Losses in Box culvert transition structures 1

[nick269]

Hello everyone

I'm on a drainage project in Vietnam, which entails converting a 4.3 km section of filthy canal into continuously connected box culvert sections (with a road on top!). The slope is extremely mild - 0.0005 to 0.001 m/m. The lower half of the system is subject to tidal backwater (flow reversals)and modeling with SWMM shows parts are surcharged for an hour or so during the design storm. Along the length of this closed conduit system, the culvert sections (all rectangular)increase in rise, span, and the number of cells, from 2 barrels of 2.5 rise x 2.0 span (total 5x4) to 4 barrels 2.5 x 2.5 (10x2.5). Connecting these variously sized sections are tailor-made transition structures - single cells which expand out to connect to the next larger section. The expansions are straight lines at varying angles, but the most abrupt one widens one meter over a length of only 1.73m.

My problem is this: I am trying to calculate the head loss through these transition structures, but I am not sure of the best approach. The classical energy method - a coefficient times the change in velocity heads - seems too simple for this situation. The change from multiple cells, to one cell, back to multiple cells may induce some turbulence despite the low velocities (0.5 to 1.4m/s) and the exits and entrances are square edged, further increasing losses. I thought of modeling with HEC-Ras, but again I'm not sure of the right approach. Any advice or suggestions would be helpful, as I am all alone here (the Vietnamese engineers don't seem to concerned, but I need to be assured).

Nick

 

[cvg]

an expansion coefficient multiplied times the change in velocity head is an acceptable method to use. (HEC-RAS will use the same method) However, the choice of coefficient to use is more critical. You have a very abrupt transition and the factor should be 0.5 to 0.8. A recommended maximum angle of divergence is 6 degrees per side. This is a standard recommended by the APWA. To meet this criteria, your structure would need to be about 5m long.