Neutral vs DFCO - 2008 Hyundai Accent

[GasSavers_maximilian]

I realized today I had all the information needed to directly compare DFCO vs coasting in neutral. The scenario used was on a level with a start speed of 60 mph and I just computed how far you'd have traveled by the time you had reduced to each speed (assuming that'd be your minimum allowable). DFCO distances are less, so you need to travel the difference under power, and since I know the total gasoline used in neutral, I can compute the mpg (makeup distance / fuel used in neutral) to get the same fuel economy. It's well above what I can achieve with my vehicle, so neutral wins hands down. Not exactly surprising.

[GasSavers_maximilian]

Realized I could just compute the makeup MPG for each 5 mph drop, rather than as a running total from 60 mph. The high value for the 55-50 drop next to the low one from 50-45 probably indicates a measurement error (or errors in the test track, etc.) and they should likely be in the middle. Trend is still pretty clear: neutral wins.

[FrugalFloyd]

Yep. The only times I can justify DFCO are when I have to scrub off speed fast. Usually it's on steep downhills where I'll build up speed too quickly in neutral

[bowtieguy]

Originally Posted by SentraSE-R

Yep. The only times I can justify DFCO are when I have to scrub off speed fast. Usually it's on steep downhills where I'll build up speed too quickly in neutral

uh, it's been WELL documented that DFCO has distinct advantages in certain applications.

[theholycow]

Originally Posted by bowtieguy

uh, it's been WELL documented that DFCO has distinct advantages in certain applications.

Yep, those situations match his description...when you need to lose speed.

[theholycow]

On a similar subject, I was thinking...it's been a long time since I've done any testing/measurements of acceleration gear choice efficiency. I previously did it when the car was new (and new to me) by trying a stategy for a whole tank, then a different strategy for the next.

I think it's about time for me to do some testing. I know for sure that cruising in the highest possible gear is always best (except uphill at 30mph where 4th works better than 5th). I'm beginning to question if I should experiment with accelerating harder or (god forbid!) higher RPM with less throttle.

I wouldn't mind a quicker, more definitive test. The test procedure and resulting data should be almost the same as your DFCO vs. neutral test, just replacing DFCO and neutral with a few different shift points.

Any suggestions how I can do a similar test and gather some decently accurate data?

[bowtieguy]

Originally Posted by theholycow

Yep, those situations match his description...when you need to lose speed.

i was thinking of many times when this is required, not just down hill.

[GasSavers_maximilian]

Originally Posted by theholycow

Any suggestions how I can do a similar test and gather some decently accurate data?

For what you're describing, it almost seems like just zeroing the current trip and getting a average for your test run might be all you need. I only gathered raw data so I could split things into pieces in order to compute a lot of different P&G scenarios. The DFCO vs neutral comparison was just a byproduct. I'd meant to do that comparison for a while but hadn't thought of how to account for the powered part of the DFCO leg. Seems obvious in hindsight (doesn't it always?).

As far as general tips for raw data collection, nothing that isn't pretty obvious. Having the flattest, longest stretch you can is a big one (unless you're testing hills of course). I just took a movie of my SG2 while going in both directions. If you can stand the tedium, more runs is always better. Then I averaged the runs for both directions, after first smoothing out any obvious errors. Also obvious, but make sure you have plenty of space on the ends of your test run so things settle down to steady state before you enter the range you're interested in.

I toy with the idea of getting a netbook and doing some data logging, but that's never going to happen.