D2R2 2012

The Story:

  As the day approached, my nerves were on edge. Last year CP, GS and I had ridden the full 180k D2R2, and the second half was very hard for me. I gave up on the Corn Maze then, a short, delightful cool-down on an optional spur past the finish, but which I think is officially a part of our 180km route. One could build an excuse around hurricane Irene which had been drizzling on us for the last 10mi, and which would submerge the dining and parking areas under a foot of mud by morning. But really, last year's ride simply beat me and I couldn't wait to get off the bike as soon as the finish was in sight.
  Two days before the start, the forecast mentioned thunderstorms and I told another rider that if that threat persisted I'd drop out. Completing the full ride would establish my new longest ride ever, and since most of the miles are on dirt roads, a storm ride sounded too hard.
  The morning of the ride, I woke naturally, 15 minutes before my 2:30 alarm, after some overnight tossing at the sound of thunderstorms. I thought again of CP who was in a tent at the start sleeping through this weather. A quick kale omelette and an nearly two hours of driving through drizzle got me to the start a little ahead of my plan, where I met CP for some food. I was feeling pretty good and decided to ride with CP until I fell off his pace. We had plenty of time to get ready for the 6:00 start time, and CP had a chat with Sandy, the dirt road guru and fan of covered bridges who designed the various routes of the event. I would learn through the day that CP knows many of the people here. The drizzle was tapering and Sandy predicted that in an hour the weather would clear. We're off!
  Riding through Deerfield, MA at the start, we're on paved roads. CP immediately met Ca, a friend and fellow bike racer, and we became a group of three. A fourth joined the back and the conversation picked up. Ca told me about her ride last year with a friend who went on to win an important race the week after D2R2. We took a left over a long bridge which I didn't recognize, but our fourth said he was sure is correct. I tried to keep up with my cue sheet, listing the turns and distances, because D2R2 has almost no signs on the road. But inevitably I lost track as Ca and I traded stories, until my odometer said we were past the point of a T intersection I didn't see. By the time we backtracked to the correct trail, we'd enjoyed a 6.5 mile detour on relatively flat pavement. But we also landed behind a large number of riders who started after us but kept to the route.
  On the first hard climb, Ca pulled ahead of me and that was the last I saw of her until dinner after the finish.
  CP and I snaked up and down beautiful roads through forests and along farmland as the sky gradually lightened. While there were a couple wet areas on the gravel roads, I noticed that the moisture bound them together everywhere else so the dust and shifting stones of past years were gone. Climbing traction was perfect throughout the day.

  After the famous "bobsled run" descent and over a short bridge, the cue sheet said hard right, and I vaguely remembered it from past years, but a pair of Jersey barriers closed that road. Another rider suggested a detour and I was able to confirm it with my GPS. Just after rejoining the cue route, we came upon a wire strung across the road and a small herd of cattle crossing just past it. A farmer was moving them to pasture and had a temporary fence system to guide them over the pavement. After they all crossed he took down the wires and we resumed.
  After a water stop and a few more hills, we found the base of Archambo Rd. This may be the most famous tiny dirt road around, because it boasts a 27% climb, steepest grade of D2R2. I've learned a couple bits of physics about this slope. Even pro riders cannot start a bike on this, so once a foot goes down the rider walks to the top, as does everyone stuck behind him or her. Second, the rider must stay in the saddle or else the rear will slip and a foot goes down. Loose gravel will also cause a slip, and my first attempt 2 years ago scraped up both of my elbows that way. CP and I rode without other riders for miles after the water stop, guessing the clump we saw got stuck in the water line they had created there. But they swept along with us a mile or two before Archambo, so they hit the the steep ahead of us and all were on foot by the one-quarter mark. I dropped into my ultra-low gear and managed to stay on the saddle, but a clot of walkers blocked all of the road just ahead. I tried to say "let me through," but it sounded like a soft mumble mixed with my gasping for breath. CP came up behind me and called out properly, opening a path for us and he accelerated through the gap. That was the most impressive display of strength I saw all day.
  In past years I walked the last half of Archambo, but the solidly wet gravel and my lower gearing got me all the way up. I might have given up at around 2/3 of the way, but a pair of cheerleaders there with ersatz trumpets encouraged us to stay with it. Too soon after that came an endless and steep climb that takes it out of me every year. Two years ago a cheerful stranger in a wool jersey told me that climb is the worst of the whole ride in his opinion, though most give Patten Hill that title. I now agree with him.
  Then the difficulty moderated and I was enjoying the break when I recognized  DB, a friend who was on a different D2R2 route with a group of other friends. I hadn't realized we were near lunch, but that's where all the routes converge, beside a red covered bridge in Vermont. We all rode together for the rest of the way, parked our bikes and enjoyed the unusually nice meal there.
  The long climb immediately after lunch felt unnecessary, but a chocolate-covered espresso bean and another story with CP got us to the top and we were back into the ride. After a while there's a segment that makes me laugh, where we divert off a major road, go through a covered bridge then over two other bridges and rejoin the same road - an obvious longcut. But I'd forgotten that we almost encircle a beautiful old cemetery between the bridges, and I was moved to connect with those stones. And I did it in pigeon sign language in case CP was watching, and because silence felt right. I pointed to my chest, then to the headstones, then held up four fingers - one for each brief decade (max) which separates us.

