jray3
Posts: 1333
Joined: Tue Dec 06, 2011 1:05 am
Location: Tacoma area, WA
Contact: Website

Re: Regen vs Neutral and speed for max range

Sun Apr 30, 2017 10:25 pm

Llecentaur wrote:Thank you for the information.

Wonder if that would mean pulling the car while regenerating would actually pump up the battery really quickly :)

Yes, as this is essentially what happened when I took road trips with the pusher trailer. Considering that I could add bars while cruising at highway speed, it made for far faster progress than a CHAdeMO road trip, equaling the gasser trip times, but as a 20 mpg bastardized hybrid.

I still think the I-MiEV could make a great RV Tadpole, providing not just local transportation at one's destination, but boosted hill-climbing and acceleration in addition to sweet regen.
2012 i-SE "MR BEAN" 70,000 miles
i-ES traded at 21,648 miles
2000 Honda Odyssey
1987 F250 Diesel
1983 Grumman Kurbwatt,170 kW, 32 kWh
1983 Mazda RX-7 EV 43 kW 10 kWh
1971 "Karmann Eclectric" EV 240 kW 19 kWh
1965 Karmann Ghia Cabriolet

Aerowhatt
Posts: 318
Joined: Wed Jan 28, 2015 1:52 pm
Location: Albuquerque NM

Re: Regen vs Neutral and speed for max range

Mon May 01, 2017 8:38 am

JoeS wrote:Aerowhatt, I can tell this is going to be a protracted discussion, and I appreciate your presenting the math which I need to stare at a bit. One of the variables compounding the problem is the non-linear relationship of aerodynamic drag with speed.

In trying to find an example that might accompany this discussion, I went back to a trip I made five years ago soon after I bought the car (and before CaniOn) that might be of some use - it briefly shows the fuel gauge going up by one bar on a fairly short stretch of highway in the Santa Cruz mountains. There was nothing special about this trip, as it had no serious hypermiling associated with it nor was I trying to maximize regen as regen was simply used to slow the car down to keep from exceeding the 50mph speed limit in the hills. The datataking was simply me reading the pertinent i-MiEV gauges and Garmin GPS readings into a voice recorder. FWIW, here it is (I just added the GPS Visualizer graph and you need to click on the link to see the primary pdf graph):
http://myimiev.com/forum/viewtopic.php?f=28&t=269&p=993#p993

I'll be taking your comments into consideration and also try to take some CaniOn readings to see if I can develop a scenario that will quantitatively further develop this topic. For now, we are at an impasse. :cry:

Although the ball is in my court, everyone is welcome to join in.

To be continued...

[All this Marching for Science and Climate Change over the last two weekends has me motivated to play with the numbers!] :geek:


Initially I had trouble with my data too because of the relationship of aero drag with speed. That's one reason I kept it to myself and used canion on every permutation of the mountain drives. It goes against our intuition that going faster in (N) could be as or more efficient. It took a lot of corroborating data to sway my intuitive preconception.

The post just above where I bumped this thread was about a 500 meter elevation climb and descent with up to 8% grades. Four bars to climb and one bar returned on a regen required for a speed controlled descent. Your drive where you had one bar return was a ~1,800ft decent where "regen was simply used to slow the car down to keep from exceeding the 50mph speed limit in the hills" some of the grades were steep enough for effective (efficient?) regen. There seems to be a rough consensus about how much power can be returned by regen based on elevation traversed since these numbers roughly agree with mine.

Since we know that there is a baseline power usage to have the inverter maintain an active zero torque field in the spinning motor. It follows that regen is more efficient the harder it is braking the car. Hence it yields good results on steeper grades. For example with made up numbers - Lets say that 10 battery amps are need to maintain that rotating field and run the accessory loads. The stronger the regen the more efficient the energy capture process is. Don said early in this thread that being in D with the power needle in the zero position down a grade is the same as coasting. Reasonable enough I thought at the time. It turns out though that what it really is 0% efficiency regen! All the regen is feeding the overhead of regenerating to begin with. Slip that car into (N) and it will pick up speed! In the example it takes ten amps worth of braking force on the car to break even. On my steep mountain grade I see spikes of -68 amps on canion. Efficiency (electric only) would look like 68/68 +10 or 87% -- pretty good right! At lower regen amp rates though. Say 10 amps indicated going into the battery looks more like 10/10 +10 = 50%. With 5 amps indicated going into the battery becomes 5/5 +10 = 33%. So for gentler grades it takes 15 amps worth of regen braking force to net 5 amps of current back to the battery. 15 amps of braking force feels reasonably significant by the seat of my pants. Which qualitatively helps to understand how it might be roughly equivalent to the significant difference in wind drag force between 35 and 65 mph??

