Potential engine mod? Otto to Atkinson conversion

[wwest40]

"Atkinson cam is bunk.."

Well, since the actual Atkinson cycle engine does not make full use of the cylinder's intake "charge" capability it also sacrifices power in favor of efficiency. And remember that during the period the intake valve is left open is exactly in time with the least force capability (just leaving BDC) of the crankshaft acting to force the piston upward.

Also take note that the Atkinson cycle doesn't come up "on the step" except at something like 70% or more cylinder charge. An elongated power stroke does no good unless there is enough charge to be still producing power at or near the bottom of the power stroke.

An ideal system would leave the compression ratio at 13:1 for light cylinder charging, partial throttle, and only go into Atkinson cycle mode at 70% or more of charge.

And by the by, the Atkinson cycle is at its MOST efficient at or near WOT when there is little of NO vacuum in the intake manifold.

[Christ]

Originally Posted by wwest40

"Atkinson cam is bunk.."

Well, since the actual Atkinson cycle engine does not make full use of the cylinder's intake "charge" capability it also sacrifices power in favor of efficiency. And remember that during the period the intake valve is left open is exactly in time with the least force capability (just leaving BDC) of the crankshaft acting to force the piston upward.

No, it doesn't. This is the part that makes most people walk away from any discussion about Atkinson engines. It has the same VE during intake stroke as any Otto engine with the same attributes. Nothing has changed about the intake cycle of the Atkinson ENGINE. It fills the same amount of space with the same amount of air. The part where Atkinson is supposedly better is the second "crank". It effectively makes for a longer power-stroke. That's all it does. Each piston is making power for more time. Each rod has more leverage on the crank, but only when the rods are actually pushing on the crank. When the crank is pushing the rod back up, or pulling it back down, there is less leverage, so the crank isn't working as much, and less work is wasted doing the intake/compression strokes. You can't logically compare this to an Otto engine with an Atkinson cam.

Originally Posted by wwest40

Also take note that the Atkinson cycle doesn't come up "on the step" except at something like 70% or more cylinder charge. An elongated power stroke does no good unless there is enough charge to be still producing power at or near the bottom of the power stroke.

Source? I believe you're still mixing terms here. The Atkinson cycle engine is mechanically driven so that each power (combustion) stroke takes place on the longer stroke on the other side of an ellipsus on the crank shaft. Therefore, there is never a time when it's not in "Atkinson mode".

Originally Posted by wwest40

An ideal system would leave the compression ratio at 13:1 for light cylinder charging, partial throttle, and only go into Atkinson cycle mode at 70% or more of charge.

See above.

Originally Posted by wwest40

And by the by, the Atkinson cycle is at its MOST efficient at or near WOT when there is little of NO vacuum in the intake manifold.

Moot. Any engine is most efficient at lowest vacuum. It's called Volumetric Efficiency. This is the point of boost, initially. It increased the VE of airplane engines, so they could run better in higher atmosphere. It allowed thinner air to be compressed so that it would resemble air at or closer to sea level, and didnt' involve parasitic drag on the engine.

Using an Atkinson-style cam is no different. The point of what I said was simply to give light to the fact that the limited reversion from the cylinder isn't enough to defeat the inherent vacuum in the manifold. The fact is, even if there is LITTLE vacuum in the manifold, there is still SOME vacuum in the manifold. Since there is SOME vacuum in the manifold, there is still a high pressure spot outside the TB that keeps positive flow. This means that the atkinson cycle can work. The limit to this is when cylinder reversion causes an influx of pressure into the manifold. This prevents the pressure differential on either side of the TB plate, and allows airflow reversion from the TB. Boost solves this, hence the Miller Cycle.[/QUOTE]

[MechEngVT]

Originally Posted by wwest40

Look at it this way...

Using the current implementation of the Atkinson cycle, delayed intake valve closing and 13:1 "native" compression ratio, look at what would happen with just a single cyclinder engine. You would have reverse airflow OUT the intake path/duct for each compression stroke, not exactly viable for a MAF/IAT equipped engine.

I meant to address this in my reply yesterday since IIRC it came up earlier in the thread...but I think christ answers correctly.

This WOULD cause a problem with either an incorrectly designed intake or if you had a single throttle plate per cylinder with no "manifold" per se. In the first situation a proper chamber volume or resonator would buffer the pressure pulses and allow a continuous inflow through the throttle (and therefore the MAF sensor) the way that normal intakes already do this to a lesser degree. In the second situation (one butterfly/cylinder with no resonator) you don't even need a MAF sensor and can map fuel/spark from TPS and RPM.

