Controllable Valve to Maintain Constant Flow Rate

[hanson]

Hi all.
I am doing a design of controlling a fluid flow.
We have some fluid which is subjected to some weight, which will serve as the external force to drive the flow.
The objective is to maintain a constant flow rate.
Do you guys think I shall constantly changing the opening of a valve to achieve a contsant flow?
Since the driving force will produce acceleration rather than a constant velocity, we need to adjust the opening of the valve if we are aiming at a constant flow?
I am wondering if there is any valve which suit this situation?
The valve control should provide ways for me to change the value of the flow rate.
Say, when I want a constant flow rate of 200ml/min, I may enter 200 onto a panel, then the valve my self-adjust such that a constant flow rate of 200ml/min is achieved.
Is there any commercially available valve that is suitable for me?

 

[FredGarvin]

When you say "subjected to some weight" do you mean that your system will be gravity fed? I am assuming this because you said your head will change over time which, to me, means that you have a decreasing tank level as you flow.

I don't know of any reasonably priced valves that do what you want as a single package. When we do fluid systems the entire system is designed and calibrated to give you what you want. What is normally done is to have a control valve with a flow meter downstream. The flow meter gives you your flow display (sent to a remote display) and then the valve is controlled to the desired flow using the display. If you want to enter a flow rate and have the valve go there, then you'll need something that has a complete feedback loop and controller. That tends to get a bit expensive. You can get by relatively cheap if you can go completely manual.

Another thing to consider would be to eliminate the gravity fed system. Use a tank that is large enough to hold enough liquid for any kind of tests you want to do and blanket the tank with a pressure from a shop air source or the like. That way you have a pretty close to constant head pressure in your system despite tank level.

 

[hanson]

When you say "subjected to some weight" do you mean that your system will be gravity fed? I am assuming this because you said your head will change over time which, to me, means that you have a decreasing tank level as you flow.

I don't know of any reasonably priced valves that do what you want as a single package. When we do fluid systems the entire system is designed and calibrated to give you what you want. What is normally done is to have a control valve with a flow meter downstream. The flow meter gives you your flow display (sent to a remote display) and then the valve is controlled to the desired flow using the display. If you want to enter a flow rate and have the valve go there, then you'll need something that has a complete feedback loop and controller. That tends to get a bit expensive. You can get by relatively cheap if you can go completely manual.

Another thing to consider would be to eliminate the gravity fed system. Use a tank that is large enough to hold enough liquid for any kind of tests you want to do and blanket the tank with a pressure from a shop air source or the like. That way you have a pretty close to constant head pressure in your system despite tank level.

Thanks FredGarvin.
What you mentioned are exactly things we are considering.
We fristly got the idea of gravity fed, but as you mentioned, the tank level will decrease and the speed tend to be varying.

But for the external air pressure applied to blanket the tank, would this really produce a constant flow? I don't mean the tank level decrease, which I know would be negligible compared to the pressure we applied. But will applying a pressure, hence force, onto a fluid flow, gives a constant flow rate? I have in my mind the Newton's second law, force <=> acceleration. I am kind of confused. Please help.

 

[stewartcs]

But for the external air pressure applied to blanket the tank, would this really produce a constant flow? I don't mean the tank level decrease, which I know would be negligible compared to the pressure we applied. But will applying a pressure, hence force, onto a fluid flow, gives a constant flow rate? I have in my mind the Newton's second law, force <=> acceleration. I am kind of confused. Please help.

In a gravity feed system like the one you described, applying a constant pressure to the tank would yield a relatively constant flow rate.

Think Bernoulli equation instead. The flow rate is constant due to the large applied pressure at the top of the tank. P1 is disproportional as compared to the hydrostatic pressure developed by the fluid column in the tank.

Just for a quick example, if you assume the tank is discharged to the atmosphere (like the examples normally given for Torricelli's equation) you'll see that if the pressure at the top, P1 is greater than P2 (discharge to the atmosphere or relatively low pressure), the flow velocity remains relatively constant. Hence the flow rate will be relatively constant.

Example:
P1 = 100 psi (689475 Pa)
P2 = 0 (assumed to be essentially zero even though 1 ATM is present - this makes the calc quicker for me but you can use P2 if you like for a more accurate answer)
h = 10 meters
p = 1000 kg/m^3 (water density)

At 10 meters the flow velocity at the outlet would be ~39.68 m/s
At 5 meters the flow velocity at the outlet would be ~38.43 m/s
At 1 meter the flow velocity at the outlet would be ~37.39 m/s

As you can see the flow rate (Q = A*v) is relatively constant.

Chris

 

[FredGarvin]

But for the external air pressure applied to blanket the tank, would this really produce a constant flow? I don't mean the tank level decrease, which I know would be negligible compared to the pressure we applied. But will applying a pressure, hence force, onto a fluid flow, gives a constant flow rate? I have in my mind the Newton's second law, force <=> acceleration. I am kind of confused. Please help.

Stewartcs pretty much spelled it out for you. I just wanted to add that instead of thinking of f=ma, think of the tank blanket pressure as controlling the delta P of your system. Flow is proportional to the delta P. If you keep that relatively constant, your flow should stay pretty constant.

 

[hanson]

Thanks FredGarvin and Stewartcs.
But why flow (you mean flow rate, right?) is proportional to the delta P?

 

[FredGarvin]

Look at the Bernoulli equation as Stewartcs showed. P1 is the tank blanket and P2 is whatever is downstream (atm in his example). That is the delta P responsible for the flow in this system. Pressure difference is what causes flow.

 

[Homer Simpson]

additionally, in piping that is full of liquid, the only way the velocity (edit: not flow rate) can go up is for the piping diameter to decrease. The only time I can think that F=ma applies in piping systems is when a 'slug' of water is being accellerated down a pipe (as in water hammer) by pressure behind the slug, like a bullet down a gun.

If you consider all the water in the head tank and piping to be one large 'slug', why does it not accellerate with a cover gas pressure applied?? If you drew up a free body force diagram that at the top surface of the water all forces would be balanced. the upward and downward force at the surface are the same, because the pressure is the same throughout the whole tank, and the force = pressure x area. So with no unbalanced force at the liquid suface, no accelleration of the surface.

 

[FredGarvin]

additionally, in piping that is full of liquid, the only way the flow rate can go up is for the piping diameter to decrease.

Are you sure you didn't mean "velocity" and not flow rate? I am assuming that you are referring to a situation where inlet and downstream pressures are held constant?

 

[Homer Simpson]

Are you sure you didn't mean "velocity"

Sorry, you are of coarse right. Velocity goes up as pipe diam decreases, flow rate remains constant.

 

[stewartcs]

Sorry, you are of coarse right. Velocity goes up as pipe diam decreases, flow rate remains constant.

Until the flow becomes choked.