Pulse&Glide vs. Constant speed?

I do not Pulse and Glide deliberately, although I see my mileage improve in rolling hills where the land itself does pulse and glide. Toyota seems to have made it as hard as is possible to pulse and glide.

To Pulse, you want the throttle as open as possible, (to reduce pumping losses) but not so open that fuel enrichment is activated or that battery power is used. I see no instrumentation to make avoiding fuel enrichment easy.

To Glide you want no regen and ideally no electric assist, although Toyota give no tools for finding this state. (if you find it on your own, the MFD can show 'no arrows')

Avoiding pumping losses (while avoiding converting to or from battery chemistry) does make sense to me, but I have never learned 'the touch' nor do I ever wish to make my fellow drivers any more irritated than they are. (dieing in a 'Duel' road rage incident is a false economy)

Duel (1971) (TV) - Synopsis

 

ICE spin requires about 2 kw as a fixed cost; if say an example P&G has the ICE off 50% of the time, on average you save 1 kw.

Second, at the low speeds P&G would be considered a constant speed would require quite a bit less than 10 kw to move the car and I will guess will be an inefficient ICE output compared to moderate power demand during the pulse phase of P&G. I cannot quantify this for you, and ymmv.

 

I think it is a mistake to consider pumping losses with regards to pulse & glide. Particularly in an Atkinson type engine pumping losses will be very small. But more appropriate to consider overall engine efficiency at different power levels. The beautiful graphs of engine efficiency provided here shows, that in the 20 to 30 KW range the Prius engine operates in the 220 g/KWh region (about 35% efficiency) but at reduced power the efficiency drops rapidly. At the power required to cruise at 35 or 40 mph the efficiency may be down to 20% It is more economical to use the engine to accelerate using 30 KW and coast (glide) back to some slow speed.

I recall in the Honda Insight pulse and glide example the top speed was only 32 mph, and they would coast down to under 5.
The Prius gets pretty good efficiency at highway speeds and as the OP notes at higher speeds wind drag dominates. So P&G works best at low speed.
-mort

 

Hi All,

Pulse and Glide is better below 45 mpg in a Gen III Prius.

But, the technique really takes training from someone who knows how to do it, or instrumentation and reading what to look for. The pulse needs to be about 2000 RPM on the Gen II, a N glide works best. As it saves battery drainage. The issue is, how heavy to make the pulse. Once you get used to it, its a pretty broad band, (1800 to 2200 rpm in the Gen II), but until you know what that sounds like in the car, its easy to be all over the place.

In my Gen II, I would do about 75 mpg P an G from 34 to 40 mph, flat terrain, maybe a little up hill even. Holding steady on that section of road was about 50 mpg. For steady speed mileage similar to this, I had to run the Gen II Prius between 49 and 53 mph in SHM on a nice humid warm summer morning, after some driving to get the tires warmed up, and the battery to 66% SOC.

On the Gen III I am going up to the high side of the central portion of the HSI, maybe a little into power, just keep from using the battery for accelleration. That seems better. But, I am still learning my 2010. Second tank is at 65.4 mpg (indicated) and 540 miles.

One avoids pumping losses by doing the P and G once the car has warmed up, and will kill the engine with proper pedal manipulation (full off, then up a little), and being below the speed the Prius will kill the engine (which is < 40 mph in a Gen II, and < 45 mph in a Gen III).

 

I'm more interested in reproducible results so my testing has been based around using cruise control 'resume' for acceleration, the pulse, and shifting to "N" for the glide after engine stops. For the 1.8L ZVW30, this imposes limitations:

• set cruise to max EV, 46 mph, +2 mph -> 48 mph
• low-limit 23 mph + 2 mph -> 25 mph

I did some testing with our 1.5L NHW11 and satisfied my curiosity. There are practical problems with pulse and glide and many have found solutions.

With our 2003 NHW11, I use pulse and glide on 25 mph, neighborhood streets with little or no traffic during warm-up of the NHW11. But the coolant heating system of the ZVW30 has makes this unnecessary.

GOOD LUCK!
Bob Wilson

 

Does the Prius have much pumping loss?

In theory the combination of Atkinson cycle engine and eCVT could have been nearly eliminated it. The few times long ago that I watched manifold pressure on the ScanGauge, it wasn't gone, especially at very low engine power. But its pattern wasn't immediately obvious. It it time to back to monitoring it again to get a better understanding of pattern and magnitude.

 

Hi Fuzzy1,

If the engine is not turning over, there are no pumping losses.

During the pulse, you need a medium high power level to get the Prius to open the throttle all the way, and ramp the load up to a high torque portion of the engine map. This minimizes the throttle plate losses. I think Jimbo misspoke about Pumping Losses. The Atkinson engine has minimal losses getting air into the cylinders from the plenum. But, it still has losses pulling air through an almost closed throttle valve.

 

As long as the engine of car is larger than needed for steady state driving (they all are), pulse and glide can improve mileage. Pulse and glide allows the engine to be used intermittently at its most efficient output. Constant speed driving requires the engine to be used continuously at a less efficient output. It's pretty simple physics.

The real question is how much can be gained. This depends on all of the factors, including the design of the car and ambient conditions.

Tom

 

If I am reading the graph correctly, GenIII Prius is still running close to peak efficiency at 10KW. Efficiency doesn't drop rapidly until power falls below 8-9KW, i.e. on the vertical operating line on the left side of the graph.

