Hacker News Reader

Comment by i_am_jl | Hacker News Reader

>more fuel means less compression is tolerated before knocking

Generally, no.

Knocking happens from pre-detonation, that's usually caused by heat from compression causing the fuel/air mix to ignite before it's triggered by spark.

To avoid this engines will run a fuel/air mix that is not stoicheometrically ideal, to make the mixture less likely to ignite early.

It is safer to run a fuel rich mix than to run a fuel lean mix as it keeps combustion chamber pressures low (unburned fuel takes heat out of the exhaust). It is more economical and more ecologically friendly to run fuel lean since you paid for that unburned fuel and it's kinda gross.

In general, more fuel than ideal means more resistance to knock. But these things are complex.

EDIT: Knocking happens from pre-detonation. Knocking can also happen from predestination, like in the case of turbocharged Subarus.

View on HN

Friendly correction for others because your auto correct failed you...

"predestination", pretty sure parent meant "pre-detonation"

A.k.a. Autoignition aka "it goes boom before you planned on it"

Only adding this as it's a pretty crucial word for understanding the comment.

That fuel is totally depraved, and has no hope of salvation.

This is why TULIP oil makes a great fuel additive.

I totally, unconditionally, agree. My own limits do not account for the irresistibility of this combination. I hope you persevere with this message.

Ah you're right, thank you!

>Knocking happens from pre-detonation, that's usually caused by heat from compression causing the fuel/air mix to ignite before it's triggered by spark.

No, it usually happens because the normal flamefront from the spark causes a rise in pressure that triggers compression-ignition in other parts of the cylinder. It's not solely from the compression, usually. That scenario is rare primarily because as you reduce the knock margin, you'd hit knock from what I said before you get to the state where it's so bad it ignites from compression alone.

https://www.researchgate.net/figure/In-cylinder-pressure-tra...

Look at this picture; this is a typical waveform of cylinder pressure vs. crank angle. The spark happens 28 degrees before TDC, so basically the left edge of each of the graphs. As the flamefront consumes the air-fuel mixture inside the cylinder after the spark, the cylinder pressure gradually rises. During knock events, the cylinder pressure as risen by the normal combustion process gets to a point where it starts igniting the fuel elsewhere in the cylinder, away from the gradually expanding flamefront. This causes rapid combustion which causes the pressure to rise suddenly, which causes damage to the engine (if severe enough)

> EDIT: Knocking happens from pre-detonation. Knocking can also happen from predestination, like in the case of turbocharged Subarus.

Edit of the week IMO.

> caused by heat from compression causing the fuel/air mix to ignite before it's triggered by spark.

I would think the usual cause is the timing was off and the spark plug fires before the cylinder reaches TDC.

This edit is brilliant!

Ah yes, read that engine knock article the wrong way around, guess my mind was drawing on intuition from things like gun powder where more stuff crammed more tightly together is worse.

So then, if they're doing a non-ideal initial burn it would have to be a lean lower-compression burn, followed up by a higher-compression secondary burn?

Pretty sure it's just a way to get more expansion from the same air charge.

It's a similar idea to the Atkinson cycle. You have dissimilar compression and expansion strokes. In normal engines, there's a limit to compression ratio because if it's too high, it causes knocking. But a bigger expansion ratio lets you extract more energy out of the combusted gas, which leads to higher efficiency.

The original Atkinson cycle idea was to use some complex linkage to get dissimilar compression and expansion strokes, but the way it's implemented in things like the Prius is to have a high compression engine, but mess with the intake valve timing such that you only use a small part of that compression during the intake phase so you effectively handicap your compression ratio to avoid knock, while still retaining the full stroke during the expansion phase.

I am sure that I don't understand the specifics of what Porsche is doing, but the article says the cycle is intake-compression-power-compression-power-exhaust. For that to be the case they have to be burning the same fuel and air mixture twice, and I don't understand the chemistry there enough to even speculate how it works.

I think it's safe to assume that the second stroke is burning incomplete combustion products left over from the first stroke. I think that the second compression stroke would have to be higher compression than the first in order to get more complete combustion of what's left behind.

That actually makes a lot of sense merging with some comments from above.

