High Speed Valves

What size pipes? What's in the pipes, its temperature and its pressure?

Oh, and what size is the "small volume"?

I am looking at a pressure difference of 15 bar (217 psi) and varying temperatures of -50 to 200ºC (-58 to 392ºF). The pipe size depends on the valve connection, so I am thinking 1/2 to 2" diameter pipes. The volume would also then depend on the connection so on the scale of 100 to 500 cm3.

...and the material in the pipe and its pressure?

oops.. It would be air flowing from a 15 bar source with variable temperature to the volume initially at ambient pressure and temperature.

Of greater concern than the opening speed of the valve is the pressure rise time of the air entering the vessel. How quickly does it have to get up to what pressure to qualify as a successful process step?

to be clearer.. It is the pressure increase rate which matters most to me and not the final pressures. I can obtain the same results when increasing my pressure from 1 to 2 bars as from 1 to 5 if the pressure rise curve follows the same path.

I need a minimum mass flow for this pressure curve to be least affected by the poppet's opening...

A 1/2in solenoid valve, when energised, will go from fully closed to fully open in about 0.1sec. Is that fast enough?

Unfortunately I have to be quite nitty gritty for this project.

0.1 seconds is ok but what is important is the shape of the opening curve in the first few milliseconds. In other words I need to maximize the flow into the volume primarly for the first ~5% of the valve's opening.

So in other words, my question is: what kind of valve would accelerate the most upon initiation of the opening.

Interesting point. The OP seems to only be interested in how quickly the valve can be opened, not how quickly the valve can be closed.

...and less than highly interested in what qualifies a successul process pressure rise criterion from an unsuccessful one.

<unsubscribes>

We can lead a horse to water....

Sorry for not noticing your previous post PWslack...

That was a good question, and I realize that is a major factor determining the test's success. Unfortunately I haven't quantified this p. rise as the mass flow, volume also affect my outcome. Basically, the pressure increase's function is to accelerate a poppet (placed after the volume) as fast as possible to create high impact speeds.

So I am merely determining my success by testing various possible valves in a Modelica based simulation to see if I get a desired output in my test results.

Now that you pointed it out I'll try and quantify the required pressure increase rate and maybe this will help me be more precise in my search..

hmm isnt this a none reclosing device? I'd find out more quicker but the wikipedia blackout is making things harder...

I was thinking of building a custom disk valve but that would be a lot of work and I am trying to avoid that.

The more I think about the idea of a rupture disc, the better I like it. Using a rupture disc will clearly be faster than nearly any reusable valve. More to the real point of the matter, a rupture disk can be made cheaply and to many different orifice sizes. By popping a hole in a rupture disc and measuring the rate of pressure increase with different orifice areas, different cavity volumes and even different initial pressure differentials, our OP can quickly determine what kind of valve speed is really necessary for the pressure control desired.

This will certainly be much more fun than reading about Van der Waals' law or even Boyle's law or even the Ideal gas law. This just might have something to do with thermodynamics, too. You know it might even be a good idea to measure the gas temperature on both sides of the orifice and at a distance away from the orifice. Who knows, this may even involve differential equations somewhere.

Totally agree with redfred on this one. A rupture disc can be purchased in virtually any size, and when used with a squib, will instantaneously open full bore.

I use these in systems to suppress an explosion after it has started, and can stop the flame front with these speedy devils.

A phrase that comes to mind when I read our OP is "Explosive Release". Thus my suggestion of rupture disc. A chemical explosion would be more entertaining, but is probably less practical.

We have not discussed how often this process occurs. If it needs to close, another valve upstream of the rupture disc can (presumably) handle that chore.

yea.. I omitted that detail. The final apparatus needs to run at the highest rates possible, i.e. aim for 1000 cycles per minute. It therefore has to be highly repeatable and that's why I must use a real valve.

What does OP stand for btw? I am quite new to discussion forums..

