Well, maybe not a reply exactly.

This evening I drove 51 miles (round trip) down the 405 freeway in Southern Calif. The first stage was driven from 6-6:45 pm. Temp was 70F so I drove with the air conditioning off, and the windows closed.

At the start of the drive (in my driveway) my avg mpg was 51.3.

About 1/4 of a mile and I was S. bound on the 405. Traffic was moderate, i.e., slow (25-30) in spots, otherwise moving at around 50-60 mph. Where I could, I set the cruise control at 59. Any climbs pulled the mpg down to 25 mph, but on the reverse slope, it went up to 99.

So, anyway the trip S. brought the mpg from 51.3 to 52. Hmm. Not too shabby.

The return was at 9 pm, with light traffic. About 1.5 miles on the street and I was N. bound on the 405. I set the cruise control at 61. Once again, on the hills/overpasses, the mpg dropped to 25, while on the reverse slope it climbed to 99.

Once back in Long Beach, 1.5 miles to home, I pretty much ran on the battery (99 mpg).

The climbs seemed to be - at most - .1 miles, so perhaps we would want to consider this a flat route.

End of run=52.3 mpg. As I said, not too shabby.

 

Correct. I included a link to the spreadsheet so folks could check my work and make local modifications:

http://hiwaay.net/~bzwilson/prius/pri_hill_climb.xls


Thank you. I didn't have a good formula for the aerodynamic drag effects apart from rolling and transaxle effects. Do you have handy a simular formula for the NHW20?

What I might do is put the "190" and "0.42" as separate vehicle constants. Once we have these values, we can show the efficiency for both vehicles.

I found it from a Google search. My understanding is gasoline energy content can change by a couple of percent and the 121 MJ seemed a reasonable value if possibly a bit low. Here in the States, we often have an E85 blend which has a slightly lower energy value. Consider it a SWAG (Some Wild *** Guess) useful for a relative efficiency number.

I agree but am uncertain about the distance needed before and after the hill. In a steady state condition, our vehicles like to maintain a 60% battery SOC (if I understand the papers on this correctly.) What I don't know is how long it takes for our vehicles on a nominally level surface to reach the steady-state battery charge. This would define the minimum run-up and run-out distances (speed related) needed to normalize the battery at the base of the hill and after the hill climb.

Some of my other experiments may result in getting battery SOC from my NHW11. Since the NHW11 has a greater battery capacity than the NHW20, the run-up and run-out distances should have plenty of margin for an NHW20.

Bob Wilson

 

Hi Bob,

Thank you for your comments.
You're welcome.

I don't have the formula for the NHW-20.
However, I think we can guess about that.
The "190" part is based on the rolling resistance and it is proportional to the vehicle weight.
The weight for NHW-11 and NHW-20 are almost same, so we can use th "190" for the NHW-20 too.
The ".42" part is based on the air drag and it is proportional to the Cd value.
The ratio is 0.26:0.29, so it can be 0.42 * 0.26 / 0.29 = 0.38 .
Ah, I mean how much fuel was used.
Is that from your reading of the mileage number on MFD?
Anyway, I believe the longer distance is better.

Ken@Japan

 

Yes. The protocol was to approach the hill from the bottom on cruise control at the target speed. Then at the lowest point, hit the MFD 'reset' and ride up the hill. When the landmark at the top was reached, read the MFD's MPG value. Then to calculate the fuel burn, we used the linear distance measured with a trip meter. A simple division of the linear (road) distance in miles by miles-per-gallon gives the fuel burn.

Given the short distances and potential errors in the running distance, there is plenty of opportunity for rounding errors. However, for comparing performance going up the hill, it works.

Bob Wilson

 

Thank you for your comments.
I think the MFD's MPG value is updated not so often, every several seconds to a minutes.
It takes only 40 seconds to drive 1.5 miles at 80 mph.
I believe we need more data to get the accurate (within a few percent error) results.

Ken@Japan

 

Hi Ken,

Thanks again for the comments. I too have some concerns about the MFD's MPG display values:
My experience has been that after RESET, the display updates are seconds apart. But as the duration of the run increases, the 'averaging' reduces the apparent update rate. BUT I suspect the MFD display number is accurate only over a finite interval.

On the weekend, I have combined highly efficient urban driving with longer highway trips. But when I compared the MFD display to the pump values, the disagreement strongly suggested that the most recent driving had more to do with the MFD display value when I filled up than the total weekend driving.

I suspect the MFD display is relatively accurate only for 30-35 minutes because the running bar graph only covers a 30 minute interval plus the current 5 minute interval being calculated. From the behavior I've seen, I suspect it is a running average that uses this formulation:

MFD_Avg = ( (n-1)*(last_average) + (sample_average) ) / n

n = number of samples in a 30-35 minute interval

To test this, I need to run a series of tests:

(1) RESET: 30 minutes at a given fuel consumption rate, R1
(2) RESET: 30 minutes at a given fuel consumption rate, R2
(3) RESET: 30 minutes at R1 followed by 30 minutes at R2

Expected results:
(A) if the MPG display is (R1+R2) / 2 -> there is no 30 minute, running average
(B) if the MPG display is exactly R2 -> there is exactly a 30 minute, running average
(C) if between R1 and R2 -> solve for the unknown interval and retest

Unfortunately, this is a difficult protocol 'in the field.' I need a flat road that I can run two distinctly different speeds for one hour each way and a weather condition with no wind and relatively stable temperatures. This suggests a night test under cloud cover after midnight. This is a much easier protocol if I can get time on a dynometer.

The formula I used was:

(distance/speed) * (minutes/hour) * (seconds/minute) = 67.5 seconds

Bob Wilson

 

My assumption is just...

MFD_Avg = miles_driven / gallons_consumed
(done by unknown intervals)

Anyway, my SuperMID M-1 does samething every 0.5 second.
You won, I lost.
Mine was for 1 mile at 90 mph by unknown reason.

Ken@Japan

 

Hi Ken,

Thanks, I make so many keystroke errors, it is good to know I'm not alone.
One thing, driving up the hill at 80 mph was somewhat exciting. I was passing the pickup trucks and SUVs like they were 'standing still', unlike their usual relationship when driving to work. The car handled fine but it was starting to get a little 'bouncy'. I suspect 85 mph would be OK but 90 mph would be too exciting.

That stretch of road is posted at 65 mph but going more than 15 mph over the speed limit becomes 'reckless driving' with a substantally increased penalty. Worse, the last curve at the top is not really banked for speeds above 70-75 mph. At 80 mph, the tires worked fine but higher speeds are likly to do a side pressure test I'm not really interested in conducting.

Bob Wilson

 

Using a mini-scanner to record the data, I replicated the hill climb and got this data:


One of the curious things was the fall-off in battery power contributing to the climb and subsequent reduction in kinetic energy. Also, I was able to measure engine braking power going down the 6% grade. BTW, the shape of the ICE power between the end of the hill climb and the turnaround nicely matches the topology of the road.

Bob Wilson