Please log in.
Before you can vote, you need to register. Please log in or create an account.
Vehicle: Car: Engine: Valves
Rotary Valve System   (+5, -2)  [vote for, against]
for engine aspiration

This idea is partly inspired by [rasberry re-tart]’s Dual-Purpose Valves idea and [acurafan07]’s Dual Purpose Valves idea and the annotations in them, and also by the Coates Patented Rotary Valve System. Etc.

The idea is to have one large rotary valve per cylinder, which rotates in the same axis as the crankshaft, at half the speed (or an average of half the speed). In the computer model I made, the valve is the same diameter as the cylinder, and there is a square air passage through the valve which curves 135 degrees (I believe). Each of the two openings would be 1/6 of the valves circumference.

Since the valve is in the center of the head, the spark plug(s) will have to be off to the side(s).

The valve opens the intake port directly to the exhaust port during part of each rotation; this means that if the engine has multiple cylinders then there is a chance that the exhaust from one cylinder could go through another cylinder’s valve into the intake manifold which could then go back into a cylinder yada yada… So that means it would probably need a way to stop that from happening, which could be: a one way valve in the exhaust port, or a supercharger, or maybe a turbocharger, or maybe just an exhaust system that inhibits the flow of gases back into the manifold, or possibly a combination of some of these.
-- BJS, Feb 02 2007

Cutaway Image of the Valve http://www.putfile....pic.php?img=4812022
It explains a lot. [BJS, Feb 02 2007, last modified Feb 19 2007]

Coates Patented Rotary Valve System http://www.coatesen...taryValveSystem.htm
[BJS, Feb 02 2007]

[SublimeGTP]'s idea Simple_20Variable_20Valve_20Timing
It would work out great in this idea. [acurafan07, Feb 21 2007]

Rotary Valve System 2 Rotary_20Valve_20System_202
Basically the same idea, but with more performance. [BJS, Mar 10 2007, last modified May 15 2007]

An engine animation http://en.wikipedia.../wiki/Engine#Modern
which shows an intake valve which is larger than the exhaust valve. [BJS, Mar 10 2007]

adjustable movie camera shutter http://en.wikipedia...dak_Spezial_II).JPG
movie camera adjustable shutter http://en.wikipedia.org/wiki/File:Verstellbarer_Umlaufverschlu%C3%9F_(Cin%C3%A9-Kodak_Spezial_II).JPG [Birdlives, Oct 09 2013]

One valve per cylinder no matter what the design doesn't seem to be a good thing. I'm pretty sure that this would negatively affect the velocity of the intake air. In the 8th paragraph of the link, they explain about intake velocity and why 1-valve per cylinder wouldn't be good. Although since this type of valve would be different, I'm not sure it would apply here.
-- acurafan07, Feb 02 2007


I don't know if I trust that poster's knowledge.

I think most of the reason the 2 valve engine has less low-end torque than the 4 valve engine is because the engine has to move more mass linearly, which puts more stress on the engine creating less output torque.

My idea wouldn't have that problem because it only rotates mass instead of moving it linearly.

If that really is a problem, would a turbocharger solve it?
-- BJS, Feb 02 2007


Besides the inevitable flow issues, I think the bigger problem is going to be friction. Rotary valve require a large bearing surface to seal properly(say area of the port times 1.5 or something) because the entire opening needs to be covered and have sufficient surface area to form a seal, also since the area to be sealed is so large the pressure will be lower over the sealing surface to allow for sealing. In normal valves the seal is made at very high pressure over a very small area. In your design every valve will have its own bearing surface and then the gear train will also need bearing surfaces. In an OHC design there are really only 4 or 5 bearings on the cam and the small area of the valve stem.

I think you would also have issues with wear as the valve timing will shift due to gear lash, but it may not be consistent over all the cylinders, in a Cam engine the timing could change to but will change equally for all cylinders. You also need to address how you are going to Lubricate the Valve as the seal is a sliding seal as apposed to a press seal the sealing surface and the bearing surface are one in the same. normal valves are not this way which prolongs life of the engine.

What do you see as the advantage in this type of a system?

*EDIT* I missunderstood what I was reading here so I retract some of my commentary here.
-- jhomrighaus, Feb 02 2007


Does it really need as many bearings as you say it does?

