solenoid valve design

Discussion in 'General Electronics Chat' started by strantor, Jun 7, 2012.

  1. strantor

    Thread Starter AAC Fanatic!

    Oct 3, 2010
    4,302
    1,988
    I'm thinking to design some solenoid valves at some point in the future (probably a long time from now) but I've been puzzling over it for some time now and recent questions about electromagnets spur me to post this.
    You may ask, why make your own solenoid valves? Well, they're meant to be for steam, to control a radial piston steam engine I'm dreaming up. Steam valves aren't cheap, especially when you want direct-acting valves and not pilot operated ones. So if I am successful in building my lathe, these solenoids will probably be my first machining project.

    I want my valves to shift as fast as physically possible. I know I need a lot of amp turns to achieve this, but I'm thinking that I'll want to keep the 'turns' part to a minimum because the more turns, the more inductance, and I figure, the slower the valve. So if I just wrap one heavy guage magnet wire around the plunger and shoot a couple hundred amps through it real fast, would that be better than wrapping it 400 turns and shooting 500mA through it?

    Is speed and power the same thing? I mean, is the amp-turns parameter the only thing you play with to achieve more power and to achieve more speed? or is there a seperate parameter that you can play with to get more speed while leaving power the same? I want speed. Power is somewhat important, but not overly; I don't need crack coconuts, I just want to slam the thing open and closed really really fast.
     
  2. strantor

    Thread Starter AAC Fanatic!

    Oct 3, 2010
    4,302
    1,988
    also, a mechanical question, if anybody knows the answer....
    Why do valves and pneumatic actuators (pistons) have seals? If you machined the parts with tight enough tolerance so that they fit air-tight, wouldn't that preclude the use of seals? What am I missing? I don't want to use seals if I can get away with it.
     
  3. spankey666

    Member

    Nov 30, 2011
    91
    3
    quire simply to stop by pass under high pressure, i doubt you could machine anything without highly specialist machinery to obtain the tolerances you speak of, but once heat is bought into the equasion , thermal expansion will alter everything.
    under standing how a o-ring / seal works is one step to your design intent.
    take for example a simple piston in a tube with a tight fitting o-ring. and you exert 1000psi of pressure to one side. if the o-ring has an interference of 4 psi, the required pressure to blow past will need to exceed 1004 psi so will never leak, that 4 psi makes all the difference :)
    have you thought of making it a barrel valve ? the turning on could be quicker and require far less energy to operate
     
    strantor likes this.
  4. GetDeviceInfo

    Senior Member

    Jun 7, 2009
    1,571
    230
    Seals typically isolate a closed system from the environment and visa versa. Within the closed system, fits, or clearances, often are the sealing factor.

    power is speed x torque. It goes back to the transformer question, do I need current, or do I need volts. Altering the mass of your armature is another variable you could consider.
     
    strantor likes this.
  5. crutschow

    Expert

    Mar 14, 2008
    12,984
    3,223
    You might google info on "electronic engine valve control". There has been much research in converting mechanical engine values to electronically controlled valves.
     
    strantor likes this.
  6. shortbus

    AAC Fanatic!

    Sep 30, 2009
    4,003
    1,521
    Steam engine valves don't normally use seals or 'O; rings. And they do leak. They use a series of grooves in the piston to seal. The groove traps condensed water from the steam and stops the pressure that way. The water is heavier than the gaseous steam so it seals pretty good but not completely. Good enough for the job though.

    Ever take a little Cox .049 model air plane motor apart? They use the same principle in the pistons in them. No rings or seals just a thin groove behind the piston head.

    Why go for electric valves? With out a pilot type valve its going to take a pretty heavy spring to keep it centered, against the steam pressure. All steam engines I have been around use a "shuttle" type valve to let the steam in and out of the cylinder. Not a separate intake and exhaust like a gasoline engine.

    The Cyclone radial steam engine uses a cam and push rods to drive its valve. Did you get the patent drawings from the web site? Shows a lot of how their made.
     
    strantor likes this.
  7. shortbus

    AAC Fanatic!

    Sep 30, 2009
    4,003
    1,521
    In car engines most of them use a valve to shut off the oil pressure to the hydraulic valve lifter. Not the valve them selves. Or the another way is to move a "wedge" out of the path between the valve rocker arm and the push rod, to keep the valve from opening
     
  8. crutschow

    Expert

    Mar 14, 2008
    12,984
    3,223
    I'm not taking about the various types of variable valve or valve shutoff mechanisms. I'm referring to designs that eliminate the camshaft entirely and use electronics to control the valve directly (in some cases using electro-hydraulic control).
     
  9. shortbus

    AAC Fanatic!

    Sep 30, 2009
    4,003
    1,521
    Sorry, misunderstood what you were posting.
     
  10. strantor

    Thread Starter AAC Fanatic!

    Oct 3, 2010
    4,302
    1,988
    ah yeah, good point
    Right, I used the improper term. Torque is the thing I'm less worried about. I want my power equation stacked heavily on the speed side.

