摘要: |
Approximately 450 scour critical bridges are maintained by Tennessee Department of Transportation (TDOT)
across the state. Many of these bridges are located in river or stream beds and banks where cohesive soils are
prevalent, particularly in western Tennessee. In the Hydraulic Engineering Circular No. 18 by the Federal
Highway Administration, Evaluating Scour at Bridges (FWHA-HIF-12-003), existing HEC-18 equations for
predicting scour depth at bridge piers and contractions in river beds for with non-cohesive, e.g., sand and gravel,
tend to perform well. However, when cohesive sediments consisting of consolidated silts and clays commonly
present on river beds, the HEC-18 equations tend to over-predict scour depth, although under-prediction also
occurs as reported in published studies. The uncertainty of scour depth prediction can led to over design of bridge
piers increasing construction costs, or under designed piers which may lead to bridge failure or future costly
repairs. TDOT engineers identified that these equations need improvement and initiated the request for study in
the state to better understand the variables that govern river bed scour in cohesive sediments near bridges. The
objective of this research project is to enhance our understanding of river/stream bed scour and bank erosion
behavior with cohesive sediments near bridges, and its relationships with τc and kd used in HEC-18 equations,
including the time-dependent scour behavior and associated cohesive physical-geotechnical properties, and
measurement of these erodibility parameters. In addition, this research focused on characterizing differences in
physical-geotechnical properties of riverine cohesive sediments across the different physiographic provinces of
Tennessee. The initial research investigated how field data collection and computational procedures for the insitu
mini-jet tester influenced kd and τc estimates. In order to meet the objectives, the following studies were
completed: 1) statistical multivariate predictive equations were developed for τc (and kd) across the state’s
different physiographic provinces as a function of significant physical-geotechnical properties; 2) measured
bridge scour data were correlated with cumulative effective stream power data from long-term river continuous
flow records as a product of hydrological model simulations in order to test whether ‘cumulative effective stream
power’ could be used as a predictive variable for scour depth in cohesive bed sediments; and 3) in a large open
channel experimental flume around a physical model and test box consisting of a cylinder in natural cohesive
soils, evolution of scour depths were measured for several multiple flow sequences to illustrate the scour timedependency
of cohesive sediments. This study identified that several factors were related to the erodibility
parameters estimation for cohesive sediments: i) variability related to the device operation, ii) variability related
to sediment source, iii) device dependent variability, and iv) soil heterogeneity among study sites. HEC-18
equations could be improved through more accurately measured τc and kd values using the mini-jet device and
multiple-pressure setting procedures compared with a single pressure setting approach. A key finding of this
study was that τc and kd were related to different physical-geochemical sediment properties, and unique to
physiographic province representing different surficial geological formations. The physical-geochemical
parameters found to be statistically relevant were moisture content (WC) and % finer particles passing a #200
sieve (Pass200) dominantly, but also cohesion (CC), dispersion ration (DR), liquid limit (LL), sodium adsorption
ratio (SAR), and organic content (OC). Many different predictive models for the erodibility parameters have
been developed in the United States, but were limited to one or only a few physical-geochemical parameters.
Physical-geochemical parameters govern the time-dependent scour behavior in riverine cohesive sediment/soils.
The time-dependent scour behavior correlated with cumulative effective stream power, a surrogate for shear
stress duration over τc, in addition to flow history. The open channel flume experiments also documented the
importance of flow history on scour, in addition to the sediment bulk density (BD). In order to develop a more
accurate predictive equation than reported in HEC-18, further research is required. Applying the finding from
this study, recommendations proposed for future research include: 1) conduct more in-situ field tests with the
mini-jet device to verify the τc values among similar physiographic provinces, 2) implement a long-term field
study at newly constructed bridge sites, and continuously and consistently monitor both flow and scour depths,
and 3) conduct additional flume experiments incorporating more varied flow sequences and sediment types with
measured physical-geochemical properties. |