Take a look at the photo below of Nairo Quintana. Maybe it’s a trick of the eye, but it seems like his rear wheel is a deeper rim depth than his front. Regardless of whether it is or isn’t, it made me think that if a wispy climber like Quintana was riding a set of mixed depth aero wheels, maybe mere mortals who weigh more and don’t possess the same climbing prowess should consider doing so too? And so the seeds for this PEZ article were planted. [I’ve subsequently confirmed that Quintana rode a 35mm front and 50mm rear Campagnolo Bora Ultra wheel set in this year’s Tour de Suisse].
Is Nairo riding mixed depth wheels?
But why would you want to ride mixed depth wheels? The idea is that a shallower depth front wheel is less susceptible to getting snatched and blown around by the wind, so better handling. But a deeper depth rear wheel provides more aero benefit. Such a setup will not be as light as shallower depth wheels front and rear (what you would expect to see in the high mountains), but also not as heavy as full-on aero front and rear (what you expect sprinters to ride). And it will be more stable than deeper depth wheels front and rear, but not as aero. So a compromise, but one that – at least on paper – seems to make sense for lighter riders like myself who would rather not go sailing in the wind riding deeper depth wheels both front and rear.
Creating a mixed depth wheel set is easy enough if you’re able to buy same model separate front and rear wheels from a manufacturer. Clearly, sponsored pro teams such as Movistar can easily mix and match wheels to their liking. But for the rest of us, at least a few manufacturers offer mixed depth wheel sets. The Enve SES 3.4 wheel set has a 38mm deep front wheel and 42mm deep rear wheel. This wheel set is actually spec’ed on the Cervelo R5 Di2. Not in their current catalog, but Reynolds used to offer an Assault/Strike combo wheel set with a 41mm deep front and 62mm deep rear wheel that can still be purchased from a variety of vendors.
Since the beginning of spring, I’ve been riding a set of Irwin AON TLR 38 wheels. Irwin also makes an AON TLR 38/58 combo wheel set and they were kind enough to provide one for apples-to-apples comparison riding.
The AON TLR 38 and 38/58 wheel sets are the same except for the rear wheel. The AON TLR 38 has 38mm deep carbon rims front (20 radially laced Sapim CX-Ray spokes) and rear (24 Sapim CX-Ray spokes 2X on both sides). The rims have a 18mm internal width and 26mm external width. The hubs use sealed cartridge ceramic bearings (2 in the front and 4 in the rear). Claimed weight is 1525 grams. The AON TLR 38/58 has the same front wheel but the rear is a 58mm deep carbon rim, and tips the scales at 1640 grams – so 115 grams heavier.
Same depth (38mm) carbon aero wheels
38/58mm mixed depth combo wheels
So is a 115 gram weight penalty worth the aero benefit?
Well, I can’t say definitively based on hard scientific data. Sorry, but I don’t own or have access to a wind tunnel! But I can give you my riding impressions and “analysis” based on riding both wheel sets.
My AON TLR 38 wheels were my initial foray into carbon fiber aero wheels. I was previously riding Shimano RS81 C35 wheels – 35mm deep with an alloy brake track and carbon fairing (16 spokes laced radially up front and 21 spokes – 14 spokes 3X on the drive side and 7 radial on the non-drive side – on the rear wheel). Tolerably “heavy” at ~1700 grams, but a great all-around set of wheels that I’d ridden for three seasons on all sorts of terrain, including some extended climbs on Skyline Drive. They were (still are) pretty much bulletproof and I had no complaints about them whatsoever.
That said, the switch to aero carbon wheels was a revelation of sorts. Yes, my carbon wheels feel lighter, but the truth is that they’re only ~200 grams lighter per the spec sheets. But because the AON TLR 38s are “true” aero wheels, they come up to speed faster and hold speed better. And they just feel faster, but I have no way to validate that. However, for PEZ readers who are data geeks, here is what Irwin has to say about how fast their wheels are in terms of drag and aerodynamic benefit.
For me, the biggest difference is that the AON TLR 38s are more stable in crosswinds. I’ve actually felt my front wheel get “snatched” by the wind riding my RS81 C35s, but not with the AON TLR 38s. For a lighter weight rider such as myself, that’s a big deal and why a shallower depth aero wheel set seemed to be a better choice than a deeper section (for example, 50mm) wheel set.
I’ve put several thousand miles on my AON TLR 38s – including riding the incredibly hard Fleche Buffoon that is 75 miles and 6,000 feet of climbing parsed out across 14 Ardennes-like climbs and is described as a half version of Liege-Bastogne-Liege – and I’ve been more than happy with them. And I would expect that comparable carbon aero wheels from other manufacturers would be much the same.
