There's lots of debate on which frame is most aerodynamic; the debate itself is not helped by the publication of manufacturers' tunnel data, which often relies on a series of shaped assumptions and protocols.  What's needed is an agent who would test the frames independent of manufacturer involvement.  This is a very difficult proposition, even apart from cost--there still is the issue of how variables between frames can be normalized so that meaningful data can be collected (for instance, how should the tunnel test for different height head tubes among frames that are nominally the same in size?).  An individual on one of the large discussion forums has decided to independently test a wide range of time trial/triathlon bikes, with the support of the forum community members, who are loaning bikes for the cause.

It has been my practice to isolate major variables first, and then look a finer details of granulation.  Here's what I've suggested:




Bicycle aerodynamics is an incredibly complex phenomena that is dependent on oftentimes individual variables.  It is important to remember that the benefits of even the most aero frames can be negated by poor body position or equipment choice.  It will be interesting to see the results.

What makes an aero helmet aero?  Is its long tail or some other factor?  Do aero helmets offer a performance advantage over standard road helmets?  According to most experts, after getting aerobars and establishing a solid aerodynamic position, the next biggest gain a rider can make is by purchasing an aero helmet.  Depending upon the study cited, an aero helmet will provide an approximately one minute savings over 40km; of course, the raw time savings will be greater for slower riders. 




John Cobb, aero guru and bicycle designer, has spend a good amount of time in wind tunnels testing all sorts of factors that impact on cycling aerodynamics.  Joe Friel summarizes Cobb's findings as follows:

This is a critical point, as most riders assume that the purpose of an aero helmet's long, sweeping tail is to smooth airflow transition over the back. Cobb himself notes in his original article that

In other words, the original design of aero helmets arose prior to the widespread use of aerobars and aggressive steep seat angle bike positions.  Cobb continues on by observing:

Cobb's testing seems to contradict popular wisdom:  If this point of an extended tail on an aero helmet is to manage airflow over the back, then why would sticking the tail of the helmet vertically into the airstream make negligible impact on aerodynamic performance?  Clearly, something was amiss.

In later in his article, Friel offers a compelling hypothesis: 

What Friel is saying--and what is suggested by Cobb--is that the benefits derived from an aero helmet arise not so much from its shape, but rather from the fact that aero helmets offer a smooth and solid aspect to the wind.  The reason why aero helmets were faster with their tails up than when their tails rested flat on their backs is because tilting the tail of the helmet up takes the front vents on the helmet out of play.  Accordingly, a smooth, ventless, and relatively round helmet should test faster than an aero helmet with a long tail and lots of vents.  (NB:  It was later theorized that the long tail of an aero helmet does have a practical aero benefit for riders who place water bottles behind their seats--the tail-up position directs airflow a bit higher over the back, which takes that messy airflow of seat back mounts out of play.)

Note how raising the tail of the helmet down removes the front vents from play.

My own field and aero testing confirms the observations above--that helmet vents and aerodynamics don't work well together and that smooth helmets are better than helmet surfaces with many discontinuities:

• If you are going to use an aero helmet with a long tail, then choose one that transitions well to your back. 
  
• I suspect that the whole tail pointing up stuff really has to do with the front helmet vents being removed from the wind stream.
• Basically, vents, holes, discontinuities in the helmet's surface = less aerodynamics.
• For my position, I found that a round, smooth helmet has close to the same drag as a long-tailed aero helmet (close enough to be in the range of error).
  
• I also found that covering the front of a standard road helmet with mylar can be fast, too, though road helmets tend to sit a little higher than some TT helmets.
• Most riders overstate the value of vents in most aero helmets in terms of heat reduction.  If one wants heat reduction, then one needs to think in terms of a road helmet (which won't be as aero).
• Shrugging the shoulders and extending the chin forward drops the head, making any helmet at least marginally more aero.

When in a good aero position, the results of a full season of testing yielded the following data, arranged from fastest to slowest:

1. Simple tight fitting lycra swim cap (or fully bald head--but I couldn't test this!)
   
2. Plastic aero head fairing (Catlike, etc.)--very low profile, no vents, but no crash protection
3. Bell Vortex helmet, with visor (can still get these on ebay)
4. Modern aero helmet (giro, spiuk, etc.)
5. Standard road helmet, front vents covered by plastic/mylar, etc.

The take aways:

Once you get rid of the vents in a helmet and make it as low profile as possible, they all are pretty good. Some might be a bit better for you position, but I don't think that any aero helmet will be dreadful. Those who extend their neck and drop their head low have a lot more flexibility in helmet choice than those who sit with their heads high above their shoulders when in the aero position.  If your aero helmet has vents, consider taping them over.

This season’s observations:

• I won a total of 16 races this season and placed top three in two; one race was an outlier and I was way off the winning time.  
• My threshold power at the beginning of September was 407 watts, which is slightly higher than last season’s.
• My ideal cadence appears to be in the high 70s.
• I was pleased with my performance in each of my 56 mile+ races.  The first I won by a 12 minute margin, even with a 8 minute penalty for passing an official’s motorbike on the left; the second I average over 30mph for the middle 34 mile section of the course, but almost had to walk the hills at the beginning and end of the course because I was running a single 60t chainring.  These hills were steep!
• In September’s 56 mile race, I was very, very happy with my performance.  I had settled on a 56x46 chainring combination and even used the small ring three times on this moderately hilly course.  I hit the 51 mile mark at 1’54”; at this point, I flatted my rear tire (it was a wet course and there was a lot of junk on the road).  Support reached me after 10 minutes or so; I got a tire change and support took off, but I flatted again about 100 years down the road.  It took support over 30 minutes to get back to me.  Despite the two flats, I was excited by the quality of the ride—I’m finally figuring out how to do long TTs and was on pace to do a 2’06”.
• My aero position has proven to be both comfortable and extremely effective.  Too bad that it isn’t UCI legal.  I’m going to head to the tunnel again this fall to see if there is a way that I can get a UCI configuration and still retain the efficiency that I gained this season.
• I’m still have a lot of issues with pain and impingement in my hip and there is an impact on my riding.  I think that I’ll get another season before I need to get some surgery.

Goals for next season:

• Ride a sub 2’05” 56 mile TT
• Win the pursuit at Masters Track Nationals
• Ride a 46km hour for a master’s performance hour record
• Be smarter regarding equipment choice!
  

 

In a recent article, I looked at how moving to a shorter crank length could have a positive impact on power output by opening up hip angle while in an aero position.  The idea is that by shortening crank length by x-amount  and by raising aerobars and seat heights by the same amount, a rider can maintain his/her current aero position with less hip impingement.  For those who are struggling to get into a lower aero position, reducing crank length enables a lower frontal position for the same hip angle.  The concept is relatively simple:

Shortening crank length means that seat height needs to be raised in order to maintain the same overall extension of the leg.  Once this is done, the rider can decide whether to raise his/her aerobars the same amount, thus opening hip angle and possibly enabling greater power generation, or to leave the aerobars at the same height, which will result in a lower, more aerodynamic position.

Selection of crank length is somewhat of a black art and science, combined.  There are many who support the proportional crank length argument, which is that crank length should be match to a specific morphological characteristic, such as inseam length.  For instance, Dr. Michael Fararri (who has some degree of notoriety for reasons other than his biomechanical analyses of cycling)  recommends:

• from 75 to 80 cm "inseam" = 170mm crank length
• from 81 to 86 cm = 172.5mm
• from 87 to 92 cm = 175mm
• 93 cm and higher = 177.5mm

 He then further refines the recommendations above by factoring in a femur/tibia ratio—that is, if “the ratio femur/tibia is superior to 1.13, it might be suggested a slightly longer crank (usually by 2.5mm) than the one indicated by the inseam measure.” 

 Finally, pedaling style needs to be accounted, too.  Dr. Ferrari  suggests that “rouleurs,” those who prefer to pedal in a seated position, might opt for slightly longer cranks, as opposed to those who climb out of the saddle (typical of lightly built climbers).

