Friday, February 28, 2014

Solar PV parts

Two parts for the solar electric system for the second bike have arrived- the panel and the solar controller. They are both an improvement over the first bike, although still not what I think would be optimum.  The bakfiets has 3 regular 12 volt panels in series to produce 36 volts, but for this bike I'm using one flexible 12 volt panel with a DC-DC voltage step up converter.

The panel is a Grape Photoflex 100 watt:
It is frameless and glassless, and in this picture is draped over a piece of firewood to show how it bends.  It has standard 12 volt PV panel output parameters, and I've checked the Open circuit voltage and Short circuit amps and found them both to be about 5% above the rated specs.

It has good efficiency monocrystalline cells, embedded in a 1/16 inch thick layer of encapsulant that is between a white plastic backsheet, and a clear PET plastic (similar to beverage bottles) topsheet, (in place of the glass top of a regular panel).  It has already started to scratch, so I'm not sure how efficient the cells will be in a few years, but the plastic itself should take many years to degrade.  One nice feature that Grape has done is to put the electron collection grid traces on the backside of the cells, leaving the face clear of shadowing to improve the cell efficiency.

The wiring junction box (thicker black section glued onto the center of panel) is the smallest I've seen on any panel, which is good for weight and size, but bad for wiring.  The wires will need to face perpendicular to the panel edge and then run down the bike rack, so there will be some adapting required, the box is filled with sealant so this may end up just being a support for sharp radius bends on the wires.

The panel weighs about 4 pounds.  In comparison, each of the bakfiet's 20 watt regular framed panels weighs about 5.5 pounds, and the plywood lid holding them together is another 5 pounds, for a total of 22 pounds (for 60 watts of power).  I expect to fit a frame to the new panel for attaching it to the bike rack, but it will still be a drastic improvement in weight and output.  The new panel will hang out the back one to two feet, but the width is just about perfect as a cover for the luggage bags on each side of the longtail.

The solar controller has an MPPT (Maximum Power Point Tracking) circuit that updates at 15 times a second, for panels that are on the move.  (My solar controller at home updates once every 5 minutes.) It also has an output circuit that changes the panel's 12 volts up to the battery's 48 volts (both numbers are nominal), as well as a charging profile (CC-CV) meant for lithium batteries.   The upper voltage set point is one volt low for my particular battery, but I don't think this will matter that much on the road.  I would have preferred pigtail leads or a crimped terminal plug rather than a screw terminal block (on the right side of the housing).  It's kind of old school, I used to fuss with them when I worked on 1960's Volkswagens.  They are slightly susceptible to moisture, but mostly once in a great while they loosen up due to thermal cycling and/or mechanical vibration.

The panel and the controller were $300 each, which is twice the price of the solar setup on the bakfiets.  It does have 166% more output than the bakfiets, is much lighter, more efficient, and should provide much more energy.  However I've been asking myself if it is worth it, and the best I can say is it depends.  I don't think you can compare solar versus plugged in based on the LCOE (levelized cost of energy produced) versus grid price, as this is more of a usage defined comparison.  If I was in an open terrain, doing longer distance riding, and plugging in was a challenge, it would definitely be good.  But for short trips around town with a lot of shadows from buildings and trees it would be only a minor help, and a bigger battery would be a better option.  From a historical viewpoint, my 22 year old home panels were $328 for 50 watts each.  But then again, currently 100 watt panels with standard frames are now selling for $200.  Assuming that the cost of the PET and glass frontsheets are similar, and the rest of the parts are the same, then a panel made without an aluminum frame should be less expensive.

There is one very significant milestone for me about this panel- I bought it from Home Depot.  I've been waiting 25 years to be able to go into a hardware store and buy solar equipment.

From the viewpoint of selling bikes with PV, these prices are high.  There are probably more than a few people that will be willing to spend this much for solar power, but not average riders.  The panel price will probably drop over time, but the solar controller might have a problem.  I still think running off the sun is a good idea, and have 3 more PV ideas I'd like to try out on the third bike which may change the controller issue for the better, but first to finish this bike.