Second stroke goes lower and there are some ports to add air which are not accessed during the initial compression stroke...so the stroke is longer (higher compression) and more air is added to help with the reburn.

Kinda sounds like combining the idea of the Miller cycle with a variable compression/stroke setup (see Nissan).

There are a lot of ideas out there that create gains individually...glad to see them being combined more and more in modern engines. (ex. VRIS, VVT, DFI,) I personally think there is still another few decades of playing around with ICE to be done...not sure it will be viable for the market...but the research will lead to a lot more interesting engineering.

Ah, extra ports, that makes a lot more sense. I had missed that detail.

So the first phase is like a regular 4-stroke engine, and the second phase is more like 2-stroke engine, where extra air (and possibly fuel) is introduced into the cylinder, like a 2-stroke, through ports located below the position of the piston during the bottom dead center of the first phase.

So I guess you have something like intake (high), compression (high), power (high->low), compression (low), exhaust (low->high), where in parenthesis is the adjustable piston height?

So a richer higher-compression first phase, followed by a leaner lower-compression second phase?

Finally got time to read a bit of the patent, which talks about two TDCs and BDCs, which I'll denote as "high" and "low", as related to the distance of the piston to the crankshaft. So the high TDC has maximum compression, and low BDC has the maximum cylinder volume, and the scavenging port(s) are between high BDC and low BDC.

Unless I screwed up my notation shift, the strokes from the patent are as follows:

1. Intake (low TDC -> high BDC)

2. Compression (high BDC -> high TDC)

3. Power (high TDC -> low BDC)

4. Compression (low BDC -> high TDC)

5. Power (high TDC -> high BDC)

6. Exhaust (high BDC -> low BDC)

So during the first power stroke, stroke 3, the cylinder moves from "high" to "low" and thus is the longer power stroke. Also during the second compression stroke, stroke 4, the cylinder position moves from "low" to "high". So technically leading to higher compression ratio. I was thinking it would cost too much energy to do so, hence dismissed that alternative, but I guess not.

The patent also notes that the extra scavenging ports are not needed, fresh air-fuel mixture can be introduced via the inlet valve(s) while the piston moves between the two BDCs.

Would be fun to try to simulate it in Ange's engine simulator[1].

[1]: https://github.com/ange-yaghi/engine-sim

I would imagine that pulling the air from the bottom would help more with scavenging the cylinder than the top but yeah...I can imagine it wouldn't matter.

That all makes sense. Pretty much what I assumed. The compression stroke at #4 is the longest stroke. Then it also has the benefit of the Power stroke at #5 being shorter which will mean it will have the benefit of a short(er) stroke motor on the power (more torque?) during that cycle.

The shorter stroke on #1 is desirable since you are using boost (I am assuming Miller cycle here) to control the beginning of compression as opposed to the intake valve.

Once the mechanical timing is down I assume it isn't too bad to keep all in line though...but a mess if it gets out.

I suspect the second stroke is higher compression. First stroke has the compression "shape" controlled by the boost (miller cycle).

With direct injection they could even be injecting more fuel into the cylinder for the second stroke...but I suspect the second stroke F/A ratio is determined by the first stroke remnants combined with the extra air allowed in at the bottom of the second stroke.

All of this with the cam variators, timing control, boost control and fuel setup that VW already runs would be fairly easy to control with the proper sensors and code.

Just speculating at this point...but it makes sense to me.

> Miller cycle

Heh all along I'm wondering, what kind of thermodynamic cycle is it? By six strokes, does it mean there are actually distinct new phases to the PV graphs compared to Atkinson/Carnot/Miller?

Or is it just masquerading a well-known cycle underneath six strokes, only some parts are being optimised?

Not sure myself...

Seems to be like it is a Miller cycle (or could be) on the initial 4 strokes. Which would allow you to control the timing of the "Compression max point" in the stroke by varying the boost. That may also vary the amount of spent fuel remaining for the second "scavenge" stroke...which if I am reading comments above correctly it pulls air from ports lower in the cylinder which would help clear the cylinder for the next 4 stroke cycle.

Seems to me more focused on reburning/scavenging to make a "cleaner" burn than anything else though.

*Not an engineer...just a shadetree mechanic who reads too deep into engine papers.

Reply on news.ycombinator.com