ADFM Placement in Gated Culverts for Unbiased Discharge Measurements

Abstract

Acoustic Doppler Flow Meters (ADFMs) are suitable to measure the flow discharge in culverts for the purpose of rating calibration and validation. However, best-measurement guidelines for ADFM placement to measure the flow through gated culverts are not available. Gates in culverts induce a complex flow field consisting of a jet below a hydraulic jump or a submerged eddy depending on whether the tailwater level results in free or full-pipe flow. The architecture and operational principles of ADFMs are suitable for measurements in well-developed flows; however, the distance from the gate where the flow becomes well developed depends upon the ratio of the gate-opening-area to the full-barrel-area (A G /A 0 ) and the gate Reynolds number (Re G ). Presently, neither a characterization of the distance from the gate where well-developed flow occurs, nor an assessment of biases in ADFM measurements of full-barrel, gated-culvert flows due to placement of the ADFM where the flow is only partly developed, are available in the literature. Tropea et al. (2007) reported that the reattachment length of a flow at a backward facing step occurs at a distance ~ 8 times the height of the step. Here we show that discharges measured in two gated culverts with an ADFM at distances (L) equal or larger than 8(D-G) from the gate (D=diameter or box-culvert height; G=gate opening) using as reference discharge measurements by Acoustic Doppler Current Profiler (ADCP) from a moving boat in the discharge canal, have biases of less than 6.4% (estimated as relative errors |Q ADFM -Q ADCP |/Q ADCP ). Moreover, our ADFM discharge measurements show biases between 0.4 and 6.4%, depending on the gate opening and culvert geometry. Additional analysis is underway to assess whether Re G strongly varies with the reattachment distance of the jet-induced eddy pocket and the non-dimensional distance from the gate, where flow becomes well developed L/(D-G).