  This route is the only place I've ever ridden such long and tiring descents that I want to rest before bottom. Picture endless swooping gravel, where the rider must be vigilant for rock, ruts and potholes. In fact, there were places where we passed groups stopped to fix flats, lining both sides of the road! Yet the most fearsome of these is reserved for the end of the ride, which many call the 'mountain bike descent'.
  But first we arrived at the base of Patten Hill, site of misery last year. One school of thought is this beast should be ridden as three individual climbs, separated by spans of less-steep climbing. But this year my gearing got me up the first (paved) section, and the damp gravel got me up the rest. Nearly to the water stop at the top, my homemade aluminum cue sheet holder broke off, certainly a victim of metal fatigue on the dirt descents. It fit nicely in my jersey pocket for the rest of the ride, and I felt lucky it didn't snap off at much higher speed.
  Several slices of watermelon later, we were into the last segment of the ride, well ahead of last year's pace and in line with CP's informal goal for the finish. Early in the 'mountain bike descent' I built up confidence in bunny-hopping rocks and small erosion trenches, and took small jumps off the tips of boulders sticking through the surface. I decided I could hop anything there and eased off the brakes, swiftly passing a more cautious rider. Then a trick of the light hid the scale of a large erosion cut until I was nearly upon it, so my hop was late and underpowered. I managed to get the front wheel onto the road surface but the rear fell way short - ka-toom! After a moment's recovery, a roar escaped me, reasserting dominance over the road, and I finished the descent without further trouble.
  Back on the pavement of Deerfield town, CP dropped down with his forearms on the bar, casual time trial style. I asked if he wanted to push to the finish and he said yes, shooting for an even 12 hours wall time. We missed that target by 3 minutes, then looked for our friends who had already finished dinner and gone home. Then we tacked on the corn maze which I found I could almost ride from memory of two years ago. And we were done.

The Stats:

  Miles: 121 horizontal, 3 vertical.
  Hours: 10:25 moving, ~12:20 wall. (11.6 MPH moving avg)

Technical Jargon:

