thefoos

I would suggest that folks that are interested in this sort of thing get a scangauge or datalog fuel consumption. Maintaining a constant speed (45 mph) on flat ground, us FXTer's will note that you get better mileage in 4th instead of 5th.

__________________
Scott
04 MT FXT STG 2 5 spd
02 WRX STG 2 5 spd

etiennelau

What you deam as efficiency is not Fuel Economy efficiency. You talking about producing max., power/torque. WIth the appropriate gear ratio, as in todays vehicles it is not necessary to run the engine at this max., efficiency to obtain max., economy.

RPM=bad is true. More RPM means more gas is consume for a given time and and a given distance. If I drove my Subaru at what you call most efficient, I would be consuming gasoline at 3-4 gallons per hour. I know when my engine is running at or below 2000rpm, it is below what you call max., efficiency, and the engine is border line lugging. But I am using very little fuel and that is all that matters.

Are there more variables at work here. Most definitely, gear ratio, weight of vehicle, etc....

But the reality is your average vehicle today is designed for low rpm to achieve max MPG. Not high rpm (3000 rpm) to achieve max., MPG.

alelser

Okay just thinking out loud here, but your average driver measures what we call mileage or fuel economy in distance driven per volume of gas consumed (not per unit of time).

Just using some fake numbers to throw this question out there: if I use 3 gal per hour in gear X at 30 mph and 4 in gear Y at 60 mph, and travel a distance of 60 miles, then X uses 6 gal in 60 miles (10 mpg) and Y 4 gal in 60 miles (15 mpg). So while the engine may be more efficient in gear X, in the real world it burns more gas over the same distance.

I know it's over-simplified, but I used it to illustrate that maybe if we are talking about the kind of fuel economy that most drivers care about -- that which costs them the least for their total miles driven -- then measuring in volume of gas consumed per unit time isn't telling us anything.

It's Sunday, and I usually turn off my brain on Sundays, so please correct me where I am wrong.

-Al-

cgettel

My $0.02 after a couple of months of working very hard on fuel economy in two vehicles - one a GMC pickup with the 8/4 cylinder engine, 4-speed auto, and more recently a 2006 Forester X, automatic. The only mods I've made were to switch to a K&N air filter, and run tires at 42 PSI. When I traded in the truck I was running a fraction over 19 MPG overall, with about 75% in-town driving. Given a full-size pickup with an 8 cylinder engine driven mostly around town, that's pretty good. Right now, I'm running about 28 MPG in the Forester with the same kind of driving, and slowly learning how to increase it. The truck had an instant MPG readout, and I found I couldn't live without it so I got a ScanGuage II for the Forester. Very helpful! It's taught me some things that are counterintuitive to say the least, and always interesting.

One of them is that different vehicles need to be drivien quite differently in order to approach their potential best fuel economy. Another is that we're not necessarily looking for max efficiency of fuel energy-to-propulsion at any moment but *overall*, and that matters. One of the ways it matters is that the guideline then becomes "moderation in all things". The learning curve is long, not so much in the general principals but in exactly how to apply them at any moment.

The questoin about how much acceleration is a case in point. I've tried it both ways in both vehicles, and I can say that in both of them, slow easy acceleration results in significantly lower MPG, and hard acceleration does too. I think it's very simple - in general as long as the engine's running, the slower you go the lower the gas mileage. If you stay at low speeds, the lower the MPG and the more gas is used per unit increase in speed. This has to do with RPMs so it'll make a difference whether it's stick or auto, but the principal is the same. Watching the ScanGuage makes this very clear. In general, I think hard acceleration will result in a little better gas mileage than creeping, as long as you don't floor it out of a dead stop. But moderate acceleration is usually the best combination of staying in low MPG the shortest amount of time and not letting the MPG get too low at any point from accelerating hard.

Accelerating uses a LOT of gas. Accelerating includes not only pulling away from a stop, but any increase in speed - of course. It also includes going uphill. Even a rise so slight as to be unoticeable will drop the MPG quie noticeably - 2-3 or more MPG. I can tell if a stretch of road which seems flat and level is rising or not by watching the MPG readout, as it's quite sensitive. It doesn't matter so much how strongly you're accelerating, just that you are.

At the same time, dumping the gas wide open at any speed seems not to get as much gain in speed as there is drop in MPG. Therefore morderation is key. The ScanGuage has a readout for throttle position and engine load (the amount of power being used compared to what's available at that point). Accelerating at around 65%-75% load seems to be the sweet spot for the Forester, but I'm still playing with that, and haven't done any specific testing. This is what I would call "moderate" acceleration, not "brisk", not "easy". It's less than you'd think. But 75% load at 50 MPH is a lot stronger acceleration than 75% load at 30 MPH - at higher RPMs there's more torque available. The message is don't just slam your foot down, accelerate your acceleration so that pumping loss and so-on are less of an influence. You can start to increase your throttle and rate of acceleration once you get up to about 2000+ RPM while still maintaining the same load percentage and roughly the same MPG.

