Move over granny, BSFC is coming through!

[#SaveTheManuals]

I was trying to think of an inflammatory title for this thread to lure people in, but I got lazy and settled for moderately humorous instead.

To date (i.e. my entire driving life until last week) I've only employed granny-techniques in my quest for better MPG. Except when I had a Miata...then I just drove like a maniac 100% of the time and had an amazing time wasting fuel. This forum introduced me to BSFC, and the idea appealed to me immediately -- not only because it makes sense, but also because it sounds like a lot more fun to drive that way.

For the final 1/5th or so of my last tank, I accelerated fairly hard in the 2020 Corolla SE (2.0L 6-speed MT) to get up to speed. Not pedal-to-the-metal, but pretty spirited. I was able to slowly increase the MPG on that tank toward the end, so I think I'll abandon the granny-style for a while and see where BSFC takes me.

I don't have the BSFC map for my car's engine, but Toyota published a chart that looks suspiciously similar:



I think this confirms my SOTP feel that < 2,000 is a little low for peak efficiency; in 4th gear at moderately high load it feels like it's almost on the verge of lugging at 2,000 RPM. It looks like the sweet spot is about 2,700 RPM (presumably under moderately heavy load?). I'm really not sure why everything falls off a cliff at ~3,000 RPM and the chart ends at 3,200 RPM. It looks like incomplete data to me. Peak torque on this engine is at 4,400 RPM.

Is there anything to be gleaned from the chart to help me figure out optimal RPM for shift points? My main concern is when to shift from 3rd to 4th. It seems like I need to shift in 3rd later (close to 4,000 RPM, probably non-optimal efficiency) in order to get into 4th significantly above 2,000 RPM so there's no lugging at high load.

[Daschicken]

2000 RPM should be well above lugging. It may not be efficient and smooth acceleration, but it ain't lugging. If you don't have a scangauge yet, or any other OBD 2 gauge, it would be great to get one. The scangauge at least has a engine load feature which shows percent load. It is generally accepted to accelerate in the 70-80% range to maintain efficiency without going into fuel enrichment. It is pretty easy to get outside of this range without a gauge.

In regards to your name, you should be happy to hear that I have taught eighteen people to drive stick so far. Some of which actively expressed interest in getting a manual car.

[Ecky]

I can't say why that chart ends at 3200rpm, but I can talk a little about why BSFC peaks where it does.

BSFC is better at higher loads, because the ratio of work to parasitic losses are better. At a quarter load, your friction and accessory losses will be nearly the same as they are at WOT - things like friction, turning the alternator, water and oil pumps. Additionally, as you open the throttle plate, there is less vacuum generated, which is itself a loss. So, at higher loads, a larger percent of the fuel burned does something useful (move the car) because many of the losses in an engine are relatively fixed.

In some engines, BSFC tends to drop off above ~75-85% load. I imagine this is commonly because of either 1) many engines tend to run rich at high load, because the extra fuel cools the combustion chamber which promotes longevity and reduces the chance of knock, and 2) many engines run more retarded (later) ignition timing, to reduce cylinder pressure and prevent knock, at the expense of sending some of the combustion energy unused out through the exhaust.

As for RPM, there are a few factors that come mind.

Friction goes up exponentially with RPM. So, you'd think that the lowest RPM possible would result in the best BSFC.

When spark ignites the contents of a cylinder, it takes some time for flame to propagate and for cylinder pressures to build. Nearly all engines fire the spark plug and ignite the contents of the cylinder while the piston is still rising on the compression stroke, meaning that for a short time, part of the energy from combustion is actually trying to push the engine backwards and doing "negative" work. If your piston is moving very slowly, you can fire the spark plug later (more advanced) and do less negative work.

A piston moving too slowly, however, will have hot combustion gases in the chamber for a longer period of time, allowing for a greater amount of energy to be lost through the cylinder walls and piston head as heat, rather than doing useful work.

There are a lot of other factors which have to do with the geometry of the engine and rate the piston is moving relative to flame front. There are also factors of the quality of combustion - e.g. an engine with a valvetrain designed for high volumes of air and high RPM operation may have poor mixing and poor combustion quality at low RPM.

There are almost certainly other factors I'm forgetting.

The original engine in my car peaked around 2000RPM. Many seem to peak in the 2500-3000rpm range. It really depends on the size and geometry of the engine.

[#SaveTheManuals]

Originally Posted by Daschicken

2000 RPM should be well above lugging. It may not be efficient and smooth acceleration, but it ain't lugging.

You're right -- it's not really lugging. I left a lot of RPM wiggle room with "almost on the verge of lugging". I guess what I meant was that the car's acceleration was noticeably less enthusiastic in 4th gear around 2,000 RPM than in 3rd or lower, with about the same amount of throttle.

