Welcome to Escape Collective. Please select your language.
Please note that this is an automated translation and it will not be perfect. All articles have been written in English and if anything appears to not make sense, please double check in English.
A few weeks back on the Geek Warning podcast we discussed what and how the Escape Collective tech crew tinker with review bikes we are sent. We dial in our preferred fit, and quite often, we’ll change a saddle or tyres, we swap wheels in and out, and the odd time we might change a stem length. None of us mentioned a complete strip-down and rebuild with an almost entirely new set of components, but that’s exactly what I did with the recently reviewed Cannondale SuperSix Evo.
The bike arrived to me in late April, just weeks before I was due to ride the Ras Tailteann, Ireland’s premier, and legendary, stage race. I had planned to race my S-Works Tarmac SL6 with mechanical Record 12 speed, rim brakes, and Hunt 50 mm clincher rims. A fine bike, my own bike, but not a modern so-called “superbike.”
The Evo had landed and the lure of racing the latest iteration of a do-it-all premium bike was 1) too much to resist and 2) exactly, I told myself, how I should use the bike to ensure a thorough review.
That’s not quite what happened, though. I’d decided I’d race on the Supersix Evo, but where’s the fun in racing it if I don’t also “optimise” it? This was, of course, to be “my last Ras,” and so I wanted to give myself every opportunity to perform well. I knew the Evo was a great race bike, but I could make it faster.
Nothing was off limits, although I didn’t bother to consider changing the shifting elements of the groupset, the time and effort spent in upgrading Ultegra Di2 could probably be better spent training a little extra. Everything else was fair game, though. Narrower handlebars, lighter (and TT-specific) saddle, deeper rims, and the pièce de résistance – an aero-optimised drivetrain.
I will delve into the exact components and spec list in a moment, but before I do: the results. The results at the Ras were not what I’d hoped for, but the bike certainly delivered the gains. Everything considered, the optimised version of this Cannondale SuperSix Evo did balloon a bit in weight to 8.1 kg, but in return aero testing of the frankenbike versus the SL6 I was due to ride suggests an aero saving of around 28 watts! That’s practically a .4 watt / kg increase in threshold power. And the testing was done at 32 km/h, so chances are the savings are even greater at race speeds, especially when attacking.
Due to the logistics of building/rebuilding the SuperSix Evo I wasn’t able to do same-day aero testing comparing the two versions, nor did weather cooperate with an attempt to test the SL6 against the stock SuperSix Evo. So we don’t have a direct comparison of savings between the two versions of the Cannondale. But we do have a significant savings to the Tarmac SL6, already an aero-focused frameset, built with similarly high-quality but not aero-optimised parts to the SuperSix. I also tested against a Ridley Falcn RS and found slightly smaller but still impressive gains.
How did we find a 28-watt saving? Read on to find out.
The frame, fork, seat post, shifters, rear derailleur, and stem all survived from the original build. Oh, and the aero bottles also. Let’s start with the basics. Out went the 50 mm-deep HollowGram R-SL 50 wheels, replaced with a 68 mm-deep Parcours Chrono front wheel and the matching 75 mm-deep rear Chrono. While the word “chrono” is more typically associated with time trials in cycling, these wheels are not the last thing more typically found on a time trial or track bike to feature on this road racing build.
