Non-Steady State Venturi Flow 1

[TheDarnoc]

Most of my fluid dynamics classes dealt with nice and neat steady state flow problems, but it turns out that now I need to know about how a system will start up.

Dilemma: I have a system where I have an inlet and then two flow paths (the fluid is water, say at room temperature). The goal is to have the water in branch A reach the 'same' endpoint/reservoir before branch B. However, I need branch B to have a much larger flow rate than branch A when they finally meet at the reservoir. I have limited space so just lengthening branch B could be an option, but it would be very difficult. Furthermore, the process being used to create the geometry is difficult to control (+/- 0.05").

One thought I had is to have a series of venturi orifices in branch B order to slow down the flow initially, but when it reaches steady state, the losses will be minimal. Is it worth it to perform a test?

Just to give you an idea of the dimensions: branch B has a diameter of ~0.75" and a length of ~18"; branch A has 2 sharp-edged orifices of diameter ~0.25" then goes into a chamber and then into 20, ~0.2" diameter sharp orifices (imagine a hose with holes in it).

There exists a proper 'ratio' of the two flow rates, but that is during steady state.

Alternate ideas on how to solve this problem would be wonderful BUT I'm mostly interested in an evaluation of my series venturi's idea.

-Con

 

[TheDarnoc]

By the way, I have entertained placing some sort of check valve creating a kind of high-pass filter for pressure, however having any sort of mechanical insert is undesireable due to the process.

 

[TheDarnoc]

Here is a picture of what I was thinking. Diameter sizes are approximate. I just wanted to make sure people understood that the venturi's would be expansions to the inlet diameter, not contractions (mini reservoirs). Also, I didn't show all 20 outlets coming from branch A, just few so you could get the idea.

Thanks in advance,

-Con

 

[BigInch]

You need to add as much volume to branch B as possible. One would normally do that by increasing the diameter of B, but if you want to make it fatter in several places and call them venturis, why not? I think I might call them small reservoirs. Problem is that when you increase diameter, even though velocity is slowed, you also tend to increase flowrate, due to a smaller pressure loss/length of travel (less resistance), so you need to increase B's diameter, but with some added resistance. If you can get the pressure loss to velocity ratio correct inside B, it might work. Its possible that during start-up you could get some of A's fluid entering the outlet of B, until the flow in B develops towards the end of B's length and mixes/forces it out again. Its all about where the downstream pressure in B and A equalize and the length of the streamline from A to that point and B to that point both during startup and afterwards, when that point starts to move downstream and off to the right of your sketch (i think). I don't suppose you could flow laminar in B and turbulent in A, to screw around with the friction factors??? Oh well... just another idea. Actually this sounds like a job for nanomachines. Good luck.

http://virtualpipeline.spaces.msn.com

"What gets us into trouble is not what we don't know, its what we know for sure" - Mark Twain

 

[katmar]

You have mentioned your start up conditions, but no restrictions on shut down. If it is allowed by the process you could have the "A" line bowed downwards so that it would remain full on shutdown. The "B" line could run sharply upwards and then gently downwards so that it would drain on shutdown.

On startup "A" would be full and fluid would exit from the discharge as soon as new fluid entered the pipe. You might need some sort of weir with a drain hole in it at the end of "B" but it should be possible to use the whole sloping section as a reservoir that has to be filled on start up.

Or just open the valve to "A" before you open the valve to "B" ;-)

Harvey


Katmar Software
Engineering & Risk Analysis Software

http://katmarsoftware.com

 

[TheDarnoc]

BigInch,

I understand what you are saying about the pressure losses vs. resulting velocities. I was trying to minimize pressure loss with the venturi's (perhaps I don't really show that in the sketch, sorry) but keep the steady state flowrates the same. My goal is really to just delay the flow in the beginning and not introduce any real restrictions for the flow.

Another correction about the sketch: the resulting flows from branch A and B end up going out to atmosphere are there really isn't any backpressure at the exit of B other than a few inches of water.

In terms of laminar vs. turbulent, I don't think that's possible since the inlet flowrate is >10gpm and the ratio of flow through branch A to B is between something like 1:4 to 2:3. That ratio is important.

Thanks!

-Con

 

[BigInch]

Then just try adding storage volume for start-up to B. Once the volume is full, velocity and flowrate will begin to steady out. If you need more velocity in B, you could try necking down B towards the outlet. If you need more flow in B, scale B up so its proportionally larger everywhere.

http://virtualpipeline.spaces.msn.com

"What gets us into trouble is not what we don't know, its what we know for sure" - Mark Twain

 

[TheDarnoc]

Katmar,

Yes, having the water in the reservoir on shutdown would help. That is a good idea. We have generally believed that having standing water, however, would invite bacteria, mold, or a fungal bloom. BUT, we already have standing water at the output of the two branches and do not have issues/complaints. The difference is that the branch A reservoir does not receive the proper coating that the reservoir of the two branches receives and cannot in the process.

This is worth looking into. It's nice when people can look at a problem objectively!

Now, tell me about my friend Venturi.

-Con

 

[TheDarnoc]

BigInch,

Having little accumulators was the idea, it's just due to the lack of space I drew them the way I did. You are correct.

-Con