This article is also available in German. Suchst du nach der deutschen Version? Folge diesem Link.
What is faster: Dropbar MTB or gravel bike? Or monster gravel vs. gravel? Or full suspension 29er vs. gravel bike?
I have already asked and answered a similar question before. Now, with my new dropbar MTB, which currently brings me so much joy, I wanted to know again:
By how many watts is my Canyon Exceed in its dropbar version slower than my my No 22 Bicycles Drifter – my titanium gravel bike or all-road bike, respectively ?
Am I wasting unnecessary watts when I’m riding it on good paths or even on asphalt? How much slower am I at the same power? Or am I just as fast or even faster (which would be unexpected)? I don’t need to ask the question for really bad roads: it’s clear which bike is faster when I have to hold back or want to hold back with the gravel bike because I fear for the tires, rims, my wrists, or even a crash. So, it’s about the (possibly?) remaining disadvantage of the dropbar MTB compared to a (race) gravel bike or even an all-road bike on asphalt or so-called „champagne gravel.“ Champagne gravel? This refers to super fine gravel or firm, cohesive surfaces on which it rolls smoothly without rough potholes, ridges, or nasty rocks that one would have to watch out for. And on which you could almost ride just as well with a road bike and even enjoy it.
But wait! Could it be that a dropbar MTB is ultimately faster than a gravel bike everywhere? Even regardless of the issue of extremely bad roads, where one holds back the speed out of concern for the equipment (primarily the tires)? Or even if you don’t hold back are simply slower due to too narrow tires and no suspension?
We will be also considering topics like cornering speed, suspension, comfort, and – currently back in the limelight thanks to events like Unbound and other mud battles – mud clearance aka tire clearance. We will only „measure“ aerodynamics and rolling resistance here, though. I will address the other aspects anecdotally.
The result may surprise you and also highlight the importance of the tire topic once again.
And since I have already published so many fundamental articles on this topic, we can get straight to the heart of the matter: the results of my test rides with both bikes and a total of 3 different tires, including analysis of aerodynamics and rolling resistance.
And as a preliminary note for the naysayers, who always exist, and for those people for whom knowledge of physics and paying attention to a well-rolling bike seemingly contradicts enjoyment or relaxation: Faster in this context also means:
- You’re faster to the cake. Or…
- You’re equally fast but more rested and ready for digestion when it comes to cake! Or…
- You’ll arrive in time for cake before the bakery closes! Or…
- You don’t have to overeat on cake and still have enough energy to complete your long tour. And so on and so forth.
According to the motto: More fast is more fun. But more fast affords also more lazy!
Some background information as preparation, if you want:
In the following articles, you can look up and find descriptions of the applied methodology, as well as the fundamentals of aerodynamics, aerodynamic testing methods, tire rolling resistance, and more. However, if you want to skip ahead and see the results immediately, you can quickly scroll to the next chapter. It’s also worth noting, that the mentioned articles in this chapter and throughout the course of this article are mostly only available in German.
The following article was my first one on the specific question, where I compared my Rose Thrill Hill (full-suspension XC MTB with a flat bar and mini aerobars) with my No 22 Bicycles Drifter: MTB vs Gravelbike – was ist schneller? Aerodynamik und Rollwiderstandtests für Jedermann (Chung Methode, GoldenCheetah Aerolab, Aerotune.com und co).
In this article, you will already find a comprehensive overview of all aspects, pros and cons, as well as the possible methods for quantifying the differences. This allows for reliable and weather-independent conclusions.
The suitability of a particular bike for a specific purpose or course does not always depend solely on rolling resistance and aerodynamics. Tire type, tire width, suspension or no suspension, type of handlebars, and whether or not to use aerobars all influence how well and how quickly you can cover a certain distance. What good is the best aerodynamics if the course is filled with roots and you can’t really pedal? Or if one curve follows another, and it’s much more important to maintain high cornering speeds with confidence – in which case a wider 29-inch tire is superior to a thinner 700c tire.
The longer a race or a bikepacking trip is, the more important the topic of comfort becomes! It is always worth considering whether a more narrow gravel tire and a lightweight race gravel bike are really faster over a certain course. But if they are: yes – you can power through it for 3 to 6 hours. You may end up completely destroyed but you tackled the course as fast as possible. But – as events get longer, possibly even lasting for days, the person who did not exhaust their energy by constantly struggling with the bike and the (too) rough terrain will be faster in the end.
However, it doesn’t always go in one direction towards increasingly thicker and more aggressive tires or more upright riding positions (which is not a good solution for everything anyway) being always better and better. It can be just as effective the other way around: if the terrain is not so rough, with many sections of fast-running asphalt or gravel, and it may be windy as well, then the person on the MTB or the bikepacking SUV (excuse me, the gravel bike with situationally oversized tires and too many used mounting points and unnecessary bags) will also struggle more than they would like to if they have to haul those maybe poorly performing heavy knobby tires against headwinds or heave the resulting weight up steep climbs… Perceived comfort or being prepared for all eventualities can then turn into the opposite and dampen the fun. The good thing is – you won’t know until you compare it with a lightweight, well-rolling bike without all the extra fuss and notice how much more carefree and agile you suddenly can zoom over the trails with it.
Anyone can easily determine and evaluate the weight of the bike. However, rolling resistance and air resistance, which are much more important than weight except on steep climbs, are not so easily measured. Yet, they can be readily determined and quantified. The corresponding methodologies can be applied by anyone. Admittedly, it requires some technical expertise and effort.
In the linked article, I also introduced and explained the two tools I have extensively used since a while: the Chung Aerolab (available in the freeware Golden Cheetah) and the AeroTune website.
I further delved into these and applied those methods in the two following articles:
Especially in the first of the two articles, I delve deep into the background and pitfalls – but ultimately, I also show the successful application of the corresponding tools.
Conclusion: Yes, unfortunately, it is still relatively time-consuming and requires a certain understanding and in-depth engagement with the topic. However, in the end, you will have reliable results from methodologies that rightfully stand alongside wind tunnel tests, rolling resistance test stands, and computer simulations in the arsenal of scientists, bike and component manufacturers, and professional World Tour cycling teams.
Conclusion: Yes, unfortunately, it is still relatively time-consuming and requires a certain understanding and in-depth engagement with the topic. However, in the end, you will have reliable results from methodologies that rightfully stand alongside wind tunnel tests, rolling resistance test stands, and computer simulations in the arsenal of scientists, bike and component manufacturers, and professional World Tour cycling teams.
Finally, when it comes to tires, you have probably already heard that wider tires with lower pressure roll faster or more easily than narrow tires with higher pressure. Yes, but… This is, of course, a very simplified statement and, in this absoluteness, very incorrect! Unfortunately, it takes such (sometimes overly simplified to the point of distortion) messages, which are hammered into people’s heads for years, until their core reaches the recipients (in this case, parts of the cycling industry at first, but eventually also end-users and consumers). In reality it’s much more complex.
So, it might be interesting for you to take a look at the third and subsequent sections of this article on tires and rolling resistance: Der optimale Reifen für Rennrad-Langstrecke, Ultracycling-Events und Bikepacking-Rennen (the link will take you directly to the relevant section)).
But enough of the preamble now. Let’s get to the results of my latest test. How big is the difference between my dropbar Exceed and my Drifter?
What is the speed difference between my Drifter and my Exceed?
Surprise! There isn’t actually much difference. If any. Or… is there?
But wait: Over what type of surface or terrain are we talking about? And in terms of which criterion?
And what is the current setup of both tested bikes? How do they differ from each other?
The competitors:
In the left corner: My Canyon Exceed in my dropbar conversion, as I presented it in this article: Dropbar-MTB, Teil 5: der konkrete Aufbau, Teileliste und das finale Ergebnis.
Tested with the Newmen Advanced wheels and Continental Race King Protection tires, 29×2.20″ at 1.8 bar. Weight including Garmin 1040: 9920 g. System weight on the test day (rider, kit, bike with filled bottle): 82.27 kg.
In the right Corner, Setup 2: My No 22 Drifter. Well-known from various articles on this blog. Tested with the Nextie Premium AGX 45 wheels and Schwalbe G-One R tires, 700 x 40c at 2.0 bar. Weight including Garmin 1040: 9080 g. System weight on the test day (rider, kit, bike with filled bottle): 81.39 kg.
