Acceleration slow or fast? 2003 Chevy Impala
 

Hi everyone, I just heard about hypermiling a few days ago and thought it would be a fun thing to try. I've read up a bunch, but there's one thing I can't seem to find a good answer about.

Wikipedia says 75-100% throttle is most effecient for acceleration, while multiple other sources say as slow as possible, and a few say inbetween. I read a couple good threads on this forum and different drivers seemed to get different results between different vehicles.

My question:
Does anyone know or have an educated guess for what will work best for a 2003 Chevy Impala? It's a mid-size car with an automatic v6 and no rpm gauge, unfortunately.

Tips on easy effective mods would appreciated as well.

Welcome to the site justjohn. Good to see a fellow Wisconsinite. :)

Slow acceleration is a good general starters tip. If you slowly accelerate you build up less momentum that you may waste by braking during light to light city driving. However, there are more advanced driving techniques.

75-80% is ideal. You want to keep your rpms low, but engine load high. Unfortunately, with an automatic you get one or the other since you can't control engine rpm, just load. Low load means low rpms and vise versa. So, you have to compromise. Usually, moderate acceleration is what I have found to be best. The exact engine load will differ between vehicle as they are designed differently. That is one area where a scangauge would help out.

Thanks for the quick reply! I've gotten fairly good at anticipating in just a couple days so (hopefully) that will be somewhat of a non-issue.

As an update, I just found this article which seems to advocate full throttle acceleration, though I haven't finished it yet. Thoughts?

If you look at a lot of the bsfc maps (we have a bunch in our wiki), you'll notice an increase in bsfc (bad) at very high loads. This is most likely due to open loop operation of engines where air/fuel ratios are enriched to avoid pinging. That is why the 75-80% load rule came around.

Ah, that makes sense I think. Is there an audio cue to tell when you hit open loop?

Also, when you say high load would that equate more with high throttle, high rpm, or something else? (I did notice the drop in eff at high rpm)

Some of us use the Scangauge to tell us whether the loop is open or closed.

load = throttle basically

As 99metro said, the scangauge is the best way to monitor open/closed loop that I know of. Autospeed does have an article on a "closed loop monitor". I built one a while back and found the scangauge to be much more accurate.

Awesome. Thanks for helping me get started!

If you can tell what your shift points are, then you've got a good start. My take on it is when you're under 35 or whenever your TCC kicks in, take it easy. Pushing it harder just churns more fluid and is more wasteful. But once your TCC kicks in probably right after it goes into 3rd gear at 35 or so, push it a little harder to get it up to 45-50 and into OD. Just dont push it hard enough to downshift or your gains are lost for the moment.

justjohn-

When you're accelerating from a dead stop, remember that the greatest amount of friction occurs when you are starting from a stop. Moderate acceleration from a dead stop is a good way to eat up tires and gas. It's better to take your foot off the brake to allow the car to move forward a little and slowly apply gas until you're in 2nd gear. Remember 1st gear is the least efficient. Once you're in 2nd gear, you can apply a little more gas to achieve your moderate acceleration.

My dad's car 2001 Impala, seems to do best with "hard" acceleration. (80% or so)

As long as you don't brake, your car will keep similar gas mileage. Everytime you brake, you are turning momentum into heat.

What's TCC? You'll have to go easy on the acronyms with me. I know very little about cars to start with.

No problem noticing gear changes at least. I'm not deaf. :)


I've been doing the rolling stops. According to those graphs though, shouldn't I be holding the throttle near max and just keeping the rpm low?

Impala specific info always welcome, thanks!

Technically that's the perfect situation, but how do you plan on doing that with an automatic transmission?

Not possible, as I understand it. So you're saying that the method you suggested will be most efficient because it will keep the rpm lower, is that right?

Is there a way to tell if that will make up the penalty for staying in a lower gear longer?

This is an interesting discussion. I still feel you use less fuel if you take off from a slow start. From what I can see on my little gauge on my 944, using far less throttle and going to a SLOW acceleration puts me in the "20mpg" area with my foot down just a hair from a dead stop, than if i accelerated quickly to speed and then leveled out, my meter goes to into my black area which is "Errrr what miles per gallon...." zone.