  This year I brought full mountain bike gearing, with a 22:34 low gear. In past years I kept my road crank, yielding 26:34, but I needed lower to finish in relative comfort. Last winter it struck me that if I used the parts my MTB is currently wearing out, the long-term cost of the drivetrain for this ride is almost zero - it's all spare parts for the MTB. Top gear is 44:11 which was plenty. Changing over all this (twice!) for one ride is a lot of overhead for one ride, since it requires changing out the bottom bracket, front derailleur, chain guide and all the rest. I wouldn't do it for any other event.
  I've long been praising my frame for versatility, an Independent Fabrication  Club Racer built around Shimano long-reach caliper brakes. It can be stripped down to a competent road configuration or can carry full SKS fenders over wider tires for winter commuting in reasonably good weather. But it only barely fits the 32mm Vittoria Cross EVO XG tires I used for this ride. I was sure I couldn't get fenders over those tires, until I mounted them the day before the ride, listening to the radio forecast of possible thunderstorms, and noticed there was some room under the brake arms. The fenders rubbed the tires badly at first, but after some persuasion with a heat gun and other adjustments I was happy with them and that's what I rode. At almost the last minute I noticed the front derailleur arm would crash into the fender in low gear, so I shifted into that position and got out the heat gun again, while pushing the fender forward over the derailleur. After it cooled and I let go, the fender sprang back a little to give the derailleur just enough clearance, and the fender retained a perfect bend to accept it. Custom fit!
  Those tires are the same ones I've used for all 4 D2R2s I've ridden, plus about two prep rides each year. At this point the center knobs of the rear tire are nearly worn flat, but I think that was perfect for this year's wet conditions. I never needed rear knobs, and the remaining thin smooth contact strip was quite efficient. Knobs remain on the sides and front center, all helpful for cornering and stopping. I ran 72 PSI rear, 60 front, slightly higher than last year because of my fear of pinch flats. But standing pedal work felt bouncy enough that I wondered several times whether I was loosing air and should add some. Even after the ka-toom, the tires held air perfectly, so it's good that I never did add any air. I really like all the Vittoria tires I've tried. Count me lucky they don't make mountain bike tires, because I go through those faster than anything else and Vittoria prices would be rough. (And if they do make MTB tires, don't tell me!)

* Top picture is by CP, impressive camerawork while climbing! Foreground rider 448 is your humble scribe.

Spoke Tools

The first phase of wheelbuilding can use a screwdriver-like tool to tighten the spoke nipples, and this is much faster than the later phase using a spoke tool. The classic screwdriver tool looks like a regular screwdriver bent into a right-angle Z. I have no doubt that a seasoned builder could use such a tool with admirable efficiency, but as an occasional builder I haven't tried to find one.
Instead, I ground down a standard hex screwdriver bit, as shown. I used the edge of the grinding wheel to leave a pin in the center, which fits into the spoke hole in the center of the nipple. This works far better than a standard screwdriver bit because it keeps the bit centered and allows the bit to hold the nipple fairly well. (This shape is borrowed from the classic tool.)
I like this bit a lot. Its small size makes it a nice handle for inserting the nip and engaging the spoke, then a finger-roll or two engage enough threads to move on to the next. I can't think of a tool design that would make this process faster.
Later, after all spokes are installed, I use the bit to perform initial tensioning, until I need more torque than fingertips can provide. Then I put the bit into a small ratchet driver and work until that too becomes slow. Then the rim strip goes on for safety and I switch to the standard spoke tool which works on the flats exposed next to the spoke (and visible in a finished wheel).

Spoke Lengths

It finally happened: I laced a wheel with the wrong length spokes. I noticed just after lacing, before any time spent tightening, but still a costly mistake in time and money.
The main culprit is the database in the on-line calculator tool I found. I believe it thinks an M756 rear hub is centerlock, but it has the large flanges that come with an ISO rotor mount. I share the blame, though, because I didn't check the dimensions before I bought the spokes.
In the good old days, say around 2000, there was a functioning on-line community in rec.tech.bicycles which put together accurate and complete tools for this as contributions to the world. These days profit-seeking seems to have destroyed those tools and replaced them with hit-and-miss or deliberately incomplete tools.

I bought the first set of spokes from an LBS which cuts and threads spokes to length. While this ensures the length is correct, the cut ends are very sharp and the threads less clean than the factory rolled ones. They're also more expensive, perhaps due to the labor of cutting.
At the second shop, I asked them to check my revised calculations. This is the most modern shop I deal with, in the sense that I get good response to email. Their calculation was even more wrong than my original set. This time I think the rim size was incorrect, though the lookup matched the sticker on the rim itself! I got a look at the calculation process at this shop, too. For the last decade or so, I've used spreadsheets or web calculators, and electronic databases of dimensions for the hubs and rims. Here there was a programmable Casio calculator taped and zip-tied to the inside cover of a binder holding pages of dimensions. (I suspect much of these are out-of-date by now.)

