The need to design bridges to withstand flood and debris loads has long been recognised in Australia with the current and previous design codes providing guidance. Limit state design philosophy was adopted in AUSTROADS (1992), the 1992 Australian bridge design code. The ultimate limit state is defined in the code as 'the capability of a bridge to withstand, without collapse, the design flood associated with a 2000 year return interval'. The association of the ultimate limit state with a 2000 year return interval design flood means that the majority of bridges over waterways will be designed for overtopping and that the design flood loads are more realistic.
The recommendations in AUSTROADS (1992) for design debris loads and flood loads on superstructures are derived from limited data. The recommendations for flood loads on piers are based on extensive research in the 1960's and earlier, and are considered reliable. Therefore, further research on debris loads and flood loads on superstructures was required for use by designers working according to limit state philosophy.
In recognition of the requirement for an improved knowledge of debris and flood loadings, the hydrodynamic and debris loadings on bridge superstructures and piers have been measured in a comprehensive laboratory program for ranges of flood conditions and geometric arrangements likely to encountered in practice. The effect of the Froude number (F), degree of submergence (SR), and proximity of the superstructure to the bed (Pr) on the forces and moments were investigated in a parametric study. Also studied were the effects of turbulence intensity, superelevation, and skew.
The forces and moments were measured on scale models of six different superstructures and three pier types, and the debris loadings on five superstructures and three pier types. The loads were measured using a custom designed dynamometer system. On one of the superstructure models, the loadings were also obtained by measuring the pressure distribution around the centreline of the model and then integrating the distribution.
Flow visualisation and the pressure distributions were used to investigate the bluff body fluid mechanics phenomena associated with flow around bridges. Boundary layer separation and reattachment, free surface effects, and wake blockage effects were studied in detail, and related to the trends in the measured loads. A number of unusual flow patterns were observed and documented.
The results are presented as drag, lift and moment coefficients. They constitute the first comprehensive set of data of this kind, and they provide the basis for more accurate estimates of design forces and moments associated with flood and debris loadings for submerged and semi-submerged bridge superstructures and piers. The coefficients were found to be dependent on the parameters F, Pr, and SR and that under some conditions a inter-dependence existed between the coefficients.
The maximum drag coefficient for the superstructure models was typically in the range 1.9 to 2.2 occurred when the water level was about the depth of the model above the top of the model. The lift coefficient was typically negative and in the range o 0 to -8.0 with the most negative values occurring just above overtopping of the model. The maximum coefficient of moment of about 4.0 occurred when the water level was about the depth of the model above the top of the model. These generalisations are for a typical case with a Pr of 3.5, i.e., the distance from the floor of the flume to the underside of the model was 3.5 times the depth of the model.
The superstructure debris models were found to be strongly dependent on the above parameters. The drag coefficient had a range of about 0.5 to 3.0, the lift coefficient a range of -6.0 to +3.5, and the moment coefficient a range of about 0.0 to 2.0.
The pier debris models were dependent on the depth of the approach flow and F. The drag coefficient varied from about 0.6 up to 5.0 across the range of test conditions.
From the data sets, a series of design charts and tables are developed and a new methodology for the calculation of overturning moments that accounts from the correct line of action of the drag and lift forces is presented.