC RECOMMENDATIONS First, it is recommended that contraction scour be viewed as a reference scour depth calculation as suggested in several recent investigations of abutment scour, and that abutment scour be taken as some multiple of contraction scour rather than additive to it. In this context, further refinement of contraction scour equations may not be the most productive approach; rather, the incorporation of contraction scour into abutment scour formulas may be a more realistic and useful goal. Second, it is recommended that a small subset of abutment scour formulas, each having its desirable attributes, be unified into a single formula in order to develop more realistic and robust procedures for abutment scour prediction. Reducing these formulas to a common form and establishing upper and lower limits of expected abutment scour depending on the limitations of each formula would seem to be a practical path forward. The formulas judged to be most promising in this regard, and with respect to the established criteria, are the following: 1. Ettema et al.- It is the only formula that considers an erodible embankment; it has the desirable attributes of reflecting the physics of the abutment scour process both in terms of flow constriction and turbulent structures of the flow separation process albeit in a rudimentary form; and it includes experiments with compound channel geometry although a wider array of experiments is needed. It could in theory be applied to scour Classes I, II, and III channels. 2. Sturm - It includes a method of accounting for flow re-distribution due to compound channel geometry with similar independent variables compared to the formula of Ettema et al., and it represents the other extreme of a solid-wall foundation as opposed to an erodible embankment. It is most applicable to Class II channels, Scour Condition B. 3. Melville - It is most applicable to short, solid-wall abutments and depends on abutment length rather than the flow distribution in the contracted section, but it can be viewed as comparable to the first two formulas if some width of contracted flow, which is related to the width of the scour hole, is established in the contracted section through which all of the approach floodplain flow passes. It also is at the limit of a solid-wall foundation rather than an erodible one. It is most applicable to Class I and Class III channels. 4. ABSCOUR - It contains the desirable attribute of including the direct effect of flow re- distribution on the floodplain through the Laursen contraction scour formula in terms of q2/Vc, although the adjustment factors for spiral flow and velocity should be re-evaluated in the limit of severe contractions as discussed previously. In addition, the correction
68 factor related to floodplain width seems to be an ad hoc adjustment for the specialized data of Benedict (2003) that applies only to South Carolina. These adjustment factors should all be re-considered in the process of developing a unified formula that is more generally applicable. ABSCOUR also has the useful feature of a user-friendly computer application that minimizes to some extent the mistakes that can be made by hydraulic engineers without an extensive background in the area of bridge scour. Although the Briaud (2009) formula does not satisfy the criterion for best parameter framework, it is one of the only databases for cohesive sediments, and the data could be useful in expanding the range of applicability of the final unified formula. It is suggested that a unifying formula or family of formulas can be formed from the list above with a common set of independent variables, preferably of the form of the Ettema et al. formula. The Ettema et al. and Maryland formulas directly use idealized contraction scour as the reference variable for nondimensionalizing the flow depth after scour at the bridge section while the others use the approach flow depth. Each of these approaches has desirable attributes, and each one should be tested in the effort to develop a unifying formula. Using the contraction scour depth as a reference length scale is very attractive if further work can elucidate the limiting case at the left-hand boundary of Figure 5-8 as discussed previously. It is beyond the scope of the present project to develop a unifying formula, although Figure 5-7 may provide a useful starting point. Such a formula could provide an interim update to the HEC-18 formulas, which currently must be used with informed caution, until the proposed research needs in the next section can be fully satisfied. Third, it is recommended that a flow chart be developed to be used as a guide to evaluate abutment scour in an informed manner and to assist the judgment of design engineers. Where a unified abutment scour formula predicts very large abutment scour depths or possible embankment failure, appropriate scour countermeasures should be indicated. Geotechnical evaluation of scour could become a routine part of the analysis. For more complex problems, hybrid numerical and laboratory hydraulic models should become a readily accessible option. Fourth, it is recommended that in the near term, abutments should have a minimum setback distance from the bank of the main channel with riprap protection of the embankment and a riprap apron until better methods are available for estimating the erodibility of the embankment itself. The minimum setback distance would then be that recommended for the width of riprap aprons (see Lagasse et al. 2009: HEC-23). Other scour countermeasures, especially guidebanks, should be seriously considered for protection of the embankment as well. Fifth, it is recommended that further development of an educational curriculum for hydraulic engineers be undertaken in order to emphasize the proper choice of parameters that go into any scour calculation and in the use of 2D and 3D numerical models to better evaluate the hydraulic parameters. At least in the short term, 2D numerical models should be used on all but the simplest bridge crossings as a matter of course. These issues are discussed further in the next section on research needs. The prediction of critical velocity and the estimates of flow distribution in the contracted section are examples of parameters that are
69 crucial to the success of any abutment scour formula. Furthermore, implementation of a computerized procedure as in ABSCOUR or HEC-RAS with controls on reasonable values of input parameters would be very helpful. Finally, it is recommended that a long-term field program of obtaining high-quality, real- time field data be undertaken. Simultaneous measurement of bed elevations and the flow field are possible with in-situ sensing devices that record the data and transmit them for real-time bridge monitoring on the internet. Sites without a large number of complicating factors could be identified, and full reliable data sets of simultaneous hydraulic conditions and bed elevations could be obtained to better understand field scaling issues and the simultaneous interaction of various scour processes driven by the hydrodynamics of the flow. While embarking upon such a program will be expensive and require patience, the results will move the ultimate solution to the abutment scour problem forward more effectively than less-expensive post-flood surveys.
