Probably in an auto your BSFC has to increase more than 10% to get over the difference between having the TC locked up and not, as it will most likely unlock under anything other than light acceleration. Also you probably need to stay between 2000 and 3000 in an auto, depending what your TC stall speed is, low RPM in an auto is like trying to paddle a canoe up a fast moving shallow stream, vs punting it with a stick.

On my way home from work I have to start at the bottom of a steep hill,flattens out,back down the other side and turn to the right. I accelerate briskly enough to reach 45mph and the top, continue accelerating on about half the flat part,go into neutral and coast halfway down the other side. Then I put it back into drive and let the car low to at least 37mph.Once I've made it around the turn without flipping over I accelerate again.
Putting the car back into drive slows the car enough not to flip and turns the motor so I've already got my RPMs up for the next hill.
you can go slow and still be reckless

That's a pretty good explanation there, Monroe.

Here's more discussion: https://www.gassavers.org/showthread.php?t=4455

It sure is! It's a challenge, to put it lightly.

Here's the link that Pale provided: https://www.autospeed.com/cms/A_110216/article.html

Look at the graphs.

After re-looking, I guess it depends on what chart you look at - the 3rd chart reflects what you are saying - low RPM (below 3,000) is good, regardless of throttle angle.

What I'm thinking about is the 2nd graph that shows SFC vs. RPM, and has 3 curves for engine load. It's pretty clear from that graph that the most fuel efficient operating condition, regardless of load, is between 3,000 and 3,500 RPM.

Are you sure you are using 70% throttle? That's a lot. If I'm using 1/2 throttle, the acceleration is very brisk, and I'm concerned about using "full" (or close to full) for the sake of FE. Maybe I'm worrying too much.

Anyways, as Pale said earlier, I'm probably overthinking things and need a ScanGauge II.

According to the Scangauge, I'm using 80% or higher throttle. My main target is to keep it just under the point where it adds extra fuel. You can see that on the SG by watching when it goes into Open Loop mode.

From my experience, the 4th, 5th & 6th graphs in the article match best with my engine. Yours may be different. You'll have to experiment and see.

disco: "Here's the link that Pale provided"

Thanks for reminding me where to find that article. I think it's very helpful.

"Are you sure you are using 70% throttle?"

I think there's a magic number, and I think 70 is probably about right, but I don't know if 70 is exactly right. I don't mean to imply that it is.

The issue is that at full throttle, the computer opens the injectors 100%. I think that's what's called open-loop mode. In the absence of a SG or similar instrument (like a DMM), it's hard to know exactly when this starts happening.

Open-loop mode wastes fuel. So the most efficient setting is somewhere slightly short of full throttle. And the right answer probably varies for different cars. What you want is the maximum throttle opening (because this reduces vacuum and therefore reduces pumping losses), just short of sending the ECU into open-loop mode.

"It's pretty clear from that graph that the most fuel efficient operating condition, regardless of load, is between 3,000 and 3,500 RPM."

I think you're misinterpreting the graph. The graph doesn't say that 3200 RPM is highly efficient "regardless of load." The graph says that at any given load, 2800-3200 RPM is highly efficient. Here's the key to understanding the graph: "load" basically means "throttle opening."

You are noticing, correctly, that given a load of 100%, that the engine is most efficient at 3200 RPM. This simply means that if the engine is running full throttle at 3200 RPM, it is more efficient than when it is running full throttle at, say, 1000 RPM. Fair enough. Here's what that means in terms of driving.

I'm climbing a steep hill. I'm in top gear. I have the throttle wide-open (for the moment, let's set aside the open-loop issue). The engine is at 1000 RPM. My speed is probably about 20 MPH. The speed is constant; i.e., I'm producing exactly enough power to fight gravity and travel up the hill.

Now imagine that I downshift to 3rd, or 2nd. RPM is now 3200. The throttle is still wide-open. Speed is still 20 MPH, and is still constant. I'm still producing exactly enough power to fight gravity and travel up the hill. But I'm not accelerating, because the hill is very steep.

The graph shows that in this circumstance, I am indeed better off in the lower gear. In other words, when heavy throttle in a low gear is not enough to raise my speed (assuming I am not already at the redline), maybe that means the hill is so steep (and/or the trailer is so heavy, and/or the headwind is so great) that I actually belong in that lower gear, and shouldn't upshift.

But this is an unusual circumstance. For almost all modern cars (even with low-power engines), if we assume a flat road and a vehicle carrying only one or two people, then full throttle (or 70%, if we want to take the open-loop issue into account) is enough to produce acceleration, even in top gear. Unless we're already at a high speed (like 90 mph), where air resistance is providing the equivalent of a steep grade.

In other words, if we're using the throttle to keep the engine at high efficiency, that probably means we're accelerating. That's the essence of P&G: acceleration, alternated with coasting.

Back to the graph. Yes, given a fully-open throttle, 3200 RPM is more efficient than 1000 RPM. But our cars do not stay at 1000 RPM, if the throttle is fully-open. Even in top gear. They accelerate. Unless you're climbing a grade, and then it indeed makes sense to downshift. But the important thing is to keep the throttle fully open (70%), except for when you're coasting. If I'm climbing a hill in 3rd gear, with full throttle, and this is not producing acceleration, then you're right, I probably shouldn't be in a higher gear, even if that would let me lower the revs.

"If I'm using 1/2 throttle, the acceleration is very brisk"

Life is much more complicated with an automatic. It's probably brisk because your tranny decided to downshift. That's what an auto trans usually does, when you give it a large throttle opening. A stick is much better, in this regard, because it lets you choose a large throttle opening without downshifting.

Theoretically, we want full throttle 100% of the time, except when we're coasting. I say 'theoretically' because we have to back off just enough to avoid open-loop mode. And with an automatic, you also have to back off just enough to avoid unwanted downshifting.

It is. I found another link to a very long paper that expounded on the graphs at the speed site. I'll dredge it up again, if interested ... but it basically says the same thing as the speed article.

Perhaps my choice of words wasn't so great, but the graph clearly shows that at the various load conditions (25%, 75%, 100%), the most efficient rev range is 3,000 to 3,500 or so.

I agree with you, though, so no argument.

Indeed. The very essence of P&G driving. I think we're in agreement.

I used to drive a stick, and am very conscious of what the auto trans is doing - thus my focus on both RPM and throttle angle.

I think the "brisk" part is b/c I've been driving with small throttle openings and low accelerations for so long that it seems very odd to be "flooring" it (not literally) and accelerating briskly.

This is where a CVT for a gasoline+spark engine comes into play... ...

Or should... unlike the shining example of the dodge caliber :rolleyes:

'95 Neon using venerable torqueflite hydraulic 3 speed transmission. 26mpg combined.
'05 porked out final year Neon, with electronic 4 speed ultradrive, 25mpg combined
'08 Caliber, CVT transmission, 24mpg combined.

On that evidence I should be grateful I've got the old 3 speed torqueflite technology auto in my minivan, and not go wishing for CVTs or a 4th gear.

Oh, I don't wish for one either. It reminds me of the SCCA F400 (?) class that used snowmobile CVTs. I can't say that I cared for those things at all.

dosco: "I found another link to a very long paper"

If you found that link on this site, I think maybe I saw it too. I vaguely remember some other similar paper, that I found via a link cited in this forum. If you found the link on another site, I think I'd like to see it, if you have it handy.

"the graph clearly shows that at the various load conditions (25%, 75%, 100%), the most efficient rev range is 3,000 to 3,500 or so"

Yes, the graph shows that high revs (about 3000) are more efficient than lower revs. But this is easily misunderstood, because the graph also shows that at any given engine speed, 100% throttle is always more efficient than 25% throttle. Following the graph, we would seek to be at 3200 RPM and 100% throttle. Trouble is, even a low-powered car is typically too powerful to maintain this state, unless it's climbing a steep hill.

The essential message of the graph is this: the engine is most efficient when the throttle is fully open. It's very counter-intuitive, but the key to saving gas is to open the throttle! (Almost fully, except for the issues I cited, about open-loop and about how an auto trans likes to downshift.) And then coast.

"I think the 'brisk' part is b/c I've been driving with small throttle openings and low accelerations for so long that it seems very odd to be 'flooring' it (not literally) and accelerating briskly."

Good point. I see what you mean. P&G is a challenge because it violates our old habits and expectations. It also has to be done skillfully, and with planning. I'm wasting gas if I accelerate and then have to turn that energy into brake heat, instead of coasting distance.

"This is where a CVT for a gasoline+spark engine comes into play"

Exactly. Grasping those BSFC charts makes it easier for me to understand why CVT is important. But I think we can see that it's possible to get good results without one, if the driver understands the principles we've been discussing.