Fluid flow test bench idea

[antithesis]

Have you ever seen how some guy on mythbusters or something drops dye into water flow to assess the aerodynamics of different objects?

I was wondering how closely we could mimic this principle when designing the pressurized components of our experimental pieces, particularly poppets, valve bodies, regulators or even complete systems.

What would have to be done to somewhat accurately mimic compressed gas? What fluid medium could we use, how fast would it need to flow to draw assessments, would it likewise need to be pressurized?

I was thinking of some things could be made in a polycarbonate or pmma mock-up, think tubes here, and mount the piece in question in this tube, making a closed loop of flowing liquid with dye injection ports strategically placed according to where it needs to be observed.

Well let me have it... could this method be used in some capacity to arrive at some general conclusions?

Edit:. Now that I think of it, if we could make a transparent test chamber with sufficient strength to support. the pressure, one could even use HFC refrigerant and leak test dye to quite possibly very closely mimic liquefied gas propellant, ala co2....or maybe even better, green gas

[Motorhead]

Coming from a background in small engine mechanics ... mostly very high performance 2 strokes in my later professional career.  Getting high speed cameras to view the flow dynamics was pretty straight forward being no real pressures outside of combustion were involved, only HEAT. So short run times could be observed noting flow dynamics not only in the lower cases, transfer path but combustion and expansion chambers too.

In a PCP we're dealing with VERY HIGH PRESSURE and I don't know of any Clear / Translucent materials that could handle the pressure in use today on a size scale we're actually using.   While I'm sure Larger exaggerated models could be made with porting specs we use to get the visual data, but not sure where the practicality exists doing so

Also the fact we're dealing with a pressure release into atmospheric pressure spaces, paths etc ...  The flow dynamics are already well known that sharp corners, abrupt change in cross section and distance all create drag.
The MATH of gas dynamics and that of differing gasses is also known ... how fast they expand, energy they carry and such stuff ( Of which all in over my head !! )

Thoughts and nothing more ...

[oldpro]

 Easier than you think! Heres a CFD program I've used for valve flow data and the like as d once you get a hang of it quite useful in many ways. https://openfoam.org/

[rsterne]

As long as the program code is written so that the high pressures and densities work as intended.... and also modified so that the speed of sound at those pressures is correct.... a CFD program may work properly.... It also has to allow for the fact that the HPA is expanding into a space that is essentially at atmospheric pressure (most programs I have seen assume constant density at inlet and outlet).... and that the barrel is corked by a movable pellet being accelerated by the flow....

I guess a Cray supercomputer could handle it OK.... a desktop PC, not so much.... but that is JMO.... The problem is that you may get what looks like a perfect solution to your problem, but you will have no idea if it is the correct solution....

Bob

[antithesis]

As long as the program code is written so that the high pressures and densities work as intended.... and also modified so that the speed of sound at those pressures is correct.... a CFD program may work properly.... It also has to allow for the fact that the HPA is expanding into a space that is essentially at atmospheric pressure (most programs I have seen assume constant density at inlet and outlet).... and that the barrel is corked by a movable pellet being accelerated by the flow....

I guess a Cray supercomputer could handle it OK.... a desktop PC, not so much.... but that is JMO.... The problem is that you may get what looks like a perfect solution to your problem, but you will have no idea if it is the correct solution....

Bob

Cray supercomputer huh? Ok bob go dig that thing out of the closet and fire her up...we all know you got one.

[Raider03]

Easier than you think! Heres a CFD program I've used for valve flow data and the like as d once you get a hang of it quite useful in many ways. https://openfoam.org/

Thank you for the info Oldpro!

[antithesis]

So in general, infeasible.

But in certain cases, if you are keeping the conclusions very basic and pressures and pressure differential reasonable, could one draw Amy very generalized conclusions about fluid flow, even if they are rudimentary ones that are well documented and the numbers crunched, say the best shape poppet for a particular application, or maybe something like the question I asked you a while back about a tube feeding pressure closer to a valves exit, or some basic visual conceptualization to merely illustrate principles that many of us are already familiar with? I do think a low pressure lp gas could be a viable option if these lower pressures have any relevance to the slightly higher ones of co2...of course the gas laws, mathematic principles, and speed of sound etc mean we could only approximate at best.

[Motorhead]

Problem also lies in the fact WE ARE NOT USING FLUID which is not compressible, but a gas which is compressible and does not act the same as a fluid 

[Mole2017]

I'd have to look it up, but flow experiments can be scaled to deal with conditions that are difficult to replicate exactly in full scale. Physically scaled models can be used, but sometimes changing the fluid as well can scale the problem. For example, water can be used to look at air flow problems. Conditions in a PCP valve might be tough, but water may yet provide valuable visualization if software is totally a non-option.

