Quantitative Pulse and Glide

With respect, what effect do you think this sort of repeated, wide speed variation (+68% to -41%) has on surrounding drivers and traffic?

I think it is nuts with other traffic around. It leads to unnecessary tailgating at both ends of the speed range . . . depending upon the posted maximum speed.

Both the SAE paper and my own testing confirm these should only be done isolated from other traffic. So in my case, it was a Saturday morning on a straight, flat road at Redstone Arsenal that connects the river side recreational areas and the main base. If other traffic showed up, I aborted the test by pulling on to the shoulder and parking. After traffic cleared, I proceeded to the next start point and set it up again. But I believe low-traffic roads and parking lots exists that can be used for such tests and this is why I'm looking for others who might measure their Pulse and Glide performance versus constant speed.

Likely places to f"I got MPG using pulse and glide at elapsed speed."

Without the control:

"I repeated the same route at a constant speed and got MPG."

This second statement is too often missing so no one can really tell what pulse and glide accomplished.

Bob Wilson

Agreed. Experimental design is important. As is limiting experiments to the lab.

But I bet he gets REAL cranky when the outside world doesn't cooperate and let him play his rolling video game according to HIS rules.

It's amazing how many people think they don't have to live in the real world.

"The down side is that it will annoy anyone stuck behind you when you are doing it. I know that and make it a point not to play that game when someone is stuck behind me."

Did you not understand the above statement?

You backpedal well.

I'm less interested in what people do in traffic as there are plenty of 'bad examples' out there. For example, this morning there was an upscale sedan doing between 40-45 mph on a cross-town, 50 mph, divided road followed by a service van towing a trailer that was rocking back and forth. Traffic was backed up behind them and I took advantage to follow the van. So I got the mileage advantage by exploiting the poor judgement of two other vehicles . . . sweet. I parked at work with just over 50 MPG indicated and the temperature was just under 40 F.

What I still don't see posted is anyone else's pulse-and-glide performance over the same route, same average speed versus the constant speed equivalent. This test can be conducted late at night, say after midnight, if nothing else to answer the question:

"What mileage improvement does PnG provide over the equivalent, constant speed?"

As for the safety aspects, there are enough poor drivers we don't really need to 'flog our own.' But if someone thinks PnG is so great, answer the question. Give some quantitative numbers:

a) minimum speed - glide end b) maximum speed - pulse c) equivalent constant speed d) MPG in PnG e) MPG at constant speed

Several years ago, some folks spent a week in Oklahoma driving section lines in a Honda Insight endurance test. Using their mileage log, they'd been driving at a constant speed equivalent of 16-18 mph. This also corresponds to the maximum speed range of a Prius (and probably quite a few other hybrids.)

If PnG boosts the equivalent constant speed mileage say by a factor of two:

20 MPG @40 mph - constant speed 40 MPG @40 mph, 5 to 75 mph - pulse and glide

(40-20)/20 = 100% improvement

This would be a game changer but I never measured more than 12% improvement. In contrast, the SAE paper claimed some curious numbers:

89.1 MPG @25 mph - constant speed (0.0% SOC decrease) 128.9 MPG @25 mph, 20 to 30 mph - pulse and glide (0.5% SOC decrease) (128.9 - 89.1)/89.1 = 44.7% improvement

81.0 MPG @35 mph - constant speed (0.0% SOC decrease)

151.5 MPG @35 mph, 30 to 40 mph - pulse and glide (1.5% SOC decrease) (151.5 - 81.0)/81.0 = 87% improvement

80.9 MPG @35 mph - constant speed (0.0% SOC decrease)

99.9 MPG @35 mph, 30 to 40 mph - pulse and glide (0.5% SOC decrease) (99.9 - 80.9)/80.9 = 23.5% improvement

What I suspect is many lay reports about how great Pulse and Glide performs may be just the discharge of the traction battery providing the additional energy that is not counted in the total energy picture.

So the challenge remains, the question remains:

What is the payback of Pulse and Glide versus a constant speed?

BTW, a route will hills changes the physics. Climbing can move the engine into power regions that are fuel inefficient. In reality, any test should be against a constant ICE power setting and let speed follow load. But more of this advanced topic later ... if we can stop bashing poor driving including going the minimum posted speed limit on an Interstate.

Bob Wilson

Thank you for demonstrating that you don't even know what that term means. In my first post to this thread where I mention pulse and glide I wrote "The roads around here are mostly 2 lane, and coasting down to 20 mph in a 35 zone with traffic behind you isn't really an option."

So I made it clear from the begining that I didn't condone getting in the way of others. You are a troll to keep implying that I do.

On Feb 3, 6:18 am, Bruce Richmond wrote: . . .

All I'm interested in is a comparison of pulse and glide versus the equivalent constant speed speed. BTW, there is one place I use 'pulse and glide' without being a traffic hazard . . . a 'rolling warm-up.'

On cold mornings, I cut through my 25 mph posted, neighborhood streets taking the 'back roads' to let the engine warm-up with a minimum fuel burn:

The characteristics are:

17-25 mph, PnG, 'glide' in "N" ~30-35 MPG, ~1.5 miles 21 mph, constant speed ~20-22 MPG

Because I'm running parallel to the main commuting route, there is very little traffic 'going my way.' With four cross streets leading to the main commuting route, the vast majority of traffic turns at the next intersection to commute to work. In those rare cases were someone is following, it is because they are going to park at their home in that block. But about once every four months or so, I pull over to park at someone's house to let a following vehicle pass . . . following traffic is very, very rare.

Coasting in "N" minimizes the fuel burned to warm-up the engine to a rate of ~0.30 gallons per hour. There is no throttle setting that has an equivalent, consistent, low fuel burn rate measured with my ScanGauge. Regardless, the goal is to arrive at the commuting route intersection when the car has warmed enough to enter "stage 4" or hybrid mode. It is more accurate to call this a 'rolling warm-up' because I don't use it once the engine coolant reaches 70 C.

I have a somewhat similar protocol leaving work where I can exploit catalytic converter warm-up. There is a 30-40 second window when the NHW11 Prius uses traction battery power and idles the engine at ~0.60-0.70 gallons per hour. By parking near the parking lot exit, this is long enough to reach 35-40 mph before the catalytic converters light off and normal engine warm-up begins. But at 35-40 mph, I'm keeping up with traffic and by follow downgrades, can maximize the use of the low-fuel burn rate, "N" during warm-up. Again, as soon as the car reaches "stage 4" or normal hybrid mode, I'm back to normal, constant speed mode.

BTW, in warmer mornings, I often take one of the earlier cross streets once the engine coolant goes over 60 C. Again, this is a rolling 'warm- up' using "N" to minimize fuel burn when rolling. In contrast, most 'pulse and glide' advocates claim the engine must be off or throttle feathered during the glide. My scangauge indicates there is no feathering with a lower rate of fuel burn than "N".

Bob Wilson

Never gave much thought to p&g during the warm up cycle. Doesn't the battery have to get up to some minimum charge before the engine will shut off? If you are shifting into "N" it may use less fuel at the moment, but even out in the long run because it takes longer to get the battery up to charge so the engine can shut off. I don't know this, just speculating.

After warm up the idea with p&g is to get the engine to shut off during the glide. Back in the day, drivers in economy runs would accelerate at full throttle in high gear, then step on the clutch while shutting the engine off with the key, then slip into neutral. They would keep the throttle wide open with the engine off to keep the accelerator pump from squirting when they stepped on the gas again. When they had coasted down to their min they would step on the clutch, slip it back into high gear, turn on the ignition, and let the clutch out to bump start the engine, still WFO. The whole idea here was to eliminate the losses due to pulling a vacuume against the throttle plates. The same applies somewhat to the Prius. During the Pulse the throttle is open wider than it is during steady state cruising. During the glide you make the engine shut off so it uses no fuel at all. Feathering the pedal is to prevent regenerative braking from kicking in. There are losses every time the battery is charged or discharged, so we do our best to prevent any charging/discharging. It would be nice if we could just flick a switch and let the computer do the feathering for us.

I haven't done much in the way of formal record keeping for p&g vs steady speed. I know that by using it I have averaged 70-75 mpg for a

40 mile trip on several occasions. That was done on two lane secondary roads where the speed limit ranged from 35 to 50 mph, so much of the time I was well below the speed limit, averaging about 35 mph. A steady 35 mph is usually good for 55-60 mpg, so the p&g gives about a 25% improvement. Over a distance of 40 miles I don't think a few % on the SOC makes much difference. When I get the chance I will try to come up with some specific numbers under more controlled conditions. Right now isn't the best time though since the temp has been running in mostly in the single digets and teens.

Bruce

Hi Bruce,

On Feb 3, 10:50 pm, Bruce Richmond wrote: . . .

Not in practice. By doing a 'rolling warm-up,' we're getting the engine into temperature regions where it can efficiently produce power. A cold engine has higher internal friction from the oil and parts. So I don't really have to worry about the traction battery since within a fairly short distance at constant speed the traction battery fully recovers. Engine efficiency is the key.

I know but there are inefficiencies in today's car including our Prius that makes this nuts. For example, the NHW11 and NHW20 both have to use fuel enrichment to prevent excessive, exhaust gas temperatures from burning out the catalytic converters. The ZVW30 uses cooled, exhaust gas recirculation but there are other inefficiencies that happen at high power settings. Another ZVW30 trick is a valve in the muffler that open under maximum power to reduce back pressure. This was not an option on the NHW11.

You're starting to see the problem with lack of documentation because you can only 'estimate' a savings. Without hard numbers, it is easy to 'estimate' the way you want the numbers to be and that isn't mean, it is just human nature.

It doesn't take much to log the results and on longer test segments, you'll need a collection of data . . . say five . . . and throw out the best and worst and average the middle three. But try to keep a record of winds, directions and temperatures so unusual conditions can be documented. BTW, be sure and record the start and stop times. It is very easy to miss-estimate the actual block-to-block speed versus the equivalent constant speed.

Do you have any type of recording, OBD device? I like to use a laptop to record critical vehicle parameters and then analyze the data later. This way I can concentrate on safe driving with my eyes out the windows instead of trying to be a human data recorder of a ScanGauge.

Thanks, Bob Wilson

impressive

my two cents...

Environmental and Experimental Factors that would effect a fuel efficiency P&G experiment:

1) Type of gas formulation: Winter Fuel Formulation whose distribution in DC/MD/VA starts about Mid October lower fuel efficiency by about 7 to 15% summer/spring time gasformulation has more energy in it. In addtion, if you can get 100% pure gas instead of E10 - the Prius fuel efficiency - Mileage would also increase (as would any car). E15 would lower the MPG results.

2) Cold/Winter operational Temperature/inital warmup stage: During the first few minutes after initially turning on the Prius the fuel efficiency drops ( about 24 mpg for the first 5 minutes ) because the ICE needs to warm up the coolant/catalytic converters - outside temperatures below

45 F which occure more often in the winter can also cause the ICE to run more oftent to warm up if it is standing still ( stop and go urban traffic). Once the ICE coolant warms up to over 110 degrees F - fuel efficiency improves dramatically on the Prius ( you can monitor the temperature with the ScangaugeII FWT Xgauge). This is characteristic of all gas- electric hybrid. My 2010 Prius III MPG usually doesn't achieve over 50 mpg (AVG) until the ICE is over 140 F degrees(FWT). Grill blocking can be used to decrease the number of times an ICE is turned on to warmup the coolant/catalytic converters during the winter time temperatures in urban driving conditions. Grill Blocking is less useful for highway driving conditions since the ICE is running harder and more often in this driving condition. MPG results can be biased by the car's intial startup/warmup temperature and a low battery SOC (which could trigger the ICE to run)

3) Tire Pressure - overinflating the tire pressures by about 9 psi over the door panel recommendations decreases the tire rolling resistance and can increase the fuel efficiency by about 3 to 5 mpg - this is true for all tires/cars to some degree. Increasing the tire pressure over the door panel recommendations can lead to a harsher ride with more road irregularities being transmitted via the tires this is less of a problem with good roads and very punishing with poorly kept roads. Increasing the tire pressure over the maximum tire pressure designated on the side wall can result in a lost of road traction(increased stopping distance) especially in wet/icy road conditions.

4) Route selection - selecting a flat driving circuit/road with no stop signs would optimize P&G fuel efficiency. Traffic lights along the route would require a hypermiler also employ DWB techniques. Route selection can improve fuel efficiency on any vehicle but would improve fuel efficiency more for a hypermiler using a hybrid.

5) Timing/Traffiic - While higher MPG can be achieved via a longer P&G where the velocity is allowed to drop greater than 10 mph between cycles - in heavier traffic a shorter P&G cycle with a 6 mph range would be more practical. .

6) Gallon per Hour Monitoring vs Velocity - From what I noticed, the fuel efficiency of a P&G cycle is dictated more by the gallon per hour (GPH) peak usage than by speed. Fuel Efficency (MPG instantaneous) is a trade-off between instantaneous fuel used (GPH) and the current top velocity (MPH). The optimum fuel efficent velocity being the velocity whose kinetic energy translates to the potential energy which equates to the max distance the vehicle would coast on the road without braking regeneration if it needed to come to a full stop. The optimum MPG inst thus is maximized when the mimumum GPH is used to achieve that max velocity within this cycle.

HTH

Walter Lee

2010 Toyota Prius III, Blue Ribbon/Dk Grey/ OEM floormats Washington DC/MD/VA region

Odeometer 7800 miles, 58.0 mpg overall best tank 66 mpg, worst tank 52 mpg, last tank 52 mpg (tank = about 9 u.s. gallons of E10, about 500 miles/tank) Best route MPG (81 mpg/15 miles as measured by a ScangaugeII)

Yokohoma Avid S33 ( 44psi/42psi) grill blocking top 100% bottom 100% ScangaugeII (AVG, FWT, GPH, RPM)

Most of my daily route is hills with stop lights at the top or at the bottom of the hills. If I know can make the light by speeding up it sometimes is more fuel efficient to speed up just so I can make the light. If the light is not visible until I get over a hill - it is often more fuel efficent to limit the speed to less than 30 mph for I get to the top of the hill. The main challenge is to avoid using fuel to create kinetic energy that would be wasted if one has to stop. There are only a few sections of flat road where I can do a classic P&G but if there isn't any traffic - I can increase my mileage in those sections by either doing a P&G or doing a SHM if the SOC has atleast 3 to 5 bars. When the SOC is only 2 bars, then P&G and DWB are the only options I have. Warp Drive and SHM take almost no skill. DWB and P&G require much more skill, timing, and luck.

The data of your experiment would be even more useful if the P&G cycle were from 50 mph to 65 mph range because then it would useful for the highway. Late spring I do a 600 mile plus trek from DC to Michigan via the super highway. The last time I did it I experimented with using regular non toll roads. The 2010 Prius got 5 mpg better mileage on the high speed toll roads (58 mpg, OH/PA Turnpike) than on the slower speed regular non-toll roads ( 53 mpg - via Garmin Nuvi designated route). The weather was mostly sunny and warm, my tire pressure was set at the normal 35psi /33 psi - but it surprised me that the lower speed non-toll roads were less fuel efficient. This year I will use only the high speed toll roads (OH/PA turnpike) but I plan to test out and monitor the use higher tire pressures

44psi/42 psi at these speeds. I will also have a Scangauge II to monitor my progress. I also hoping to hone my P&G skills then.

On Feb 4, 9:50 am, Neo wrote: . . .

What were the equivalent speeds?

This is the classic problem of PnG advocates who consistently report "MPG" without the "mph." Mileage claims without:

o "mph" for each test, one PnG and the other the equivalent constant speed o a record of altitude, temperature and wind effects

If you live within a reasonable distance of an interstate, it would only take four runs, two with PnG and two with the equivalent steady speed, to quantify the effect. Many interstates have exits at least 10 miles apart, the perfect distance for doing a compare and contrast. Then you'll have some data others can use.

BTW, about a year ago, I tried comparing PnG at highway speeds, 45-70 mph, to the equivalent constant speed, only to find a significant loss of MPG. The test was conducted between 2:00-5:00 AM between Decatur and Courtland on a divided highway. Given the increased driver load and poor, high speed results, I can't find any reason to use it.

Good Prius friend Hobbit said it best that the goal is to operate the engine in efficient modes and that can be easily monitored using engine rpm. The surest path to higher MPG is to drive slower down to the 15-18 mph range. Below this speed, the vehicle overhead becomes the limiting factor.

Bob Wilson

I realize the battery will recover in a short distance at constant speed. The point was that more fuel will be needed during that recovery than would have been used to maintain the constant speed without having to charge the battery. Also since there will be no load on the engine when it is in "N" won't it take longer for it to warm up into those more efficient regions? It would be interesting to see the total fuel used in the first few miles using the different techniques. BTW I have been trying "N" when cold since reading your post and it was certainly an eye opener seeing 99+ mpg 100 yards from my driveway at 9°F.

Agreed, I wasn't saying to use WFO for pulse acceleration. On those old cars the carbs were also re-jetted for a lean mixture when wide open. Normal jetting is slightly rich at wide open because it makes more power, and that's what most people want when they hold it wide open. Just curious, have you done any experimentation to find out what rate of acceleration during pulses might work best. From what I have read it doesn't seem to make a whole lot of difference one way or another.

Up till now most of the things I have looked at made big enough differences so that they stood out with informal testing. Changing tire pressure by 10 psi makes a big difference, and the fact that it stays the same from then on tells me that it is a result of the change, not just some fluke of the day. But I am interested now in looking at things in finer detail and logging my measurements.

I have a laptop. What hardware/software do I need to go with it?

Bruce

Not that I can tell from my data records:

All heat comes from combustion and exhaust gas but the area exposed to these high temperatures remains constant and exposure in the engine block very limited . . . measured in fractions of a millisecond. Then the hot exhaust gas passed into the tail pipe and can no longer heat the block. Only the ZVW30 has an exhaust gas, coolant heater loop.

I'm pretty much driven by my measured, brake-specific fuel consumption graphs but generally speaking:

o 2,400 rpm or less - very good (I believe Hobbit prefers 2,200 rpm or less . . . ) o 2,400 - 3,200 - Ok, don't make a habit of it o 3,200 - 4,100 - make it brief o 4,100 - higher - better be to avoid a traffic issue or 'blowing off steam'

FYI, I was updating my software this weekend trying to get an ELM 323 ($50) to work . . . so far, no joy but I was testing MacOS. It may work with Windows. I'll test it later this evening.

Bob Wilson

On Feb 5, 1:10 pm, bwilson4web wrote:

No.

I could drive a constant 55 mph on the toll roads for several hours. but the non-toll roads had stop signs and traffic lights every 15 minutes to 30 minutes so my average speed dropped to something closer to 40 mph on the non toll roads even with light traffic. The non toll roads were less efficient because of the more frequent acceleration and deceleration events ( where I would lose kinetic energy every time I had to stop and and where I had to spend extra fuel to accelerate to a proper speed) Also some of the non toll roads were no well maintain while the toll roads were for the most part well maintained. So the rolling resistance/drag was significantly more on the non-toll roads. My Garmin Nuvi generated my alternate non-toll road route from Maryland to Michigan. So while I saved a small chunk of change using the non toll roads - I spent slightly more at the pump ( a decrease in about 5 mpg) and had to deal with a much rougher ride

P&G works best between 25mph to 35 mph (when wind resistance is not a big factor) and the terrain is flat or nearly flat. I use it on several segments of my commute when I have about a 1/2 to 1 mile segment of flat surfaces or gentle downhill. While P&G is only useful in a limited driving environment - it has the advantage of being independent of the traction battery's SOC

In "ECO" mode, the Prius does not have much torque and it's heavy curb weight becomes a liabilty when going uphill. In Maryland, my testbed road is a stretch of Norbeck Rd. Westbound Norbeck Rd (Rt 28) from Georgia Ave (Rt 97) to Rockville Pike (Rt 355). Westbound Norbeck from Muncaster Mill to Baltimore Rd is basically a gentle downhill route where the Prius FE increases from 35-45 mpg to 50 mpg to

63 mpg. However, going westbound on Norbeck Rd after passing the Rock Creek Trail bridge there is a steep uphill climb (about 12 degrees uphill ) to Avery where the Prius fuel efficiency can rapidly drop. At the top of the hill at Avery there is a traffic light which might require a full stop.. My current technic is to give a initial pulse as I am going downhill to achieve a desired top speed is upon arriving at the bottom of the hill ( e.g. 45 to 55 mph). If I think I will have to stop at the traffic light at the top of the hill my max speed is 46 mph is I think I will drive through the traffic light at the top of the hill my max speed at the bottom of the hill is from 50 to 55 mph. Before the Prius has climb up the next hill, I start up the gas engines about two seconds before the Prius needs to climb up the hill ( I start a bit earlier because my Prius in "ECO" mode and this compensates for the " ECO" throttle delay ). I am using a Scangauge II to monitor how much gasoline the ICE is burning.- on the uphill climb I am trying to maintain the initial speed as long as I can without having the the gasoline burn rate exceed over 1.15 gallons per hour (GPH). This gasoline consumption rate limit the ICE to roughly 1200 to 1400 rpm, wrt to the HSI display the HSI bar will be slighty pass the middle marker ( slightly to the right of the "C" of the "ECO" indicator). If I allow the velocity to drop about 15 mph ( e.g. from 50mph at the bottom of the hill to 35mph at the top of the hill). then I have minimized the negative impact of the uphil climb on the route's fuel efficiency. If I attempt to maintain speed while going uphill then there is a signifcant impact (e.g. loss of 7 mpg) on that segment' fuel efficiency. At the bottom of the hill, before starting uphill, my AVG mpg during the summer is normally about 58 to 65 mpg. After climbing uphill, if I try to maintain a speed from the bottom of the hill my AVG drops to about 52 to 56 mpg. After I climbig uphill, if the Prius throttle is restricted to less than 1.15 GPH and the speed is allow to gradually decrease then the AVG drops to about 58 to 62 mpg. Even if the Prius had an initial speed of 50 mph at the bottom and the traction battery SOC was at 100% and I have my tires overinflated to 44psi/42psi -- the Prius electric motors cannot pull the Prius to the top of this particular hill without the ICE. All my number, represent ideal driving conditions, when the temperature drops - my fuel efficiency drops significantly. (9_9)

. . .

. . .

I was hoping you might have considered doing a series of tests with and without P&G to document the differences, the scientific method. But it sounds like you are still trying to figure out how P&G applies to your specific environment, the DC area where I and my wife lived

1973-1985.

My wife and I used to live on Sherier Place near Arizona and MacArthur Blvd in DC, also Fairfax, Arlington, and Lanham Maryland. Many of the routes you've described are familiar but remembering DC traffic, there may be no opportunity for constant speed benchmarks, much less testing. It is a lovely area but with that traffic, you have my sympathy.

In contrast, Huntsville has a 'rush minute' that is easily avoided by leaving a little early or late, +/-15-30 minutes. I even have base access to Redstone which provides another set of low-traffic routes. So when the 1989 tornado blocked the main north-south roads, Redstone opened up Pershing and everyone got home that night without having to drive to the river bridges at Decatur or Scottsboro.

Regardless, you are welcome to look over my mileage records:

The cold weather is ending and once temperatures rise to 50F, mileage significantly improves.

GOOD LUCK! Bob Wilson Huntsville, AL

On Mar 15, 5:02 am, bwilson4web wrote: . . .

One late thought, you'll notice my early records were running two per week. I would fill-up the car Friday and Sunday evening to separate the weekly commuting mileage from the longer, weekend errands and drives. This allowed me to quickly understand what was happening with the car and how it behaved in different conditions.

With the newer, ZVW30:

I don't drive it that often so we're using refueling normally.

Bob Wilson