School Me: WAI vs CAI

[serialk11r]

Originally Posted by ChazInMT

OK, so I've read through some articles (find below) and educated myself on CVVL. I see it is mentioned that there is potential to eliminate the throttle, but, I don't see where anyone has done so completely yet.
..
The main take away I'm getting from all this is that by closing the inlet valve before the piston reaches BDC (er... Bottom Dead Center) you can control the "charge" that way and thus eliminate pumping loss. It appears a small amount of "loss" is good in that it creates a healthy swirl to promote better combustion which off sets the loss and then some.

Valvetronic, Valvematic, and VVEL are capable of 100% throttle independent output control. The throttle is backup only. They can cut duration to under 20% maximum duration, which is enough to idle. Engines that do so need pumps and motors to drive traditionally vacuum boosted things.

The pumping loss that early intake valve closure induces is not so bad because the pressure drop happens at the valve, and the air at low pressure is contained completely within the cylinder and so a greater percentage of the initial power consumption drawing air past the restriction is recovered on the compression stroke. The bigger downside is probably that on the intake stroke as the charge is expanding, it absorbs more heat from the cylinder walls and thus incurs some more pumping loss, but I am not sure how big of a deal this is.

Joeggernaut, throttle by wire is not actually very different from cable throttle. Most engines are not equipped with any kind of variable lift system, and the only variable valve timing is cam phasing. The advantage that electronic throttle control allows is for the cam phasing system to work together with the throttle, rather than responding to a throttle input (and lagging behind, throwing away some of the efficiency benefits). However the amount of control that intake cam phasing has over volumetric efficiency is not nearly as good as Valvetronic. The range of positions for cam movement has been increasing, but retarding the opening of the intake much past TDC starts to incur significant pumping loss. As long as there is significant vacuum to deal with a warm air intake has the potential to improve fuel economy because throttling losses tend to be quite serious.

[Daox]

I read most of the articles Chaz posted. I don't think we can say variable lift technology 'virtually eliminates pumping losses'. It sounds like they've just moved the restriction from the throttle body to the intake valve. Something still has to control how much air gets in, and they're still using the intake valve as a restriction. Its definitely an improvement, but I just don't think we can say pumping losses are virtually eliminated. I think reduced is a much better word.

The only exception I can see is if they're using something similar to Toyota's atkinson-like cycle (used on the Prius/Camry hybrid) where the intake charge is partically exhausted out of the intake valve.

[serialk11r]

Originally Posted by Daox

I read most of the articles Chaz posted. I don't think we can say variable lift technology 'virtually eliminates pumping losses'. It sounds like they've just moved the restriction from the throttle body to the intake valve. Something still has to control how much air gets in, and they're still using the intake valve as a restriction. Its definitely an improvement, but I just don't think we can say pumping losses are virtually eliminated. I think reduced is a much better word.

The only exception I can see is if they're using something similar to Toyota's atkinson-like cycle (used on the Prius/Camry hybrid) where the intake charge is partically exhausted out of the intake valve.

Say you had a variable lift system that controlled the rocker ratio rather than using a secondary rocker to vary lift and duration together. That would essentially be giving the engine independent throttle bodies at the valves.

This would still reduce pumping loss because 100% of the air under vacuum is contained within the cylinder and will thus return more energy to the system upon compression. With independent throttle bodies, the throttle is leaking air through all the time vs. only during the intake stroke, so it doesn't have this benefit since the throttle valve must be closed further to reduce pressure to compensate for the "leakage". With the typical plenum and single throttle body setup, you essentially just have a consistent reduced pressure atmosphere after the plenum, and a quick PV diagram lets you see why it doesn't recover as much energy on compression.

The "best" solution is considered to be fully variable duration late intake valve closure since it has the least pumping loss, but it is the most difficult to implement. Early intake valve closure is easy because the systems that reduce lift basically "waste" a portion of the cam lift, all you need for that is an extra rocker that can engage at varying "heights" from the cam's center, but the extra valvetrain mass increases friction, especially at higher engine speed (which is why BMW does not use Valvetronic on the S65 and S85).

Luckily, if we're willing to give up low end torque, cam phasing alone with Atkinson cycle (aka high duration intake cams) can get us pretty low throttling losses. Say you have a 285 degree duration intake cam. If you retard the intake valve opening point to say 10 degrees after TDC (the higher the rod/stroke ratio the better to reduce pumping losses here), the intake valve closes at 295 degrees after BDC. This means you're bleeding off ~60% of the intake charge! Thus to idle you only need to have maybe 50kPa vacuum instead of 80, and when cruising on the highway you only need 20kPa or less vacuum.

[Joeggernaut]

Originally Posted by serialk11r

Say you had a variable lift system that controlled the rocker ratio rather than using a secondary rocker to vary lift and duration together. That would essentially be giving the engine independent throttle bodies at the valves.

This would still reduce pumping loss because 100% of the air under vacuum is contained within the cylinder and will thus return more energy to the system upon compression. With independent throttle bodies, the throttle is leaking air through all the time vs. only during the intake stroke, so it doesn't have this benefit since the throttle valve must be closed further to reduce pressure to compensate for the "leakage". With the typical plenum and single throttle body setup, you essentially just have a consistent reduced pressure atmosphere after the plenum, and a quick PV diagram lets you see why it doesn't recover as much energy on compression.