Aerowhatt
2014 cool silver ES, acquired new 4/2015
2014 Labrador Black Pearl ES, acquired new 3/2016

jray3
Posts: 1333
Joined: Tue Dec 06, 2011 1:05 am
Location: Tacoma area, WA
Contact: Website

Re: Regen vs Neutral and speed for max range

Mon May 01, 2017 10:10 am

I like the turn this thread is taking. If there is a lower amperage limit on regen efficiency, below which you do nothing but create heat in the motor and inverter, perhaps there is a good reason that Mitsu makes regen go away below about 12 mph. If you're just making heat, far better to do that in the brakes than in the electronics.

One thing that has been appealing to me about other EVs like the BMW i3 and the BOLT is their ability to smoothly regen down to a nearly complete stop. Maybe that's mostly for show...
2012 i-SE "MR BEAN" 70,000 miles
i-ES traded at 21,648 miles
2000 Honda Odyssey
1987 F250 Diesel
1983 Grumman Kurbwatt,170 kW, 32 kWh
1983 Mazda RX-7 EV 43 kW 10 kWh
1971 "Karmann Eclectric" EV 240 kW 19 kWh
1965 Karmann Ghia Cabriolet

JoeS
Site Moderator
Posts: 3148
Joined: Thu Dec 15, 2011 5:47 am
Location: Los Altos Hills, California

Re: Regen vs Neutral and speed for max range

Mon May 01, 2017 12:01 pm

The following is Off Topic but related to this topic, so I thought I'd share…

----------------------------------------------------------------------------------------------------------
Science Saved My i-MiEV (and a Deer)

I've been lucubrating on Aerowhatt's insightful writeups and thought it's about time to start getting a feel for the data we have available on CaniOn.

Where I live, from the point my side road enters the winding country road there is about a 1-1/2 mile (2.4km) gentle downhill on that country road with a 30mph(48km/h) speed limit before hitting the stop sign at the bottom. My habit for 40 years when getting onto this road is to accelerate to about 35mph (56km/h) and then drop the car into Neutral and simply coast all the way down to the stop sign, using regen in B to slow down just before the stop sign. The i-MiEV holds that speed constant very nicely, whereas my previous Gen1 Honda Insight and now Tesla accelerate to the point where I needed to start slowing them down. I attribute this to wind resistance and their lower Cd than the i-MiEV.

This morning on my way to Jazzercise, after accelerating and kicking the i-MiEV into Neutral I wondered what would happen if I simply dropped into D and let the car slow down of its own volition. Since there was no traffic, I did just that and was dismayed as the car slowed down more and more (after all, it is a gentle grade) … as it got down to 15mph I was just about to accelerate back up when a deer jumped right in front of the car! My reactions are still pretty good, and both the deer and car escaped unscathed (LOVE the i-MiEV brake rate application response which seriously applies heavy braking in an emergency). Had I been traveling at my usual 35mph, both the deer and the car would have crunched! Had to share…

----------------------------------------------------------------------------------------------------------

Aerowhatt, I appreciate what you wrote this morning, I'm still digesting all your information. I'm considering formulating a scenario that would enable us to perform some controlled tests. What's funny is that almost all of my (now aggressive) i-MiEV driving is invariably interspersed with stints in Neutral, and I've never faced a situation where I needed to eke out every last mile(km) of range and actually put this downhill regen vs. N into practice. That said, I will admit to occasionally restraining the Tesla on downhills and holding down its speed with slight regen in the belief that it would improve that trip's Wh/mi (Wh/km) readout.
EVs: 2 Wht/Blu SE Prem., '13 Tesla MS85, 3 156v CorbinSparrows (2 Li-ion), 24v EcoScoot(LiFePO4)
EV Conv: 156v '86 Ram PU, 144v '65 Saab 96
Hybrids: 48v1kW bike
ICE: '88 Isuzu Trooper. Mothballed: '67 Saab (orig.owner), '76 MBZ L206D RHD RV

Aerowhatt
Posts: 318
Joined: Wed Jan 28, 2015 1:52 pm
Location: Albuquerque NM

Re: Regen vs Neutral and speed for max range

Mon May 01, 2017 1:33 pm

The Tesla may work very differently in this (N) vs regen respect. I would be happy to do my best to quantify the Teslsa in this regard. Just loan it to me for 6 months of mountain trips ;)

Aerowhatt
2014 cool silver ES, acquired new 4/2015
2014 Labrador Black Pearl ES, acquired new 3/2016

Aerowhatt
Posts: 318
Joined: Wed Jan 28, 2015 1:52 pm
Location: Albuquerque NM

Re: Regen vs Neutral and speed for max range

Mon May 01, 2017 6:10 pm

JoeS wrote:The following is Off Topic but related to this topic, so I thought I'd share…

----------------------------------------------------------------------------------------------------------
Science Saved My i-MiEV (and a Deer)

I've been lucubrating on Aerowhatt's insightful writeups and thought it's about time to start getting a feel for the data we have available on CaniOn.

Where I live, from the point my side road enters the winding country road there is about a 1-1/2 mile (2.4km) gentle downhill on that country road with a 30mph(48km/h) speed limit before hitting the stop sign at the bottom. My habit for 40 years when getting onto this road is to accelerate to about 35mph (56km/h) and then drop the car into Neutral and simply coast all the way down to the stop sign, using regen in B to slow down just before the stop sign. The i-MiEV holds that speed constant very nicely, whereas my previous Gen1 Honda Insight and now Tesla accelerate to the point where I needed to start slowing them down. I attribute this to wind resistance and their lower Cd than the i-MiEV


Really glad you missed the deer, and it you, Really weird how something seemingly unrelated can affect your behavior, or timing, saving you a hassle, or catastrophe. Then again if you had just gone your normal 35 mph you might have been well past there when the deer crossed. Just strange how often this comes up in my universe.

So this trip down your road got me thinking. Since Canion only reads what the car gives it and mostly that is what is happening at the battery interface with the rest of the system. It can't show what the regen efficiency is at different speeds. I know what that rotating 0 torque field uses up at freeway speed but I would doubt that the requirement is a constant. Makes some sense that at lower motor RPM the field might require less power to support. Your road offers a possible opportunity to get a reading for 35 mph. So, you just match the steady state (N) coasting speed for the grade and take a few Canion battery amp draw numbers while in (D) and feathering the accelerator as steadily as possible to maintain the exact same speed. Average the readings and we have a 0 torque rotating field amp draw for a slower speed. I don't have any grades like you describe, that I know of. Everything around here is too steep, too flat, and/or too short. Just please have someone else read Canion so you can see the deer coming :)

jray3 wrote:I like the turn this thread is taking. If there is a lower amperage limit on regen efficiency, below which you do nothing but create heat in the motor and inverter, perhaps there is a good reason that Mitsu makes regen go away below about 12 mph. If you're just making heat, far better to do that in the brakes than in the electronics.


Just guessing here, but I think Mitsu purposely made the car act like a normal automatic ICE as best they could. Being first to market and an electric car being such a foreign concept to most of the population. Likely a good Idea to make it "feel" as normal as possible??

Aerowhatt
2014 cool silver ES, acquired new 4/2015
2014 Labrador Black Pearl ES, acquired new 3/2016

JoeS
Site Moderator
Posts: 3148
Joined: Thu Dec 15, 2011 5:47 am
Location: Los Altos Hills, California

Re: Regen vs Neutral and speed for max range

Tue May 02, 2017 12:04 am

Aerowhatt wrote:...Your road offers a possible opportunity to get a reading for 35 mph. So, you just match the steady state (N) coasting speed for the grade and take a few Canion battery amp draw numbers while in (D) and feathering the accelerator as steadily as possible to maintain the exact same speed. Average the readings and we have a 0 torque rotating field amp draw for a slower speed. ...
Good idea, and I'll start doing test runs first to see where and how exactly the car behaves in Neutral (I like 35mph, for various reasons) and then simply start recording the data each time I go to town at that exact speed. I recall how difficult it was when, in the pre-CaniOn days, I had a Hall-effect ammeter clamped onto the battery cable and tried to hold the car at a steady speed while measuring amps. Amazingly close, when later validated by CaniOn.