[MechEngVT]

Originally Posted by Christ

The idea of the "Atkinson cam" is bunk, essentially. It may increase FE, but it sacrifices power to do it, which is exactly the opposite of the true Atkinson design.

That's one way to look at it, but I think there's another perspective that would question whether it is indeed opposite of a "true" Atkinson.

The Atkinson-style cam is intended to simulate the operating parameters of the Atkinson cycle. The Atkinson cycle is fundamentally a "normal" compression ratio with an "enhanced" expansion ratio built into an otherwise Otto cycle engine.

In the "true Atkinson" engine this is done with a really trick crankshaft and/or connecting rod journals that lets the engine have a mechanical compression ratio of say 10:1 and a mechanical expansion ratio (i.e. "power stroke") of say 13:1. Since the swept volume will be less on the 10:1 intake stroke you have a smaller displacement engine than if you were to look at the swept volume of the power stroke. Personally I would rate the engine as the displacement indicated by intake stroke swept volume. In the "Atkinson-cam" engine of the same mechanical size with a 13:1 compression/expansion ratio the swept volume is larger than of the "true Atkinson" engine. Since intake charge is bled off during compression in the "cammer" then its specific power output is less than a "true Atkinson" but it is very likely to produce equal power to the "smaller true Atkinson" that has a lower volume.

So one may say that the "true Atkinson" doesn't sacrifice power the way the "atkinson cam" does, but the end result is essentially the same because the "true" engine is a smaller displacement not capable of producing as much power while the "cam" engine produces the same power because it's slightly less displacement-efficient. All in all, the two lumps o' iron (or Aluminum) are the same package size and weight, produce essentially the same usable power, and the cam engine is easier to build and control.

What's bunk about that?

[wwest40]

Where to start...

Okay, take a 2L displacement Otto engine and convert it to the TRUE Atkinson cycle engine and its EFFECTIVE displacement now becomes something less, as a ratio between the shortened intake/compression stroke and the power/exhaust stroke, than 2L. Giving up the POWER of a 2L OTTO engine infavor of the efficiency, longer "burn" cycle of the Atkinson engine.

The current implementation does exactly the same thing, sacrificing POWER for FE.

Pumping losses.

When the piston is at BDC the travel direction of the crank is at 90 degrees to the angle required to move the piston upward in a compression cycle. Delaying the closing of the intake valve during this early period in the compression stroke results in lowering the pumping losses since the travel angle of the crank will be closer to linear for piston travel once compression actual begins.

An ideal engine would be able to vary the compression ratio as a function of cylinder A/F mixture charge level. An engine with a fixed compression ratio, say 10:1, is not very effective at/with low charge levels. SAAB is currently testing an engine that can vary the compression ratio as a function of charge level.

My point with the Atkinson cycle is: what is the use of an elongated burn cycle if the charge level was so low that the "burn" was complete at 2/3 piston travel..?? So it would be better to not delay the closing of the intake valve for partial throttle and thereby make more efficient use of the native compression ratio of 13:1.

What drives the tubine in a turbocharged engine...??

Partial or moderate throttle = No excess energy left over to be exhausted into the turbine, NO turbine power.

Think you're ever see a turbocharged Atkinson cycle engine..??

[wwest40]

Originally Posted by MechEngVT

I meant to address this in my reply yesterday since IIRC it came up earlier in the thread...but I think christ answers correctly.

This WOULD cause a problem with either an incorrectly designed intake or if you had a single throttle plate per cylinder with no "manifold" per se. In the first situation a proper chamber volume or resonator would buffer the pressure pulses and allow a continuous inflow through the throttle (and therefore the MAF sensor) the way that normal intakes already do this to a lesser degree. In the second situation (one butterfly/cylinder with no resonator) you don't even need a MAF sensor and can map fuel/spark from TPS and RPM.

"..buffer the pressure pulses...."

At WOT how do you "buffer" the pressure pulses in a way that prevents a pressure wave from travelling back up the intake path...??

With a 4 cylinder engine you will have ACTIVE SUCKING of the "opposite" cylinder to neutralize the pressure wave....

[Christ]

Originally Posted by wwest40

"..buffer the pressure pulses...."

At WOT how do you "buffer" the pressure pulses in a way that prevents a pressure wave from travelling back up the intake path...??