Casually deliberate contemplating the 14 climbs of Flèche Buffoon
Going more aero
But then I saw that picture of Nairo Quintana and kept asking myself if there would be any real benefit by going to a more aero rear wheel. We all know that aero wheels provide a real and tangible benefit. And that deeper section aero wheels are more aero. A lot of the aero conversation focuses on the leading edge of the bicycle to make it as “slippery” as possible through the wind because it’s the front of the bike that creates wind resistance and less is better. So it would seem that a more aero rear wheel would, by definition, not make a whole lot of difference. But the trailing edge is important too. That’s where the airflow separated by the leading edge rejoins. The idea is to reduce turbulence, which also makes the bike faster as well as more stable.
So in a world of marginal gains, a more aero rear wheel makes some common sense.
I am a willing guinea pig
As this is written, I’ve got not quite 500 miles riding a mixed depth aero wheel set. But it was “you had me at hello” starting with the first ride. I know it was probably just all in my head and the “something new” phenomenon, but the AON TLR 38/58 combo just felt faster. Going uphill didn’t seem any harder. And they were just as stable in the wind.
Or was that all just placebo effect because that’s what I was wanting and expecting?
I said at the onset that I wouldn’t be able to say anything definitive based on hard scientific data. But that’s not the same as not having any data and some evidence.
To wit, I decided to do some back-to-back riding at Hains Point, which has the second most ridden Strava segment in the U.S. Hains Point is a flat loop that’s a little over 3 miles all the way around, with 2.5 miles of it uninterrupted by stop signs. It’s a popular place for flat out riding and pace-lining. And even on the calmest days, there always seem to be some wind (on this particular day only ~5 mph). So a close to perfect place to test ride ride aero wheels.
After a warm up lap, I did two 3-lap efforts – first on the AON TLR 38 wheel set and then on the AON TLR 38/58 wheel set. Neither effort was a flat out FTP or time trial effort. I wasn’t trying to see how fast I could go. Rather, I tried to keep each effort below my anaerobic threshold and inside my aerobic sweet spot. And I did this purely on feel as I actually forgot my heart rate monitor strap. Doh!
Here’s my Strava summary riding the AON TLR 38 wheels:
And here’s my Strava summary riding the AON TLR 38/58 wheels:
Looking at estimated average power, energy output, and cadence, the efforts aren’t exactly identical (it would probably be impossible to do that) but they’re pretty close. That’s at least an indication that I wasn’t trying to go faster on the second run. Yet I was faster … all of 0.2 mph faster. But still a marginal gain. In theory, the differences in speed would be magnified at higher speeds.
And here’s a look at comparing the gate-to-gate segment, which is 2.5 miles of uninterrupted riding. Lap 1 was in the small (36T) ring and laps 2 and 3 were in the big (50T) ring. Again, faster on the 38/58 combo.
I want to stress that this was obviously not a controlled scientific experiment so the data – such as it is – shouldn’t be viewed as definitive and conclusive. However, it’s at least some piece of evidence that a mixed depth aero wheel set is somewhat faster than a same depth aero wheel set. But I’ll let you interpret the data as you see fit and draw your own conclusions.
So if you’re willing to believe a mixed depth aero wheel set is faster – even if just ever so slightly – than a same depth aero wheel set, what about in the hills where weight and gravity are more of a factor than aero?
For that, I created a ~10-mile route with ~1,000 feet of elevation gain. The climbs varied from short and wall-like (a tenth of a mile and 10 percent grade) to longer and more steady-state like (over a mile and 4 percent average grade). But since I live in a dense, urban area and not the Alps or Pyrenees, no “big” climbs. Here’s the route I put together that also shows the profile.
As with my rides at Hains Point, I did my best to keep my efforts as “equal” as possible. But I know that’s inherently more difficult climbing. At least I remembered my heart rate monitor (the Cyclops heart rate/power meter that’s the older version of the PowerCal that uses an algorithm to interpolate power from heart rate) to help me monitor my effort on each ride. Mostly I wanted to make sure I didn’t go so hard that I put myself into the red. As far as riding on the hills went, I tried to stay seated as much as possible and spin rather than push gears, but I did get out of the saddle and up on the pedals a few times.
Here’s my Strava summary for riding AON TLR 38 wheels in the hills:
And here’s my Strava summary for the AON TLR 38/58 wheels:
So the first thing you’ll notice is that my elevation gain for both rides is less than the 1,062 feet Strava calculated when I created the route. And that the elevation is different for each ride (but only by a difference of 16 feet). I subsequently used Strava’s elevation correction feature which gave me 977 feet. I have no idea why it’s not the same elevation that Strava calculated when I created the route, but at least it’s closer to the predicted 1,062 feet.
Both efforts are pretty close, but it looks like the second effort was just a tad more looking at heart rate, cadence, and power. (BTW, ignore the max power number for the second ride – that’s obviously a glitch that happens every now and again because my power meter really isn’t a direct force power meter but a heart rate monitor that uses an algorithm to approximate watts. If you want to learn more about it, read this.) I can’t say that I felt like I was working harder on the second ride and the numbers may reflect that it was getting hotter (and it was already hot and humid to begin with).
As with riding at Hains Point, neither of these rides were attempts to chase PRs or KOMs. I wasn’t going slow either (I readily admit that my older self is slower than my younger self and that my younger self may not have been that fast), but more of a “right side of comfortable” pace to consciously avoid going into the red.
Overall, my second ride on the mixed depth wheel set was slower – coincidentally, by 0.2 mph which was how much faster I was on the flats versus the same depth wheel set. But was actually I slower climbing?
Here is the data for the different climbing segments on the route (note that the white rows are segments that are actually within the larger segment above them).
Looking at the individual segments, the data is mixed. I was slower on some segments and faster on others. And I was “interrupted” riding Molte Piccolo Mortirolo the second time by another rider who saw me and wanted to ride along for a little bit. Because I’m a big believer in not being “too pro to wave,” I didn’t wave him off or otherwise ignore him. I told him what I was doing and we had a nice chat as we rode part of the climb together until he veered off to go home. So that probably accounts for some of going slower on that segment.
As another data point For Molte Piccolo Mortirolo, I rode this segment on the AON TLR 38/58 wheel set the previous Wednesday on a weekly hill ride that I lead. On that ride, I “lead from behind” to make sure the last rider doesn’t get dropped or lost. So I’m not trying to go hard. As such, my effort is not unlike how I rode the test ride. And I was a little faster at 6:37 for the segment. So it’s not the wheels per se that are faster or slower.
For me, the most interesting segment comparing wheel sets is the Tri360 QOM/KOM Challenge. It’s a mini Tre Capi in that it consists of three distinct climbs as part of one overall climb: an initial steep section of ~10% average grade (the 31st/26th steep segment), which is followed by a short descent and then a very short ~10% climb, a false flat, and then a longer, more gradual mid-single digit climb to finish (the last two sections comprise the 26th St N segment). For whatever reason, I never feel like I ride it well from start to finish. So the fact that I wasn’t slower on the mixed depth wheels came as a bit of a pleasant surprise (and it was my second time up so you would think I might be a little more tired, especially in the heat and humidity).
Once you go mixed, you never go back
Again, my riding the two different wheel sets does not constitute anything close to a scientific experiment. And you would want a lot more sample points riding the different wheel sets on the same segments to have a large enough data set to approach something even remotely statistically significant to quantify any differences and make a truly data-driven decision.
Just going by the numbers based on very limited data, it would appear to be a wash. I was 0.2 mph faster on the flats with the AON TLR 38/58 wheel set, but 0.2 mph slower on the hills.
But even in an age of power meters and riding by the numbers, cycling is also about “feel” – “sensations” as Alberto Contador would so often say. So there’s a subjective – even emotional – part of the equation to consider. Especially if the data isn’t conclusive one way or the other.
That said, for me and the type of riding I do these days, the mixed depth aero wheel combo just feels “right.” It might be just psychological, but they feel faster. Maybe if I was doing multiple long climbs over the course of a single ride I’d feel the cumulative effect of those extra 115 grams in the rear wheel. But I get out to the mountains for that kind of riding only a couple times a year, if that. (I’m actually hoping to get out to Skyline Drive to ride the AON TLR 38/58 wheels on some long climbs before the end of this riding season.) Probably most importantly for me is that I don’t compromise handling in the wind with a shallower depth front wheel and deeper depth rear wheel. For a smaller, lighter rider such as myself (5’8″ and 134 pounds) who rides a variety of terrain, I think a mixed rim depth wheel set makes a lot of sense. Especially if you only have one set of wheels to ride, think of it as a best of all worlds or jack of all trades set up.
Not convinced? Well, there’s always the most important criteria of them all … I think they look badass! Of course, there’s no accounting for taste.
But it’s not just me. If you watched this year’s Tour de France, you might have noticed that more than a few riders were on mixed depth wheel sets. One of those was Julian Alaphilippe who rode such wheels to victory on Stage 10 to put himself in the coveted Polka Dot jersey.
Who am I to argue with that?
Just like Julian Alaphilippe. Or maybe he’s just like me?
PEZ contributor Chuck Peña is a former weekend warrior racer who now just rides for fun, but every once in a while manages to prove Fausto Coppi’s adage true: Age and treachery will overcome youth and skill. He lives in Arlington, VA with his wife (who is his favorite riding partner), his daughter (who takes great joy in beating him at golf all the time, but at least he’s still faster on a bike), and their dogs (who are always there to greet him when he comes home from a ride). You can follow him on Twitter and on Instagram.