Saturday, August 23, 2014

Longtail construction (2/3) Rear rack

In the first post about building the Longtail frame, I covered extending the chain stays and putting in a platform for the battery box to sit on.  The next step was adding on the rear rack and then tying the top and bottom horizontal tubes together to form a truss structure.
After installing the wheels, the frame was propped up along a line drawn on a flat and level floor.  (You can do this work at any angle if you are willing to do a lot of measuring and calculating offsets, but it's much easier and more foolproof to just hold up a level.)  I used an oxy acetylene torch to heat the tubing while bending human friendly curved ends on the rack, as the radius was too tight for any cold bending forms I had available.  It was then rested on top of the battery box with 3/8" thick spacers to create space for a tie down rope to fit through, leveled, propped in place at the rear, and then notched to fit to the seat tube.  There are wedges centering the front of the tire in the chain stays so that the tire tracks straight, and the boards clamped on each side of the tire are the right width to hold the rear of the rack centered.  At this point the old seat stays are not connected to anything at the top.

The sequence for connecting the chain stays to the rack was based on working from the fixed point at the seat tube towards the tail end, and fixing any errors from welding heat warping as I went along:
1.  Tack weld the front of the rack to the seat tube to stop things from moving around.
2.  Cut and tack on uprights that just clear the back of the battery box.
3.  Readjust the rack to be square and level, and add diagonals from the drop outs to support the back of the rack.  The upper ends are notched as usual with a grinder to fit, but the lower ends were slit for an inch and squeezed in to make a taper on the tubing, and then hammered square to fit over the existing luggage rack mounts on the drop outs.
4.  At this point the assembly is starting to be self supporting, and cutting tools can now be used on the top of the old seat stays to fit them to a new cross stay inside the rack tubes.
5.  A diagonal is then run from the cross stay to the main chain stay junction.

All of the fitting work had to be done inside on the flat floor, which is OK for one project, but if I were going to build several bikes I'd make a table or jig in a better space or outside.  Each diagonal was briefly tack welded in place with many plywood shields around it to contain sparks, and when enough triangles had been formed to stabilize the structure it was moved outside for more complete welding.  Tack welding an assembly first also helps control warping from welding.

The cross stay that connects the top of the old seat stays to the rack tubes.
It also links to the diagonals running down to the chain stay extension junction,
and holds the rack tubes at the right width.

After making several mock ups, I decided that the board on top of the rack should be about 5" to 6" wide.  If it was wider the corners cut into the insides of a passenger's thighs, and if it was narrower it wasn't as useful for supporting a load.  Since standard 1"x 6" lumber is 5.5" wide, I used that width so it could be an option even though I planned on using thin plywood, as 5.5" would also allow strips to be cut from a sheet of plywood with a margin for a kerf but not too much waste.  The rack tubes are 5 5/8" wide, so that the tubes stick out 1/16" further than the plywood on each side to act as a rounded bumper.

With diagonals welded in to form triangles, the rear section of the frame was pretty stable, and I went back to finish up the front of the rack.  (Although some parts look yellow in this photo, there was no brazing, only welding.)  I added two strips for attaching the top of the battery box, and finished the seat tube welds.  Unfortunately the rack landed on the original seat tube exactly at the top, so I removed the old seat clamp lugs, and then welded a few inches from the top of another tube on top.  

I try to work as cold as I can to avoid changing the properties of the steel, while still achieving a good weld penetration.   But even though I also used a piece of aluminum tubing (it doesn't stick to a weld) inside the seat post to align the top extension, all the welding in this area deformed the seat tube enough that a seat post would not go in.  I don't have the proper size reamer to fit this bore, so I taped a 1" wide strip of coarse grit belt sander cloth on a 3/8" rod, wound it around to build up the width until it just fit inside the tube, and spun it with a drill for about 20 minutes inside the problem section to open it back up.

Looking at the rack attached to the lower chain stay assembly from the rear.