With a 4 cylinder engine you will have ACTIVE SUCKING of the "opposite" cylinder to neutralize the pressure wave....

You have the same thing with a 6 or 8 cylinder engine. Even 5 cylinder engines, and 10 cylinder engines will always have a cylinder on intake at the same time that one or more cylinders is on power/exhaust/compression. The only exception to the rule is the single cylinder engine, and any variant of rotary engine.

The only difference is the speed and timing at which the next piston draws back the unused (expelled) intake mixture.

2 cylinder engine, pistons are working 180 from each other (normally)
3 cylinder engine, pistons are working 120 from each other (normally)
4 cylinder engine, pistons are working 90 from each other (normally)
5 cylinder engine, pistons are working 72 from each other (normally)
6 cylinder engine, pistons are working 60 from each other (normally)
8 cylinder engine, pistons are working 45 from each other (normally)
10 cylinder engine, pistons are working 36 from each other (normally)
12 cylinder engine, pistons are working 30 from each other (normally)

In every iteration of a standard balance multi piston engine, each cylinder is working opposite another one such a way that the expelled intake mixture would be recovered by another cylinder.

MechEngVT - What I meant by bunk was more like "Wasteful". I mean this because the cam causes outflow of intake mixture back into the exhaust manifold. This causes issues with intake harmonics, etc, and screws up ideal mixtures in just about any gasoline engine. What's to say that the mixture pushed out wasn't the richest part of a layered homogenous mixture? Now the mixture in the cylinder is lean! Subsequently, the engine has had another pulse of air through the induction system, which in carb'd apps already has gas in it, or in FI apps has fuel added as it flows into the cylinder, but either way - the new cylinder's mixture is ideal, but it's sucking in the richest part of the last cylinder, so now it's rich! Now you have a lean cylinder, and a rich cylinder. Who knows what part of the rich cylinder will end up in the outflow from the second piston's compression event, and so on/so forth. See what I mean? In the Atkinson Cycle engine, this isn't a concern, as there is no outflow.

Ideally, to use the Atkinson cam in place of the Atkinson engine, the outflow would either have to be precisely controlled, or fuel precisely metered AFTER the ouflow event. This is where GDI comes into play. Gasoline Direct Injection engines are similar to diesel engines in that they can directly inject the fuel into the compressed air already in the cylinder. Therefore, the Atkinson cam's outflow event would only expel unnecessary air, and the mixture could be leaned by the GDI computer (ECU) to account for the outflow from the Atkinson cam, so that each cylinder could be individually fed it's own fuel supply without an outflow of fuel from the previous cylinder to suck in.

In that sense, the Atkinson cam is less complex, and easier to create/manufacture, and wins. In any other type of engine, I don't see it being a very reliable system. (Apparently, neither do many auto mfgrs.)

wwest40 - Turbocharger exhaust turbines don't require exhaust gasses to expand to reach full spool. They only require the flow of exhaust gasses. Regardless of the expansion ratio being too large or too small for the engine, the expansion of the exhaust gasses being complete, etc... there will still be exhaust flow and heat energy (not expansion, heat... the exhaust flow will still be hot.), so the turbo will still spool and create the feedback loop that turbochargers create in order to reach maximum speed.

Guess what happens if you have a turbo running off an engine's exhaust, but feeding it's compressed air to something other than the engine that it's leeching exhaust gas pressure from? The turbine will never reach full speed, and the compressor will never reach max potential boost. Turbochargers require a feedback loop to reach their full potential, except in extreme circumstances. (Obviously, if you put a T3/T4 hybrid on a submarine engine, it's going to hit 100,000+ RPMs without the feedback loop.) So I guess I should restate to include the obvious - a turbocharger sized properly for the engine will not fully spool without the feedback loop it creates.

That essentially means that your theory on why you don't see Atkinson Cycle or Atkinson Cammed engines with turbos is a myth, and holds no water in logical thinking or applied science.

If you don't believe that turbos don't require expansion of gasses (as opposed to flow) to operate, put a vacuum cleaner nozzle on one some day. I bet you a dollar it starts to spool up.

[wwest40]

And you think an the high volume exhaust flow, enough to drive a turbocharger into overboost, isn't energy that could be better/different used...???

[Christ]

The volume of exhaust flow isn't any different than the Otto engine w/ the same initial displacement as the intake cycle of the Atkinson engine. The volume of gasses is the same.

