Wednesday, August 26, 2009

District Carbon: which district?

Well, since I got an eye infection around a month ago, I've been feeling up and down, but mostly down. Progress after antibacterial drops from the optometrist, but since then, I've been feeling under the weather again. I needed to pass on the University Road Race, which is a favorite race of mine. My power numbers suggested I should be able to hang on that suffer-fest, and placing at University probably means a top-15 result, which is good for me.

So when I'm not riding, my mind invariably drifts to equipment. This bike in particular catches my attention:

Trek District Carbon

First, the rational self: this thing makes no sense at all. Urban bikes get ridden through potholes, leaned against posts, stacked on bike racks in overcrowded Caltrain cars, wrapped with chains outside Rainbow grocery, and shouldn't stand out in a crowd of possible theft targets.

This bit of bike boutiquery begs to be babied: one serious scratch or crack on carbon and it needs to be tossed, or else sent to Craig Calfee for expensive repairs. For urban, steel is the real choice. Even a direct hit from a car and all it takes is a few choice mallet whacks from a good frame builder and it's ready to go again, as I experienced first hand with my Ritchey Breakaway (thank you Mr. Mikkelsen!).

The one dose of sense: the belt drive. These things have been proven to be efficient even compared to a clean chain, and chains often are less than photo-ready clean. Furthermore, while I wouldn't want to rub a dirty rubber belt against my pants leg, it certainly beats a direct hit with an oiled chain, even a properly oiled one where the excess has been wiped away.

So with all the silly aspects, why do I keep staring at the thing? Excessive, sure, but excess directed at something other than winning a UCI-sanctioned race. Excess at riding to work and home again. Excess in a bike instead of a car.

And while it's obviously not an all-around racing bike, it would work nicely for certain hillclimbs. For example, I used a 2:1 gear ratio in my recent Old La Honda PR. This same ratio is available as a 50/25. A range of ratios are possible, with the adjustable drop-outs handling belt tension.

Plus, I love the black-on-black look.

Will I buy one? No way. My rational self has too much sway. But even that rational self admits it's pretty cool.

Any less excessive?

Tuesday, August 18, 2009

cost-mass tradeoff on 2010 groups

I did a few minutes of data mining of the Competitive Cyclist road group configuration page, collecting datapoints in each of the three available full-group manufactures, SRAM, Shimano, and Campagnolo. The page offers convenient totals on mass and price. Their mass numbers aren't always great, and certainly even local shops often beat Competitive Cyclist prices (especially with bike coalition or racing club discounts), but the idea is to compare the relative cost of different groups, and it is assumed discounts are available for each.

First, on this comparison, it's clear Shimano is uncompetitive. SRAM comes out the best, with Campy a close second.

But second, sticking with complete groups, the jump to the highest end comes with a hefty price tag. Each is over $5/gram. Among the second-tier groups, Force shows itself as the real weight-weenie bargain. With performance essentially equivalent to Red (with exception of the ceramic bearings, which might be important), it sacrifices less than 150 grams.

And it's easy to make up that weight for less money than the Red upgrade. For example, I show how swapping the Force crankset for a Vumaquad comes out lighter than Red at less money, even though the Vumaquad is a boutique-level crankset. The Lightning crankset could yield the same mass as the VumaQuad at around $350 less. However, I was sticking with parts Competitive Cyclist sells. In any case, the Lightning gives up some stiffness to the SRAM cranksets, and while crank stiffness isn't a big deal for my power and mass, the stiffer Vumaquad makes my point better.

Competitive Cyclist price-versus-mass comparison

Saturday, August 15, 2009

House of Tacks

My last attempt at the House of Pain ended in a minefield of puncture vine, which took out both tires less than a mile after I joined the proto-race. Today, I was back, ready to check off another goal in my riding. I'd already PR'ed Old La Honda, I'd already stuck with the Valley Ride to the sprint. Now I needed to stick with a weekender: Spectrum or the HOP.

After that first attempt, as I was in Pegasus purchasing the two tubes I needed to replace the pair which had been perforated well beyond the patchable threshold, another customer in the shop mentioned puncture vine was a common problem on the roads around Danville. But he'd also heard of tacks being spread in the bike lane or shoulder. I didn't think much of it.

So today I rolled out of Walnut Creek BART, heading south. As I crossed under 680 on Main Street, a car honked aggressively as it passed, the adjacent lane completely empty. After a brief Spike Bike fantasy, I put it behind me.

I continued southward on Danville Boulevard. The HOP starts at a park and ride, which I'd just passed by, but by the time I'd realized my error I was far enough on my way I figured meeting them at the Danville Peet's was a better idea, anyway. Surely Peet's is a nicer place to wait than an isolated parking lot. And I had plenty of time, I decided, as I cruised south in the bike lane.

But then that unmistakable feeling: the subtle loss of control of a rear flat. A quick glance back verified what I knew: the rear tire was almost completely empty.

Once after I'd swapped a punctured tube and completed my descent of Old La Honda where Jobst Brandt and his ride companion were waiting for me (I'd joined them on the road), he'd asked pointedly, "what caused the flat?" It had been something of a test of intelligence and inquisitiveness that a rider should always determine what had caused a leak, preferably with Holmesian detail. In this instance the brilliance of Sherlock was hardly required: a silver thumbtack was firmly embedded in my tire.

Based on the time the HOP pack had rolled through the last time I'd been here, I figured I was good on time, so I decided to take the easier but slower path of patching the tube (requiring moving only the small punctured segment) than to replace it and patch the old tube as my new spare. Patching requires waiting five timed minutes for the glue to dry (to maximize adhesion), but I can stretch. Total elapsed for the fix: 12 minutes.

During the last few minutes of these minutes, as I was inflating my tire with my small hand pump, a few groups of fast-looking riders, maybe six each, passed. No way either of these were big enough for HOP, I decided. This was a very popular route for cyclists: the principal north-south route on the flat-lands west of Mount Diablo. When I'd finished I glanced at my watch: 9:07. Surely they hadn't passed yet, given they'd only arrived at Railroad Road last time after 9:20.

So I continued onward. I paused at a group of three riders on the shoulder, two of whom appeared to have just finished a flat-repair job themselves. Then I came across yet another rider, this one fully in the act. "Tack?" I asked, to which he responded in the positive.

Finally at Peet's, where I arrive at 9:12 am: a small group were also waiting for the HOP, including Mike whom I knew from the Memorial Day Tour. One of them had also taken a tack. I encountered another at Peet's with the same story.

Okay, dude, whomever you are, you won. Those tacks you dumped in the bike lane caused a bunch of riders grief on an otherwise wonderful morning for a ride. You're so clever.

The HOP ride never did show. Or maybe those small groups had been it? I suspected perhaps the tacks had deterred them to an alternate route. So instead, after waiting until 9:40 am, then stopping at a nearby shop (not Pegasus, which was still closed) to borrow a floor pump, I set off to climb the south side of Diablo. It proved to be perfect conditions for the climb: the growing heat of the day offsetting the effect of the increasing altitude. And I had some nice encounters at the summit. One, fellow Mouse Ben, told me the HOP ride has been rolling out faster than they used to. So hmmm.... And I got to see my first example of a 110 mm BCD Cinqo. Cool.

So a nice day after all. But next time I'll likely stop at the Park and Ride.

Today I saw my first specimen of Quarq's new 110mm BCD Cinqo: finally supporting a "compact" crank for those of us without a ProTour-level power/mass ratio. The lighter Lightning version should be available within two months.

Friday, August 7, 2009

Hanging with the Valley Ride

I hadn't yet done a Valley Ride this year.

The Valley Ride is in the same category as the Tue Noon Ride and the Spectrum Ride: a race simulation which if, if you can survive, is a good indication of race readiness. Other than my extended climbing (I recently PR'ed Old La Honda), I've had precious little indication of race readiness this year. Certainly I've shown none at races. The frenzied, lactic acid saturated hammer-mania typical of the shot climbs often encountered in these rides or races is a much different beast than an aerobically dominated climb like Old La Honda Road.

And while I'd not done the Valley Ride, my several attempts this year at either the Tue or even the somewhat tamer Thu Nooner had all been failures. Refusing to take short-cuts taken by 2/3 of the packs at Canada College or at Corte Madera, I'd been gapped by the leaders, who are rejoined at the bottom of the associated descents by the low-roaders and are then gone for good. "Unfair?" Yes or no, if I'm fit to race, I should be able to stay with the leaders.

Unlike the Tue-Thu Noon Ride, the Valley Ride has a regroup: after the descent of Kings Mountain Road following the climb of Huddart Park. Still, in the spirit of competition, I made a maximal effort on that climb: 301 watts for 5 minutes, 17 watts off what I'd managed on a 4.5 minute interval in July and a whopping 32 watts below what I'd averaged in a 4.1 minute interval in 2008 (I'd done a lot more intervals in this time range last year), but this time I had legs that had already been run through the "espresso" setting of the grinder. After having done sitting-standing intervals on Old La Honda and W Alpine Road earlier in the day, I definitely wasn't working with a full tank of glycogen. So I was pleased.

Gentle approach to Huddart, then crossing the gate and beginning the climb. X-axis is ride time in minutes (including riding to start), y-axis is watts w/ 5-second smoothing. Seriously now: riding at the back is like an interval session before the real interval.

In comparison to the approach, the climb was quite civilized. I did a decent job of smoothing out the power spike which invariably comes at the steep section: it feels ridiculously slow at 350 watts, but "maintaining momentum" as most riders do comes with a high rate of interest. I passed several others when the grade lessened since my power didn't fade as much as theirs did. I tried to keep my cadence up on the final straight, and as a result I was able to maintain my power to the finish of the climb, despite hurting legs.

Where I got into trouble was on the climb of Sand Hill which follows. This climb is clearly shorter and less steep, but the shorter length simply implies riders attack it even harder. I was systematically losing position while at around 6 W/kg: my legs simply had nothing more to give. Fortunately, while my legs were done, I had position to give away at that point. There were still a few wheels left to follow as we crested the top. I'd avoided the wake of destruction left behind on Sand Hill....

As the group winded up for the final sprint, I was finally gapped for good. My legs had had enough. But this was well into the final kilometer of the race, I mean ride, and so I'd basically survived, which had been my goal, despite my hard ride that morning.

Good, right?

No. My pack position through much of the ride had been wretched. Too far back, sometimes all the way at the back. Draft is draft, right? Again no. At the back acceleration-deceleration cycles are amplified. Furthermore, on climbs there's no margin for error or weakness. It was only pack position which saved me on Sand Hill. Had I been at the back, hasta la vista, Dan. I'd made considerable efforts on Mountain Home, for example, which I would not have had to make had I been further up.

I get nervous on these rides, especially if I haven't been a regular. All that separates me from the operating room is some guy I don't know losing concentration and overlapping a wheel or riding onto a dirt shoulder. Or worse, some aggro driver who thinks the roads aren't the place for pseudo-race group rides, and is going to enforce his own brand of righteous intimidation.

All true, but the fact is life is never without risk, and anecdotal analysis suggests these rides, despite the craziness bordering on chaos, are fairly safe. The vast majority of tragedies on the roads have been individuals or small groups. Even minor injury crashes are relatively rare, typically limited incidents. The ratio of danger perception to danger evidence is on the extreme high side. There's some safety in numbers. Even Critical Mass, which is designed specifically to annoy people, and is populated by riders with essentially no pack skill, has a low race of injury.

Linda Jackson once said to me you race like you train. You're not going to ride like a wuss in training rides then be up there in the action in a road race. I need to get out there and do better.

But at least I met my primary goal this week.

Sunday, August 2, 2009

Powertap analysis of back-to-back Old La Hondas

On 22 July I did Old La Honda twice. The first time I rode it at a steady, hard tempo. The second time doing 60 second intervals each four minutes, zone 3 between intervals. So this isn't a direct experiment. Yet the interesting thing was my total work reported by the Powertap differed in the two cases:

min from ride start
km into climb
min into climb

The difference in work is 3.65%. But it makes sense work should be less if I was riding slower, right? Less wind resistance.

So I converted each data set into power versus normalized position, and plotted the difference in retarding force (power / speed) versus the difference in the square of speed. The standard model is that in the absence of headwind (and the winds were light that day, especially on sheltered Old La Honda Road's eastern slope) the wind resistance force is proportional to the square of speed.

A linear regression yielded a quite reasonable value for CD A of 0.36 meters². The curious thing was the intercept: 1.6 newtons force, on average, less in the second run after accounting for the difference in bike speed.

It's possible I sweated away mass between the two reps. But suppose I sweated away 500 grams, a stretch. To account for this force difference, the total mass difference should have been more like 2.1 kg, assuming a 7.3% grade with CRR = 0.5%, leaving an extra 1.6 kg effective weight reduction unaccounted for.

Robert Chung recommended I calculate altitude from power using my derived CD A and estimated mass. Assuming again CRR = 0.5%, the model does a quite nice job of reproducing the actual hill profile. The following plot shows the derived profiles (left axis) with the difference (right axis). Curious: in the middle of the climb suddenly the second rep begins deviating linearly from the first rep. Powertap flatulance, perhaps? The plot also shows Lucas Pereira's data, which is slightly longer than either of my two runs: perhaps he took care not to cut any corners.

Powertaps are rated to an accuracy of 2% or so. Yet I'd expect this would be over worst case conditions: for example comparing two runs on different days at different temperatures. Weather conditions were stable both near the top of hill and in the valley. But maybe there was a subtle wind shift during the second climb. It doesn't take much to swing power by a few percent. It would have been nice to have an iBike...

Saturday, August 1, 2009

position check

An eye infection has me sleeping poorly and feeling generally crappy. In these situations, I prefer to take a few days off from worrying about training (except for stretching), get plenty of rest, and focus on getting better.

A good chance for a sanity check on my position.

CyclingNews did a nice article about the Saxo Bank Squad team getting fitted by Andy Pruitt. Specialized, who sponsors the team, likes to represent itself as experts in ergonomics, and so really play up their relationship with Andy Pruitt, a respected bike fitter who wrote a nice book on the subject (not as "complete" as the title would imply, but definitely worth the read).

In the cyclingnews story, there was this side shot of Frank Schleck, showing a side view. I'll take this as an example of, for Frank, a relatively exemplary position.

Frank Schleck fitted by Andy Pruitt (CyclingNews).

So, how do I compare?

Leaning the bike against the wall (not the best: a trainer would be better) I took photos of myself in various positions, extracting angles with Gimp.

Well, a few observations:
  1. My leg angle of 33.5 degrees is within the range recommended by Pruitt, which is 25 to 35 degrees, but near the upper end. I tend to ride with my toe down, though: 26.8 degrees in the photo.
  2. My knee is slightly more forward than Pruitt suggests. But rotating the body forward shouldn't make much difference as long as all the angles are good. I like being in a slightly more forward position. Compensating for an extra cm on the stem length? No big deal in any case. Christopher at PK Cycling established my present set-back, and I trust him.
  3. Frank appears to be a lot lower than I am (he's on the tops and he's already fairly low). Well, he's a pro and I'm not... I clearly am more aero in the drops than on the hoods. Were I to lower my bars I wouldn't be comfortable in the drops, and would spend more time on the tops and less on the hoods. This position seems about as aggressive as I can go and still be comfortable.
  4. li>My head position looks okay. I've been focusing a lot on relaxing my shoulders, even shrugging a bit (not doing that here), and letting my head drop as I look up to see the road. I had a tendency to ride with my face perpendicular to my direction of motion, which increases wind resistance.

Conclusion: nothing seems terribly broken here.