    Well hey, that's good news. I feel like less of an idiot, not being too far off base about not wanting seals. :)
    I want to use electric valves so that I can change the timing while it's running. And because I want to make solenoid valves, because it seems like it would be fun to me. I'm planning on using spool valves - requires no power to hold it in any position, open, closed, or anywhere in between. also the simplest to machine, and simplest to attach a coil to. only drawback I see is that the gas has to make a few 90° turns to get in and out of the valve and I'd like to have nice, unobstructing type of valve.

    That's exactly what I'm after. I had no idea it's already been done in gas engines. Is there a damn idea left out there that hasn't thought of yet?
     
  11. strantor

    Thread Starter AAC Fanatic!

    Oct 3, 2010
    4,302
    1,988
    So any advice on making the solenoid operate at warp speed? Should I wind it like paper towell roll or flat like a washer?
    Should my plunger be made out of laminations or can it be just a regulat old piee of steel? (or stainless?)
    Should I go for less turns and higher amps or more turns and lower amps?
    Would it work at all if I only made one turn?
    I know there's more to it I'm not thinking of at the moment.
     
  12. shortbus

    AAC Fanatic!

    Sep 30, 2009
    4,003
    1,521
    I have some plans for an electrically operated valve, model gas motor I can scan for you, but the valves are just like regular poppet valves in a car motor. Also have quite a few plans for radial cylinder model engines to. Have the complete set of the old "Strictly IC" magazine. Was going to build some of the engines in them after retirement, until health got in the way.
     
  13. shortbus

    AAC Fanatic!

    Sep 30, 2009
    4,003
    1,521
    Don't know that there would be any advantage to changing valve timing on a steam engine. It's not near as important as it is in an internal combustion engine.

    Are you going to make a single cylinder one first as a 'proof of concept' ?
     
  14. strantor

    Thread Starter AAC Fanatic!

    Oct 3, 2010
    4,302
    1,988
    Sure, if you've got the time, I'd like to see any inspiration :).

    I haven't 100% convinced myself that there would be any advantage of it, but let me pose a couple of examples so you know where my head is on this.

    Details:
    I'll be utilizing both sides of a double acting linear pneumatic actuator. pushing on the cap end for extend, pushing on the rod end for retract. May even try a regenerative setup where I apply pressure to both side simulateously. Also considering an intermediate "dwell" or "float" period between extend & retract to smooth things out.

    examples:
    #1: Let's say I have a single actuator acting on the offset crank of a shaft. I'm operating very slowly, let's ssy 1RPM. I want to keep pressure applied to the cap end all the way through the extend stroke until I pass top dead center, at which point I'll apply pressure to the cap end to bring the crank back around with the retract stroke.

    #2: now, let's say I've got that same setup but now I'm running 3000rpm. Right now I'm applying pressure to the cap end, extending the cylinder. If I wait until I get all the way past top dead center before I start moving valves around, inertia will have likely carried the crank a considerable way through the 180° retracting phase. By the time the valves are done moving and the gasses have exhausted from the cap end and are starting to propogate into the rod end, who knows where the crank will be? I envision that it could be all the way 180° through the retract cycle and into the next extend cycle. So by the time I get my gasses right for retract, I'm already supposed to be extending, and the engine ends up fighting itself. So here, I would need to advance the timing so that valves start shifting and gasses start flowing before the point where they're needed. So I do this, I advance the timing, and it works well.

    #3: same setup from #2, with the previously advanced timing. Now I'm running 1RPM again. As my cylinder painfully slowly extends out to the top dead center position, it gets 90° through the extend cycle and then valves shift and the motor reverses direction. Why? because I advanced the timing to run @ 3000 RPM.

    So, I'm thinking I'll need to advance the timing proportionally with speed to maintain flawless operation. But I could be wrong. Then throw in my idea for the intermediate float position, this would exhaust both sides of the cylinder, say 75% of the way through the extend stroke so that by the time it passes TDC and we start applying pressure to the rod end for the retract stroke, there is less resistance, as the pressure on the other side of the piston is already ambient and incoming pressure does not need "push" against a higher than ambient pressure out the exhaust. Not sure if that's needed or practical, but that's the point of this thing, to learn.

    I have a strong suspicion that I'm overthinking this whole thing, as is what usually happens when I think about something for too long without actually doing it.

    and yes, I'll probably make a single cylinder version first. The final version that I envision uses all 5 of the pneumatic actuators that I already have, arranged radially like an old airplane engine, all acting on the same offset crank of a shaft. Shouldn't need a flywheel I think. I've also been brainstorming about making it variable stroke as well as variable timing. So the distance from the center of the offset crank to the center of the shaft would be adjustable (while running). The crank arm itself would have a cannel cut into it, through which the crank would slide. the sliding crank position could be adjusted by a lead screw turned by a servo motor. This would give it variable torque; crank far from the shaft - high torque, low speed, crank close to the shaft - high speed, low torqe (for a constant pressure). For a variable pressure it would give you the ability to to always run as efficiently as possible (minimize wasted gasses) at a constant speed and changing load (I think).
     
  15. shortbus

    AAC Fanatic!