Over all these years of driving the i-MiEV, I've always "felt" that keeping the power needle at the zero point when in one of the drive settings was simply not the same as kicking it into Neutral. Now perhaps I'll have some data to either validate or disprove this supposition. Edit: I should have noted that this is yet a different test that I would like to run in addition to the one described above

I keep my CaniOn set to metric, as the primary figure of merit I look for on each trip (when I have CaniOn on) is Wh/km, with my goal being less than 100 (including freeway driving) if I'm trying to drive efficiently. I've never bothered with the instantaneous reading, but perhaps now should start looking at it just to get a feel for it. For the first 8000 miles of ownership I was obsessed with recording my i-MiEV's wall-to-wheels consumption, so when you say miles/kWh I think in terms of end-to-end and not just battery-to-wheels.

Will post here when I have some meaningful data.
EVs: 2 Wht/Blu SE Prem., '13 Tesla MS85, 3 156v CorbinSparrows (2 Li-ion), 24v EcoScoot(LiFePO4)
EV Conv: 156v '86 Ram PU, 144v '65 Saab 96
Hybrids: 48v1kW bike
ICE: '88 Isuzu Trooper. Mothballed: '67 Saab (orig.owner), '76 MBZ L206D RHD RV

JoeS
Site Moderator
Posts: 3148
Joined: Thu Dec 15, 2011 5:47 am
Location: Los Altos Hills, California

Re: Regen vs Neutral and speed for max range

Tue May 02, 2017 11:13 am

So, as I'm doing my Jazzercise (I'm the token male in the class), led by a beautiful blonde cheerleader, I try to keep my mind occupied and thus this morning formulated the following summary statement attempting to define what it is we are attempting to prove or disprove. I'm putting this on the table for discussion and refinement:

The Hypothesis:

On any given downhill, placing the i-MiEV into Neutral and coasting is more efficient than leaving the car in "gear" and regenerating.

I would like to define Efficiency as based on the difference in State of Charge (SoC) at the bottom of the hill vs. SoC at the top of the hill. This thread's title uses the term 'max range', and I submit that the simple concept of having a larger SoC at the bottom of the hill using one or the other technique will demonstrate this.

I recognize that this statement encompasses uncontrolled variables such as time, speed, distance, altitude, and road grade, to name a few. May I suggest that, in order to bound our investigation, the following variables be limited:

Speed - lower end of 35mph, which also happens to be the truck speed limit on many Interstate steep (≥6%) downgrades; I'm suggesting an upper limit of 70mph so we can both stay legal and not over-rev the i-MiEV motor.
Note: one of the parameters I hope we can establish is a grade % vs. coasting speed for the i-MiEV at any given altitude.

Time and distance - I am putting no restrictions on these two as they will be bounded by the distance travelled and the car's speeed.

Road Grade - this will probably be naturally limited by the availability of grades on which to perform our tests

Altitude - this will be significant, as the vehicle coasting in Neutral will achieve much different losses for any given grade depending on altitude.

Other variables such as air temperature, windspeed (and its relationship to the car), etc. should simply be noted.

Ok, I'll shut up now and open this up for discussion - do we have agreement on the Hypothesis (or whatever we wish to call that statement)?
EVs: 2 Wht/Blu SE Prem., '13 Tesla MS85, 3 156v CorbinSparrows (2 Li-ion), 24v EcoScoot(LiFePO4)
EV Conv: 156v '86 Ram PU, 144v '65 Saab 96
Hybrids: 48v1kW bike
ICE: '88 Isuzu Trooper. Mothballed: '67 Saab (orig.owner), '76 MBZ L206D RHD RV

JoeS
Site Moderator
Posts: 3148
Joined: Thu Dec 15, 2011 5:47 am
Location: Los Altos Hills, California

Re: Regen vs Neutral and speed for max range

Wed May 03, 2017 8:29 pm

Well, this was an interesting, but frustrating, exercise. Don't bother reading any further, as the only one who might be interested in the following is Aerowhatt.

Here's the elevation chart of the road I take:

Image

The objective was to first get the car up to 35mph, kick it into Neutral, and then coast all the way down. The idea was to repeat this with the car in D (or Eco or B) and hold the accelerator steady to hold the 35 mph speed and then record the power consumption (ammeter reading). Simple, huh?