I prefer priming the work after I've finished a section, mainly because it splits hours of sanding and cleaning for paint preparation into smaller sessions, but also because it immediately protects the newly worked metal from oils and salt from my hands and dirt, so there is less prep needed.  With this method the primer needs only a light cleaning with a degreaser, a scuff sanding, and then a quick tack wipe when I'm ready to paint.  I've always preferred a self etching primer for automotive work as it adheres a bit better, but I found out on the Bakfiets that the chemicals in it do not work well with an acrylic latex top coat.  The water based paint reacted through the primer and rust spots bled through.  For this Longtail I used a regular plain lacquer style primer and had no problems.

A side view of the rack assembly.

The weak spot in this design is the hole for the battery box, as it has only a short height diagonal brace under it that lands in the center of a seat tube section.  However overall this design is overkill, as the loads in a truss are carried mainly in the top and bottom chords, which in this frame are well braced straight tubing spaced 17" tall.  To give you a rough idea of the strength, compare this to a 10" or 12" high floor joist in your house- the load carrying capacity of a beam increases proportionally to the square of the height.  Without considering the effect of width, a 17" high beam is 2.9 times stronger than a 10" high beam.  A stiffer frame will also give better handling (more precise steering).

Wednesday, August 6, 2014

The Strolling of the Heifers

The Strolling of the Heifers was founded in 2001 to benefit local food, farmers, and healthy agriculture.  It has grown over the years, and now has a Slow Living Summit to go with it.  Brattleboro cargo bike advocate Dave Cohen put out a call for cargo bikes for the parade, and I took the Bakfiets down to it.  A full range of bikes showed up, which included panniers, racks, trailers, trikes, longtails and bakfiets.  I tried to take some pictures while riding, but there is a more complete video of the parade on their website (the bikes are at minute 41). My overwhelming impression was amazement that so many people wanted to watch cows walk up main street.  Bikes are much faster.  But in keeping with the spirit of things, we rode at a heifer's pace, (which I think is just a little faster than the pace set by the preschoolers in my town's 4th of July parade).
I parked on the south side of town and rode in to the parade.  On the way I had to stop at the offices of Building Green, the energy efficiency building journal, next to the Estey organ museum, for a Green Photo.  The row of old mill buildings along the street all have slate siding.

This was our staging area, the hill in the background is across the river in NH.
(Click on pictures to enlarge.)

Among the bikes were Kathleen's trailer/canoe,

a local newspaper bike delivery service,

the frame of a new vendors cart bike being built by Hannah,

 and a nice cargo bike with a front box built by Jesse of Seneca Bicycle Works.

This is Todd's custom ebike with a 2 speed front hub, pulling a trailer.

The drive system of an Outrider trike in the parade.  Note there are only 3 large wires coming out of the motor, so there is no rotor position feedback.  This makes the motor harder to control and not as efficient, I'm in the middle of planning the motor and gearing for the third bike right now, and plan to cover this more in a future post.

Peter explaining his motor upgrade in his very dairy (decorated with cow spots) ELF.

Two Yuba el Mundos with a Stoke Monkey and a Bionx, and a passenger who is ready to ride.

Heading up Main Street.

Looking back down Main Street, (with 1/4 mile of people already).

Rolling UnThunder.  Quiet saves lives?

Stephan and Eliah in their Quadracycle.  Each rider was connected to one rear wheel,
with no link connecting the two sides.  Photo courtesy of Dave Cohen

My Bakfiets. The person riding the bakfiets next to me is Dave Cohen.
His passenger (sitting backwards) is reading the book "Life is a Wheel".
Photo courtesy of Dave Cohen.

About 2/3's way along the route.

Sometimes you don't want Dad's help, you just want to peddle yourself.
She actually rode the whole parade!

My word, the people!  It's heartening to think that most of them are in favor
of local food and agriculture, and so were probably very interested in biking too.

