Archimedes screw throttle

[ronarprfct]

I may be blowing smoke on this one, but it occurred to me and I wanted to put it out there and find out the opinion of people much more knowledgeable. I have read that one of the things that limits engine efficiency at less than wide-open-throttle positions is the lack of efficiency of air moving past the throttle (plate, body? I don't remember the word for it for sure, just the thing that throttles the air intake). It occurred to me that something like an arhimedes screw whose rotation could be completely unhindered or even augmented via an electric motor for wide open throttle might be more efficient if you had to slow it down, ie hinder its rotation for less open throttle. Just free wheeling or with electric motor augmentation, I would think it would lose much less energy than the standard-type throttle. Any thoughts? I admit to having much less knowledge than some, but I am not going to stop thinking for myself just because of that, even if I occasionally look like an idiot. For all I know, this type of throttle has been implemented before or someone attempted to implement it before.

[Frank Lee]

A restriction is a restriction no matter what form it takes.

[MechEngVT]

I think Frank's right, it's just a different (more complicated and expensive) method to achieve the same result. With a single screw in a pipe it isn't even a positive-displacement device, so you would have air bleed (drawing air through the passage) and would not be able to super-charge. A twin-screw Lysholm supercharger would be able to be slowed down as a restriction or sped up to provide positive pressure, but the control system to do this as a primary engine speed control device would be VERY complicated (not to mention the safety aspects of a motor-driven device as the motor speed control?).

The best system I have heard of yet was BMW's Valvetronic, where electronic valve lift control was the "throttle" limiting the amount of air allowed into the cylinders. This was better because it kept port flow velocities high and allowed better mixing of the air and fuel at light engine loads.

[ronarprfct]

Originally Posted by Frank Lee

A restriction is a restriction no matter what form it takes.

Sorry, but that is like saying that, dragwise, a car is a car no matter what form it takes. If you have some actual data to support the idea that the air flowing over a partially closed plate and through a tube has less turbulence than the air flowing through a freewheeling or motor-augmented archimedean screw, then fine, but just putting forth such an oversimplification without justification is pointless.

[ronarprfct]

There are ways around the airbleed problem. I was just putting forth an idea, with the notion of discussing its efficacy if properly implemented, not discussing what would have to be done to properly implement it or the difficulties in implementing it. I am not trying to be offensive, I was just asking for an evaluation of the concept and perhaps further discussion, if necessary, of ways of improving upon the idea-assuming it has any validity at all. I'm just saying don't be so quick to throw out the bathwater because it is a little dirty unless you are certain there is no baby.

[dcb]

I think it is an intriguing idea myself. you have a pressure drop there with the throttle, it would be nice to put it to use

[Frank Lee]

Originally Posted by ronarprfct

Sorry, but that is like saying that, dragwise, a car is a car no matter what form it takes. If you have some actual data to support the idea that the air flowing over a partially closed plate and through a tube has less turbulence than the air flowing through a freewheeling or motor-augmented archimedean screw, then fine, but just putting forth such an oversimplification without justification is pointless.

No, it isn't.

Think of the point of the throttle... think of the meaning of the word "throttle". It is to restrict. It is to prevent a runaway engine; to regulate engine rpm and power.

Why is intake turbulence a problem?

[johnmyster]

That's no oversimplication. A pressure drop is a pressure drop. Kinda like "HHO is a Scam" isn't an oversimplification. It's the truth.

Now, if you use a turbine to capture the energy of that pressure drop, then you'd be onto something.

Turbulance schmurbulance. That's why the "tornado" air intake dohickey we all remembered works so well. A third year mechanical engineering student learns that it doesn't take much to disrupt laminar flow. You don't have to induce turbulance - it's already there, I promise.

[dcb]

I think Ron misread your post Frank. Ron, I think you misread Franks post

Frank just said a restriction is a restriction is a restriction.

I take that to mean that cylinder doesn't care as long as the mix is right. It could be a roots type restriction or a throttle plate. Throttle plate has been done Something else is going to take some head scratching, but if there is a bolt on for existing gas cars that can help close the efficiency gap with diesels, then it's definitely worth discussing.

[Vince-HX]

You guys should check out the honda r18 engine. Its drive by wire and under load load conditions it switches to a delayed intake valve timing. In this mode the throttle is opened much farther and gains of up to 16% were seen by minimizing pumping losses.

Only easy way to do it is drive by wire though

2006 Honda Civic 1.8 SOHC i-VTEC engine

[johnmyster]

That's the same "two intake settings" vtec setup that my '98 accord used. Mine used it for additional power/torque. They're using it to have a "low load efficiency" setting with drive by wire throttle control.

I thought the "i" in i-vtec stood for infinite. Looks like it stands for something else.

[dcb]

Hmm, it says they retard the intake timing and open the throttle more. Hmm...

I know some modders advance the valve timing to move the torque band lower. Hmm... I guess it would be a pretty efficient detuning to retard the intake cam. I might have to try it on the bike and see...

[tasdrouille]

With careful valve timing pumping losses can be reduced by 80% at low loads. No wonder recent engines have CVVT.

As far as the throttle plate in concerned, it's designed to restrict flow so I don't see the point of improving it's aerodynamics.

[metromizer]

There are many improvements made on the 'throttle blade' opening over the years, from 'rolling cylinder with an oval hole' on this Aurora indy car engine (shown), to some single and double guillitine types, an iris type, and double opposing blade design I've seen. A friend's sponsor applied for a patent on one he's used/using on his 4 cylinder in-line Ford USAC midget. They are mostly useful in some types of race engines, ones where part-throttle acceleration and low end torque for exiting a turn are useful. But the only engines I've seen them on are one throttle per cylinder (no plenum type intake tracts). I am thinking that a plenum is busy enough, load of turbulence but the ports and intake runners 'straighten out' the air flow after the throttle on a plenum design.

Still, I think about this stuff too. What if you had a 'revolver' type throttle body design. Where, like a revolver pistol, the correct venturi size was quickly rotated into position for optimum venturi diameter for a given engine load?

If you think about GM's XFi, where they used a slightly smaller throttle body than a standard Metro, there must be something to be gained (I assume GM wouldn't have bothered if there weren't mpg gains to be had for specifically the XFi modle of Metro)

enough of my rambling...

[Johnny Mullet]

I actually would have to side with Frank Lee on this one.

A restriction is a restriction.

[ronarprfct]

I believe intake turbulence is a problem because it increases pumping losses. I had read that the partially opened throttle caused inefficiency due to the work done sucking air into the engine. I would think lowered turbulence in the intake stream would lower pumping losses.

[johnmyster]

That's not what the term "pumping loss" refers to.

[dcb]

Just grabbing the first numbers I can find, a chevy 305 at idle:
assuming: 75 cubic feet of air per minute at 20 inches of vacuum

(75 cfm) = 2160 cubic inches per second

imagining a 1" square piston in a suitable "cylinder", it can lift 20 lbs, 2160 inches in one second. I figure that is about 6.54 horsepower in pumping losses at idle across the throttle plate. (1 hp = 550 foot*lbs/second)

[MechEngVT]

metro, I always thought a really cool throttle would be a sphincter. Lacking anything else to call it that's how I envision it...an idealized convergent-divergent nozzle (i.e. a venturi) that can smoothly and continuously change the area of the minimum opening to vary the mass flow through the nozzle. Haven't quite worked out the practical aspects of that yet.

Ron, I don't think your idea was taken as offensive. It's just an idea. We can discuss and debate it if you wish, but I *am* a mechanical engineer and in product development ideas only make it to fruition if there is a reasonable tradeoff of performance gained to effort expended on development. I think what folks have tried to establish is that there's no "100 mpg carburetor" and that any throttling device improvements are pulling off-the-wall concepts out chasing a gnat's-ass potential improvement. That's a recipe for "little to gain, lots of effort to gain it" and while there may in fact be some gain there (doubtful in this specific instance) it isn't worth chasing.

[johnmyster]

Originally Posted by dcb

imagining a 1" square piston in a suitable "cylinder", it can lift 20 lbs

20 inches of mercury isn't anything close to 20 psi. 20 inches is (ballpark) 2/3 of an atmosphere, so roughly 10 psi.