Saturday, October 4, 2014

A Speed/Wattage Comparison Tool for GPX Routes

Cycling Calculator for a Specific Route 

The side effect of me now having a much better handcycle, aerodynamically speaking, is that I have new questions:
  • What should I expect for average speed given stop signs and wind for a given route?
  •  If I expend a given wattage of work (let’s say a constant 125 watts) over a route with a zero wind speed, what should I expect for bike speed if the wind is xx miles per hour ?
  • What happens to speed when I add stop signs to a route?
  • If I know my bike’s aerodynamic coefficient, the wind speed, and the number of stop signs for a give GPX route, can I calculate my average wattage?

One reason I want this information is, I am thinking about doing the 24hr Sebring.  I have completed the 12 hour Calvin’s Challenge for which I rode 193.5 miles at 16.4 mph average (with an aerodynamically awful bike).  The handcycling WR for 24 hour is 403 miles at just under 17 mph average.  The Sebring course is VERY flat whereas the Calvin’s Challenge is far from flat.  So an enhance bike calculator can show me the expected difference.

I could not find a cycling calculator that answers the above questions given a GPX route – so I wrote my own.  This calculator takes a GPX route and does all the hard math along the many waypoints.  Most of the GPX routes (like those from MapMyRide or RideWithGPS), show a latitude/longitude/elevation for every 10 meters or so along the route.  With the lat/long/elev, one can then calculate the distance and slope from waypoint to waypoint.  With a known wind speed and wind direction, and the above information, one can then calculate the average speed for the route for a given constant energy (wattage) output. 

With all the waypoint-to-waypoint and wind information, the calculations become “interesting.”  Let’s say from one waypoint to the next waypoint, there is an upward slope. In order to maintain the constant wattage (again for this example – 125 watts), the cyclist is going to slow due to the climb.  But when the cyclist then moves on to the next waypoint that is for this example, flat, the cyclist speeds up.  But the acceleration from one speed to another requires work. That work might limit the speed in order to maintain a constant 125 watts.  Also, the rolling resistance will increase or decrease with speed.  All of this can be calculated. 

Acceleration and speed are manipulated all along the route in order to maintain a constant 125 watts of power.

With my calculator, one can also throw in any number of stop signs for the route.  For each stop sign, the cyclist is slowed to 0 mph and then then cyclist must accelerate up to 125 watts.  The average speed drops at the stop and then moves up during acceleration. The slope may help (or hinder) acceleration.

All of the above calculations take into account the wind direction, cyclist/bike weight, rolling resistance and drive-train efficiency. 

So what is the result?  For my bike, I found that the aerodynamics is so good, that in order to improve the performance of the bike, the bike must lose weight!  I am 6”-3” and 190 lbs.  I can lose a few more pounds – but not many.  As well, relative to handcycling, the amputee handcyclist is always going to beat the legged one – unless the amputee is really fat – given that everything else is equal.

Here is the result for standard workout without wind and without stop signs.

If I add the 15 stop signs:

What if I use the same input data applied to the Sebring 3.6 mile course with zero elevation change:

So, as you can see from the above, I “should” be able to go about a 1.66 MPH faster on the Sebring track with the same effort as my nightly workout.  Of course I do not expect to average 19 MPH.  So what do can I expect for average output if I were to maintain 17.4 MPH on the Sebring track?

If I were to change the front fork on my bike in order to accommodate a regular derailleur (I have a heavy I-Motion 9 transmission now), I will drag around far less weight.  Currently, I use the transmission so that I can shift down at a stop sign.  I have over-stressed my shoulders to many times by getting caught in the wrong gear at a stop with fast cross traffic.  But I should not have that problem on the Sebring track.  Here is the calculations with less weight:

As you can see, the weight difference (10 lbs) allows me to go about a ½ MPH faster (or use less energy) on the Sebring track.  That is a huge change!  Anyone that has done long distance events, know the difficulty of gaining just a tenth of a MPH increase over many hours.

Sunday, September 21, 2014

Update on Newest Base...

After putting over a 1000 miles on the newest edition of the handcycle, I find that each time I come back home after a ride, I tell my wife, "This bike is fast!"  I guess I am just not used to going this fast considering the effort I am putting in on the rides. It is not that I am going easy -- it is that I am going as fast with less effort or faster with equal effort.

On this past Friday night ride, the first half of the ride was basically slightly uphill.  On my old handcycle I usually expected that the I would average 15.0 to 15.6 mph by the turn-around before the downhill section.  On the new bike, I had averaged 16.8 on this half of the route the one time I cycled it.  Last Friday, I averaged 17.8 mph -- and I was not pushing particularly hard.  After the turn, I was on a long, very gentle decline for the second half of the route.  So I averaged over 20 mph -- with a total average speed of 19.2 mph for the completed route.  There was no way I could even get within a mile per hour, drafting, of that on my old handcycle. And this is with many stop signs!  So that is why when I arrive home, I proclaim, "This bike is fast!"

Ok -- my bike computer must be off?  On Saturday, I went on a 84 mile ride only to find that indeed it was off when compared to MapMyRide-- but the difference would only make my speed faster rather than slower.  That is, the bike computer was reporting too little mileage.

I did make the bike faster after my last post.  The area under the headset is a separate closed volume for which I originally did not add the 2-part urethane.  By adding the urethane to that, I added another degree of stiffness to the handcyle.  I did this Thursday night. So on Friday, this stiffness translated into more speed.  Albeit, I do miss to some degree the slight "forgiveness" (comfort) of the base before adding this final touch of stiffness, I enjoy the extra speed.

What's next?  The front fork's (originally designed for a different base) leg holders have to be reshaped in order to drop the knees slightly and improve the alignment of the leg position to the base. This should bring the bike closer to the design's Coefficient of drag (Cd) -- thus faster. Also, I am going to redesign the hand grips.  I find that my hands cramp on the long rides.  This may take a couple of attempts.

My dream is to reach the 20 mph threshold on a 25 mile workout -- repeatably. Maybe it is possible...

Sunday, September 7, 2014

Handcycle -- New Base with Improved Aerodynamic Design

Small changes in shape can have a big impact on aerodynamic drag.  Surprisingly so.

Cd of 0.18
The 3D design:

I am testing a new base to my handcycle for which the design has been put through hundreds of iterations of Computation Fluid Dynamic (CFD). That is, I take the 3D design, put it into a virtual "software" wind tunnel, see the results, make changes to the 3D design, and do it again -- hundreds of times. As small changes improve the aerodynamics, I maintain those changes in the 3D design.  My challenge is to reduce the aerodynamic drag on the handcycle.

The base is connected to my existing front fork (the front fork takes much longer to build).

Some of the rules of thumbs of design that I thought would hold true, did not.  For instance, a "longer tail" did not give the best results.  A "narrower wheelbase" did not give the best results.  "Closer to road" did not give the best results.  That is not to say that the rules of thumbs are wrong -- but only that the best overall design results may be achieved with otherwise.

That being said, the best measure of aerodynamic design is of course, "How fast is the handcycle on the road?"  This handcycle is easily the fastest I have ridden.  


In the first three days of taking the new bike out on the various rides (that I have done many, many times) I broke three of my course records easily.  The first ride was a 25 mile "time-trial" with something like 14 stop signs and a 450 foot elevation gain. I averaged 18.4 mph.  The next ride was a 42 miler for which there is about 500 foot gain and too many stop signs to count.  I averaged 17.1 mph,  The third ride was a 72 miler with a 800 foot ascent with 30+ stops including numerous stoplights.  I averaged 17.4 mph. None of the rides were "all out" -- but they were definitely above a leisurely pace (for which I know little about).

Of course I do not stop completely at all stops -- but I have to slow to a a max of 5 mph since the corn, soybeans and grass is high -- and I am low to the ground.  Most stop-sign corners are not mowed -- so I have to slow considerably.  And I do literally stop at stop lights. Of course then I have to speed back up -- the hell of handcycling.  On the flats, without wind, I am guessing that I average in the high 19's -- maybe even into the 20;s.  But there is nothing really flat -- even 20 or 30 feet ascent over a mile makes quite a difference.

Why faster:

So why is the bike faster?  Through the many iterations of CFD on the 3D design, I found that the small changes added up quickly to a better overall drag coefficient.  Most importantly, I was able to: 
  • Reduce the trailing drag by reducing the compression of air under the handcycle;
  • Replace the rear axle with a "wing" shape;
  • Contour the headrest/hydration bag area (many! iterations);
  • Keep rear wheels vertical.  A 5-10% slant of the rear wheels increased drag.
  • Reduce frontal area under the legs.  The base comes to nearly a point at its front edge.
  • Much lighter.  The base lost 15 lbs!


The above times were completed with a fair amount of flex in the base.  I added four more layers of  carbon fiber (two on the top, two on the bottom).  This helped -- but the bike still had too much flex when  I was climbing hills.  So I added a 2-part urethane foam to the interior of the base as a structural element. The urethane can withstand 20 lbs./sq. in.  At all out, I might be producing 200 ft-lbs/sec. Given the huge surface area of the base, the urethane foam appeared to be more than adequate. The urethane (2 lbs. per cu.ft.)  greatly reduced the flex -- and was a simple, simple solution to the problem,  The addition of the urethane only added a quarter pound to the base and required about an half-hour of work -- mix the 2-part liquids, pour it into the base, and in 5 minutes, it expands about 20-25 times the liquids volume. In fours hours, the urethane is hard.

Added features:

2 rear disc brakes!!!  Man, this bike can slow or stop much quicker.  Thus I can come into a turn (or stop sign) at higher speed and reduce the speed quickly.

Large hydration bag.  I now have a 6 liter hydration bag instead of two smaller ones. 

Positionable hydration line holder. I can place the holder out of the way for when I enter or exit the bike.Plus the water line is always available by moving my head forward on inch.

Regular wheels (thus cheaper!!!). These wheels cost a total of $290 with the Shimano hubs.  I used a 1/2" x 20 tpi hex-head SS bolt going into an aircraft-aluminum axle holder wrapped in about 30 layers of carbon fiber fitted into the "wing" axle.

I have hit some big-ass holes, ridden over miles and miles of newly "tar and chipped" roads (for which my teeth nearly vibrated out of my jaw).  The axles appear to be quite strong.  My axle holders are but 1/2"x 20 tpi automobile aluminum wheel hub nuts -- 2" long, costing about $4 apiece.

This may be my last base I build since it might be very difficult to improve upon it...

USB Rear Light:

The base holds a large lithium-ion polymer battery that can hold a week or two of charge for the 15-LED rear light.  There is a USB line hanging out of the rear (see pictures above) to allow for EASY charging!

Monday, May 12, 2014

Add two DIY Carbon Disc Wheels

Finished Wheels with Carbon Fiber Covering

If you have read this blog previously, you probably get a sense that I hate to spend good money on expensive frames and parts if I can figure out a way to build those myself for a much more reasonable price. The one area for which I may be able to gain the greatest benefit (speed) for the least amount of cash, is to build disc wheels.

My initial search of what others have done relative to DIY disc wheels was to find that many had bought sheets of plastic, cut a circle from it, slice the circle from its center to its perimeter, and make a "cone" or disc that fits from the hub, to the spokes and to the rim.  Then the do-it-yourselfer would then glue, tape, and/or mechanically fasten the plastic disc in place relative to the hub, spokes and rim.

Some studies even propose that DIY disc wheel covers perform equal to the expensive carbon fiber disc wheels.

As I was searching the subject, I found that many disc wheels actually contain spokes.  Well my existing wheels contain spokes.  And those wheels do not have rim brakes...

After ordering 24"x24" - 3/32" ABS plastic sheets, I retrieved the delivered box from the front door, only to find that the box of plastic was actually pretty heavy.  I do not need more weight for the handcycle! So I decided to use a a similar technique as mentioned above, but with a single layer of carbon-fiber substituted for the ABS plastic.  This would reduce the weight considerably.

The process of building the wheels was to use one layer of carbon fiber for each disc (one disc per side of each wheel).  I laid out the carbon fiber fabric on the ABS plastic I had bought (with the intention of building wheel covers). This ABS plastic was cut to a diameter that slightly larger than the wheel rim. I then applied the West System 206 epoxy to the carbon fiber fabric. The side of the carbon fiber fabric on the ABS-side was to be the finished surface.  After the epoxy cured, I was able to remove the carbon fiber from the ABS plastic.  It did not stick (to speak of). This gave me a fairly clean, flat and finished piece of single-layer carbon fiber.

Each  carbon fiber disc, after I sliced it from the center out and cut the hole for the hub, was glued to the rim. Of course the hub and spokes were in place on the rim.  The discs were oversized initially  There is an overlap now where the disc was sliced from its center outward as it is fitted to the hub outward to the rim. This is to accommodate the dishing of the wheel.  I applied epoxy to the overlapped area of the disc.

After the epoxy attaching the carbon fiber disc to the rim cured, the extra carbon fiber that extended past the edge of the rim was removed with a dremel tool.  