Our line up of bikes during the afternoon after the parade at the Bike Petting Zoo.  We answered all sorts of questions about our bikes.  I took a picture for a mountain biker from Virginia while he sat on the Bakfiets, and I had a great conversation with a couple from New Jersey about urban versus rural biking.

Yes, the LED flag staff lights up!

Friday, July 25, 2014

Bicycle bodywork (4 of 4), bike fairings

In Bicycle bodywork part 3, I ended with the 2014 Porsche 919 LeMans Prototype Hybrid Electric:

Drawing credit: Porsche, from article in Excellence, May 2014
It turns out this design has over 2000 hours in a wind tunnel tweaking the airflow.  Aside from steeper leading surfaces and more gradually sloped trailing surfaces, it is not the 1930's classical smooth teardrop shape, it is discrete sections where the airflow is shaped piece by piece for different requirements along the car.  Because of their odd shapes, I think the same methodology can also be applied to cargo bikes.
Photo credit: Porsche, from article in Excellence, May 2014
From the front it is easy to see the fenders are just big enough to cover the tires (no turning the tires sharply!), the cockpit is just enough for the driver, and instead of fairing in a smooth bodywork between the two, the body has been dropped down to just cover the suspension links.  The emphasis is on minimizing the frontal area, and smooth, continuous, streamlined shapes are secondary.  Streamlining works best in laminar air, not so well in traffic or off axis gusts, while minimal area works all the time.
Distance record bike Overzealous with rider Jeff Nielsen, Photo credit: Trisled
Note minimal frontal area.

It is relatively easy to cover smaller bicycles such as Velomobiles and human speed record vehicles in smooth, full bodywork, and it is highly effective at their traveling speeds of 20 to 80 mph.  Cargo bikes are much harder to fair in, but they ignore aerodynamics at their own loss.  If you only ride from 4 mph (walking) to 10 mph (medium run), then you can discount air resistance and go for fat tires to reduce rolling loss instead, but air starts to make a difference when you go above 12 mph.  Since my Bakfiet's average speed was 14.9 mph last summer and the Longtail's about 17-19 mph so far, with maximum speeds on both around 40 mph, aerodynamics is important.  I'd also like to develop a layout that will use the solar panels to provide some basic weather protection.  Taller bodywork, (such as Cedric Lynch's bike and the Velerique shown in my first Bicycle bodywork post), do reduce air turbulence loss while allowing a rider to sit more upright than a recumbent and carry a small load, but 2x4's, surfboards, child seats, etc. will not fit, and it is much more susceptible to cross wind effects.  One solution to load capacity and crosswinds is to make the bike into a Velerique with three wheels:

Twike, original body with human power + electric on left, and newer TW4XP all electric at right.
Photo credit: Foro Coches Electricos
Note that the human powered Twike has motorcycle tires, and the TW4XP has car tires- load capacity and rolling resistance will be roughly proportional to tire size (working air pressure, tread design and compound, and carcase construction also come into play).  The base 10 Ah x 384 Volt battery pack in the original Twike is good for 75 km, or 82 Wh per mile.    Using the Wh/m consumption figures to indicate relative drag/rolling resistance, this is 6.2 times as much energy use as my solar electric Bakfiets at 13.2 Wh/m, which I have found is okay to pedal on level flats, but I prefer to use the electric drive going up hills.  The new Longtail seems to be using about 16-18 Wh/m average, and I can pedal it along a flat up to about 1/3 mile, and then the drag starts to be a little bit tiring and I use the motor.  You probably won't pedal the Twike very far without turning on the motor, (see The Hyper-Efficient TWIKE Human-Electric Hybrid Vehicle, Treehugger for a review), and this means it's range is the battery range.  I haven't found performance figures for the TW4XP, but it was designed for the Progressive Insurance Automotive X Prize which had a contest minimum requirement of 100 mpg, or maximum energy use of 330 Watt hours per mile, and it does not have pedals. My estimate is 2 to 2.5 times the energy use of the original Twike, or 160 to 200 Wh/m.  These levels of energy use will not work well with mobile solar, or for people pedaling with limited physical capacity.

Scaling back to two wheels, the Technische Universität München did a  Mobilität neu erleben mit dem Reisefahrrad Läufer (New mobility experience with a rotor pedal bicycle) project:
Photo credit:  Unternehmer TUM GmbH
Fabric side skirts can be fitted for water protection.

Treehugger also photoshopped the front clip of a Honda Civic onto a Giant Revive Spirit electric assist bicycle:
Photo credit: Treehugger, 2005
Nice fit of the hood over the front wheel, but headroom might be a little tight.

Another version of a streamlined body on a bicycle.
Also known as "A woman needs a man like a fish needs a bicycle." (Irina Dunn)
Photo credit: The MPG of a Human, Tech News Gigaom.com 

Fish position, Photo credit: Unknown

Since it is so hard to package a practical fairing onto a cargo bike, let's look at riding position. My upright Bakfiets is almost straight up and down, like an Oma (Dutch for Grandmother) bike.  This is very comfortable and has great vision, but is not good for aerodynamics.  Leaning forward for racing reduces frontal area, and adds the use of gravity to a good leg angle for pushing, but it is hard on the neck, arms, wrists, and back of riders who may have arthritis or injuries, and the seat height isn't good for stop signs.  Moving the seat backwards and down also reduces frontal area, lowers center of gravity, and is a comfortable natural chair posture that allows both feet to be flat on the ground at stops, with the disadvantage of not being as effective for power production up hills.  If the bike is designed with an electric assist in mind, power becomes less of an issue.
Cruiser style, Malmo, Sweden.  Photo credit: Olaf Lundgren

I've had over a dozen people tell me they could not test ride the Bakfiets because they were not able to sit on a bicycle seat.  The narrow bicycle seat necessary to fit between the thighs for a forward leaning bike leg angle would also be uncomfortable for most riders over longer distances.  However as the leg angle from crank to seat is leaned backwards the rear part of the seat can be made wider for better hip support without interfering with leg movement, with the extreme example being the sling seats used on recumbents.  I also know from motorcycle experience years ago that there were many shorter people, mostly women, who preferred the lower Honda Rebel 250/450 motorcycle seat height over a regular motorcycle seat.  RANS calls their bicycle version of this Crank Forward, other bike makers have named them Feet Forward, and there is the historical term SofaCycle.
RANS Hammertruck, Photo credit: RANS
Notice the wider rear portion of the seat, and also the
stub seat post so that a standard front shifting mechanism can be used.
If laid out well it is still possible to stand up on the pedals for hills.

In the last four bikes with the seats moved backwards, you can see a space opens up under the seat where some narrow profile electric mid drive components could fit.  It should also be possible to put a bit of the third bike's mid cargo box under the seat, so that the length could be shortened.  However if the cargo box is large it doesn't package as tightly:
Peter Gibbs, The Jobbing Gardener, with his 8 Freight cargo bike
from race bike designer Mike Burrows.  Photo credit:  8 Freight
Notice the steep head tube angle on this bike- is it for speeding up the
response of a long bike?  (Added 23 August, 2014- Now that I've ridden
the longtail with some loads, my guess is that Mike is trying to reduce
self steering effects (particularly at low speeds) by reducing trail, but the
front fork layout is constrained.)  Another part of this design is that both
the front and rear wheels attach on only one side.

Adelaide "Ho Chi Minh" Longbike, by Ian Grayson and Bruce Steer, 1988
Photo credit unknown.  The length of this bike is similar to an Xtracycle,
(one rear chainstay, or about 15", longer), and the whole fits together very nicely.
However both the 8 Freight and this bike use standard upright leg angles,
and have either a pronounced forward lean position or full air resistance.

