By on April 5, 2010

Joel writes:

I have a manual transmission car and am traveling on the highway at a speed that has my RPMs low for mpg. Now say I come to a hill, and my car can hold its speed and rpms, but I have to floor it or nearly floor it to keep my speed. I could downshift, where my engine rpms would go up, and I wouldn’t have to give as much gas through the pedal. My question is, is downshifting using more gas than keeping it in the same gear?

Now, I do realize that this is pretty much an OCD kind of question, but then again, if the answer is rather large, it would be good to know. And, if the answer is negligible, that would be cool too. So, if you know the answer, or is the kind of thing to put on the website, I’d love to know the answer. Thanks in advance.

FYI: the cars in question are a 2001 Honda Civic and a 1985 Volvo 245, and a 2007 Honda Fit, and all are sticks. That said, I even converted the wife to driving a stick, the Fit is hers!

Sajeev Answers:

I answered this question once, well before there was a Truth About Cars.  Years ago I installed a somewhat desirable, vacuum fluorescent display’d Ford “Tripminder” computer in lieu of the bean-counted clock in my 1988 Mercury Cougar. Proud of my accomplishment, I religiously monitored my instantaneous fuel economy to ultimately answer your question.  And the answer was most inconclusive.

You have three factors in play: gearbox ratios, engine torque, and long term durability.  Let’s stick with that final point, because going uphill at low rpms with high throttle effort isn’t a good idea for the engine.  The extra load can have a negative impact on any number of bearings, shafts, etc. And while you might not notice any problems today…

Then there’s engine torque: if you have a gutsy V8 or turbo diesel at the helm, the motor is far less likely to “bog” at lower rpms when you hit the throttle.  But most importantly, it’s game over when the motor bogs, downshift to save your engine all that unnecessary stress.

Finally there’s gearing: wide ratio or short ratio gearboxes in particular.  When I did my fuel economy tests on my (wide ratio) four cog automatic-motivated Cougar, the rpm jump out of overdrive was quite significant. And mileage went down considerably compared to nursing the big V8 up a crest in top gear.  After I made the switch to a six-speed stick, the rpm jump from 6-to-5 was far smaller, keeping the rpms low while avoiding the “bog” was no big deal.

Grain Of Salt Note: Houston is a relatively flat city, so my Tripminder’s calculations have little to no relevance in places with real terrain changes. No driving condition is the same, so I’d recommend this: avoid engine bog at all costs, downshifting accordingly with no regard to the trivial change it might have on fuel economy.

(Send your queries to [email protected])

Get the latest TTAC e-Newsletter!

46 Comments on “Piston Slap: Bogging, Tripminding, Economizing...”

  • avatar

    Did I miss something … or was that a complete non-answer ??

    Let me try to restate the question … appologies to Joel if I’ve mis-read his question.

    I’ve actually wondered the same thing in my more OCD moments. (Also have a 7th gen Civic 5spd) Forget bogging the engine. Not the issue here at all.

    I actually encountered this exact situation climbing Hogback in VT yesterday. 5th gear I could hold speed at nearly 90% throttle @ maybe 2000rpm. 4th gear I could more easily hold speed at maybe 50% throttle @ maybe 2800rpm. (don’t remember the exact speeds and rpms)

    So, which would burn more fuel ?? Higher rpm with less throttle input, or lower rpm with almost full throttle. My gut feeling is the lower gear, but I don’t know.

    FWIW … I dropped to 3rd and ran up to 5000+ rpm 8-) Big truck … wanted to get ahead of him so as to enjoy the twisties ahead.

    • 0 avatar

      To be fair, he said it was inconclusive. But he brings up the perfectly valid point that either way, you’re putting quite a bit of mechanical stress on the engine if you go uphill at low revs in a high gear. So basically, Sajeev’s saying it’s not worth it to ruin your engine and transmission, OCD or otherwise.

  • avatar

    Using a higher throttle setting in a taller gear minimizes your pumping losses. Having a wide-open throttle is one of the reasons diesel engines are so efficient (most noticable at idle). Lower RPM also means fewer parasitic losses for a given power output.

    Finally, and this one is just speculation, it seems lower rpm might give the mixture more time to expand and really extract all the heat you can.

