Is "Pulse & Glide" riding more efficient?

[xyster]

Discussion branched from Andrey's question here:
http://endless-sphere.com/forums/viewtopic.php?p=4416#4416

Compared to maintaining a single, intended speed, is it more or less efficient to "pulse" the throttle to get to a faster speed then one intends to ride, then coast down below that intended speed, and then pulse again ad infinitum?

http://www.metrompg.com/posts/mpg-miscellany.htm#pulse

"> Does pulse and glide help EV's?

Yes, it can if done properly. But you have to understand why it works to do it properly.
ICEs have a strong efficiency peak at some particular horsepower, which is usually some substantial fraction of full horsepower. For example, a 100hp engine might be most efficient at 50hp. But if you run it at 10hp or 100hp, it is less than half as efficient (uses twice as much fuel per horsepower).
The trouble is, the vehicle itself is most efficient at low speeds, where wind resistance is negligible. So the slower you go, the more efficient the vehicle gets but the less efficient the ICE is.
If you drive at a constant speed, then the best fuel economy occurs where the product of the vehicle and engine efficiencies reach a peak. This typically occurs at 30-40 mph. But the vehicle only needs 5-10hp to drive at this speed, so its efficiency is less than half its peak
efficiency.
Fuel economy champs found a strategy to beat this problem long ago. The trick is to always run the engine at its peak efficiency (say 50hp), or shut it off entirely. But keep the vehicle speed as low as possible, so aerodynamic losses are negligible. Start at some very low speed like 10 mph. Put it in gear so the car's momentum starts the engine. Run the engine at 50hp, which makes the car accellerate strongly. When you reach the speed where wind resistance losses begin (like 40 mph), shut off the engine, take it out of gear, and coast back down to 10 mph. This is "pulse & glide". On any car, you can at least *double* your fuel economy by driving this way.
The same principle applies to an EV, except that the electric motor has a much flatter efficiency vs. horsepower curve. A motor might peak at 85% efficiency at 20hp, but is still 75% efficient from 1hp to 50hp. This means there is less to be gained by pulse & glide. So pulse & glide would give you an extra 10% range, not double the range."

Is this correct? Seems to me the extra power demanded during acceleration is not completely countered by the power saved during the coasting phase. Perhaps it works for ICE engines with a narrow powerband, but if so, why don't racers and cost-conscious drivers use the technique more?

There's also the Peukert effect to consider with EV's. This is discussed somewhat here, from where the above quote originated:
http://autos.groups.yahoo.com/group/ev-list-archive/messages/57787?threaded=1&m=e&var=1&tidx=1

 

[Leeps]

Its a tough cookie. On one hand you have the constant loss of no load current. On the other hand you have the increased resistive loss of making up for lost kinetic energy when your accelerating. I would have to spend some time with a calculator. I guess best case senario would be a motor with a current limit that places it near its efficiency peak, then you coast down to min speed, and reaccelerate. Your saving no load loss during the coast time and when reaccelerating your doing it as efficiently as possible.
My verdict, it could be, depends on you

Joe

 

[knoxie]

Hi

Its a good question, I think generally it has got to be more efficient to always keep the speed below 17mph and power and cruise as the power required to accelerate (whilst moving above 5mph) at lower speeds is much less due to wind resistance.

I can get amazing range with light peddling and light but constant throttle if I keep the speed to about 16mph, I can do 40 miles on 12Ah without even breaking a sweat, its kind on the batteries and controller also.

Good question though

P

 

[Mathurin]

In racing games, fuel economy is kinda secondary, you can tank up when your tires are worn out anyways, and you want to carry as little gas as possible because cars are best when as light as possible, they eat up their tires slower, accelerate faster, corner better... The only time you're not floored is while braking or exiting a curve. I see no place for pulse & glide there, to make up you'd need a more powerful (bigger) motor, bigger tires etc, surely this would dwarf the weight in fuel savings.


According to Darin, pulse and glide is too impractical to be useful.
http://www.metrompg.com/posts/pulse-and-glide.htm

 

[fechter]

If you're talking about a hubmotor with no freewheel, then the drag due to core loss and friction is always there regardless of whether there's any power going to the motor or not.

Add to this the Peukert effect and I2R losses that increase with the square of current, I'd say you'd be better off by maintaining a steady throttle.

If you had a freewheel that allowed the motor to stop while coasting, then you might be better off pulsing, since there would be no motor core loss while the motor is stopped.

 

[lemmiwinks]

As mentioned pulse and glide is far too impractical to be of much real world value. I'm a firm beliver in gentle acceleration and a light throttle application to maintain cruising speed to give maximum range. I rarely actually do this on my EVT, I normally peg the throttle and leave it there, but gentle acceleration and feather throttle is how I drive my ICE smogbox and get very good economy.

