Optimizing acceleration

Discussion in 'Fuel Economy' started by brick, Mar 13, 2006.

  1. brick

    brick Answers to "that guy."

    Now that I have a good grasp on how to coast, (in-gear terrible, NICE-on good, FAS best), it's time to tackle the bigger challenge: acceleration. So far it looks like nobody really knows if it is better to accelerate very gently and take a very long time getting up to speed, or to accelerate briskly so that you can close the throttle and cruise. My hand calcs and spreadsheet are not surprising, indicating that the losses to drag are greater when you speed up quickly because more time is spent fighting the drag at high speed. These losses seem to be fairly small for the kind of speeds that I'm dealing with, however. (For now let's restrict it to the 0-40mph realm of my ugly 5 mile commute.) Accelerating from 0-40 in 20 seconds only reduces energy lost to drag by ~2% vs. if it took five seconds for a hypothetical 1000m stretch of level road.

    Of course, this does not take into account the efficiency of the drivetrain and I suspect that it is the most important factor. My current technique is to accelerate as slowly as possible and jump gears somewhere between 2k and 2.5k RPM depending on traffic and road conditions. Checking the instantaneous FE on the Scangauge is a painful experience while doing this because the readings are so low. We're talking 18-28mpg under the best of circumstances, 14-20mpg under the worst. I can't help but think, why not just goose the throttle a bit, get up to speed quicker, and then either coast down or cruise with the throttle feathered? Once I'm in a load-following mode at 40mph+ I can keep the instantaneous FE in the 40s or higher as long as I don't come up to a very steep hill or lose too much speed. So it stands to reason that average economy might be better if I take the hit early on and make up for it with extended cruising. Clearly I need to start driving loops and see what works best.

    Any thoughts, especially from those without hybrids? Suggestions on the best way to use the Scangauge readouts would be at least as helpful as direct experience regarding rapid vs. slow acceleration. Right now I use the iMPG and GPH readouts and adjust my right foot for the best numbers. Is there a better number to watch?

    Thanks in advance for sharing your thoughts!

  2. krousdb

    krousdb Defiant NX-74205

    There has been lots of discussion on this topic on a variety of forums, but no definative answer as far as I can tell. During the Prius marathon last summer, we did some experimentation with acceleration rates. Rick Reese preferred more aggressive acceleration, 15-20 MPG instantaneous while mine were more moderate, 25-35 MPG. But when you compared our performance, we had equivalent results, about 117 MPG IIRC. Wayne had by far the lightest foot with 40-50 MPG acceleration rates, but he also tended to let his glide speeds go down lower than Rick or myself. So his 120 MPG segment was probably due more to extended glides rather than acceleration rates. But hybrids are different animals because they are able to optimize the use of gas/electric on thier own. For non hybrids, the answer may be different.

    One intriguing arguement is that WOT (wide open throttle) is actually better to get up to speed because with very little vacuum, you eliminate the pumping losses in the engine. The flip side of the coin is that WOT usually means an open loop condition where the ecu cannot use the O2 sensor to adjust the air/fuel ratio to stoich. In that case you get the preprogrammed target AFR of approx 12.5:1 (in my case with the Del Sol).

    On the Del Sol I have just begun experimentation with a more aggressive acceleration to cruising speed. This is because I noticed that in 5th gear, at 17 in hg and higher, my instantanous MPG spikes, sort of like what happens on the Honda hybrids when entering lean burn. At 35 MPH, 5th gear and 17 in hg, the instantaneous is 60 MPG. At 18 in hg it spikes to 75 MPG and at 20 in hg it shows 105 MPG. Of course at 20 in hg I cannot hold my speed so I gradually loose speed. This is a form of pulse and glide, probably more appropriately called pulse and bleed.:D

    Previously I would accelerate at 10-15 in hg. Now I am accelerating at around 3 in hg. The theory is that I can spend 15 seconds accelerating at 40 MPG instantaneous or I can spend 5 seconds acceleration at 20 MPG instantaneous, which gives me 10 extra seconds at 100MPG.:cool:

    It is difficult to tell if it makes a difference because the temps are getting warmer and I don't know if the gain is because of warm weather or aggressive acceleration. But I can say that aggressive acceleration is no worse that gentle acceleration, at least with the Del Sol. This morning at 54F, wet roads and an 8 MPH headwind I rolled into work at 60.13 MPG.:D

    Now only if I could get dry roads and a tailwind.
  3. brick

    brick Answers to "that guy."

