Steeds of Roubaix: The Good, the Bad and the Ugly

As the National Capital Region's very own Paris-Roubaix slowly fades into memory and other challenges move to the fore of consciousness, I thought I'd keep the ball bouncing across the cobbles with a feature of this year's 3rd place bike. Granted, I was going to start a series of posts on the bikes of Roubaix anyhow, but now that mine has attained a podium spot, I figure I have even more reason to blather about it. I'm hoping other Tall Tree riders will follow suit and post on their set-ups to convey the spectrum of rides employed on Sunday. Its clear that some were good, but I'll leave it to the others to claim the Bad and Ugly. This post will be part Roubaix tech report and installment three of the series of writings I'm doing on my Steelwool Secteur 18.

Ottawa's longest running and most popular Spring Classics inspired sportif event, the Paris-Roubaix, named after the...Paris-Roubaix in France. One cannot fault the tireless organizers of the Ottawa Bicycle Club for borrowing the name. After all, sometimes this is what it takes to convey the spirit and flavour of an event; its easier to draw a direct line to the inspiration. Despite sharing a name, these two events are not really all that similar. France's Roubaix feaures twenty-odd sectors of cobblestones. In contrast, our Roubaix features none. Instead, about 90% of the route is dirt, mostly dirt road. While riders are spared the boneshaking abuse of the cobbles, we are treated to three offroad sectors that can only be described as trails. These sectors throw hazards at riders that pose the risk of both crashes and punctures, perhaps even sheared derailleurs and bent rims. However, these sectors are not taken at 50kph, unlike the cobbles.

Another key difference between the races is the elevation profile and distance. France's Roubaix is nearly completely flat, while out Roubaix features around 500m of rolling hills, some as steep as 15%. Then there is the distance, 260k versus 90. That's a massive difference, yet the tactics employed are not all that different. In both races many riders hope to finish, and many don't. So what sort of bikes are suited to our home grown Paris-Roubaix? Read on.

Regular readers of the blog will know that Tall Tree Cycles and Steelwool Bicycles co-owner, Will Ficner, and I worked together to design my first custom bike, the Secteur 18, last summer. The bike was to be the first Steelwool all-road bike, designed around 30c tires and fenders, mid reach caliper brakes, 80mm of bottom bracket drop, 72.5 degree head angle, 73 degree seat angle, and moderate trail. We indicated to Sam Wittingham, the builder, that this was to be a bike used for a variety of riding, including races like the Roubaix. Truth be told, I had designs on racing the bike in the amateur Paris-Roubaix, the real deal. The name, Secteur 18, pays homage to the infamous Forest of Arenburg sector of the Roubaix, the 'trench.' While it is not always the 18th from last sector in the race, that is the number it has most often held. The trench is hallowed ground, often decisive, always respected. Inspiring. The bike was delivered in September, mostly constructed from Zona tubing, a choice that came down to Sam, based on the information we provided about my size and riding style.

Last fall I didn't have a great deal of opportunity to put the Secteur 18 through the paces. The big events had come and gone, and sights were now set on cyclocross. I did end up racing the bike during one 'cross race, and was very happy with it. However, my general impression of the bike was that it was perhaps a tad stiff in the seat-tube. I was happy with the accuracy and torsional stiffness of the front end under moderate efforts, but felt a good bit of twisting under full sprints. This was no surprise to me, as it had not been designed with a bias toward sprinting. It was rather to be a bike for long rides with steady effort in the saddle. For this reason, it needed to be very efficient from the seated position. I felt it was perhaps a little stiff in the seat-tube to plane as much as I'd like. This observation was mostly based on the performance of my Pinarello cyclocross bike, constructed from standard diameter tubing (25.4tt, 28.6 dt and st). By the end of the fall I knew I had to put in a lot more riding on the bike to come to ground on the performance of the tubing. Ditto for geometry, as I was trying to figure out why the bike felt a little less stable in cross winds than my other bikes, and also exhibited a bit of wheel flop.

Fast forward to April. By this point I'd logged about 3000k on the Secteur 18, and had confirmed that it did indeed flop a bit more than I liked, and was a little less stable in cross winds than I liked too. But the question remained: how would it perform under race conditions? The Tour of the Battenkill was my first opportunity to discern how the bike responded to the surges and broad power output of a hilly mixed surface course, much like the routes we tend to ride.

Battenkill features numerous grinding climbs. My bike set-up was identical for both Battenkill and the Roubaix, save the gearing. I ran Mavic Ksyrium Elites shod in Brand Bois Cypres 30c tires and 28-43mm Continental tubes. Up front I used a 36/50 compact double crank with an 11-28 for Battenkill and 12-25 for the Roubaix. I always run Specialized Bar Phat under my tape, Yokozuna Scott-Mahauser compound brake pads in my mid-reach Shimano calipers, SRAM Force/Rival controls and drivetrain, 3T bar, stem, and post, Shimano XT pedals and a Ti San Marco Regal saddle. All in, the bike weights about 22lbs. This is definitely heavy for a road bike, but actually super light for a randonneur (if I were to set it up with a bar-bag I could call it that!). The low gearing for Battenkill was my typical set-up, but I didn't actually need the 28. The tire choice was larger than necessary, as the dirt roads were buff, but they are not slower than 28s anyhow. I experienced no negative handling traits during the race. THe bike was completely stable at 80kph dirt descents, and being in the pack most of the time cut down on potential for cross wind influence on the steering.

For the Roubaix, none of the climbs would require a lower gear than the 36x25, and in fact, I remained in my big ring for the whole race except two climbs. Compact gearing is good for that. The large slicks were ideal for the route, as I don't find tread worthwhile with so little trail, and no turns that are particularly difficult.

This is a real melange of compnents: Shimano pedals, SRAM Force crank, SRAM 1070 chain, Campagnolo Record front derailleur. It works flawlessly.

Even with the 30c Grand Bois, there is still ample room under the Shimano mid-reach brake and bi-plane fork crown.

Tire clearance in the back is pretty good with the 30s. I've run a 32 knobbie on this with good results, so the 30 has no trouble. With the fender installed I run a 28 for better clearance. Pauls Racer M brakes afford better clearance for fenders and 30s. This bike will be modified to accept the Racer Ms before too long.

Ok, so that's a run-down of the set-up. But for those uninitiated into the strange world of steel, the burning question might be just how a 22lb bike can be competitive in a race like the Roubaix? Isn't the extra weight detrimental?

In a word, no. What I am coming to better understand all the time is that the choice of frame-set, that is, when there is an actual choice, should come down to the specifics of the route in question. Here is a weak analogy: consider stage 12 of the 2009 Giro d'Italia, a time trial that featured a hair-raising descent down a very narrow road under a canopy of trees. Most, if not all, racers opted to use their regular road bikes over their TT bikes. That was the intelligent move, as handling was the priority. They had the option, and they exercised it.

During the Battenkill race, there were moments when I wished my bike was lighter. This does not mean I wished it was carbon, just lighter. This is entirely achievable in steel. The issue was the relentless climbs at 70 and 90k. Since my seat-tube is a little stiff, in my estimation, the seated powering I was doing was not entirely rewarded with flex side to side, what is describes as planing. What I should have done was stand a bit more often to keep from hurting my legs. That seems to be the best approach on a bike that's stiff in this way. Bikes that are too stiff hurt the legs more than bikes that plane well. My Pinarello confirms this. So the roads were smooth enough to mix standing and seated pedaling, but I fooled myself into thinking I was better off 'conserving' in the saddle. Wrong, I should have mixed it up.

In contrast to the Battenkill route, the Roubaix features numerous loose sections, as well as full on trail. The climbs are neither as steep or long. A resilient frame and fork will translate into greater efficiency over this route to a degree that exceeds Battenkill. The climbs are not long enough to bog me down in the saddle, so the seat-tube stiffness is less of a concern. On the other hand, accelerations are a significant concern. There are normally numerous instances in this race where you have to respond quickly to attacks. The less than optimal flex in my frame for long steady rides becomes an asset on this course, as the bike responds better to jumps than a softer frame could. Bear in mind here that I am speaking from a very 'picky' point of view. That's the primary reason I am involved in the R&D for Steelwool. Not all riders will ever care to think about this stuff or even be able to discern it. Its up to us to figure out what works.

The degree to which a frame planes has a great deal to do with the rider's pedaling style, including both their power curve and rate of cadence. When I was alone trying to chase after flatting, I got into a very good rhythm on the bike. Even though my lower back became intensely painful, and my legs and lungs were nearly maxed out, I did not have any desire to be on a lighter, stiffer bike. My bike was working flawlessly, rewarding my power. So the question that comes to mind is whether this bike necessarily performs better under higher power outputs? I believe it does. The wind was blowing, and I was putting in enough power to hurt my back from the bracing, so I think I was getting more of a planing effect out of the bike than while climbing at Battenkill. The bike was in its element. So we are talking about matching the flex of the bike to the terrain AND, importantly, the amount of power output expected. A randonneur bike would not likely be ridden this hard, so it probably shouldn't be as stiff.

When I attacked into and through the last sector, the Sugarbush, the bike was steady and predictable, responding well to my power output. The low bb instilled confidence, as always, and we came out with enough juice left to keep the pressure on to the line. I was completely happy with the bike's performance.

I'm continuously learning about the nuances of frame design, material selection and spec. Jan Heine's writing in Bicycle Quarterly is contributing greatly to this learning process, more than any other source by far. Padraig of Red Kite Prayer is the other writer contributing to this ongoing conversation about frame design. His observations about handling are tinged by his penchant for high speed descents, a nice compliment to Jan's longer distance bent.

While there is much to learn from here, I'll offer a couple conclusions I've come to with regard to tubing choice. These are not set in stone, just the ideas I'm marinating right now.

1. The first step is determining how and where you ride. One rider might ride the Gatineau Parkway and never get out of the saddle. S/he would not need or want the same sort of tubing as a rider who spends a lot of time out of the saddle, and goes for the sprints. The roads are relatively smooth, so comfort would not need to be addressed to the same degree as a bike being ridden around Wakefield, for example.
2. The next step is determining how much emphasis you want to put on comfort/efficiency. Do you want the bike to feel fast or be fast? There is a difference. 23c tires feel fast, but they are usually slower than 25s.
3. The third step is deciding where you want your handling bias to fall: should the bike be optimized for moderate speeds, or should all out speed on descents be a priority? If you like to plummet descents at maximum velocity, you might want to forego a bit of pedaling efficiency in favour of front end accuracy (read, stiffness). So while a 25.5 top-tube might provide better suspension, it might not feel precise enough in the turns. This is a tough one to get to the bottom of, as suspension and precision cannot be taken separately. Companies like Specialized are using oversized head-tubes and steerers and focusing the flex in the fork blades like suspension on mtbs. This can be achieved with steel forks too, but whether oversizing the head-tube and isolating flex to the fork would be optimal is unknown to me.

I've come to the point where I am relatively certain that more fork rake will reduce if not eliminate the flop and cross-wind instability I am presently experiencing (see the latest edition of the BQ for an article on cross-wind instability). This conclusion is based on my theoretical understanding of trail, which is mostly informed by the tests and calculations conducted by Jan Heine at Bicycle Quarterly. I hope to have a new fork build in the near future to address these quirks, all part of the process.

So what did you ride for the Roubaix, and what did you like and/or dislike about your set-up? Would you consider a high quality steel bike as an addition to your quiver (if you don't already have one)? What would you want to use it for? Perhaps we'll hear from David Bilenky, proud owner of a special aluminum Alan cross bike that is reputed to exhibit very lively sensations. Don't feel the required to geek out to the degree I do here. All comments, no matter how technical, are welcome. Its really all about the sensations...