Archives

Simple question. How hard should I accelerate? As my FE is nearly zero when accelerating, the logical conclusion is to accelerate as hard and as fast as possible to my desired cruising speed, obviously assuming there is nothing in front of me, e.g. traffic lights, possible events that I will need to slow down for. But this sounds like a whole load of fuel would be used up, or is it just being counter-intuitive?

Or is my logic faulty, and I should be accelerating as gently or possible? Or somewhere in between?

Thanks very much for your time in advance!

I have read a lot of people recommending engine load around 80, and shift around 2k.

I have found load in the 70’s and keep rpm’s under 2150 works best for my car.

I think I'm with Allen on this one. Moderate acceleration seems to give me the best results. I've been thinking about doing some calculations on how much fuel is burned overall during a pulse. I would be using the GPH feature of the Scanguage and a stop watch to see how much fuel is burned to go from a lower target speed to a higher target speed during my P & G routine.

Call me a salesman for Autospeed. :rolleyes::p

http://www.autospeed.com/cms/A_110216/article.html

that artical is great,
a lot to understand on a 1st read, I find my self rereading that one every month or so.

It is indeed a GREAT article. I've read it several times and I have a general understanding - I think. But, I am always left with some of the same questions. The main question I have is, how do we apply the article results to our individual cars. The article clearly indicates that best BSFC is going to occur at different loads and different RPM for different cars. In other words, there are different graphs for each car. How do we figure out, except in a very general sense, what the graph looks like for our own cars.

It would seem that one could gather experimental data with the SGII on GPH times time, as diamondlarry proposes. By doing that at different loads, one could graph the results and spot the optimum. As in all testing, a level road and a windless day would be required. In his case, it would seem that the electric assist might complicate the results, but I know very little about Prius.

My personal experience shows the rainbow colored chart to be fairly accurate for my car.

There's also a thread at ecomodder where people post whatever charts they've found. http://ecomodder.com/forum/showthread.php/bsfc-chart-thread-post-em-if-you-got-1466.html

Yow Andrew, you are just a fount of great information! The 1.0L Metro engine is one of the engines I was most interest in an there is the info I was seeking. I do note from the chart that the "sweet spot" is very narrow in terms of torque, or load, so driving a 1.0L Metro in the sweet spot, over a wide range of conditions, is gonna be very difficult. Installing the long gear and doing a lot of shifting would probably be required. I think some experimentation, as diamonlarry suggests, would help nail the actual load, but without SGII capabilities, even that approach has problems.

The site you ref has no info on other cars so apparently this type of vehicle specific information is hard to come by.

The map from autospeed is exactly what i was thinking should happen. Lower gear, and higher throttle increases peak cylinder pressure, which tends to improve thermodynamic efficiencies of any cycle. Nice to have suspicions confirmed. Thanks for the post.

So what is the practical implication of this article for me? Is it better to floor it while pulsing? Does this not go against the adage "accelerate like a granny"?

You should get a ScanGauge to monitor your engine load while you're accelerating. To achieve best BSFC, you want to keep LOD between about 75 and 85, and RPM between about 1500-2200 rpm (for most 2.0 liter class engines). Full throttle wastes too much gas. You need much less throttle climbing a hill, and much more descending a hill, to achieve the same load. We're talking about aggressive accelerating during pulses, not maintaining speed - which should be done with light throttle.

For steady-speed driving, yes, you want the minimum throttle possible, like a granny.

But we want more. So we pulse and glide. This article hits on the technique for the pulse part.

I have a question. I have done 75-80% throttle and had great results. I have now been pulsing from around 40 mph at full throttle and achieved the highest numbers in a year. Someone said that it is my scanguage reading incorrectly because it is wot. Can't thr SG read wot correctly? The mpgs have gone up significantly by doing this. I accelerate from a stop at about 40% throttle at 1st gear then about 100% from 2nd through 5th gear to a max speed of 60 then FAS on the glide. The tires have been at 75-85 psi so the glides are very very long. The change seems logical to me, but what say you all?

If it goes open loop at WOT, then SG may not be accurate. Need to confirm with tank averages.

Can you seen on the sg if it goes open loop?

You should be able to see it on the LP gauge. I'd pull it up and see what happens while you are accelerating with your foot planted. If it opens, the air/fuel ratio the SG uses to compute consumption won't be valid anymore.

at wot, loop is open, sg wiill most likely read 9999mpg.

SG won't read 9999 mpg unless you've got open loop and fuel cutoff. My SG is set to monitor fuel cutoff at idle thottle position sensor (1) plua 4 (5). At WOT, the TPS would be 35 or so, and I'd have <5 mpg fuel economy.

