Potential engine mod? Otto to Atkinson conversion
 

So this one hit me as I was looking around wikipedia last night. I seem to recall reading somewhere a long time ago that the CRX HF engine ran on a different "cycle" than a normal engine, partially accounting for it's increased fuel efficiency.

A few minutes of searching turned up the Otto cycle (the "normal" cycle) and the Atkinson cycle (the high FE cycle).

It looks to me as if the only appreciable difference is in the lift duration of the intake cam. If it was possible to have a new cam ground to provide this increased intake duration, you could effectively convert your engine to the Atkinson cycle, thereby raising your FE.

So the question I'm posing for all of you engine gurus out there is this:
Would this be a feasible mod for all us folks out here with Otto cycle engines?

Good idea -- I wonder what driveability is like: is there reduced power (especially torque)? I wonder if a learning curve exists in driving with this type of engine. Since it's found in some hybrid applications, the electric motor makes-up for the "getting started" part.

I'm curious if a custom cam would achieve the result...

But, I see that the original Atkinson engine had a different stroke on the intake and exhaust cycles...

RH77

I've been pondering the same thing for awhile now....

If I had the money - I'd have a custom ground camshaft machined.... As I recall, that's the key difference between an Otto and the Atkinson in several hybrid vehicles on the road at th moment...

The concern I have is backfiring... So stock setup is to spray fuel on the intake. But with a cam that leaves the cylinder open - some of that fuel fortified intake will be pushed back. Of course, under normal conditions - some fuel is left in the intake track - so I imagine it's up to the valve seals to ensure there's no backfiring...

I'm also curious as to how the manifold vacuum will change :)

I honestly think it's possible.... Just someone needs to fork up the funds for the cam (unless we can find someone that'll do it for free or does it for a living :p). If no one does before I'm out of college (and earning some dollars), I'm probably going to give it a shot :p

Yeah, about the intake charge being pushed back into the intake manifold...
I can't imagine that being a huge issue. I mean, cars with using the Atkinson cycle don't (usually) have direct injection, so they are using the same limits we are. For example, only spraying fuel when the valves are open.

Still, I'm having a hard time convincing myself that the air sensor(s) are only going to count the air that's passing through one way (into the engine). I can't help but think it'll end up enriching the mix by "re-counting" the air that has been ejected back from the engine, and re-injecting the fuel for said amount of air.

Or, that air could just be all compressed and stored in the manifold somehow (TB closed), the MAF could be too slow, etc.

I know this is a long-shot (but we're brainstorming here)...

Is any of this worthy of Grant Writing? Call me crazy, but someone might pony up a few bucks for the sake of engine research.

Any word on the emissions profile as well?

RH77 :turtle:

So I've done a bit of thought on that... Actually, that's what most of my thoughts focused on as that's arguably one of the key sensors for fuel trim that will be affected.... The conclusion I came to, for a MAF sensor, is that it will not be a problem.

So MAF sensors operate VIA the use of a heater wire (typically). Bulk flow over the wire cools it in a predictable manner. As long as the manifold vacuum is negative - there will always be positive (into engine) bulk flow. The degree of which depends on engine speed and pressure.

In my thought experiment - pressure is what will change. Rather than the typical 22" Hg (or so) - I think there will be slightly higher pressure which in turn results in less bulk flow. WHICH makes sense as we're rejecting a portion of flow per intake stroke.

That doesn't sound like a bad idea actually.... I'll stay by what I had said before about my timetable (school for me, right now, takes up all my time - despite that it also takes all my money too :p). But someone else should totally go for it if they have the means :)

No clue - that's one thing I don't even know where to begin on how to think about it :p Intuitively, I imagine that the engine management will compensate for the lower airflow (MAF) in addition to normal a/f metering (O2 sensor). I would think that exhaust temperature change would stay nominal BUT I'm not sure. Given the lower compression and longer power stroke (compared to compression stroke) - the rejection cycle might have lower temperatures (better for NOx?). Exhaust content has been a voodoo type subject for me... I personally would build it, then measure :thumbup:

Turtle
http://gallery.photo.net/photo/6065333-md.jpg

So, there may be problems....
Can't seem to find any difinitive word on the HF's particular cycle (Atkinson/Otto)

I too have been thinking about this off and on, but haven't done a ton of research on it yet. From what I've read though, the Atkinson engine is simply less power dense. I believe (although I'm far from an expert on this) that the main difference between the two engines is cam timing (probably added duration too). The intake valve closes late to allow reversion of the air/fuel back into the intake manifold. This reduces the amount of air/fuel in the chamber, but also allows higher compression ratios and that helps compensate for the power loss. The higher compression is where you get the boost in effeciency. So, it has less power, but is more effecient. If you can live with less power, your good. How much less power? I guess it all depends on how you design the system. The Prius uses a modified 1NZ-FE (11:1 CR) engine that normally puts out about 106 hp. Once the Prius engine is done being modified into the 1NZ-FXE (13:1 CR) it has 76 hp. So, you loose roughly 25%. I'm sure having an engine with VVT and some way to control this would make things infinitely easier to play with. However, adjustable cam timing pulleys would be a cheaper alternative.

Cvt
 

With detonation and torque issues, I'm wondering if a CVT transmission would be the best option. I know, cart before the horse...

Reason-being -- from driving the Prius and noting the engine speed, it seems to turn faster, and in bursts. It begs the question of the Atkins' ability to have enough power in the standing-start RPM range in either a manual or TC auto.

Variable timing may be the key, as mentioned.

BTW, Does the HF have VTEC? That might achieve the Atkinson-like operation...

RH77

Nope, HF was pre-vtec. It was a SOHC 8-valve job too.

When I bought my HF it had a cracked valve at 75k (about 1/4 of the valve face was gone:eek:) makes me wonder about the detonation thing...

And I can't help but wonder if a little low rpm detonation would be too much for an engine to cope with. They've all got knock sensors now, so the ECU just dials back the timing, and you've got a little less power to take off with.

I think the Honda D-series (B's to a lesser extent) may be good candidates, they're all undersquare, so good for torque. They have cheap parts, so easy to get the custom grind and bump up compression. And they have all of the electronics that you'd need. Plus, they're cheap and plentiful, so you could easily replace it if you broke anything :D

Oh, and the way I understood it, the Atkinson gets its efficency from retaining more of its kinetic energy from the compression stroke since it is compressing less volume than an Otto cycle engine. Sure the nomenclature remains (compression ratios) but the meaning is changes when you're not actually compressing 30% of that volume.