Around the world on a solar ebike

I find this interesting as I've found the opposite.
After I first put a solar panel on the front of my semi-recumbent I noticed a significant improvement in cruising speed (45-50kph) efficiency of about 50W+. I've increased the size of the panel over the years from 15W to 40W now and the efficiency gains only seem to increase with it.



Maybe it's related to the 'semi'-recumbent nature of my bike as well as the higher speeds I cruise at, but the irony is for me I get better gains from the efficiency gains than I do from the solar energy generated while riding. :lol:

Cheers

Cowardlyduck said:

I find this interesting as I've found the opposite.
After I first put a solar panel on the front of my semi-recumbent I noticed a significant improvement in cruising speed (45-50kph) efficiency of about 50W+. I've increased the size of the panel over the years from 15W to 40W now and the efficiency gains only seem to increase with it.

Maybe it's related to the 'semi'-recumbent nature of my bike as well as the higher speeds I cruise at, but the irony is for me I get better gains from the efficiency gains than I do from the solar energy generated while riding. :lol:

Click to expand...

That is interesting. I can't explain why you seem to be getting significant gains from your front fairing. Air resistance increases as velocity squared so it would be about double at your speeds compared to the 20 mph (32 kph) I have been testing at (45^2/32^2=2.0). Small reductions in drag or area would result in more noticeable reductions in power needed to maintain the same speed.

I spent a lot of time reading through posts in human powered racing forums where enthusiasts discuss the benefits of various home-made semi-faired and faired recumbents. The consensus seemed to be that tail fairings are far more effective than nose fairings with anecdotal evidence like "0.5 mph gain from nose fairing, 1.5 mph gain from tail fairing" being common. The explanation was that vortex shedding from the back of the vehicle causes a low-pressure region of turbulent air which "pulls" on the vehicle from behind whereas a long, smooth tail box results in more laminar flow, reducing this effect. Even the maker of the Windwrap nose fairing I used emphasized increased rider comfort in winter and downplayed expectations of speed gains. This all seemed consistent with my own experimental findings so I've focused on tail boxes and reducing the frontal area.

I even paid for a subscription license for a CFD (Computational Fluid Dynamics) software package from AutoDesk for a couple of months hoping I could model the effects of different tail box designs before building physical prototypes but I never mastered the software well enough to get meaningful results. From what I understand, the most effective design strategies are often not intuitively obvious. Just because something looks fast doesn't mean that it has low drag. Small, seemingly insignificant changes in draft angle or the gap between the rider's body and the fairing can make a surprisingly large difference.

I also found the testing procedure more difficult than I anticipated. It's hard to find a long stretch of flat road what is shielded from the wind. I tried doing multiple runs in each direction with and without the fairing(s) and found that the variance between individual runs was greater than the average difference between faired/unfaired runs. I could only hope that the wind was the same throughout the experiment. Even something simple like maintaining a constant speed with no acceleration/deceleration was tricky. I never got the Cycle Analyst dialed in so that it wouldn't oscillate about the target speed setting. Next time, I think I'll just set it to maintain constant power and compare the difference in speed.

solarEbike said:

Small, seemingly insignificant changes in draft angle or the gap between the rider's body and the fairing can make a surprisingly large difference.

Click to expand...

This is what I think makes the difference in my case.
I think that in my semi-recumbent position, my body catches a substantial amount of air basically turning me into a wind sock. With the panel upfront it forces a large amount of air to balloon around me and rejoin at my sides or just behind me. This effect is obviously greater at higher speeds.
I can even feel this when riding, particularly on very cold (-7C) mornings I can stick my head or hand our past a certain point and feel a stream of dense air streaming around me at speed.
So while the flat panel up front may not seem like it helps much just looking at it...under the right circumstances I think it greatly improves things in the right direction. All this is anecdotal I must admit, but I hope it helps inspire none-the-less.

