If I need two wheels in the back...

After two test rides, I have found myself correcting some details on the handcycle.  The steering stops constrict the turning to too large a radius.  I have reworked the stops to allow a smaller turning radius.  The carbon fiber has required a few days of curing.

Additionally, one of the inline-skate wheels, when under a lot of loading, was scraping against the carbon fiber.  When I designed it, I probably should have given couple of millimeters of clearance.  I cut it pretty close.  No matter,  I think the problem is solved now.  The rubbing really slowed me so it was difficult to approach a speed in which I could balance the bike.

With these two problems solved (hopefully), I will give the handcycle another spin.  If the bike is too difficult to control at low speeds, then I have a backup plan.  I have designed a parallelogram tilting rear wheel assembly that will allow the rear wheels to tilt into turns.  The benefits of the tilting rear wheels are quite numerous.

I am a novice at the physics of the problem, but here is an explanation that makes sense to me. When the center of gravity (CoG) moves outside the triangle formed by contact points of the three wheels, the handcycle rolls over. When cornering, the front wheel turns and that triangle is now shifted to the side of the turn since the contact point of the front wheel has shifted.  The CoG then may fall outside that triangle -- to the opposite side of the turn.  That is when I roll.  I attempt to lean to the inside of the turn thus shifting my CoG along with the triangle.  In doing so, I am normally limited because my head hits the rear inside wheel.  If I can keep the CoG within the triangle, then I would spin out instead of rolling (if I am going too fast).  I have yet to spin out -- but I have rolled far too many times.

That triangle does have a height formed by a single point from the three corners. Think of a pyramid with a  three-cornered base.  By keeping the CoG very low (I have 1.5" clearance), then the potential of rolling is lessened.  But as the CoG moves higher (as it moves up the pyramid), then the triangle cross section becomes smaller -- thus I have a greater chance of rolling since the CoG may fall outside the pyramid.

So why a parallelogram tilting rear wheel assembly with two wheels?

(1) Rear wheels can be much closer together -- even within the widths of the shoulders.  If the rear wheels are within the width of the shoulders, then the rear wheels are no longer considered "front wheels" (unprotected).  The difference between a protected and unprotected wheel (two rear wheels in this case), can add up to many watts of power depending upon the type of wheel. In my case, I would think that I would be saving 35-50 watts or so with two protected rear wheels in comparison to two unprotected ones.  That is not a lot to a legged cyclist, but for a handcyclist, it surely is. At race pace, I am probably generating about 235 watts of power.  Fifty extra watts of power would be the world.

(2)  Higher speed in turns.  My last race had 60 turns over 42K.  I rolled the bike once when I did not slow down below 17 mph on that turn.  That cost me probably 30 to 45 seconds.  I probably had to slow down for another 40 turns.  One can only imagine if I never had to slow, what type of speed advantage I would gain.  Additionally, the process of sprinting out of a turn takes a lot out of me. Maintaining a single speed is far easier.

So I am attempting to put together a list of materials required for the above design. It will be constructed mainly from carbon fiber (I still cannot weld).

Inline skate wheels work -- but two wheels do not...

When at a stop, the two inline skate wheels work well in that I can easily rest on one or the other of them.  Also, I can easily ride on either one when attempting to get up to speed.  Unfortunately though, when I do get up to speed (5mph let's say), I cannot move my center of gravity enough to get off the inline skate wheel.  The center of balance of the bike is just too low as compared to the inline skate wheels.  If my center of gravity were a few feet higher like on a regular bike, I could probably just shift the position of my head or shoulders a couple of inches and that would suffice.

As was learned a few weeks ago (http://www.futurity.org/science-technology/top-heavy-bugs-show-how-to-hover/), flying insects can more easily hover if their center of gravity is higher rather than lower relative to total height.  Well, my center of gravity is so low that it takes huge movements to shift my weight over enough to move off the inline skate wheel.  The funny thing is, in 2006, a designer of recumbent bikes recognized the same problem.  I am too slow in doing so.

What are my options:
(1) Lower the back of the bike so that the inline skate wheels are closer to the ground and therefore require less lateral angle of rotation of the handcycle in order for the inline skate wheels to touch the ground. Thus, I would not need to shift as much weight to move off the inline skate wheel.

(2) Go to larger inline skate wheels (requiring a rebuild of the housing) in order to do the same as (1).

(3) Cut off the back of the bike, cut a couples of holes in the carbon fiber below the headrest, add an axle, and throw 2 wheels on the back.

(4) Go back to the design board...

(5) Pull out the bottle of single-malt scotch and a good cigar and at least enjoy those.

I think I will opt for (5) tonight...

Relative to (1), I am a little hesitant to lower my center of gravity even more.  It is very low now.

The steering stops worked quite well. Albeit, I need to allow a smaller turning radius for low speed.

My wife thought she took a picture of my test... instead, she turned off the phone.  But hell, she now knows that when pushing a mouse "up" on a mouse pad, the cursor on the screen goes up vertically as well.  She believed the cursor should have gone down on the screen.  Go figure. Now, she understands far better the interface of man and computer, so I cannot bitch too loudly. Man and smart phone -- that is another step in evolution...   So, no pictures.

Cables going into rotating handgrips

The shifter and brake cables break easily on a handcycle if those same cables have ends planted into brakes levers or shifters that reside on a rotating handgrip.  The usual cause is that the cables repeatedly bend at the spot of entry into the lever or shifter.  I have built a long holder for the cables that always keeps the loop of cable above the shifter and brake lever.  In addition I have added spring coverings to the cables at both places where excessive bending can occur.  The springs should extend the radius of the bends thus slowing the pace of breakage.

Getting ready for the second road test...

After I rode the handcycle on a trainer for a few hours in order to "get the kinks out" relative to cabling, braking, changing gears, and spinning, I was able to diagnose a few minor issues.  Specifically, I had to align the disc brakes and calibrate the cables. Additionally, I decided I had to build a more substantial cable holder since I was able to break the temporary one off pretty easily.  And finally, I decided to build, with carbon fiber, a holder for the speedometer/odometer pickup that sits on the fork.  The speedometer unit that I purchased is a cheap wireless one.  Therefore -- due to wireless nature -- the pickup has to sit fairly close to the rear of the fork.

But the bigger problems displayed their ugly faces when I removed the handcycle from the trainer and attempted to get on it and move it forward.  The front wheel easily "flopped" to one side.  This is not an unknown problem with handcycles.  Since a handcycle's front fork usually has an integrated crank, bracket and steering, the fork can have much more severe problems than those encounter by the typical bike.  I found it quickly: A flopping fork in which the bike's front wheel attempts to lay itself flat to the ground...

I decided upon two courses of action to relieve this problem: (1) Adding a self-centering steering mechanism and (2) adding "stops" on the steer tube holder that only allow the fork to turn a maximum number of degrees before confronting the "stops".  Therefore, the maximum flop can be controlled by the size of the "stops".

Both the stops (left and right) and the self-centering mechanism are in place.  I have some cosmetic work to do on the frame after the changes -- but those are much more minor as compared to the additions.  Here are the pictures: