Pulse and Glide vs Constant Speed

Why not try both and see what you get? I'm sure terrain makes a huge difference.

 

When I'm driving a constant 35 mph, I'm not in the "gas" part of the range (i.e. I'm in the first fat bar, not the second). This drains the battery down and requires the gas engine to kick on to recharge it.

This is inefficient, because of the conversion losses of converting gasoline to electricity.

Pulsing up to 40 mph, using the gas engine, and then gliding down to 30 mph (with no gas being used) yields a higher MPG value.

 

Many years ago, I ran a series of tests with our 2003 Prius:

• multiple passes along the same, level route in both directions
• constant entry point and exits for each
• recording data with a Graham Miniscanner
• same day, low wind, same temperature
• random mix of "pulse and glide" and "equivalent constant speed"

1. Cruise control pulse and "N" glide
   1. Enter the route on cruise control set to 43 mph, mark in the data the entry
   2. "disengage" cruise control
   3. when engine stopped, ~41 mph, shift into "N"
   4. when speed reached 25 mph (cruise control engage minimum is 23 mph), shift to "D" and hit resume
   5. repeat steps 1-4 until exiting the course and "mark' the data.
2. Cruise control steady speed of end-to-end time of the pulse-and-glide
   1. Enter the route on cruise control set to the block-time, speed equivalent of pulse and "N" glide
   2. Mark the data entry
   3. Mark the data exit

Here is the data and summary: Wilson PnG Test

I did do some 2010 Pulse and Glide experiments, again using the cruise control for acceleration and "N" for glide. The details are: P&G 45-60 mph | PriusChat

Also, here is another study thread: P&G or steady state driving? | PriusChat

CONCLUSION

I still drive using "CC" at every opportunity. The driver overhead of trying to P&G and consternation it makes on other traffic means it is really a technique best applied on a closed track. The incremental fuel savings is not enough to justify the headache.

Bob Wilson

 

I like the reproducibility you get by using the CC, but I don't think that's an accurate way to emulate how most hypermilers do P&G; the cruise control will tend to accelerate much faster (and running at much higher RPMs) than I would when trying to save gas. It would be interested to see the test repeated with a tachometer attached, and accelerating at 1800, 2000, and 2200 RPMs, for example.

That said, if you're not a certifiable nutcase like me, I suggest just driving the car however you're most comfortable.

 

I used to wonder why we don't see more accurate data from pulse and glide advocates in a world that has:

• GPS mouse records - there were and are serial, USB, and blue-tooth USB mice that make one second, records of vehicle location by longitude, latitude, altitude, and velocity. I have used them to record complete energy states for hours of data. Even the Garmin nuvi record vehicle ephemris which I have pulled from my own to measure trips including speed and altitude changes with 1-10 second intervals (Garmin does not use a fixed recording of GPS ephemeris.)
• OBD recording - I have both a Graham miniscanner and AutoEnginuity which have been used to record hours of vehicle metrics that have been combined with GPS data.
• three axis accelerometers - built into every Apple iPad, iPhone, and touch are three axis accelerometers and there are an abundance of apps to record and report these values. Better still, Gulf Coast Data Concepts makes an excellent, recording accelerometer, I have two, that sample up to 320 per second for hours of very accurate data. I used them to measure the 'brake pause' and later confirm the software patch fixed the problem and loaned them to other Prius drivers to document their braking symptoms.
• marathon drive logs - the driver logs for the 2005(?) Prius marathon and the original, Honda Insight, along with the protocol described in the SAE paper.
• all Prius have an electronic accelerator - this could be operated by a computer to give exactly, reproducible driving profiles as a function of just about any other computer detected sensor including the ones already mentioned . . . completely reproducible profiles.
• Excel, MathLab and probably a host of software packages exist that could model, fairly accurately, what pulse and glide advocates propose.

There are an abundance of metric and modeling systems that could be used to document what pulse and glide advocates do . . . yet I have managed to miss all except two marathon logs and one SAE paper. All of these data recording and modeling systems are 'off the shelf' so I remain surprised that we don't see more detailed, pulse and glide studies than the few I've found and my own. My studies use the cruise control built-in to every Prius and a simple, reproducible protocol and is something any Prius owner can follow.

Perhaps you might do what you've just proposed and share the results? Include the cruise-control protocol over the route to give a relative metric.

I am not totally happy with the built-in cruise control. It drops out at any speed under 23 mph and ignores the "Eco," "Normal," and "Pwr" modes. I had started looking at what it would take to make my own and the accelerator interface is the easiest part. But to build a practical system, the complexity rapidly increased and this was during the 'unintended accelerator' dust-up. I did not want to talk to a lawyer.

Bob Wilson

 

In perfect conditions I was able to do well over 100mpg with a 35mph average. I did this for 6hrs one time and for nearly 3 hours another time. This is using Pulse & Glide from 45mph to 25mph.

 

Did you still have a record any machine captured metrics?

Bob Wilson

 

I've found pulse & glide gives me at least a 10% gain. Well, pulse and glide and driving with load since it's hilly. I don't get that much opportunity to do it though, since I drive mostly on 2-lane roads and p&g requires either no following traffic or risk-free overtaking.

 

All I was doing was watching the Scangauge to monitor RPM for the most part. I used the built in display to show my mpg over time. It maxes at 100mpg though. Hand calculations showed 109mpg. This was with no net elevation change.

The 95mpg trip included some steady state freeway driving for 50miles at 60mph. That it what kept me from getting over 100mpg on that trip.

 

I was looking for recorded OBD or GPS metrics. The photos are nice but don't give enough information to replicate the exact, driving protocol. A movie might if the camera were recording the speedometer but an OBD or GPS would provide a much better source.

The OBD metrics need to include: vehicle speed, rpm, and either MAF or fuel injector timing sampled about once per second would be enough to understand the specific protocol. GPS data normally comes in one sample per second, fast enough, and also includes altitude changes.

The reason why photos are inadequate, is they do not show how anyone, even a computer, could replicate this one:

Pulse and glide was not used.

Bob Wilson

 

It was not a data collecting trip. It was an endurance trial.

I'm curious if there would have been any major changes in average mpg if you had gone longer. 10Min should be enough to cover a couple discharge/charge cycles but I'm not certain.

 

Homeland Security around the airport near F8L should be able to provide some security footage of his endurance trial.

 

I've found pulse and glide most useful at lower speeds in traffic, and constant speed better on the highway. There's a technique called "warp stealth" that is basically gliding for highway speeds, but I don't know how to pull it off, and gliding on the highway doesn't work well. Your speed drops from 90 too fast to make it practical, and accelerating to 90 from 60 or 70 requires a lot of energy. At low speeds though, pulse and glide has really helped my record, especially on short trips.

 

There was a span of 3 years when I really tried to eke out my best mpg driving the same road to work and back. I tried both, but got my better results driving constant speed between 40mph to 50mph. I didn't use cruise as the constant inclinations of the road did not make it as efficient.

The way i did it was once I get to speed I'd wait until i feel the cars inertia is stable. Then ever so slightly (but very quickly) i'd back off from accelerator to just where it barely maintains speed. This slight move would make my mpg go up from around 60mpg to 70mpg.

Then i'd slowly tilt my accelerator foot to it's right side so I can brace it against the floor. This lessened the minute accelerator spikes caused by road irregularities...(those were the days when Pedalogic wasn't available).

Doing just these 2 improved my full tank bests from 62mpg to 72mpg and stretched my visits to the fuel station from 2.5 to 3.5 weeks. I'd regularly make 750miles per tank.

Then I got lazy, lol.

 

That is consistent with Prius ICE operation.

P&G exists only because engine efficiency is often quite variable depending on operating conditions, and normal highway cruising was typically not at the peak efficiency 'sweet spot'. Better efficiency could be obtained by Pulsing near the sweet spot, then gliding with the engine either off or idling.

P&G worked best with old fashioned non-hybrids with fixed valve timing and transmissions geared only for performance, not efficiency. Once we add in better transmission gearing choices, variable valve timing, Atkinson-cycle engines, and ICE shutdown at lower speeds, then most of the potential P&G gains from the old days are already harvested automatically by these more advanced propulsion systems with much better EPA ratings.

There is far less to be gained from P&G with today's cars than with 20th Century cars. Those gains are not actually lost, but rather now rolled into much of the normal operation.

... probably because that was meant for Gen2 Priuses and Honda Civic Hybrids (HCH). While much better than traditional non-hybrids, they still had a smallish sort of 'sweet spot' on their engine BSFC charts. Warp Stealth put that to 'sweet spot' to good use.

The Gen3 Prius has a noticeably flatter (and slightly higher too) 'BSFC' chart than Gen2, effectively getting nearly 'peak efficiency' over a fairly wide range of operating speeds. That means the driver has to work even harder to find any gain at all, and when (or if) it is found, it will be significantly smaller than was possible on Gen2 and HCH. This is not something to lament, but rather to celebrate because the gain is (mostly) already built in, (almost) full time, without any driver effort.

If engine BSFC charts become completely flat, and propulsion systems achieve a uniform high efficiency across their full operating range, then P&G and Warp Stealth will be dead.

The Gen3 Liftback must spin its ICE above 46 mph, and often doesn't shut down until you drop several tics below that. Gen2 had a slightly lower threshold speed.

Most Prius P&G gains occur below those threshold speeds when the ICE can quit spinning completely during a Glide phase. When the ICE must keep spinning, there is far less potential gain to be harvested.

So, if you want to P&G at highway speed, get a Gen4, which has an ICE shutdown threshold speed high enough to use on the highway.

 

On a flat highway maybe pulse-and-glide benefits wane, but in typical around-town driving, you can constantly pulse-and-glide: with all the slow downs, stops, red lights, corners, grade ups-and-downs, left-turners, whatever, and no one will even notice.

Maybe that's not considered pulse-and-glide though, just strategic driving?

 

In hindsight, I realized that this in-town version is normally labeled by hypermilers as 'DWB' -- Driving Without Brakes. (Terminology note: It isn't really driving without brakes, but more of a planned reduction of brake use by letting off the accelerator earlier to burn less gas before the brakes will be needed, or delaying arrival to a traffic signal or slowdown late enough that a full stop is no longer even needed.)

But DWB and Pulse-and-Glide naturally stack together well, enhancing the benefits.

 

Maintaining a generous following distance helps a lot with DWB: you can "eat" some of it instead getting on the brakes. It's also good with waning reflexes, increased distraction.