  • avatar

    My old BMW 325i had an mpg meter in the tach. Throttle position was much more a factor than rpms. The car had a close ratio 5-speed with no overdrive, 3000rpm at 70mph. That car had a similar torque curve to a Honda, no low end grunt. Downshift it.

    The pumping losses theory does not matter on a gasoline engine as much as other factors. Generally, gasoline engines are most efficient at or near torque peak. That was under 2000rpm on the old Chevy six, more like 4000rpm in a Honda.

  • avatar

    Won’t help on the Volvo (too old), but for the Civic…

    Buy/ rent/ borrow a good code reading device or scan tool and connect it to the data port. Fuel flow in gal. per hour (or minute) is on the OBDII bus. (Might be a metric reading depending on where the car was made…)

    You should be able to get real-time readings while you experiment with different driving styles.

    From personal experience, Hondas are most MPG efficient between 2000 and 3000 rpm. Below that they bog, and above roughly 3000 rpm they begin to “come on the cam” and make more power. Great fun, but not really efficient.

    Good luck, and keep us posted as to what you find.

  • avatar

    I’d recommend investing in a ScanGauge II (or similar) so you can read the OBD instant mpg (as well as numerous other engine and trip stats).

    I know that doesn’t answer your question, but I will add one piece of advice: When you come down the other side of the hill, it’s better to be in top gear (injectors off) than in neutral (injectors idling). However, if top gear slows you down as you descend, neutral is the better option.

  • avatar

    I have a Fit and Scangauge, so I can answer for that case: stay in 5th, unless you can’t. (If you can’t, you’re probably climbing a back road in the Rockies.)

    Also — and this one I checked carefully — all else being equal, you’re better off accelerating quickly to get into high gear sooner.

    • 0 avatar

      I should add that the Fit has a rather short 5th, giving around 3K rpm at 60 mph. Supposedly that was chosen to allow climbing mountains in 5th, for those who want to turn on cruise control and go to sleep.

    • 0 avatar

      As long as you are not wasting all the energy you got accelerating by having brake to a stop at the next light, yes… by all means accelerate
      at half to 3/4 throttle and upshift to your tallest gear as soon
      as you can.

      Just look at what modern automatics do. Just after you pass the RPM that would result on the engine not being bogged after the upshift, they upshift. It seems weird at first, but it saves fuel.

      On the highway modern engines with wide band sensors and smarter ECUs also open the throttle more and reduce power via leaning or valvling to save on pumping losses…

  • avatar
    Ken Magalnik

    I’d love to know for sure, but I’m guessing engines are more efficient at higher rpm due to valve overlap. In most engines, there is a point in the cycle where both the intake and exhaust valves are open at the same time, the idea is that the air coming thru the intake has build up enough momentum to flow into the cylinder and displace the exhaust. In order for this to happen, the air flow has to be fast, IOW rpm has to be relatively high. At low speeds the exhaust can get pushed into the intake plenum, and then sucked back into the cylinder, where it is injected with fuel that cannot burn as completely due to the lack of oxygen.

  • avatar
    Paul Niedermeyer

    Gasoline engines are most efficient at wide open throttle (WOT) at the toque peak. With adjustable valve timing, that is obviously more of a plateau than a peak. I lean towards higher gears and a more aggressive throttle setting. Ultimately, the difference is going to be very little, as long as you avoid the extremes.

    • 0 avatar
      Darth Lefty

      While this is true it doesn’t matter much for mileage unless the engine must be WOT to go 65 mph. And that’s not usually the case these days…

    • 0 avatar

      Taken at face value, the engine may be most efficient at WOT at peak torque, but what do you do with the excess torque? WOT at Peak torque, where peak torque is sufficient to sustain a given speed, maybe a truer statement.

      But, I don’t believe the initial statement is true either, WOT generally running rich, which is less efficient then something less then WOT.

    • 0 avatar

      Gentlemen your discussion is reminding me of the rationale I orginally heard for CVTs in cars. Namely that the CVT would keep the engine at or near it’s tourque peak at all times. I know that’s what it does with my 150cc scooter. The engine is capable of somewhere north of 9000 RPM but when I’m at max speed and no longer accelerating the engine is pegged right around 6,500 RPM. (Top speed BTW is approximately 60mph.)