So, essentially, we have two random gravel bikes with different tire clearances here. One has a non-aerodynamically optimized carbon frame (the Exceed), while the other has a non-aerodynamically optimized titanium frame. Both have drop bars. However, the Exceed has a front suspension fork and wider MTB tires on low-profile rims, while the Drifter has a sleek carbon rigid fork and narrower gravel tires on aerodynamically efficient rims.
Pre-conditions and resulting reservations:
This was a quick test conducted over three test days. Well, others may extract much more (perceived) information from just three laps around the church square, but that doesn’t meet my standards. To conduct truly reliable tests, it is necessary not only to repeat a setup three times in one day but also to confirm it on different days to ensure that the differences between the tested setups are repeatable and comprehensible. Only this way it allows me to confidently say: yes, the difference between Setup A and Setup B averages x watts at y km/h. This x value may be slightly larger or smaller or even equal to the number obtained from the initial test. But it provides the opportunity to learn about the range of results and see how variable they are. I have explained this in detail in the articles linked in the previous section.
It is also permissible to use two different power meters (which is necessary when testing different bikes), but in that case, it is important to try to align them beforehand. This can be partially achieved, for example, by using a pedal-based power meter and a crank-based power meter. Then, you can ride with the pedals on the bike equipped with the crank-based power meter and check if the same values are recorded. In a previous test, I went through this process with my Powermeter P1 pedals and was confident that they remained consistent when switching between two bikes. However, I am not so sure about my current Garmin Rally XC200. On the contrary, based on initial tests last winter, I found that even the latest version of the Garmin power meter pedals still have some issues regarding accurate (or at least sufficient) torque on the pedal screws. Or in the absence of that, they require 2 to 3 longer rides until the values stabilize even during peak power outputs.
At the moment, I have no reason to distrust my Garmin XC200 compared to my various other power meters. However, it is worth noting that the accuracy of the power measurement is crucial for any analysis, whether using the Chung method or the Aerotune procedure. Whether the Garmin Rally XC200 on the Exceed displays 1% fewer or 1% more watts at the same power level compared to my Sram Red AXS Quarq power meter on the Drifter, I simply do not know, and I would never be able to feel the difference.
In summary:
- Two different power meters were used to compare the Exceed and Drifter!
- Only three tests per bike and setup. That’s two more than many others do. That is exactly the number recommended by Aerotune. It is sufficient if no adverse circumstances or unnoticed errors were introduced. However, more tests are better because there are always adverse circumstances, variable environmental conditions, variations in the rider’s position on the bike despite the best intentions, and the desire to identify such sources of error.
The first partial result:
Partial result because I was almost surprised or rather relieved by how small the difference between both bikes is, and then I therefore tested another setup and further examined the recorded data.
Let’s first look at the result summaries from the Aerotune test.
On one hand, Aerotune is super easy to use, but on the other hand, it can still be tricky. I extensively explored aerodynamics and rolling resistance tests from summer 2021 to winter 21/22 (linked already – here again for your convenience), and in recent weeks, I hoped to find a more advanced and improved version of Aerotune. There have been significant developments, particularly concerning the physiological power test and a new beta evaluation of activities, among other things. Some of these exciting advancements were revealed in a recent podcast episode by the creators behind Aerotune (Aerotune Episode 8). That’s great and fascinating. Unfortunately, the beta version of the crr estimation introduced for the power test back then is still entirely unusable even now in summer 2023.
This is more than unfortunate, and Sebastian must definitely resolve this issue if he wants to realize what was teased in the last podcast episode: an automatic aerotesting based on every activity automatically synchronized with Aerotune. However, there are workarounds, and there’s still the Chung Aerolab in Golden Cheetah.
One possible workaround is as follows: Despite using different tires, the coefficient of rolling resistance (crr) will simply be entered as the exact same fixed (reasonable) value for both bikes. It acts as a known pre-condition in this way. In this case, it was set to 0.007, which corresponds to 7 rollingPOINTs (rP) in Aerotune terminology. All differences between the two setups (including the tire rolling resistance) are then completely reflected in the resulting coefficient of drag (cdA). So, I still know exactly how many watts more or less power Setup A requires compared to Setup B and how much longer it takes me to cover a certain distance. I just don’t know how much of that is due to aerodynamic drag and how much is due to rolling resistance. It’s by far not ideal, but it’s sufficient for the first step.
So, what does the first result from Aerotune show?
On the left is the result graph for the Canyon Exceed, and on the right is the graph for the 22Bikes Drifter. The green arc on the right side of each graph represents the crr (rP), which is not a result but a preset value. The red arc on the left side represents the resulting aeroPOINTS, i.e., the cdA, and in smaller text above it, the error range.
The cyan-green bars at the bottom show the corresponding watts required for three speeds: 35 km/h, 40 km/h, and 45 km/h. And between both graph contents, there is also the modeled time over a predefined course.
On average from these three test runs, the Drifter is indeed faster than the Exceed. The difference between both results is 3 aP, which corresponds to 0.03 m2 of cdA. In terms of required watts, that’s 18 watts more at 35 km/h on flat terrain. Converted to 30 km/h, it’s still nearly 11 watts that I need to exert on the Exceed to maintain the same speed as on the Drifter. This was on a smooth but unpaved bike path with a light, partial layer of gravel and sand.
On one hand, this is significant (in layman everyday language), but on the other hand, I had expected a larger difference between the Exceed with its suspension fork and wide MTB tires and the Drifter with its aerodynamic tire-rim combination and narrower tires designed for gravel bikes.
Let’s also look at the error range before we further examine the results:
With the crr fixed at 0.007, the Drifter has 45 aeroPOINTS (+- 4.9%), while the Exceed has 48 (+- 4.6%). This means the true result for the Drifter could be anywhere between 42.8 and 47.2 aP, while for the Exceed, it could be between 45.8 and 50.2 aP.
This implies that both could be exactly the same. Perhaps both could have 46 aP? That is possible, although not probable. Further test repetitions would provide greater certainty.
But first, let’s discuss possible reasons for the result before delving into more tests and analyses to get to the bottom of it.
Reasons why the Exceed might not be bad at all or almost equal to the Drifter:
I’ll start with this because it’s where the biggest and most important differences or similarities lie.
- Rider position on the bike. This is the most significant factor (accounting for about 70-80% of the total aerodynamic drag). Despite having a similar position due to the drop bars and other setup elements, the Exceed intentionally has a slightly higher handlebar and a lower saddle-to-handlebar drop (sag-corrected). Instead of a 6 cm drop on the Drifter, it’s 3 cm on the Exceed (sag-corrected). (See also the table in the section „Bike Fit Comparison between Flatbar MTB, Dropbar MTB, and Gravel Bike/Road Bike“ in this article). –> Expected slight disadvantage
- The Exceed has a wider handlebar than the Drifter, although it’s not an excessively wide flare handlebar. However, this difference in width might be offset by the resulting shoulder position, potentially neutralizing the effect. There is much to be considered in the ramifications of positional changes and the effects (literally) downstream lifting your shoulderblades, widening or narrowing it or leading to a slight different head position thus affecting the cDA positively or negatively. You really have to test this in detail. So – this is unknown. But nevertheless: –> Expected slight disadvantage
- Exposed cable routing or brake lines: This is likely a tie between the Exceed and the Drifter. Both bikes have clean cable management, and both have wireless shifting. The Exceed has a better setup for the rear brake line, which runs directly into the headset. The Drifter follows a more traditional approach but likely has slight advantages for the front brake. In any case, the drop bar Exceed is far superior to the typical cable mess on a flatbar MTB, even if it had electronic shifting. And much better than a flatbar MTB with mechanical shifting and damping lockout cables. –> No expected disadvantage
- The Continental RaceKing tires on the Exceed are among the fastest-rolling MTB tires available (according to Bicycle Rolling Resistance rankings, and on those we can rely on, as we’ll see later in this article). –> No disadvantage, rather an advantage expected
- Weight of the bike: On flat terrain, weight is the least significant aspect, but it still affects rolling resistance. Both bikes are relatively light and have minimal weight difference: Exceed including Garmin: 9920 g, Drifter including Garmin: 9080 g. That’s a difference of 0.84 kg. –> No real measurable difference expected on flat terrain
- During the test, the Drifter had an (empty) Quadlock mount on the handlebar. The iPhone was in the pocket, so the weight of this mount is not significant. But there is a difference. Whether the mount has a noticeable impact is debatable, but it won’t have a positive effect. –> No measurable difference expected
Reasons why the Drifter should be faster:
- Rider position: See the previous section on the slightly higher handlebar position of the Drifter.
- Smaller frontal area of the tires: 57 x 740 mm (Exceed) vs. 41 x 700 mm (Drifter). This results in a frontal area of 422 cm² vs. 287 cm².
- Aero advantages of the tire-rim combination on the Drifter (41 mm tire width on a 40 mm wide and 45 mm high gravel-aero rim) compared to the box rim on the Exceed.
- Smaller frontal area and likely better drag coefficient of the fork on the Drifter compared to the suspension fork on the Exceed, which has cylindrical (poorer drag coefficient) and thicker stanchions and an additional bridge between the stanchions.
The decisive reason might be: the tire rolling resistance.
As explained at the beginning when looking at the Aerotune results, the rolling resistance is included in the comparison. However, the specific contribution of each tire is still unknown since we set both tires to 7 rP as a constant.
My working hypothesis, before conducting detailed research for this article, was that both tires, the Continental RaceKing on the Exceed and the Schwalbe G-One R on the Drifter, are relatively good and intentionally chosen as the best in their respective categories. At least for the Continental RaceKing this is true. In particular, I selected it not only for its rolling resistance but also for its low weight. As for the Schwalbe G-One R, I experienced it as a very pleasant and well-performing tire that I’ve enjoyed riding and considered one of the best gravel tires I’ve used in terms of feel and cornering. And I also had in mind that it is also relatively good in terms of rolling resistance.
However, refreshing my memory on the actual test results for the G-One R made me realize otherwise. Compared to the RaceKing, it’s quite sluggish! Oh my! (Here’s the link to the gravel tire overview on BRR).
The difference in rolling resistance between the two tires is more than enough to overshadow all the previously mentioned reasons why the Drifter should be faster. In other words, if I were to fit the Drifter with a similarly fast tire as the RaceKing but in a 700x40c size, the difference for the Drifter would be even more pronounced, making it even faster!
From this, you can already see that the Schwalbe G-One R must have significantly worse test results in terms of rolling resistance compared to the RaceKing. The tests on BRR are conducted on a drum testing machine and purely measure hysteresis losses. While this is a good and common test method that provides reproducible results, it doesn’t show the whole picture (see the section on hysteresis and impedance losses in my tire article). Nevertheless, it serves as a more than adequate criterion to determine which tires roll faster or slower, regardless of the surface. More on that later in the article.
At a speed of 29 km/h and a load of 42.5 kg on the test tire, the following results are achieved on BRR for tires with appropriate pressures and tubeless setup with 40 ml sealant for MTB tires and 30 ml sealant for gravel tires:
- Continental RaceKing ProTection 2.20 x 29: with 2.4 bar: 0.00435 crr, 14.5 Watts
- Continental RaceKing ProTection 2.20 x 29: with 1.7 bar: 0.00471 crr, 15.7 Watts
- Schwalbe G-One R Super Race 40 x 700c: with 2.6 bar: 0.00724 crr, 24.2 Watts
- Schwalbe G-One R Super Race 40 x 700c: with 1.9 bar: 0.00827 crr, 27.6 Watts
Wow, I had a better values in mind for the G-One R for sure! But yes, it’s not that impressive. The G-One R is (barely) better than previous Schwalbe offerings in this field and definitely rides very comfortably and better than the G-One Allround and others. However, in terms of rolling resistance, it’s only average.
Interestingly, the Schwalbe G-One RS, which is a year younger, performs significantly better. Before you say, „Well, it doesn’t have any or barely any center knobs, so it must be faster!“ Not so fast! Compare it to the Conti RaceKing or other gravel tires with center knobs. Knobs may, but don’t have to, significantly increase rolling resistance. Sometimes, a slick profile or a profile with a continuous center strip can have stiff rubber that performs worse than knobs. Ultimately, the casing and rubber compound are playing the most significant role here. Knobs or no knobs.
Knobs may, but don’t have to, significantly increase rolling resistance. Sometimes, a slick profile or a profile with a continuous center strip can have stiff rubber that performs worse than knobs. Ultimately, the casing and rubber compound are playing the most significant role here. Knobs or no knobs.
So, is it just the slick or rather slightly treaded profile in the center of the Schwalbe G-One RS, or did Schwalbe make changes to the casing or rubber compound between the G-One R and G-One RS?
In any case, here are the results for the G-One RS for comparison:
- Schwalbe G-One RS Super Race 40 x 700c: with 2.6 bar: 0.00486 crr, 16.2 Watts
- Schwalbe G-One RS Super Race 40 x 700c: with 1.9 bar: 0.00591 crr, 19.7 Watts
Even higher on the list is the Schwalbe Pro One TLE 34. This is not a gravel tire per se, but rather a wide slick tire. With a width of 34 mm, it’s the widest option in Schwalbe’s Performance road racing tire lineup. It’s considered the fastest and smoothest tire that Schwalbe produces for professional road cyclists and enthusiasts (excluding special time trial tire models, which should be used only on well-maintained roads). It’s great and very convenient that BRR also tested it as a gravel tire, which makes the comparison easier. Here are the corresponding results:
- Schwalbe Pro One TLE Addix 34 x 700c: with 3.0 bar: 0.00465 crr, 15.5 Watts
- Schwalbe Pro One TLE Addix 34 x 700c: with 2.3 bar: 0.00567 crr, 18.9 Watts
This places the tire high in the ranking of all tested gravel tires, but it still only ranks fifth. It falls slightly behind four other slick tires in widths ranging from 32 to 44 mm, including the Conti GP5000 S TR32 and the René Herse Snoqualmie Pass TC Extralight 44.
But this is very convenient because I have personally ridden the Schwalbe Pro One 34 for the first time in the Mittelgebirge Classique last year and I like it. Mounted on my Nextie Premium AGX rims, it should be a (aerodynamic) force, at least in terms of providing the conditions for minimal flow separation, especially considering the almost constant crosswind a cyclist experiences. The 105% rule is often used here, which suggests that the outer width of the rim should be at least 105% of the tire width. I already had this tire and I also wanted to return my Drifter to a more road-oriented configuration anyways. Therefore, I formed a third setup with the Schwalbe Pro One 34 and tested it afterwards.
Wattage loss of the tires according to BRR test results
But first, let’s go back to the coefficient of rolling resistance or the power loss in watts applicable to one tire at a speed of 29 km/h. That means we need to double the corresponding watts or differences for each of our two tires on the bike. Here are the BRR results I mentioned earlier with the pressures that come closest to my setup, and I’ll also include a fifth tire:
- Continental RaceKing ProTection 2.20 x 29: at 1.7 bar: 0.00471 crr, 15.7 watts
- Schwalbe G-One R Super Race 40 x 700c: at 1.9 bar: 0.00827 crr, 27.6 watts
- Schwalbe G-One RS Super Race 40 x 700c: at 1.9 bar: 0.00591 crr, 19.7 watts
- Schwalbe Pro One TLE Addix 34 x 700c: at 2.3 bar: 0.00567 crr, 18.9 watts
- Maxxis Ikon 3C MaxxSpeed TLR 2.20 x 29: at 1.7 bar: 0.00809 crr, 24 watts
What stands out? The Continental RaceKing, as an MTB tire (supposedly the category with the worst rolling characteristics), is by far the fastest tire among all the tires listed! Although the Continental RaceKing is not a heavily treaded mud or downhill tire, it is an XC racing tire. However, it is robust and race-proven, not only for short races on technical XC courses but also for long and challenging off-road bikepacking races such as the Tour Divide or the Atlas Mountain Race. And it definitely has more knobs than a Schwalbe G-One RS or even a Schwalbe Pro One. The latter being a pure slick and primarily a road tire.
I also included the Maxxis Ikon as another representative of MTB tires to show that it is not automatically true that MTB tires, even when designed for hardpack and cross-country use, roll faster or lighter than gravel or road tires. The Maxxis Ikon is also a popular and reliable tire used in bikepacking races. I personally rode this version during my first Atlas Mountain Race and had no issues and was very satisfied on ride feel as well. However, I was aware that it was not the fastest tire. This is clearly evident here. Nevertheless, it is still faster than the Schwalbe G-One R, and it has a larger volume and is intended for much more demanding terrain than the G-One R to boot!
Let’s go back to the Continental RaceKings. At 29 km/h, a pair of these tires requires a power output of 2 times 15.7 = 31.4 watts to overcome rolling resistance (on firm ground or on the roller test stand).
In the case of my Schwalbe G-One R, it would require 2 times 27.6 = 55.2 watts! As I always say, tire rolling resistance plays a significant role! Thus, the Drifter is „worse“ than the Exceed by 23.8 or almost 24 watts just because of the choice of tires. From the Aerotune test results above, we were able to deduce an overall advantage of 11 watts (at 30 km/h) for the Drifter. If the full 24 watts are considered as a tire penalty for the Drifter, it means that the Drifter would have to be aerodynamically 35 watts faster than the Exceed.
What I keep saying: there is a lot of music in the tyre rolling resistance! 24 watts difference, simply because of the choice of tires!
This aerodynamic advantage should result from the aforementioned reasons: body position (not significantly different, but still, a 3 cm lower handlebar and a 2 cm narrower handlebar on the Drifter) and, above all, from the fork and the tires or wheelsets (larger frontal area and worse form factor with MTB tires and MTB wheelsets).
Let’s take a closer look at that and expand the tests to include the mentioned third setup with the Schwalbe Pro One in a 34 mm width. We will also use the Chung Aerolab in Golden Cheetah, specifically the Chung Virtual Elevation Method, as an additional tool from now on.
Test Extension: Setup 3 and Tire Pressure Excursus
The third setup consists again of my Drifter. For the third day in a row, with stable weather conditions, at the same time of day. Everything remains the same, including the wheelset. Only difference: I removed the G-One R tires and installed the Schwalbe Pro One 34 tires instead. Also tubeless, with approximately 55 ml of sealant in each tire.
Unfortunately, the rear tire caused me some trouble. It didn’t seal properly (converted the night before) and had a slow leak at the valve. This is certainly not ideal and doesn’t provide stable and reliable conditions for an accurate test. Fortunately, since I planned to do only 3 test laps per day and only one day per setup (as mentioned in the introduction), I wasn’t aiming to determine precise and absolute crr (coefficient of rolling resistance) values anyways. However, I did want to find out if and how much faster my Drifter, with the Schwalbe Pro One tires, would be compared to the Schwalbe G-One R tires, assuming similar aerodynamics. Are the differences observed in my test consistent with the rolling resistance tests conducted by BRR?
So, Setup 3 consists of my No 22 Drifter, tested with the Nextie Premium AGX 45 wheelsets as before, and equipped with Schwalbe Pro One TLE tires in 700 x 34c size, inflated to 2.5 bar in the front and 1.6 bar in the rear. Weight including Garmin: 8720 g. Total weight on the test day (rider, kit, bike with filled bottles): 80.59 kg
Exkurs Reifendruck:
Tire Pressure Excursus:
By the way, optimal tire pressure was not the focus of this test. I didn’t conduct various test runs with different tire pressures to find the optimal pressure for the test surface and determine the lowest overall rolling resistance resulting from hysteresis and impedance losses. It would have required additional trial runs in advance. Therefore, each of the tested tires might perform even slightly better or worse on my test course. The differences between the tires could converge or become more pronounced. For instance, if I had inflated one tire too hard for the test course and another tire much too soft. However, I don’t expect significant variations. Firstly, I ride each tire with adjusted pressures that fall within a comparable range based on recommendations from tire pressure calculators (one good example is the one by Silca). Secondly, while I prefer lower pressures, I wouldn’t want to go significantly lower even if it resulted in slightly better rolling resistance. That’s because what good is it if it compromises cornering stability and precision or causes the rear wheel to bounce up and down with the slightest imperfection in pedaling due to excessive softness? If I wouldn’t ride a particular tire significantly differently on the corresponding terrain, I need to determine the corresponding rolling resistances and rankings accordingly.
Aerotune Results with Constant Rolling Resistance
For convenience sake, I will once again present the Aerotune results for Setup 1 (Exceed with RaceKing) and Setup 2 (Drifter with G-One R), along with the addition of Setup 3 (Drifter with Pro One).
Again, the rolling resistance coefficient is set as a constant value of 7 rP (rolling resistance points) or crr 0.007.
The ranking of the setups is as expected. Since the rolling resistance was kept constant, any advantage for Setup 3 will solely be reflected in the cDA (drag coefficient) or the aeroPOINTS (aP), and it does.
With 43 aP, the Drifter with the Pro One tires is on average 2 aP faster. Considering the error margins, the following list emerges:
- Setup 1, Exceed + RaceKing: 48 (+-4.6%) –> 45.8 – 50.2 aP
- Setup 2, Drifter + G-One R: 45 (+-4.9%) –> 42.8 – 47.2 aP
- Setup 3, Drifter + Pro One: 43 (+-4.9%) –> 40.9 – 45.1 aP
As mentioned before, it is possible that all three setups achieve the same „true“ values. None of the setups falls completely outside the error margin of the others. However, it is still expected that the true values will be around the indicated 48, 45, and 43 aP (as the overall result of cDA and rolling resistance).
In terms of power requirements, we can read the following values for 35 km/h:
- Exceed RaceKing: 332 watts at 35 km/h
- Drifter G-One R: 314 watts at 35 km/h
- Drifter Pro One: 302 watts at 35 km/h
This is a significant difference between Setup 1 and Setup 3, between the Exceed and the Drifter. There is a 30-watt difference at 35 km/h and still an 18-watt difference when converted to 30 km/h between the two setups. So, even if the Conti RaceKing has excellent rolling resistance, the Drifter makes up for it aerodynamically. This is despite the fact that I am positioned more aerodynamically on the Exceed (not at all comparable to a flat-bar mountain bike) and despite the potential slight disadvantage of the Schwalbe Pro One rear tire, which may have lost some of its low rolling resistance due to its lower pressure of only 1.6 bar.
However, one thing is clear: if I were to mount less efficient tires on my drop-bar Exceed within the same range of rolling resistance as the Schwalbe G-One R gravel tire, the differences would be even greater.
Experimental Aerotune Results with Adjusted Rolling Resistance
I attempted to achieve meaningful absolute cDA (drag coefficient) values in the Aerotune model results by varying the rolling resistances within the range expected from the BRR tests. One goal was to reach an equal cDA value for both Drifter setups. It is likely that Setup 3, due to the even better tire-rim interface aerodynamics of the Pro One 34 mm with the Nextie rims, will achieve a slightly better cDA. However, I cannot confirm this or determine the significance without further test runs.
Here are my results:
- Setup 1: RaceKing set at crr 0.0047 –> 52 aP & 336 watts at 35 km/h
- Setup 2: G-One R set at crr 0.0083 –> 43 aP & 313 watts at 35 km/h
- Setup 3: Pro One set at crr 0.0073 –> 43 aP & 304 watts at 35 km/h
As you can see, when manipulating the values, the order of magnitude remains the same, but due to the necessary equation solving in the model, it doesn’t perfectly match the exact wattage values. Despite using the same recorded Garmin data, Setup 1 now shows 336 watts instead of 332 watts. Conversely, Setup 2 now shows 313 watts instead of 314 watts. And Setup 3 shows 304 watts instead of 302 watts. However, these differences are within an acceptable range and much smaller than the overall error margins that need to be considered.
This approach provides a more realistic picture of the differences attributed to rolling resistance and aerodynamic drag. However, it can still be doubted whether it is a precise representation, especially regarding the breakdown of the components.
Analysis with the Chung Virtual Elevation Method
The Aerotune website is cool, but as an end user, it’s essentially a black box for me. I can’t see the detailed effects and variations along the length of my test or measurement route. During the test runs for this article, I also found that Aerotune seemed to have gotten, shall we say, a bit more temperamental. While I had no issues with ingesting my Garmin and Wahoo files during my bikepacking bag test 2 years back, many of my test runs during this particular test were not considered valid by the Aerotune website. Out of good reason, do I have already extensively discussed the advantages and disadvantages of both methods (Aerotune and Golden Cheetah Chung Aerolab) in this article.
Therefore, although it is more time-consuming, delving into the Chung Aerolab in Golden Cheetah provides even more insightful results.
Without referring to the Aerotune results, I entered the relevant parameters into the Chung Aerolab (air pressure, total weight, 2% drivetrain loss (eta 0.98)), and performed multiple visual adjustments. Since it is not a loop course but rather an out-and-back route with turnaround and braking sections, there can be no automatic best-fit adjustment for the entire route occuring within the Chung Aerolab. You can do it if you choose a continuous loop where you must’nt brake. For an out- and back course each test section must be examined separately (this can be done nicely using the EOffset field).
Here is an example for Setup 3 with the Schwalbe Pro One, and in the image, I am evaluating the parameter fit for lap 38:
Differentiating slightly higher rolling resistance coefficients and slightly lower aerodynamic drag is difficult to determine. Therefore, I provide two sets of values for each setup. If one deviates significantly from the presumed real distribution between crr and cDA, the virtual blue model curve cannot be overlaid on the differently paced test laps. It only fits well for either the slower or faster sections, but not both. Therefore, the Chung Virtual Elevation Method is still the method of choice when it comes to rolling resistance alone or primarily (more details in the previously linked article on aerodynamics testing fundamentals).
Here are the Chung Aerolab results:
- Setup 1, Exceed + RaceKing:
- Aerotune result: 52 aP for 4.7 rP input
- Chung result from: 47 aP for 4.7 rP
- Chung result to: 44 aP for 5.7 rP
- When using higher crr values, e.g., 6.7 rP and 41.6 aP, I cannot achieve satisfactory agreement for the fast out-and-back lap when adjusting for the slow laps and vice versa.
- Setup 2, Drifter + G-One R:
- Aerotune result: 43 aP for 8.3 rP input
- Chung result from: 39 aP for 7.8 rP
- Chung result to: 38 aP for 8.3 rP
- Setup 3, Drifter + Pro One:
- Aerotune result: 43 aP for 7.3 rP input
- Chung result from: 39 aP for 5.6 rP
- Chung result to: 38 aP for 6.7 rP
Insight #1: The overall rolling resistance values derived through visual adjustment (and verified or at least evaluated based on the data) are within the same order of magnitude as the experimental evaluation with Aerotune.
Insight #2: Although the method used here aims to determine the total rolling resistance (hysteresis + impedance), it aligns quite accurately with the rolling resistance coefficients determined through roller testing by BRR. Consequently, the ranking of the test results from BRR remains intact.
Insight #3: The final Aerotune results for aerodynamic drag are significantly higher for each setup compared to the results from Chung Aerolab. This is somewhat surprising and may be related, on the one hand, to the modeling of the wind component in Aerotune. In the Chung Aerolab, wind is unknown, but gusts of wind or tailwinds/headwinds would still be reflected in the integrated blue virtual „elevation“ line and must be interpreted comparatively. Additionally, the tests were conducted on days with low wind, although an influence can never be completely ruled out (another reason for conducting more than three test runs per setup, ideally on different days for the same setup). Other systematic differences could be attributed to varying approaches in drivetrain losses. However, both models assume a 2% drivetrain loss input.
But Insight #4: The relative cDA differences between the setups are very similar, and that’s what matters to us. According to both methods, there are either 9 aeroPoints (in the Aerotune model) or 6-8 aeroPoints (in Chung Aerolab) between the Drifter and the Exceed.
For a more precise analysis, I would suggest two things:
1) Conduct at least one additional test day and three more test runs for each setup.
2) Install aerobars. They greatly help maintain consistent body position and directly lead to lower error margins: both through repeatable body positioning and higher test speeds at the same power output, resulting in higher resolution of the desired parameters.
Conclusions regarding rolling resistance (and another data point from the literature):
Fast XC MTB tires are fast! Faster than some performance road or gravel tires.
On smooth, unpaved, hard bike path surfaces with occasional light gravel, a high-quality 29er mountain bike tire (such as the Continental RaceKing Protection, which ranks in the top 3 out of 49 tires tested on BRR) is faster or at least as fast as a good road bike tire or gravel slick tire. The same can be assumed for asphalt surfaces.
My Schwalbe Pro One TLE (in the 25mm version) ranks 27th out of 126 road bike tires tested on Bicyclerollingresistance.com, and my exact model (in the 34mm width) has also been tested as a gravel tire, ranking 5th out of 75 gravel tires. For comparison, the Conti GP5000 TR 28mm ranks 8th out of 126 road bike tires, while the Conti GP5000 TR 32mm, which was only tested in the gravel tire category, ranks 1st out of 75 gravel tires.
The rolling resistance coefficients (crr values) obtained from these tests are determined for all tire overview lists (MTB, gravel, and road bike tire lists) using the same testing procedure but tested at pressures adjusted for the tire’s intended use and construction. In other words, the top crr values of a road bike tire cannot be directly compared to the top crr values of an MTB tire because different pressures were used. However, the relative ranking of rolling resistance test results is largely unaffected by the tested pressures. The increase in rolling resistance on the roller decreases similarly for different tire models as the pressure decreases. There may be some convergence between two models at lower or higher pressures, and yes, there may also be the occasional close switch up in rankings. Overall, however, the ranking remains relatively constant. This means that a tire that rolls excellently at 3 bar pressure will not suddenly drop significantly in relative performance at 2 bar pressure. At the same – a poorly rolling tire will not suddenly become a star performer. The decisive factors of tire construction, rubber compound, and tread design remain constant regardless of the pressure used.
Another aspect that should not change in regard to the ranking (at least I am not currently aware of any corresponding analyses) is the contribution of impedance losses to the overall rolling resistance coefficient (see my tire article). These losses should behave similarly to pure hysteresis losses (which are measured on the drum) across different tire models. This means that a tire that rolls faster on the drum should also be faster in reality on rough (yet still hard) surfaces. My test results confirm this.
They also show that I apparently don’t need to add much, if anything at all, to the crr values obtained from drum testing at the relevant chosen low pressures by BRR in order to achieve consistent results in both the Aerotune and Chung models. However, without further tests on different surfaces, I cannot quantify this.
Fast remains fast, regardless of the surface mix.
For truly rough surfaces, including deep surfaces like grass, followed by forest floor and then a gravel road, the results are different. Maier et al. used the Chung model to investigate the rolling resistance of 9 different 29er XC tires over a test lap on these surfaces.
For each tire, 5-7 test laps were ridden by one rider, and the Chung model was then evaluated by 3 different individuals who were blind to the tire being evaluated. The authors found that the Chung Virtual Elevation Method was highly reliable (typical error = 0.0006, 2.8%; limits of agreement <0.0005, r ≥ 0.98). The average crr value obtained for all 9 tires was 0.0219, ranging from 0.0205 for the best tire (Continental Race King) to 0.0237 (Ritchey Shield): Thomas Maier, Beat Müller, Lucas Schmid, Thomas Steiner & Jon Peter Wehrlin (2018), Reliability of the virtual elevation method to evaluate rolling resistance of different mountain bike cross-country tyres, Journal of Sports Sciences, 36:2, 156-161, DOI: 10.1080/02640414.2017.1287935
A crr value of 0.0205 for a mixed XC course with forest floor, gravel road, and grass sections, compared to my crr of 0.0047 to 0.0057 on smooth hardpack (mineral-bound bike path surface). In the more easily interpretable rollingPOINTs of Aerotune, this corresponds to 20.5 rP for off-road terrain and 5 to 6 rP for an unpaved bike path. These are the order of magnitude differences. But fast remains fast, whether on the drum or on a cross-country course. Whether we consider only hysteresis (tire material properties) or hysteresis and impedance losses (the energy loss from constantly accelerating on rough surfaces and the damping of these accelerations in the bike and rider’s body) combined, the conclusion is the same. But of course: for determining overall losses and calculating the required power for a given course, it is necessary to consider those impedance losses, too. I.e., the total rolling resistance. Which – just to be clear – you already and always get when you test a tire out in the field with the methods I described and used here.