Sunday, February 23, 2014

Multimodal long distance travel

If I did not have a car...

I know- a lot of people don't have cars.  A recent short essay about this is George Black's "The Road not Taken", in the December 3, 2013 issue of OnEarth:
"Carless, they're approaching the future by stepping back, at least for now, into some of the pleasures of the past."

Taxis and buses are a limited option where I live, it is best if I provide for some mobility, even though I know people have lived with limited access in these hills for more than 300 years.  My bakfiets is good for a 60 mile range, which will get me two towns over and back, and I've just ordered a 33% larger battery for the second bike. Better solar parts for the next bike should extend the range even further, so it might have a 90 to 100 mile range. Since the next goal is raising the average speed, some of the battery range will probably now go into climbing hills.  I'm also thinking of how to put some basic weather protection on the third bike.  So my local mobility needs can be met by a cargo bike.

However I do travel further.  To start, I've thought of two example distances for pleasure trips I'd like to be able to do, that would also be the distance of some more routine trips:
-Hartland to the BALE (Building A Local Economy) Fest, in South Royalton, at the VT Law School, about 35 miles one way
-Hartland to SolarFest, in Middletown Springs, about 65 miles one way

One obvious solution would be an enclosed body- the streamlining would raise my average speed, lower fuel use, have surfaces for solar cell mounting, and provide weather protection:
Aerorider, Delft, Netherlands

There have been many who have thought about this problem, starting in the early 1900's, a good summary was written by Peter Cox at the University of Chester, UK, for presentation at Energy and Innovation, Seventh International Conference on the History of Transport, Traffic, and Mobility, (T2M), Lucerne, Switzerland, November 2009:

1934 Nordisk Cyklefabrik A/S photo credit:
The modern RANS HammerTruck is designed along these lines.

Although I like the idea of an electric assist Sofacycle, (the ultimate couch potato), full bodywork is something I've been avoiding, similar to a mid drive motor setup.  Good design is the most function and redundancy with the least material, which is a beauty of bicycles, but I may have to change my view yet.

Then I thought about longer trips, for example- visiting friends in Connecticut.  There could be some help for this, by bicycling to the Amtrak station in Windsor, VT, riding the train, and then getting off in Windsor, CT and bicycling to their house, (125 miles one way).  A couple of weeks ago I wrote to Amtrak asking about bringing my bike on a train, and they replied:
"Regrettably, it is not possible to transport a bicycle on Amtrak services between Windsor, VT (WNM) and Windsor, CT. Neither the Vermonter train from Windsor, VT to Springfield, MA, nor the shuttle from Springfield to Windsor, CT have 'wheel on' bike spaces. Also, neither of these trains offer checked baggage or checked bike box service."

In direct contrast to Amtrak's statement, the Vermont Rail Action Network just stated in an article by Christopher Parker for Green Energy Times (issue 24, February 15, 2014):
"VRAN has been actively advocating for carry on bikes on Amtrak trains, working with the Vermont Bike/Ped Coalition and the State of Vermont.  We can now report that Amtrak has tested a car with bike racks, and details are being worked out for this service on the Vermonter and Ethan Allen."

Short of getting a Brompton folding bicycle and trying to carry it on Amtrak in a knapsack, or trying to fit a cargo bike on a Greyhound bus, does this mean the only carless option is riding a bike the whole way?  Ironically, this distance is beyond the range of most pure electric cars or motorcycles, but with a second battery on board is possible on an eCargoBike.

Comment Feb 28- It's been pointed out to me that I can always put a raft on the Connecticut river and float down...