Cheers

Here is a good article , I first saw pictures of a vetter full fairing scooter many years ago , Have been waiting for a commercial product to come out like it for some 20-30 years .
In this article the faired electric Zero S gets great efficiency despite such a rear tail that does not look better than one that comes together at the rear .

https://ridermagazine.com/2013/12/20/the-last-vetter-fairing/
and
a picture of the blue scooter bodywork .
https://www.pinterest.com/pin/705094885386226659/

https://en.wikipedia.org/wiki/Vetter_Streamliner

There was lots of research done from about 1975 to 2005 on the aerodynamics of recumbents, by university engineering departments, and independent developers. They found that, as cowardlyduck suggests, an upright seating position acts as a parachute. Laying the seatback back reduces that effect. Also, putting the windmilling cranks and legs directly in front of the torso does wonders to reduce drag. This is why a full front fairing is most effective on a LWB, low bottom bracket bike like the BikeE, but does little to improve performance on a bike like solarEbike rides.

Warren said:

There was lots of research done from about 1975 to 2005 on the aerodynamics of recumbents, by university engineering departments, and independent developers. They found that, as cowardlyduck suggests, an upright seating position acts as a parachute. Laying the seatback back reduces that effect. Also, putting the windmilling cranks and legs directly in front of the torso does wonders to reduce drag. This is why a full front fairing is most effective on a LWB, low bottom bracket bike like the BikeE, but does little to improve performance on a bike like solarEbike rides.

Click to expand...

That makes perfect sense and explains exactly why I see the difference I do where others with lower bikes do not. Thanks for explaining!

Cheers

Fabric tail fairing mock-up. What do we think? Is this worth pursuing?



[youtube]k7IBL_LM2Iw[/youtube]

solarEbike said:

Fabric tail fairing mock-up. What do we think? Is this worth pursuing?

Click to expand...

Over decades of riding, and lusting after higher speeds, I found that my average speed on a long day never got over 15 mph. At that power level, adding fairings didn't help. Maybe crossing Kansas into a head wind, but in the rolling hills of central Virginia, the added weight offset any gains. Life is not fair. Real athletes benefit from this stuff. Guys like me...not so much.

Now that I ride e-assist, I could benefit from fairings, but the speeds I could attain and the odd appearance would just make me a target for law enforcement.

If I could get it licensed as a LEV, as Cedric Lynch did with his famous bike in the UK, I would be tempted. But even then he got stopped, and hassled for going to the grocery store.

Warren said:

solarEbike said:

Fabric tail fairing mock-up. What do we think? Is this worth pursuing?

Click to expand...

Over decades of riding, and lusting after higher speeds, I found that my average speed on a long day never got over 15 mph. At that power level, adding fairings didn't help.

Click to expand...

Thanks, Warren. I keep coming back to this every couple of years and keep coming to the same conclusion. Each time, I think "maybe there's something I overlooked last time?" or "If I could just tweak the design a bit or improve the test methodology." Weird, I've never thought of myself as pathologically optimistic.

It really just comes down to the "speed squared" term in the equation. At 15-20 mph (24-32 kph), the theoretical gains are too small when combined with my "easy to fabricate, light weight and removable" design constraints.

A gave it another shot today and I think I'm good for another couple of years. This was my best execution yet. Unfortunately, I didn't get any good photos before removing the fairing for the "unfaired" portion of the test but it was a masterpiece of spandex, double-sided tape and safety pins. I even removed the roof and shaved my legs wore form-fitting clothing to give it the best chance of success.

Instead of trying to maintain the same speed and measuring the power required to do so (like I did last time), I ran the motor at a constant 300 watts to simulate average power input from both the motor and pedaling on a long day. I spun the cranks with my feet to maintain any windmill effects but I wasn't adding any power to the wheels.

While a wind tunnel or indoor track might be ideal, I don't have access to either so I found a 1.1 mile (1.8 km) stretch of road near the Port of Oakland with a wide turn-around at the end and did several out and back runs both with and without the fairing. A typical run looked like this.



I captured the data at 1Hz resolution using a Grin GPS Analogger but used wheel speed instead of GPS speed for better accuracy. I trimmed the acceleration/deceleration at the start/end using the Grin Trip Analyzer. I was also able to separate out the one-way figures. Here are the results from all the runs.



