[bwilson4web]

SAE 2009-01-1322, "Vehicle Inertia Impact on Fuel Consumption of Conventional and Hybrid Electric Vehicles Using Acceleration and Coast Driving Strategy," Jeongwoo Lee and Douglas J. Nelson, April 2009. For $15, anyone can get a copy and it is a fine paper discussing 'Pulse and Glide.' For example, Table 4, Summary of Results of 2004 Toyota Prius:


In May 2007, I also looked at Pulse and Glide with my 2003 Prius and used the following profile:

My test protocol was built around using the Prius cruise control so anyone could replicate the procedure (i.e., "Johnny cab" driving style)

• acceleration - use cruise control "resume" to provide a consistent acceleration profile that is independent of the driver.
• stop cruise control - at 43 mph use the 'stop' cruise control function so the vehicle will slow down and pass through 42 mph and auto-stop the engine. This is a control law of the NHW11 and NHW20 Prius.
• shift into "neutral" at 40 mph - for a no-human glide
• shift into "D" and hit resume - puts the Prius back in acceleration to the target speed.

So looking at the field parameters, it turned out to be:

• 15 seconds acceleration - midway between the 10 seconds in the second row of Table 4 and 20 seconds in the third row.
• 43-25 mph - compared to 40 to 30 mph in the second row.

I'd plotted my results and got:

So we can now compare table 4, second and third row, 10 and 20 second acceleration, with my 15 second acceleration results:

• PnG - 151.5-99.9 (paper) vs 87.8 MPG (NHW11 road test)
• Cruising - 81.0-75.9 (paper test) vs 79.0 (my test)

The only notable difference is the traction battery state of charge. I didn't bother to record it but simply averaged five runs. But Table 4 shows the State of Charge for all PnG simulations and tests. Carried to its logical conclusion, Pulse and Glide either in the simulation or the Test would eventually discharge the traction battery to the point it would no longer run. Certainly discharging a traction battery to supply motive power would increase the apparent MPG of the SAE 2009-01-1322 Prius.

I also notice that the largest difference between the paper and my results, 151.5 MPG versus 87.8 MPG, occurred concurrent with the largest change in the State of Charge, 52.0 to 50.5 (row two.) We already know that traction battery energy can substitute for fuel and lead to unusually high MPG. I've had private e-mails with Dr. Nelson but I never brought it up and we never discussed the State of Charge impact.

Now i can repeat my NHW11 tests and record state of charge. I would also increase the number of runs, say 35, so we would have a statistically significant set of data. Finally, I would also make sure the milage and speeds were GPS calibrated and adopt a 40-30 PnG speed protocol. I may yet do this but I seriously doubt it would change the results.

What may be equally interesting is to replicate the test using a ZVH30, the 2010 Prius. I have one and it is just a question of repeating the test protocol and gathering the data. I'll see what I can get later this evening.

There is one other area were . . . Jeongwoo Lee, Douglas J. Nelson and I agree:
. . .

Lee_Nelson said:

. . .
There are some known safety issues using the PnG driving strategy on the road; traffic considerations, unstable braking performance, and unstable power steering system if the engine is off, so this driving technique should only be used on a closed course under controlled conditions.
. . .

Wilson said:

. . .
I agree that Pulse and Glide needs to be limited to closed tracks or very low traffic roads. The speed differences required for pulse and glide means as the number of vehicles increase, the probability of having to brake and need to maintain larger vehicle safety distances increases. This reduces either the effectiveness of Pulse and Glide or road vehicle capacity. It is a technique best practiced solo.
. . .

I've been waiting for the right time to discuss SAE 2009-01-1322 and serendipity, the time has arrived. There was one comment and additional data point:

#19 (permalink)
Sun, Jun 21st, 2009, 11:10 pm
bwilson4web

HybridDriver said:
. . . In the article by Jeongwoo Lee and Douglas J. Nelson how many miles did they test PnG for?

They used a four-wheel, dynamometer from Argonne National Laboratory:

HybridDriver said:
. . . If you, me, they, do a test for a long enough number of miles ( 50 to 100 miles ) then the battery SOC becomes insignificant, dare I say, irrelevant. The pack SOC will either go up-down-up-down or, reach a low point, and maintain there. Regardless, it will reach stasis after some number of miles.

I'm a little worried about using a dynamometer for an acceleration test. You are dependent upon the dynamometer replicating the inertial mass. This is possible but I would have preferred seeing the results replicated in the field.

HybridDriver said:
. . .In my own tests, I can get 10% higher FE with PnG vs. steady speed driving. But, as you know, conditions have to be right for it. And the work involved, is sometimes not worth an extra 10%. But it's fun to experiment.

My data shows closer to an 11% improvement at the maximum 43-25 speed range. Sad to say, when I tried to wrap the optimum speed 18-20 mph, it looks like PnG was worse.

Bob Wilson

[bwilson4web]

Hi,

I ran a preliminary test with our ZVW30 (2010 Prius) this evening at 1:30 AM on a circular route near work. Again, using cruise control at 77F:

1. 66.2 MPG - PnG40-30, using cruise control on a 2.8 mile circular route.
2. 77.3 MPG - steady 35 mph, cruise control, same 2.8 circular route
3. 77.2 MPG - PnG42-30, using improved cruise control on 2.8 mile route

The key to improving the PnG performance was to 'over shoot' the target glide speed and pull the cruise control stalk to trigger an engine stop. At 40 mph, I shift into "N" and the results were much better than just hitting "N" when reaching 40 mph.

This route has two turns that require slowing down to 23 mph. My regular test route has no turns. So I still need to repeat these tests on a better route. However, this preliminary data doesn't look so good.

Bob Wilson