I have also read reports of Velomobiles, without electric assist, being able to maintain high speeds (20 to 50 mph), and even passing racing bike pelotons.  Hmm, maybe a streamlined body is looking better:
ROAM- Roll Over America, Velomobiles traveling coast to coast, at the starting fountain in Portland, OR, August 2011, Larry Varney blog, photo credit Lonnie Morse

My friend Andrea has sent a presentation by Susan Zielinski about Multimodal research at the University of Michigan Ross School of Business.  Their SMART program has 4 sub projects, and one of them is partnering with colleagues in Beijing to look at Multimodal options:
I know that California's BART rail system already allows bikes on the train, and Portland has trams that have hooks on the left side for bikes, but this could go much longer distances.  Ford and Alcoa are supporting the U of Mich programs.  I always knew Detroit could do better than the Corvette.

photo credit: Audi

Saturday, February 15, 2014

Measuring the front fork trail (part 3/3)

I've been explaining about the front fork geometry in the last two posts, and doing some measurements to check for anything weird before cutting and welding on the second bike.  I explained how to interpret front fork geometry in the second post, and showed that there is a pretty wide range that can be ridden, the harder job is making it comfortable with neutral handling.  The first post was about measuring the head angle, and the hardest part of that is finding a level and flat spot on the floor for measurements.  The head angle turned out to be mid range to slightly leaned back, which should give slower handling.  But it's important to check trail too, which is a little more difficult to measure.

My first thought was just to put a string parallel to the front edge of the head tube, and mark where it hits the floor.  There are a couple of chances for making a large error with this though:
-Lining up the string is similar to winding sticks used in carpentry, it takes some practice to look at the sticks square on so that there is no parallax error.  Because the string would need to be out 3 inches to clear the fork, the chance for error would be worse.
-The string would be at the edge, not the center line of the steering shaft, and when it hits the floor the difference would need to be subtracted.  Since it hits at an angle, the chance of error increases.

I decided to use the center line from the start, and attach little right angle squares sticking out for lining up the string.   To find the center line, I took a piece of cereal box, and cut a slot in it that just fit over the head tube:

Then tape this onto a yardstick so that it sticks out square (perpendicular) to the side.  Mark the center of the slot on the masking tape on both sides of the yardstick (so that you can flip it over for left and right sides).  Then put pieces of tape at the top and bottom of the head tube on both left and right sides.  Slide the slot over these pieces of tape while holding the yardstick against the seat tube (this keeps everything lined up square), and use the slot center marks to mark the center line on all four pieces of tape.
The left end of the yardstick is resting on the seat tube, to keep everything lined up square.
Do this on both left and right sides to get the center line at the top and bottom of the head tube.

I decided that holding two little squares on the center line marks while lining up a string was too hard, so I made a jig that could be clamped on the frame.  First I cut a cereal box template that lined up with the center line marks, and ran back along the frame 10 inches.  It's symmetrical so that it can be flipped over for the other side, and I had to cut a notch in it for the shifter cable ferrule.  I traced this onto a piece of plywood and cut it out.  Then I cut a board, that was wide enough to stick out past the fork sides, to be as long as would fit above the top of the fork (the inside corner had to be angled to clear the steering bearings).  Glue them together, making sure they are square:
Cereal box template, finished jig, and square used while gluing the boards together.

The jig clamped on to the bike, lined up with the center marks on the head tube.

The board sticks out far enough to clear the forks, and it turned out so solid for taking measurements that I used a yardstick resting on it, instead of a string.  Here is the basic setup for measuring:
Note: This is on a level, flat floor, and the bike is propped standing straight up.
(The cereal box on the yardstick near the brakes is from marking the center lines earlier, and is not part of this step.)

I've used a screen door spring going back to the crank to hold the wheel straight, a bungee cord would also work.  You need to measure both sides because the wheel will almost always be resting turned a few degrees left or right, and taking the average of both sides removes this error.

