Archives

Up until about 6 months ago (when I got the Matrix and started looking into ways to increse fuel economy) I had always heard that accelerating slowly gives best fuel economy. After I started reading about things I heard that accelerating at WOT (manual transmission) would return best engine effeciency since you eliminate the throttle plate restriction. Of course you have to look at closed/open loop here too to make sure your air/fuel ratios aren't going into the toilet at WOT. But, you couple WOT with shifting at low rpms and you get the most effecient acceleration. This all made sense and I had been using this technique, and it seemed to be working. Now, at the Milwaukee meet I had a conversation with RLC and he suggested accelerating at roughly 2.5 liters per hour. I didn't get real in depth with it because he was instructing another driver at the time. So, now I'm asking what is the reasoning behind this?

Your economy is sub-optimal until you get into top gear (or the highest gear for teh speed limit). Charging hard to get there will waste gas and so will staying sub-optimal for too long. You have to find a sweet spot in the middle where your engine accelerates the mass most efficiently.

I try to accelerate with the LOD @ 50-55 until top gear speed can be achieved and then slower from there.

Hi, Daox! I remember that statement and what I was referring to was 2.0gph as seen on the ScanGauge. This equates to keeping your instantaneous mpg at half of your speed -- a rate suggested by Prius drivers as ideal.

In practice, I have found the best results to be around 1.7gph in my car. 2TJB is right -- there is a sweet spot and it is defined by getting yourself to the highest (most efficient) gear at a reasonable speed. Any faster and you use more gas because you've got your foot in it, any slower and you use more gas because you are taking longer to get to the cruising speed.

This is very much a seat of the pants sort of thing, and you should try some acceleration rates over a known stretch with a known starting state to see what works best. I used the 2 FASes out of my garage every morning up to the first light to try out some things.

I hope that gets the idea across? If not, keep asking more questions until the picture crystalizes for you.

You've been full of very good questions lately -- keep it up!

Yeah, I'm gonna need some clarifying. I can kinda see where your going with this, but still wanna clear it up a bit.

If the engine is running without restriction it is running more effeciently. This is assuming a stoimetric air/fuel ratio which my car pretty much keeps 95% of the time. So, if our air/fuel ratio is static and our throttle is not a restriction, then the only controlalble variable is rpm (via shifting). Torque will vary dependant on engine design and will effect fuel useage per rpm. Therefore, you still need to find where that sweet spot is as you say. However, I don't see how restricting the throttle could return better mileage. These are my thoughts on the subject. Please mention anything you see here that you find wrong.

Hm. I'll try but a lot of this for me is trial and error.

You'll do better with higher gears used sooner because you can keep the RPMs lower -- assuming that (within reason) lower RPMs over a given distance equates to less gas burned (fewer revolutions per unit distance?). If you travel too slow, the revolutions per distance could catch up to what you would have done by accelerating faster. Also, if you try to leadfoot your car you'll get to a certain point at which pressing the pedal harder won't actually get your car accelerating any faster. At this point, you are still increasing the flow of fuel but not getting any more power -- so are you still burning everything you are pumping in? Even if you are, your effective power is the same with less gas so you are using more than necessary -- right?

Those are the extremes that I can come up with, but I think that they should scale inward as you reduce the behavior from leadfooting and idling in gear to moderately aggressive and leisurely acceleration. To throw a math term at the problem, I see this as approaching an asymptote.

Back to you -- it's your turn to poke holes in my logic now. ;)

All I can tell you for sure is that you can run some tests and you'll find that the sweet spot will eventually surface in your results.

Sorry, but I'm gonna disect your post to answer.


Hm. I'll try but a lot of this for me is trial and error.

You'll do better with higher gears used sooner because you can keep the RPMs lower

Yes, I agree completely here. The lower rpm reduces the amount of friction the engine needs to overcome to put power to the wheels.


-- assuming that (within reason) lower RPMs over a given distance equates to less gas burned (fewer revolutions per unit distance?). If you travel too slow, the revolutions per distance could catch up to what you would have done by accelerating faster.


Again I agree. The equation is miles/gallon. Increasing miles faster is of much benefit.


Also, if you try to leadfoot your car you'll get to a certain point at which pressing the pedal harder won't actually get your car accelerating any faster. At this point, you are still increasing the flow of fuel but not getting any more power -- so are you still burning everything you are pumping in? Even if you are, your effective power is the same with less gas so you are using more than necessary -- right?

I have to disagree here. If the engine is in closed loop the oxygen sensor is controlling the air/fuel ratio. More air = more fuel, period. If no more air is going in, no more fuel is injected. The ratio stays the same. However, (I just thought of this while writing the response) ignition timing may be effected by throttle position... This could retard timing and result is lower power output. I'll have to do some testing to see. If this is the case you'd want just enough throttle to ensure the throttle plate is not the restriction. Any more than that and you'd start retarding the ignition and loosing power.



All I can tell you for sure is that you can run some tests and you'll find that the sweet spot will eventually surface in your results.

I definitly think that is what it'll come down to. I could type away all day, but the best way to find out is to do it.

What about Pulse accelerating on a strictly ICE P&G? Full throttle or something in the 1.7-2Gph range (4cyl. engine)?

Yes, I agree completely here. The lower rpm reduces the amount of friction the engine needs to overcome to put power to the wheels.

True the slower speed reduces the friction in the ICE and the transmission.... The Friction increases are not linear with RPMs... Lubricating fluids are less effective the faster the RPMs ... so the frictional coefficient goes up as speed goes up.... Thus more friction per Rotation.

Additionally the faster the ICE turns the more of the combustion event is lost, as the fuel has a limited flame speed.... Injectors are only so fast...etc...

These forces among others, result in every ICE having an ideal point of efficiency of converting the chemical energy of the fuel into mechanical torque.... It usually is well above idle but well bellow full throttle. in order to reach the ICE's best conversion efficiency point.

I have to disagree here. If the engine is in closed loop the oxygen sensor is controlling the air/fuel ratio. More air = more fuel, period. If no more air is going in, no more fuel is injected. The ratio stays the same.

Many vehicles intentionally change the air to fuel ratio... so more or less fuel can be injected at different times or under different conditions.... for instance many vehicles with fuel injection systems will injected a richer fuel to air ratio during start up of the vehicle than later, when the engine is up and running at speed and or temperature... Some vehicles reduce the fuel to air ratio at higher speeds for highway cruising.