Good point about a ScanGauge. Something like that will at least get me in the right ballpark for load; could be that I'm over- or under-shooting the intended load.

Of course I'll still be flying blind without the BSFC map, unless the torque/RPM/MPG map published by Toyota is (accurate and) more or less equivalent. In that case, per my annotations below, it sure looks like the most efficient region is between 2,200 - 3,100 RPM, with load between ~76% - ~95%. The sweet spot is 2,700 RPM at 88% load:



76% - 95% load is higher than I expected based on the general consensus around the forums, but I suppose there are always outliers. At some point I'd like to verify or debunk whether I can really drive my car around at 90% load for peak efficiency. That seems a little like having my cake and eating it too. Therefore I'm skeptical of this chart.

In regards to your name, you should be happy to hear that I have taught eighteen people to drive stick so far. Some of which actively expressed interest in getting a manual car.

nice work! My kids are just a few years away from learning to drive, and I look forward to teaching them how to drive stick on the Corolla.

[#SaveTheManuals]

Originally Posted by Ecky

I can't say why that chart ends at 3200rpm, but I can talk a little about why BSFC peaks where it does. ...

Great stuff Ecky -- thank you for sharing! Some of this is new to me and I'm slowly absorbing it. But I'm getting the gist and I appreciate the level of detail you shared.

In some engines, BSFC tends to drop off above ~75-85% load. I imagine this is commonly because of either 1) many engines tend to run rich at high load, because the extra fuel cools the combustion chamber which promotes longevity and reduces the chance of knock, and 2) many engines run more retarded (later) ignition timing, to reduce cylinder pressure and prevent knock, at the expense of sending some of the combustion energy unused out through the exhaust.

This is particularly of interest to me. I was reading a bit about the M20A-FKS engine in my car, and I came across an ECU-tuning site that did some analysis on it. In their words:

Engine runs a paltry... 10 Degrees of ignition advance at peak horsepower and 5* at peak torque, and manages 168HP. Not to mention runs rich (like most toyota engines) at full throttle.

So if running rich and retarding timing under heavy load tends to cause a drop-off in BSFC, that doesn't seem to mesh with the chart above that shows the highest efficiency between ~76%-95% load. Could be the chart is misleading, could be the ECU-tuning people are mistaken, or could be something else entirely going on. I'm a little out of my element here, just trying to connect the dots.

[Taylor95]

I try to drive a little bit faster when possible, because my engine is pretty inefficient at low rpms. I get the best gas mileage by far when I'm around 50 mph for flat roads and about 60 mph for going up hills. I believe my engine reaches peak torque around 2000-2500 rpms. At 60 mph, it runs just below 2000.

[gumby79]

I am noticing this phenomenon a faster is more fuel efficient with my new transmission and Allison 6-speed.I went from a 2.44 final (od gear 0.69x axel raito3.55) to 2.16finar(0.61) puts me at 1608 @70 and 1263@55. Peak tq is 1846rpm

[MeteorGray]

Yeah, my six-speed Mazda3 automatic has a final overdrive ratio of 0.60, and when combined with the diff ratio of 3.59, it achieves an overall final drive ratio of 2.15, which is quite a benefit on long highway trips.

I like those tall legs on long highway trips.

[slowmover]

Originally Posted by gumby79

I am noticing this phenomenon a faster is more fuel efficient with my new transmission and Allison 6-speed.I went from a 2.44 final (od gear 0.69x axel raito3.55) to 2.16finar(0.61) puts me at 1608 @70 and 1263@55. Peak tq is 1846rpm

The aero wall is at 60-mph. The very few who can prove higher bets better FE, are rare indeed.

Gearing & tire size are a “problem” you’ve created with that trans swap (as you know).

Will be interesting to see what you come up with.

Was any baseline 60-mph testing done?

.

[ever_green]

Depends on the car. For example, all the new BMW engines do not have a throttle body. instead they use continuous valve lift. I know this is only like 5% of the cars out there, but every car is different. what I have found with all my new cars is that one must be gentle with the throttle. anytime you attack the throttle or give it too much it either goes into positive boost pressure (turbos) and/or enriches the mixture (turbos and NA). Modern cars run especially rich to avoid damage on cheap 87 octane gas in America. On the other hand, electronic wastegate turbos with direct injection almost eliminates pumping losses (even without continuous valve lift). There is never any vacuum even under slightest throttle depression on modern turbo engines. At the same time spending too much time accelerating counteracts fuel savings you could have realized coasting or cruising. with modern cars I would just flow with the traffic acceleration wise to target speed and then do pulse and glide. With older traditional cars I would just get up to speed as soon as possible. If an average journey is 30min, you probably spend only 5 min of it accelerating and 25min cruising/coasting (so focus on the latter).