Fuel Saving Idea? 3

[XR250]

I have an idea for a way to save fuel.
I am sometimes a hyper-miler. My car is a 2023 Mazda3 6MT. I notice when I coast on long downhill highway stretches, my indicated MPG definitely increases versus driving the same stretch in gear. Same for long downhill section of backroads. I assume this is due to the lower pumping and friction losses in the engine running 2,200 RPM versus 700 RPM idle.
I also look for stop signs and red lights on my GPS and try to coast accordingly to minimize the use of my brakes (without holding up traffic of course). These strategies gain me at least 4 MPG.
This gave me the idea that software could be written to coast the vehicle based on GPS terrain, traffic and traffic sign data. This might have to be implemented in some sort self-driving software to be effective. I would assume the car would require a transmission that can coast. I know the new Mazda CX90 doen't use a torque converter in the automatic but a wet clutch so that could be held open- same with a DCT.
Is this just a crazy idea?

 

[lucky-guesser]

When you get the timing just right and your coasting allows you to hit the light just as it turns green, yes that will definitely benefit you, as is the case for neutral coasting on down hills, but I HIGHLY doubt you are getting an honest to god 4 mpg overall improvement.

Could it be implemented in a self driving car? With enough R&D I'm sure it could, but in my experience you might see a 1 mpg improvement if you are lucky.

 

[XR250]

I basically try not to use my brakes. It is def. a 4 mpg increase when I am careful about my driving in this manner.

 

[CWB1]

Cruise controls with automatic transmissions have slipped neutral going downhill for years. Part of the reason "tow mode" was added to modern pickups was not only to adjust shift points for heavier loading, but also to ensure the trans stays locked into gear to provide engine braking downhill.

 

[XR250]

Cruise controls with automatic transmissions have slipped neutral going downhill for years.

Are you sure about that? Do you have a reference?

 

[BrianPetersen]

Any modern automatic can be made to coast (regardless of torque converter) by de-energising whichever clutches are normally engaged for whatever drive gear it is otherwise in, because they're all solenoid controlled under ECU command nowadays.

The tricky bit would be getting it to do this seamlessly without leading to driveability issues and complaints. People complain about idle-stop/start ... this would be way trickier.

Or you can simply leave it in top gear during coastdown as long as possible. Modern automatics have really tall top gears anyhow. There's not going to be a whole lot of difference between 700 rpm idle in neutral and (let's say) 1000 rpm with top gear engaged.

Taking up the load again without incurring a gearshift jolt is the trick here. How do you distinguish between the driver just barely requesting taking up load for steady speed (no downshift needed) and requesting slow acceleration (perhaps warranting a downshift from 10th to 9th gear), and progressively requesting faster acceleration (downshift 10-9, oops not enough, downshift 9-8, oops still not enough, downshift 8-7, oops the driver keeps on pressing further, downshift 7-6, STILL not enough, downshift 6-5, there we go!)

A momentary delay to evaluate how quickly the driver is stabbing the accelerator, thus enabling a guess at the right gear to pick, is a common strategy (and it also "debounces" abrupt accelerator movements due to travelling over a bump), but then people complain that the transmission is slow to respond!

This "calibration" thing, is not easy.

Perhaps you are understanding part of the motivation to use CVTs now ... or forget the whole thing and just go electric, no more multi-speed transmission and shift logic needed ...

 

[BrianPetersen]

Addressing simul-posts. Lots of automatics drop out the torque converter clutch when the torque request (basically, accelerator position, kinda-sorta) drops below zero, i.e. coasting. It doesn't drop the engine to idle, but it might drop it a couple hundred rpm. You can tell if you carefully watch what the tach does relative to what the speedometer does when you gently move the accelerator pedal between zero-input and just-a-wee-bit input.

 

[XR250]

That makes sense but I don't know of any that actually open up one of the planetary clutches ro freewheel it using cruise control as CWB! suggested.

 

[TugboatEng]

I don't think want to disconnect the engine from the driveline as the engine is also running the pumps and alternator. The engine is much less efficient at driving the pumps and alternator than the driveline. What you may want to do is shut off fuel injection and open the throttle wide to eliminate the pumping loss.

 

[BrianPetersen]

Leaving it in gear at somewhat higher revs but with deceleration fuel cut (so zero consumption at that instant, but with a little bit of deceleration) might turn out to be better in most real-world circumstances than coasting with the engine at idle speed but using some fuel in order to maintain idle speed. And, people expect a certain amount of deceleration when they take their foot off the accelerator. Actually shutting off the engine in these conditions while coasting is never done. For one thing, power steering and brakes (although there's ways around that nowadays). For another, transmission oil pump!!

 

[XR250]

More reason that it would be implemented in a self-driving scenario.
I did not realize cars with CVT's and million speed auto's were doing 1,000 RPM at highway speeds. Must only work for larger engines.
My Mazda is doing 2,200 RPM at 60 +/-
I'll take the hit of my 2004 Tundra with a 4 speed auto doing 2k at 60 versus the questionable reliability of new trucks!
I suppose a Hybris would be the best candidate for this as it regen's the braking so using intelligent braking or coasting would not have as much effect.
More to the point of EV's
Thanks for thinking thru this with me.

 

[EdStainless]

I wish that cruise control had a standard 'hold speed' mode for when you are in traffic and looser 'eco' mode that minimized acceleration and such.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[TugboatEng]

It's counter intuitive but you want to maximize acceleration for best fuel economy. Your engine is more efficient at higher power levels so you want to maintain a high power level while doing the work of acceleration. Then, minimize power while cruising and do as little braking as possible.

 

[XR250]

I have hear that and it makes sense, but my experience is that I get better mileage when I accelerate gently.
Isn't the air-fuel ratio richer at WOT?

 

[TugboatEng]

The whole purpose of overdrive is to force a wider opening of the throttle plate.

At WOT the engine operates in closed loop mode so the O2 sensor is ignored. However, you would spend very little time in this condition because you'll accelerate quickly.

Also, leaner does not mean more fuel efficient.

 

[GregLocock]

At higher power settings the engine runs very rich to use fuel to cool the exhaust valve, and, to my surprise, the cat. However, somebody logged the actual driver behavior on customer cars for several thousand hours and I spent a happy few hours dissecting the results. None reached the red line, WOT was scarcely touched.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[BrianPetersen]

Most modern powertrains will allow full-torque operation in closed loop at lower revs (in the highest gear that the current speed and load conditions will permit) so as to be as efficient and emissions-friendly as possible. My van stays in top gear with the torque converter locked as much as it can when going up a hill, and only does a downshift if it's putting out full torque and unable to maintain speed in that gear. Scangauge says it's still in closed-loop when doing this - it would not comply with today's emission standards if it had to go rich in that situation. (Chrysler 3.6 Pentastar, 62TE automatic)

In the old days, the most-efficient load at any given RPM was usually around three-quarters of full torque, presumably due to enrichment beyond that and potentially delayed ignition timing to avoid detonation, which carburetors and old-skool simplistic logic had to do. On the newer engines with variable valve timing, it's probably still most efficient a little below full torque, because (for example) the Pentastar shifts its cam timing below full load to emulate the Atkinson cycle (effective expansion ratio higher than effective compression ratio), but if it actually has to put out full torque, it can't do that any more.

 

[TugboatEng]

Maybe we need to compile a list of engines that are dual Atkinson and some other cycle. My wife's Camry has a big 2.5 liter engine. It may be Atkinson only.

My engines at work are Miller Cycle to the extreme. It's interesting that the BSFC doesn't vary and any extreme way. .315 lb/hphr at 355 psi BMEP at peak torque speed. At idle speed, .365 lb/hphr at 103 psi BMEP.

 

[BrianPetersen]

It's fair to say that anything with variable valve timing is going to head towards Atkinson cycle at part load, because that's what the standard optimization process is going to lead to, the only question being to what extent. Prius is Atkinson all the time.

 

[TugboatEng]

I think Toyota is Atkinson all of the time on many of their engines which is why the 2.5 engine currently in the Camrys is only 176 horsepower. I will say that it runs bette than rh 2.5 V6 in my Lexus. Same power but it picks up at a much lower RPM