    • 0 avatar

      The CVT doesn’t necessarily keep the motor at peak torque, just at the rpm required for current torque demand. Which allows them to drop rpms significantly when torque/power demands are low. The Sentra that I once rented was often below 1400rpm during my suburban commute.

    • 0 avatar

      Thanks for the clarification srogers.

    • 0 avatar


      The CVT on your 150cc scooter is probably a rubber belt mechanical CVT rather than the electro/hydraulic wonders in modern automobiles. The two are vastly different.

      Rubber belt CVTs are effectively a horsepower governor. One could easily mistakenly say “torque” but it’s actually HP, and it’s keeping your engine at its power peak during acceleration while the CVT shifts ratios to maximize the available torque at the ground for maximum acceleration. The clutches won’t downshift so far as to allow the engine to over rev its horsepower peak, but they won’t upshift so quickly as to drag the engine down below horsepower peak. By backing off the throttle and hitting cruise speed you no longer demand the acceleration, decreasing belt tension, allowing the clutches to shift to a numerically lower ratio and decreasing engine speed.

      cardeveloper, when you have excess torque at WOT/peak torque you can further improve fuel mileage by the hypermiling technique known as pulse-and-glide, whereby you accelerate at WOT to a point above a target speed and cruise in neutral (to prevent engine braking, and with the engine off if you’re nuts enough) to below your target speed and repeat. This is the technique used by folks who build the ultra-efficient vehicles for fuel mileage competitions.

  • avatar
    John Horner

    KPS has real data, and KPS’ data makes sense. You are not going to hurt the bearings driving in top gear up a hill under any but the most extreme circumstances. If you are not hearing any ominous noises, no worries.

    Pressure lubricated main and rod bearings ride on a constant oil film. They wear primarily during the brief period at start up before oil pressure is developed.

  • avatar

    100% agree with kps.

    Sajeev mentions “bogging” the engine, but I believe the correct terminology is “lugging.” The engine only really “lugs” when you are at/near full throttle and the terrain load is causing you to lose speed. That degree of lugging requires a downshift, but if you’re accelerating, even ever-so-slowly, it isn’t dangerously lugging your engine.

    Also while engines are theoretically most efficient at WOT in the real world they aren’t. You need to stay far enough out of WOT to allow the computer to run in closed loop mode, which is typically around 85% throttle. Having a ScanGauge II or other OBDII reader helps as they are capable of displaying whether or not the computer is in open-loop mode where it injects extra fuel for full throttle knock prevention.

    So hill climbing, as well as accelerating from a stop on level ground, is best accomplished by maximum closed-loop throttle input while shifting to keep RPMs at or below the torque peak. There is a more accurate method of determining shift RPM based on mean piston speed, but I can’t recall the optimum mean piston speed from memory right now (plus there’s wiggle room/debate about the exact number), but most engines wind up needing to shift between 2000-2400 rpm according to piston speed. Longer stroke engines shift earlier and short stroke engines shift later.

    • 0 avatar

      hmmm …… You wrote: “The engine only really “lugs” when you are at/near full throttle and the terrain load is causing you to lose speed.”

      By that definition I could be lugging the engine at moderate rpm, if I’m just in too high a gear. Like on a long slow incline where I’m gradually losing speed, but not yet downshifted.

      The reason this caught my attention is I have a 5spd Matrix. According to a number of Matrix forums this car has a high falure rate for transmission bearings, particularly on 03-04 models. The Matrix is geared way too tall for the engine. Many reviews for this car refer to it being underpowered, but I suspect the real reason is poorly designed gear ratios. You have to downshift for mild headwinds, or any hill over 1% grade.

      This could mean that during normal driving in a Matrix people may be lugging the engine much of the time because of the poor gearing (or screaming the engine just to maintain speed on a hill). If lugging can damage the bearings, I wonder if this could cause the high failure rate of the transmissions?

    • 0 avatar

      @rtt, My girlfriends Vibe destroyed the bearings in third gear of the transmission by 40,000 miles but it was an 2005. Local GM dealer replaced the entire third gear gratis. Her Vibe is the base model with the tiny engine and 0 options. I’ve noticed that it cruizes at about 3000 RPM in 5th and going about 80mph on flat sections of interstate. I always felt those RPMs were a little high but then I grew up with 80s Detroit iron that was lucky to hit 5500 RPM without turning into a grenade. Plus my daily driver is a 2004 F150 with the 4.6 modular engine that has enough torque and the gearing to not have to rev very high.