There are two levels of fairings that I am thinking about. The first is bolt on, and I'm starting with a windshield on the Longtail, and then getting a little more sophisticated with a better windshield and lowered seat on the mid drive third bike.  The second level is more comprehensive, fitting solar panels as permanent bodywork, or "solar panel over my head" on the fourth bike, and will offer more weather protection.  Unfortunately this will probably require more of a recumbent leg angle from crank to seat, and I'll test some of the layout on the third bike.  Some examples of bolt on devices are:

Front fairing for a Longtail, by Zzipper.  Photo credit: Thomas David Kehoe
This type of fairing should be positioned about belly height when you are in a semi
upright position, so that it diverts air away from the pocket formed by your body.
It should be as close as possible to your body, while still allowing you to 
lean forward and tuck in behind it for coasting down hills.

A nicely done canopy by Metrofiets, diverting air from the rider's belly area,
protecting hands and lower extremities from cold air, and keeping the load dry.
All corners could be smoother.  Photo credit: Unknown, probably Portland area.

Another Bakfiets canopy, on a Harry vs Larry Bullitt.  Photo credit: Unknown
A BMW K100 sport fairing style curved top.  I would guess that the box is big
because there was a specific load it was designed to carry.  A side comment-
it is good to have rounded shapes and corners not just for smoother air flow,
but also for safety.  Automakers got rid of sharp mirrors and door handles inside
and out, and other similar bodywork in the 1970's, to reduce pedestrian and other
injuries during accidents.  Even a small radius helps.

The bolt on aerodynamic aids above will still allow cargo loads.  They are all leading surfaces, as I think that on a cargo bike the air flow is guaranteed to be turbulent, and a trailing edge device will have limited effectiveness.  They will offer some cold air protection, but only a little water protection.  The next step up would be a semi enclosed bodywork, halfway between a regular bike and a velomobile.  Several of the bikes on The Sun Trip during the summer of 2013 used solar panels as a canopy overhead, including several recumbents, however most were not particularly aerodynamic, and some of them probably had a poor reaction to crosswind gusts.

Lionelle Candelle with 300 watts of Enecom solar panels forming a canopy
over a Bullitt cargo bike.  He was running a 250 watt Ludo Technologie mid drive motor,
but the wind was probably stronger sometimes.  The canopy actually cracked
from flexing due to lack of bracing, and had to be welded back together
in Ukraine, but it did finish the trip to Astana.  Photo credit:  The Sun Trip

Team Declic Eco, Angélique and Guillame used a solar canopy and front fairing.
Most of the bikes on this trip were using 250 watt motors because of legal limits
for many countries, and were set up as mid drives to get around the limited power
in mountains.  This bike used a direct drive front hub, and finished about the
middle in a field of 2 dozen bikes.  Photo credit: The Sun Trip

The Velomobile of Josselin Bonmartel, with about 350 watts of solar panels that can
be tilted.  This bike didn't finish the tour, which was just as well as I think some of the
roads across the steppes would have trashed the body.  Photo credit: The Sun Trip

For comparison with the above semi enclosed bikes, the bike that covered
The Sun Trip's 7500 km the quickest and arrived in Astana, Kazakhstan first 
was Raf Van Hulle's Bilenky (or Hase Pino?) framed front cargo bike and trailer.
Notice the minimal frontal area.  Four solar panels were spread out at stops,
and then the 2 removable panels snapped over the 2 fixed panels for traveling.
I estimate these are about 125 watt panels, for a total of 250 watts moving, or 500 parked.
The front panel could be angled left to right.  Raf used a direct drive rear hub motor.
Photo credit: The Sun Trip

Added 23 August, 2014:  Raf Van Hulle's ideas for his 2015 bike.
Photo credit: The Sun Trip

Performance comparisons of the Sun Trip bikes need to be done carefully because of external factors.  For example the direct drive hub motors (but not mid drive kits) could probably be run above the 250 watt limit if the country allowed it, as most of those motors have insulation good enough for 300 to 500 watts or more.  Then it becomes a question of the motor controller's abilities to change parameters, the battery and panel capacities, and driving strategy:  If the weather forecast calls for a cloudy period during the day, should the beginning speed be slow to save the batteries for the cloudy interval, or should it be fast while the sun is out and then revert to pedaling during the clouds?  What if the cloudy period then turns out better or worse?  It's not a straight forward comparison.