The Evo was delivered with Continental’s GP 5000 tube-type tyres, those were replaced with the 28 mm GP 5000 S TR. If I was to build this bike again, I’d consider the GP 5000 TT version of the same tyres. Front-wheel upgrades complete. I skipped the disc rotor fairing … because that would be illegal 🙁 Staying up front for a second, as that is where the bike and rider the wind hits first, I had to change the Vision Trimax Carbon Aero handlebars I was finding so difficult to get along with. In came the Aerocoach Ornix hyper-aero bars. the Ornix are 325 mm wide at the hoods, flaring out to 375 mm at the drops to provide a wider position there, and, crucially, circumvent the UCI minimum handlebar width rule. The Ornix bars are narrower than anything else this aero profiled. They are not the lightest, weighing in at 361 grams, but that’s not the point. The Ornix bars are designed to include a little extra reach as standard to compensate for the decreased overall reach from saddle to hoods from a narrower handlebar. The idea is that the extra reach built into the bars should mean many riders coming from wider sizes will not need to purchase a new stem to correct their lever reach. The Ornix offer both internal and external hose routing options, but these upgrades are nothing if they are not aero, and so we stuck with the fully internal routing through Cannondale’s Conceal stem. Neat and aero. I was very tempted to get the trusty hacksaw out and chop the last 6 cm or so of the drops off. The extended drops do give the bars their UCI-compliant width, but I can’t help but feel they are needlessly long and, at this angle, provide additional unnecessary drag. Again, the frame, fork, and seat post all remained the same, moving back the next upgrade we get to is the Wove V8 TT saddle. The one-piece carbon shell and rail help make for a lightweight saddle, but, for me, the real gains are in the options the V8’s wide nose, grippy cover, and deep central cutout offer in increasing my ability to access the most aero positions. It looks sweet too. Wove have since released the Mags, a road and gravel-specific saddle that I’m currently trying to get some time on with an eye to a longer-term review. It has its work cut out , though, trying to top the V8. The Gripper aero bottles and cages also stay. That lump of powder in each bottle is just the tip of a humungous carb-berg I worked through in five days of racing. Now, the main event, the pièce de résistance, the bit you’ve all been waiting for, the “talk of the Ras,” a 60-tooth 1X narrow-wide chainring from Pyramid Cycle Design mounted to a Miche Pistard Air 165 mm track crankset. That crank isn’t a perfect match for a road bike; with a slightly narrower chainline, 138 mm Q-Factor, and 144 BCD, not much about these cranks matches the Ultegra crankset that was on the bike. Luckily, Pyramid offers (or at least they did) a 60-tooth aero chainring with a 144 BCD, it, which just barely cleared the Cannondale’s chainstays, and if anything, the narrower chain line helped as I inevitably found myself a cog or two further up the cassette than usual. That huge chainring and aero crankset were matched with an 75 mm rear wheel from Parcours to match. Basically, if it was aero, I was having it. Why, though? I had looked at the Ras route and identified plenty of wide main roads, very little climbing, and almost certainly incredibly fast racing; hence, I wanted a larger chainring on the front. Add in the friction efficiency gains of a larger chainring, and thus I wanted to go as large as possible.
But, with a larger chainring comes an increase in aero drag. there’s simply more of it catching the wind. Thus I wanted (read: needed) a 1X system, dropping the front derailleur and the inner-chainring drag, plus the most aero-optimised crank I could find. I also knew that reducing the Q-Factor would provide another aero gain. Overall, I was hoping the gains from the crank itself and the 1x configuration would offset the penalty from the chainring.
Truth be told, I don’t actually know if the Miche Pistard Air with a single ring is more aero than an Ultegra 2×12-speed crankset, but it sure looks it and the Miche includes immeasurable psychological gains as standard. I included Drag2Zero’s aero-profiled chain catcher and I didn’t have single issue or dropped chain with the 1X setup, until the last stage. That final day proved a nightmare with dropped chains starting a few kilometres into the stage and repeating sporadically throughout. My on-the-bike reaction was “damn 1X,” but investigations after the stage identified a sticky chain quick link as the issue. Just to cap things off, I opted for a CeramicSpeed bottom bracket and UFO-coated chain plus the new OverSized Pulley Wheel Aero system. Again, bigger is “gainier” and draggier. Again, I hadn’t tested if the OSPW Aero is faster than the standard OSPW or even if it’s less draggy than the standard Ultegra pulley wheel hanger, but it sure looks fast. The Ras wasn’t flat this year; there was even a category-one climb on the first day, so you may be wondering, “How the heck did I get over the climbs?” Well, before revealing the final aspect of this build, let me first explain that I optimised for the parts of the race I felt I could be in the moves, making the selection, and be in the hunt. In current form and weight, that was not going to be on the climbs, so I was willing to sacrifice some outright climbing optimisation in order to maximise my potential on the flat and rolling roads where most of the racing would happen. Still, I was not going to get up any climbs in a 60×32, and of course I had to get over the climbs to stay in the race. That’s where Classified and its Powershift planetary gear system built into the rear hub comes in. I wasn’t going to need anything other than the 60-tooth chainring and the midsection of the cassette for the vast majority of the race, but when we did hit the climbs, Classified provided me with an effective 41-tooth “small chainring” hidden inside the rear hub where the gods of winds and aerodynamics couldn’t touch it.
It’s not that the Classified system is more aerodynamic, or even lighter (it was all but impossible to compare weights on this build with so many changes and even different rims), and it’s not that the Powershift hub made me faster in anyway, but the optimising potential it unlocked was almost unrivalled. Without the Powershift reduced-ratio gearing option, I could not have run 1X on this course and I certainly could not have run a 60-tooth chainring. The Powershift button isn’t the best aspect of the Classified shifting, as Dave Rome touched on in his much deeper dive review of Classified’s Powershift two-speed hub system. I opted to mount it below the Ultegra shifter, and it did require more thought and effort than a standard shfiter. This a chart displaying all the shifts I made for one of the five stages. The 11-32 Classified cassette doesn’t include a 17 sprocket, so I assume this is some auto-loaded ratio, but what is clear is I spent most of the time around the middle of the cassette, rarely cross-chaining all that much. This was, of course, speed and course-dependent, but it was also exactly what I was expecting when I opted for the 1X and larger chainring setup. Analysing the various stage files in WKO, I can see my average cassette sprocket was usually either the 14-tooth or 15-tooth after factoring in the sprocket combination discrepancy mentioned earlier.“What’s that button in there” … “Oh, that’s for my motor.” 🙂 Some have raised concerns about the efficiency of a planetary gear system such as that in the Powershift hub. Classified emphatically claims that its system is “more than 99% efficient in both ratios.” Truth be told, I don’t know if the Powershift is 99% efficient, but I also didn’t care. As mentioned earlier, I optimised for the flat and fast sections of the race. One thing that is not disputed is that the Powershift probably has no more efficiency losses in the “locked” 1:1 ratio (which I’d be using the majority of the time) than a standard road drivetrain setup, that is to say, it’s about 99% efficient. Factoring in the larger 60-tooth chainring creating less chain articulation, there might actually have been a small bump in drivetrain efficiency.Now, having a 60-tooth 1X setup and a Classified rear hub, the last thing I wanted was a puncture. I did of course, have a spare bike on the car, but I wanted to be on the optimised bike from flag drop to the finish line. The solution was, of course, tubeless, but I also included a Vittoria Air-Liner Road tyre insert: one of those “green foamy things” that is compressed when the tyre is inflated and, supposedly, expands in the event of a puncture deflating the tyre. Apparently, they work so well that the tyre is still rideable. Unfortunately (or fortunately for me), be it the tubeless sealant or just good luck, I didn’t actually suffer a puncture that week, and so thankfully didn’t find out how well the inserts work. So, did the optimised bike provide gains in the race? I’d say, again, I didn’t get the results I’d hoped for, but I was doing things my legs shouldn’t have been capable of. Bridging gaps to breakaways, stringing out groups, and most shockingly for me, I was attacking, bridging, and distancing the field on descents, combining the Evo’s handling and the speed of the optimised setup, not to mention the option to add power with the 60-tooth chainring when I’d normally be spun out in a 53.
That was my optimised Ras bike. Sadly it has been stripped down and rebuilt again since, and the Evo now sits in its original state, ready for collection for return to Cannondale.
As for that 28-watt saving, that is real and measured. As listeners to the Performance Process podcast episode on aero testing will know, I have quite a detailed protocol to ensure accuracy. Truth be told, that level of attention to detail isn’t entirely required for most testing, but what I didn’t mention on that episode was that I was testing different bikes rather than smaller changes in equipment, helmets, or clothing. As such, and with all the potential pitfalls that could scupper bike vs. bike testing, I believe this attention to detail is required.
Long story short, I used my trusty Tarmac SL6 as the baseline setup and tested against the optimised Cannondale seen here and the new Ridley Falcn RS. The bikes were mounted with Aerosensor’s new aero meter, with a mount on each bike, meaning I could simply switch the sensor from bike to bike from run to run.
I normalised the tyres, pressures, and weights across all three bikes (although the Ridley was fitted with Hunt 32 mm rims, a much lower profile than the rims on the other bikes), used the same power meter, adopted the same “standard, upright riding position on all three bikes (although with the Ornix bars being so narrow the standard upright position was inevitably more aero) , and completed six out and backs on each bike. In fact, I did an additional six out and backs on the Tarmac SL6 between Ridley and Cannondale runs. So if the Specialized is bike A, Ridley is bike B, and Cannondale is bike C, I did ABACA testing, with six out and backs for each run. It was a lengthy process.
Testing complete, I sent the anonymised data to Aerosensor, who ran the numbers, not knowing which bike was which or even the doubling up on baseline runs.
The results were pretty clear, the SL6 with exposed mechanical cables and a six-year-old frame at this point, was the slowest. The Ridley (with the lowest profile wheels) was 8 watts faster. The optimised Cannondale? Well, it tested 28 watts quicker than the SL6 and 20 faster than the Ridley! Ideally, I’d have liked to complete the tests again to double-check such a huge gain. There was some drift through such a lengthy test, but with such a large gain, we can be sure the Cannondale was signifcantly faster. And, of course, a large proportion of that gain was likely down to the narrower position (not to mention the significantly deeper wheels), but that’s the very point. I didn’t fit the narrower bars and the narrower and shorter crankarms to stay in the same position; I fitted them to go faster. With that 28-watt gain coming with just a 595-gram weight gain, it was mission accomplished at least as far as the bike is concerned.
