Practical range/power consumption testing (serious)

How-Tos & Esk8 mechanics
1

After reading the garbage fire that is a certain thread on german hub boards, I have decided to share some research that I have been doing with the forum. I was always curious how different aspects of our boards and riding styles affected how much range you can actually get from a board. So I read a bunch of threads on here and the old forum and other stuff on the internet. The conclusion that I came to was that no one has released reasonably scientific practical testing results for Esk8.

So I decided to do just that. I found a nice testing area, setup a couple of profiles in Metr, and tested one aspect of range: Wind speed.

For any reasonably scientific test, the objective is to change one variable and keep all others as constant as reasonably possible. Considering that I donā€™t have access to a wind tunnel, large enclosed course or dynamometer, I canā€™t easily get separate aerodynamic drag and mechanical/electrical loss figures. So I have to use the total system power consumption. This test is meant to measure the power consumption in a ~13mph headwind and a ~13mph tailwind at different cruising speeds.

Hypothesis: Power consumption should massively increase when riding into the wind. When riding with the wind, power consumption will be very low until riding faster than the wind, then it should increase like normal.

For this test I used:

  • DIY dual gear drive with Metr. (exact specs are not relevant to this test)
  • 95Kg rider (me)
  • Quarter mile long, flat, outside parking lot
  • eRPM speed limiting using Metr (Note: real speeds were reduced slightly by voltage sag. Both speeds will be listed in results.)
  • All runs were done in the exact same riding position, standing straight chest forward.
  • All runs were done in a straight line on the same section of the lot.
  • Results are an average of the power consumption at the test speed.
  • All runs (except 5mph) were of the same distance
  • The reference battery is a 10s2p 30q pack (216Wh). Most chinaboards that use this have a listed range of 12-14 miles at 15mph.

Now with that out of the way, the results of my test:

Results
Metr log
Headwind

Speed (mph) Power consumption (Wh/Mi) Range (Mi)
5 11 19.6
9.6 16 13.5
14.3 23 9.4
18.9 32 6.8
23.3 39 5.5

Tailwind

Speed (mph) Power consumption (Wh/Mi) Range (Mi)
5 3 72
10 5 43.2
14.7 7 30.9
19.6 9 24
24.1 18 12

Conclusions
Headwinds and tailwinds can massively affect power consumption. It appears that the hypothesis was proven, as all the headwind data is much higher than the tailwind data, and the tailwind data starts increasing significantly above 15mph. Riding fast in a stiff headwind can halve the range of your board.

Future posts with more tests should be forthcoming, as I have already tested them, I just need to write them up. @longhairedboy

If you have recommendations for additional tests, comments on my methodology, or other comments for me directly, please reply to this post directly, tag me, or PM me as Iā€™m not that amazing at following threads.

Edit: It occurred to me that I left out the most important part. The aerodynamic drag equation.
Drag = 1/2(air density) X relative wind speed ^2 X how aerodynamic you are (CoD X frontal area)
Please note how your speed relative to the wind is squared. This means that double the relative speed isnā€™t double the energy requirement to push the air out of the way, it is four times the energy.

20 Likes
2

Good stuff! It is quite surprising how much effect wind can have on your range, something I didnā€™t consider at all when I first got into this. Great write up. Calculated range and real world range there can be such a huge gap! And there are soooo many factors to consider. This is a great test, thank you!

1 Like
3

Music to my ears.

5 Likes
4

What is this witchcraft you speak of?

4 Likes
5

Original thread

6

That data sure was statistically significant. I was concerned it wouldnā€™t be because of start/stops.

Efficiency falls off massively the faster you go

1 Like
7

Regarding methodology and the drag equation since it has to do with cross sectional area I imagine your size and stance will effect things but I can certainly feel the extra drag effect from the wind and standing vs crouching or reducing my profile vs flat chest forward seems to make a difference on extra windy days.

8

Air drag is exponential of speed, this is not news, what needs to happen is a realistic range estimation at real riding speeds on flat ground assuming no wind. As EPA rate MPG on cars, board range should be measured at a set speed, say 20mph, from full charge until it drops from that speed.