The "best" solution is considered to be fully variable duration late intake valve closure since it has the least pumping loss, but it is the most difficult to implement. Early intake valve closure is easy because the systems that reduce lift basically "waste" a portion of the cam lift, all you need for that is an extra rocker that can engage at varying "heights" from the cam's center, but the extra valvetrain mass increases friction, especially at higher engine speed (which is why BMW does not use Valvetronic on the S65 and S85).

Nissan and Toyota do use higher rpm engines though. BMW systems are groundbreaking but they are never simple like the Japanese who later improve upon their ideas.


What do you think about electromagnetic controlled valves (magnetic actuators)? I think it is feasible and could completely eliminate the mechanical linkage (camshaft/springs/etc). The ECU would only then have to control the electric pulses. The only downside I see is what happens if the engine loses electric power then you might run into bent valves as they might not return to their previous position. I guess you could use capacitors to store enough energy to limp the car to a complete stop while still having valve movement.

EDIT: Here is a cool little article that someone combined sources from into one answer: http://wiki.answers.com/Q/Working_of...valve_actuator

Then all we would need is magnetic bearings for the crankshaft and we could almost eliminate the need for oil except for piston lubrication.

[serialk11r]

Originally Posted by Joeggernaut

Nissan and Toyota do use higher rpm engines though. BMW systems are groundbreaking but they are never simple like the Japanese who later improve upon their ideas.


What do you think about electromagnetic controlled valves (magnetic actuators)? I think it is feasible and could completely eliminate the mechanical linkage (camshaft/springs/etc). The ECU would only then have to control the electric pulses. The only downside I see is what happens if the engine loses electric power then you might run into bent valves as they might not return to their previous position. I guess you could use capacitors to store enough energy to limp the car to a complete stop while still having valve movement.

EDIT: Here is a cool little article that someone combined sources from into one answer: Working of camless engine with electromechanical valve actuator

Then all we would need is magnetic bearings for the crankshaft and we could almost eliminate the need for oil except for piston lubrication.

Not really. Nissan VQ37s hit 7500rpm but they seem to have very high oil temperature when run at high speed (and their specific torque isn't all that great at the top end) suggesting there is some serious friction going on. No Toyota Valvematic engine runs over 7000rpm.

Camless valve actuation sounds nifty but yes, bent valves when power is lost is a problem.

[Joeggernaut]

Originally Posted by serialk11r

Not really. Nissan VQ37s hit 7500rpm but they seem to have very high oil temperature when run at high speed (and their specific torque isn't all that great at the top end) suggesting there is some serious friction going on. No Toyota Valvematic engine runs over 7000rpm.

Camless valve actuation sounds nifty but yes, bent valves when power is lost is a problem.

7500prm is no joke on a 3.7L v6. That is pretty high but I get what you mean. Too-bad toyota doesn't use the valvematic design and make it run like the old vvti-L. Those little celica gt-s pulled pretty good once lift engaged.

Most applications of this design seem to be targeted a slightly smoother torque curve and fuel consumption instead of performance output. Wonder what the long-term liability is like with these cvvl technologies.

[ChazInMT]

Regarding pumping loss, I think it does practically eliminate it, here's why. I hope the logic I have can be followed, cause once I figured it out, I was like, Oh yeah.

1st, keep in mind a deactivated cylinder has no pumping loss, if the valves don't open, the air in the cylinder just acts like a spring and returns whatever energy it took to compress it back on the next downstroke. We can agree on this. If you think otherwise, read this.

2nd, following this idea, if you only let in enough air to operate at 30% power, then opening the valve till the cylinder is half full, then shutting it prior to when it would normally be, you in essence only pull in the air you need, you then essentially draw a vacuum on the cylinder at the bottom half of the intake stroke, and when you compress the half cylinder of air, you have the same result.

The advantage is, you pull in air for only half the stroke when the piston goes from zero to full speed then you quit drawing by closing the valve.....does this make sense, the valve is open when the air velocity is lowest, makes it to where the velocity is about to be at its max, and closes, so there is only a short time that the air is trying to squeeze past a restrictive opening, and this short time is actually very beneficial because it creates a lot of swirling to aid combustion.

Another advantage is the lower cam load, by only working half the time, the cam doesn't need as much power to turn it.

So to paraphrase, the valve at low load is only open at the top part of the stroke when the air velocity is low, so low loss, then for a brief moment there is a high loss/swirl producing intake, then you're back to no pumping loss because the valve is closed, and whatever pressure differential occurs during the bottom half of the stroke will not be a factor.

In my mind I see this as being a way to vary the displacement of the engine without cutting off cylinders.

Hope this didn't further confuse and If I'm missing something I'd sure love to talk more about it cause I'm learning too.

[serialk11r]

Originally Posted by Joeggernaut

7500prm is no joke on a 3.7L v6. That is pretty high but I get what you mean. Too-bad toyota doesn't use the valvematic design and make it run like the old vvti-L. Those little celica gt-s pulled pretty good once lift engaged.

Most applications of this design seem to be targeted a slightly smoother torque curve and fuel consumption instead of performance output. Wonder what the long-term liability is like with these cvvl technologies.

Actually I think the performance is "at its best", because they can utilize a higher maximum lift and duration at the top end with zero low end issues. The Valvematic ZR engines have more power than the non-Valvematic engines.

I think it's useful to think of it as similar to a pushrod valve engine, except the pushrods are lighter and adjustable length. The added mass puts more stress on the cams and requires stiffer springs, and creates more friction. You can rev them up, but the cams can suffer more wear (compared to an OHC engine).