Rest the yardstick on the center line board, let it slide down slowly to hit the floor, and mark the spot on masking tape stuck to the floor.  Hold up a level that lines up with the center of the axle, and mark where it hits the floor.  Measure between the two marks, repeat on other side, and calculate the average to get the trail.

One other thing you can measure while the yardstick is set up is the fork offset:
The bottom edge of the yardstick is on the fork stem centerline, and the distance from there to the center of the axle is 1 3/8", or 35 mm fork offset.  Again, use the average of both sides.

With a little care this method should get you pretty good results.  It can be improved by watching the details, such as adding a shim to keep things square if the head tube is not the same diameter as the seat/down/top tubes.  The head bearings need to be in good adjustment too.   On a bike without a round head tube you might have to use the bearings or handle bar stem instead.  It might also be possible to make the jig out of foam core artist's board and hot melt glue so that you don't have to do wood working, but clamping it to the frame will need extra TLC and shims to avoid squashing the foam.

Getting back to this bike's measurements:
-Stock, the crank center line is 11 3/4 inch off the ground, head angle is 68.5 degrees, and trail is 3 5/8 inch.
-With the pedals dropped to a crank centerline of 10 1/4", the head angle is 64.7, and trail is 4 7/16".

Dropping the pedals will put the bike on the slow side of the recommendations I've collected.  I could cut the head tube tube off the down tube and reweld it back on steeper, but I think I'll leave it as is.  It isn't as extreme as many choppers, and I actually desire slower handling.  When riding along after 25 miles, I don't want the bike to dart left in front of a car if I scratch my back, the emphasis is on going straight. The extra trail will also make it correct it's path a little more if it gets bumped sideways by road garbage when I'm tired.  The cargo bike will have a long wheelbase (which is slower), and is heavier (slow handling again), so I may be pushing my luck by adding slower steering, but this can be an experiment, it will be good to find out how much of a difference it makes.

Saturday, February 8, 2014

Front fork geometry 101 (part 2/3)

Before I write further about the second bike's front fork geometry, it seemed like there should be a front fork 101 to help put it into perspective.  Many times you will see a fork geometry diagram like this:
This is a very nice diagram, but I thought I'd add a little background to it to help with interpreting the new longtail's numbers.  Photo credit: Calfee Design

A caster is stable when pulled in the direction where the wheel trails the pivot axis, and wobbles around when pushed in the other direction, with constant steering corrections needed.  Also, a small amount of trail responds to direction changes quickly, and a large amount would pivot slower and take more effort to change direction.

However if the steering axis is tipped, the trail can be moved to the other side of the contact point, and the caster becomes stable while moving in the other direction.  Another factor is added though- gravity also tries to make the wheel flop over.

This fork has a 35 mm axle offset (a unicycle would have none).

With a vertical steering axis (90 degrees head angle with the ground), the steering is very quick.  If the front axle was not offset, the bike could literally turn on a dime.  This is great for stunts, but not pleasant at speed, and if an obstacle catches the wheel it can flip it sideways easily.

The other extreme is a horizontal steering axis (0 degrees head angle with the ground).  This is very stable at speed, and cannot be knocked to the side by an obstacle, however it takes a whole parking lot to make a turn. (The bike will turn sort of when the front tire leans over onto it's sidewall.)

At 0 degrees head angle, the tire naturally wants to flop to the side, which is 9" lower in this example.  To straighten the bike, the rider has to keep the bike lifted up 9" with the handlebars.

When the head angle is lowered and front forks are extended flop becomes a problem.  Above the pedals on this bike you can see a black spring device that is connected to the forks, to help keep the bike lifted up.  At more normal head angles flop is a much smaller effect than the forces from head angle and trail, but at speed it can make an instability problem worse.  Photo credit: Rob Lalonde, Ontario, CA, RHL Customs

Normal production head angles have ranged from mid 60's up to high 70's degrees.  On a motorcycle this angle is called rake, and on a car it is caster.  Besides this range of 15 degrees in production bikes, the head angle will also change a degree or two on a bike with suspension as the frame tilts when one end or the other compresses or extends.  There isn't any narrow band of head angles that is correct because other factors also affect handling, and common angles have changed as rider preferences change.