    Thanks for the input. You make a good point about WOT reducing pumping losses since you get rid of the intake restritcion. That theory matches my observations of the past couple of days, which is that I almost always do better in a higher gear. This applies even when I think it might be a little too high, such as dropping into 5th at 42mph up a slight grade. Sometimes it means the difference between tooling around at 28mpg instantaneous vs. 44 or 50mpg.

    I just went out for a little joyride at lunch (I feel guilty burning the gas but it was in the name of conservation over the long-term). About 25 minutes, maybe 15 miles total. The Scangauge showed 32.4mpg round-trip on wet roads at temps of ~50F. "Pulse and Bleed" is a great term, and something that I tried to use as often as I could. There seems to be a sweet spot for the throttle at any given speed where FE maxes out. Not enough throttle and I dump speed, too much and the FE starts to tank. I guess finding the balance is what some folks refer to as "surfing." I have a hard time believing that I'll be able to get this car into the 40mpg range but mid 30s might be doable with more practice. Hell, the EPA only rates it at 32 highway. :)
  4. tbaleno

    tbaleno Well-Known Member

    I think slow acceleration is always best. I wouldn't even think twice about it. Some people don't know any better and just guess and say "If I get to cruising speed faster then I will be at my high mpg numbers for longer." It is kind of logical except the car burns much more gas than is made up when they get to their cruising speed. This is especialy true in city driving where they will start decelerating almost imediatly. Another thing is that that kind of person tends to try to hit the speed limit as their cruising speed where as with slow acceleration you may not ever reach the speed limit before you know you have to slow down.
  5. krousdb

    krousdb Defiant NX-74205

    I guess I'm hoping that you are wrong Tom. Otherwise I will never get to the 10% rolling average torture target that you have dreamed up.:eek:

    It won't hurt to try I guess.:D
  6. tbaleno

    tbaleno Well-Known Member

    Heh. Humm. If I had to bet between you and me being right when I look at our respective numbers I might have to side with you ;)
  7. xcel

    xcel PZEV, there's nothing like it :) Staff Member

    Hi Brick:

    ___I am in Tom’s camp for the most part. When I saw you speaking of > 2 - 2K RPM acceleration rates, I just do not get a good feeling for a lot of reasons … A top FEH driver has gone through the test scenario I designed below and found the slower rate accelerations improved his FE under ICE-On conditions but given the strong EV of the FEH, he only uses that with fake shifts to bring back his SoC for another lengthy EV/glide distance. Look up Gary G’s posts in the LGA (Low Gear Advantage) thread over at GH for his results IIRC? Dan has given you about all that needs to be handed in terms of his automobiles and has a lot of data to back him up as well. Let me add what and where I can …

    ___Given the IMA based Insights best FE came about by riding leanburn (even during acceleration!), I was always the go-slow, no-assist, bring her up in a low RPM driving scenario kind of guy. I never ever touched VTEC and never had I seen > 4 bars of 20 of assist during my own brand of low G accelerations … Unless something needed to be ran away from or traffic was building behind for example. On warmer days (80 + degrees F), she would allow me to reach 100 mpg in less then 3 miles using this approach. No one using the WOT, 1-2-5 shifts, or heavy assist w/ maybe 1/2 throttle all the way to WOT had ever seen those kinds of numbers using those higher G acceleration rates that I know of. Because of the slow and steady accelerations, I know the mechanical friction losses are at a minimum and the strain on any given component was minimal at best let alone did I rarely if ever touch the pack. I am an absolute believer in longevity from whatever we drive and the go slow approach should yield superior results to a larger throttle plate opening/lower pumping losses at higher RPM’s.

    ___Now lets consider the Ranger w/ a stick. She is driven much differently then the Insight was of course and might be a better fit by comparison to your Accord w/ a stick. I assume you have an LX/EX. Let’s start off by comparing HP, torque, weight, and the EPA estimates between the two.