How important is it to have tight seals?
-- BJS, Feb 02 2007


Actually it could have no bearings, but it still has a surface that rubs against another surface(a Bearing Surface)

Tight seals are vitaly important to the operation of the engine, if pressure is allowed to leak out through a seal then it can not do work on the piston and thus you lose power(a lot of power). This is what people are talking about when they discuss a "Compression Test" on an engine.
-- jhomrighaus, Feb 02 2007


Ok, I take back what I said about why 1 valve per cylinder wouldn't be a good thing. But if I understand correctly, you propose using a supercharger to flush the exhaust gases out and prevent them from flowing into the intake of other cylinders. Wouldn't this result in a loss of boost if it is scavanging the manifold? Also, what would be the advantage of this over the Coates system you linked to? Oh, and your first link doesn't work.
-- acurafan07, Feb 19 2007


I believe you understand me correctly. I don't remember what scavenging means for engines (could you explain it?). The advantage over the Coates system is more air flow volume since it is one large valve. I fixed the link.
-- BJS, Feb 19 2007


Scavenging is when the compressed air is used to blow the gases out rather than providing boost for the engine. The problem with that is that it will always have to be in use, resulting in low to no added pressure to the cylinders while still costing HP. Your idea sounds very possible and creative, but like I said if I understand the part with the supercharger correctly, you would sitll have that problem. You'd be better off to have a conventional valve system and harness the supercharger's boost that to waste it and use this.
-- acurafan07, Feb 19 2007


Rotary valves would eliminate the possibility of variable valve timing, a technique commonly used to increase efficiency or power as conditions demand.
-- nuclear hobo, Feb 20 2007


No they wouldn't.
-- BJS, Feb 21 2007


Please explain how a single rotary valve can provide variable intake and exhaust openings.
-- nuclear hobo, Feb 21 2007


Well, you can't make make the valve openings any larger without making the entire valve larger, but you can change the duration of the valve being open by making the valve spin more slowly or stop while it is in the open position (and then more quickly while it it in the closed position to compensate). You can also change the timing by offsetting the valve's angle of rotation a few degrees in relation to the crankshaft.
-- BJS, Feb 21 2007


I like [SublimeGTP]'s idea for simple variable valve timing, and think that it would work perfectly in this system since it doesn't involve using camshafts (which, on this engine, wouldn't exist).
-- acurafan07, Feb 21 2007


For [SublimeGTP]'s idea to work for his application, the "pulley" would need to have a 4:1 gear ratio to the crank shaft.

For it to work in my application, I think the worm gear would also have to turn the differential casing back and forth every revolution of the valve shaft, so that the exhaust would be open long enough.
-- BJS, Feb 21 2007


What kind of mechanism would one use to change the duration and timing of a single rotary valve? This seems to be the crux of the problem, though assuming it would work, heat dissipation would be a serious problem, especially keeping fuel from igniting upon contact with the hot exhaust valve as it rolls around to beome an intake.
-- nuclear hobo, Feb 21 2007


//This seems to be the crux of the problem, though assuming it would work, heat dissipation would be a serious problem, especially keeping fuel from igniting upon contact with the hot exhaust valve as it rolls around to beome an intake.//

Many old pushrod engines were designed so that the intake and exhaust ports and valves were on the same side (meaning that intake gases were heated from the exhaust gases) and those never had any problems with pre-detonation.
-- acurafan07, Feb 21 2007


"heat dissipation would be a serious problem, especially keeping fuel from igniting upon contact with the hot exhaust valve..." It could have direct injection, couldn't it?
-- BJS, Feb 21 2007


Yeah, as opposed to fuel being injected through the hot valve during the air intake stroke!? Mwahahahaha....Although, that would be a halfbaked idea in itself at it would increase fuel diffusion into the air and thereby giving it a more combustable mixture than with direct injection.

The problem would be that a rapidly changing hot/cold valve wouldn't last very long unless it were made of some kind of material that can withstand rapid changes in temperature. (cold intake~hot exhaust cycles)
-- quantum_flux, Feb 23 2007


Actually, rotary engines us clearances of 4/1,000ths of an inch to eliminate the need for lubrication but maintain a seal that doesn't allow blow by. This idea is half decent, but the fact that it opens up makes it kind of useless unless it can be cured, a two way valve would work, but it's overly complicated. The idea is to keep it simple. Also, the exhaust valves are ALWAYS larger than the intake valves, which you can see in any engine. I have another idea that works all of this out in a way that is timed the same as a dohc engine is now with the same cam gears, slightly different idea on the valves, but same general principle. I worked on it for 3 months while I was in prison once. It's quite possible.
-- MarcStinebaugh, Mar 09 2007


MarcStinebaugh, I think you mean a one way valve. So you think 2 moving parts compared to more than 10 is "overly complicated"? I don't know where you heard that "exhaust valves are ALWAYS larger than the intake valves", but on wikipedia there is an .gif animation that shows otherwise (which I've provided a link to).
-- BJS, Mar 10 2007


I think in fact the opposite is true, the viscosity of the intake air generally dictates a larger intake valve area as it is easier to push out the hot exhaust gases than it is to pull in the cold intake gases. There are some 5 valve engines out there that have 3 intake and 2 exhaust valves per cylinder.
-- jhomrighaus, Mar 10 2007


And some 3 valve engines with 2 intake and 1 exhaust.
-- acurafan07, Mar 10 2007


Why did no one tell me that the geometry of my 3D model (shown in the first link) would not work at a constant speed for a four stroke engine?
-- BJS, Oct 25 2007


WOW, am I the right guy for this! I ran Coates for about 9 months so ask me anything. The idea is actually his son's, he stole it. As many of the submitters surmised, one of the big issues is the sealing of the valve and the friction of the seals. Coates' solution was, I worked there MANY years ago, to use some of the hardest materials known to man, which would then battle it out to see which would destroy the other. The problem with your design is a slow moving valve won't open and close fast enough. Also you don't want to have intake and exhaust running thru the same passage as it one glowing hot piece of soot would cause the intake charge to ignite. (Bad) The other problem with the idea in general is the surface area of the valve that is exposed to the combustion heat. So if you want to do this make the valves out of a slippery ceramic.
-- MisterQED, Oct 26 2007


Believe it or not this exact idea is already in the process of being baked by a guy in Cairns (Queensland, Australia). He based it on the Family II block (Astra / Barina / Camira / Vectra / Commodore / Corsa / Calibra and so on).

He previously need 20ft/lb of torque on the old engine cam to open the valves. On the rotary valve head a cordless drill was able to spin the 'cam' at high speed (at a gues, something that equated to about 4000 engine RPM).

The engine runs, but only just, due to valve timing issues.

The concept of altering the valve opening length is easy to achieve with planetary gears. If you fix the centre gear, drive the middle gears, and have something driven from posts on the outer edges of the middle gears, you will note that the driven item distinctly slows down (almost to a stop), then accellerates quickly to the next dead spot. Much like a spirograph.

Ideally you would have two valves per cylinder so you can vary the timing and phasing of the valves (incedentally the reason that most engines are twin cam now is so that they can do this to pass emissions and keep their balls attached)

Early attempts at rotary valve engines used oil for sealing. At the cost of emissions (which no-one cared about) and oil useage.
-- BLSTIC, May 22 2008


What you have described is basically a Crossley valve engine from the 1920's (I think). These have continued in developement up to the present day. One engine designer even built a V10 version for F1, which was then promptly banned.

The failure of this device it more a case of maintaining the status quo than any inherent problems with the idea. Its very difficult to go against 100 years of development of an orignally poor idea (the poppet valve).
-- eight_nine_tortoise, Mar 05 2009


I'm not sure I would classify the poppet valve as a poor idea considering it is still almost universally used by all engine makers and is simple, durable and reliable.
-- jhomrighaus, Mar 05 2009


The problem is poppet valves have several advantages, such as longer life, less heat loss and simpler sealing.

I couldn't find any Crossley rotary valve engines only slide valve engines which are different animals.
-- MisterQED, Mar 05 2009


Hey consider the variable shutter in film movie cameras as a way to adjust the length of intake impulses.
-- Birdlives, Oct 09 2013



random, halfbakery