The head angle combines with both axle offset and tire size to produce trail.  (A note: sometimes the front axle offset is called rake by bicyclists, but to avoid error because of motorcycle terminology, I use offset.)  Road bikes commonly have offsets in the 40 to 55 mm range, with 26" tires using the smaller and 29er tires the larger end.   Although changing the offset will affect the front end height vertically and thus the head angle, at 70 degrees the change horizontally in wheelbase is 2.7 times larger (tan of 70 degrees).  Looking at the 70 degree head angle picture above you can see that increasing the axle offset would move the Point of Contact forward, which would reduce the trail.

Just like the caster at the beginning, smaller trail is quicker, larger trail is slower handling.  I've looked up some trail recommendations to fill in some actual numbers:
-50 to 63 mm (2" to 2.5") for road bikes (57 considered ideal for stability and agility) (Calfee Design)
-63 mm (2.5"), as more trail is sluggish and makes the bike wobble more when pedaling while out of the saddle.  Less trail is more precise ( Dave Moulton and 
-80 to 90 mm (3" to 3.5"), (Pink bike)
-touring bikes can have 4" or more trail
-56 mm (2 3/16") gives neutral handling with 700C wheels (Spectrum Cycles)
-69 degrees with 60 mm offset for 29er wheels, as 29ers tend to be sluggish at low speeds and a more standard offset of 45 to 50 mm would be good at high speed, but sluggish at low speed (Cannondale)
-3" stock, 2" to 4.5" range for regular motorcycles, 2" to 9" range for choppers (Chopper handbook)
-3" to 5" range (triker don)
-minimum 2", maximum 14" (North Dakota state motorcycle law)
-4" Harley Davidson FXR, 6" Harley Davidson FLH stock motorcycles

Bicycles (lower speed) seem to use a quicker handling 2" to 4" trail range, and motorcycles (higher speed) use a slightly slower 3" to 5" range.  These numbers should be taken as general guidelines, because many of the authors leave out a description of tire size (larger tires are slower handling, and they also scale the geometry up), and what they consider normal handling.  70 years ago racers preferred no trail and very squirrelly bikes (see Dave Moulton's blog ).  To prove that people can learn to ride almost anything, there are choppers with head angles at 45 degrees or lower:
California chopper with about 45 degree head angle to help clear airplane motor.  This probably has 8" of offset, but I can't estimate the trail because the head tube is hidden.  Photo credit Steampunk

Neutral handling bikes travel through a corner consistently, a different requirement than being stable in a straight line.  This is a bit beyond Front Fork 101, but I'd like to mention it.  Low trail bikes tend to resist being leaned into a corner or straightened up, and tend to climb out of a corner (straighten out) by themselves.  High trail bikes will lean better and feel solid at speed, but also tend to drop into a tighter curve than the rider intended.  This set up depends on several other factors, such as tire radius and profile, wheelbase, weight distribution, riding position (for sensory feedback), so I'll just give an example of how difficult it is to predict:
The best handling motorcycle I've ever ridden was a BMW K75S, (sport), during the late 1980's.  I didn't ride the bike, it read my mind.   On the other hand, I've ridden a BMW K75T, (touring version), which is the same bike, weight, geometry, and tires, but with a different set of front fork springs and riding position, and it needed correction during turns and leaning.  I spent a lot of time trying to adjust it to match the S.  The S would be a safer bike than the T.

I think I'll finish with this nostalgic news story.  I've put a ruler on the head tube and it looks like the trail is only a couple of feet off, I would guess that he is having trouble balancing because the fork is flexible and he can't accurately correct for a flopping problem, look at the twist between the front wheel, the handlebars, and his position over the rear wheel:
Photo credit: UPI news service