Apparent changes in wind speed and direction really muddle the results but the first faired run (log file 669) and the last unfaired run (log file 683) have very similar speeds in both directions with only 0.2 mph speed difference between faired/unfaired overall speeds and only 2.5% difference in energy consumption.

If you guys can glean any more insights from these numbers, please let me know but I don't think it's worth the weight and hassle. Too bad. The red color is fantastic for improved visibility on the road and the black fabric is stretchy yet completely waterproof... would have been nice to keep all my stuff dry.

I decided my wind gust problem was mostly hardware related. The adjustable driveshaft clutch I'm using to protect the tiny gearbox of the motorized tilt mechanism was not strong enough to hold the panel in place against a sudden wind gust. Even when the panel was level, a slight updraft could suddenly flip the panel, which was dangerous at any speed.

A stronger clutch wouldn't fit inside the trailer tube so I explored a couple of externally-mounted solutions. The first is a solenoid-actuated locking mechanism. Effectively, a secondary clutch that would fit around the outside of the tube. The Arduino tilt motor controller would active the solenoid before each motor movement to pull apart the Hirth joint and re-engage it afterward. The joint is spring-loaded so it remains locked when the solenoid is de-energized. I even got as far as ordering the parts for this one but didn't love the idea because I really wanted to keep the entire mechanism inside the tube for better dust/moisture protection and for better aesthetics. The second scheme below is a larger external motor with external gears but suffers from the same limitations.



That's when I hit upon this bit of cleverness: I added a second planetary gearbox ("Gearbox 2" below) on the other side of the clutch. An external force on the panel (wind gust, crash, manual tilt adjustment by hand) now goes through a 3.7:1 gearing change so the max torque of the clutch is effectively multiplied by 3.7. The tilt speed would have been reduced by 3.7 so I changed the Gearbox 1/motor combo from a 5.2 RPM model (670:1 reduction) to a 9 RPM model (387:1 reduction). The maximum panel tilt speed is reduced but I have more torque to move the panel against a stiff breeze and the slower speed actually helps with small positioning adjustments. I'm driving the 12V brushed motor with a PWM controller but if I go slower than about 1/2 speed it doesn't have enough torque to move the panel and the motor will stall.



This new version has been rock solid for several hundred miles now. I was gifted the opportunity to crash test it on a recent overnight trip. This was a short, steep, loose grade. I had the wrong tires and wrong tire pressure for something like this (the trip was 90% paved roads and paths) and 2 wheel drive wasn't enough to overcome the challenge. I managed to come to a safe stop and dismount but dropped the bike when I tried to push it up the hill while applying throttle and the front wheel spun out. There's just not enough weight on the drive wheel in this situation.

I wonder if I set the Cycle Analyst to limit speed to something like 5 mph would that prevent the spin-out? Speed signal is via one of the motor hall sensors so I may not have enough rotational resolution to detect the slip?



Here's a close-up of the custom mounting brackets for the panels. There are two low-friction UHMW plastic brackets in front which are designed to slip where the panels connect to the stationary trailer tube and two solid metal brackets which clamp to the rotating tube section in the rear. I've found that polished 6061 aluminium holds up surprisingly well to corrosion. I have parts that still look brand new after 2 years of riding near salty sea air though they rarely get wet as it doesn't rain much here and the bike has been garaged at night. There's no clear coat and it scratches very easily. I'll experiment with anodizing in the future.



The timing of the current California wildfires was an inconvenience but I had a non-refundable reservation. I wore my 3M 6503QL respirator with P99 filters to deal with the smoke. The rubber pressed against my face all day isn't what I'd call comfortable but it's easier to breathe through than any cloth mask I've tried and is more effective at blocking particulates than N95 masks. I could still smell the smoke when I got to more heavily polluted areas.



The whole rig worked like a champ. I had it fully loaded with extra gear and water to approximate the weight I will need to carry on remote stretches. Total gross vehicle weight: 340 lbs (155 kg) including food, water, gear, rider, bike and trailer. I had to max out the rear shock at 275 psi to get a 35% preload but the bike handled it with ease. I'm very pleased.