    03 Ford Ranger 2.3L w/ a MT

    143 HP @ 5250 RPM
    154 Ft.-Lb’s @ 3750
    3,002 #’s
    EPA: 24/29 mpg’s

    02 Honda Accord 2.3L w/ a MT

    150 HP @ 5,700 RPM
    152 Ft.-Lb’s Torque @ 4,900 RPM
    3,057 #’s Curb Weight
    EPA: 26/32

    ___The 03 Ranger to 02 Accord specs are close except for the higher RPM - peak torque of the Accord and the Ranger’s terrible aerodynamics. Below 40 mph, the Ranger’s ungainly Cd can be overlooked somewhat and that is where the bulk of acceleration lyes. Anyway, they both probably have very similar feel and if this is the case, might I suggest another approach to accelerate that may help reduce both pumping losses AND mechanical friction losses with lower RPM accelerations. The way I accelerate the Ranger is with about 1/3 throttle up to ~ 1,500 RPM and short shift to the next higher gear. Because the Ranger has such good pull, the acceleration rate is what I call brisk (slow to average for most more then likely ;)). She never sees over 2,000 RPM and her pumping losses are reduced somewhat with a throttle plate that is open by a larger percentage then when using a much slower acceleration. To go with the above, once up to a target speed, she is shut back down (FAS) and coasted to the next light, sign, traffic obstruction and restarted via clutch or gear reduction starter motor depending on my 12 V’s reading in a warm up P&G back to an even higher target. I have ~ 290 miles on her with hundreds of less then 5 minute drive segments over the last 3 months and the tank is still > ½. Maybe a 38 - 40 mpg tank is in the offing with temps ranging from - 2 degrees F to 58 degrees F over those last 3 months.

    ___Because the 05 Accord has an Auto; her acceleration description would be of little use to you. I will discuss her idiosyncrasies in an Accord review I am in the process of writing up later on.

    ___As for watching the Scan Gauge, it doesn’t like to see a FAS and IIRC, the 98 - 02 Accord’s had a straight up ISO bus, not a CAN-Bus. If this is the case, a base reading without a FAS would make sense but adding any kind of ICE-Off time to the mix throws the darn thing into a tizzy from everything I have read. My discussions with the owner/designer of the Scan Gauge by phone a few months ago told me he really wasn’t going to do much to try and help those of us that push the FE envelope no matter what we drive. His words were, “I can’t do anything about communication lag” whereas I think a few programming tricks would have solved the issue nicely … This is where Yoshi’s SuperMid would work out nicely ;)

    ___To continue, lets try some warmed up ICE-On acceleration experiments. Mark off a 1/3 - 1/2 mile segment from a stop sign to a marker of your choosing.

    #1) Start w/ a the ICE at idle, reset the scan gauge, hit the accelerator with a low G acceleration rate (22.5 - 25 mpg) and pulse up to let us say 40 mph. Maintain 40 mph over the rest of the distance until you cross your 1/3 or 1/2 mile marker. Record the gph just as you cross the marker. Head back to your starting point.

    2) With ICE-On at idle, reset the scan gauge, hit the accelerator in a slightly higher G acceleration (20 - 22.5 mpg) and pulse up to the same 40 mph and maintain that speed over the exact same distance of 1/3 - 1/2 mile. Record the gph just as you cross the marker. Head back to your starting point.

    3) With ICE-On at idle, reset the scan gauge, hit the accelerator with yet an even higher speed acceleration (17.5 - 20 mpg) and pulse up to the same 40 mph and maintain that speed over the exact distance of 1/3 – 1/2 mile. Record the gph just as you cross the marker. Head back to your starting point.

    4) Repeat until you are at almost WOT for a short period and maintaining the 40 mph to the end of the 1/3 – 1/2 mile finish line. Record …

    ___The above should help you narrow down your optimal FE acceleration rate w/ a warmed up ICE on a flat section of road … If you add ascents and descents to your trials, that is a whole other ball of wax and one that would need a whole lot more scrutiny.

    ___Next we can talk about semi-steady speed FE saving techniques. Dan and I (well at least I do anyway ;)) use a ton of them depending on what we encounter!

    ___I have to ask about your setup as that can skew your results mightily. How much and what type of crankcase oil are you using and what are your tire pressure? Are you using the OEM tires or have you replaced them with something else? Just a heads up but most replacements are harmful to your FE vs. the other way around in many cases.

    ___Good Luck

  8. philmcneal

    philmcneal Has it been 10 years? Wow

    omfg you got your scangauge already? i ordered mine before brick and i'm still left in the dust...

    i really hope the ICE off bug isn't apparent in my car otherwise... I"LL BE PISSED!

    until then i have nothing to offer for this thread except when i drive the nissan quest 99, I NEVER go over 2000 rpms. When a car is programmed for more torque than horse there is no such thing as lugging in an automatic.
  9. krousdb

    krousdb Defiant NX-74205

    This looks like an excellent test to me. I would be very interested in the results. MY gut teels me that the optimal is at one of the extremes and not in the middle. But thats just my gut talking. I need to try that test if only I could think of a good place to do it.:D
  10. brick

    brick Answers to "that guy."

