CAN-View Data Analysis, Complete with Charts

Finally got it done. This is from the commute this morning to my part time job site. I'll see if I can get posted later the same charts as those from the full time ride, with annotations. It's a totally different route in more or less the opposite direction and through the middle of an urbanized area. But for now this shows the relationship between current and speed for neutral glides.

The Y axis on this has the opposite orientation as that of previous charts showing glides. Curiously, N-glides sometimes show positive current flow (more on that below) so I plotted positive values. Then it seemed more logical for numbers to increase going up the Y axis rather than down. Maybe I'll redo the previous chart later to make the Y axes match.



As stated, current flow moves within a more narrow range than with pedal-controlled glides, at speeds below 30 MPH. Interestingly as speed approaches 30 it appears to creep up a bit, and then above 30 it's all over the map. Finally, what's most interesting to me is this blip:



This is a neutral glide that I let transition to warp neutral coming down a fairly steep bridge approach. It carries me across the bridge and then up a short incline to a toll booth, where the last data point is.

On this and other steep hills previously I've noticed a brief surge of positive current at speeds around the mid-30s, often spiking to well over 10 amps, along with a simultaneous sensation of increased drag as though I had let it slip into regenerative coasting. (It might have been over 10 amps here too but CAN-View captures data only every two seconds or so, and that's about the duration of the positive current flow.) Then something I hadn't noticed before are the fluctuations of current flow in warp neutral being more erratic than most of those at slower speeds.

So, once again calling upon the experts: What explains this change in current flow above 30 MPH, especially the blip, while in neutral?

 

No, I keep it in neutral going downhill to avoid the ICE spin-up as the car crosses the 40 MPH threshold. Warp neutral is a term Hobbit uses (and may have coined) presumably to make the distinction between this and warp stealth.

 

Done. The test consisted of four runs with each method on a four-lane road through a lightly traveled industrial park. The route is shown here. There was a net elevation decrease of approximately 36' from start to finish.

I aimed to begin each run at approximately 28 MPH. Each glide was allowed to continue to a defined ending point 0.6 mile from the starting point.

Outside air temperatures were in the low 90s, and winds were out of the WSW at around 15 MPH -- both as reported by Weather Underground's almanac information for zip 23832.

Both front windows and the right rear window were opened for driver comfort and ventilation for the hybrid battery. Climate control was off to avoid the significant current draw from air conditioning.

Results, with charts for speeds and hybrid battery current flow:





As suspected, there clearly is more drag with neutral glides than with pedal-controlled glides.

The CAN-View data capture records battery SOC in half-percent increments. All glides began with SOC between 55.5% and 59%. SOC predictably dropped more with PC glides than with N glides: 1%-1.5% with each of the former and 0%-0.5% with each of the latter.

I will probably post these results in a new thread, with added discussion, for the benefit of the P&G regulars.

 

Thanks for the quick calculation.

I'm trying to get a handle on your formatting suggestion, but it's not quite sinking in. How specifically might you have done this one differently?

I'll work on enlarging the charts a bit. I've been taking the path of least resistance -- simply pasting into Paint and saving as a JPG file.

Unless I'm missing something, it seems to this somewhat uneducated mind that with the lower current flow, N-glides are more efficient. It seems generally accepted that use of battery power and subsequent replenishment introduce inefficiencies from conversion losses, and better from an efficiency standpoint is to avoid the battery as much as possible. I can't quantify it, however, beyond what the charted data show. (Well, maybe I can, but I can't put my hands on that spreadsheet at the moment -- must have left it on a different computer.)

 

Yes, I'm familiar with the mechanics of changing line and marker formatting. I was curious how you might have chosen to reformat mine. I'm always open to suggestions for improvement.

My first goal in formatting these most recent charts is to provide a quick at-a-glance visual of overall trends. Then by varying the colors and marker styles, folks can distinguish between individual runs for closer inspection.

Increasing the zoom level in Excel, or either the zoom level or image size in Paint, didn't change the size of the finished file. Maybe I need to learn a more sophisticated image editing application -- an area I haven't spent much time with.

 

Another test completed yesterday and today, this comparing MPG results using the EV switch with those not using it. These are from my morning commute (described in detail in post #5 above) the past two days.





Many of the PC regulars insist that the EV switch's benefit is marginal at best. My own experience suggests otherwise, but I wanted to document it with a (semi-controlled, at least) test. The main argument against its use is that depleted battery energy has to be replenished, with conversion losses suffered through discharge and recharge. But my main use is not for propulsion, but instead to force ICE shutdown during its warmup when propulsion is not needed.

Nonetheless, I sought to control and account for SOC. Though the goal is not to propel the car on battery power, ICE recharging is lost with EV mode. Therefore, I attempted to begin and end each run with identical SOC. The non-EV run was on Monday, and I simply began and ended it with what the car gave me: 54% to begin and 59% to end, both as displayed by CAN-View. Initial SOC Tuesday was 56%, so on first startup (with EV activated) I ran the AC until the display showed 54%, at which point I began data capture. SOC was displayed as 60% as I approached work, so I ran with a short EV mode segment to bring it down to 59%. Turns out, the recorded data showed the EV run beginning at 54.5% and the non-EV run at 53.5%. More on that in a minute. The data showed ending SOC for both runs to be 59%.