OP stands for Original Poster

1,000 cycles per minute? yes, an important detail. In your comment #8 you tell us that 0.1 seconds to open is fast enough. If close time is the same, that is about 300 cycles per minute maximum.

OP is shorthand for 'Original Poster'. In general, the forum dislikes text shortcuts, but OP is one that seems to be acceptable.

I was saying its enough as in I can settle for 300 rpm.. but 1000 is my goal :)

Redfred and doorman have given you the best solution for a fast opening valve. The burst disk goes from fully closed to fully open in the smallest amount of time (dependent upon pressure and flow rates).

But if you want it to be able to close again, you need something like a pilot valve. One could be designed with a surge protection setup that when flow increases it closes and resets the valve. A simple example of a surge protector is a water hammer. A good example of a pilot valve and burst disk is found here. I used the burst disk to launch a tennis ball about 150m.

If a surge protector is installed upstream of the pilot valve and a heavy enough spring used, the surge protector will close and the pilot valve will close at almost the same time, allowing the surge protector to open again. (here is a good article on surge protection valves)

The best way I can think of to get the kind of cycling you want would be to use a spinning disk with open holes and have it pass through your pipe. It will block flow momentarily then open fully. The circles can be elongated into ovals to allow longer open periods. It will be tough to get a seal that will allow the disk to spin and not allow leakage.

Drew

Mass of poppet ? Stroke of poppet ? Expected end velocity at impact ?

I made years ago an oscillating cylinder totally mechanical control with a commuting end of stroke time of less 0.005s. I used a system with positive and negative stiffness springs which allowed a very high acceleration as soon as a threshold was reached.

Due to the alternative movement you need it could be a solution.

Mass = 8g

Stroke = 4-10 mm

Impact Velocities: 15 m/s one way and 5 m/s the other way

aim of 1000 cycles per minute..

I thought of making an oscillating cylinder but then its difficult to get different impact speeds on both ends. It is then also difficult to perform with poppet temperatures varying from -50 to 200ºC.

Good way to evaluate the required dimensions. However is there even a valve that will offer comparable opening speeds? I will think about using this..

Assuming that the valve opening rate gradually increases, I realize now that the best way to catch the highest pressure increase is by selecting a poppet downstream pressure which will time with a higher valve opening rate (where the mass flow is higher)

I am not sure if this is clearly written...

I seriously doubt that the timing of your valve opening will be the critical factor of a yet to be well defined manifold and pressure control. Looking for a reusable valve as fast as a rupture valve before you determine if you need this speed is a waste of time. IMHO.

Remember, the speed of sound of your gas media will be a limiting factor. Because of this velocity limitation and the unknown manifold dimensions, you can very easily reach a transmission line condition where the pressure front does not reach your end cavity before the valve is supposed to close.

Redfred, so you are proposing that I first test with a rupture disc.. That would mean building a system to be tested only once (per disc).. I am sure that such a disk would work and hence why I don't see the point in testing with it. It will give me a favourable result but won't tell me how a the solutions will fare. except maybe for helping me with pipe sizing..

My point is that it appears to me that you are looking to make some measured and controllable mass quantity transferred relationships with this 1000 cpm mechanism. I expect that you'll find that by doing a single action burst opening test that your manifold before and after this valve will significantly restrict your range of controlled mass movement. I believe the convolution equation of the orifice area opened during your 1000 cpm cycle will be a useless parameter to know. The burst opening function itself will be close enough to a true step function that data collection from this test will be valid for telling you how much of an effect the manifold will have in your design. I expect that you'll also find that as your downstream manifold lenghtens that spreading of your 8 milisecond puffs of gas from transmission line effects will be significant enough that you might as well just vary your orifice opening size in a DC fashion instead of pulsing.