Seems the same on my little escort. Starting and stopping is a large amount of the fuel consumption on the trip, even cutting down on this very subject can make an enormous change in your mileage.

1. Sounds plausible

2. Isn't that saying basically the opposite of the article I linked? Are you saying it's wrong then?


If you mean that accelerating harder creates more drag I'm pretty sure that's wrong. Drag is dependent on speed, not acceleration.

justjohn - what he might mean is if you accelerate harder, the weight distribution changes. For instance if I accelerate lightly, the nose of the car raises slightly, raising further if you accelerate harder. However, since the co-efficient of drag is the most important thing, and you are correct on speed vs wind speed vs drag is the biggest variable, I think rolling resistance would be a bigger factor. But i think he is talking about the angle of the car when the weight distribution changes. Sort of like if you took a brick in the wind, with the wind driving at you straight forward the surface area exposed is simply the front. Change the angle and the other portion exposed will be underneath, then the wind would be also flowing along other variables under the car. However, at a low speed, I don't see how this would be a very large variable. I think a light foot is better still :)

My method keeps you slow in 1st gear and then moderate acceleration throughout acceleration after that. By moderate acceleration, I mean that you should accelerate as fast as possible with the shift points staying around 2000 rpm. If you accelerate REALLY slow throughout, you're going to be spending too much time just short of the shift points, which wastes gas. Generally with a V6, anything over 2000 rpm is going to eat up gas quickly.

The only way to see the penalties of different acceleration is by using a Scangauge, but this is what has worked for me and several other people.

You could try letting the car roll a bit, then stabbing the throttle hard briefly. These are large cars, they roll very well.

Do you have the 3.4 or the 3.8?

I think he has the 3.8.

I think it has been said that most of the time, highest gear and lowest engine RPM's at cruising speed is the best combination to high MPGs.

Alright good, so essentially what I thought. This does seem to make sense to me for an automatic transmission. Unfortunately I don't have an RPM gauge, so this is going to be pretty tough.

Oh, and I don't actually remember if it's the 3.4 or 3.8. I can check though.

my 80% statement is probably a little stretched. All I know is, I accelerate to the speed limit, and then let the TC lock up. I think it shifts at 2200 or so, I may start accelerating at 2000 and then see the mileage.

I'd suggest trying a few consecutive tanks at approximately the same exceleration rate then try a new acceleration rate keeping track of what seems to be best for you and your car. Personally I have always benefited most by being light on the pedal and always keeping the rpm's as low as possible without overworking the engine, but all but one of my cars have manual transmissions. I have found that going just as light on the throttle as possible always works best for me and I've tried several different methods.

Double checked and it's the 3.4, rated 21/32 city/highway.

OK guys, I'm going to ask a REALLY dumb question, because I'm not an engineering type at all....

What is load? is that the amount of work the engine is under at a given RPM?
Ergo, load is higher at 2k rpm going up a hill than going down? Or is that a different concept?

So there's no way to really know other than scangauge/mpguino? Are they 100% accurate on short term mpg?

SG is pretty good over 2 miles. I do a number of 4 mile trips through downtown Columbia every week. It's pretty accurate, within half a mile per gallon.

Like all gauges, the SG is better up until your average tank length. I've had a 600-mile tank where it was 2 mpg off. It's accurate at 450 miles/tank, though.

If your car has a built-in instant or average MPG gauge, use that to see the results. The instant MPG gauge is by far the more useful for day-to-day use. The average MPG is good for comparing driving techniques across tanks.

Those 3.4's will easily get 33-34 highway if driven sensibly, and 25-26 combined. At least that's what the indifferently maintained 3.4 in my brother's GM minivan gets.

Alright, I'm reviving this thread. I have been pondering this very same question for.... well, a long time.

It seems sometimes as though we drivers of automatics are shunned from the efficiency discussion. Unfortunately, my good ol' 3100 with the 4 spd is all I have for the time being and I have to make due with what I can get from it. Nevertheless, I see no reason not to at least TRY for better fuel economy, even if the transmission isn't the best tool for the job.

In order to settle this once and for all, I intend to employ various techniques of acceleration over the course of a long-distance trip, while monitoring certain parameters in Scangauge, to determine the method of acceleration that yields the least amount of fuel used.

I have tried monitoring AVG (average miles per gallon) before, with very broad results. I believe that either the TFC (trip fuel cost) or the trip total gallons consumed would be the best bet for this.

My idea is to record the starting cost or quantity at a stop. Then accelerate to 55 MPH and record the ending cost or quantity, repeating this for various acceleration techniques. Quite simply, the technique that yields the least fuel consumption from 0 to 55 MPH should be most efficient.

The techniques that I will test are:

• Acceleration at a given throttle setting (10%, 20%, 80%, etc.)
• Acceleration at a given MAP (manifold absolute pressure) setting (20 in. Hg, 22 in. Hg, 26 in. Hg, etc.)
• Induced shifting at a given RPM (2000, 2500, 4000, etc.)

I'll be making the trip tomorrow. Not sure how many of these I'll be able to check out, but I will update when I get some results.

Yes, I'm surprised more people haven't done this already. I'm planning on doing it when I eventually get an MPGuino.

Don't forget that slower accelerating methods will cover more ground so just going by how much fuel is consumed to get to 55 isn't a "fair" method.

I was debating on how to account for this and I was just going to compare mpg after accelerating to 55 and holding at 55 at least long enough to reach the distance required for the slowest method. I'm not sure that's a perfect solution but at least it accounts for the main differences.

Sometimes we drivers of automatics are deliberately shunned from the discussion here. :rolleyes:

I'll tell you the best tips I know for getting better FE with older vehicles that are automatics and less sophisticated (read: Chevys). (I have two of them.)

For highway driving the idea is to accelerate moderately briskly to get into the highest gear and enter torque converter lockup ASAP and stay there. Usually that is around 45 MPH on a Chevy, so maintaining a constant speed of 45-55 will do best. If you go much faster than that you are using more fuel than at that ideal speed for efficiency. If you are driving slower than that the automatic is also inefficient due to the shift points of it.

City driving is a real challenge, and your FE will automatically be lousy (no pun intended). Any time you can avoid coming to a total, dead stop, do so, for that really eats gas up, especially on a heavy vehicle. (But watch out for cops at stop signs and red lights, as getting a ticket is not worth the gas saved!) What does help some is coasting in neutral. But I recommend you only do it when you know you can coast to a near-stop. Most automatic transmissions are potentially susceptible to damage if you re-engage them back into drive at higher speeds.

Let's say there is no traffic and you are driving along a long block with a stop sign at the end. Accelerate briskly enough to cover the distance needed to coast, then shift it into neutral and coast to the stop sign. However don't do this when approaching a red signal, because if the light turns green your shifting back into drive at speeds of over 10 MPH might damage your tranny.

That's about all you can do with an automatic without putting it at risk.

On one hand I agree with you, and on the other I don't. Basically, the idea is simply to determine how much fuel each technique uses to accelerate to cruising speed. Once there we make decent MPG, so any amount of time or distance after that is irrelevant.

However, the overall goal is to use less fuel over the entire distance of the trip. Hypothetically, if harder acceleration was more efficient than gradual acceleration, there should be less fuel used over the same amount of distance. The problem with this method is that although it would determine which technique is more efficient, it adds a certain amount of distance at cruise to the techniques that take less time to achieve cruising speed. This could be considered "unfair" to the harder acceleration techniques.

I think the best way to go about this is to drive a set distance for each test. It shouldn't be too much longer than the technique that yields the farthest distance covered before reaching cruise speed, so as not to skew the results by variations in cruise MPG. I think 1/2 mile should be sufficient. If the slowest technique yields the best results, longer testing distance may be required.

Getting to cruise faster is part of the advantage that harder acceleration has in real life vs slower acceleration. The question is whether it makes up for the fact that more gas is used in the short term.