    Sep 30, 2009
    4,003
    1,521
    There are a couple of things you maybe missing.:) From what I read, you seem to be thinking "double-action engine", with steam on both sides. Don't think that works on multi-cylinder engines. Not saying that it hasn't been done, just in all the reading I've not seen it.

    Your also looking at the engine like a internal combustion engine. By that I mean that it sounded like the piston was drawing the steam into the cylinder. It's not, the steam is pressurized, the valve is controlling it. The intake must close before BDC,but not by much. And that timing is always the same. The same with exhaust, it opens after BDC and stays open until the intake reopens.

    Unlike an IC engine there is no compression stroke. The power, compression and intake are all one stroke. The exhaust is the second. At least this is true on a single acting engine.

    Have you ever heard of Lindsay books? You really need to get one of their catalogs.
    http://www.lindsaybks.com/ The books are mostly reprints of old books, but they have information that is timeless. And a lot of what your interested in is represented. The books aren't that expensive and will give you some great ideas and in-site, I'm sure.

    You'll enjoy this- http://www.lindsaybks.com/gallery/Jorg/index.html
     
  16. strantor

    Thread Starter AAC Fanatic!

    Oct 3, 2010
    4,302
    1,988
    I'm not surprised that you haven't seen it. I haven't done any reading about steam engines, so this isn't based on any preexisting understanding of how they're supposed to work. Not studying previous designs - This is how people set themselves up for failure, but also how new things are discovered (not approaching the task from inside the box of what's been done already).

    I don't see why it wouldn't work to utilize both sides of the piston. My thinking is, that if it works on one side of the piston, should work on both (only less torque applied when pressurizing the rod end). And if it works on both sides with one cylinder, it should work on both sides with multiple cylinders. But I'm sure I'll find theres some caveat (hence why it's not been done already, that we know of).

    I didn't mean to make it sound as though the steam is being drawn into the cylinder. No, the steam is pressurized and is the force which pushes the cylinder. What I mean to describe is that in addition to pushing the cylinder, it must also push the gasses on the other side of the cylinder. The higher the pressure on the other side of the cylinder, the more resistance, and the less efficiency (working against itself). So that's why I propose venting the other side of the cylinder at some point before the changeover.
    I realize that in the traditional steam engines the timing is always the same. Also, every steam engine I've seen operates slowly in comparison with other types of engines. I suspect that the reason they are slow is because of this fixed timing; any attempt to go faster would result in the engine timing being retarded (and fighting itself) so a natural max speed is established. Keep in mind this is my own conjecture based on an admitted lack of knowledge so if I'm off base, let me know. I think that by gaining control over the timing, faster speeds can be realized
    I'm confused by the term stroke; When I use the term stroke, I use it to describe the piston action that correlates to 180° rotation of the crank, But I think others may use it to describe 360°. The way I propose to do it (double acting) every 180° is a simultaneuos power and exhaust stroke; so, "constant power stroke" if using 180 degress definition or "constant double power stroke" if using 360. with the exception of being within a couple degrees of TDC or BDC, the (single cylinder) engine would be outputting torque some torque through all the rest of the ~350 degrees. The torque would be highest when the crank is perpendicular to the actuator and decrease as it gets closer to TDC, then as it passes TDC begin to rise until perpendicular on the other side and then decrease approachin BDC and so on. This might almost negate the need for a flywheel. When you add just one more cylinder at a perpendicular angle to the first cylinder, I believe that it does in fact negate the need for a flywheel. I plan to use all 5, so as to get the torque output ass smooth as possible.
    I cam across a page that smelled like this the other day, but not sure if it was the same page. I had a look around there just now and looks very cool. Why are all these excellent references only found in the past? Why does nobody make books like this anymore? Why does nobody need books like this anymore? Why Does nobody go out into the garage and create things anymore?

    Well, I've requested their catalog so I'll be getting as many of those books as I can before I quit my job;)
     
    Last edited: Jun 9, 2012
  17. strantor

    Thread Starter AAC Fanatic!

    Oct 3, 2010
    4,302
    1,988
    FYI I found a place online you can get old books that are out of copyright. Sort of like the library of congress, online.
    I searched for "steam engine" and this is the first promising book I found: "Steam Engines, Theory & Practice" - By William Ripper, 1922 - 546 pages
     
    shortbus likes this.
  18. shortbus

    AAC Fanatic!

    Sep 30, 2009
    4,003
    1,521
    Don't understand really myself why people don't want to build things them selves today. It really is easier now the back in the day, with the better access to power tools.

    The thing with a double acting engine is the need of a "crosshead". make the engine so much bigger with that. http://en.wikipedia.org/wiki/Crosshead

    As to the variable engine stroke, the problem I see with that, is without also varying the cylinder length too, what will happen to the extra steam that can't get out of the cylinder? As the stroke changes the piston will not sweep the whole cylinder length and that steam will cool and condense. Condensed steam is water and water is non-compressible, so after a while somethings got to give/break.
     
    strantor likes this.
  19. strantor

    Thread Starter AAC Fanatic!

    Oct 3, 2010
    4,302
    1,988
    excellent point. I hadn't given that much thought. May have to back to the drawing board on that.
     
Loading...