On another point, one valve approach that nobody has mentioned that easily moves at the cycle rate you're looking for are the intake and exhaust valves of an internal combustion engine. Much design work has already been done on this controlled gas movement system. You should explore this documentation to see what is already well known to apply to your research. I remember that somebody explored the practicality or lack of it for a motor to have solenoid driven manifold valves. I vaguely remember that they found that as engine speed was increased that a similar phenomena to the valve floating of a mechanical valve occurred. (I think I read this in a NASA Tech Brief article, but that could be wrong.) I do remember that the experimenters were initially surprised at the large amount of power they had to drive into these solenoids.

Thank you for the lead engine lead, I will look into this. However I predict that these solutions will not offer my desired high flow rates for volume filling..

Otherwise, I am honestly a little confused when it comes to the whole convolution equation.. but with this burst I guess you are saying that I should determine experimentally the necessary manifold diameters to obtain the necessary mass transfers.. This makes sense to me. I am still an undergrad so experience is not on my side. Also, by using DC you mean using direct current with step function pulsing?

Actually what I meant by DC was an effectively zero frequency, steady state modulation approach to your valve for gas control instead of a pulsed modulation technique. (Being an electrical engineer I naturally and in this case wrongly insert electrical terms in my discussions.) This may not be a valid analysis because I do not understand how your valve modulation concept fits into your project.

As for my explanation of a convolution, I suspect that since you're an undergrad you may not have covered this mathematics in your classes yet. (At least I hope that's the case.) Let me start with stating a few given conditions that you've yet to determine. You found out that with the manifold you must use because of the geometry limitations and the pressure differentials you anticipate with your system that the orifice opening must have a certain area (A) for 8 milliseconds. If you fabricate a circular orifice with a radius of (A/pi)^0.5 and open this with some valve mechanism for 8 milliseconds you'll find that less than the amount of gas you expected will travel through your orifice. This is because your valve will not open and close instantly to make you an orifice of only A or 0. Instead you will get a non-linear but contiguous function that slides from 0 to A to 0. This obnoxious, tedious mathematics that describes the orifice area is known as doing a convolution of two functions. In your case one function will be the orifice area A and the other function will be the kinematic function of the movement of the valve that nick name was describing.

I see a refresh in the mathematics of a convolution of two functions in your future.

use 2 valves in series. Solenoid+flow control valve

Why not see for yourself:High Speed Valve | Products & Suppliers on GlobalSpec

Thanks! I was partly hoping for links like this

I have used a fast rise time pressure transducer calibrator made by PCB Piezotronics. Rise time was in the pico second range if my memory serves me correctly and it was a slide hammer on a poppet valve arangement. Maybe this technology would work for your project.

A picosecond is 10^-12 second. I think your memory is off. I cannot see hundreds of megahertz for a pressure transducer. Now you may get hundreds of kilohertz with a piezo-transducer and likely our OP will need a sensor this fast to see the responses that they are looking to see these effects of a fast valve.

Yes you are right. The transducers are 0.2 microsecond response time and the calibrator is capable of 35 microsecond rise time at 1000 psi. Very fast but not quite picosecond.

is this what you mean?

http://www.pcb.com/TM_Pressure_Calibration.asp

yes

Process scenario aside. Obviously the bigger the valve, the more volume. If it's purely an SDV with no control function. Then spade, butterfly or ball valves is the way to go. The bigger the valve and the higher the DP's the bigger the actuator. The bigger the actuator the more volume required for fast opening. Check out Fairchild Pneumatics. They make very simple and very robust pilot operated volume booster/quick exhaust arrangements for just this sort of application.

Another option would be a rotary union style with outlet at 90 to the inlet.

If you need to exhaust as well, use two driven from same motor/transmission set 180 apart.

1000 Cycles per minute would be achievable in this configuration- just use gearing and a motor running at least 1760 although 3600 would be easier to maintain more consistent action.

It is an intriguingly complex question. As it evolves, I do not think, it is the last of it we see. The two valve setup is unavoidable as you ask it. I have a problem with the original question.

Why?!?

Why does the whole control be IMPLEMENTED in pneumatic, when hydraulic, and in particular electric are magnitudes faster. No, I am not questioning the result, but the presumed process.

Waw amazing. I never thought posting on a forum would yield so many suggestions.

Anyway, ideally I would like to buy a valve (or two valves..) which accomplishes this task, this way I can save on design time. But it seems there might not be such a solution. The rotary mechanisms are very attractive from my point of view, I have to read up on them.. I don't yet understand the rotary union too well for example. I will also have a look at the pressure transducer calibrators mentionned and the possibilities that Drew mentionned.

To answer your question Leveles:

I've looked at mechanical/electric possibilities of creating the high impact speeds I need (15m/s) and found many obstacles, also the testing environment wouldn't be as representative as the poppet to be tested wouldn't be under high pressures.

I didn't look at hydraulic options as I am really unfamiliar with them. Also I am doing this for a pneumatic group so they are mostly familiar with pneumatic systems and have such ressources available. Maybe you can throw something at me to consider but I have accepted to stick to using pneumatics at this point.

Pinch valves will get you the volume per the pipe size, but the 1000 cycles pm will not be possible

Are you dealing with compressed air? I cant see any reference to what gas it is (if not liquid).

I have experience in compressed air. For a faster built up of flow/pressure I would use a (much) larger valve. Although in my work I have to achieve the opposite where a very slow build-up is required.

But before rushing off to buy a bigger valve you need to consider the pipe size before-and-after the valve because that is where the restriction to flow will occur once the valve is open. And the shape of the flow at the valve orifice will tend to form a 'nozzle' profile (whatever the 'shape' of the opening itself) and choke at sonic flow (remain constant) when the up-stream pressure is more than double the down-stream pressure.

Yes I am dealing with compressed air.

What your saying about using a larger valve makes sense to me. However, wouldn't a bigger valve also mean a slower cycling rate?

Faster pressure build up is indeed more important to me than cycling rate but down to a certain limit.. 0.1 seconds would be about the slowest I am ready to settle to I think.

The first job is to get the air into the system fast enough. So with pipework to allow this with a larger valve, the problem will soon become one of getting the air out of the system if you want to speed up the cycle rate.

I like the idea of a mechanically driven poppet/shuttle valve connected to a rotating crank on an electric motor. The shuttle can then be adjusted to reciprocate at a rate set by the speed controller of the motor.

A valve on an internal combustion engine can open and close when the engine is revving at 6000 rpm.

I've recently worked on some big diesels with solenoid operated valves, engines doing 1500 rpm. Pretty sure they open and close in a fraction of a rev. Take a cylinder head and see what your mechanics can do with it.

Till now only qualitative aspects have been discussed. I would suggest you have a look at some quantitative values:

assuming acceleration constant over the stroke =a [m/s²] Speed = a*t [m/s] and stroke will be s= a*t²/2 what we know is the stroke = 4e-3 ...10 e-3[m] and final velocity v=15 [m/s]. The required acceleration will be a= v²/(2*s) with above values a= 1.125E4 [m/s²] for s= 10[mm] and a= 2.81 E4[m/s²] for S= 4[mm]. If we consider that g= 9.81 m/s² you intend to reach between 1128 and 2810 g. The required force will be for the indicated mass of only 8E-3[kg] F= a*m = 90...225 [N].

If you use compressed air at 1.5 MPa then you need a piston area of 6 to 15 cm². So that to the mass of you hammer (8 E-3 kg) you have to add a piston mass a lot more important. Which of course will increase the force requirements ans so on....

I sincerely doubt that only a fast opening valve will be the solution especially when you look at the stroke times for 4 mm stroke it is about 0.5 ms and for the 10 mm stroke about 1.33ms. With valve opening times in the range of 10 to 20 mas (the extremely fast ones ) you are still far from needs.

Of course if I understood what you gave as answers to the various questions.

Thank you for the quantitative analysis, here is what I see to it:

I think the effective areas you have calculated should be 0.6 cm² and 1.5cm² (225N/1.5MPa=150mm²=1.5cm²). I am working with about 3 cm². Also, I said 1.5 MPa but I am free to use higher pressures if need be, the limitation being that high flow valves can't often handle higher pressures (from what I found on the internet that is..).