One thing that was very counter-intuitive was to find out that higher RPMs doesn't necessarily mean lower MPG. When pulling up a pretty significant hill, I was able to improve the MPG 2-3 points by downshifting. This raises the RPM, increasing torque, and so one is able to back off the accelerator quite a bit, and maintain the same speed. Equating uphill to accelerating going down the road, staying in a lower gear longer may improve your azcceleration rate, thus get you into high gear and high MPG sooner.

Someone asked about torque and horsepower. The way I understand it is that torque is what accelerates you, while power has to do with how much ground you cover in a period of time. Like MPG and GPH, they're clearly related in actual practice, particularly when driving an automatic with less control of RPM.

I'm learning how to play the terrain. One feels like maintaining a steady speed is the best way to go, but this means accelerating for every uphill, in effect. You want to arrange things so that you accelerate strongly downhill, where you have gravity's help, and let speed bleed off on the way uphill. This way the acceleration is short, and you spend the least amount of time in lower MPGs, and you're not accelerating uphill just to maintain speed. Avoid acceleration as much as possible, but when you do, use gravity if possible, otherwise get it over with as soon as possible wihtout flooring it or downshifting more than one gear.

The so-called P&G - Pulse and Glide - technique is indeed a wonderful way to improve gas mileage. Here in New England I use the terrain to do this. As a rule, I acelerate up to something a bit over the speed limit at the very start of even a slight downhill, then pop it into neutral, or back way off on the gas, and coast back down to some speed that's the lowest I'll accept given conditions, traffic and how much of a hurry I'm in. Sometimes I'll put the Cruise Control at this min speed, ortherwise I almost never use it any more. Sometimes if I have a lot of speed to make back up, I'll accelerate even up to 99% load (the max readout available - I assume it's really 100%) but for a very short time. The MPG drops like a stone - down to 4-5 MPG - but only for a couple of seconds, and then one can glide in neutral or foot off for a long time with gas mileage in the 70's, 80's and even well over 100 MPG if at highway speed. Where traffic is not in the way, I found I had some mental adjusting to do to allow my speed to vary as much as it needs to.

I turn the engine off if I catch a stoplight that's going to be more than roughly 10 secs. long. Sitting still, your gas mileage is zero - and you can sit there and literally watch your average MPG number tick down. Very discouraging. Other than that I don't ever turn the engine off. Coasting in neutral with the MPG at 70-100+ MPG is good enough, thankyouvery much, and I usually do it as much as possible, even if it's only a few seconds. It's nice to see the ScanGuage reading 0.25 gallons per hour, too. Again, that's good enough for me.

Of course, what we're after is not necessarily lower MPG, it's using the least amount of gas possible to get where we need to go. This is about GPH as much as MPG. However, MPG is definitely time-related, too. And it's not just the number, it's how long you keep it up. They are not independant numbers, but MPG is a lot easier to understand and work with while still at the start of the learning curve. Soon I'll start comparing different driving methods and speeds over the same course to see how or if total gas used varies, and if there's any usable correlation with instant MPG going down the road. For example, I recently found out that going to the grocery store across town via a 4-lane divided highway over a big hill used more gas than going through town with traffic, stop signs and stoplights. Go figure. I think it's because the in-town route offers far more opportunities for P&G driving and being able to stay longer at more efficient speeds, and even though there's a long coast at 130 MPG down the hill on the highway, it doesn't make up for the grind uphill at 50-55MPH at 12-14 MPG and a high GPH figure. This is the opposite result, by the way, from my GMC truck.

I've also been able to corroborate that the most efficient speed for my Forester is about 40-45 MPH. I think of it as a range from about 37 - 47 MPH. Above and below that the MPG drops at least 2-3 points. On a recent 160+ mile round trip on 2-lane country roads and small towns, I ran 31 MPG overall. I think it would have been better if I hadn't felt like I needed to maintain 50-55 MPH or so on the more major roads. I personally feel like it gets very risky the more one deviates from traffic flow, so I feel constrained not to move either too slow or two fast on major highways where one becomes a major hazard if speed departs from general traffic flow by more than a few MPH. An accident would be a far greater disruption to my day than the loss of a potential mile per gallon.

In general I've been getting almost 400 miles per tank - a little less.

Sorry for the lengthy post. I thought it'd be better to have it all in one place than piecemeal in reply to others' points.
Regards -

guroove

Thank you cgettel, your post was the most informed and informative one in this thread, and possibly in the fuel economy forum I have seen to date. Discussions of hypermiling techniques on a subaru forum are often brushed off, but you seem to do it in a very safe and sane manner which makes sense to a lot of us Forester owners. The scangauge seems to be the best fuel economy related modification out there. Your insight will definitely benefit us all.