    Wow. Once again I am amazed at the amount of useful information that you guys have. I'm actually encouraged that I know so little because that means there's so much to gain :).

    Will do! I remember coming across that thread but definitely didn't absorb it thoroughly.

    An excellent point among others. If drivetrain efficiency really is a major player, slowing down the parts would be one way to combat losses.

    I'm more than willing to give that a shot. Same goes for your test plan. I would like to do some more focussed testing after hours, and think that I might invest a little time in it this evening.

    The oil is filled to ~1/2 way between the min/max marks. I do know that the oil is one grade heavier than manufacturer spec due an inability to find the right kind last time. I believe honda wants 0W20 (but don't hold me to that) and I think that I had to put in 5W. The manual says that it's alright in a pinch, and I know that I need to change it out for the correct weight to reduce friction. I will do that when the car is due for a change in April, after the bulk of the cold weather/rich running is over. I use Castrol dyno oil.

    I think that the tires are stock. They are Bridgestone Touranzas, and I pumped them up to 40psi when I swapped out the snow tires on Saturday. I believe the size is 195/65R15. The wear appears to be nice and even all around.

    Thanks again for all the advice!

  11. brick

    brick Answers to "that guy."

    Well, I'll be ****** if you guys don't know what you are talking about. I took advantage of the minimal traffic at this time of the evening and made a little test loop to one of our offices, which is ~20 miles round-trip. Ambient temp is 55, roads are dry, and I started off with a cold engine. I used xcel's technique of shifting out of first at 1500RPM and shifting every other gear right at 2000RPM. 1st and 2nd gear instant MPG numbers were in the teens, but everything from third on up was in the mid 20s while getting up to speed. Jumping from fourth to fifth on a relatively flat road is really wierd. If I keep the same throttle position the instant FE numbers jump up from 30-35mpg to the low forties, and the car literally feels like it's being sucked down the road by some mysterious force. Crazy! But I suppose you would all prefer that I stop talking about dark energy and get on with the numbers.

    Once I got to the office I checked the Scangauge and saw 34.6 mpg. I decided not to reset it and just head back and see the average. The result? 36.1mpg. :D I'm impressed!
  12. brick

    brick Answers to "that guy."

    And another big gain on the way to work this morning. This is the five mile commute up and down hills with several traffic signals, so I don't expect much. Yesterday's trip mileage using the simplistic "take it easy" technique got me an abysmal 20.xmpg. Today's commute in the same conditions but using revised shifting and acceleration came out to 27.1mpg. Not bad considering that the engine hardly has time to get warm!

    My hat is off to you guys for knowing your stuff!
  13. xcel

    xcel PZEV, there's nothing like it :) Staff Member

    Hi Brick:

    ___I wouldn’t say any of us know anything but we guess real good ;)

    ___Congrats with the increasing segments. There are quite a few more MPG’s where those came from so just keep working the Accord on the basics and then you can punch her through the roof with the more advanced techniques! I know she is worth 40 + and you will be the first to take a 6th gen Accord to 40 + for a tank, I am quite sure.

    ___Good Luck

  14. brick

    brick Answers to "that guy."

    Thanks for the vote of confidence. ;)
  15. hawkgt647

    hawkgt647 Well-Known Member

    I'm not sure if this would help, but my ScanGauge has a "hybrid" setting that allows it to stay powered during autostop in the Insight. I don't know if this will help Brick, maybe it would work during a FAS in his Accord.

    I need to put the ScanGuage back in my Acura TL and test this.

    Mobil 1 0W-20 synthetic is still available, you may have to do a little searching, try NAPA if your interested. I highly recommend it, I believe in the stuff.
  16. brick

    brick Answers to "that guy."

