Separation of liquid flow

[Nicotik]

Hi !

I need help on the following problem : I've got a liquid in a pipe that splits into 2 in a "T-junction" (see drawing below) :

I want to know if the liquid will split equitably (50% to the left and 50% to the right) or not.

Length after the elbow = 3m
Inner Diameter 0,27 m
Flow volume 0,04 m3/s
Section 0,057 m²
Velocity 0,7 m/s
Density 1050 kg/m3
Dynamic viscosity 0,0011 Pa.s

Thank you very much !

 

[anorlunda]

The images in your post are not visible. Use the UPLOAD button to include images.

But speaking of any T junction, the split in flow between the two paths depends on the total flow resistance of the paths. If they are identical, then the flow should split roughly 50% 50%. Note that flow resistance is not just for the junction, but it must include whatever they are connected to down stream.

 

[Nicotik]

Thank you Anorlunda. My knowledge in fluid dynamics is very limited. Could you explain what you mean by the total flow resistance of the paths ? Is there any formulas available to calculate it?
Thank you very much for you time !

 

[anorlunda]

Calcuations would be difficult for you. There are many factors that influence resistance. Even with a formula, you don't have the values of the constants to put into the formula.

What are you trying to accomplish? Maybe we can help that way. Why is are equally divided flows important? Are you making an irrigation system to distribute water?

 

[Nicotik]

I want the flow to be as laminar as possible and the fact that the flow isn't divided equally may have an impact.

It is a ferromagnetic fluid that's used to separate different materials depending on their densities. The more the flow is laminar the best is the seperation.

 

[anorlunda]

In that case, instead of calculating, I suggest valves in the output lets. Start with both valves wide open. Then partially close the one in the leg with the most flow until the flows balance.

Turbulent versus laminar flows are more a matter of Reynolds number, not dividing a flow into two equal subflows.
Are you familiar with Reynolds number?

 

[Nicotik]

Oh I didn't precise but I don't have the machine so I can't do any experiments ...

Yes I already calculated it (170 000 so very turbulent)

 

[boneh3ad]

At a Reynolds number like that in a pipe, any discussion of laminar flow is nonsensical anyway.

 

[Nicotik]

I just talked with someone about that and he said the flow doesn't necessarily have to be laminar ! Because the fluid is redirected to a laminator block (where we make the flow laminar because of the honeycomb structure in the block).

So I'm just trying to know if the flow will perfectly split 50% to the right and 50% to the left. Do you know how I could calculate that ? Thanks !

 

[anorlunda]

Do you know how I could calculate that ?

You already have that answer (flow resistances) but you said you didn't understand it. Unfortunately it is not so simple as using a standard formula.

 

[boneh3ad]

Also at such a high Reynolds number, the flow may relaminsrize due to your "laminator block", but it is still highly unstable and is going to transition back to turbulence fairly rapidly downstream of that.