Additionally it is actually impossible to get the same fuel to air ratio as the air velocity changes coming into the engine as the vehicle's speed changes... modern computer controls do a good job of their job... but as the vehicle speed changes the air density will change and the air flow rate will change the fuel injectors are not variable enough sensors and controls are not fast enough to be able to keep the same fuel to air ratio even if they wanted to.... they can do a good job... but it can not be kept the same.

------------------------

The Above only really deals with the ICE part of the mix....

Trying to get the best FE or MPG is much more complex than just trying to get the best ICE conversion efficiency.

Hard accelerating is a gamble ... it can pay off with improved FE / MPG under the right conditions... under other conditions the slower accelerating will net better FE / MPG.... learn your route and see which method you prefer.

P&G... is very effective in many conditions ( but not all conditions ) at significantly improving your FE / MPG... but it is more hazardous and dangerous than driving at closer to a constant speed... depending on who you are and where you drive the increased risks may be worth it or may not be worth it.


just my 2 bits.

This is the reason why slower acceleration (to a degree) is better: assuming flat engine efficiency and no drag, it takes the same amount of energy(fuel) to accelerate a car to a given speed regardless of how quickly it is done. This is because regardless of how you got to 65 mph, you have the same amount of kinetic energy at that speed. This is obvious given that the braking distance when you're traveling at 65 is always the same regardless of how you accelerated to get there. What this means is that by accelerating more slowly, you cover more distance while you are accelerating before reaching maximum speed, but because the speed you reach is always the same, you always burn the same amount of fuel. This means that under these idealized conditions an infinitely long acceleration would be ideal.

However in the real world cars don't have flat engine efficiency, and have high amounts of drag. Engines often have the highest efficiency at around 50% of their capacity, however this is a fairly fast acceleration in most cases. However an engine might have 90% of its max efficiency running at 25% of its capacity, so a compromise is made. Obviously you want the running at as an effecient capacity as possible, but you also to accelerate as slowly as possible, but the engine efficiency drops down to 0. These forces are contradictory, and so a balance must be found, for which the maximization is unique for each car. Throw in different engine speeds from the transmission, and you have yourself quite a mess.

The big area you want to stay out of is when you push the engine past that peak efficiency point to where the engine efficiency is dropping from pushing it over 50%, and you're losing potential distance from accelerating so quickly - a double loss

I don't really have much experience in accelerating properly to improve FE, but I understand the physics of it. Hopefully this will help you out:D

Elixer, I have to disagree. It does NOT take the same amount of energy to accelerate an object regardless of speed. This has nothing to do with the drag in the engine of your car -- it has everything to do with the mass of the vehicle and its consequent momentum. Roughly speaking:

F = Ma (where mass is constant -- otherwise it would be F = (dm/dt)a where a = dv/dt)

So the force scales in direct ratio to the acceleration. Since the energy required to move an object is defined as Work in this fashion:

W = FD (Where the force is in the same direction of travel the Distance is measured in)

we can see clearly from the above that the energy required to move an object is ALSO directly proportional to the force applied. Putting it together we have:

a = W/DM

Thusly, if you want higher acceleration you need greater energy. This is independent of the final momentum (which is what you were referring to when you directed attention to the coast down distance from a speed for some particular vehicle).

With this in mind, as well as the fact that the engine does not have a "flat" or constant efficiency independent of load, to get the most efficient acceleration rate we must balance the energy used to get up to speed with the efficiency characteristics of the engine. This is precisely why P&G works -- it seems counter intuitive but that is only because the engine is most efficient at a particular load.

I have obviously ignored all the frictions associated with travel in the real world but those are modifiers of the basic behavior illustrated above.

Note that the whole thing changes when the motive force IS roughly the same efficiency at many different loads -- this is the case with electric motors. There is a fall off of torque as RPM is increased but within moderate deltas you could consider it mostly a flat curve. P&G will NOT work for BEVs and we find (as expected) that if you want to get to speed most efficiently you need to pick a slow rate of acceleration. Higher rates will always use more power and will indeed reduce the range of the vehicle because they deplete the battery more quickly.

Sorry Sean, he's right.

Change in energy is only related to mass and speed.

KE = 1/2*M*V^2

You are also right in that F=MA, but that just means that faster acceleration requires more force. Do the integral and see for yourself that energy required is the same regardless of acceleration rate. A faster rate will require a larger force for a shorter distance, and a slower rate requires less force for a longer distance. The total work done is the same so long as the initial and final speeds are the same.

These facts only hold true in a frictionless world - in the same frictionless world, zero energy would be expended to travel.

Ah... I have assumed D to be a constant which would allow v to vary -- which is out of step with the proposed scenario. I agree -- as long as the final velocity is the same (and hence the momentum) the energy (work) required is the same.

This is one of the particular reasons EV's are so exciting for me, long time speed freak. Once we ditch the ICE, we lose crap like pumping losses as well! So in theory, fast acceleration is NOT less economical, it is only limited by the rate at which you can deliver that energy (power).

Then we just have resistive losses due to higher currents.

Higher voltages maybe? :D

I had not monitored my GPH during accleration until today. I've been obtaining pretty good results in the Accord so far and seem to be using an acceleration rate in the 1.5-1.8GPH range most of the time, edging up to 2.0 on a few occasions.

I don't think you can really use the work energy theorem for this. The amount of work done to achieve the given speed is the same, but the key thing to remember here is that this is the amount of work done by the engine.

It takes chemical energy from the gasoline to give the car the kinetic energy. The thing is that the chemical energy output from the gasoline varies based on the speed of the engine.

OK - so for accelerating, should you watch TPS, LOD, or GPH??? :)

OK - so for accelerating, should you watch TPS, LOD, or GPH??? :)

YES.

;)

Pick one and be consistent... :)

I don't think you can really use the work energy theorem for this. The amount of work done to achieve the given speed is the same, but the key thing to remember here is that this is the amount of work done by the engine.

It takes chemical energy from the gasoline to give the car the kinetic energy. The thing is that the chemical energy output from the gasoline varies based on the speed of the engine.

The general assumption made was that the engine's efficiency is relatively constant. This is not the case, of course, otherwise white knuckle stomping through the gears would be just as efficient as easing it up. The key things in accomplishing this are:

1. Keeping the ECU in closed loop (lean or stoich A/F, too rich otherwise)
2. Reducing pumping losses (throttle open enough)
3. Staying close to peak torque / high engine VE

If you can satisfy all of these, then acceleration rate can theoretically be increased without too big of a hit on FE.