 

[Toorbough ULL-Zeveigh]

As far as ICE's are concerned, that verdict has been in for a long time. All the cab drivers in Bombay, India drive this way. In metro Manilla, it is so ingrained as the normal flow of traffic that no one ever comes to a complete stop at a stop sign & any foreigner who makes the mistake to actually stop gets rear ended in short order. They wouldn't be doing it if it didn't offer substantial fuel savings.

Car & Driver once had an article about VW developing an computer controlled system tied into the starter motor, shutting the engine on & off automatically & seamlessly in order to raise the mpg. I don't know if this ever came to market, perhaps in parts of the world like India & Phillipines that pay the highest prices for petrol.

On my ebike, all I can say I believe it does help somewhat. While it's hard to quantify, there are times I'm running low on the precious juice & I do this to squeeze out the last watt-hour. It's small but unmistakeable that I've gone farther than I would have if I held constant. I have often wondered why this should be so. Perhaps a function built into the controller might make it seamless for an EV as well.

 

[Reid Welch]

Perhaps a function built into the controller might make it seamless for an EV as well.

Yes for a conventional hubmotor, but nay for a freewheeling motor (shock loads).

Anyone disagree?

 

[lemmiwinks]

I'm still not believing the benefits of pulse and glide. On my ebike (which admittedly has a low top speed) it normally pulls about 120-140W on level ground at 22kph full throttle. If I rolled off the throttle, lost my momentum and then had to accelerate from say 10kph up to top speed again I could easilly pull 400+W while accelerating. Surely Peukerts effect alone dictates that a steady 5A is better than peaks of 16A (which is more than a 1C discharge on my 12Ah 24V pack).

I'm not convinced pulse and glide is a good idea or more efficient for ICE or EV and I doubt I ever will be.

 

[lemmiwinks]

As far as ICE's are concerned, that verdict has been in for a long time. All the cab drivers in Bombay, India drive this way. In metro Manilla, it is so ingrained as the normal flow of traffic that no one ever comes to a complete stop at a stop sign & any foreigner who makes the mistake to actually stop gets rear ended in short order. They wouldn't be doing it if it didn't offer substantial fuel savings.

You're describing stop-start traffic which is a whole different kettle of fish. No doubt being a little naughty and rolling through that stop sign is much more efficient than coming to a compete halt, but that's not what I'm talking about.

Lets say we're going from point A to point B, a distance of 30km with no crossroads, it's a straight shot. If you pulse and glide, and I maintain a steady light throttle not only will I get there before you, but I would have to have used less energy simply because of Newtons first law.

 

[andrey320]

Pulse and glide will probably result in a longer ride.

 

[lemmiwinks]

But how is that practical in a real world application? We both just need to get to point B, since theres no cross traffic there's no speeding up and slowing down of the traffic flow. Pulse and glide would get you rear ended (and rightly so) out in traffic while I'm already sipping a latte at point B.

 

[Mathurin]

Dunno about Bombay, but having lived in New Delhi for 2 years,
I can guarantee to you that they didn't use pulse and glide there.

[...]Lets say we're going from point A to point B, a distance of 30km with no crossroads, it's a straight shot. If you pulse and glide, and I maintain a steady light throttle not only will I get there before you, but I would have to have used less energy simply because of Newtons first law.

You didn't lookup the link I posted:


Pulse & glide overview

Pulse and glide works like this: let's say you're on a road where you want to go 60 km/h. Instead of driving along at a steady 60, you instead accelerate to 70 (that's the pulse), and then coast in neutral with the engine off down to 50 (that's the glide). That's it. Rinse and repeat. And repeat. And repeat...

By doing this, you're still averaging 60 km/h, but it turns out that pulse and glide is significantly more efficient than driving along maintaining a steady 60 km/h.

 

[lemmiwinks]

You didn't lookup the link I posted:

Gulty as charged

Pulse & glide overview

Pulse and glide works like this: let's say you're on a road where you want to go 60 km/h. Instead of driving along at a steady 60, you instead accelerate to 70 (that's the pulse), and then coast in neutral with the engine off down to 50 (that's the glide). That's it. Rinse and repeat. And repeat. And repeat...

By doing this, you're still averaging 60 km/h, but it turns out that pulse and glide is significantly more efficient than driving along maintaining a steady 60 km/h.

Ok, but we're talking a fairly short period of time here to shed 20kph on a big vehicle like a car. For an ICE there are some safety issues (maximum of one good brake application before vacuum assist is used up), plus driver distraction, plus wear and tear on motor stop/starting all the time (oil pressure nil, then start up and immediately apply load).