We had a visualization experiment that used water "seeded" with silica (or silicate--I'm forgetting which right now). A plane of laser light from a line generator (just thin glass rod held across the laser outlet) illuminated the particles much like sunlight illuminates dust in the air, except that laser plane gave a view of a thin slice through the test section. In that experiment we were watching convection currents in water. For fast transients, a camera would catch the blur of moving particles and thus give an idea of the flow through the valve.

[antithesis]

Problem also lies in the fact WE ARE NOT USING FLUID which is not compressible, but a gas which is compressible and does not act the same as a fluid 

Although part of me wants to counter with the idea that 1. I did say a few limited applications, obviously we won't be stuffing a clear tube with thousands of pounds of air.  2. Compressed air in some capacity DOES in fact behave like a fluid, although yes, for all intents and purposes replicating a pressure drop would be challenging and expensive, prohibitively so in my case and 3. Although I doubt that the medium would carry dye in the vapor phase, a low pressure lp gas could give a lower working pressure to examine the dynamics of co2 in some situations, which seems like the only real scenario in which something like this could work..

At this point in the discussion I accept that to even closely replicate conditions enough to study would be largely impractical if not next to impossible, and in many cases unnecessary.

[rsterne]

Compressed air, oil and water all act as FLUIDS.... However, oil and water are incompressible LIQUIDS, and therefore not the same as compressed air, which is still a gas (the molecules are not touching, but moving freely and colliding).... To complicate things further, above 550 psi, air is actually a supercritical fluid, but that is still distinctly different from a liquid, which is incompressible....

Bob

[JPSAXNC]

I mounted a valve inside a tube that was fed by 100psi. shop air, I had a micrometer head mounted on the valve stem to open the valve as I rotated the micrometer. With the valve that I was using, I was able to see that there was a large pressure drop when the valve had been opened .020" due to turbulence, before and after this point the flow was fine.

[Mole2017]

Very nice experiment James! Can you clarify the arrangement of pressure gauge and air source? For example did you use two pressure gauges to establish the pressure drop?

Of course, next comes the question of scalability: what happens at 550 PSI or higher, if the air is doing this super fluid thing that Bob mentions? For that perhaps a "blow down" test might be possible. Given a large tank at PCP pressures, use the micrometer to open the valve incrementally and note the pressure drop. It's the sort of test that might have to be done quickly to minimize the air consumption. (How long would it take to empty a scuba tank through a PCP valve?) With instrumentation, e.g. sensors for micrometer position and air pressures going to some sort of recorder, you can go really quick. The next fastest low tech approach is video simultaneously recording pressure gauges and micrometer position.

[JPSAXNC]

Hi Mole, I didn't use a pressure gauge after the valve, it wasn't needed. The tube was fed from a large capacity compressor, at a constant flow of 100 psi. The flow of air out of the valve was enough to hear and feel any change in flow out of the transfer port. I assumed that the turbulence was the result of some interaction between the valve head and the valve throat, and that it wasn't dependent on pressure. I had only tried 100 psi. because of convenience.

[rsterne]

I'm not quite clear on the results.... You say there was a large pressure drop when the valve was open 0.020".... Do you mean by that the pressure on the inlet side dropped at that point (indicating a large increase in flow rate) ?.... If you were only going by the "sound and feel", do you mean that it was at that point both increased significantly?.... Or, do you mean that the flow (and sound) was greater at, say 0.010" and 0.030", but less at 0.020" (due to turbulence)?….

Bob

[JPSAXNC]

Hi Bob, The flow rate out of the transfer port was much higher at .010 and .030, than at .020, due to turbulence at .020, The reduction in flow was big.  I used the valve in a gun, and it seemed to work fine, the valve may have opened fast enough with a hammer strike, that there was no effect on power??

[rsterne]

James, thanks for the clarification.... Interesting results.... I wonder if the flow was alternating between choked and not (probably not at only 100 psi)…. or just as you say, some turbulence effect....

Bob

[int3man]

Compressed air, oil and water all act as FLUIDS.... However, oil and water are incompressible LIQUIDS, and therefore not the same as compressed air, which is still a gas (the molecules are not touching, but moving freely and colliding).... To complicate things further, above 550 psi, air is actually a supercritical fluid, but that is still distinctly different from a liquid, which is incompressible....

Bob

Hi Bob,
             Finally  People who understand the difference!!  Fluid is not compressible.

Michael :- )

[rsterne]

Many fluids are compressible.... Gasses are fluids.... LIQUIDS are not compressible....

In physics, a fluid is a substance that continually deforms (flows) under an applied shear stress, or external force. Fluids are a phase of matter and include liquids, gases and plasmas.

A liquid is a nearly incompressible fluid that conforms to the shape of its container but retains a (nearly) constant volume independent of pressure.

Bob

[Motorhead]

Misnamed "Fluid & Liquid" on my OP as well.  Point and known non compressible nature of a "Liquid" was still being stated tho incorrectly named