What has become my favorite annual summer gravel race has somehow not yet warranted a post, which is bananas because I've had some decent results over my 5 attempts.
I first heard of Dirty Kitten during the pandemic, and that year it wasn't a race, so much as an isolation fondo. There were timed segments, and riders rolled out at intervals. You were allowed to ride with folks who came with you, and I brought Alastair. Though he holds no love for gravel, this one got into his heart, and we haven't missed an edition since. We always do the Sunday races, which are 40 and 20 miles, as neither of us enjoy gravel quite enough to want a 60 or 80 mile race.
I've been on the podium in the 40-miler every year since 2021, and last year was the closest I've come to victory at 2nd place.
With just over 6 weeks until the 2025 event, I'm beginning my annual panic over training and equipment for an event most folks do just for fun.
The course is just a tick under 20 miles, with no pavement anywhere. Conditions change year-to-year, but generally the gravel is stable and well worn-in, so speeds are pretty high over most of the course. It's absolutely normal to see 20mph+ on the flatter sections, even with my middle-age legs. There are 2 significant climbs that will test gearing limits. The first, Cedar Mountain, is just a deep-gear slog. It comes less than a mile into the course and generally determines who's there to compete. Usually fewer than 10 people make it over the top at the front, and fewer still get a clean run through the off-camber crumble to the screaming fast descent. Last year I nearly ate it--the trail gods demanded a bottle sacrifice.
After that, you're rewarded with 10 miles of flattish gravel. A few sharp turns, a couple bits of soupy gravel, and one section that's best described as "forest road" where the ruts and conditions are unpredictable year-to-year (but I imagine this would be MISERABLE when wet).
Then you get to the signature element of the course: the lollipop of death and the kitten crusher. This 4.3 mile portion of the course is mostly exposed to unabating July sunshine and can be broken down into a few critical elements:
- the 1.4 mile approach road that pitches upward at anywhere from 3 - 11%. This thing just sucks the life out of you before you even get to the real climb, and every year somebody decides to make this road hurt even more than necessary. But I've never seen a move succeed here, and there's more than enough pain coming.
- The Kitten Crusher. "Is this the climb?" Oh you sweet summer child... The approach is 8%. The first ramp is 15%. Each blind left turn holds an even-steeper grade, but just wait until it turns right. The Strava segment shows 154.5% and goddamn it feels like it. Whatever little group you've been riding with will be shattered here. Most will walk, but that doesn't really make it easier. Fortunately (?) there's an aid station at the top, and a dude with a sprayer to cool you down. If you stop, your race is done. I've made it over 2 times in 5 years / 10 efforts.
- The Crusher Descent is quite honestly one of the scariest descents I've ever undertaken on a bicycle. It doesn't help that your heart rate is still sky-high from the climb, but the first turn is loose and steep, and while you're in blessed shade for the first time, that just makes it hard to see. Average gradient -11.4% with peaks close to -20%...in a turn. Hang your ass off the back of the saddle, squeeze those brakes as much as you dare, and just pray for the bottom--you'll get there one way or the other.
- Though there's no segment for it, the rest of the lollipop sees you climb back up to the approach road and run it in reverse. If you got dropped on the crusher, this is where the work starts. A half-mile climb at 5% back to the road, a quick breather, a 10% punch, and then a screaming 1 mile descent in 2-way traffic of yelling at folks to stay on their side. It ends with a braking zone that would make a F1 driver weep before tackling a sharp right back onto a public gravel road.
The rest of the course is far less dramatic. At about 3 miles, it's 1 mile of flat recovery (or chasing, depending on your crusher results), 1.5 miles of undulating rollers in sunshine, and 0.5 miles of sharp rollers in the shade with an uphill kick to the finish line.
Then you get to do it all again. Yay!
I've spent years trying to get the "perfect" setup for this race. My first forays saw me on a 1x mechanical SRAM setup, always with a 40T chainring, but with various cassettes from 11-36 to one miserable effort with a 11-4(x?) and a WolfTooth goat-link intended to extend the supported cassette size for the derailleur. It didn't work well at all--the chain jumped all over the place and ghost-shifted on every climb. it was awful. Awful awful awful.
2 years ago I upgraded the gravel bike to electronic shifting, settling on a 10-44 cassette and keeping that same 40T ring up front. I was still dropping chains, but on the 2nd lap I finally made it over the crusher for the first time ever.
Last year I picked up a derailleur hanger alignment tool and corrected a massive mis-alignment, which made shifting so much smoother, but the chain *still* fell into the wheel on the way up. I ended up panicking on the 2nd lap and just got off the bike before it could fail again, costing me the potential to sprint for the win.
The only constant over the years has been at the front of the drivetrain: that 40T ring. While the gearing of the past 2 years has gone deeper than 1:1, I've still only gotten over the crusher 50% of the time, and lost the chain into the wheel 4 times. At the opposite end of the spectrum, the SRAM AXS app shows I only used the 10T cog for 54 seconds of the entire race.
That 54 seconds had an average power of 133W, as well. By contrast, the 44T cog saw 4:26 of use at an average power of 304W, with the steepest 2 ramps of the crusher at over 400W.
So this year I'm thinking of trying something different. I cannot affix a SRAM XDR-compatible cassette deeper than 10-44 (if such a thing even exists), and even if I could, through some sort of updated Goat-link equivalent, I would never want to re-invite the issues I had with the bike 3 and 4 years ago. Going XD (SRAM's mountain-bike drivetrain popular in "mullet" gravel bikes) would add tons of cost. But I can move to a smaller chain-ring for $20.
I've resisted this option for years for a few reasons. The first is, honestly, it feels wimpy to go smaller up front. I'm a Big Strong Man(TM)! I don't need a smaller ring for MonsterCross or any other gravel/CX racing I do, and the idea of constantly swapping equipment is not even close to appealing. But aside from that, there's chain-line efficiency: the 15T (4th from smallest) gets almost 45 minutes of use--far and away the most used gear on this course, followed distantly by the 17T at just over 30 minutes. Using a smaller chain-ring would put that use farther outboard, creating a less optimized chain-line, which increases drive-line resistance, even if only slightly. Then there's the compound issue of more chain movement and slap--even with a clutched derailleur--as the chain moves farther outboard. I run a chain-keeper on this bike, but when the chain comes off, it always gets tangled in the keeper, which makes it just that much harder to re-seat when you're already on the limit. I don't want to create another situation where I panic and just get off the bike.
So what size ring? I went through this exercise when setting up the old crit bike as a purpose-built climber, and ironically ended up at 40T on that bike, too, though that was driven by wildly different goals and maths.
The power-meter crankset on this bike has a 5-bolt 110BCD interface, so I could go down as far as 34T, but that would see me spun-out for almost 15 minutes of the race. The gear-ratios of 40:15 and 34:13 are nearly identical, so that 45-minutes in 15T would become 45 in 13T, and it would come with a ridiculously deep gear on the climbs.
A 36T chain ring would offer one deeper gear for the climbs while only cutting off the top gear on the flats, making me spun-out for only about a minute. A 38T ring is honestly insignificantly different from a 40T--changes at the front need to be more substantial than changes at the back.
Since I'm trying to solve for about half of 4:26 of the race, I don't want to make a change that will impact the faster bits, so I've settled on trying the 36T.
According to my calculations, that will reduce the top speed at 90RPM to 26.9mph, which will only affect the two main descents--but let's be honest: I wasn't doing much pedalling on those anyway.
No comments:
Post a Comment