But – when just choosing one tire over another, relying solely on test results from websites like Bicycle Rolling Resistance is sufficient if the other tire properties you are after meet the desired intended use.
Fast on the flats, fast in corners, fast on rough terrain – where’s the catch?
The Schwalbe G-One R gravel tire is significantly slower than both the Conti RaceKing mountain bike tire and the Schwalbe Pro One TLE 34mm road or all-road tire. This is consistent with my results and is reflected in its position in the lower third of the Bicycle Rolling Resistance (BRR) roller test results for gravel tires, ranking 52nd out of 75.
Of course, this does not provide information about other qualities and criteria for selecting tires beyond rolling resistance, such as tread pattern (slick vs. knobby), suitability for dry weather or muddy conditions, exceptional high durability, etc. It also does not account for the riding feel and cornering characteristics. However, it does show that simply going by looking whether a tire has knobs or is a slick or whether it is a narrow road tire or not or if it is a mtb tire or not just doesn’t work to find a good performing or good rolling tire.
The G-One R gravel tire (like most other knobby gravel tires) falls behind in terms of rolling resistance. This is evident from roller test results as well as my own tests.
But also in corners, despite the decent tread pattern that provides a secure feeling, the G-One R with its 40mm width does not give me the confidence and therefore the the speeds that I can achieve with the Conti RaceKing. It is of course difficult to quantify this, as it is based on my personal impressions and cornering skills. I am specifically referring to typical forest or fire road corners on level ground with a hard surface and some loose gravel or stones on top. These are situations where you cannot lean into a washed-out groove or an upward sloping shoulder of the path but instead rely solely on the cornering grip that you can generate with your tires and apropriate weight distribution.
Additionally, there is a further unquantifiable disadvantage of narrower gravel tires (and the overall gravel bike system compared to a drop-bar MTB) in terms of rough downhill capability. Where you most probably would find yourself holding back or stay quite a bit more on the brakes just because it’s too bumpy to stay in control or you just are afraid of destroying your tires, your rims, your wrists or all of the above.
So, is it all green lights for the adventure bike or the 29er with MTB tires? Maybe choose the same frame geometry and narrow handlebars as a (race) gravel bike, forego the suspension fork, and suddenly you’re faster everywhere than with a 35 to 45mm gravel tire? Depends.
One thing that cannot be changed and remains an aerodynamic and weight advantage of road bikes over gravel bikes, and even more so over mountain bikes, is the frontal area of the tire, its weight, and the ability to design an aerodynamic tire-rim interface.
With a road bike (and a gravel bike with limited tire widths of about 35mm and not too high of a „Knob-ness“ of the tire), it is absolutely no problem to design and purchase aerodynamically efficient rim profiles to match the chosen tire width. While the selection is growing only slowly, finally we see movement there. See my article on this topic: Breite ist Trumpf – Hochprofil-Felgen und Aero-Laufradsätze für Allroad- und Gravel).
However, when it comes to MTB tires with widths of 2.1 inches and above, this is no longer feasible or practical. An aerodynamically efficient rim profile would need to be both very wide, which does not harmonize with the common mtb tire cross-sections and their intended use and construction, and very deep. Roughly estimated, this depth would need to be at least 100mm, which would be very unusual and unwieldy. It would also contradict the qualities sought for typical off-road environments (compliance, durability, lightweight).
So there is not much (if anything at all) we can do to this piece of the puzzle for MTB tires. The aerodynamic disadvantage of the tire-wheel system will be a factor on long and mostly flat stretches and thus a disadvantage. In my testing here, I didn’t single out this effect but examined and quantified this for the entire drop-bar MTB system (hardtail with suspension fork).
The complete system, including an MTB tire (2.20 x 29″), MTB wheels, and suspension fork, but with a similar (though not identical) riding position, is approximately 8 aeroPOINTS = 0.08 m² slower in terms of cDA than my Drifter Gravel bike. And that’s a significant difference.
However, it is easy to offset a good portion of this penalty by choosing a good XC MTB tire, especially when the comparison gravel bike hasn’t mounted the top tier of the best-rolling gravel tires, which often come with significant compromises in terms of puncture protection and grip (both aspects no problem for the MTB tire competitors simply already due to their larger volumes). It makes no sense, for example, to choose a narrower 32mm Conti GP 5000 S TR simply because it ranks higher in the low rolling resistance category on BRR when you really think you need a tire like a 40mm G-One R as the suitable tire for a chosen route. That 32 mm Slick just wouldn’t be able to offer the same level of puncture protection and/or grip. On the other hand, choosing the 57mm-wide Conti RaceKing Pro with knobby tread instead of the G-One R would be a no-brainer if the bike allows it.
Conversely, it also doesn’t make sense to prefer the Conti RaceKing if a 40mm G-One R would have been overkill from the get go. In that case, the 32mm slick tire from Conti might be the better choice as it rolls just as fast as the RaceKing and forms a lighter and more aerodynamic overall package with a race gravel bike. Yes, this tire could even be used with a current-generation aero road bike. And that aero road bike might actually be the right tool for the job if it would be for a course like last year’s UCI Gravel World Championships.
This is what makes tire selection so exciting. And one should widen his envelope and definitely consider 29 inch MTB tires instead of just 700c Gravel tires.
With the Conti RaceKing, the pure aerodynamic disadvantage of my Exceed compared to my Drifter is reduced from 8 aeroPOINTS to only 3 aeroPOINTS when considering the tire rolling resistance between Setup 1 and Setup 2 in terms of cDA.
In addition, I get better cornering grip on hard-packed trails and much more confidence on typical forest paths and other off-road terrain!
In terms of required power, these 3 aeroPOINTS translate to an additional 18 watts at 35 km/h on flat terrain. At 30 km/h, it’s still nearly 11 watts that I have to exert on the Exceed compared to the Drifter to maintain the same speed. These values apply to a smooth, unpaved bike path with a light covering of gravel and sand.
You can always become more aero…
Can I have both? The advantages of a super fast-rolling 29″ MTB tire and but also the low aerodynamic drag of a (race) gravel bike?
By the way, why do I always put the word „race“ in parentheses before „gravel bike“ in this context?
We need to be aware that all the numbers mentioned here in terms of their exact magnitude and the difference between the setups only apply to my exact two bikes, the Canyon Exceed in my drop bar conversion and my No 22 Bicycles Drifter. Furthermore, they only apply to these two bikes and me as the rider on top. However, you can transfer the general conclusions to yourselves. But it is also clear that many more gravel bikes nowadays follow a „utility“ approach with numerous attachment points all over the bike, increasing tire clearance, and a tendency to use wider „gravel handlebars“ with a monster flare. If you’re sitting on such a bike, maybe even with quite a few spacers under the stem and you feel comfortable on it… then everything is great. But In that case, you don’t need to worry about the supposed disadvantage of aerodynamics on a drop bar MTB with suspension fork and MTB tires. Chances are that I am much more aerodynamic on my drop bar MTB than you are on a utility gravel bike as described above. And if you have fork-mounted bags attached to your gravel bike and never bother to remove them for convenience, then it’s even game over anyways (Not saying that fork mounted options don’t have their place – they have). But if that is you on the comfort utility monster wide monster flare handlebar-ed gravel bike you’re probably not even reading this text anyways. ;-)
That’s why, when making such comparisons in reference to my Drifter, I nowadays refer to it as a (race) gravel bike. Not because it’s uncomfortable or uncompromising in any way. But because it’s a sleek and agile gravel bike. It has a classic titanium frame, adequate tire clearance but not an excessive one, a slim fork, 0 cm of spacers (a result of fittings, a custom frame and tried and tested over numerous bikepacking races across Europe) and the resulting saddle to handlebar drop. Keep that in mind during these comparisons.
How close can I get to this benchmark with a drop bar MTB and suspension fork?
1.) I would need the exact same body position.
That’s easily achievable. I’m almost there already. And it feels just right when riding (as is to be expected) when I flip my stem down on the Exceed and ride with the same saddle to handlebar drop as on the Drifter. However, I intentionally chose a slightly lower saddle to handle bar drop to have a slightly more upright position for off-road or forest trail descents, which are typically steeper than road descents. And for a sense of safety and better overview. For a similar reason, I also opted for a slightly (2 cm) wider handlebar compared to the Drifter. But definitely not a monster flare handlebar. I despise those. If you really need one, a flat bar would serve you better. Confidence in a descent – if that’s the reason for such a monster flare design – should come from the geometry of the bike (wheelbase, head tube angle, trail, etc.), appropriate tire selection, and also suspension will help. Trying to achieve it solely through handlebar width, in my opinion, is the wrong approach. But tastes vary so if you like your monster flare handle bar – fair enough. I dare to say we see so many of them because brand managers think that’s the right way to promote a bike as gravel bike and newcomers to the booming gravel trend are not consulted here and are presented with such a stereotype or pre-configured bikes that they simply adopt and don’t question. But I digress. The bottom line is that the rider accounts for the largest proportion of the total air resistance (often mentioned as 70-80% or sometimes even between 60-90%), and that’s where the greatest leverage lies. Achieving identical body positions on the drop bar MTB and the (race) gravel bike is easily possible. However, the desire for a tiny adjustment towards a slightly more upright position is understandable and meaningful.
2.) The suspension fork. How can an aerodynamic suspension fork look like?
If I want to have a suspension fork (and don’t want to swap it for a rigid fork, which would actually be quite easily done with an externally routed brake line on the suspension fork – so you can this very well regard as a temporary option), then I need to work on the concept of an aero optimized suspension fork. A standard suspension fork with its cylindrical stanchions and lower legs is absolutely not optimized for aerodynamic drag. Additionally, there are two bridges (the main fork bridge and the connecting bridge of the lower legs). Not ideal.
Option 1: Use an upside-down fork. This has its pros and cons. The mountain biking community seems to have decided what makes most sense for them, as upside-down forks are rare. However, they eliminate one of the two bridges. But, if the lower (or now upper) legs or the stanchions, too, become even slightly wider compared to a normal fork to compensate for stiffness losses, it might result in a neutral aerodynamic effect. Or even a net loss. So, maybe an upside-down fork doesn’t help after all in regard of aerodynamic drag.
Option 2: Simply omit one of the fork legs. This should nearly halve the aerodynamic resistance. In this light, Cannondale Lefty forks gain even more charm as they already have. I’m not sure if anyone has ever tested a Lefty fork against a normal suspension fork in a wind tunnel or with Chung Aerolab tests to determine the cDA differences to a normal suspension fork. Another advantage of Lefty forks is their super-sensitive response. This is achieved through the internally triangular stanchion and low-friction needle bearings guiding it. Unfortunately, weight is not one of its advantages. One might think that weight would be reduced since one fork leg is missing. However, normal suspension forks don’t contain much material internally, and the load can be distributed on both sides more easily and with less material. As a result, a Cannondale Lefty Ocho (XC fork with 100mm travel) weighs 1570g (not bad), while a RockShox SID Ultimate Race Day (also an XC fork with 100mm travel) weighs only 1340g (class-leading). Another disadvantage is the need for specific hubs, which means no hub dynamo is possible. Additionally, if one wants to use the fork legs for bikepacking purposes, only one is available. However, from an aerodynamic and overall fork performance perspective, it would be interesting to use and test a Lefty fork.
Option 3: A linkage fork. In this design, the damping cylinders are smaller and operated by levers. This also eliminates or minimizes the influence of braking on the suspension, if done well. There are occasionally wild concepts that emerge, although (almost) none of them are designed with aerodynamic advantages in mind. However, this offers the greatest potential to combine suspension and good aerodynamic properties. Of course, one could also aerodynamically cover the lower legs of a normal telescopic fork. That would bring significant improvements, but it would also look very unusual, and the necessary width of the lower legs and stanchions cannot be avoided. But looking unusual is also true for linkage forks.
Or would they really look unusual? Various patent applications from Sram have emerged, including a Gravel Aero Suspension Fork. Sram literally aims to develop a more aerodynamic and aesthetically pleasing suspension fork, they say. Hell yeah – I’m curious to see what might come out of it. And what kind of travel it will provide. As the patent drawing shows, the linkages are positioned quite high in the fork legs and relatively far apart, connecting two rather conventionally looking, slightly flattened fork leg parts.
In the end, this could offer exactly what Sram promises. However, I’m afraid they will focus too much on the „Gravel“ side of thinking. That means they might provide only 30 to 60mm of travel maximum. Well, well-executed 60mm… It’s cutting it close but could be a good compromise. 80mm would be ideal, and 100mm would probably not be necessary nor achievable with such a design. If you want to delve deeper, Wheelbased (from whom I also linked the Instagram post) has written a great article about this.
Option 4: A „Headshok“. Once again, we come to Cannondale, who popularized and marketed this type of front suspension as the Headshok in the 1990s. The suspension and damping occur entirely within the head tube and slightly below it, while the lower part of the fork is essentially an ordinary rigid fork. Thus, it can do everything that normal forks can do. It can be aerodynamically designed, have attachment points for bags, and use hub dynamos, etc. My vintage Cannondale Super V 900 full sus, which I still own, has 60mm of front suspension provided by such a Headshok. For me, it’s the ideal gravel suspension principle. Apparently, not only for me, as there are now manufacturers that produce this type of suspension again, although with (too) limited travel. BMC, for example, uses it in some models of its URS. I have written more about this in a separate article: Höher, Schneller, Breiter? Ein Blick auf aktuelle Graveltrends, Gravelbikes und Federungssysteme.
3.) MTB tires and MTB rims.
Option 1, the body position is by far the most important factor. Whether the aerodynamic penalty of a typical XC suspension fork is greater or smaller than that of a typical MTB wheelset with XC MTB tires, I cannot say. My gut feeling is that the MTB wheelset is perhaps the greater, or at least equally significant, issue as the suspension fork.
Unfortunately, there is likely little we can do about it. We do want the specific tread pattern and we do want that width of the chosen tire, after all. To achieve an aerodynamically effective overall profile, starting from, for example, a 2.20-inch tire width (57mm) with potentially protruding side knobs, the rim would have to be nearly as wide (or better a bit wider still – so a whopping 60 mm at least) and it would probably need to extend at least around 100mm in depth. However, I doubt that such a wheelset would be desirable to ride or possess the qualities expected from an MTB wheelset designed for rough terrain.
An interesting exercise would be to compare my Nextie Aero gravel wheelset with a 40mm gravel tire or even the 34mm Pro One on the one side and the Conti RaceKing 2.20″ tire on a typical MTB rim like my Newmen Advanced on the other side in the same bike. In order to isolate the influence of the tire-wheelset system. Unfortunately, someone would need to provide me with a wheelset consisting of Nextie rims laced to MTB Boost hubs, or someone would need a gravel bike with tire clearance for 2.20 x 29″ and mount the RaceKing on standard box rims with hubs in road standard to compare both systems on a gravel bike.
But maybe there are test results already available, even in a wind tunnel, which could point us in the right direction? How much difference does option 3, the MTB tire-wheelset system compared to a good aerodynamic gravel tire-wheelset system, make? There are. Not for our exact question but close enough. I refer to the highly interesting wind tunnel tests and analyses conducted by Swissside: Performance Analysis Of Swiss Side GRAVON Gravel Wheels And Tyre Setups (January 2022)
You may know that Swissside and its founder, Jean-Paul Ballard, are practically the aerodynamics kings in cycling. They are the only ones, of course. But they are really „in the know“. They are also behind many developments for numerous manufacturers (DT Swiss, Canyon, Cube, and many more) and provide support to numerous world-class athletes in road cycling and triathlon.
Regarding the mentioned article, it is important to note that Swissside is, of course, dependent on keeping up with the current state of technology and market realities. They probably know that a Nextie Premium AGX rim with its 40mm outer width and 45mm depth provides an even more interesting tire-rim interface and is very likely even more aerodynamic than their Gravon 420, which they tested as the „best case.“ However, they sell the Gravon 420 and currently have nothing better. I wondered for a few minutes when the Gravon 420 was introduced why they didn’t present something more exciting until I realized: of course, they developed it in cooperation with DT Swiss and have them produce it. DT Swiss also has their GRC 1400 SPLINE 42, which Swiss Side sells as the Gravon 420. I’m not familiar with the details of that cooperation, but fair enough!