But either way, we arrived at the same conclusion for how to test, so that's probably the best option. :)

Rats. I thought I was the first one to discover those tricks... :p

I have been able to achieve about 32 in town using these and other fuel conservation tricks, compared to my all-time record highway trip MPG of 38 (thank you, no applause necessary). It never fails to amaze me when I see other drivers accelerate to get right on my bumper on a descending highway on-ramp, just to hit the brakes for the corner ahead, as I merrily coast along at 25 MPH or so :turtle:. It just goes to show how inefficient the standard driving style in this country really is.

I shift in and out neutral at 70 MPH all the time, it gets me some good time at over 120 MPG on a good stretch of highway exit lane. There is definitely some skill involved though. From steady-state cruise, lift your foot off the throttle. The trick is to shift to neutral at the exact moment that the engine crosses the RPM that would match the transmission if the TC were locked (about 1500 for my car), Scangauge comes in handy for this purpose. This is usually about a second or two delay as RPM momentarily increases as the TC unlocks, then slowly drifts down toward idle. If you catch it at just the right time, you can't even feel the change. Shift before and you'll feel the car lurch backward because the engine was still providing forward power, shift after and you'll feel the car lurch forward because the engine was providing engine braking.

Shifting back into overdrive is pretty basic for the most part, just shift to drive and the PCM will automatically find the right gear and (relatively) smoothly engage it. However, I like to try to match the speed at it engages, so as to further prevent wear and tear. Shift to D, and immediately apply VERY light throttle, this will spin the engine up just enough to match the transmission. It takes about a second at cruise to engage. This is tricky to master, too much throttle causes the transmission to "slam" into gear (not good).

I have found that below about 35 MPH, there is no engine braking effect whatsoever with the transmission in drive - in fact it actually pushes the car a bit - and fuel consumption and MPG are comparable to coasting in neutral. So the only time I shift to neutral below this speed is when I know that I will be coming to a complete stop, otherwise it's just excess wear and tear. Also, after coasting down from cruise speed to a stop sign or light where I will only be stopped for a short time, I pop the gear back into D at about 10-15 MPH. It will smoothly reengage at that speed, rather than coming to a complete stop and getting the typical automatic "clunk" when shifting into gear.

With my Malibu, same engine, probably same tranny, it coast just as good in gear as Neutral so I don't bother with N on it. My money's against slow accelerating, I don't like seeing single digit mpg, and you'll see them longer.

I'll do a few 3/4 mile runs tomorrow and record my results with the Stratus 2.4 I4 recording TFC vs max RPM.

I just got done programming the trip distance as an x-gauge. One thing that concerns me is that the distance only displays down to 0.1, and I'm not sure if I can increase that resolution. This means that there could be up to a .09 mile variation in results.

Also, I have decided to use the TFC gauge because (at $4.20/gal) it will have much higher resolution than the fuel used. However, after a quick run around the block, I am guessing that all results will still be within about a $.04 range.

If you use the odometer on your car you should be able to see when it is about to roll over to the next tenth so you should be able to start and stop exactly as you hit the mark, giving you the exact 1 mile or whatever you were going to do.

But yeah, .1 possible error is waaaaaay too much for a .5 mile or even 1 mile test.

I favor hard acceleration, but then I would- I just love stomping on it and making it go. Instant MPG suffers, but it's not too much lower than moderate acceleration and the distance spent at that lower MPG is a lot shorter than the distance spent at the almost as bad MPG of moderate acceleration- at least for my car.

It gives me a lot more distance in glide. It has the added benefit of catching back up to the truck I was drafting that pulled ahead of me on my long coast to the tollbooth. Even if it didn't work for me on its own then catching back up to the truck would make it work.

Up you gas cost to $10 a gallon, then use TFC for each run should give enough accuracy.

I'm going to go gravel road to gravel for turn around, picking something a drive or sign little short of the mile to record TFC, once each direction at rpm's yet to be determined, probably 500 rpm spreads from 2000-3500. Initial data should narrow the area to focus on for max. efficiency.