Note that terribly slow acceleration in too low of a gear will fail 2 and 3...

The key things in accomplishing this are:

1. Keeping the ECU in closed loop (lean or stoich A/F, too rich otherwise)
2. Reducing pumping losses (throttle open enough)
3. Staying close to peak torque / high engine VE

If you can satisfy all of these, then acceleration rate can theoretically be increased without too big of a hit on FE.

Note that terribly slow acceleration in too low of a gear will fail 2 and 3...


That is a wonderful explination/outline.

The general assumption made was that the engine's efficiency is relatively constant. This is not the case, of course, otherwise white knuckle stomping through the gears would be just as efficient as easing it up. The key things in accomplishing this are:

1. Keeping the ECU in closed loop (lean or stoich A/F, too rich otherwise)
2. Reducing pumping losses (throttle open enough)
3. Staying close to peak torque / high engine VE

If you can satisfy all of these, then acceleration rate can theoretically be increased without too big of a hit on FE.

Note that terribly slow acceleration in too low of a gear will fail 2 and 3... Interesting - There aren't enough gauges, or eyeballs to watch them for that matter, so maybe the open-loop/closed loop will work for me. I keep average MPG, instantaneous MPG, and RPM gauges up. This leaves just one open slot. That's why I was asking TPS, LOD, etc, as the choice for gauge number 4. Most posts I've read seem to think TPS, with LOD as a close second. Maybe for accleration, open/closed loop will work better for me. Another thing to try. Eventually (hopefully) I'll get this car figured out!! Thanx! :)

I would use Loop only until you learn what the cutoff is. How much throttle does it require to go into Open Loop? Get a feel for that, then keep below that cutoff. Once you learn that, you can use another gauge in its place.

I know most Hondas, the Insight in particular, use speed-density algorithms to determine appropriate injector pulse width and timing advance. The density is from the MAP sensor, which is a fairly accurate estimate of the load on the engine, regardless of throttle position.

I don't have a scanguage, but I assume "LOD" is related to the actual load on the engine, and this is what you should monitor. You will find that there is probably some load threshold at which the ECU will switch between open and closed loop - once you get a feel for that threshold, you may not need to monitor load anymore.

If you can find a VE map for your engine, you can pin down the RPM ranges and loads where it is most efficient.

I would use Loop only until you learn what the cutoff is. How much throttle does it require to go into Open Loop? Get a feel for that, then keep below that cutoff. Once you learn that, you can use another gauge in its place. hmmm... wonder if open loop is based on TPS or RPM...now I'll have to find a stretch of road to "play" - Thanx!! :)

I know most Hondas, the Insight in particular, use speed-density algorithms to determine appropriate injector pulse width and timing advance. The density is from the MAP sensor, which is a fairly accurate estimate of the load on the engine, regardless of throttle position.

I don't have a scanguage, but I assume "LOD" is related to the actual load on the engine, and this is what you should monitor. You will find that there is probably some load threshold at which the ECU will switch between open and closed loop - once you get a feel for that threshold, you may not need to monitor load anymore.

If you can find a VE map for your engine, you can pin down the RPM ranges and loads where it is most efficient. Thanx! I will play with both Open/closed loop & LOD to see how they work for me. :)

Shifty, where does one find a VE map for their engine?

Shifty, where does one find a VE map for their engine? Yes, where DOES one find this? I would love to see a torque curve!!!

VE maps are not easy to "find". Usually you need a custom programmed scanner that can read the 3-d maps (tables) that are in the ECM. The ECM has to be "hacked" unless you have access to a scanner that already knows these items. You have to know the memory address and it's size. That is just the beginning. You also need to know how to convert the hex to decimal and what the scalar values are to make a "map" that makes sense to a human. Then after you have found the "map" and what the values are, you have to know what the map actually is. (spark, fuel, transmission pressure, shift speed, etc.)


AN ALTERNATE METHOD!

If you want to know what your VE is then start logging with something that can read the amount of air going into your engine. Record MAP, RPM and air flow (usually in grams/sec) Otherwise you will have to use fuel injector on times or something that you can extrapolate into air amount. When in closed loop the air-fuel ratio on most engines will be 14.7:1

Take the amount of air and devide it by the displacement of your engine. Put it in a table with MAP (manifold absolute pressure) and RPM as the Y and X. Put the air flow amounts into the cells and you should have a "VE Table".

Hope that makes sense. What you will find SHOULD match the engines TQ curve pretty closely. VE and BMEP follow together, which makes the TQ curve.

Oh good, it's that easy. Kinda like finding the albedo of the moon.

Like, can you get one from the manufacturer?

Oh good, it's that easy. Kinda like finding the albedo of the moon.

Like, can you get one from the manufacturer?

Not very likely. All of this stuff is proprietary. They don't want you futzing about inside there while under warrantee. Some companies are good enough to "hack" these things. Hypertech and other companies make a business out of this.

Shucks. But, thanks for the great information. Wish I had more ability, but the only way I cook ECU is "sunny side up."

Here is my understanding of fuel injection strategies, since people seem to be invoking these arguments but seem a bit foggy on the details. This may not hold true for every car out there, but I've not seen an exception in few fuel injected cars I've worked on in detail:

Cold start: vehicle operates open loop because 1) the o2 sensor hasn't heated yet so will not give a useful reading and 2) rich a/f ratio (AFR) desirable

Warm (most operation): closed loop operation because at approximately stoichiometric AFR, closed loop is the most robust and accurate method

Warm (near full throttle operation): open loop operation; to keep combustion temps down, minimizing chance of detonation and maximizing power, rich AFR is desirable.

Let me explain why non-stoichiometric target AFRs require open loop operation. Most o2 sensors out there are narrowband (lambda) sensors that are really sensitive around a particular point and not sensitive elsewhere. If you were to plot the voltage output vs AFR, you'd get something that looks like an inverted step function with the step right at a point somewhere a tad richer than 14.7:1. If you go even a bit leaner than this, you get a huge change in voltage in one direction, if you go a bit richer, you get a huge change in the opposite direction. After that initial change there really isn't much sensitivity.