 All of this can get overwhelming.  There are, though, others who advocate a simple formula approach.  Such formulas are based on the notion of an average femur length:

 

Inseam length in mm x .219 = Crank Length .

 

So far we have accounted for inseam length and physical morphology; we still need to account for predominate muscle fiber type.  In very simple terms, there are two types of muscle fibers: Type 1, slow twitch (recruited during endurance events); Type 2, fast twitch (recruited during explosive, anaerobic events, such as sprints).  Two basic principles need to be kept in mind.  All exercise initially will recruit Type 1, slow twitch muscle fibers; if the event is long in duration and short in intensity, then Type 1 fibers might be all that is used for the entire session.  Second , each athlete has a different proportion of Type 1 and Type 2 muscle fibers; research is inconclusive as to whether the proportion of Type 1 and Type 2  fibers is a genetic or acquired trait.

 While a needle biopsy is necessary to determine with any specificity the fiber type of a specific muscle, athletes can gain a general sense of their fiber type ratios by performing the following test as outlined by Jason R. Karp:

An indirect method that can be used in the weight room to determine the fiber composition of a muscle group is to initially establish the 1RM (the greatest weight that they can lift just once) of your athletes. Then have them perform as many repetitions at 80% of 1RM as they can. If they do fewer than seven repetitions, then the muscle group is likely composed of more than 50% FT fibers. If they can perform 12 or more repetitions, then the muscle group has more than 50% ST fibers. If the athlete can do between 7 and 12 repetitions, then the muscle group probably has an equal proportion of fibers

 

What does all of this mean within the context of crank length selection?  According to Dr. Ferarri,

• A longer crank could be adopted by those riders who have predominance in red fibers (or Type I) and prefer high pedaling cadences, while suffering high force peaks.
• A longer crank in facts requires, at a given power output, lower force peaks for each pedal stroke, due to the longer arm.
• Those riders who have a higher percentage in white fibers (or Type II) instead, more used to bear high muscle tensions, could opt for shorter cranks, allowing inferior articular excursions and reducing internal friction.

 

In simpler terms, I recommend the following as a starting point:

• If you are a lightly built endurance athlete who tests primarily for Type 1, slow twitch fibers, you might want to opt for a crank length equal to (or slightly longer than) Dr Ferarri’s inseam proportion recommendations (listed above).
  
  
• If you are a muscular athlete who excels in explosive efforts (i.e., has a high percentage of Type 2 (fast twitch) fibers, then you might want to opt for a crank length equal to ( or less than)  Dr. Ferarri’s recommendations.

In a future article, I will explore the final variable in crank length selection, which is cadence and foot speed.

Photo from a Central New York Crit in the early nineties.  Course was an oval loop: straight up one side, straight down the other.  A cross road bisected the course near the bottom third of the descent.  I would blast the downhill, which was very steep; when I hit the cross road at over fifty miles per hour, my bike would launch upward and I would cover some remarkable distance.  I was dropped early in the race--I never could climb--but I worked hard to keep from getting lapped by the pack because the downhill was so exciting.

The photo below appeared in a major city newspaper.  Note the "bump" sign in the background.

This past Saturday, I was hit by a car while on a mid-morning training ride.  I was on my normal loop on which I do my intervals.  This loop has lightly traveled roads, with very broad sight lines; the surface is good and I feel that I'm as safe as I can be while riding on public roads.

At the time of the incident, there was a small mill on the right side of the road.  This mill was converted into a few small shops, none of which ever seemed to have much in the way of visitors.  As I was passing the mill, a car approached me from the opposite direction, then stopped approximately 50 yards away, signaling to turn left in the mill's parking lot. Just when I reached the parking lot, the car accelerated sharply and turned left in front of me.  I was going approximately 27 miles per hour; fortunately, I was able to turn hard to my right to avoid a head-on collision, but my path still ended up drifting in front of the vehicle.  The car hit my left hip and leg pretty much from behind; this, in turn, pushed me forward into the parking lot.  I managed not to crash (and not to hit--or be hit by--another car), and rode home with a bruise that covered most of the left side of my lower body.

While this certainly was a disturbing incident, it wasn't the first time that I was hit by a car.  In 2004, I was struck by a car in a very similar situation, only that time I was hit solid.  The car's bumper struck me in the left hip, launching me into the air; as the car continued forward,  I was hit a second time, which tossed me completely over the vehicle.  The end result was the loss of a good amount of my left hip and left shoulder, along with other broken bones and vertebrae.  I had a one month stay in a rehab hospital, a two month confinement in a wheel chair, and two surgeries.

Last Saturday's accident got me thinking about how many accidents that I've had while riding.  Back when I raced on the road, I was used to crashing, as I was a sprint specialist.  Oddly enough, the only injuries that I experience during this period were bruises and abrasions--I never did break any bones.  But when I shifted my focus to time trialing (to be safer), my string of bicycle-related disasters began.

Partial Catalog of Unusual Incidents

• Struck by car three times.  Two of the incidents are discussed above; the third occurred last year, when a car full of teenage girls passed me from behind and then turned sharply right, cutting me off.  I hit the side of the car and landed its opposite side.  Damage: broken right hand, dislocated left shoulder.
  
  
• Hit by golf balls three times.  This happened twice as I was riding through a local public golf course and housing development.  The first time was the result of an errant drive off a fairway located next to the road; this ball hit me in the left thigh and broke the skin, leaving a large bruise.  The second time was near the back nine, when someone enthusiastically chipped their ball out of a sand trap, which hit me in the left arm.   The final time occurred at another public course in an adjacent town.  I have no idea as to the source of the shot; as I was riding, I felt a painful impact on my lower back, and then saw a ball bouncing across the street.  I stopped, picked up the ball, and have it in my office right now.
  
  
• Picked up two snakes.  I like finding interesting objects that I can bring home to my daughters.  One spring day, I saw a beautifully colored toy snake on the side of the road.  I stopped, picked it up, and then immediately flung it away when it started to move on its own.  Coincidentally, an elderly woman was tending a flower pot at the end of her driveway near me when this happened. Upon hearing my scream (and seeing a snake take flight toward her), she let out a loud yell of her own and rapidly moved to her house, dragging her cane behind her.  Not learning from prior experience, I picked up a live snake a second time, only there was no audience present to participate in the event.
  
  
• Snagged by a fish hook.  This was just like the scene in There's Something About Mary.  My interval loop passes through a reservoir which allows public fishing; there's a fairly broad shoulder on this road, on which people park their cars and fish.  One afternoon, as I was finishing an interval, I felt a sharp pain, and then a strong tug on my left forearm.  Looking down, I realized that I was snagged by a fish hook. Apparently, one of the fishermen that day was in the middle of a cast as I passed by. Not knowing that he had caught a cyclist, the fisherman jerked on the line to free the hook.  The hook tore out of my skin, leaving a small scar; most interesting was how upset the fisherman was at me for disturbing his recreation.

Aftermath of an Accident, Part One (2004)




Same accident--note how the front fork was sheered off at the crown.  The front wheel and fork landed approximately thirty feet away from the crash.

After my recent experience at Timberman (see an earlier blog entry), I reconsidered the wisdom of using a single 60t chainring in front.  I realized that there may be times when it might be advantageous to be able to shift to a smaller chainring; the diffculty was that the 60t was so large that I couldn't fit a front derailleur on my bike so that it would shift correctly.  (The problem is both a clearance and derailleur geometry problem.)

The thing is, I really liked my 60t.  My optimal cadence is between 75 and 80 rpm, and the 60t accomodates this cadence band very nicely.  Also, on a subjective level, large chainrings just seem to role more comfortably for me than smaller chainrings, and I like having the option of having a really big gear for long downhills so that I can keep my cadence in control, while still being able to reach high top-end speeds.  (This is important, as I can catch a lot of riders on the descents who are better climbers than me.)