Two-day trip used 440 Wh from the pre-charged battery and 2810 Wh from solar. I had energy to spare, covering 89 miles (143 km) with 800 meters of climbing per day and felt like I could have gone 50% farther.


Fully interactive log file.

Wow, this is amazing. Nice work.
That last photo reminds me of our bush fires here in Australia over the summer. Was scary stuff!

Cheers

Cowardlyduck said:

That last photo reminds me of our bush fires here in Australia over the summer. Was scary stuff!

Click to expand...

It was surreal. At New Brighton beach, there was some light ash falling and water-dropping helicopters flying directly overhead. The total area burned so far is only about 1% of the Australia bush fires last summer but it will take weeks to get all the fires under control. The thunderstorms that started them were uncommon and intense.

I took a trip down the coast to visit relatives and put the whole rig, gear and rider to the test. I figured I would post here when I got back but I'm not finding the motivation for a detailed write-up. Here's a video I put together. How about an open question and answer session? What do you guys want to know?

[youtube]UqFfbAtS6Vg[/youtube]

solarEbike said:

What do you guys want to know?

Click to expand...

Just one question. When you gonna get out of that fire infested hellhole and get on the road? It's been 2.5 years of analysis, so time to come out of the analysis paralysis phase and do it. I've been waiting on the message "John I should be hitting CR on about X date, how to we meet up?"

John in CR said:

Just one question. When you gonna get out of that fire infested hellhole and get on the road? It's been 2.5 years of analysis, so time to come out of the analysis paralysis phase and do it. I've been waiting on the message "John I should be hitting CR on about X date, how to we meet up?"

Click to expand...

I genuinely laughed out loud. Thanks for that.

I cringe every time I see the title of this thread. Some of the delay has been me grossly underestimating the amount of effort it would take to build the rig I had envisioned. Some of it had to do with changes to my career situation. Finally, when I was "this close" to heading out toward Alaska in April of this year, the pandemic hit. I've had a whole list of travel vaccinations based on my 50 country itinerary but the one that I need most is not yet available.

At this point, I'm not convinced the Canadian border is going to open by spring 2021 in time for the short time window my route has for making it to Alaska during the few weeks when it's thawed out. They've been postponing the re-opening for months. I've been following a few travelers on InstaGram and the few people who are doing any travel now are running into significant restrictions and inscrutable new rules. The idea of visiting 50 countries seems preposterous at the moment.

I'm really looking forward to sending that email but it's going to be a while.

Since Alaska isn't another country, just call that a practice run. Then you and your bike can take a boat and bypass Canada, so the virus doesn't impede some real progress. Take a month or 2 to see Alaska by bike in late May or early June and get that advantage of very long days with your solar panels. Seems like a good virus compromise to me, and better than waiting until 2022.

How many solar watt-hours did you average a day?

Did you charge at night? Was AC power easy to find?

Any issues with wind acting on either panel?

John in CR said:

Since Alaska isn't another country, just call that a practice run. Then you and your bike can take a boat and bypass Canada, so the virus doesn't impede some real progress. Take a month or 2 to see Alaska by bike in late May or early June and get that advantage of very long days with your solar panels. Seems like a good virus compromise to me, and better than waiting until 2022.

Click to expand...

Bypassing Canada would be preferable to delaying by another year. Ideally, I wanted to do continuous overland travel from end to end on each continent but that's looking much less feasible now than it did a couple of years ago.

It sounds like maybe you were thinking that I was planning to start in Alaska? The route I've worked out starts and ends in California heading west, timed to avoid winter for 2.5 years by crossing the equator roughly every six months so that I spend summer in the northern hemisphere followed by summer in the southern hemisphere, etc. The westward heading is kind of a leftover from early efforts to incorporate ocean crossings by sail and later abandoned when I found that popular sailing routes didn't align at all with my goal of riding continents from end to end. Attempts to consider predominant wind directions over land for the whole route (to minimize headwinds) proved inconclusive.