I *still* say it's completely impractical and I remain highly skeptical that constantly accelerating and decelerating will require less energy than maintining a constant speed. See Newtons first law -

"An object at rest will remain at rest unless acted upon by an external and unbalanced force. An object in motion will remain in motion unless acted upon by an external and unbalanced force."

We're talking about the second half, an object in motion. So our car, or ebike remains in motion unless acted upon by an external force, in this case primarily friction with the road surface, wind resistance and gravity. If we don't balance out these forces we will slow down. If we balance them we maintain our speed so energy in will depend on the amount of external force acting upon our vehicle.

Pulse and glide says that the amount of energy used to accelerate the object from an arbitrary speed (say 50kph up to 70kph) is less than the amount of energy lost as the object decelerates (say from 70kph to 50kph) and less than is required to maintain the average (60kph). All things being equal (i.e two vehicles of identical mass, rolling resistance, wind resistance and on an identical gradient) it just doesn't make sense to me.

Is there a physicist in the house?! :?

 

[Mathurin]

As a general rules, ICE motors suffer badly from throttling losses at light loads, and pumping losses at higher RPM's. To avoid throttling losses you have to open up the throttle, to avoid pumping losses you have to keep the RPM's low, used this way the motor is using it's fuel more efficiently.

So what you do in pulse and glide is run the motor as described above, but this accelerates the car so the motor then gets turned off instead of throttled, the car is allowed to coast down and a certain point the process is repeated... And so you end up wasting less of the fuel's energy, but this isn't to say the car needed more or less energy to make it go.

 

[lemmiwinks]

As a general rules, ICE motors suffer badly from throttling losses at light loads, and pumping losses at higher RPM's. To avoid throttling losses you have to open up the throttle, to avoid pumping losses you have to keep the RPM's low, used this way the motor is using it's fuel more efficiently.

I can see how this might be the case for carburettored motors as they require a certain amount of air speed for venturi effect, but does this also hold true for EFI motors? I would think that a typical EFI equipped vehicle with a closed loop lambda sensor system http://www.picotech.com/auto/lambda_sensor.html, TPS (throttle postion sensor), CDI etc, would be extremely efficient (as far as those hunk of crap 25% efficient infernal combustion engines go anyway) regardless of throttle opening.

So what you do in pulse and glide is run the motor as described above, but this accelerates the car so the motor then gets turned off instead of throttled, the car is allowed to coast down and a certain point the process is repeated... And so you end up wasting less of the fuel's energy, but this isn't to say the car needed more or less energy to make it go.

So if you're wasting less of the fuel that goes in, but using more of it overall (I think Newtons law indicates that it will require more energy to accelerate than hold a steady speed) isn't that less efficient overall? :?

 

[Leeps]

Throttling losses and pumping losses are different names for the same thing in a gasoline engine. This comes from an imbalance in forces on the intake stroke of the engine. Imagine at half throttle you have 7 psi of absolute pressure in the intake manifold, full open throttle ideally would have 14.7 psi or atmospheric pressure. On the bottom of the piston you have atmospheric pressure on the top you have 7 psi this gives you 7.7 psi of pressure on the piston fighting the rotation of the engine for no good reason. The problem is the same for carburated or fuel injected engines. I would reason that the negative pressure should aid the piston on the return stroke but this is the theory that i have read in text books and what was taught to me in my mechanics classes thus it is what i am repeating here.
The losses from compression are greater at full throttle there is a greater peak pressure in the combustion chamber that the engine must fight against. However for reasons unknown to me higher compression engines are more fuel efficient so the greater apparent compression ration and effort required at top dead center actually makes for better efficiency in the long run.
There is however a tendancy for engines to run excessively rich at wot and throw efficiency out the window. The standard in fuel injected engines at wot is to ignore the oxygen sensor and to throw in a given amount of fuel based on the mass air flow meter or the manifold air pressure sensor multiplied by the rpm and a given variable. This ensures a rich mixture, lowers combustion temperature and makes for greater engine life by saving it from high combustion temperatures and a possible lean condition. This leads to greater pollution but they can get away with it because pollution tests are never done at wide open throttle.
Joe

 

[xyster]

Is there a physicist in the house?!

Really, a good physicist could explain this quickly and easily. From what I recall of my one year of college physics with calculus, which isn't much, I agree with lemmiwinks that in general pulse-glide at average speed of X could not be more efficient than holding the speed at X. The reason is the same as the reason it's less efficient to go up a hill of X height for Y distance, then back down that same hill for the same Y distance, compared to just going on the level for the same 2Y total distance. I just can't seem to explain that reason, and neither does it seem can the others who agree, like lemmiwinks and fechter...

As we've been discussing, it seems possible there could be efficiency gains particular to certain kinds of motors that could be greater than the efficiency losses.

I'll try to devise a good test to run using my ebike or scooter, reporting the results back here later. Maybe if others could do the same, we'd be able to settle this. This seems a very important debate to settle, especially for us EV'ers. If pulse-glide is indeed more efficient, there's plenty of times when I'm out on my low-mileage scooter that I'd use it.

 

[Mathurin]

Throttling and pumping losses the same thing, ok.
Then how do you call high rpm loss? Windage?

I've read throttling losses are the greatest source of loss in an otto cycle motor

 

[Leeps]

Throttling loss and pumping loss are two words for the same thing. Its the piston fighting against the negative pressure in the intake manifold.
The high rpm loss depends on what you are talking about exactly. The loss in torque that occurs at high rpm is due mostly to the inertia of the air not letting it accelerate and get into the cylinder at high rpm. This is why they need camshafts and intake manifolds tuned to high rpm, the intake manifolds are there to set up a resonance and use the inertia of the air as an advantage, the camshafts hold the valves open longer to give the air a longer time period to get in. If you are actually talking about a mechanical energy loss this would be compression loss. The pressure at top dead center is directly related to the number of molecules of air that are able to enter into the cylinder, the greater the charge the greater the pressure and thus the greater the force and energy needed to make it past top dead center on the compression stroke.
Joe
edit: you are right throttling losses are the primary loss in an otto cycle motor

 

[fechter]

On my Honda Civic Hybrid, when you want to coast, the valves are kept closed by electric solenoids and the fuel injectors are turned off. This minimizes engine drag so the electric motor can recover max regen.

Pumping losses account for a very significant amount of an ICE's inefficiency.

 

[Toorbough ULL-Zeveigh]

Dunno about Bombay, but having lived in New Delhi for 2 years,
I can guarantee to you that they didn't use pulse and glide there.

Could be a regional difference. That was a couple of decades back, so perhaps it's been outlawed since. But in Manilla, that is the situation as of a month ago.



I've said this b4, but no one else seems to share my view that I consider the battery to be the engine. I think the answer lies within how the engine operates, be it ICE or battery.

How many are familiar with this experience.
Your battery goes dead. Be it ebike or torch, don't matter.
You shut it off & leave it off for a few minutes.
You come back, turn it on & it works briefly once again.
Where did that energy come from

It's an inherent characteristic of batteries. I think this is part of the Peukert effect. Have also heard it referred to as plate re-polariztion. I do know that if I run a constant heavy drain out of a battery, when you cut the load for a while, the voltage sag recovers a bit when you start up again.

So it may simply be due to heat buildup. Whatever it is, whatever you call it, I'm sure you've all witnessed this b4. Pulsing may give the battery a chance to cool off and/or repolarize to recover some capacity & operate more efficiently. Probably not the whole picture, but it gots to play a part.

 

[Matt Gruber]

Perhaps a function built into the controller might make it seamless for an EV as well.

Yes for a conventional hubmotor, but nay for a freewheeling motor (shock loads).

Anyone disagree?

Shock loads are a symptom of poor throttle design. awful.
but quite common.

 

[xyster]

Shock loads are a symptom of poor throttle design. awful.

Can someone briefly explain 'shock loads'?

So it may simply be due to heat buildup. Whatever it is, whatever you call it, I'm sure you've all witnessed this b4. Pulsing may give the battery a chance to cool off and/or repolarize to recover some capacity & operate more efficiently. Probably not the whole picture, but it gots to play a part.

I understand your observation, have experienced it too, but think it's being offered as evidence for the wrong side of this debate.
The battery, whether considered the engine or not, rebounds after being run at a higher rate than is most efficient for it -- same thing happens with my lithiums. I'll get a rebound of a couple hundredths of a volt after a hard ride, but not after an easy ride. The SLA's on my scooter rebound much more after a few minutes at full power than at low power.
It should be axiomatic that pulse/glide requires a higher peak discharge rate than the steady-speed technique. Hence, employing pulse-glide, the battery is much more likely to run at least briefly outside its preferred discharge rate zone, and so more apt to show noticeable rebound.
The rebound phenomenon provides evidence for pulse/glide's inferior efficiency compared to the steady-speed techique, not the other way.
I like your name, btw, it sounds like 'turbo rides-all-day'

 

[Matt Gruber]

QUIZ:

what is a SHOCK LOAD?

a. when you stick your finger in an electric socket, and have to change your underwear

b. when u fall and your head hits the pavement

c. anytime your head hurts from thinking too much

d. when your g/f says NO after a marriage proposal

e. anytime a drivetrain has slack in it, and gets hammered upon engagement