Pumping loss appears far lower on my Prius than on my traditional Otto-cycle Subaru. Watching ScanGauge's MAP (Manifold Air Pressure) display, Prius's intake vacuum was typically 0.1-0.2 bar in steady conditions above this power. Vacuum increased substantially under changing conditions or lower power.

In contrast, the Subaru (twice the intake displacement, 2.5L) at normal highway speed (nearly twice the RPM) was typically running a vacuum of 0.4-0.7 bar. That is pumping loss.

 

Pumping loss is the frictional loss of energy from pulling air through an engine. If you think about it, an internal combustion engine is just a big air pump, which burns fuel to provide energy. Ideally all of the energy from the burning fuel would be used to do something useful, but in practice much of it is wasted. One of the wastes is the friction of pulling air through the engine.

Of all the pumping losses, the biggest is normally pulling air past a mostly closed throttle plate. This is one of the main reasons gas engines are inefficient at low power levels (diesel engines do not use throttle plates, so therefor do not suffer from this).

Pumping loss is not measured in Bars. Bars is a unit of pressure, so it can be used to show the pressure drop across the throttle plate. The higher the pressure drop, the larger the loss.

Tom

 

Hi All,
For back of the envelope figurin' the power lost to pumping the air into an engine is W=dP * Q, for me, using conventional dimensions, that's horsepower = pressure drop in psi multiplied by flow in cubic feet per second. Well, there's a bunch of conversions in there. Assuming the engine is in the 30% efficiency range and putting 10 horsepower into going 60 mph, that works out to about 4.4 lb of gasoline per hour. The engine will use 860 cubic feet of air to burn that gasoline. Or 14.3 cu ft per minute, which is about 25000 cu in per minute. I don't know what the pressure drop is for the intake of a Prius at cruise, but I'm sure it is less than .5 bar, about 7 psi. 7 psi * 25000 cu in per minute is 175000 in - lb per minute or 243 ft-lb per sec. or .44 hp That's 4.4% of the power output, or about 1.3% of the gasoline...
As I said before, pumping losses are not the reason efficiency drops a low power levels.

-mort

 

But they are a significant contributer at moderate levels. It's easy to see why if you look at the extremes:

1) If a throttle plate is fully closed there is zero airflow. With no airflow there is no pumping loss (obviously in the real world the engine would not be running with zero airflow - this is a thought experiment.)

2) If the throttle plate is fully open there is no pressure drop across the plate, therefor the pumping loss due to the throttle is zero.

Given these endpoints, it's easy to see that pumping loss across the throttle increases as the throttle opens, since air starts to move through the restriction. The pumping loss continues to increase with more throttle, but as the throttle opens the resistance to airflow decreases. At some point the decrease in resistance overcomes the increase in airflow and the pumping loss begins to decrease. Maximum throttle induced pumping loss occurs at one magic point somewhere in the middle, generally on the lower side of the middle for most engines.

Tom

 

Most drivers will use over a gallon of petrol to drive 60 miles at 60 mph, so the starting petrol weight is closer to 7 pounds.

Second, air is about 20% oxygen, and the mw of oxygen is 16, of carbon 12, and hydrogen 1. Petrol is approximately CH2, and combusts to CO2 and H2O. So the weight of air passed into the combustion chamber is about 5*6*32/12 + 5*1*2/12 equal to 81 pounds. Air weighs 0.075lbs/cu.ft, so we pass 1080 cubic feet of air to burn the gallon+ of petrol.

Works out to about 25% more than your estimate, or 400 watts to pass the throttle plate in more sane units

This is at 60 mph, not the 30 - 40 mph where P&G takes place.

 

Hi All,
My point is that in the most unfavorable condition the pumping losses are less that 5% of power output. If you can find a situation where pumping losses are higher please show your math.

-mort

ps while this may be true:

I used the simpler approximation of ideal A/F ratio.

 

Incidentally, not to begrudge hooves, hp and things, but metric is *so* much easier ...

One bar is 10^5 Pa, and one Pa is very conveniently one N/m^2
Since W = Force * distance, air weighs 1.2 kg/cubic meter, and we push 37 kg of air,

Work is (31 meters^3 * 50,000 Pa) joules.

 

As I indicated earlier, my Prius' manifold vacuum ran 0.1 to 0.2 bar while cruising at highway speed yesterday. Call it an average of 0.15 bar.

But at lower power levels, vacuum increased sharply.


Here is my crude estimate of pumping loss at 60 mph on the highway (note the use of Prius's maximum intake displacement, not the 1.8L exhaust displacement) --

Prius: 0.5 * (1.2 L) (1m^3/1000L) * (0.15 bar) (1e5 Pa/bar) * (1500 RPM) * (1 min / 60 sec) = 225 watts (or 0.3 HP)

Subaru: 0.5 * (2.5 L) (1m^3/1000L) * (0.60 bar) (1e5 Pa/bar) * (2700 RPM) * (1 min / 60 sec) = 3375 watts (or 4.5 HP)

 

This one is easy. Losses are at a maximum at idle, where the engine is producing zero output power. At this point pumping loss is infinite as a percentage, as are all other losses.

The trivial point of zero output power illustrates why pumping losses are considered significant at low power levels. Total pumping losses increase with speed as more air is drawn through the engine. However, as engine speed increases, so does output power (up to a point). Even though actual pumping losses increase, pumping losses as a percentage decrease (once again, up to the point where engine efficiency starts to fall off). Percentages are tricky things.

Tom

 

This sentence has been bugging me. If power required in passing the throttle is ~ 400 watts, and ICE efficiency is say 1/3, then petrol consumption is times 3, not 1/3rd -- or 1200 watts to overcome the resistance.