So, I will likely cover the entire length of the traditional Prudhoe Bay, Alaska to Ushuaia, Argentina itinerary but I'll do California to Alaska at the start and South America to California at the end. Apparently, this is allowed under my self-imposed rules.

I never worked out an eastward bound version of the route. Maybe I should look into that now? Starting mid-summer from California to be in Argentina in January might be a back-up plan for a closed Canadian border come spring 2021 but would mean traveling through many small countries earlier in the trip so less time for the pandemic to peter out...

JackFlorey said:

How many solar watt-hours did you average a day?

Click to expand...

Here are the trip stats from the end of the video. 20,104 Wh and 12 days looks like 1675 solar Wh/day but that includes solar Wh generated during my non-travel days. If I add up just the travel days, I averaged 1472 solar Wh/day with a low of 1079 Wh and high of 2118 Wh.



Keep in mind this, was off-peak for summer travel as the last day of the trip was just before the autumn equinox. I generated 2266 solar Wh in one day during the 207 mile loop I did on June 30 (8 days after the summer solstice). This is where these days fall on the modeled production estimates using weather data from a Typical Meteorological Year. The 2266 Wh value is well above the average because I was able to hand-pick a perfect day with no morning fog and no clouds. The longer trip had no such luxury so it more closely correlates with average values for that time of year. This chart also conveys how much variability there is from day to day.

JackFlorey said:

Did you charge at night? Was AC power easy to find?

Click to expand...

Except for 450 Wh in my aging LiGo battery pack at the start of day 1, I chose not to do any AC charging. Most days, I was riding from sunrise to sunset so there wouldn't have been time to wait for a mid-day charge. Most nights, I was camping with no AC outlets nearby. I have done some 2-3 day trips in the past with a much smaller solar array and found that looking for outlets along the way and relying on outlets for an overnight charge was time consuming, stressful, and became the central focus of the trip.

100% solar charging eliminated that problem but I found that trying to estimate how much distance I could cover in the next day or two could be challenging when the "clear/cloudy" weather forecasts didn't match reality and I had camping reservations at a National Park, for example. The solution to that would be to plan to average 80 miles (130 km) per day at this time of year with this rig.

JackFlorey said:

Any issues with wind acting on either panel?

Click to expand...

The most extreme 80° left/right tilt is only useful at sunrise and sunset when the sun is directly to my left/right. I tend to show this trailer position in videos because it looks dramatic and shows off the tilt range. The full tilt is most useful when I have to squeeze the bike through a tight spot or when charging while parked at sunrise or sunset but is rarely needed while riding.

The optimal trailer tilt angle is usually closer to horizontal. Light breezes are not a problem under these conditions. When wind speeds increase, I use a handlebar-mounted control to limit the maximum tilt angle until the effect is minimized. In strong winds, I simply stop the panel in a horizontal position and keep riding without problems.

On this trip, I ran into crosswinds which made riding very difficult for about 2 miles while crossing Salinas Valley next to a berm which funneled the wind across the road in a dangerous way. Admittedly, it wasn’t safe or fun. I slowed down and got through it. The remaining 99.8% of the time, wind was a non-issue.

Here's a teaser from a project that's currently in my video editing queue.

Great work! Did you encounter any unexpected difficulties, such as flat tires? What parts of the trip did you like best? Would you rather be on a road that's straight or curvy?

Also in the video I sometimes saw you cranking pretty fast in low gear going uphill, when I would have expected more motor and higher speeds instead. Some of those were probably because you had to go back for your camera anyway, but is that just to make sure you don't run the battery too low? Sorry if these questions are silly :lol:

thundercamel said:

Did you encounter any unexpected difficulties, such as flat tires?

Click to expand...

Just one flat tire on day 1, thankfully. No serious mechanical issues. My trailer's suspension "collapsed" on the way back but fortunately it was just a bolt that worked it's way loose from all the vibrations and I caught it before I lost any hardware so I was able to tighten everything up and get back on the road in a few minutes.

thundercamel said:

What parts of the trip did you like best? Would you rather be on a road that's straight or curvy?

Click to expand...