So what to do?  After Timberman, I swapped out the single 60t for a double 56,48 combination.  Even though I would seldom want to shift out of the 56, I thought that it would be nice to have a small gear option if the terrain becomes really, really steep.  I've been using this combination for two weeks now and find that it works well on relatively flat courses, but that I miss having the extra gear inches associated with the 60t when bombing down hills.    What I decided to do was run some calculations to see how fast I could go in different chainring combinations,  based on a specific cadence.  I chose a cadence of 80 rpm for my modeling, as this falls in my optimal range; I know that I can go at a much higher cadence if I have to, but this seemed like a good starting point.

Now there are two ways that you can calculate speed based on gearing and cadence.  One is to plug a formula into Excel and enter your variables; the other is to go to Sheldon Brown's Gear Calculator page (http://sheldonbrown.com/gears/) and work with his simple to use online application.  I chose the latter.



Sheldon's Online Gear Calculator

Sheldon's Online Gear Calculator is pretty simple to use.    In the gear units drop-down menu, I selected the MPH@80 RPM option; for chainrings, I entered 60, 58 (the largest chainring that I can use with a front derailleur that will shift), and 56.   For the cassette, I chose the stock 11-21 nine speed, which I normally use. 


Data Entered into Calculator


The resultant calculations are interesting:

What the chart above tells me is that at 80 rpm, I will be traveling at 29.7 mph if I'm in a 60x13; at the same cadence, my speed would be 27.7 mph in a 56x13.  Assuming that I could carry the same cadence for both of these gear ratios, the two mph difference could be significant.  Note that for a given cadence, the delta in speed grows greater as the gearing becomes larger: 35.1 mph for a 60x11 versuse 32.7 for a 56x11.

Is this delta significant in real world terms?  For me, I think so only if there are a lot of descents in a course.  On a flat course, I'll average between 29 and 31 mph; in this scenario, a 56x11 would fall nicely in my optimal cadence range.  If there is a large downhill--think the back side of Marsh Hill at Timberman, for instance--then I would want a much bigger front chainring.  Given that in most instances one needs to climb at some point in a loop course if it has great descents, a small chainring might be necessary, too, which excludes the use of a 60t (see above).

The goal, of course, is to match gearing to one's natural cadence and power bands.  By observing what gears I'm using during competition, I'll have  a better sense of whether I need to go with a bigger or smaller chainring in front.  Ideally, I like to have two bailout options: a small gear for going up hills, and a large gear for going down.  Based on the data above, it would appear that (for me) a 58t (the biggest chainring that I can use and still have a functional front derailleur), along with a 48 small, might be the best all-around combination to use for most of the courses in New England.   

Most riders today will install their aerobar shifters pointing downward, as in the image below of one of Lance Armstrong's 2009 Tour de France TT bikes:


In this configuration, to move to larger rear cogs (re: easier gears), the rider has to reach down and pull up on the right shift lever; to move to smaller rear cogs (re: harder gears), the rider has to move his hand in front of the right lever and push down.

In my experience, this is not a natural ergonomic motion, as it requires a good amount of hand movement to actuate a shift.  The awkwardness of this motion can result in a greater chance of shifting errors, even with indexing levers.


I've always installed my aerobar shifters pointing upward (i.e., reverse of most of today's set-ups) to minimize hand motion and to increase comfort while shifting:


In this "inverted" configuration, pushing forward on the right lever makes it easy to precisely shift into larger rear cogs, even when running shifting in friction mode.  To move to smaller cogs/higher gears, the right lever is simply pulled backwards.  Over the years, I have found this to be a much more natural and comfortable motion; a number of small efficiencies like this collected together can result in a disproportionately positive increase in performance.

Of course, the new generation of electronic shifters have the potential of being the most ergonomic solution of all.

I've been experimenting with a couple of different time trial bike designs.

Cervelo P3C

I posted a variant of yesterday's entry on data representation in an active discussion forum.  It generated a good amount of dialogue--over eighty posts in less than a day.  I just added the following commentary that summarizes why I posted the article in the first place: