Tuesday, December 01, 2009

The following was posted by me at a forum, but is worth repeating.

I'm going to step on some toes here. It's not my intention, but hard to avoid given the subject matter.

Anyone, including Brendan, is entitled to his/her opinion and to share same.

ATMO, not too many folks really know all that much about frame design, from established (even relished) handbuilders to product designers at major bike manufacturers. Don't get me wrong, they know enough to make a decent bike. But let me offer a few illustrations.

Years back, when I participated more on the framebuilders list, virtually no one could speak to their rules of thumb for determining the right combination of tubing diameter/wall thickness. Richard Sachs and Freddy Parr could offer insightful input on this topic, but no one else then participating could. At least not beyond the opinions that might be offered up here by riders (as opposed to builders). Mind you, all the builders here weren't participating then on the list: so please don't make any assumptions about any particular builders. Anyhow, the lack of input really opened my eyes up a bit - especially when I saw the range of solutions that recieved glowing reviews through various online/print venues.

In another example, its hard to find anyone who can really speak well to the pro/cons of different front geometries re handling. That includes builders and other industry participants who too often bluster their way through the topic. Let me illustrate some of what I mean, but first a warning that a lot of facts about specific designs are hard to find. For example, look at EPS geometries on Colnago's website: head-tube angle, rake, and trail aren't listed. It's a deep secret.

But, some secrets can be unraveled. Using a combination of resources (BikeCad, Anvil Trail Calculator, Excel), and keeping to a limited range of sizes (55-56CM top-tubes), it's possible to make fairly accurate projections of how these key measures vary by brand. I picked the size range based on it being toward the mid-point of sizes, and presumably less succeptible design anomolies that might be present at the ends of the size ranges.

NOTE: there are precision errors in published data - plus some companies may intentionally misrepresent specs. Anyone who can, is welcome to improve the accuracy of the numbers that follow.

Colnago EPS HTA=72 Rake=43 Trail=64.
Dedacciai Scuro HTA=72 Rake=44.5 Trail=62.
Cannondale's 'comfort' fit HTA=72.5 Rake=44 Trail=59.
Trek's 'performance' fit HTA=73.5 Rake=40 Trail=58.
Trek's 'comfort' fit HTA=73.5 Rake=40 Trail=58.
Cannondale's 'comfort' fit HTA=73 Rake=44.1 Trail=56.
Specialized 'performance' fit HTA=73.5 Rake=43 Trail=56.
Specialized 'competition' fit HTA=72.5 Rake=40 Trail=56.
Pinarello Dogma HTA=73.5 Rake=43 Trail=55.

We see the EPS with a trail of of 64mm while the Dogma has a trail of 55mm. Both are considered great handling bikes. An easy out would be to suggest that the Dogma is a crit bike - but I don't think Pinarello considers it such. Moreover, if there are different handling paradigms for different events, why don't more builders/manufacturers offer clearly deliniated models with different geometries for different events?

Instead, many lines carry the same geometry through broad portions of their line. Others that don't, make no attempt to suggest differing uses for models with different geo's.

In fact, looking at most of the literature, the debate re designs-for-events focuses on the question of weight versus stiffness. ATMO, that's an artificial distinction, but that's what prevails in the market today.

Meanwhile, I can come up with quite a list of venerated existing/retired builders who will simply say to keep the trail between 57-58mm and everything will be fine.

Going back to the sometimes deified Colnago, I believe that they stick to a 43mm rake throughout the size range. With the variations of HTA that occur across sizes, this will lead to different handling characteristics between their smaller and larger frames.

Where an I going with all this? The thoughtfulness and care provided by your builder is more important than tenure in role. And that's ATMO.

Tuesday, September 22, 2009

Las Vegas - Capital of the Dada Anti-life

Well… Here I sit waiting for my first seminar of my first Interbike. This, btw, is related to a new entrepreneurial venture that I’m trying to kick off for 2010 – not frame-building. Stay tuned here to be among those who first find out about it.

A rain delay in Dallas meant I got to my room at 2:30AM (my time). In Vegas, the dispatcher for the ground-transport company was convinced thought that being contrary and contradictory was really funny for weary travelers. Then, the registration desk at the hotel was very backed up – after all time of day hasn’t much meaning in Vegas. Finally I made the mistake of thinking that Vegas was only 1 hour ahead of Chicago – meaning I got up earlier than necessary.

The hotel is the Imperial Palace. The best aspect of these digs is that they reinforce my sense of entrepreneurial spirit (in other words its cheap). It’s rated 3 stars, the middle of the Vegas range. I’d hate to see the lower rated venues. Seriously, I’ve been in Motel 6s that seemed nicer. But, at the rates I got – it’s probably a fair value.

Having criticized the hotel, I should point to the objectionable elements. While my floor appears to be no-smoking, the ventilation system recycles the smells of tobacco smoke and air freshener. Not cool for anyone with sinus/allergy issue. Also, the room is old. How old? The shower still uses separate hot and cold faucets – one of which fits loosely enough that water sprays behind it and the fiberglass surround. Also, the shower and sink drains don’t – at least not beyond a trickle. The balconette faces a roof and two walls of windows – it’s almost impossible to see sky (fair enough that this is due to pricing). The cost of food is astronomical. I understand that the whole $2 buffet thing is long gone from Vegas, but $10 for a cafeteria croissant and small cupa is just a bit high. Especially in a place where one has traverse a floor of crazy, smoking gamblers to get in and out. Not my idea of the place for a vacation.

Being a newbie to Vegas, I am amazed that it exceeds any image I had of this town. In fact, I assumed that much of its reputation was a caricature – but that simply isn’t so. You’re welcome to laugh at me for being so naive, but there it is.

Yep, the characters in the hotel are pure Vegas. There are lots of bored looking people playing penny slots without really paying attention to their machine, or the people around them. It’s almost as if they’re catatonic. There were a few obvious prostitutes in the mix, some looking for action while others appear to be on their way home. I don’t know if they’re legal of not – but it feels strange to see them in a regular hotel.

On the other hand, the sun was shining (naturally) and it was 90 degrees this morning. I didn’t head to the Outdoor Demo Days: this morning was dedicated to journalists – and I wanted to attend seminars in the afternoon.

Instead it was time to get some fresh air (really important after the casino/hotel) and enjoy the dry heat. Walking past the Venetian (walking is probably a lower class thing to do in Vegas), the neighborhood looks nice and upper class. But I was hailed by a hooker, who wouldn’t give up after just one try. At 8:30AM Vega time. It was weird.

Coming back in the afternoon, there was a fellow selling "new Sony & Apple laptops" out of his shopping bags on the sidewalk in front of the Pallatzo and across the street from Dior. Down near Harrah's, someone was hawking CDs from a bag. Further down, there was a little gauntlet of people doing a rhythmic thing of slapping tickets on there wrist (making a slapping sound) and then thrusting them out in front of the pedestrians. At the last moment, they'd retract their arms if the walker wasn't a taker - starting the slapping cycle over. Based on their t-shirts, it looks like they were marketing a strip show. Again, weird.

Sunday, September 20, 2009

More on tires and wheels and such

Today I finally got out on the updated carbon bike. The core of the ride was 32.63 miles in 1:57:32 - or a bit over 16.8 mph, with a total ride of 42 miles. That's pretty good for me, riding solo, on a breezy day. And the route was as hilly as its possible to find around here (although this isn't very hilly).

Initial impressions of the SRAM group. All parts worked correctly. The brakes do have enough modulation for my taste, along with good stopping power. The shifters shift reliably, as do the derailers (Force has replaced the Campy front derailer). The chain and cassette don't seem loud, although they are a bit louder than Campy when cross chained.

The grip around the base of the levers felt good; my initial impression was that it was better than Campy (due in part to its greater width). The end of the hoods have a larger and more pronounce hump than Campy. I like to ride with my palm on the hump with Campy, but all the weight is concentrated on the tall nub with the SRAM shifters. Overall, I was having more Ulnar nerve issues today than usual - but I'll need more rides to determine if this is related to SRAM shifters, or something else. Going forward, I may put on a longer stem to compensate for not riding on the shifter nub.

Overall, I'd say that the tires lived up to expectations - but I have to qualify it for now because the wheels are also new. The tubeless ride is definitely smoother riding - with this being most obvious on sharp-edged bumps. Its not quite like a high-quality tubular, but it's clearly a step up from an open tubular.

One other observation re the tires: They are very nicely concentric on the rims. Recently, I was using some Michelin Pro3Race on test wheels while building the rear-wheel fairing on a tri bike. It was almost scary how far from concentric these (expensive) tires are - so much so that I took them off and remounted them just to make sure that the beads were seated properly. Of course, when testing for clearances, this makes a great worst case example. But while riding, concentric tires roll more smoothly with less rolling resistance. So, again, I am impressed with the concentricity of the Hutchinson tubeless tires (based on a sample of 2).

While my time today was good, I suspect that it had more to do with my current conditioning and a need to burn off psychic energy. I'm sure, at least for now, that the wheels weren't really doing their job.

There is a nice downhill run where I test my rollout. My usual wheels are either an older pair of Campy Zonda (with the steel spokes rather than aluminum), or some MA3's laced up to Record hubs. With these wheels, no one out-rolls me, and I pass up many folks going down this hill.

The Easton EA70s were clearly not competitive going down this hill. I don't know whether this is seal drag, drag in the rachet of the freehub, or a bearing issue. Note that these wheels didn't even have a mile on them when I started this morning - and less than seven miles when I reached the coasting hill. Its entirely possible that the issue is just one of breaking in the bearing seals. Subjectively, the wheels felt as if they were rolling better towards the end of the ride. Guess I'll have post further on this issue after I get a few hundred miles on these wheels.

Over all, I enjoyed this mornings ride. With the gears working correctly, I could better enjoy the ride and the handling of my frame and fork. This really is a nice frame, and handles beautifully. Today was a super morning to enjoy it. It was still a little dark when I started and about 54 degrees. By the time I got home it was 65 degrees, sunny with blue sky.

And that's a wrap for today.

Friday, September 18, 2009

Old & New

Spent yesterday working in the shop. Busy day. Finished it by updating my personal carbon ride. It's been equipped with Record 10s w/ a long cage rear derailer. Of late, it's not shifting right. The cassette is fine, the chain is new, and the derailer tab checks out as straight. So... looks like the derailer is bent. I tried straightening the cage a bit, but there are so few flat surfaces on it that I couldn't get a good read on the alignment table.

Meanwhile, I've been meaning to test out some Hutchinson tubeless clinchers. Supposedly, with Stan's tape and fluid, they work well on Easton rims. And I happened to have a spare set of EA-70s lying around (don't know quite why). The wheels, however, are set up for a Shimano cassette - so a no go-on my Campy bikes.

However, I also have a bit of spare SRAM gear lying around including some 1st gen Rival shifters and rear derailer. Which, of course, would work with Easton rims.

So, I did the only bike-geek respectable thing possible: I've converted the old bike to SRAM.

For now, I can only offer initial thoughts - I don't have any meaningful mileage on this setup. But, here's what I have.

1) Stan's tape was easy to apply - using my truing stand. Without a stand, I think it would have been a pain because the tape is really made for a wider (MTB) rim. It requires tension and a wiggling motion to get it seated properly down below the bead.

2) Unlike with regular tubes, the threaded nut on the valve stem is necessary. Without it, the stem doesn't pull in tightly enough to prevent leakage. Using the nut cranks the stem down nicely.

3) Stan's instructions are right on, but the video is the way to learn the proper install method.

4) It takes tire levers to mount the Hutchinsons (at least on the EA70 rims). I understand the concern about tires blowing off (or at least of having tires loosen enough to loose air), hence the use of a carbon fiber bead that won't stretch. But I have to wonder if the Stan's tape is contributing to this issue. I'm going to have to try tubeless on a set of Ksyriums where rim tape isn't necessary.

5) The tires pumped right up with a floor pump, even without the water and soap suds seal at the rim. After inflating them, I put them in the sink to work in the suds (see Stan's video for explanation). Actually, I couldn't pump the tires up until the valve stems were tight. But, once this was done, the tire beads seemed leak-free. BTW, the next morning, the tires still felt full when squeezed.

6) Installing the fluid wasn't too hard, but because the applicator looks like a big syringe, I thought to draw in fluid and then squirt into the valve stem. Doesn't work that way - too big a nozzle, too thin a fluid: it all spills out. So, I used the applicator more as a funnel - connecting it to the stem - and pouring fluid into it.

7) The initial test rides seemed very plush, but I need to go back and check pressures and try again with some actual miles. Also, I need to get out the tire gauge and see how long the tires hold pressure.

8) The old style (silver) Rival shifters have a nice looking finish. I don't know if its painted or polished and anodized, or? But it looks nice. It will be interesting to see how durable it is. These shifters don't have all the adjustments available on newer SRAM shifters, but seem to fit my hands just fine.

9) For now, I'm using a Campy front derailer. This isn't working well - I can't seem to dial in the shifting. Next stop, Rival front shifter.

10) The rear is shifting nicely. It's so nice to have every cog available to me again. Too bad I didn't have a cassette with a 26t or 27t available. Now I just have to train myself in SRAM shifting. Let's acknowledge here that these 1st gen, lower level, shifters don't sound or feel as nice as Campy. But they are much cheaper and seem to be shifting great with the Rival derailer.

11) The brakes are 1st generation Force. Again, attractively finished. Good braking power - still checking out modulation. Fit of the Q/R impedes easy access to adjust the brake shoe on the cable side - but not outrageously so. These brakes don't have the centering screw of later versions - a loss, but not a major one. The rear brake feels more spongy than I'd like, may have to reduce the toe-in.

12) I haven't weighed anything, but suspect that the SRAM components are lighter or equal in weight to the Campy ones that were swapped out.

13) The cassette has an attractive looking spider inside of the cogs. Otherwise, the cassette just works - nothing to write home about.

14) The ratchet in the rear freehub seems rather draggy. It's not noisy, and the issue may just be too much grease. I'll have to wait and find out if these wheels roll ok, or not. If the hubs roll, then its a great bargain of a wheel set - complete with nice (low drag) stainless spokes. But if the hubs are friction factories, I'm not going to be very happy with these wheels. Perhaps I'll have to check out the Easton website regarding how to adjust the bearings.

That's all for now. I'll provide some more feedback once I have some miles on this combo.

Tuesday, September 15, 2009

If bikes had a memory...

oh the stories they could tell. Today continued our outstanding weather - it feels like San Diego or something. With the day job starting late on Tuesdays, I scooted out for a quick ride on what is now called the Death Bike, which has some stories of its own.

Death Bike is a track bike, originally designed for one of our local racers. He, however, found a good deal on a used BH carbon frame and canceled right after the tubes were mitered. What the heck, the frame was only a little small for me, and I didn't have much time on a pure track bike design, so the build moved forward as my bike. Naturally, its set up as a fixed gear, but not like the path racer. The wheels are clincher, with 23mm open tubs from Deda - nice rolling tires. It has track forks with thick walled blades and limited rake making them stiff stiff stiff. The rest of the frame isn't much different being of 2 x oversize tubing (a 35mm steel downtube is a serious piece of pipe) - which combined with its black paint presents a very serious look.

Death Bike has relatively long trail given its steep head tube, because of the short fork rake. At slower speeds, trail quickly begins to stabilize the bike, but the steep head tube and short wheel base allow it to turn quickly and in a small space. Frankly, I think the geo is nailed for track riding. But it is a bit jarring on the streets, and there isn't room for cushier tires.

Death Bike had an old Campy mid-level (Athena?) brake on the front to keep me safe on the road. Yep, I drill the fork crown. But recently, a friend borrowed it to see if he liked fixed gear riding. For this, I swapped in a SRAM Force brake. That puppy offers right . . . NOW! stopping power - ideal for someone not used to only one brake. To simplify the brake change, I swapped in a new bar with a Cane Creek lever (leaving the old brake and bar connected to each other).

Anyhow, my rider has lots of time on fixed-gear training cycles, so I guessed he could make the step up to a fixed gear bike pretty easily. Whoops, I was wrong. Should have been there to give him pointers. Anyhow, he quickly got launched, twice, and decided fixed was not for him. Hence the name Death Bike.

The new handlebar is a Nitto Rando. I don't have much experience with these, so I was pleased to find that the shape was more comfortable than I expected. The Rando has a great bend behind the brake lever which supports the hand without leaning on the lever. This makes it ideal for a bike with only one brake lever, as both hands get support.

Today's ride pointed out that this is really a pretty stiff handlebar. Some folks take think of flex as a bad thing, but a softer bar offers more comfort than a stiff one. My ulnar nerve sometimes gives me trouble, making my hands/fingers numb. There are many factors related to this, but for me the primary ones are: fitness; & stiffness of the the bike's front end. At this point in the season, most bikes don't bother my ulnar: my weight is going into the pedals, not the bars. But, even on a short 1.5 hr ride, I experienced numb hands with the Rando on the Death Bike. Don't get me wrong, many other bars would do the same thing. But don't buy the Rando expecting it to enhance your bike's ride.

Both the roads & trails had light traffic this morning. Blue sky and temperatures in the mid-70's made for ideal conditions. I beat out a pretty good pace: both the frame and the wheels like to scoot. Meanwhile, my mind was liberated from serious thoughts. When I chose to consider it, I could tell that I was breathing hard and pushing the pace. But mostly, I just didn't notice. Other than the Ulnar thing, I could gone for hours at this level. Sort of a Zen thing - my mind and body all doing its thing in a very coordinated fashion, but without conscious thought driving any of it. I don't know if this makes sense to you, but its a cool thing when it happens.

That's it for today. Enjoy your rides. Pix of something soon.

Monday, September 14, 2009

Good Weather

Lots of work yesterday, but the weather was too nice to miss riding. I went out on the path racer (yeah I need to post some pix) this afternoon. It's always a peaceful bike to ride - even when pushed.

Along the way, some guy comes blasting out into the street, without looking, from a pedestrian path connecting to a residential neighborhood. He's all tricked out on his C'dale six/sixteen - and practically glowing with reflective bits from head to toe (despite it being a bright sunny afternoon).

A quick swerve avoids him, but the expected pass from behind never arrives. Half mile down is a stop for a busy road, and before I get across Mr. C'dale shows up. No problem, I'm just beginning to cross - but I get the feeling that he's right behind me and trying to pass. It's not a good spot for this as we are entering a trail with center barrier and a speed bump (both of which seem useless). So I keep my tempo up and the sound of him recedes as the path gets bumpy.

After a sharp turn onto another trail, and a street crossing, the trail scoots down a viaduct under the expressway, then it's back up and half a mile to the first street crossing. As I'm waiting to cross, guess who shows up and manages to time things perfectly to cross without stopping. All of which is cool, he ought to be making better time than me anyway.

But, once he's in front, he slows down. I'm not out for a race, but I like to maintain my cadence, and I can't just shift gears to adjust to his. The nice weather has brought out lots of slow riders, so passing is a bit chancy in many sections. Eventually the opportunity comes, and I'm off.

The next crossing has a stop light, and you can guess what happens. Instead of waiting for the light, I go ahead and time the traffic to shoot across before he can react. This time I up my tempo for a bit, build a gap and settle in. After a mile and a half he makes another dumb pass in heavy traffic. Naturally, as soon as we're in the clear, he slows down. So one more time I pass him, and stick to my cadence. It's a few more miles to the next crossing. While there's plenty of traffic, no one is trying dumb passes here. Cool, I get to relax and just enjoy peddling, with the subtle speed control that comes with a fixed (don't worry, I have front and rear brakes).

Before the next street, there are a couple of sharp corners. Ahead of them, I come upon a couple riding single file, a bit slowly. Judging distances, I can probably get around them before the corners, but it will be close enough that they might be scared, or pissed off. So I slow, and follow them through. And they slow more. As do I. And then they slow down even more. And so do I, again. Five-hundred feet begin to feel like a life sentence - but I'm staying cool.

Naturally, as we emerge from the trees, and try turn into the crossing, Mr. C-dale comes flying through, with total disregard for where each of us is, or where we are heading.

This time, I get to feeling pissy. Traffic holds us up from a quick crossing. At the gap, I take off but this time I spin it up and hold the up-tempo for a couple of miles. The sun is shining and the sky is clear and a bright blue. This portion of the trail rarely has much traffic, and is true to form today. While often less than 100 yards from the expressway, the path rises, dips, and winds through woods and past ponds. Its some of the best local scenery - it reminds me of Minnesota. Three miles of kicking it gets me to my 18 mile marker, and the beginning of loop around to start home. A few trails come together here, and with the expectation of traffic, I slow down.

Naturally I'm waiting for my shadow to jump out - but he doesn't. The last stretch must have finished him off. I return to my normal tempo. Fat tires (32mm), at 75psi, run quietly and absorb most of the bumps. The ride is mellow now. The light has a fall-like quality, even though all of nature is still green. The miles disappear without thought. The effort remains, but its natural - not something to notice.

People must be heading home, because I don't use either my voice or my bell for the rest of the way. Out on the roads, even the drivers seem to have chilled out. The air tempurature is in the upper seventies - its real short-sleeve weather. The three stop lights I pass are in sync with my ride- and the bike just spins on through.

Somehow, the messages are muted where I'm normally reminded that it's good to be almost home. And I spin along. Its fun to feel how happy the tires are when I do silly sharp turns. The garage door is up when arriving at my drive, and I coast right in. Opening the door to the house, I smell the first batch of chili for the season. It's a bit early, but smells great. The kids both have friends over, and are out back releasing there extra energy. I grab a shower no longer caring about missing today's Century. Instead, I'm ready to go back to work.

Thursday, July 16, 2009

Little Details make a custom frame

One of my current projects is a fast bike for someone who may yet try his hand at racing. I've been working with him to adjust his fit for some time now; and we're starting to see results. He's getting pretty quick on the bike, and can use much more handlebar drop than when we first started. That said, there's no way to get past the fact that he has an extreme body geometry where he needs about a 55CM seat-tube combined with 53CM top tube. And that's using a 9CM stem. You could say he's long in the leg.

He's a strong and heavy rider, so no wimpy frames for him. And it's going to be lugged steel. Now this presents some interesting issues. Generally, lug sets aren't available in multiple angles. So, framebuilders have to 'adjust' lugs to fit their designs.

For this frame, I've chosen Dazza's Slant Six (aka XL Compact) lugs with a Kirk Pacenti l
ugged Bottom Bracket. This combo is designed for use with a 2X Oversize tube set (in this case it'll be Columbus Life), with diameters as follows: DT=34.9mm, ST=31.8mm, TT=31.8mm. These fat tubes should keep everything plenty stiff. The Spirit chainstays will do their part to, running full size (i.e. without a taper) until the last 90mm. Most chainstays taper over the last 250-300mm - so this is a meaningful difference.

The Slant Six l
ugs create a frame with a modest slope (6 degrees) to the top tube, allowing for somewhat more standover clearance. Which is great. They are sized for a 1-1/8" steering tube - unlike most lugs which are sized for 1" steerers. Personally, I don't think that steerer diameter is very important for road bikes - but the rider wants a carbon fork and the larger steerer will leave us with more options.

So, back to the lugs, short top tube, and adjustments. The angles on the stock lower head tube lug, and on the BB between the seat and down tube, are about 3-1/2 degrees away from the plan on BikeCad. This means that it's time to adjust, in a fairly significant

I have some nice bending bars and quickly got the BB in shape. Naturally, the bent lug ports needs some hammering to make sure that they conformed to the shape of the tubes. The lo
wer head tube lug is a bit more difficult. It is somewhat like a bikini lug, in that there isn't much lug on the headtube, especially above the down tube. This means that its hard to clamp this part of the lug in place during the bending, and that this section doesn't have enough material to both fill the gap created by the bend and provide a good surface area on the head tube. That's where this piece comes in. I traced the top of the lug onto a cutoff piece of head tube, and then cut it out.

This will get braze
d into the corresponding section (see the second photo) of the lug, using brass filler. After some file work, it will fill the gap between the adjusted lug and the head tube. Some more filing on the outside of the lug will restore the outer shape (so that the shoreline of the lug doesn't become twice as thick).

The actual joint will be brazed with silver filler, hence it won't heat up enough to weaken the brass filler used to modify the lug. Cool concept but plenty of work.

That's it for this post. Cheers,b

Tuesday, July 07, 2009

They said what?

Its become clear to me that I shouldn't promise anything here - something always comes up to distract the dialog. So no promises, I think.

There many topics worthy of discussion, but what got me going tonight was this bit in Velo News

Now, I don't know about you, but I think big companies try to patent too much - especially broad concepts which they haven't invented or substantially developed. I think that Trek's 'Kamm Tail' fits this description. First of all, they are borrowing aerodynamic licks that have been employed repeatedly, for many years, and in a borad variety contexts - it's nothing new. And applying them to
a bicycle isn't nothing new.

In fact, I've been experimenting with similar tubing shapes on TT bikes for over a year. And I didn't get the idea from Trek or anyone else in the bike business. In fact, I've long wondered why the so-called aerodynamicsts of the bike industry weren't exploring these ideas all along - thus making me wonder how many of them really know anything about aerodynamics.

It should be noted that an 8:1 aspect ratio is not optimal (as described in the article) for head-on air pressue. 8:1 represents a 12.5% ratio, which is near the maximum for the size of airfoil created by a bicycle tube, and the speeds at which a bike travels. Airfoils have scaling issues, and the smaller the airfoil is, the greater the challenges presented. If we want to talk about an optimal aspect ratio for dead-ahead air pressure, then we should be looking to something in the 8-10% range, meaning somewhere between 10:1 and 12:1. These are significantly different form an 8:1 ratio.

Of course, it's not just aspect ratio, but also the shape and size of the object which determines how slippery the object is. A really big object with a very slippery shape and aspect ratio may have more drag than a smaller object which has a sticky shape and aspect ratio. Or it may not. So much more needs to be pinned down, apart from the 8:1 ratio. But lets explore this further.

First, a round tube isn't much better than a square tube regarding the drag produced as an air stream hits it. Historically, metal aero-tubes had a round front and tapered rear. The rear fairing did little to improve aerodynamics over a normal round tube (although it never stopped a manufacturer from touting aero advantages). The simple learning here is that the leading edge (front) of the tube is the most critical shape for achieving low aero drag.

Trek suggests that the tube's profile should/can be cut off just past its broadest point. So let's look at this with their 8:1 profile and the UCI limit of 3:1 size.

Usually the thickest part of a wing is about 1/3 back from the leading edge. On an 8:1 ratio tube, with a 1" thickness, the chord (front to back distance) would be 8". If we took only the front 1/3 of that shape, the actual (as opposed to effective) chord would be reduced to 2-2/3". This creates a tube with a 1:2-2/3 aspect ratio. Which keeps us in the game vis a vis UCI rules. So far, so good.

Now lets talk about the Kamm Tail. The concept of the Kamm Tail (or more correctly Kamm Back) is that an aerodynamic automobile not only has too little aerodynamic downforce for stability, but actually can achieve aerodynamic lift as speed. The Kamm Back creates a low pressure area behind the vehicle (drag) which helps to maintain directional stability. So, while it works in conjunction with aerodynamic drag reduction, it actually is using drag to keep a car traveling in the correct direction. Which is quite different from what Trek is describing.

This doesn't make Trek's shapes or concept wrong, but shows how marketing mis-uses technology to explain and sell products.

OK, so let's get back to this shape thing. It looks like Trek has 1/3 of an inch left according to UCI rules. Should they max out the 3:1 ratio?

Aerodynamics occur in 3 dimensions. Any aerodynamicist who forgets this is unlikely to be successful for long. For example, a round down tube has a salami slice like shape (sort of an oval) as it presents itself to the wind - because the tube isn't vertical.

Moreover, the airflow will chase the lowest pressure areas. As with a swept wing on a jet, this means that on the downtube, there will be airflow from the further forward points (near the head tube) towards the bottom bracket - all else being equal. Perhaps you've seen an airplane with tiplets on the ends of the wing - they often look like small rudders. On the wing, air travels similarly down its length. Air below the wing has higher pressure than on top - so when the flow of air reaches the end of the wing, the high pressure below wants to flip over the top and fill the low-pressure area. This reduces lift and causes drag. The tiplet is designed to interfere with the spill of high pressure are from the bottom to the top, thus improving lift and reducing drag.

The sides of the downtube are symmetrical. So, if the wind is from dead-ahead, then there won't be a high and low pressure side. Nonetheless, all else being equal, the air flow along the downtube will be slightly downhill - which further elongates the oval presented to the wind.

It's possible that a tube that has a actual 3:1 ratio, presents itself to the wind at some other ratio - for example, instead of being 3" deep aerodynamically, it might be 4" long. Now this is of no particular advantage to one builder or another - they will all have a chance to leverage this phenomena. But it does prepare us for the next step.

Unlike what Trek says, the optimal cutoff for the airfoil is not at its widest point. The airflow needs to be stablized into a path similar to that it would take if the rear of the foil had not been truncated. The path from the thickes part of a wing to it's trailing edge is not a straight path. But it is generally the straightest part of the outline of the airfoil. Let's say that when we truncate an airfoil, we still need 10% of the overall chord length to be located behind the widest point of the airfoil. In reality, how much wing is required behind the maximum thickness is a function of the wing size, shape, and airspeed - so we're speaking in hypotheticals here when we use the numbe 10%.

Given that the base airfoil is 8:1 (per Trek), the 10% extra would be 0.8". Now, measured using the actual tube profile, we have a chord lenght of 2.667". Adding 0.8" to this gives us a chord of 3.467" - or an aspect ratio of almost 3.5:1 - well above the UCI limits. But let's think this over again. Assuming that we present a foil to the wind that is 4" deep and 1" wide, for a tube that has an actual 3:1 profile: In this case we have more of the aerodynamic airfoil with which to play. Let's move the widest point of the tube forward, so that it is 2.667" behind the leading edge when measured along the path of air flow. That means that we have 1.333" left behind the widest point, and this 1.333" is 13% of the original air foils chord. If we only needed 10% of the overall chord behind the widest point, then we have another 3% to play with.

Let me point out that the numbers above are not meant to emulate the reality of Trek's tube shape. Rather, they give the reader some insight into the sorts of opportunities and tradeoffs which exist in designing a an efficiently shaped downtube.

Now, if we have 3% of the chord left to play with, what should we do with this? If the wind always came from the front, we would probably just use it to truncate the shape closer to the trailing edge. However, the wind rarely is truely from directly ahead. All the points of the compass have equal likelyhood has being the source of any wind, and while the bike's ground movement adds another vector necessary to calculate effective (as opposed to true) wind direction, it should be obvious that we will commonly contend with a wind that is not from dead ahead. Further componding things this challenge is the fact that a bike rarely travels in a truely straight line. Instead bike and rider are constantly adjusting inputs, causing the front end to continually move back and forth.

So what? Think of our truncated airfoil. With wind dead ahead, a sharp corner at the rear has little impact. But, turn the wind direction 15 degrees to one side, and now this corner becomes a drag riser. We want to smooth the transition of the air flowing around this edge. A rounder shape will work better in these circumstances. So, the extra 3/10" in chord length might be best used to transition air around the rear of the tube, when its coming from a direction other than straight ahead.

With the widest part of the tube being located 2.667" from the leading edge, along the actual airflow, where is it located in a 90 degree cross section of the tube? Well, its 2.667/4.0 from the front. Solving the math gives us a maximum width 2" from the front of the tube. Similarly, the end of the rearward taper is located 2.600" inches back, leaving 0.4" that can be shaped to help cross winds get around the rear of the tube. All of this has been down in the context of UCI 3:1 rule.

Naturally, this isn't the end of the story. For example, cross winds see an assymetric shape, even though the tube is symmetric from left to right. This begs the question regarding whether there might be benefits to having a tube which is symmetric from front to back? Also, there are aerodynamic tools which can manange how well the airflow follows the contour of the airfoil. Remember when we (arbitrarily) suggested that the tube shape needed to include 10% of the overall (8") chord located behind the widest point of our tube? Flow managing tools could reduce this number without limiting our drag reduction. If so, then we would have more chord length available (within UCI rules) by which to manage how cross winds wrap around the down tube. And, what if there is another place where we can truncate the chord of the tube without compromizing aerodynamics? There is such a place, but that's a story for a different day.

Be aware, Trek is sloppy in describing what they are doing, greeding in suggesting that they may patent their technology, most likely accurate in describing what they doing as an improvement, and not producing an impact that you can measure within margins of error in real practice. Yeah, we haven't touched on this yet, but you the rider are the source aero drag. Yes the bike contributes, but its contribution is small. Your position on the bike; what accessories (water bottle(s), computer, tool bag, etc.) and how carry them; these are the important aspects of aerodynamics. Until you have these nailed, don't worry about the shape of your frame tubes, or how well hidden your brake may be.

Until next time, Cheers!