I determined from the data the "SOC equalization" point where SOC from the EV run had climbed to match SOC from the corresponding point on the non-EV run. It is at that point that I consider the MPG difference to be most meaningful -- whatever battery energy was used for EV mode has been replenished. I adjusted for the 1% startup difference in defining the equalization point: it was where 60% for the EV run matched or exceeded 59% for the non-EV run.

I noted cumulative trip MPG at each of 5 different benchmark locations, including destination. As luck would have it, the SOC equalization point fell very near benchmark #3 for the EV run, with that benchmark for the non-EV run lagging by about a half minute. So I use that benchmark's numbers for my results comparison: 73 MPG with the EV switch versus 69 without. Given that the EV run was on the cooler of the two days, the MPG difference is even more significant. Wayne Brown's simulator predicts a 2.3% improvement in fuel economy at 22 MPH with an increase in OAT from 60F to 65F.

22 MPH, BTW, was the combined average speed of the two runs up to the equalization point. Other average speeds:

• EV run, up to SOC equalization: 23.1 MPH
• EV run, total: 22.0 MPH
• Non-EV run, up to SOC equalization: 20.3 MPH
• Non-EV run, total: 20.7 MPH

Skies were clear both days and winds calm or nearly so.

This also will probably find its way into a separate thread with added discussion.

 

Not sure I remember seeing that one. Do you have a link?

Back to the EV mode tests ....

After that post, something in the charts jumped out that bothered me a bit. The non-EV run had three complete stops at red lights before the second benchmark. To find them all red is typical. But on the EV run I got lucky, in a sense, catching two of them green. So not having to completely stop could have biased the results in favor of the RV run.

So I repeated the EV run this morning. OAT was closer to that of the non-EV run: 64F. Winds were calm and skies clear. Two of the three early lights were red this time. Significantly, the SOC equalization point was considerably earlier, very near the second benchmark. Starting SOC was 54% (after another brief period of AC), so I adjusted for the 0.5% difference in starting SOC in defining the equalization point. Ending SOC again was 59%.

Given the equalization point's location this time, I consider fuel economy results at the second benchmark most significant for this comparison: 70 MPG vs. 60 MPG. Not needing to stop for the third red light could account for a small part of the difference, but nowhere near 10 MPG.

Here's the chart for the second EV run, along with a repeat of the non-EV chart. Pointers for the early stops are included; that and adjustment of the equalization point are the only changes to the non-EV chart.

 

In response to a discussion in another thread, I put some math to this and came up with 0.72 hp. I averaged current and voltage (which the CV data stream also includes) for each set of runs. I multiplied them to figure wattage and then converted that to horsepower. Here are the individual numbers, with times, amp-hours, and watt-hours thrown in for good measure:

	Column 1	Column 2	Column 3	Column 4
0	Per Run Averages	N Glides	PC Glides	Difference
1	Amps	1.260	3.638	2.378
2	Seconds	100.5	90	10.50
3	Hours	0.028	0.025	0.003
4	Amp-hours	0.035	0.091	0.056
5	Volts	226.6	225.3	1.3
6	Watts	285.6	819.6	534.0
7	Watt-hours	7.97	20.49	12.52
8	Horsepower	0.38	1.10	0.72

With these calculations comes an opportunity for error. Current and voltage readings are of course continuous, whereas CV’s data capture essentially does data sampling. And the sampling interval actually is slightly less than two seconds and seems to vary somewhat. I use two seconds (and state that they are approximate time intervals) for convenience and to keep the math simple.

Is my process correct?

(It sure would be nice if this forum had the code for building tables enabled! )
(EDIT: They do now, thanks to Danny's quick response to my suggesting it! :clap2

 

A discussion on braking and regeneration in another thread got me to wondering about the relationship between brake pedal position and regeneration. These are data combined from most of my recent trips for which I have data:



The CAN data stream captures brake pedal position on a scale of 0-127. (Why the odd number, I wonder?) Those data are converted to percentages for the chart.

The data include all data points with any braking (brake pedal position of >0) and regeneration (current flow into the battery) at >7 MPH (to eliminate any effect of low-speed friction braking).

Attila Vass has tested efficiency of regenerative braking using his own CAN monitoring setup. Based on current, voltage, and time to brake from 45 MPH to 5 MPH at various pedal pressures, he concluded that the most efficient braking occurs with a pedal position of 17 (13.4%). Interesting that, though the chart appears to demonstrate the expected relationship, current flow varies considerably at most positions represented -- including the 13-14% range. I figured the data points would be more tightly bunched.

 

Given my continual focus on the car's operation and my awareness of the effect of even minor changes in pedal input (brake and accelerator) on the car's behavior and fuel economy, I'd say I have pretty good pedal control. So my braking normally is pretty steady.

I have considered the effect of speed. I thought about charting speed in some way in a variant of this chart but I didn't have time yesterday. And it will probably be a couple of days. I left the flash drive with my data plugged into another computer at another location, and I won't be able to retrieve it until tomorrow or the next day. :frusty:

Braking aside, I assume higher MG1 and MG2 RPM levels lead to higher amounts of regeneration. If that's the case, then the next challenge will be determining what portion of regeneration at any given speed can be attributed to braking. Maybe (when I get my hands back on my data) I'll chart regenerative coasting current.