Here's what happens normally -

Looking at the above graph, it is fairly level for the first 0.1-mile with a fairly sharp curve so I use that distance to accelerate up to 35mph, kicking it into Neutral at the 0.1-mile mark.

The car maintains a speed of 33mph-35mph for the distance from 0.1-mile to 0.8-mile, despite the road being somewhat curvy in places. In this mode, CaniOn reads 1.3A and instantaneous 8Wh/km. At the 0.8-mile mark the car starts accelerating and by the time it gets down to the 1.2-mile mark it is almost up to 40mph.

I realized why I never see that 40 mph in my everyday driving: at around the 0.9-mile mark (last curve followed by straight stretch) I reflexively drop it into D then Eco then B and consciously slow the car down using regen so that it is crawling by the time it hits the stop sign at the bottom (1.3mi). Also, that last stretch is occasionally a speed trap by the local sheriff (the speed limit is 30 mph), so by playing with regen I automatically avoid that issue. That's kinda my force-of-habit method of driving that road daily in the i-MiEV for the past five years.

Incidentally, with CaniOn, my figure of merit is to drop down my cumulative trip consumption to zero Wh/km by the time I hit that stop sign - it takes that long to counteract the sharp initial acceleration power draw, and the final regen really brings that number down fast.

So, what happened today when I tried it in D -

What I decided to do was make a few runs in only that nice steady first portion of that drive: between 0.1 miles and 0.8 miles, and record the data using a voice recorder. In a nutshell, in my first few attempts today, trying to keep a steady foot on the accelerator in order to keep the speed in the 33mph-35mph range and then read the Amps on CaniOn proved to be impossible. The Amps were all over the map, from a negative maybe 2A (slight regen) to a positive maybe 9A, as the car is operating in a very lightly-loaded region. The amps number was different every time I looked at it! I need to keep my foot from twitching or else I need a mechanical throttle with a vernier on it!

FWIW, it was our first very warm summer day today, afternoon temperature 87degF (30degC), battery was at 65%SoC, battery average temperature was 26degC (79degF) and RR=44 - just some meaningless data points.

On the drives back up the hill I was better able to steady out the fluctuations, with perhaps an average of 25A at 35mph on that stretch between the 0.8mi and 0.1mi markers. Have to think about it, as that info might prove to be significant.

What does all this tell us? Nothing! Over the next few days I will now see if I can develop a really gentle non-twitchy featherfoot to keep that accelerator pedal constant with the car at 35mph and then glance over at CaniOn for the amp reading, without worrying about deer jumping out in front of the car. To be continued...

Edit: I just had a duh! moment, as I forgot that I can upload the CaniOn data and can then dissect it. Also, I can take a screenshot of the CaniOn graph...
EVs: 2 Wht/Blu SE Prem., '13 Tesla MS85, 3 156v CorbinSparrows (2 Li-ion), 24v EcoScoot(LiFePO4)
EV Conv: 156v '86 Ram PU, 144v '65 Saab 96
Hybrids: 48v1kW bike
ICE: '88 Isuzu Trooper. Mothballed: '67 Saab (orig.owner), '76 MBZ L206D RHD RV

Aerowhatt
Posts: 318
Joined: Wed Jan 28, 2015 1:52 pm
Location: Albuquerque NM

Re: Regen vs Neutral and speed for max range

Thu May 04, 2017 6:28 am

I like to analyze things qualitatively and then quantify it if needed. Why? simply because it is easier. So just as a suggestion, I would tackle evaluating your descent more simply first.

Record SOC at the top and the bottom and also record Wh/km at the bottom for each case. Then compare them and see which delivered better power stingy results.

So you would have SOC difference between the top and bottom for (N) and SOC difference between top and bottom for (D). Plus Wh/km for both to compare.

It's might also be better to use Eco for the "in gear" descent since it is easier to be steady on the throttle with that modes less sensitive throttle. Then your only focus it to match the speed in both runs. This would give clean simple data that anyone can wrap their head around. If the two runs are super close in power consumption then it could be quantified more to discern the winner.

Aerowhatt
Last edited by Aerowhatt on Thu May 04, 2017 1:53 pm, edited 1 time in total.
2014 cool silver ES, acquired new 4/2015
2014 Labrador Black Pearl ES, acquired new 3/2016

Return to “Driving the iMiev - Range - Energy Efficiency - Operating Cost”