By selecting the right downstream pressure to my poppet (or hammer as you call it), I think I am able to postpone the poppet opening to a moment when the valve is closer to fully open, at which point the flow is almost at full speed (hence the pressure increase rate at its fullest). I don't know if this can be controlled realistically... all i know is it works out on my simulations using Modelica..

Another important concern, as redfred has pointed out is this manifold business. The flow restrictions due to the pipes may quickly put an end to my hopes..

Accepted, it was an error but what mass would have a 3cm² piston ? Would you have seals ? Then friction will ask for an increased piston area. Which means that the 8 E-3 kg mass is not the real one you have to accelerate. There is a possibility to increase the acceleration force making usage of a differential piston and a spring. Let say the "active movement" is from left to right. The right area of the piston can also be pressurized and generated force is used to compress a spring. Both sides are equiped with high dynamics valves (better to use 3/2 poppet valves) and when the one which control the left side opens the one at the right side does the same. So that at same time you increase pressure on the left side whose force adds to the spring force and reduce pressure on the right side. It is a well-know push-pull process which can increase in a consistent way the acceleration since the the force difference is increased. But I still doubt that even if you do all that you can reach the 15 m/s with a 4..10 mm stroke. I think you should look at a special very high speed solenoïd called conoid which was developped years ago for control of engine valves it had a high force and a small time constant.

You used several times as reference the Modelica model you made. May I say -since I do simulations with different softs since a long time - that the results depends on the model you made. The basic way to get a good result by simulations is NOT to believe the results as good if a simplified model computed by other means does not give qualitatively same results. To go more into the simulation you made. Did you consider that in the first period of the valve opening you have a chocked flow? (as long as the pressures ratio is under the critical threshold ). As Red Fred said for you the characteristic parameter is the integral of Av(t)*dt which gives the mass flow through the valve. Not only the connections are important as flow resistance but also as parasitic volume which will decrease the dp/dt and thus the acceleration. If you want to obtain the highest dAv/dt (which you need) then the best is to use several 2/2 ways valves in parallel and synchronized for opening. in stead of only one valve since the bigger the valve the higher the time constants.

As you see in above graph computed only as an indication and under the most favorable conditions as:

- valve totally open

- flow in chocked range (maximal)

no friction

piston velocity does not rise brutally but progressively and has the aspect of a c1*t^m1 where m1<1 the displacement grows as a c2*t^m2 m2>1. This is the reason I doubt you could obtain the velocities you expect on so short strokes.

What you could do is the have a mechanical brake on the piston -hammer - poppet and pre-load a chamber with a quite high volume and when the pressure in the chamber is OK then release abruptly the brake. It is a system which works as a burst plate. The valve dynamics can be less stringent since the release of a dry brake is on a very short travel and thus the full acceleration is from the start available. the basic principle is not to supply energy but to use an accumulated energy which can be released much more rapidly. In any case the mass is not the one you gave but will be a lot higher.

An other aspect is that you cannot use seals at such speeds. The friction will generate too high loss power and the thermal load at surface will be too high. You must use labyrinth sealing with a very low gap between piston and cylinder and several grooves to reduce as much as possible the leak and the risk of friction due the eccentricity (from non centric forces). If you have the possibility use aluminum with a hard anodized surface with ptfe which in case of contact has the lowest possible friction.

I would like to see your model and what results it offers. If you do not want to put it on availability for all then use the private channel. I use other softs as Simulink or Vissim

You definitely need a 2 step approach:

Charging the prechamber depending on the air temperature, to result in constant mass and/or constant speed of rise in the final chamber. Transfer curve to be designed-in, but determined late in the design cycle.

You definitely deal with gas flow at a significant multiple of the speed of sound at the opening stage of the main valve. In contrast with subsonic flow, that is a transmission line type flow, with multiple, confounding reflexions, including a steadily varying one, as the valve is moving to open. It does not matter, what type of valve it is, in that sense. The reflexions matter, as they restrict the flow.

The "fixed" reflexions wander around, as the pressure differential drops. The gradually opening valve introduces an additional variability during the opening.

The transfer time overall may be satisfactory, but it will be nowhere near linear.

On the other hand, If you need to stay close to ideal gas laws, and desire an exactly repeatable setup, consider the following: a dry piston compressor. Its compression etc. function is exactly predictable. Valves activated well outside of the "interesting" portions. Fed from the variable pressure prechamber described above. For added variability drive it with a VFD Variable Frequency Drive. There you can dial in from standstill to a very high rpm, as your heart desire. All components off-the-shelf, except the prechamber transfer function, and the final calibration procedures.

Continuing from #50, considering #51.

If this is a purely research lab setup, I would proceed as follows:

Prechamber, as I described before. Its stabilizing function is independent from the rest of the setup.

Use a free flying, magnetically driven plastic piston in a feedback loop for speed / acceleration control. Optical position readout. Piezo pressure readout. You will need the cooperation of a mechanical and electrical engineer, that the piston does not hammer itself to death at its endpositions.

Kilowatt power amplifiers. Waveform generators. Limiters.

Sequencer, to time all the pieces, including magnetic valves. With proper overdrive you can force valves under 1msec, or simply drive them outside the "interesting" time window.

Everything is adiustable. Exquisitely repeatable.

Everything is FAST: under 10 ms, if the same charge cycled in the cylinder, under 20-30ms, if new charge is needed. I bet dollars to donouts on that one.

Now, which one shall it be?

Hi cporowski,

This is all a bit over my head, (mostly the maths). I'm just a long in the tooth Sales Engineer who has spent the last 40+ years selling mostly pneumatics and would like to mention the following, based on an application at a customer of mine.

He used an SMC VT307-6DZ-01. This is a 1/8" bsp, 3/2, Direct Operated, "Spoppet" (guided poppet) valve with a 12vDC coil.

They used to "hit it" with a 24vDC spike that had the valve opening in low, (single figure) milliseconds and it lasted a long time, (a testament to SMC's quality).

They were "spitting" hot melt glue, onto the flap of a carton, that was used to contain the initial "free" staples when you purchased an office staple gun.

It was a high speed, high cycle rate application.

Best of luck with your application,

John

Thank you jes,

they work at low pressures for me(9bar) and low flow rates for me. But maybe its possible using many in parallel..

@Redfred

Understood the convolution problem. Its very interesting but I guess the easiest solution here is to test as you mentionned.. maybe with this rupture disk. Then I can find out what pressure differential is actually necessary for a chosen diameter.

The whole DC modulation story is still cloudy for me. I have zero electrical experience in design as I am in Mechanical.. I will have to read up on that a little before I set up my control.

I assumed the valve opening curve on the right in my simulations since I couldn't find any solid information regarding this.

@Nick name

To inform you more properly:

The poppet I must test is 8 grams and has the 3cm² area and is setup without sealing, only a spring. In big piston compressors these poppets may reach impact speeds of over 7 m/s with such 4 mm gaps..

Do you have any link for this conoid you mentionned? Can't seem to google anything related.

I like your brake idea, its like loading it up with a spring but using pressure instead which is better for my application. however the difficulty there is getting the 5m/s return impact speed.. I thought of the spring possibility but dropped it for this reason.

@Leveles

Thank you for your contributions. The piston compressor is a great solution and I am sure it would work. However, as much as I can I am avoiding this solution as the whole piston compressor setup is hard work for one person and quite costly. Hence the reason I am working on using high pressure source with fast valves..

The second choice, I am not sure about how you've imagined it all but it also sounds quite design intensive. Probably also a good solution but I don't understand it well enough. maybe u can clarify..

This is a working link for "poppet" ...

http://www.hellopro.co.uk/Hoerbiger_Origa_UK_Ltd-4972-noprofil-2004361-29477-0-1-1-fr-societe.html

You should give a better description of what you want to do. The Hoerbiger solution allowed in fact a tremendous increase in compressor efficiency it was developed in 1910-1920 th. If you want a similar test then the best is to use a cam to lift it and a spring to push it on its seat. Consider as well the spring mass in your total mass estimations since it can be under circumstances bigger than the one of the poppet. Take for not too long springs about 1/3 of the spring mass connected to the poppet and consider for your dynamic analysis the spring mass-less. The speed for the lifting is controlled by the cam which has a step and the spring accelerates the "poppet" toward the target.

Interesting assembly you've shown. I like the concept of a storage receiver and the multiple disc rotation. The necessity of lubrication is a point I forgot. It does bring to my mind a question of gas purity. If the gas is air then there can be added complications of suspended debris (pollen, dust, etc.) along with condensation complications from humidity.

One significant correction. The OP is looking for 1000 cycles per minute or 16+2/3 hertz.

Sorry, my error. I could never get 100Hz out of it anyway

It provides very good food for thought Thank you!

What was the application? Also testing?

It is a very neat looking design and ultimately I hope to build something along these lines. The main difficulties I feel are going to be the sealing and friction wear.. slowing down the spinning by making a bigger disk and adding more holes is good tip though!

Hi cporowski,

With regards to leakage/friction etc.

Either Norgren or SMC made "frictionless" cylinders, by using the same techniques as in making a lapped spool & sleeve pneumatic valves. The tolerances were very low resulting in minimal leakage and due to the "air bearing" effect very long life.

Best regards,

John

Your's is the application. I designed this during my lunch break based on the responses to this blog. Reading/thinking time 20min, write up 10min, drawing 30min, (inclusive of time for sandwiches, an apple and coffee).

But, I have previously designed multi-position, multi-port plate valves where each rotation of the plate through 30° reconfigures up to thirty connections. and I have previously used laser cut spacers to recreate an exhaust manifold for a ship's diesel engine in two days, when a replacement machined casting would have taken 4/5 weeks. So, old ideas reworked for a new application.

very impressive.. I guess my project fits quite well with your area of expertise . Thank you for your effort.

In what kind of position are you faced with such problems? Research?

Just asking as a reference for future job applications..

I did review the messages, and I must admit, I am more confused, than ever.

Are we chasing after the fastest mechanical valve, because that is an end itself?

OR, there is a research subiect, that needs some things to make it work in all likelihood. Walking down on a singular, logical yellow brick road, OP found a stumbling block in the opening speed of the inflow valve.

OP may declare, that I am nosy, and the fundamental question is none of my business. That is allright. The quest for fast mechanical valves is interesting in itself.

BUT, If that is not the case, stating the research case succintly, may open up other avenues. In this sense, the original question as stated is unnecessarily narrow.

Yes the original question was quite narrow. I meant only to find out about valves since in my project, I will initially test the possibilities using valves as it is the easiest and cheapest method of attempting to solve the problem. Then if the results are really bad I plan on developing something of my own probably using a rotating mechanism as shown in jhhassociates post.

I have moved passed the concept generation steps of my project and have already decided on using a pneumatic solution which is why I didn't post my question using the main project goal but instead the problem I am faced with at the moment.

In my solution, the pressure upstream and downstream of a single poppet valve must be controlled in order to translate a poppet constrained to move longitudinally in a way that its opening impact speed is 15 m/s and closing impact speed is 5 m/s. And this.. at the highest possible cycling rates.

I think with this things should be a little clearer..

Jordan Valve, and perhaps others, makes a short-stroke valve that resembles a gate valve. The gate and its mating surface have parallel slots. Moving the gate one slot width makes the valve go from full closed to full open; this could be done by some type of plunger, hammer, or cam mechanism.

I suspect that pipe friction dynamics would be more of a limiting factor than valve opening time.

Thank you! I will give them a go.