The first 30 miles on the DIY Carbon Fiber Wheel Covers went quite well.  I was surprised that the wheels were not "louder."  In the past, I have often heard cyclists with carbon-fiber wheels approaching me long before I was passed due to the vibrations/drum effect.  I expected the same from these.  But my expectations were not quite met.

Thursday, May 1, 2014

After 150 miles...

Delta-handcycle after the first 150 miles
My thoughts after 150 miles...

1 -- The handcycle feels pretty good.  That is, it fits like a glove -- or more precisely -- a recliner fit to my body.  As an example, when the weather is bad or I start a late workout I put the handcycle on the rollers, indoor, and throw a couple of  movies on the "big screen", I often do not get off the bike even after I finish the workout. Instead, I continue to lay on the bike to finish the films.  Hell, it's more comfortable than a chair.

My recliner for movies...

2 -- The disc brake is fantastic.  I will never ride a handcycle without a disc brake again.  On my next bike, I will design two of them into it.

3 -- The bike is LOW to the ground.  That makes it nice for getting off the bike.  I just slide over to the garage floor.

 4 -- Faster? I don't yet know if the bike is faster as compared to my Force-G or not.  My right shoulder is 4-5 months past the surgery.  The shoulder is getting stronger quickly but I cannot push particularly hard.  Though on a couple of small downhill areas of my "nightly" ride, I would hit 26 mph max (after a climb). In comparison, without pushing, I did hit 27 mph with this bike.  So, this bike is at least as fast as my old one.  BTW, I had modified the Force-G such that I was laying pretty flat on it as well.  I am carrying about 7 extra pounds on the bike (about equally divided between the transmission and the carbon fiber). As well, after many months of little activity I have gained 7 pounds of fat.

5 -- I had to rebuild the handles in order that my hands would be more comfortable (the other handles were too small). Additionally, I wanted the right handle to have the twist shifter built right into it.  The design works out very well!  It is comfortable for short rides.  The real test will come on the 200K and longer rides.  That is when the hands cramp and shifting becomes more difficult.

Right-side handle with twist shifter built in

Wednesday, March 5, 2014

Delta Handcyle on Rollers

The shoulder is healing.  At this point, I am allowed to ride 8 minutes a day at about the lowest pressure. It's a bit hard to suck up to the fact that it hurts more to do 8 minutes of riding on rollers than to do a 12 hour cycling event.  But then again, it better to be doing 8 minutes than nothing!  I finally have my newest DIY handcycle on an old set (25 year old!) of rollers that have been accumulating dust for far too much time.  The bearing are good, and the rollers spin like a charm.  And they are very quiet.

Over the last weeks, I have been working on a self-righting mechanism for the steering.  I have tried very heavy rubber tie-downs (not bad), stainless steel spring (not bad), and a carbon-fiber composite spring (needs some work).  I may use two stainless steel torsion springs -- one on each end of the steerer tube to self-center the steering/fork.

If the weather is warm this upcoming weekend, I will be using some acetone to dissolve the foam inside the rear horizontal cavity.  It is there that the two 100 oz. hydration bags will be placed.  The holes are drilled in the carbon-fiber to fit the mouth of the bags (see picture below).  Otherwise, the bike is ready for the road!

Sunday, February 9, 2014

Delta Handcycle Version in Carbon Fiber

Delta Handcycle Version in Carbon Fiber

The delta (two rear wheels and one front wheel) design is nearly finished.  I still have to add finish epoxy coatings to the bottom and put a UV protective clear coating over the full handcycle.  As well, I have to add the cables for the brake and shifters and the shifter itself.  

The rear wheels spin well within the wheel housings.  Each rear wheel tilts in about 5 degrees from bottom to top.  You might notice the large structural "keel" (unfinished) running the length of the base.  

The chain incorporates half link(s) in order to size the length correctly for the transmission.  Again, I am using a transmission system in order to remove some problems that I often encounter with shifting. In particular, if I have to stop suddenly and am in a high gear without down-shifting, I put tremendous pressure on my shoulders trying to get the bike moving again.  On my aluminum handcyle, I broke two welds over the last couple of years when caught in such situations.  With the transmission, I can shift when stopped.  

I am not sure when I can put this on the road since  I am still in the process of getting my strength back from rotator cuff surgery on the right shoulder.  A full tear and a partial tear has slowed me down a bit.  Luckily I am at the point of adding very light strengthening exercises in my rehab process.