So to maintain a non-stoichiometric AFR, a fuel injected car needs to either use a different type of o2 sensor or needs to ignore the sensor completely and just listen to other sensors, like the TPS, MAF, MAP, IAT. Which ones exactly depend on the car, of course. For example, speed density cars like many hondas won't have a MAF sensor. Such a car will look up a table which compensates for the variable volumetric efficiency under different running conditions and tells the injectors how long to open. Cars that use a MAF sensor primarily don't need to consider their VE.

Bottom line, don't go WOT because the car will intentionally be less efficient. However, I'm unclear on where this transition occurs. Does the scanguage tell you whether you're operating open or closed loop?

addendum:
Let me clarify my understanding of what a speed-density engine computer will generally do in open loop operation:

You want to know how much air you're going to have in each cylinder so you can add the appropriate amount of fuel.
Recall the ideal gas law: PV=nRT
You want to know n. R is a constant, V is the volume inside the cylinder, P you measure with a MAP sensor, T you measure with an IAT sensor.
So it is straightforward to calculate n, with one caveat.
This is an ideal case, and since we are in the real world we need to consider the non-ideality, which is what the volumetric efficiency represents.
Bonus question left as an exercise for the reader, is it possible for the volumetric efficiency to be >100% and if so, how?
So you scale V by your volumetric efficiency and calculate n.

There are other details to deal with such as that the amount of fuel passed by an injector is not linearly proportional to how long it is opened, and short term and long term fuel trim to account for other variations in the model but you get the idea.


This is all background, let me get to something more relevant:
Sometimes you can find a published chart of an engine's brake specific fuel consumption. That is, how much fuel is used to produce each hp. This is plotted as a function of engine speed. I'm almost certain it doesn't exactly track the torque curve but I think it's pretty close. I haven't seen enough of these graphs to confidently make a generalization. But if this holds true, then screw finding the VE table for you car (it's probably not going to happen), and just look at a standard dyno output, which you should be able to find for any engine.

Okay, I'll take a stab at it - where can we find a "standard dyno output" for our engine?

Bottom line, don't go WOT because the car will intentionally be less efficient. However, I'm unclear on where this transition occurs. Does the scanguage tell you whether you're operating open or closed loop?



From my understanding, Yes the ScanGuage can tell you open or closed loop.



Bonus question left as an exercise for the reader, is it possible for the volumetric efficiency to be >100% and if so, how?


Yes it is very possible. If you have an engine that makes more TQ in ft-lbs than it's cubic inches it likely has over 100% VE in those areas. That is a rough estimation, but it usually works. Obviously turbocharging or supercharging produces over 100% VE when running any significant amount of boost. Other ways are from acoustic tuning methods that include cam overlap and timing, intake runner length and volume, plenum volumes, exhaust manifold size and length, etc. At certain RPM's these items when designed properly can create high and low pressure zones at or near the cylinder head that can actually "force" more air into the cylinder than it can physically hold. This would be more than 100% VE.



The rest of the info given was good and accurate. The only thing missing is that now there are wide band sensors that can accurately simulate the actual AFR in the exhaust so that engines can accurately run in closed loop and use rich or lean AFR's during power enrichment and lean cruise modes. Exhuast converter designs used to limit the AFR ratio to 14.7:1 so that they would not burn out. Straying from that AFR could only be done for short periods so that it was not damaged. I have not kept up with converter designs lately and I'm not sure if this is still the case, but I'd guess that it is.

Yes, the Scangauge will give open/closed loop. On my older Honda, it only goes open loop if the pedal is ON the floor. 99% throttle or less is closed loop. I can't say about others, though.

I've seen BSFC maps of multiple engines, and they mostly agree that mid-low rpm and mid-high throttle are best. My personal experience bears this out as being the best for the Pulse part of P&G.

There's a conceptual BSFC map here (http://techno-fandom.org/~hobbit/cars/SAE-bsfc.gif), from an SAE paper discussing
the early prototypes for electronic throttle control in the
Prius. The little numbers on the contours indicate relative
consumption per useful work produced. The Prius ECUs make efforts
to stay above the thick black line, requiring that we come right
off idle into high-torque and stay there through a wide RPM
range, to crawl as high on the "hill" as possible whenever the
engine is running at all. Ecodrivers who have direct throttle
control know how to do this manually.
.
The "sweet spot" for the real-life Prius engine is actually a
bit lower than the chart implies, centered more around 2000 RPM
from what the collective brain can determine.
.
_H*

Here's a huge-long study on engines and fuel economy: http://www.tc.gc.ca/programs/environment/climatechange/subgroups1/vehicle_technology/study2/Final_report/Final_Report.htm

Search down to "Techniques for Improving Fuel Economy" and see the chart there.

Yes it is very possible. If you have an engine that makes more TQ in ft-lbs than it's cubic inches it likely has over 100% VE in those areas. That is a rough estimation, but it usually works. Obviously turbocharging or supercharging produces over 100% VE when running any significant amount of boost. Other ways are from acoustic tuning methods that include cam overlap and timing, intake runner length and volume, plenum volumes, exhaust manifold size and length, etc. At certain RPM's these items when designed properly can create high and low pressure zones at or near the cylinder head that can actually "force" more air into the cylinder than it can physically hold. This would be more than 100% VE.

The latter what I was thinking of.

Technically, compared to static conditions, forced induction will give you >100%. But for the purposes of the speed density calculation, running turbocharging and supercharging is going to increase P (and T to a lesser amount) so it's not clear that you'll get any improvement. What I mean is, the engine computer takes into account the P and T measurements, but is still going to scale V (the static volume) by a similar amount compared to the naturally aspirated case.


The rest of the info given was good and accurate. The only thing missing is that now there are wide band sensors that can accurately simulate the actual AFR in the exhaust so that engines can accurately run in closed loop and use rich or lean AFR's during power enrichment and lean cruise modes. Exhuast converter designs used to limit the AFR ratio to 14.7:1 so that they would not burn out. Straying from that AFR could only be done for short periods so that it was not damaged. I have not kept up with converter designs lately and I'm not sure if this is still the case, but I'd guess that it is.

Yes, I did mention wideband sensors, although not by name, but I think you're not exactly correct about catalytic converter design. It's actually the newer three way catalysts that must stay around stoichiometric more closely. This is because unlike older cats which were only reducing or oxidizing, the newer cats do both but can't do both at the same time. Thus you have to alternate between being slightly rich and slightly poor, otherwise either your CO emissions or your NOX emissions will skyrocket. I think in the past NOX emissions were not as tightly controlled as today so that enabled you to do things like lean-burn stratified charge and not worry so much about catalyst operation. Of course going rich on any catalyst for an extended period of time will cause it to fail eventually, but it's always been that way.

Bonus question left as an exercise for the reader, is it possible for the volumetric efficiency to be >100% and if so, how?

Available from Chrysler in 1960.

http://www.chrysler300club.com/uniq/allaboutrams/jengine3.JPG

How Ram Induction Works

This is how ram induction obtains its "free" supercharging: 30-inch-long ram tubes carry a mixture of air and fuel at a high rate of speed directly to each combustion chamber. As the intake valve on the combustion chamber closes, the inertia of the fuel-air mixture set up in the long tubes ram an extra amount of the mixture into the chamber. At the same time, a sound wave is created in the tube, with the compression wave calculated to be at the intake valve just before it closes. The compression wave sends still another extra amount of fuel-air mixture into the chamber.

These two "bonus" supplies of extra fuel-air mixture forced into the combustion chamber, account for the ram or extra power effect.

Hi All:

___After trying to determine the BSHP-fuel maps sweet spots, I said screw it and found my own with real world and open road numbers as you all will with an SG-I or II available to you. If you FAS, you can P&G with varying rates and ranges (from a given bottom speed of 20 mph for example) and after each pulse with a different rising RPM to a given speed, shut down and when you come to the same lower end, look at your aFCD. Did it climb or fall from the previous result. You can do this to figure out how “hard” you need to step into the pedal for a near peak FE for a single P&G cycle. In the Ranger for example, an 18 mpg low to a 36 mph high with a lightning fast second and third gear shift will give 65 + in the summer and the accelerations are up in the 70 – 90% load regions with an ~ 1,500 – 2,300 RPM range. This only works for P&G. If you accelerate that hard to a steady state, you have thrown away a lot of fuel to arrive at your vehicles great steady state cruise - minimal FE area. For a std. accel to a speed, setup a mile marker program and accelerate using varying rates up to said speed and hold. Once you hit the mile marker, mentally record your aFCD. Repeat with alternating accel ranges What I have found is a very lazy accel below 2,000 R’s shifting or not (Accord and MDX w/ Auto's) through every gear yields the best result.

___The Prius is a special animal as it has some interesting actions that practically make no sense as its steady state FE in the 25 – 40 mph range can be less then its steady state FE in the 43 – 53 mph range. A lot less in fact! Bring those up to 15 – 20 mph under a very light EV accel, light her off and use no more power then the mimic will allow before the pack draw shows up. For steady state cruise, 1,280 – 1,380 RPM’s at 50 mph (SHM) appears to allow the best FE with a steady state 70 + mpg during the summer months on mildly rolling hills and flats. Sometimes in the winter too ;) Running the Prius up anywhere near the 2,000 RPM range other than in a P&G scenario is beginning to throw it away and in a hurry under the same conditions. Accelerating to a highway steady state cruise other then the std. 0 – 20 + under EV is a real SOB with those and a warm-up P&G routine with ever higher speeds on the tops and bottoms appears to work well in my experience … That will not work with an on-ramp to an Interstate however :(

___Instead of trying to deduce this stuff from a particular vehicles dyno charts while in the perfect lab environment on a calibrated dyno and probably with the tires at door specs, brand new, with a std. oil etc … perform the tests on your vehicle on your roads and with your own setup. Find out where the peaks and valleys lye for yourself and I can guarantee you will blow any of the charts out of the water. Ranger P/U’s are not supposed to see 80 + mpg under any scenario but for some reason, I know one that has? Prius’ are not supposed to see > 120 + mpg’s at relatively low RPM’s but for some reason I know of at least one that has? Insights are not supposed to be able to receive 200 mpg’s at light loads or slow speeds but for some reason … If everyone were using the charts as the automobile companies do, there would be a lot less variance in FE from one driver to the next other then top speeds, ambient conditions and period to closed loop ops (warm-up hit).

___Good Luck

___Wayne

Wayne, I'll take your word on accel rates up to cruising speed. You certainly have the miles and mpg's to back that up. ;)

Once you're up to speed, I linked to this chart (http://www.tc.gc.ca/programs/environment/climatechange/subgroups1/vehicle_technology/study2/Final_report/image/Final_28.gif) from the document I linked above, to show exactly WHY P&G works so well. The dots are various points on the EPA city mpg test, and you can see that most of it is in a very inefficient region. P&G puts a lot more of the usage in the peak area. This extracts more work from the same amount of fuel, compared to what's on the chart.

I totally agree with Wayne's approach. In many situations I have been confronted with similar issues - you have a complex system that is pretty easy to measure and test. Having a solid background understanding is crucial but then you have to go out and see what the system does in real life, not just in a lab setting. Pre-facto discussion is important however to reduce the number of cases you try to a manageable amount.

I really like that canadian document, btw. Very informative, and nice charts. Although putting on my academic research cap for one minute - I'm wary of how their charts were derived. They say they used data from several representative vehicles to approximate the average car/engine. However when they do their comparisons (2v vs 4v vs vvt vs supercharging, etc), I'm concerned that the vehicles in the different groups have too many other uncontrolled differences to make the comparison accurate.

Instead of trying to deduce this stuff from a particular vehicles dyno charts while in the perfect lab environment on a calibrated dyno and probably with the tires at door specs, brand new, with a std. oil etc … perform the tests on your vehicle on your roads and with your own setup.

+1

Definitely the best thing to do!

JMH

Hi, I just noticed this topic and joined to give my $.02 worth.

Yes, kinetic energy is kinetic energy. If you are going to accelerate to a given cruising speed, you might as well get it done with the engine running in its thermodynamic sweet spot.
But, to really understand why it is so right to step on it to get up to cruising speed, you have to look at why taking forever to get up to cruising speed is so wrong.
Let's imagine a three mile trip on a highway with a 60 mph speed limit. Thinking that accelerating super slow is the most efficient way to drive, a driver takes an entire mile to get up to 60, then cruises at 60 for the next mile, then gradually slows down to zero on the last mile. To make the math simple, lets assume that the speed increase and decrease is perfectly linear. Here's what results. The first mile takes two minutes to cover, the second mile takes one minute to cover, and the last mile takes another two minutes to cover. That's five minutes for the three mile trip, a moving average of only 36 mph. Do you really think that's the most efficient way to average 36 mph? You gave the car 60 mph worth of kinetic energy and fought 60 mph air resistance for an entire mile to average only 36 mph.
Now let's leadfoot it to a cruising speed of 38.4 mph in only 1/10 of a mile and stop just as fast. This also gives a moving average of 36 mph but you are fighting only 38.4 mph wind resistance and only giving the car 38.4 mph worth of kinetic energy.
By stepping on it when you accelerate, you can stay 'on schedule' with a lower cruising speed. Remember, air drag quadruples when the speed doubles.

I first stumbled on this principle while riding bicycles. I learned that I could maximize my moving average by attacking acceleration and uphills with everything I had, and letting myself catch my breath and rest up for the next uphill on the downhills by just letting the bike coast.

Hi BLE:

___Welcome to CleanMPG!

___We are not taking about accelerations from 0 – 60 in forever but we are talking about usually slow and steady. Bailout has said it is best to ramp up to speed quickly in his Yaris but it revs relatively quick anyway so that may be an exception? Please try and put the analyzing on the back burner and get out there and prove it.

___To many tens of thousands of miles and hundreds of posts have been performed disproving what you have stated as the proper way to do things. The real world results has always spoken for themselves and we are all about real world results here at CleanMPG.

___Good Luck

___Wayne

According to my SGII .... when I accelerated at like 3/4 throttle in a high gear... I would consistently get 12.5mpg. I do much better than 12.5 mpg accelerating with low throttle and a lower gear.

I'm wanting to set up some kind of test for this theory. I have a ScanGauge, currently set up to view instantaneous MPG, current trip MPG, water temp (cause I used tips on this web site to design a radiator block - thanks guys), and trottle percent (TPS). I could do a couple of tanks with specific acceleration (using rpm, tps, lod, and instant MPG), and then do a couple more tanks with a different acceleration. We'd have to put some thought into shift points, and perhaps highest speed and highest and lowest RMPs. Could clearer minds than mine come up with some parameters that perhaps a bunch of us could drive within, to determine which is best, rapid acceleration to speed, or slow acceleration to speed?

Hi Kwj:

___This one will be easy for you.

___Mark off a 1 mile start stop on a relatively flat and trafficless road. From a stand still, reset your Today’s aFCD and accelerate using the low RPM quick shift from first to fifth and than hold at 50 mph until the 1 mile point. Check and record your aFCD once you cross the 1 mile point. Now do the same for a mid and high RPM launch and hold to the same the same. Than repeat the series over one more time. Always in the same direction to account for wind and repeat the test within minutes of one another so ambient conditions do not change and skew the results. I used this acceleration test method for Tarabell’s HCH-II article and it works like a charm ;)

___Good Luck

___Wayne

This is an excellent thread boys. Now could someone translate it into English?


--des
mechanically challenged

Hi Des:

___Which part :D

___Good Luck

___Wayne

Hi BLE:

___Welcome to CleanMPG!

___We are not taking about accelerations from 0 – 60 in forever but we are talking about usually slow and steady. Bailout has said it is best to ramp up to speed quickly in his Yaris but it revs relatively quick anyway so that may be an exception? Please try and put the analyzing on the back burner and get out there and prove it.

___To many tens of thousands of miles and hundreds of posts have been performed disproving what you have stated as the proper way to do things. The real world results has always spoken for themselves and we are all about real world results here at CleanMPG.

___Good Luck

___Wayne

I'm a new owner of a 2008 Yaris and I got 45.29 mpg on my last tank in the real world. The car isn't fully broken in yet and I'm still learning what it wants.
I usually accelerate rather briskly to a lower than average cruising speed instead of slowly ramping up to a normal cruising speed. My goal is to keep my top speed as low as possible while keeping up with the average speed of the traffic. The brisk acceleration makes it possible to achieve the same average speed with a lower top speed. I guess you can call this a variation of pulse and glide.

In the example I gave a few posts ago detailing how a car that ramps up to 38.4 mph in 1/10 of a mile achieves the same average speed as a car slowly ramping up to 60 over a mile on a 3 mile trip, I decided to plug in some numbers just for fun.
Let's assume a 3000 pound car that has 100 pounds of air resistance at 60 mph.

Energy needed to accelerate to 60..360,745 ft lb
Overcoming air resistance during cruise, 528000 ft lb
Air resistance during the ramp up, 176,000 ft lb
Total 1,064,745 ft lb

Energy needed to accelerate to 38.4 mph, 147,761 ft lb
energy needed to overcome air resistance, 2.8 miles at 38.4 mph 605,553 ft lb
air resistance during acceleration 7,209
Total 760,523 ft lb.


I know this is all academic but this thread seems to be about optimum acceleration with a lot of details about what loads the engines can work at most efficiently. But you also have to look at the load end of the equation and there, the laws of physics favors the lowest top speed for a given average speed. Will you get better mpg by ramping up slower? Sure you will but you also will lower your average speed and there is an even more efficient way to lower your average speed.

Keeping your average speed up without high peak speed means it's just as important to never go slow as it is to never go fast and most people have no idea just how much going slow impacts the average speed.
Try this brain teaser on your buddys. You walk to the store at one mile per hour, how fast do you have to jog back to raise the round trip average speed to 2 mph? I bet most of them will say 3 mph because the average of one and three is two but that answer is wrong.

Hi, I just noticed this topic and joined to give my $.02 worth.

........

I first stumbled on this principle while riding bicycles. I learned that I could maximize my moving average by attacking acceleration and uphills with everything I had, and letting myself catch my breath and rest up for the next uphill on the downhills by just letting the bike coast.

I've noticed the same thing on bikes--it makes a big difference when you're riding a couple hundred kms. I suspect that this is one of the secrets to success in obtaining good FE around high hills because it reduces the wind resistance cost of the high energy (including gravitational potential!) portion of your pulse/glide cycle. Put simply: for big hills, pulsing uphill and gliding downhill is a good trick to reduce total wind resistance. Unfortunately, the wind resistance math is against you on level ground, because the lowest total wind resistance for any given average speed occurs at constant speed at the average speed (this is also consistent with my empirical bike riding observations).

Trying to figure things out using the dyno maps is difficult for a number of technical and mathematical reasons I can think of. (If somebody wanted to pay me to try though.....) I'm lazy, so I'm gonna use Wayne and anybody else's empirical data--particularly if they have info about , say, 2003 4 cyl MT Rangers:p.

Thanks Wayne. I'm now thinking of where I could do as you suggest. Things like "low," "mid," and "high" are fine for you pros. Would it help to select a constant throttle setting and constant shift points, so others could do this with some assurance that our results would be reliable, thereby better supporting one theory or the other?

Hi BLE:

___I took Brian’s Yaris out for a 50% highway/50% city/suburban drive from dead cold in Reno, NV last week and hit 58.0 the first time I had ever driven the vehicle. It is worth 65 around town and 55 on the highway from my quick take behind the windscreen. Climbing the mountains like Brian does, whatever you get climbing and add the descent for free about does it …

___Rich, you have done quite well in your own Ranger this year IIRC too ;)

___Kwj, I cannot wait to see your own results!

___Good Luck

___Wayne

I'm a new owner of a 2008 Yaris and I got 45.29 mpg on my last tank in the real world. The car isn't fully broken in yet and I'm still learning what it wants.
I usually accelerate rather briskly to a lower than average cruising speed instead of slowly ramping up to a normal cruising speed. My goal is to keep my top speed as low as possible while keeping up with the average speed of the traffic. The brisk acceleration makes it possible to achieve the same average speed with a lower top speed. I guess you can call this a variation of pulse and glide.

In the example I gave a few posts ago detailing how a car that ramps up to 38.4 mph in 1/10 of a mile achieves the same average speed as a car slowly ramping up to 60 over a mile on a 3 mile trip...the laws of physics favors the lowest top speed for a given average speed. Will you get better mpg by ramping up slower? Sure you will but you also will lower your average speed and there is an even more efficient way to lower your average speed.

Keeping your average speed up without high peak speed means it's just as important to never go slow as it is to never go fast and most people have no idea just how much going slow impacts the average speed.
Try this brain teaser on your buddys. You walk to the store at one mile per hour, how fast do you have to jog back to raise the round trip average speed to 2 mph? I bet most of them will say 3 mph because the average of one and three is two but that answer is wrong.

As for the brain teaser, it is impossible to average 2mph unless you can get back home instantaneously.

I don't see how you are achieving about the same average rate of speed as traffic, most of the drivers out there are already using your method accept they keep going full throttle until they hit 5-10mph above the speed limit and then jam on their brakes at the next red light, leaving your Yaris in the dust.

I don't think anyone is prescribing accelerating a mile to hit 60.

I did some tests last year on my daily commute of 9.8 miles using gentle, 3/4, and full throttle each 3 times, here are the averages.

Gentle throttle 1-2-3-4-5 shifting 2500 or less 33.4 mpg.
3/4 throttle shifting at 4000 1-2-5 shifting. 28.8 mpg.
WOT shifting at 5000 1-2-5 shifting 27.8 mpg.

I didn't, as you are suggesting, cut me top speed back on the 3/4 and full throttle runs to lower my average speed to that of the gentle throttle but I see no way that that would have improved the difference by enough. I'm only going 45 as it is, if I went any slower I need 4th gear not 5th, and that's a net loss.

It is possible that I could improve the brisker throttle FE by shifting lower but that much throttle at my usual low shift points would produce more lugging than I want.

Do I understand you right that maximum braking is also the best method?

Alright, I've been reading up on this stuff a bit now and it has reinforced what I claimed at the beginning of this thread. Lower BSFC (brake specific fuel consumption, or power per amount of fuel used) is achieved at the highest possible torque output assuming a constant air/fuel ratio. This means accelerating at WOT (or very near it so you stay in closed loop mode) WILL achieve the best fuel economy. This is why pulse and glide works so well fellas. Your engine is running most effeciently when it is able to avoid pumping losses, and operate at maximum torque.

Thankfully, most engine today have a fairly flat torque curve. Here is an example from my 1993 Toyota Tercel. Going from this I would achieve the best engine effeciency keeping my rpms between 2250 and 3600. This seems a bit high for some of you, the truth. I'm sure friction does have some say in this, so I would definitly say to stay toward the lower end of this if at all possible.

http://www.tercelreference.com/tercel_info/engine_info/3ee_dyno_1.jpg

Hi Daox:

___Forget the torque curves or BSFC numbers and let the real world tell you what is worth it or not. You have the SG-II and with less than 30 minutes of testing, you will have the numbers in your hands ;)

___Good Luck

___Wayne

I agree testing to prove theory is useful. Later today I'm planning on taking some time to procure the Paseo's actual torque curve. From there I'll be doing some runs to compare fuel consumption at different rpms.

The horsepower/torque graphs we usually see show the engine's torque and horsepower at WOT because they are measured by people who are interested in how fast a car can go.
When engines are run with the throttle only partly open, both the maximum torque and maximum horsepower peaks shift to a lower rpm. To give you an idea of just how dramatically the torque curve changes when the throttle closes, consider an idling engine. An idling engine can indeed deliver some torque and some horsepower to an external load but to do so, it will be running at an rpm that's lower than its free idle speed. Above the free idle speed, the engine is actually braking.
A really comprehensive dyno test should show about 10 different torque curves taken at 10 different throttle openings and have them displayed on a 3-D map.
Once you study such a map, you begin to understand why it is a mistake to always have the engine running at its "torque peak rpm" when the engine is throttled down to 10 horsepower or so.

Hi BLE:

___I took Brian’s Yaris out for a 50% highway/50% city/suburban drive from dead cold in Reno, NV last week and hit 58.0 the first time I had ever driven the vehicle. It is worth 65 around town and 55 on the highway from my quick take behind the windscreen. Climbing the mountains like Brian does, whatever you get climbing and add the descent for free about does it …

___Wayne


I doubt that I will be able to match that mostly because I live in Austin, TX which is about 600 ft above sea level. Reno is about 4500 ft above sea level I believe. (I Googled it)
Some of the best highway fuel mileage my wife's Honda Element ever got was in New Mexico which is 5000 to 8600 ft above sea level. Not only do you have less air drag at that altitude but the less dense air also effectively reduces your engine's displacement.