The rolling hills with no signs of human development around Vandenberg Air Force Base, north of Santa Barbara, and around Pinnacles National Park were sublime. I like any road that allows me to maintain a constant pace for a long while without constant stops and starts for traffic lights and such. In fact, I preferred the brief sections of freeway riding (Hwy 1, 101 and the I-5) on this trip over the beachside bike trails which were often choked with joggers, dogs and cruiser bikes oblivious to their surroundings.

thundercamel said:

Also in the video I sometimes saw you cranking pretty fast in low gear going uphill, when I would have expected more motor and higher speeds instead. Some of those were probably because you had to go back for your camera anyway, but is that just to make sure you don't run the battery too low?

Click to expand...

I naturally prefer to maintain a higher pedaling cadence of 85-95 RPM and from what I've read that is a good target range for anyone who wants to minimize the likelihood of knee pain on long rides. If you mean the clip where I ride past the Malibu sign, that was a particularly tough morning because my battery was completely empty and I was struggling up the 5-6% grades with a heavily loaded bike with almost no electric assist. I got a very early start that day and the cloud cover didn't lift until late morning. I had 111 miles to do that day because I had pre-paid for a motel room because camping options were unavailable and I was on my way to meet up with a fellow solar bike enthusiast in Santa Monica that morning.

thundercamel said:

Sorry if these questions are silly :lol:

Click to expand...

Not silly at all. Thanks for asking.

Having just started my solar bike project, I'm grateful to have found this discussion. Since l have regen breaking, my concern is that the pv current coming from my mppt controller would somehow throw off my system, since there would be 2 simultaneous sources of charge during regen breaking. I'd be trying this with the stock battery and electronics that came with my RadPower ebike. My other option would be to use the solar panel exclusively to charge my backup battery. Having read this entire discussion, it looks like people don't run into trouble with this kind of thing. Is that correct? Thanks for your help with a newcomer and beginner!

Paulfin said:

Since l have regen breaking, my concern is that the pv current coming from my mppt controller would somehow throw off my system,

Click to expand...

Your regen brakes will charge directly to the battery without interacting with the solar charger controller. There will be no problem. All the solar system sees is an increase in battery voltage.

Chalo said:

Your regen brakes will charge directly to the battery without interacting with the solar charger controller. There will be no problem. All the solar system sees is an increase in battery voltage.

Click to expand...

Exactly right. There's no meaningful conflict. When my battery charge is at 90% or higher, regen braking may briefly drive the voltage up to 4.2V/cell so solar charging is briefly interrupted but it immediately resumes when braking stops and the battery voltage drops.

Paulfin said:

I'd be trying this with the stock battery and electronics that came with my RadPower ebike. My other option would be to use the solar panel exclusively to charge my backup battery.

Click to expand...

Can you connect both batteries in parallel and charge/discharge them as one big pack? This is how I've always done it. You'll prolong the life of both packs because the current per cell will be lower both for charging and discharging. With a bigger battery, you can set a higher regen current so you recapture more energy/prolong brake pad life... assuming this is something you can adjust on your bike.

Also, you never need to think about which pack should be on the solar charger. Having to constantly swap them sounds like a pain in the ass. Imagine if you start the day with both packs fully charged. If they are separated and only one is connected to the solar charge controller then your solar panel is just dead weight until you ride a few miles, stop and swap packs. You'll need to keep doing this throughout the day to ensure you're capturing every available solar watt-hour. With both packs connected in parallel, solar charing begins as soon as you start riding because battery voltage sags under load. No swapping necessary.

You also extend battery life and increase system efficiency if the solar charge controller's output can be directly connected to the motor controller. If I'm getting 200 watts from solar and the motor controller is drawing 200 watts, I avoid the round trip losses (around 10%) incurred by putting those solar watt-hours into the battery and I reduce the battery's cycle count. This is easier with my Grin LiGo batteries because the charge and discharge leads are the same. It's a little trickier with ebike batteries with separate charge/discharge ports.

Some proprietary systems like Bosch make it difficult to implement solar charging so separate batteries may be easier to implement there but it's less efficient and less convenient.

Welcome to the forum.