If there is a significant change in throttle position, wot to no throttle, or vice versa, the sg has a tendency to read 9999 for a brief period.

All that can be said for certain is that in WOT circumstances most vehicles depart from the stoich air/fuel ratio and dump enough fuel in to make sure that the cylinders don't burn (ie, they run very rich).

The SG measures air flow and uses that to calculate the fuel consumption using the stoich ratio and will continue to do so during WOT. This means that it will report less fuel used than what is actually burned -- a higher mpg readout than reality would dictate.

I think the advice here is good for manual transmission vehicles, but that a warning needs to be offered for vehicles with automatic transmissions because many automatics require a lighter throttle to keep the engine from revving too high before the transmission up-shifts. In an automatic I try to accelerate at medium loads up to about 2000 rpm and then back off the throttle a little to try to get an up-shift.

For steady-speed driving, yes, you want the minimum throttle possible, like a granny.

But we want more. So we pulse and glide. This article hits on the technique for the pulse part.

To confirm, I should being accelerating like a granny to my chosen steady-speed? I'm not ready to do something as advanced as P&G yet.

I also am still figuring out efficient acceleration.
Wayne said somewhere to keep accelerations in the 20's to 40's on LOD.
For my xB that works pretty good in gears 1, 2 & 3.
But in 4th, engine torque below 2000 RPM is really crappy and LOD under 50 gives painfully slow acceleration, traffic builds up behind me like crazy.
In 5th, acceleration under 50 LOD is just shy of nonexistent unless I'm on a slight downhill.

My solutions to this so far go against much of what I've read here;
-Sometimes I'll keep a light LOD in 3rd and let the RPM's go to 3000 before I shift to 4th, thats helps the overall acceleration rate and especially helps it in 4th gear.
-Sometimes I'll go to a heavier LOD (50, 60, 70) and try to get to the high gears quickly and then go to hang time once I'm at speed.
-When I'm going down a slight decline it's easier to keep LOD under 50 and RPM max at 2000 because my poor acceleration in 4th and 5th gets a gravity boost.

I have a pretty good feeling for the light foot on the pedal and am doing much of that automatically, but I still don't feel like I have the sweet spots dialed in.

Comments please!

Looks like you're seeing the effects of aerodynamic drag there. Obviously, you have to give enough power to accelerate. I'd use the lower LOD values in the lower gears, and progressively increase the LOD as speeds increase.

Aim to keep the rpm low. That's more important than keeping LOD low. 2000 rpm and 80 LOD is better than 3000 rpm and 50 LOD. You can find that from the BSFC charts.

I'm not buying this 70-80% LOD business, at least not above 2000 rpm. On my gas Jetta that puts me well into the open loop range, somewhere I do not want to be. In town I've been using 30-50% LOD values for acceleration -- not babying it by any means, but definitely not half throttle either.

In the higher speed P&G tests I've done, I have found that even pushing up to just shy of open loop burns the tank down much faster than 30-40% LOD does. Unfortunately the SG's fuel consumption calculations are unreliable on my car, so it's hard to get a really conclusive answer.

In any event, that's on a supposedly "torquey" VW engine. It may well be that more traditional (peakier) Japenese engines are more efficient when pushed harder.

On both of my Hondas, I have to be over 95% lod before it goes into open loop.

I'm not buying this 70-80% LOD business, at least not above 2000 rpm. On my gas Jetta that puts me well into the open loop range, somewhere I do not want to be. In town I've been using 30-50% LOD values for acceleration -- not babying it by any means, but definitely not half throttle either.

Hmm. 30-50% LOD on my car would mean an acceleration of 0.5mph per minute! It idles at 25 LOD. Maybe it's because you've got really large engines in your car, while I've got an ickle 1.6? But, I'll try this out anyway. I've never seen SG go into open loop when accelerating, no matter how hard I do it, it only goes into open loop when entering DFCO. But again, I'll double check this.

Yeah, my Tacoma 3.4 V6 would burn rubber at 80% LOD, while the Fit cruises at 70% LOD and climbs at 98% with its little 1.5L. Very car specific, in my experience.

On both of my Hondas, I have to be over 95% lod before it goes into open loop.

Just tried that out in my Honda, I found it does the same when I floor it.

Hmm, obviously my LOD reading works differently than others. I show a LOD down in the low to mid 20s cruising along at 60mph, and can get up 5-6% grades with LOD values in the mid 30s.

(The TDI has what others here might consider "normal" LOD values -- near 50 when cruising down the highway).

My Nissan is always in open loop when pulsing and gliding. It doesn't enter closed loop until it's been running at least a minute.

LOD when cruising at highway speeds is in the 20s or 30s, depending on whether I'm on level ground or climbing a grade.

So far the sg doesn't seem to be in open loop when I accelerate at wot. It reads around 15mpg not 9999. That only happens during fas. Sounds right. Ill check the LP reading too, to confirm, next time I drive. I can't imagine there could be open loop with a 15 or so mpg reading the whole pulse.

You might be surprised. Remember, the SG is only really measuring airflow.

It seems to me that for the SG to read 9999 there is an RPM threshold where the engine has to be above(1200 rpms?) and a fuel flow below the selected level in the "cutoff" parameters. There may even be a throttle position requirement for the 9999 reading as well?

It'll be interesting ....I have drive it later today n Ill post the results.

I've confirmed my open loop during P&G with the SG.

Dfco (9999 mpg) in my car comes with open loop, >2500 rpm, and no load (throttle position sensor equivalent to idle). Once at 9999 mpg, it'll stay there while rpms drop well below the initial triggering rpm.

Sensors involved are vehicle speed sensor, PNP (park neutral position) sensor, TPS, engine coolant sensor, crankshaft position sensor, and camshaft position sensor.

The Scangauge figures you're in DFCO if it sees open loop AND the TPS is below (and maybe at, not sure) the value of the fuel cutoff setting.

Ok, found out whats goin on with open loop with my car.

Full throttle does indeed go open loop.
75-85% also goes into open loop at times
Nice-on coasting goes into open loop at times.
fas is always open loop

good news is ... sg reads about 16-17mpg during full throttle and reads about 15-16mpg at 75-80% throttle. Seems a realistic reading.

Also, the sg obviously shuts off for a second or two everysingle time I fas which is constantly. So I wouldnt worry about it. The scanguage doesnt read in many brief circumstances, so very brief wot is not skewing the numbers any more than the rest of the time. The miles on the tank are agreeing so far. I think many of us as other posters are reporting are in open loop at many times during accel or coasting anyway.

the sg is a close measurement, but as we know not accurate down to the molecule. but that is all we want and need. a close reading within even 5mpg is great for me and i think all of our needs.

after you are in the 50-80mpg range a few mpgs off is nothing. the car is already doing amazing things and saving amazing amounts of fuel.

ps the sg doesnt seem to be even more than 1 or 2 mpgs off most of the time

I think I have a way to take some data...
I regularly drive the same section of road leaving home, it's a 3 mile stretch with 3 lights.
What I'll do is make a check list to track my trip mpg at the end of this stretch and also keep track of which lights I get stopped by and my acceleration style engine temp...

Hopefully after I have a dozen or so data points for each particular condition I can begin to draw a conclusion for my car.

I took the time to work through the math of BSFC and posted on another forum.

the thread (http://www.fitfreak.net/forums/eco-fit/39993-shift-points-best-fe.html)

QUICK question, what is lod and does it show this on the scan gauge? I want one.

LOD = engine load in percent. Scanguage shows this.

Polaski, from your thread you are shifting way too high. 4K RPM is way the heck too high.

You want your shift points around 2K RPM (2300 RPM is okay for a pulse) and load values in the 75-80% range. Even my guzzling Elantra is good for well over 60mpg using that rate and shift points combined with P&G. On the highway, P&G has to use higher RPM (because of the speed and gearing) but I still use comparable acceleration rates and get over 50mpg consistently.

Here's an example of someone who shifts a Fit much lower. I believe he's shifting at about 2,000 rpm. This is what's possible in your car.

Use high load on the engine, but control the acceleration by keeping the rpm low. Fixed foot position, adjust with the gears, rather than the other way around.

http://www.cleanmpg.com/garage/images/977.png (http://www.cleanmpg.com/index.php?page=garage&displayunits=MPG(US)&viewcar=977)

Keep in mind that my Civic has VERY similar specs to the Fit. Similar engine, even down to the power/torque curves, similar gearing, similar vehicle weight. See what I do with it. My commute this morning maxed out at 2100 rpm.

I know a lot of folks will go with higher load, especially during their P&G segments - my Paseo seems to be much happier with gentle acceleration all the way around. The smoother I keep my whole process, gentle shifts, gentle rev matching, gentle bump starts - the better my FE is.

Of course every car is different and my observations are based purely on MPG figured at the gas tank - same pump every time, fill it til it nearly spills.

i have been trying not to go into open loop with my pulses and can go up to 87% throttle and keep closed loop on my car. hoping to get higher mpgs that way.

Well, the Woo seems to like it, whatever you're doing.

yeah, Pale, i read the Chevy Aveo is a Woo too, made South Korea by Daewoo with a Chevy tag on the hood. Dodge has at least one Daewoo under the Dodge name. I believe it was formerly known as the Daewoo lanos. I guess the Aveo would respond to what Im doing similarly? Anywho, so you think there would be a flaaw in doing about 87% tps on a pulse as long as loop stays closed? of course keeping rpms low and lod high. what you think, i mean you are the one among others who got me to the big time numbers so far!!!

What I mean is you're doing it right. That's pretty similar to my approach, and it gets good results for both of us. In my case, 95% lod or higher is the trigger for Open Loop.

Like I said, the high shift points I'm trying are just that, a trial. I'll run 2-3 tanks at that and see what happens, then try a couple with your 2000 rpm suggestion since it's lower than I tried before. Just pointing out, my last 2 tanks were a low rpm test (2500 shift with high load) and I got 32.747 and 31.916 mpg respectively, my worst two tanks since I bought the car.

I put over 79,000 miles on a 1996 civic ex manual before I bought the fit BTW, though I couldn't do much high-load low-rpm acceleration on a tank. Sucker needed new rings, burned oil, and if I kept the load down it wouldn't contaminate the oil quite so quickly. Easy decision really, high-load and change the sludge out at 1000 miles, or low load and change at 3000.

So you guys talking high load but low shift points are talking manual trans right? it is not possible to do in a auto it seems, my auto will shift at 2k rpm fine but only if im very gentle with the acceleration.

What does a engine do in open and closed loop?

Here's a link (http://www.cleanmpg.com/forums/showthread.php?t=17788&page=6) to my experiments with P&G in my MT car. Start reading at page 6.

P&G with an AT is really tough. I still haven't found a really good routine in my wife's AT Elantra. In fact, the only regimen that shows any measurable improvement over steady speed cruise control is a 30-45 mph P&G delta, and it's only 2 mpg. 25-40 mph gives just a 1 mpg improvement over CC.

Do any current autos stay in closed loop with stoichiometric mixture at full throttle?

How much do they typically enrich the mixture? Down to max power mixtures of 12.5-13:1, or even richer? (I well remember how most American cars used to smoke on WOT.)

I have a 2009 Yaris sedan with AT, and no Scangauge yet. I'd be particularly interested in hearing anyone's observations with a similar vehicle about where it goes into open loop, and how much it enrichens the mixture.

I think you'll find that it downshifts long before it goes into open loop. As long as it's not downshifting, you won't be hitting enrichment.

Good point, thanks!

I'm still curious about when it goes open loop, but I'm not likely to encounter that much. With the AT I'm unable to load it down with high manifold pressure at lower RPMs.

I'm confused by this article. SFC is lowest at 100% LOD (see link). should I be jack-rabbitting through my gears at 2-3k rpm?

http://us1.webpublications.com.au/static/images/articles/i1102/110216_6lo.jpg

edit:

I just got to this image- I guess so. http://us1.webpublications.com.au/static/images/articles/i1102/110216_10lo.jpg

so the common thought that turtle vs rabbit fast starts was false? its actually the latter assuming the gear shiftpoints remain the same?

Article (http://www.autospeed.com/cms/A_110216/article.html)

Look at this chart instead. Aim for the red zone.
http://us1.webpublications.com.au/static/images/articles/i1102/110216_9lo.jpg (http://www.autospeed.com/cms/gallery/article.html?slideshow=0&a=110216&i=9)

If you go all the way to 100% load, you get into fuel enrichment and MPG goes out the window. From personal experience, 1500-2200 rpm is better than 2k-3k. High, but not maximum, throttle - more like 75%.

what is 100% Load equivalent to the yaxis variable on the graph? ref ( http://www.autospeed.com/cms/gallery/article.html?slideshow=0&a=110216&i=9 ) is this the same load as our SG "LOD"?

100% LOD on the Scangauge would be the top of the chart - where the color stops. So, at 2000 rpm, the red zone is 80-120 out of a max 140. That's 57% - 85% LOD.

thanks for your help! is there any qualitative way to measure SFC? Obviously every engine is different- is everyone following a general rule of thumb such as shifting at 2500 while keep the SG LOD at 70%?

You'd need a dyno and a lot of time to figure out the specifics for your car. Very few specific charts are available "in the wild". I've found that chart to be pretty close for my car, and then I've polished my technique from that starting point.

See how the yellow zone on the chart is only 7% worse than the red, and it's a much wider target.

not to be smart, but how did you "find" out that the chart was close to your civic?

Trial and error. After a while with the scangauge, you can get a feel for the car's performance. I did start from that article, and especially from that chart. Experience tells me that the lower rpm range generally works better, unless hills are involved.

For what it's worth, I concur. See my last tank in the Elantra.

Sean - it's worth a lot. If you were to disagree, I'd go out and try your way tomorrow before replying. Since you agree, well, I told you so! ;)