Designing a 2D Nozzle for Fluid Testing

[M.Pesh]

HI,
I'm constructing a 2D nozzle for fluid testing. I have made my nozzle X5 larger and taken a 1/8" slice. Now I'm having trouble designing the 2d nozzle so that the exiting air will have a velocity of 3ft/s. I must maintain an Re number of 1000 as well. My problem lies in that fact that. I can only buy pressure regulators. So what pressure must I use to achieve a exit velocity of 3ft/s
A=.000128 ft^2 (rather small)?

 

[Q_Goest]

Hi M Pesh,
Can you provide dimensions for your nozzle? Attaching a drawing would help.

 

[M.Pesh]

I have never loaded a picture into a forum before I hope this worked.
-M.Pesh

 

[Q_Goest]

You should be able to determine flow rate by making some assumptions on the fluid's (air? water? other?) density at the outlet. For example, if this is venting to atmosphere, and since you want a very low velocity, then outlet pressure would be atmospheric pressure (ie: there will be no significant pressure change at the outlet). So the first step is to determine flow rate required. Note that velocity and Re are dependant on each other, so you may or may not be able to obtain both Re of 1000 AND velocity of 3 ft/s.

Once you know your flow rate, you need to determine a method of controlling this flow rate. Regulators don't control flow rate, they control pressure. However, if you have a controlled pressure, you can then put a restriction downstream of it to control flow rate. If you can provide more detail now as to what fluid you are using and how you intend to control pressure, the next step is to determine how to restrict flow to give you the flow rate you require.

 

[M.Pesh]

The working fluid is Air and it exits to atmosphere. So my flow rate world be
Q=(v)(a)
(.000128)(3)(60)=.02304 CFM
I plan to control the pressure via a standard adjustable pressure regulator (0-100 psi), I am also open to suggestions.

 

[Q_Goest]

Since you can't measure V directly, and the geometry of this device is a bit too complex to calculate dP accurately for the given flow (and because dP across the nozzle will be so extremely small), I'd suggest measuring flow rate by putting all the gas through a flow meter prior to or right after the nozzle. Rotometers are cheapest option, and should work fine for this application. Get a rotometer rated for air from McMaster Carr or other supplier. Should cost around $30 for an accuracy around 3%. Rotometers are only accurate however, when they're venting to atmosphere. I suspect if you put the rotometer upstream of the nozzle, the pressure on the rotometer outlet will be essentially ambient, so that might work fine, but you'll need to measure downstream pressure in this case to verify it's ambient. Alternatively, you could have the nozzle blow into a chamber (ex: air tight cardboard box) and then have the meter inlet connected to that chamber so you can get a steady state reading of all the air going into/coming out of the chamber.

To vary the flow rate, I'd suggest putting a valve downstream of the reg, set the reg at some pressure, and then adjust the valve till you get the flow you want as measured on the rotometer. This method might be a lot more versatile and reliable than simply trying to set a reg. Pressure will likely be very VERY low with just the reg, so minor flucuations in regulator pressure will cause large fluctuations in flow.

Best solution is to get the pressure up to 25 psig or so, then have a small metering valve that has this (relatively) large dP across it so that minor pressure fluctuations in the reg setting don't significantly alter the flow. If P upstream of the valve is 15 psig or more, the flow will be choked and thus dependant only on the regulator set point and metering valve adjustment. If you set the reg at 25 psig, the metering valve will need a Cv of 0.00097. If you need help finding this valve, just shout.

 

[FredGarvin]

You got it right there Q. Measuring the flow up stream is the easiest way to go. That's the way we flow test our nozzles and other items. The key is the valve just downstream of the regulator. Once you get that guy choked, you're in business. That way any weird pressure fluctuations due to the nozzle don't make their way back to the regulator. If the OP has a rough idea of the expected flow rates, in stead of a rotometer, they could go with an orifice and just measure the delta P across that to get the flow. However, one trades off the inexpensive option for a slightly more detailed one. It all depends on what they want to do I guess.

 

[M.Pesh]

The rotameter up-steam will do the job.
As for the pressure reg and metering valve, I get what your saying. but I'm not sure how to find a metering valve with a Cv of 0.00097. A few suggestions would be a help!
Thanks for you help thus far, It's much appreceated.
-M.Pesh.

 

[M.Pesh]

Also since I made my nozzle 5x bigger do I have to divide my flow rates by 5 to keep everything scale?

 

[Q_Goest]

If the OP has a rough idea of the expected flow rates, in stead of a rotometer, they could go with an orifice and just measure the delta P across that to get the flow. However, one trades off the inexpensive option for a slightly more detailed one. It all depends on what they want to do I guess.

This is a good point, and it should be mentioned that if this is for a mass produced item, putting a regulator, metering valve and flow meter on it is going to be a very expensive proposition. Instead, integrating this nozzle into the system by having a known restriction such as the orifice as suggested, is a much more economical way to go. Also, looking through ways to eliminate the regulator would be advisable. But if this is just a one-off experiment you're doing, then having the rotometer is your best bet.