    That's good to know. Do you really think the synthetic is worth the extra cost?
  17. xcel

    xcel PZEV, there's nothing like it :) Staff Member

    Hi HawkGT647:

    ___It is not so much the hybrid mode that is the problem, it’s the keying off which drops the older ISO bus and then the Scanguage loses itself for ~ 15 - 20 seconds. The CAN bus can pick itself back up in < 1 second after placing her into IG-II per my conversation with the scan gauge owner but it locked into the pre-FAS speed and gph readout during this period. I have read of some that are going after the ignition system rather then the key with a relay (see Calpod’s FAS it works thread at IC) which bypasses the key’s electronic reboot and it should keep the Scangauge live but I am not sure? I don’t think I want to hack my Accord or your TL but it does appear to work to keep the electronics live and reading properly? Let me know how it works in your TL without any mods as I have to pick one up for future CAN bus based automobiles and would like to know what I am up against before hand …

    ___Brick, you can move your intervals up with Mobil1 to help cover some of the expense. It will not cover all of it but it will some. It is worth it to me for automobile longevity, extended change interval’s, with the FE savings a bonus to help cover some of the extra expense. Now that it has been re-introduced, you should be able to ask your WalMart Automobile parts and supply associate to place it on his/her list for re-order. If not, NAPA and AutoZone are two outlets to consider in your search for the best oil made to date ;)

    ___Good Luck

  18. philmcneal

    philmcneal Has it been 10 years? Wow

    after reading this thread again, and i'm the kinda guy always aiming for 20 - 25 mpg accleration rates. i find I go up to 35 mph and then i make a choice of bleeding my speed (60 mpg +) or keeping it consant at 35 mph (42.8 mpg). Now I'm going to try 15-20 mpg rate for my acceleration and see if that makes a difference on roads I know I don't have to see another light for a bit. If I know there's a light ahead then slower acceleration could be more benifcial as you can have more time to make that light green for yourself. From my experience for every 3 seconds I idle at 0mph, my average mpg drops by .1 mpg.

    edit: i'm still battling this aspects that kills my mpg everytime I come to a complete stop (lately its been too many).
    Last edited: May 4, 2006
  19. philmcneal

    philmcneal Has it been 10 years? Wow

    indeed this is a topic i'm trying to figure out on my own.

    1/3 open throttle and 2000 rpms shifts? (my engine does not like low luggy rpms like 1500 in 3rd will consume more gas than in 2000 rpms)

    or following the recommended shifts points in my manual book?

    15 mph (24 km/h) 2nd gear 3300 rpms drops to 1700 rpms
    27 mph (43 km/h) 3rd gear 3000 rpms drops to 1800 rpms
    39 mph (63 km/h) 4th gear 2500 rpms drops to 2100 rpms
    53 mph (83 km/h) 5th gear 2500 rpms drops to 2100 rpms

    seems like honda really wants me to to keep it between 2000 and 3000 rpm i can see why. If you lightly touch the throttle at anything higher than 1700 rpms the MPG meter will spike for the better. Anything lower than that rpm (with the exception of 5th gear and sometimes 4th until it goes below 1500 rpms) and any input on the throttle and i'll be using more gas than usual just to pull the engine (so lugging in theory). I have a bad habit of trying to keep rpms too low and maybe that's why when I stop I really mess up my average?

    argh one of the only bottlenecks from preventing me from getting good mileage besides the DIY mods.

    it seems at low rpms accelerating, i use more gas to produce a small amount of pulling power which is pretty low if you ask me. BUt at a higher rpms, I eat almost as same amount of gas but the pull is much greater.... hm.........

    no real answer argh!
  20. brick

    brick Answers to "that guy."

    Hmm. I hadn't thought of looking in the owner's manual to see what they recommend. You would think that Honda would suggest shift points that result in the best mileage and least wear-and-tear. I'm going to go grab my book and see what it says for the Accord.

    Ok, here's the chart it gives me:
    "Normal" Acceleration:
    1->2 @ 15mph (24km/h)
    2->3 @ 28mph (24km/h)
    3->4 @ 41mph (66km/h)
    4->5 @ 52mph (84km/h)

    "Cruise" Acceleration:
    1->2 @ 7mph (11km/h)
    2->3 @ 22mph (35km/h)
    3->4 @ 33mph (53km/h)
    4->5 @ 48mph (77km/h)

    I can tell you that "Normal" corresponds to ~3000RPM shift points in all gears (maybe 2700-2800 in 3rd and 4th). I can allso tell you that "Cruise" corresponds to shift points just slightly later than my current 2000-2200RPM target. The written instructions are:
    "Drive in the highest gear that lets the engine run and accelerate soothly. This will give you the best fuel economy and effective emissions control."
    Last edited: May 4, 2006

Share This Page