This plastic bottle was fully inflated in Taos, NM. The picture is what it looked like when I got back to Austin TX.

http://i170.photobucket.com/albums/u260/ben_eberle/Sipapu07093.jpg

Hi BLE:

___How can I put this gently … The 3,400 # Accord w/ the 2.4L and an Auto pulled a 45.1 last night in 28 degree temps at 600 ft. of elevation. Her last tank was 45.1 over 840 miles of snow storms, blizzards, Minus 10 to + 30 degree temps and an Autobody shop pulling an amazing .3 mpg over 4 miles :angry: 55 + is a gimme in warmer temps with any distance of note. A Yaris w/ a stick will bury the Accord in every traffic condition or scenario without even trying.

___Do not sell yourself short as you seem to be missing what we actually do here. I hope for your sake you are realizing what is available irregardless of any preconceived notions about HP and Torque curves or BSFC maps. Once you let all the tech detail go and begin driving for FE, the artist in you will bury the tech every time. The Yaris is easily good for 60 + in warmer temps so please do not believe for a second an elevation number is holding you back. The excuses have to be thrown into the garbage yesterday and its time for the real world EPA busting actuals to begin.

___WRT a crushed plastic bottle, have you read "The Centennial Greatest Automobile Race Reenactment" thread? What I found interesting is not the crushed and bloated plastic bottles as we climed and descended but how fast the carbonation disappears at higher altitudes ;)

___Good Luck

___Wayne

For Daox/BLE,

What I have found w/my Celica is that it's steady state MPG sweetspots coincide w/the torque dips, not the peaks.

But, what Wayne says is true. Using hypermiling techniques like an artist will bury the tech. They certainly prove it every day;)

As for the brain teaser, it is impossible to average 2mph unless you can get back home instantaneously.

I don't see how you are achieving about the same average rate of speed as traffic, most of the drivers out there are already using your method accept they keep going full throttle until they hit 5-10mph above the speed limit and then jam on their brakes at the next red light, leaving your Yaris in the dust.

I don't think anyone is prescribing accelerating a mile to hit 60.

I did some tests last year on my daily commute of 9.8 miles using gentle, 3/4, and full throttle each 3 times, here are the averages.

Gentle throttle 1-2-3-4-5 shifting 2500 or less 33.4 mpg.
3/4 throttle shifting at 4000 1-2-5 shifting. 28.8 mpg.
WOT shifting at 5000 1-2-5 shifting 27.8 mpg.

I didn't, as you are suggesting, cut me top speed back on the 3/4 and full throttle runs to lower my average speed to that of the gentle throttle but I see no way that that would have improved the difference by enough. I'm only going 45 as it is, if I went any slower I need 4th gear not 5th, and that's a net loss.

It is possible that I could improve the brisker throttle FE by shifting lower but that much throttle at my usual low shift points would produce more lugging than I want.

Do I understand you right that maximum braking is also the best method?


Bingo! You got the right answer to the brain teaser!

No, the guys accelerating full throttle to the next red light don't leave me in the dust. The red light makes them wait for me. Really, I try my very best not to get to the next light early enough to need to stop. Nothing blows your average speed like spending half a minute going zero mph.

The example I gave was more to show the physics involved with two different ways to achieve the same transit time, I didn't mean to claim either one was optimal, only to show that you can accelerate too slow for optimal FE and it's not all about where the engine is most efficient.

As for stopping, I try my best not to use my brakes to stop. Letting a hill turn my kinetic energy into potential energy is good. Using my remaining kinetic energy to finish the trip is good too. A brake stop from 60 throws away as much energy as 16 brake stops from 15 mph. There's a diminishing return here and so I don't sweat it if I still have 10 or 15 mph left over when I reach the stop, but I absolutely hate to do a 60 mph panic stop.


The biggest improvement in my FE was finding an alternate to taking 12 miles of interstate highway home. The "scenic route" is also about 2 miles shorter. That 45.29 mpg I stated was from a tank that included several trips done the old way so my actual improvement may be more. I have to wait for the next fill up to be sure.

I'm really more interested in achieving a normal commute time as efficiently as possible than I am in setting a new FE record while winning a slow race so I may not reach some of the records posted by others.

Y'know, it all depends on the car and the engine. Every one
will have different sweet spots. The *general* theory can be
had from the BSFC maps and such, but the Yaris will be different
from the Accord and the Ranger and the Prius and the Insight
and everything else in the fine details.
.
Note that the maps do not in any way imply "full throttle", either.
They imply making the engine work, certainly, but not against
the wall. In fact 75% throttle is a figure often bandied about
on here, and is pretty much how the fly-by-wire throttle control
on the Prius is done and that system seems to work quite well.
The driver can still help by anticipating the upcoming terrain
that the car can't predict and mentally calculating the "energy
envelope" required to get from here to there [such as the next
red light]. It's managing that overall energy to move from
point A to point B where the big gains come in. If that means
brisk or gentle acceleration FOR YOUR OWN CAR, so be it, but
that will take some exploration.
.
_H*

Bingo! You got the right answer to the brain teaser!

It was obvious once he mentioned it, at least to me. Then again, the "3 mph" suggestion stuck that number in my head... But that's another show. ;)


The example I gave was more to show the physics involved with two different ways to achieve the same transit time, I didn't mean to claim either one was optimal, only to show that you can accelerate too slow for optimal FE and it's not all about where the engine is most efficient.

So it sounds like we're all on the same page, then. Taking forever to get up to speed is sub-optimal, but so is running with your foot on the floor. There is a "happy medium" (though in this case, more of a "local maximum") for MPG over time.

There's a thread on one of these forums about a guy who changed from gentle acceleration and shifting very short, to a (slightly) more aggressive acceleration and somewhat higher shift points. He tried to maximize the instantaneous MPG gauge in his dash, and the overall effect was better FE over all. (It was some form of BMW, I think.) This is another example of the fact that you can err on either side.

-soD

Hi BLE:

___If you were interested in max FE, you are only about half way there. If you want to drive for high FE, again, you have to throw out the BS and begin to drive for FE instead of relying on a number as a crutch. If you are happy with 45, great. I can say there is a lot more but it depends on when you are willing to learn and practice and if you want it or not?

___Hobbit, with the hybrids, the pack gets in the way at anything near 75% TPS or throttle plate opening, so I would not recommend going there unless it’s an emergency in a Prius or IMA equipped hybrid. Your Overall FE will be nowhere near peak at that rate :(

___Good Luck

___Wayne