__________________
2005 Obsidian Black Pearl XT w/4eat

Quicksilver

cgettel-
I agree, well done!
One question...you've talked mph, mpg, and gallons per hour...did you have any correlations with rpms (like a range of most effective engine speeds)?
-Quick

__________________
2004 FXT 4EAT - Platinum Silver Metallic

thefoos

There is a direct relationship between engine load and fuel economy. That relationship doesn't exist between rpm and fuel economy. If it did, why don't you shift at 2k all the time and bury your engine under extreme load?

No sense in arguing though.

__________________
Scott
04 MT FXT STG 2 5 spd
02 WRX STG 2 5 spd

thefoos

Thank you for confirming this. This is because there is a stronger relationship between engine load and fuel economy than there is between rpm and economy.

__________________
Scott
04 MT FXT STG 2 5 spd
02 WRX STG 2 5 spd

cgettel

Thanks for the kind words, folks. Glad it was helpful.

To answer your question, Quicksilver, I have seen no such relationship. I don't usually pay RPMs much mind in an automatic except in a general way. It's on my list of things to check out, but I don't expect to find anything useful in terms of economy, though.

Scott, I agree completely about the load. It's pretty clear with a ScanGuage. That said, since load and acceleeration are also closely related, it's still high on my todo list to do some testing to derive some principals about under what circumstances to accelerate how hard in order to mminimize gas consumption. I'll have to get back to you about that.

As I said, I've been focusing on MPG, which is kind of the simple answer, and I think it's time to start looking also at GPH. I'm not sure what kind of relationships to expect here. Time to do some algebra, maybe. I'm wondering, for example, in what circumstances, if any, you're better off shooting for low GPH, vs. low MPG -- or if just going to max MPG is sufficient. I just want to understand that whole thing better.

Regards -

cgettel

I need to revisit this fuel economy vs RPM question. I should have said not that there's no relation, but that th relationship is not simple. Torque and horsepower are both functions of RPM, and are clearly related to both acceleration and load, which are both, in turn, related to fuel economy. I'm trying to figure out if there's any way to get to a more-or-less simple set of basic principles that you can make use of when actually driving. This is not a simple thing.

Regards -

adc

As others have said. Miles per gallon is not a function of engine efficiency..not even close.

If you get your best engine efficiency at 50 mph and 40% of that energy is lost bc of air friction and then you are 5% less efficient at 40 mph and 30% of your energy goes to wind resistance................. do the math.

__________________
Be careful what you wish for.

robzr

FWIW, torque is a measurement of force, with no regard for time. HP is a measurement of power, which is force applied over a given period of time. The formula to calculate HP at a given RPM from torque would be HP = RPM * torque / 5252. Measuring torque at the engine says nothing by itself of potential for acceleration. HP (not just peak, at any RPM) is much more relevant, since relatively low engine torque at a high RPM geared down may produce more torque at the wheels than relatively high engine torque at a lower RPM. A torque curve will show you at what points in an engine RPM continuum you will accelerate more if the gearing is a constant.

There is a measurement BSFC (Brake Specific Fuel Consumption; units are lbs of fuel per HP * Hour) which may be relevant to this discussion. BSFC does not necessarily correspond with Volumetric Efficiency (a measure of the air pumping efficiency of an engine). Peak VE in a well tuned engine does generally occur near peak torque, which makes sense when you think about it. However, BSFC does not seem to have such a correlation. Even if it did (in theory), there are a lot of competing factors and concessions being made during design and production of engines which prevent them from being delivered in an optimally tuned state.

I'm not really qualified to speculate much here but I believe internal losses will go up proportionately if not exponentially with RPM (friction, fluid dynamics on bearing surfaces, and other stuff that I have no clue about). So I would think peak (low) BSFC would occur at a trade off point of when the engine is beginning to get into an efficient combustion range (ie; moderately high dynamic compression ratio, decent swirl/turbulence in the combustion chamber making for a uniform fuel dispersion, and of course reasonable fuel metering, ignition timing), while being at as low an RPM as possible in order to lessen internal losses.

If you guys are really interested in this, there is a program called Engine Analyzer which I believe has a free demo you can download. You can plug in all kinds of data for engines, and it will model the engine in software and you can experiment with different with modifications, like lowering the CFM on the intake to mimic a lower throttle opening. I ran a few simulations through with a 2003 Ford Focus Stock engine model provided by performance trends. The peak (low) BSFC appears to occur around 1500 rpm, while peak torque occurs around 4500 rpm and HP around 6000 rpm.

It would be great if someone had an engine definition for our boxers...

Rob