The rotameter up-steam will do the job.
As for the pressure reg and metering valve, I get what your saying. but I'm not sure how to find a metering valve with a Cv of 0.00097. A few suggestions would be a help!
Thanks for you help thus far, It's much appreceated.
-M.Pesh.

For a flow of 0.023 CFM (~ 1.38 SCFH) you might use a rotometer with an integral valve such as P/N: 5079K63 from McMaster Carr (4% accuracy):
http://www.mcmaster.com/

or perhaps a Brooks meter if you want more accuracy:
http://www.emersonprocess.com/brooks/category10.asp

There are plenty of others out there. I have to believe Omega makes one too.

If you need a metering valve which is independant of the flow meter, I'd recommend Parker, Swagelok or possibly a Generant.

Regarding the 5X bigger nozzle, flow is by area, so if area increases by a factor of 5, then flow increases by 5. If linear dimensions increase by 5, area and flow increase by 25.

 

[M.Pesh]

Ok, let's say I went with the pressure difference approach. How would I get the flow velocity?

 

[Q_Goest]

You can't measure velocity. Calculate flow needed to obtain the velocity you want, then measure flow.

 

[M.Pesh]

The venturi will work for vacuum correct?

Also, To keep my laminar flow and an Re number of 1000 with a nozzle 5x bigger, I do have to lower the velocity by 5 times. This means I now need a flow rate of .276 SCFH.

The Rotameter made by McMaster uses 1/8” NPT thread. This will allow me to run a 6mm(ID 4mm) tube into the valve. On the exit side I will use an expander to bring the 1/8 NPT thread to ½ NPT. This will allow me run a 16mm(ID12mm) tube. The expansion from 4mm to 12mm will slow the gas up by 4X times. Allowing me to run a higher valve reading and use a standard valve1/8", in and outlet

 

[Q_Goest]

The venturi will work for vacuum correct?

You're confusing me now. What are you referring to as a venturi? You only mentioned a nozzle up to this point. Also, you haven't mentioned anything about a vacuum yet. What vacuum? I was under the impression you had a nozzle with discharge to atmosphere.

Also, To keep my laminar flow and an Re number of 1000 with a nozzle 5x bigger, I do have to lower the velocity by 5 times. This means I now need a flow rate of .276 SCFH.

I was assuming you knew how to calculate flow given a velocity and outlet conditions on the nozzle. If not, please provide all information about the new nozzle you want to determine a flow rate on.
- Velocity
- Fluid (air?)
- Outlet pressure
- Temperature
- Flow area at outlet
- Any other pertinant information

 

[M.Pesh]

I wanted to know about vacuum because it's involved in another project. One to similar to this, only requiring vacuum.

The nozzle. (requirements)
-a required velocity of 0.6 ft/s (I would like to be able to vary it from .2 to 1.4)
-the working fluid will be exiting to atmosphere pressure
-the pressure in the nozzle is unknown
-temp is room so 70F
-The flow area at the outlet is .000128 ft^2
-I need to keep an Re number of 1000
-The working fluid is air.

 

[Q_Goest]

Because velocity is so low, the pressure drop through this nozzle will be negligable and the air can be treated as an incompressible fluid. You say you don't know the pressure in the nozzle, but the pressure drop to atmosphere is so slight that you can safely assume the pressure upstream of the nozzle is also atmospheric pressure.

Flow rate can then be taken from:
Q = VA
Where Q = volumetric flow rate (cubic feet per second)
V = Velocity (ft/s)
A = Area (ft2)

You have A and V so calculate Q. That's the flow you need to measure with the rotometer. Generally, the rotometer is in SCFM or SCFH, but you're already at standard conditions, so just multiply Q (cubic feet per second) by 60 or 3600 to get SCFM or SCFH respectively. (Ex: For your stated conditions of air at 14.7 psia and 70 F exiting an area of 0.000128 ft^2 with a velocity of 0.6 ft/s I calculate a flow of .27648 SCFH, so you need a rotometer that can measure that flow accurately.)

Note that this is average velocity, not peak velocity. Due to fluid shear forces, the velocity along the walls of the nozzle will be 0 and there will be some maximum velocity farthest from the walls. Between there will be some parabolic velocity distribution.

As for Re, that's simply set by conditions. There's nothing you can do to adjust that. For this case, Re (assuming a circular nozzle) comes out to 47. That's laminar flow as you would expect, but far lower than the 1000 you say you need. Not sure why you say you need an Re of 1000, perhaps you could explain what you're doing with this.

 

[M.Pesh]

I wanted to stay around 1000 for my own personal preference.
I'm obliviously not vary fluent in my fluids, but I was under the impression an Re number under 1500 was laminar. So I picked 1000 because it safely below the 1500 mark and it would allow me to run one of two choices:
1) Higher V
2) Smaller A
Now, I'm not so sure if this is true. (review of my textbooks needed...)
But I have decided to use the McMaster rotameter

Do you have any suggestions on vacuum flow meters?