  • avatar

    One key factor that has not been mentioned so far is if the engine fueling is running closed loop around stoichiometric (0.95 < lambda < 1.05) or open loop.

    Basically under most operating conditions the oxygen sensor in the exhaust is used for feedback control to cycle the air fuel ratio just above or just below lambda of 1. This allows 3 way cataysts to work.

    Under high load conditions often additional fuel is added to aid cooling of key components (such as exhaust valves or close couple catalysts). This obviously uses more fuel….

    So as stated previously, lower rpm & higher % open throttle position will give lower fuel consumption but only while the engine is running around lambda

  • avatar

    By lambda, do you mean when the there’s roughly enough oxygen to burn all the fuel, but no more? What do closed and open loop refer to?

    • 0 avatar
      Ken Magalnik

      In an open loop system, the computer consults a chart that gives it the theoretical optimum fuel quantity for a given rpm, throttle opening, air temperature, load, etc. The value is pre-determined and not verified in situ. In a closed loop, the computer verifies the amount of fuel burned and makes further adjustments on the fly.

    • 0 avatar

      David, lambda ratio is the percentage rich/lean of a combustion mixture. By definition lambda of 1 (100%) is equal to stoichiometric mixture. a lambda of .95 would be 5% rich, and 1.05 is 5% lean.

      Closed-loop defines the computer control scheme for determining the amount of fuel to be injected. Closed-loop means there is a sensor providing feedback monitoring the results of the calculated fuel injected, allowing the computer to make adjustments to more accurately control to the proper air/fuel ratio. Open-loop means the computer is not consulting a feedback sensor (oxygen sensor) and is more or less guessing at fuel injection based on pre-programmed data and the pre-combustion sensors such as pressure, temperature, speed and throttle position.

      Ken is correct in that during closed loop operation many EFI vehicles alternate between rich and lean to alternately activate the oxidation and reduction catalysts. Catalytic converters require a lean mixture to oxidize hydrocarbons and CO into CO2, but a rich mixture to reduce NOx.

  • avatar

    I know this wasn’t intended to be a CC edition, but the picture, the mileage computer and vents appear to be definitely from a Ford, is it a mid-late eighties T-bird or Cougar?

    • 0 avatar

      Google is amazing. The picture above is actually my Cougar, as I was the first (on the Internet) to do the conversion.

      And, true to my pre-Piston Slap roots, I posted the entire conversion for everyone:

    • 0 avatar

      Nicely done. I had an ’86 Thunderbird ‘elan, so I recognized the computer, but the dash looked different. The T-bird’s was surrounded in fake wood grain. Mine also had the digital instrument cluster; I learned to like it, though it annoyed my girlfriend when I complained about following someone who was going 59 mph in a 60 zone.

    • 0 avatar

      Sweet. You’d like that digi-cluster even more if you knew that adding a $0.05 resistor to the circuit board allowed it to read a full 188 MPH instead of stopping at 85, per federal mandate.

      Not that the car was good for much more than 115 MPH, but whatever.

  • avatar

    I used to have a late 80’s dodge turbo and lived in a hilly (but not mountainous area)… lightly accelerate before you start up the hill, so you can very gradually lose momentum, yet be at/above the speed limit when it the crest… this used to keep me from ‘getting into boost’, and crushing the effiency… I would suspect that this would be helpful with normally aspirated engines as well… of course how much you allow yourself above posted limits is determined by your risk aversion to LEO intervention…

  • avatar

    On a 4200lb AWD 5-spd manu-matic with a non-DI 3.6L v6, it’s much better to let the engine rev up a little bit according to the instant MPG meter. It’s several mpg improvement over putting your foot way down while maintaining too high a gear. In the range of 3 ~ 3,500RPM, depending on speed and hill steepness obviously, there seems to be a sweetspot.

  • avatar

    My gut reaction is to downshift. Lugging the engine always struck me as a bad idea and when I had manual transmission cars I always tried to keep the revs around the torque peak of the engine when climbing hills and just below for cruising. This worked out to around 2000-3500 for a VW Scirocco.

    • 0 avatar

      Good answer. That was what I tried to say, but I think you did it far better. Who cares if you lose 5-10MPG for the few seconds you go up a hill? That’s not the main concern here, no matter what engine/transmission is in question.

  • avatar

    I’m glad somebody finally asked the question. On my cars, I always get to the 5th speed as fast as i can, running around on the 5th at cruising city speed 50-55 km/h. I figured that since the car is already in motion, it did not need much torque to keep it up. Too bad my vibe doesn’t have a 6 speed, at 100 km/h the torque is at 2800 and i could use a 1500 rpm 6, downshifting to 5 for passing. I really wish GM or Toyota could offer that as an option…

  • avatar

    I have an Acura TSX with a 2.4L/6-speed manual. The owners manual suggests shifting approximately every 11 mph (11, 22, 33, 44, and 55 mph) for best economy. This keeps the RPMs peaking at about 3000 in each gear. Following this shift program will give you moderate acceleration.

  • avatar
    Greg Locock

    Didn’t read all the above but some people seem muddled

    IC engines are typically most efficient at their max torque speed, at about 90% throttle. You will not internally damage an unmodified production engine running at full throttle at relatively low rpm. You may damage the engine mounts if you run at too low a speed, say below 1300 rpm typically.

    The point about running closed loop vs open loop is good, but this is accounted for in the BSFC map (see wiki) for your engine.

  • avatar

    Hey guys! Thanks for answering my goofy/OCD question. It’s been banging around my head for a while with no real good place to ask it. So, it seems that the basic answer is, accelerate a bit more at the bottom of the hill, then downshift when appropriate, and don’t worry about the mileage (because for that one hill, it won’t really matter).

  • avatar

    Congratulations on training the wife to drive a stick. You sir, are my hero.

  • avatar

    I installed a factory trip computer in our ’90 Spirit. We frequently drive up a 2000′ foot climb with about an 8% grade. The Spirit has a 4-spd automatic. Left alone, it climbs the hill in 4th. With only one person in the car, putting it in 3rd raises the revs but drops gas consumption by about 15%.

    However, with 3 people in the car, lowering the gear has no benefit. So my answer would be that it depends on the car, engine gearing, load, hill steepness, and probably more. I’d guess that the best mileage would be at the engine’s sweet spot, but with only so many gears and an infinite variety of possible loads, it’s usually not possible to be in the sweet spot.

    btw, you know the advice to save gas by going slower uphill? If this were true, you’d get infinitely good mileage by climbing infinitely slowly, wouldn’t you? Try creeping up a hill and observe on the trip meter that mpg is worse than moderate speeds – up into speeds where you get a LOT of wind resistance. Then you need to use more gas to regain speed over the top of the hill. Most gas savings from driving slowly uphill are due to increased trip time, not by violating the laws of physics.

  • avatar

    Been here, done this. I live in a hilly area, and my mileage gets slaughtered in round-town driving from it. The answer, as you can guess, is not so simple. I have a 2004 Subaru WRX, 2.0, turbocharged, AWD with a MPG gauge that works off the OBD port. As the load on the engine increases, the benefit for the low gear drops off QUICKLY and you should downshift to get better mileage sooner rather than later. What I found was best was to keep a constant throttle position going up the hill, and downshift when you drop 5 or 10 mph. It was different for ever hill at every speed, and that was for my own car. There is no general rule of thumb that will work.

  • avatar

    One of the gaping problems with getting the most out of a tank of gas is the matter of time. Most aircraft pilots do not refer their range with the term ‘miles per gallon’ rather it is ‘pounds of fuel per hour.’ Which means that the most effecient setting of an aircraft engine might not yeild the least amount of fuel used, over a given distance because you’re in the air longer and therefore are consuming more fuel than if you actually were moving at a higher speed.
    Those of us old enough to remember experienced this first hand when the Carter administration inflicted the 55 mph speed limit upon us. What we discovered was that on a long trip of say 300 miles we started with the same amount of gas as we did with the 65 mph speed limit, but got to our destination with less gas in the tank. Why? Because the difference in gas mileage between the two speeds was offset by the amount of time spent on the road. So in reality the difference between 65 and 55 was about one or two miles per gallon, but the time difference to travel the same distance was 45 minutes to an hour, and so I could actually use more fuel on a long haul at 55 than at 65.