A comparison of the frontal area of my solar electric Bakfiets and an ELF
at the Strolling of the Heifers.  (The Tesla S on the left had too much frontal area
to fit in the picture.)  My Bakfiet's long term average last summer was 13.2 Wh/m
at an average speed of 14.9 mph.  Unfortunately the ELF data hasn't been recorded,
so the best guess of energy use is between 20 to 25 Wh/m at close to the same speed,
which is quite a bit closer to a bike than to the Twike's energy use.  (The Organic Transit
website lists a 14 mile range with no pedaling at 15 mph at 75 F, or 30 mile with pedaling,
for the standard 11.25 Ah x 48 Volt battery.  This works out to 38.6 to 18 Wh /m,
with your guess on the level of pedaling and road grades.)

A more enclosed bike with a solar panel canopy could reduce it's frontal area by moving the seat to the rear to lower it, and then adding another panel over the pedals like a Bilenky style cargo bike rack.  The question for me is how much lowering is possible, without becoming a recumbent and losing some of the cargo carrying ability?  I will do some mock ups to test this with the third bike.

For those of you who can't wait for me figure out the fourth bike's
dimensions, I'll leave you with this option...   Photo credit: Unknown

Monday, July 7, 2014

July 4, 2014 parade and booth

The new solar electric cargo bike Longtail was in the 4th of July parade.  I'm going to post some of the pictures, and then follow up with a few more posts of pictures taken while building it.

The basic frame is a Longtail with a battery box in the center and rack on the back.

The bike with the glassless and frameless solar panel on it.  I've attached it 
temporarily, and am designing a clip-on hinge, so that it can be used as a cover
over cargo bags, and also be removed for carrying people or large packages.

Last year I was also put in with the kids bikes, but I don't like riding with them
because they have more energy and will run me over.  I exercise town energy 
committee privilege and wait for a few other groups to go by, and then start out.
It's OK, as long as I'm in front of the horses.

A young builder thinking about making his own bike!  Photo credit: Peter Allison

A lot more fun than the School Board, and no school budget discussions either.
Photo credit:  Katie Rawson

My Hartland Energy Committee booth before the crowds arrive.  On the left is a poster about cargo bikes and data from the bakfiets last summer.  The center poster is on weatherization, highlighting the current Neighborworks HEAT Squad program and $100 household energy audits.  The right poster is about solar, announcing 3 solar workshops being held by Suncommon at the library (the building in the background), and the Vital Communities Solarize program, a group purchase bulk pricing drive for residential installations.  And of course the bike- either most important, or bait for handing out weatherization info, depending on your point of view...

The first test ride on the Longtail.  Chiho's comment was "that has a lot of power!"
I haven't kept an exact count, but the Bakfiets probably reached 100 test rides
at the Upper Valley Electric Vehicle Forum on May 15, and now I'm ready
to start loaning it out for longer trials while I test the Longtail.

I'm not going to say much more about the Longtail until I've done 10-12 rides to be sure that the numbers are consistent, other than to say the Longtail seems to be a bit faster than the Bakfiets, and also use more power.  The direct drive motor also idled along at a 4 year old's parade walking speed nicely (about 2 mph).  An experiment will have to be holding the speed equal and seeing if the energy use becomes equal.  It is also much different riding at the front, after being way at the back of the Bakfiets for the last year.  In a review of the 8 Freight cargo bike, the tester mentioned a distinct feeling that he was moving sideways when turning, and this bike is similar.  On the Bakfiets it was more like watching the front end pivot around the rear wheel.