Flow in noncircular ducts

[Daamz]

Hi all,

I am currently working on the design of a closed channel that is to convey flow from the Sand Ballasted Flocculation (Actiflo) basins to GAC filters in a 110 MGD water treatment plant. The original design called for two (2) 72” pipes. Now we are looking at concrete channels instead of pipes to lower costs.

So far, I have been using Darcy-Weisbach for my headloss calculations since the channel is full and under 20ft (8.65psi) of pressure. I have used the hydraulic radius with a wetted perimeter equal to the perimeter of the channel. I modeled 90° bends by assuming they has about the same K value as a 90° mitre bend (=60*friction factor, according to Technical Paper No. 410 by Crane).

Beside asking if anyone has had any experience with noncircular ducts to tell me if i'm the right track, I’m also wondering about flow control for a larger channel splitting into two smaller ones. I’m pretty sure weirs are no use here since the channel is full.

 

[RWF7437]

Confusing!

You say you are using concrete "channels". By definition channels are not under pressure. Are you using circular pipes then ?

If you are using circular pipes then the Darcy Weisbach equation may be used but you don't need to caculate the Hydraulic Radius or wetted perimeter.

If the pipes or "channels" are not circular in cross section then you may use the Manning Equation and calculate the slope "S". Remember that when the conduits are flowing under pressure the slope is the slope of the hydraulic grade line, not the slope of the conduit.

Finally, what is the shape of these "ducts" ? How will they save you money compared to circular pipes ? Might be so, but it seems unlikely that they will.

good luck

 

[kchayfie]

I've never looked at full pipe flow for anything other than a circular section but your approach looks like the one I would take if it came up.

In terms of flow control, can you make the two small channels follow identical (e.g. symmetrical) flow paths? If so, that should take care of splitting the flow. Or you could use a combination of downstream flowmeter & modulating penstock to generate a positive flow split for the two channels. Weirs would only help if you can afford to break the pressure in the pipe.

 

[Daamz]

Thank you RWF7437 and kchayfie for your responses.
Just to clarify, these "closed channels" are basically rectangular (or square) concrete pipes. Having a water level about 20ft above the inlet makes the "noncircular ducts" under 20ft of head until the water get to the filters where it enters an open channel through a vertical shaft.

 

[gbam]

Just use Manning's to compute the headloss due to friction.

Q=1.49/n A R^(2/3) S^(1/2)

Q= K S^(1/2) --> K = 1.49/n A R^(2/3)

Sf= (Q/K)^2

hf= L Sf --> Friction Loss

Minor Losses are the Same!

Refer to HEC22, FHWA, see link.

http://www.fhwa.dot.gov/engineering/hydraulics/library_arc.cfm?pub_number=22

Should be straight forward.
Have Fun!!