In the end I re-learned the old saw, measure twice, cut once.

Mystery of Shimano Tall Flanges

I really like Shimano hubs. They've served me well for over a decade, through all sorts of bad weather and other abuse. I have looked to them for my relatively recent interest in mountain biking (MTB).

(Well, that Nexus internal hub wasn't very reliable, but the freehubs are all I'm interested in these days, and they've all been very good.)

I built up Bumblebee, my current MTB, using their XTR hubs. I like everything about these except one thing, their low flanges. The XTR flanges look like those of a road hub, which is nice for a lightweight hub. However, I think it's bad for a reliable MTB wheel, because both front and back disc-braked MTB wheels run much higher peak torques than (rim-braked) road wheels ever do. For example, the chain drives at least 50% higher torque in an MTB than a road rear wheel: just compare the low gear ratios. And my vote for highest torque in any bike wheel is a disc front.

Torque is conveyed through the spoke web by differential tension: the pull side spoke tension increases and the push side decreases, both to exactly balance the torque between hub and rim. The taller the flange, the smaller the tension changes because they operate over a larger lever. (The most extreme case of this would be if someone built a fully radial rear wheel. With no lever arm at all, the hub would be pulled apart by any significant torque.) In other words, a low-flange MTB hub requires a very good wheel build, and is still more fragile than a less carefully built wheel with a high-flange hub.

The frustrating thing about the XTR hub flanges is that I have an XT rear hub with tall flanges. It's lower-end, weighing almost twice as much as the XTR. But why wouldn't Shimano offer a tall-flange XTR when there is one in XT?

First I noticed that Shimano's centerlock hubs use low flanges while their ISO 6-bolt hubs have high flanges. I tried to turn this into a demographics situation - maybe ISO users are thought to be poor wheel builders?

It hit me the other day riding in: the receptacles for the 6 ISO bolts would cover the spoke holes of a low-flange hub, making the wheel very difficult to build! So ISO hubs need high flanges for the builder's sake, while centerlock ones do not.

My current plan is to use the XTR hubs and just keep after the wheel tension. I think this will work for me - though I wouldn't recommend it for someone doing big drops or very fast descents. And avoid thin spokes! The DT Revolutions I sprinkled in the first build proved to be a real problem here, and I've since replaced them all in the rear wheel.

Zipper Insertion

While most of my favorite homemade or modified bike projects come off the milling machine, many are done with a sewing machine instead. My fabric projects fall mostly into two categories, adding reflective tape to road outerwear or improving cool/cold weather clothing. This project is in the latter camp.

I've long enjoyed Nike's running shirts as road undershirts or mountain bike shirts. They make a long-sleeve running shirt that I really like, featuring excellent fabric and visibility. The bright orange is great for hunting season, and it has sleeve reflectors that I'd like to see on my road clothes. The only drawback is that I heat up a lot while riding, and Nike's design has no zipper at all. My favorite road jerseys have full zippers so they can be opened completely when needed.

I decided to add a full zipper to this shirt. I have limited sewing experience and am self-taught since a short class in 7th grade. This project was somewhat tricky because the Nike fabric is a very stretchy synthetic, while basic sewing technique assumes rigid materials that can be ironed, such as cotton. I've been developing some techniques for this over the years, and I have access to a rather good sewing machine for difficult fabrics.

Finding the zipper proved harder than I'd expected. A coarse, jacket-style zipper was easy to find, and that's what I used in the end. I searched the internet for a finer one, more like the excellent one used on Assos jerseys. I couldn't find anything like this, except for a hint that there are more options available for mass-production style work. Maybe another time I'll explore buying continuous zipper by the foot and assembling its termination components myself.

I'm happy with the results. For this project I wanted a wide final stitch to prevent fraying and distribute forces from the stretchy fabric, but the stitch itself doesn't need to stretch because the zipper tape can't. I found a stitch on the sewing machine that I'd never used before, and I really like both its look and technical properties in this application.Like most successful projects, I learned a bit and had fun doing it, then wound up with a piece of equipment that will help me for a long time. All these benefits are the justification for the time spent doing the work.

D2R2 2010

The Deerfield Dirt Road Randonee, or D2R2, is a challenging bike ride in western Mass, riding up into Vermont. Last year was my first time riding it, a 100km (65 mi) ride which was more demanding of my abilities and equipment than any other. It was so much fun that I wanted to try a longer version this year. My goal, if all went very well, was to ride the longer, 180km route to the lunch stop shared by the two routes, then switch to the 100km route for the way home, making a medium-sized route of 93 mi.

Training did not go perfectly, of course, and as the ride approached on the calendar, I wrestled with whether to swap my early start time for the later 100km start. All of my friends from last year's ride either stayed with the 100km route or had to skip it altogether, so the longer ride would also be solo. But a pair of dirt road prep rides went better than I'd expected, and I decided to go for it.

My 4am alarm was unnecessary after I naturally woke at 3:15, and this was a lucky stroke because it took longer than I'd planned to get out the door. Drove 4:25-6:20, watching the sun rise and the temperature drop to 50 degrees F. After weeks of heat, I hadn't imagined it could be cold, so I didn't have any extra layers. After registering, I found a pair of bags the OJ cups had come in, and fashioned them into ersatz arm warmers.

"Sea smoke" floated on the river next to the parking area as I started riding at 6:53. The earliest start was 6, but the organizers wanted untimed folks like me to wait until at least 6:30. I hoped to start about then, so was behind schedule already. But I was determined to ride at a natural pace, to avoid burning out early, so I pushed time out of mind and enjoyed the scenery and some conversation with other riders. One of the main draws of this event is the alternating river and hilltop views of the beautiful rural landscape of the Connecticut River watershed.


The first water stop was at mile 21, where I learned I only needed two (XL) bottles and emptied the third. I also learned that I didn't need the helmet light I was wearing, when two of the staff asked whether I rode to the start before dawn. Later in the ride my neck muscles would also tell me I shouldn't have loaded that weight onto my helmet. A rider next to me commented that he'd already had 3 flats, and was out of tubes and into the patch kit. This was just partway through the first quarter of the ride, designated a warm up!



Randonees don't have marked turns, unlike most big road rides, so cue sheets are very important, and a GPS can be quite helpful. We came to a place where the cue sheet said, "CAUTION: wicked downhill next mile - steep, rutted, narrow, stony". This sounded good to me as I've always enjoyed descending, and this ride feels like a nearly constant climb. After a bit of dirt road descending, a fork was marked with one of the few D2R2 signs of the day. I'd have gone left without it, but saw the arrow, hugged the adverse camber and kept the momentum going right. A little levelling was followed by one of the longest, loosest, fastest descents of the whole day. I thought, 'I'm glad the riders have thinned out so I don't need to worry about other riders here!' as I hung on. I saw bike tracks in the dirt, a reassuring sign, but noted that the path was wavy, as if a roadie were skidding the brakes downhill. I again thanked my cyclocross tires for their wonderful grip. The cue sheet and GPS were beyond illegible with the handlebar's vibration. It crossed my mind that it would be nice to check them, but there weren't any real options since that D2R2 sign, so no chance of error until I passed a little road that had forked right. I stopped slowly as I was still on the loose slope, and the GPS read "U turn when possible". Oh no. I struggled back up the hill to the little fork, and the GPS didn't want me to go there either. Oh no. I had to go all the way back to the D2R2 sign, finding I'd misread the direction of the arrow. I'd backtracked 1.7 miles, including the first walking of the day. When I rode, I wobbled a bit for traction - just like the track I'd seen on the way down! This was not the right way to prepare for the big hill coming soon on the cue sheet.

"15% climb next 3/4 mile", read the cue. I'd been bracing for this all morning, and sure enough, it was a dirt road slog. The next cue included "super steep", which I'd pictured describing the descent which would naturally follow a climb. Nope, it was more dirt climbing, and here I did some more short walks.



A little after the next waterstop lay the most fearsome cue, "27% grade", a dirt climb. The Wednesday before, some friends and I practiced on a local gravel trail and succeeded climbing some very short climbs that were steeper than this, so I had confidence that I could ride this at least for a bit. But the reality today was a tall climb, perhaps 200ft, and a surface of loose dry sand and rounded pebbles. Almost as soon as the slope hit, my rear tire slipped right, tipping me left. I was unable to clip out of the pedals while falling over, so I landed on my elbow and started rolling backward, the bike coming over me. To cut through the frustration, I noted that this was the mirror image of a fall I took Wednesday on a steeper climb, and now both elbows were symmetrically scuffed. After walking around 100ft, the slope seemed to drop just a bit, so I tried again and immediately fell the same way, to the right, re-opening Wednesday's scab. The rest of the climb was a walk, during which I promised myself I'd get rid of these pedals. There were bike tire tracks in the sand, though I can't figure out how anyone could ride it. There were also shoe prints and one hoof print, perhaps from a donkey, the proper vehicle for this road.

At a tiny water stop I started fretting about lunch. It was 1pm, and lunch was to close at 2:30, with 12 miles between. My average speed indicated it would take an hour, but the ride had so many surprises so far that I couldn't bank on that. Mechanical troubles seemed to be plaguing the other riders too, and I couldn't be immune. If this were a novel, you might expect the devil to appear about now, and that's just what happened, though it wasn't clear until afterward. A few turns later, I saw a rider ahead of me take a left, then a pair went right. I stopped and checked cue sheet and GPS and decided the left was correct, so I caught up with the pair and said I thought we should go back and go the other way. The guy in front said, 'this is the quick way to lunch'. What a relief to find a way to stop sweating over missing it! And I'd seen that left led to yet another big dirt climb. But I shook myself out of it and turned back, rejoining the route. I hadn't given up the real route yet! And anyway, I didn't want to have that guy be my only guide to lunch. The more I thought about it the less I appreciated his invitation to join his shortcut.


One gravel descent here gave me what I think of as a "pinch kiss". (If there's a real name for this, please let me know.) It's related to the infamous pinch flat, in which the tire is squished hard against the metal rim by a rock or similar bump, cutting through the inner tube and causing a flat which is difficult to adequately patch. Lots of riders got pinch flats on this ride. A pinch kiss is similar but doesn't cut all the way through the tube, so it's more a warning than a problem. It makes a loud, bad sound like a cartoon kiss. Also, I believe a "kiss cut" is an industrial term for cutting partway through something, like cutting the shape of a sticker without cutting its backing paper.

I arrived at lunch at 2, and there was still plenty of good food. Like last year, the stop was at a covered bridge in Vermont, next to a small waterfall. This is a peaceful spot to eat and rest a little before the long ride back.


Navigation for the return was a bit different, except for the cue sheets. Amy had loaded the GPS with a turn-by-turn route for the first half, investing significant effort converting Delorme data for my Garmin unit. While this was great for many turns, in others the GPS indicated turns that it fabricated, probably due to differences in map details between the companies. Garmin didn't recognize an apparent logging track as a road? I can't say I'm too shocked about that. Anyway, I had no such route for the return. Last year's track, just a colored line on the map, should have helped, but it turned out that track was not loaded, so the GPS did little after lunch. On the other hand, I remembered much of it, so didn't need as much help.

At the end of the 100km ride, there was an optional 5mi flat loop called the "corn maze". (This was a mandatory part of the 180k route.) I don't think it was even available last year. I wanted to ride it, and felt pretty good when I reached it, so I stopped for a quick drink and bite of the sandwich I'd been carrying, and went into it. This was just a ride on the tractor path through a corn field, a fun cool-down after the hills. In the quiet, though, I could hear the grinding of the chain and gears loaded with the day's road grit. I finished at 5:40, getting home before 9:30.



It was a great ride, the biggest of my life. At 8:36 moving time for 103mi (2.4mi vertical), it was about 1.5 times the work of a road century. I recommend it! There are other routes available, down to 40mi flat.

Russel Mills

After colleagues GC and SM showed me these wonderful trails, I brought E there Thanksgiving weekend. Today was our third ride there, and E has rediscovered the joy of mountain biking.