Globe Valved IC Engine: Possible with Modern Materials?

[wolram]

The idea seems possible with modern materials, the globe is controlled by a
hydraulic stepper motor (which is also possible) according to engineers i have spoken to, the only way to find out if the whole idea is good is to make one,
now i am in the sticky stuff, it requires MONEY, should i carry on to a prototype? can anyone see a major pitfall in the idea?

 

[FredGarvin]

What would be the benefit of this design? It seems you are trading off having lifters vs an increase in complexity. Am I understanding your premise correctly?

 

[wolram]

What would be the benefit of this design? It seems you are trading off having lifters vs an increase in complexity. Am I understanding your premise correctly?

The (globe valve) would allow an un interupted path for gasses, the hydraulic control would allow maped variable valve timing, at the same time
doing away with cams,springs, chains , push rods, the down side is the energy to drive the hydraulic pump, which i think is a small price to pay.

 

[turbo]

You are going to need an encoder on the drive shaft to have real-time readings of rotational position and you'll need stepper motors to control the valves. There is also the question of flow rate. When conventional flanged valves open (especially with steep cams) they can get to maximum flow pretty quickly. Can you get better control over the flow rate with globes? One suggestion - instead of steppers, have you considered solenoids that can snap the globes from fully open to fully closed very quickly? That might cut down on the cost and complexity - at least for prototyping. I can envision advantages in performance and/or economy that might arise from using microprocessors to control the valve timing and duration.

 

[wolram]

You are going to need an encoder on the drive shaft to have real-time readings of rotational position and you'll need stepper motors to control the valves. There is also the question of flow rate. When conventional flanged valves open (especially with steep cams) they can get to maximum flow pretty quickly. Can you get better control over the flow rate with globes? One suggestion - instead of steppers, have you considered solenoids that can snap the globes from fully open to fully closed very quickly? That might cut down on the cost and complexity - at least for prototyping. I can envision advantages in performance and/or economy that might arise from using microprocessors to control the valve timing and duration.

The idea for the (globes) is that they require only a 90 degree rotation, and no wasted return stroke, and i am almost sure flow rates could be vastly improved with no interuption as in the poppet valve head, and yes the hydraulic stepper would be controlled via an encoder black box.

 

[brewnog]

Not to shoot down the idea, but conventional head designs are so well developed that valves don't really pose much of a flow restriction. Volumetric efficiencies well in excess of 100% are readily available, even on naturally aspirated engines, by the use of porting, multiple valve configurations, and tuned manifolds.

I'm not really familiar with these globe valves, are they capable of sealing well enough to withstand combustion pressures, and are they durable enough to withstand exhaust flow temperatures?

I think I mentioned before that the solenoids for a fully variable valve timing/lift system I've seen on a prototype engine were in excess of £18,000 each, not including the control system. So don't underestimate the moolah required for the venture! But keep me updated on how it goes...

 

[wolram]

Not to shoot down the idea, but conventional head designs are so well developed that valves don't really pose much of a flow restriction. Volumetric efficiencies well in excess of 100% are readily available, even on naturally aspirated engines, by the use of porting, multiple valve configurations, and tuned manifolds.

I'm not really familiar with these globe valves, are they capable of sealing well enough to withstand combustion pressures, and are they durable enough to withstand exhaust flow temperatures?

I think I mentioned before that the solenoids for a fully variable valve timing/lift system I've seen on a prototype engine were in excess of £18,000 each, not including the control system. So don't underestimate the moolah required for the venture! But keep me updated on how it goes...

The activation of the globe valve was a major concern, but i am lead to believe that a hydraulic stepper could be very cheap if mass produced,
and that the control electronics, are scalable to complexity, the globe valve would rely on modern materials and very close tolerances, but again
I am told that the requirements are possible, i wish it were otherwise then i would not be in this dilemma.

 

[Q_Goest]

Hi Wolram,

according to engineers i have spoken to, the only way to find out if the whole idea is good is to make one,

I'd have to disagree with this premise. There's no reason to have to build something like this in order to determine if there's any benefit to the idea. That's what engineering is for.

Valves of any kind are designed. Not only are all the flow paths easily calculable in terms of flow and pressure drop, but so are the dynamics, stresses in the parts, sealing ability, etc... There's nothing about a valve design that can't be analyzed, you don't need to make one to test it.

Also, you don't need any kind of FEA or CFD either. Those things may be of benefit for some types of more refined analysis, but they're not needed to get a good estimation of a design.

I'd suggest as a first cut, to model the valve flow and dynamics using an iterative model, say using a spread sheet, or making a computer program. There are some things you may need a bit of experience with, such as flow coefficients of various valve geometries, but those are fairly simple once you know them. Get a good book on valve design and go from there if you're really serious about this. Try "Lyons' Valve Designer's Handbook", you might find it in your library. Or look into other valve design handbooks.
I realize there are generally no classes in valve design in college, but like anything it is a matter of knowing how to model something mathematically. Just a suggestion any way.

 

[wolram]

Hi Wolram,

I'd have to disagree with this premise. There's no reason to have to build something like this in order to determine if there's any benefit to the idea. That's what engineering is for.

Valves of any kind are designed. Not only are all the flow paths easily calculable in terms of flow and pressure drop, but so are the dynamics, stresses in the parts, sealing ability, etc... There's nothing about a valve design that can't be analyzed, you don't need to make one to test it.

Also, you don't need any kind of FEA or CFD either. Those things may be of benefit for some types of more refined analysis, but they're not needed to get a good estimation of a design.

I'd suggest as a first cut, to model the valve flow and dynamics using an iterative model, say using a spread sheet, or making a computer program. There are some things you may need a bit of experience with, such as flow coefficients of various valve geometries, but those are fairly simple once you know them. Get a good book on valve design and go from there if you're really serious about this. Try "Lyons' Valve Designer's Handbook", you might find it in your library. Or look into other valve design handbooks.
I realize there are generally no classes in valve design in college, but like anything it is a matter of knowing how to model something mathematically. Just a suggestion any way.

And very welcome, thank you Q, the LVDH sounds very good.

 

[FredGarvin]

I see what you're driving at Wolram, but I only have more questions. Why globe valves in particular? Globe valves would increase the back pressure on the system and would, most likely, be tougher to seal. Perhaps the sealing part isn't too terribly important, I'm not sure on that part. Also, globes usually involve a change in flow direction of some kind which can play with the flow charcteristics. The Cv for a globe has to be less than an existing exhaust valves. Also, I think you are understimating the required complexity of the hydraulic system that would be needed. It would be a cool set up, but I would really question the benefits gained versus the complexity and guaranteed lesser robustness of the system.

 

[wolram]

I see what you're driving at Wolram, but I only have more questions. Why globe valves in particular? Globe valves would increase the back pressure on the system and would, most likely, be tougher to seal. Perhaps the sealing part isn't too terribly important, I'm not sure on that part. Also, globes usually involve a change in flow direction of some kind which can play with the flow charcteristics. The Cv for a globe has to be less than an existing exhaust valves. Also, I think you are understimating the required complexity of the hydraulic system that would be needed. It would be a cool set up, but I would really question the benefits gained versus the complexity and guaranteed lesser robustness of the system.

Fred the (globe valve) is simply a sphere with a hole running through it, if you looked into the inlet when the valve was open you would see the piston, a quarter rotation of the valve would close the port off.
I am in touch with an hydraulics engineer who has worked on high speed
hydraulic systems, he thinks a prototype can be rigged up with some mods
to off the shelf components.
All this would be a huge gamble, what with ceramic coatings having a critical role,and getting an electronic control system together

 

[FredGarvin]

Wolram, what you are describing is not a globe valve. That is a ball valve. A globe valve is a different beast entirely.

Globe valve: http://www.valvediagnostics.com/media/pictures/globe.gif

Ball valve: http://www.superklean.com/webphotos/Ball%20Valve%20Drawing.jpg

 

[brewnog]

Ahhhhhhhhhhhhhhhh! It becomes slightly more clear!

 

[Mech_Engineer]

A worry I would have for a ball-valve (not a globe valve) would be the fact that there is a lot of friction against moving the valve no matter where it is in its motion, and that the flow characteristicts for a partially open valve would be ugly to say the least.

Current mechanical valve designs' seals are inside the cylinder, meaning combustion pressures help you achieve a seal when it presses on the valve, where as with this design it would have to work against the pressure, increasing size and weight of the valve. What about electrically actuated valves based on current designs?

What is the point of trying to use a hydraulic system to open the valve instead of an electric system? It seems like additional complexity with no benefit.

 

[FredGarvin]

The flow characteristics could be helped by using a v-notch ball valve. I'm not sure how much flow turbulence and an uneven flow distribution have on the engine operation though.

I think the infinitely adjustable valve timing would be a benefit.

 

[Mech_Engineer]

The flow characteristics could be helped by using a v-notch ball valve. I'm not sure how much flow turbulence and an uneven flow distribution have on the engine operation though.

I think the infinitely adjustable valve timing would be a benefit.

The main concern is assuring an even mixture of fuel and air in the combustion chamber to assure even flame propagation through the cylinder. If the mixture is too heterogenous, it could result in incomplete combustion, or too slow a burn possibly causing backfiring.

The adjustable valve timing would indeed be a large benefit, many modern cars utilize variable valve timing to increase power over the powerband (sometimes seen as VVT or VVT-i). But, could a hydraulically-actuated ball valve open at the necessary time constant? 6000rpm seems pretty fast for a ball valve...

 

[Danger]

I thought briefly about a disk rather than ball valve set-up for this, but that led me to another idea. Since I really know next to nothing about valves, somebody's going to have to critique this pretty mercilessly.
What about a slider somewhat like a guillotine? Think of the neck-hole as the port, and the blade as the 'shutter'. Any feasibility to that approach?

 

[brewnog]

Danger, we discussed that idea a while ago. I seem to remember we decided that sealing was the major issue, let me see if I can find the thread...

 

[wolram]

Danger, we discussed that idea a while ago. I seem to remember we decided that sealing was the major issue, let me see if I can find the thread...

Trust me guys, i think i have read up every engine configuration ever invented, the good points about a ball valve (sorry Fred) is that the only sealing problem, other than the port, is the drive shaft, otherwise the valve is totally enclosed, i did uplink a url where 15000 rpm with poppet valves
lifting 6mm was in production, using hydraulic activation, the idea behind my
engine is, no cam shaft, chains , pushrods, etc ,and a mappable engine off the production line, i think as long as the engine tends to be more a long stroke, rather than a high reving engine, up to 10,000 rpm, this valve design
has more advantages than dis advantages.

 

[Danger]

Thanks, Brewski. It'll be quite embarrassing if it turns out that I'd participated in that discussion and then forgot about it.

 

[wolram]

Thanks, Brewski. It'll be quite embarrassing if it turns out that I'd participated in that discussion and then forgot about it.

Hey, Danger, partner up with me, i am sure you would make a good sales man/negotiator and would be handy around the workshop, no wages yet though.

 

[brewnog]

Danger, I think you did mate, I can't find that link though!

 

[Danger]

Hey, Danger, partner up with me

Too late; Brewski and I did that months ago, remember? (See 'Occupation' in my public profile.)
I hate to disappoint you, but I'm absolutely useless at sales or negotiating. It's rumoured that I've been known to turn a wrench now and then, though.

Brew, sealing could definitely be a problem, but I can't see it being any worse for a slider than for a ball. What I'm thinking of (just off the top of my head, of course) is a 'piston ring' built into the port, with the shutter being grooved down the edges to accommodate it. (Picture a rectangular yo-yo.) The fourth edge, where the actuator rod attaches to the slide, could have a much deeper groove so that it actually slides over the 'ring' during its whole stroke. The rod itself could be attached above or below the groove, and be O-ringed into its bore.

 

[NateTG]

Electronically controlled valves are an area of active research for several manufacturers, and there is a whole slew of fancy technology for variable valve movement on the market already.

The poppet valves (which are currently the popular technology) are quite similar to the globe valve. In general, the drawback to valves like the one in the illustration is that there is 'dead volume' between the valve and the cylinder that limits compression. Optimizing away that issue essentially leads to a poppet valve.

Alternative valve technologies have been used in engines. Sleeve valves were used in some engines in the first half of the 20th century with some success, but sealing was a problem.

Hobbyists have built engines using rotary and sleve valve engines that run without significant problems. I'm sure that with adequate effort, a gate valve can also be made to work.

 

[eaboujaoudeh]

i'd worry about backfire, uneven flow through the valve. just because some technology has advanced far doesn't mean its too late to try new ways. but I'm worried about backfire and even distribution inside the cylinder, cause there's no spraying control in such valves.