To answer your question: No, the energy from pulsing exhaust gasses isn't better used than to create more power with less spent energy. The idea of turbocharging has been openly displayed in earlier posts. Given that the actual volume of gas exiting both engines is the same (for comparative purposes), the only difference is that the Atkinson engine has already spent more heat from the expansion of gasses than the Otto engine has. That said, the Atkinson engine already has both a performance and an efficiency advantage over the Otto engine. Being that turbochargers generally exaggerate common efficiency of engines, the Atkinson only stands to benefit from the turbo by a ratio linear with it's advantage over the Otto engine, based on the second "stroke" length of the Atkinson engine in reference as a percentage to the first, i.e. 115% stroke means the Atkinson has a displayed advantage of 15% over the Otto.

What this means for practical application - While an Otto cycle engine has a marginal advantage in complexity and cost of manufacture, the Atkinson engine still proves to be numerically more efficient in an exponential fashion when compared to the Otto. Turbochargers, as suspected, compound the VE advantage that would normally be seen in the Atkinson engine, as do they compound the VE of either engine upon installation.

Unless you can come up with something that hasn't already been discussed, this thread will have reached the end of it's useful life, and I will no longer entertain the same argument without basis for it's existence.

[MechEngVT]

wwest40:

I don't think any engine's combustion event has fizzled out before BDC on the power stroke, by which time the exhaust valve has already opened. Even at partial throttle there is enough air/fuel mixture in a cylinder for very substantial expansion as the fuel vapor is combusted. Expanding the combustion cycle in the Atkinson engine doesn't "waste" movement/momentum on the elongated power stroke, it merely extracts more energy from the expanding gases.

And you don't just compress it to a "native" 13:1 because of detonation. ESPECIALLY at partial throttle. While it is full-load spark knock that will kill an engine in a heartbeat, it is part-throttle pinging that will kill an engine with 10,000 paper cuts. Until we get to high anti-knock index fuels (methanol) you'll be stuck at 11-12:1 max with naturally aspirated Otto cycles.

Christ is right re: turbos' operation, only the reason that you'll never see a turbocharged Atkinson engine is that a "turbocharged Atkinson" is called a Miller engine.

Buffering pressure pulses: ever heard of the analogy between intake air flow and acoustics? The intake "manifold" is somewhat of a muffler on the intake side of the engine. On ALL engines, not just Atkinson, there is unsteady intermittent flow going into each cylinder and when the air mass in the intake ports flows it gains momentum. When the intake valve closes the momentum of the port flow stops and sends a pressure wave reverting up the port opposite the "flow" direction. These pressure pulses leave the intake port into the plenum, which is essentially an open volume common to all (or a good number of) the intake ports and is fed by the throttle body or carb. Since not all reversion pressure waves occur at the same time and each of them is small in magnitude relative to the volume or mass of air in the plenum the plenum pressure is more consistent than the port pressure. Therefore, the plenum acts as a resonant chamber or "buffer volume" to help equalize the effects of the unsteady flow between the cylinders. In the Atkinson engine there is actually a small amount of reversion mass flow instead of just inertial pressure. This reversion flow is small relative to the inlet charge and therefore small relative to the port volume. Port volume is typically small relative to plenum volume, so in this respect both the intake port and plenum act as "buffers."

Christ:

I think you're reading too much into the intake reversion flow. Contrary to popular opinion port fuel injection systems typically inject fuel into the intake port onto a CLOSED intake valve (to give more time to inject fuel and to ensure full vaporization). The head of combustion/exhaust of the preceding cycle starts to vaporize the liquid fuel droplets and the turbulence of the intake event homogenizes the mixture. In a homogeneous mixture there is no "rich" or "lean" part as that would indicate it is non-homogeneous. Since fuel is injected toward the valve the "rich" part of the inlet charge would be drawn in first and turbulence would mix it well or if not the rich portion would swirl near the piston during the intake stroke. With the Atkinson-cam it is the last portion of inlet charge that is expelled. This would either be fully mixed near stoichiometric or worst-case be the "lean" portion of the charge. Since the expelled volume is small relative to the intake port most of this air should remain near the valve of the same cylinder to be re-used during the next intake event in that cylinder. Over multiple cycles the closed-loop ECU control system would learn from the O2 sensor how much fuel needs to be injected PER CYCLE which would already account for the expelled mixture from the previous cycle. Engines and electronic control systems are pretty robust in this regard; they can run sub-optimally pretty well and can be adjusted incrementally.