    • 0 avatar

      Ahem… If you get more miles per gallon at 55 rather than 65, time has nothing to do with it. You have fallen into a classic trap, I’m afraid. Time is irrelevant to an mpg figure. It’s miles per gallon, and miles and gallons have no time component in them. By your “argument”, at 30 mph, the amount of gas used on the trip would be even higher!

      I had a friend who thought that driving home, which is due north from here, was obviously uphill. So he always expected to get worse mileage going home for the weekend, rather than when he came back south for the work week. Don’t think he grasped the concept of sphere.

      Meanwhile, I conducted an experiment up a long fairly steep hill today. Speed 60 km/hour. Consumption in fifth, 19.2 litres/100 klicks. Fourth, 22.4, Third, 25.2. Coming back downhill, consumption in the same three gears, 2 l/100km. You guys can do the math. Fifth was most efficient. Subaru Legacy GT.

      I’ve noticed this many times on my instant economy readout. Lower gears eat gas, despite torque peaks, etc, etc. There is a reason why car companies make their automatic trannies grab that next highest gear as soon as possible, and it’s better fuel economy. Hardly a revelation. As long as the engine and gear combination can provide the necessary torque to maintain speed up the hill, the lower engine speed, and hence higher gear will give better economy, unless the fuel injection cycle has to “overcompensate” for a higher gear.

      LJK Setright used to chortle at the usual old rubrick about damaging a “lugging” engine. Old Wife’s tale. Mr. Locock above, an engineer in the automotive industry in the Antipodes, I believe, points out the real problem. The engine leaps around on its mounts and cannot run smoothly at too low an rpm under load (350 rpm in an old Chev stovebolt six, I’d say). My legacy will glide along at 1100 rpm in fifth lockup at 31 mph (50 klicks) on my secondary road behind Toyotas and Buicks, and accelerate smoothly from there if you don’t goose it too much. If you do, it shifts to second and rockets the hell out of there.

      I really fail to understand the mystery in all this. Pumping losses are lower at wide throttle openings. For the average modern overpowered car, a very high gear will usually provide all the torque required to go up most hills, and the lower engine speed means fewer injection cycles per second. The time to shift down is when, with foot on the floor, the car slows down anyway.

  • avatar

    There’s not one totally correct answer here. On older cars with carburetors and distributors then maximum efficiency was when the engine was at full throttle, at it’s torque peak RPM, and fully loaded. Those conditions are rarely met and seeking them in the real world is usually futile. It gets more complicated with fuel injected engines. I haven’t studied all of them, but most of the common electronic injection systems run the engine lean under most conditions; the computer seeks the air / fuel ratio that is best for the catalytic converter, not fuel economy (for what that tidbit is worth to you). When you floor the throttle, the computer will shift into a “power enrichment” mode and temporarily drop out of closed loop operation – this gets you a nice burst of power but it isn’t very efficient. On these cars, running full throttle at the torque peak and fully loaded results in very low efficiency and lots of unburned fuel going out the exhaust.

    If I had to come up with a “one size fits all” rule – use the gear that provides the lowest RPM as long as there’s still sufficient torque available to accelerate the car. If you hit full throttle and can’t accelerate it’s time to downshift. And try to avoid holding the throttle fully open for any length of time – it’s very wasteful.

Read all comments

Back to TopLeave a Reply

You must be logged in to post a comment.

Recent Comments

  • ajla: The Sienna is a much more expensive vehicle.
  • mcs: “Battery powered snowblowers, are darned near useless even for front porches in snow’y areas.” That...
  • mcs: “Having an electric mower would add a whole new dimensions to “Oops, I should have mowed this a few days...
  • lsstal01: A hybrid Maverick gets literally twice the MPG of a Taco. Not sure you want to bring Toyota into this lol
  • bullnuke: 45 years ago I lived in Tijuana and commuted to NAS North Island at Coronado, CA. Going south into Mexico...

New Car Research

Get a Free Dealer Quote

Who We Are

  • Adam Tonge
  • Bozi Tatarevic
  • Corey Lewis
  • Mark Baruth
  • Ronnie Schreiber