Constant (low) pressure reservoir

[undapresha]

Ok here is the deal. I am looking at constructing a reservior (say 3L in volume) that can act as a constant pressure source (say 100mbar). The theory is that the reservoir will be filled with air with a pump then used as needed. While the volume of air would decrease when a tap was opened, the air pressure would be kept constant by way of a constant force (weight) applied to a moving boundary, either:

a) a flexible membrane or
b) a rigid disc (imagine a piston being the moving boundary in a cylinder for example)

Therefore the air pressure in the reservoir would depend on the weight used.

Are these ideas feasible? I realize that the idea (b) would require some good seals to prevent leakage, but this would not be too problematic as long as the pressure was kept constant and the leakage was a small percentage of the total volume.

Any thoughts? Any advances on these ideas (reservoir, not pumping solutions)? Cheers.

 

[Topher925]

It does sound feasible but rather complicated. You do realize that as the water level changes so will the pressure right? In other words, you will have to keep adding weight to top of your reservoir in order to keep the pressure constant as the volume decreases.

 

[undapresha]

Sorry I meant for air, post edited.

 

[FredGarvin]

It seems like it would be a heck of a lot easier to just use a regulator valve. You're going to have a science project to get this tank idea to work and it will be more complicated.

 

[undapresha]

Haha yes well a science project isn't always so bad. :)

Re:regulators - if I'm looking to achieve flows of the order of 0.5ml/sec, pressures around 0.1bar and tubing diameters around 1mm all at the output, would I be able to accomplish this with regulators? It is important that these pressures/flows at the outlet are constant - I don't want any drift.

I understand that 'droop' can be a problem with regulators, but I don't know how this applies to my situation. The volume of my tank/reservoir is only of the order of 5L and does not act as a constant pressure source.

Thanks for all the answers so far.

 

[undapresha]

So does anyone have any ideas?

I admit that using a pressure regulator seems like the easy (and therefore best) engineering solution. However while some regulators meet my pressure requirements (output~100mbar), most are designed for flow rates of the order of m^3/hr (SCFM) and not cc/min. I require < 30cc/min.

I'm all ears!

 

[FredGarvin]

Sorry about letting this one drop. I hadn't realized that your flows and pressures were so low. With flows that low, usually peristaltic pumps are used. The ggod ones are used as flow metering pumps. None of my sources touch the ranges you need. I will keep looking around though. I would think something in the medical field would do the job.

 

[undapresha]

No problem, I appreciate the feedback as I'm a bit new to this - the number of posts you have would indicate that you've got a few questions to answer anyway! Plus I didn't outline my flows clearly at the start, which would have saved everyone some time.

You're absolutely right about peristaltic pumping. The issue in this case for me is that flows change as the tubes age due to the compressive action of the pump, so I'm looking for a lower maintenance solution.

Re: regulators, the closest thing I have found so far is the low range model available from Bellofram, which can output 0.03-1.7bar, I'm just trying to find out the flow limitations.
http://www.bellofram.com/PCD/T-10.htm

Thanks for the input!

 

[Q_Goest]

hi undapresha. is that from the queen song? :tongue2:

The flow you want is so small, you probably won’t find a valve that can do it reliably. If you knock pressure down to just a few psi, then one last time to 100 mbar, the Cv you need is ~ 0.00001. That’s a bit smaller than the smallest metering valves made by Parker or Swagelok. You could use a sintered orifice to get the flow (22 SCCM) or maybe a ruby orifice (synthetic ruby with a hole in it). I don’t know of anything else that can control air flow to the rate you want it.

But do you really need to control the flow, or just pressure? If you’re going for pressure, then a small, fixed volume downstream of your flow restriction might be sufficient without having to go into sliding pistons. The idea would be to control pressure using a small solenoid valve and limit flow through a flow restriction such as a very small metering valve, sintered orifice or ruby orifice. By measuring pressure in the fixed volume, you could then open or close the solenoid valve to maintain the pressure within the pressure band.

The analysis for that is fairly simple. You could assume constant temperature gas going in/going out of your fixed volume, a fixed inlet flow rate, an upper and lower pressure control point, and with that information you should be able to easily calculate the on/off time for your solenoid valve. As long as your on/off time seems reasonable, (on the order of a few seconds or so) then knowing how many cycles you can put on your valve before it fails, you can optimize the system for length of service, volume size, and type of restriction.