These extended range estimations of a board going 5mph down a slight incline for the next 200km is just useless.

3 Likes
9

Agree about the instantaneous value being extrapolated not being realistic but imagine it is being done to try and emphasize the point and use the limited data available from the runs.

Regarding ā€œstandardizationā€ of testing finding an open road with no intersections where I can just go 20mph consistently is pretty tough, I usually peak around 20mph and average 12mph on rides to the local convenience or grocery stores. 5mph vs 10mph vs 15mph tests seem more reasonable without riding at some dangerous speed (unless you run more than 20mph there is at least some risk of sliding out)

Also for a how long does board sustain mph would be effected by the wind and other factors as the test here is showing so you want to do it like rc pilots etc. Half the test with wind to your back half the test and inclination 0 or one way then second half of test other direction then second half opposite direction and average things to negate the wind and other environmental factors that should 0 out mostly when changing direction.

1 Like
10

For gentle riding at 15mph, I agree with this. If you ride at a higher average speed, this number is not very accurate however. This test and some others that I have not written up yet suggest that aerodynamic losses are far greater than electrical and mechanical losses combined at the speeds we ride at. Although there are some component choices which consume vastly more power (foamies, certain AT wheels) than a standard street build.

I have actually tested that, but I havenā€™t written it up yet. At 25mph, a racing tuck is 36% more efficient than standing straight up.

Not only is air drag the exponential of speed, its also the single largest cause of energy consumption by a huge margin. Air drag is really a function of you and speed, not the board at all. So if you know how much energy your riding style consumes in Wh/Mi you can extrapolate to any similar board. In addition, my consumption numbers during these conditions will not match your consumption numbers as my frontal area is different than yours.

The extrapolated range numbers serve to emphasize the point. Itā€™s not easy for people to wrap their heads around Wh/Mi if they donā€™t have prior experience with this sort of testing. My target for this article is people new to esk8 through to experts. It should be pretty obvious that my quarter mile parking lot is not 72 miles long and that I canā€™t afford to spend fourteen hours riding to achieve this particular result in practice. But lets be real, the 72 mile figure isnā€™t the important one, the 5.5 mile figure is. In a stiff headwind, you can loose half your range easily. If you didnā€™t know that in advance, you could easily run out of charge in an inconvenient place while riding in those conditions.

I have tried this previously, and I found that it reduces the magnitude of environmental effects, but I still got 1 - 2 Wh/Mi variations between tests on different days. When you are looking for changes with a very small performance impact, this variability will render any data collected largely useless. The only solution I have come up with is to conduct tests on calm days or days with steady wind, conduct all tests in the same direction, and assume a 10% measurement variability between days. Testing inside a large warehouse would largely eliminate this issue but I donā€™t have access to one.

One other thing that I am considering doing is using an anemometer to measure the wind speed and ride in the same direction as the wind at the same speed to largely remove aerodynamic losses to find the real magnitude of electrical and mechanical losses. Unfortunately, gathering enough data points to really measure that at different speeds will probably take multiple months.

4 Likes
11

To the range discussion:

For me personally itā€™s legit to state 10Wh/km for average consumption for range. From an engineers perspective eboards are (almost) all the same regarding efficiency. They all use the same cemistry of batterys, same motor architecture, have similar weight (sure some boards can be double as heavy but in a very small area) and use similar wheels. Itā€™s a very different story for AT wheels, sure! It is not like in cars where you have a 3 ton pickup with 7 liter v8 and a small city car with 1.2 liter turbocharged engine. Differences in architecture (hub, belt, gear) have different pros and cons (less friction but worse heat dissipation for hubs). All in all the equal out themselves in normal testing conditions. At least from my experience. When Iā€™m going for a long ride Iā€™m not bombing all the time. Also I donā€™t have hills in my area. 10Wh/km is something bioboards, all chinese brands and so on use for their range estimation. So the guys stating here 12Wh/km which is completely fine imho. Look at the reviews of daniel kwan or press reset, when I have that right in mind you can estimate the range by using the Wh of the battery quite accurate (Ā±5% seems ok).

Different story with AT wheels, personally i would half the range for an AT board (20-25Wh/km).

For the speed:

Without GPS it is nothing worth. You could also have lifted up the board, go full throttle and the remote would show 54km/h.

For all the rest stated about facts, claims, talking bs and so on:

You re goddamn right peoples :slight_smile: But for the range discussion I disagree with you.

2 Likes
12

I completely disagree with this statementā€¦ If i drew less current with more packs in parallel, i have less losses from ohmic heatingā€¦ if I used thinner wires which results in higher resistances, again, more lossesā€¦

Different drive systems and speed controllers also plays a big role in the resultant efficiencyā€¦

Since you mentioned wheels, i assume youā€™re not just talking about power efficiency of the drive system, but overall workdone (in energy) by the vehicle given the amount of energy (batt power) put into the system.

If thatā€™s the case, riding style, speed, riderā€™s weight, terrain, weather, wind speed, air density and riderā€™s surface area all play a part in determining the systemā€™s efficiency.

So, no. All boards have VASTLY different efficiencies.

3 Likes
13

tenor (16)

Riding style & terrain are king though, doesnā€™t matter how much you crunch numbers. Riding style really matters. There is no such thing as an average consumption for range that is anything better than guestimation.

Iā€™m averaging 40-45Wh/mile, which is considered decently high. Iā€™ve seen others get 25-28Wh/Mile on similar boards. Thatā€™s a pretty huge spread.

3 Likes
14

Barring size/heat theyā€™re mechanically identical to hub motors. Curious why you would think they wouldnā€™t be applicable for comparison?

1 Like
15

Weā€˜re talking about range testing and comparing. So yes you re right, when i draw a lot of current from a hub it will be less efficent than a belt drive motor.
But what I had in my mind: several boards, riders with the same weight start together for a ride. Same riding style, same route and Normal cruising. The differences in range wonā€˜t be fundamental.
If you go to the extremes the slight differences in efficeny multiply and get more displayed.
I think riding style makes huge difference for range, never said anything against it. But for range test you have to assume an average rider. Therefore my experience is that 10-12Wh/km seems pretty accurate for PU wheels on flat area. Just my experience. If I ride hard in a hilly area it s a different story.

16

With losses comes heat, with heat comes more losses, heat inside a bed of urethane = more heat, its almost an exponential formula.

Comparing a closed solution to one spinning in free airā€¦yea its different.

1 Like
17

Sure, but Iā€™d say that still makes them an applicable comparison point as they are still the most mechanically similar. You can generally account for losses due to heat/resistance much easier than physical drag introduced via belt/gear/chain.

So we might be splittin hairs, but Iā€™d call them a totally valid comparison.

2 Likes
18

Put it this way: you do not have an overall efficiency. You have spots where you can tell under which load the system has which efficiency. The load includes all factors (rolling resistance, wind resistance, uphill; which is all influenced by speed, acceleration and weight). At high loads efficiency matters. Hub motors with small battery would be worse because of the heat which generates more resistance and so on like @linsus said.
But for the average efficiency spot at a moderate load the values are not differing that much from my experience.
When you compare loads it is all near together. The higher the load the more the gap widens.

1 Like
20

Since weā€™re talking about hubs and especially ā€œthe most powerfulā€ hubs out there, I canā€™t help but think of the Raptor 2 and wonder how many riders have hit 10Wh/km at any speed, under any riding conditions, besides full downhill runs. I think itā€™s probably between zero and none, and itā€™s probably a tall order to find anyone whoā€™s even done 20Wh/km on those hubs under any kind of ā€œaverageā€ riding conditions. Sure, those hubs have issues, but the point is that assuming efficiency numbers that low for any kind of ā€œaverageā€ rider under ā€œaverageā€ riding circumstances is not realistic, especially for hubs.

21

@hummieee and devin published real world testing figures on the other forum doing laps on a flat circuit. they did test runs hoping to maximize mechanical output, and test runs hoping to maximize electrical effeciency. iā€™ll try to dig it up all the juicy data is hidden behind the ā€œpost flagged inappropriateā€, the thread didnā€™t end well with devin pissing off everyone including jason but the data is still valid

4 Likes