Both wheelsets are undoubtedly nice, robust, and, as one would expect from DT Swiss, highly compatible and easily convertible to various freehub standards. However, even for 2022, they fall significantly short in terms of effective outer width, with it only being 32mm wide (see also: Breite ist Trumpf – Hochprofil-Felgen und Aero-Laufradsätze für Allroad- und Gravel).
So all this must be taken into account when the Swissside article speaks of the „best combination“ and when comparing this best gravel tire-wheelset system in the first part of the article to a very good road wheelset system (25mm Conti slicks on a HADRON2 Ultimate 625 rim with 30mm outer width). So, 25mm tire on a 30mm wide rim (superior cross-section, excellent aerodynamic behavior) compared to a 40mm tire on a 32mm wide rim. It’s clear what the result will be. One could easily ride a 35mm gravel tire on a 40mm wide and 45mm deep rim (or develop and test even more fantastic rim profiles—Swissside has all the tools for that. However, they also need a viable business case understandably so). Well, they didn’t actually test this configuration I just described because they wanted to compare a significant difference between an aero road setup and a standard gravel setup (tires, wheelsets, and bikes). Therefore, they used the Gravon 250 wheelset.
Now, the second part of the article is much more interesting! They investigate various tire widths on the Gravon 420 wheelset and compare them to the Gravon 250 wheelset. The Gravon 250 is a standard aluminum wheelset with modest 25mm depth and 28mm outer width. It closely resembles a typical MTB wheelset – great.
Furthermore, further down in the article, Swissside tests differently treaded tire models of comparable width and determines the power losses caused by the „knob-ness“ of different gravel tire profiles. Also super exciting.
The aerodynamic difference between a 35 mm tire with a very smooth tread (Schwalbe G-One Speed) on the 32 mm wide Gravon 420 rim and a 45 mm wide tire with small and uniform knobs (Schwalbe G-One Allround) on the Gravon 250 rim is 5 watts at 30 km/h!
Swissside
Taken together, the answer to my question is: The aerodynamic difference between a 35mm tire with a very smooth profile (Schwalbe G-One Speed) on the 32mm wide Gravon 420 rim and a 45mm wide tire with low and uniform knobs (Schwalbe G-One Allround) on the Gravon 250 rim is 5 watts at 30 km/h! That is a already a significant difference. And: it’s measured for a system difference which is still less pronounced than we are trying to determine here. It can be expected that the drag result for the Gravon 420 wheelset with the 35mm tire is not as good as it would be with the my Nextie wheelset using the same tire. On the other hand, the 45mm tire is still quite far from the 57mm width of a 2.20 x 29″ tire. Based on the values provided by Swissside regarding the influence of tire width and „knob-ness,“ one could conclude that an additional 2 to 4 watts has to be added to the 5 watts of the Swissside comparison to account for the still more pronounced difference we are looking for in our case.
Therefore, the sought-after difference between an optimized gravel aerodynamic wheelset and an MTB tire-wheelset system is likely in the range of at least 5 to 9 watts or even more at 30 km/h. 9 watts at 30 km/h correspond to 2.6 aero points (aP) when converted to cDA. In total, I identified 8 aP as the aerodynamic disadvantage of my Exceed compared to my Drifter. That means 5.4 aP would be accounted for by options 1 and 2, body position and suspension fork. And 2.6 aP would be attributed to the MTB tire and wheelset system.
As already shown, my Exceed gains back 5 aP compared to the Drifter through the lower rolling resistance of the tires alone. Contrary to the approximately 2.6 aP I wouldn’t be able to remedy. So in total – given all else equal (i.e. mainly body position) I would still be 2.4 aP faster on the drop-bar MTB. Even compared to a (race) gravel bike. Given the respecitively used tires with their respective rolling resistance. Furthermore, the topics of cornering speed, grip, and downhill performance have not even been accounted for yet, which adds further advantages to the drop-bar MTB.
However, if I pay a little more attention to choosing a tire with excellent rolling resistance on the (race) gravel bike, while neglecting doing the same on the MTB, the significant advantage in rolling resistance quickly evens out or even reverses! The aerodynamic disadvantages of the MTB, even if it’s a drop-bar MTB, remain. But the handling advantages remain as well, especially when it comes to rougher terrain.
Summary
I tested the aerodynamic resistance and rolling resistance in a comparison between a drop-bar MTB (or monster gravel bike, or 29er adventure bike) and a gravel bike using a total of 3 setups.
Specifically, I tested my Canyon Exceed, a hardtail MTB that I converted to a drop-bar setup, and my No. 22 Bicycles Drifter, a classic gravel bike with a titanium frame and gravel-aero wheelsets. The tested tires were the Continental RaceKing ProTection 2.20 x 29″ on the Exceed, and both the Schwalbe G-One R 700 x 40c and the Schwalbe Pro One 700 x 34c on the Drifter.
The testing was conducted on an unpaved, hardpacked, mineral-bound path with a light partial layer of sand and gravel. This route has been previously used in aerodynamic tests and is excellent for reliable comparison tests. I chose this route because I wanted to answer the question: How many watts am I unnecessarily losing with my drop-bar MTB on a good surface? This refers to a surface where the grip, suspension, robustness, or the width of an MTB tire or MTB itself is not truly required. So like on so-called „champagne gravel“ or even smoother surfaces, perhaps even asphalt.
The hypothesis (supported in this article with additional references and examples): the advantages of an MTB are evident on rough off-road terrain where you cannot ride the gravel bike as fast as you could with an MTB due to concerns about equipment or personal safety. Or you could, but you choose not to. Or you could and would but because of lack in control and grip due to narrower tires and no suspension you still end up slower. But how does it look like in terms of pure rolling resistance and aerodynamic drag on every surface and especially on good surface mixes like the aforementioned „Champagne Gravel“?
The analysis of aerodynamics and rolling resistance reveals:
The overall system of an MTB tire (2.20 x 29″) with MTB wheelsets and a suspension fork, along with a largely similar (but not identical) body position on both bikes, is approximately 8 aeroPOINTs = 0.08 m2 cdA slower. And that is a significant difference. It corresponds to an additional power requirement of around 28 watts at 30 km/h on the Exceed compared to the Drifter.
My own tests and a review of rolling resistance tests based on drum rolling resistance devices (Bicycle Rolling Resistance, BRR) show that very good (in terms of rolling resistance) MTB tires can easily compete with the top performers in the gravel tire test overview. Moreover, they are more practical for off-road use.
My own tests and additional tests from the literature confirm that although drum tests only measure the hysteresis losses of tires, the values obtained are highly applicable in real-world scenarios and remain valid even on rough surface mixes (gravel, grass, forest paths of an XC race course). Therefore, you can fully rely on these tests from e.g. BRR when selecting the tire with the lowest rolling resistance.
My Schwalbe G-One R tires require an additional 24 watts (at 29 km/h) compared to my Continental RaceKing in terms of rolling resistance.
Taking the excellent rolling resistance coefficient of the Continental RaceKing into account, the 8 aero points of pure aerodynamic disadvantage between setup 1 (Exceed + Continental RaceKing) and setup 2 (Drifter + Schwalbe G-One R) decrease to 3 aero points when considering tire rolling resistance in the cDA.
But in exchange (and addition), I gain better cornering grip on hardpack trails and much more confidence on typical forest paths and other off-road terrains! And thus additional free speed.
Converted to power requirements, these 3 aeroPOINTs mean an additional 18 watts at 35 km/h on flat terrain. At 30 km/h, it is still nearly 11 watts that I need to exert on the Exceed to maintain the same speed as on the Drifter.
Despite the aerodynamic disadvantage of having a suspension fork and the MTB tire-wheelset system, I can easily make up for the remaining 11 watts difference by incorporating the same body position on the Exceed. This is something that is easily achievable due to the minimal differences between the two setups. However, I intentionally omit this aerodynamic gain to be better prepared for the actual field of use of my Dropbar MTB for slightly rougher and steeper forest trails.
On the other hand, it would be more challenging for the gravel bike to select a tire with similar low rolling resistance to the Continental RaceKing without compromising puncture protection and grip. But with such a tire and typical gravel bike applications – namely „champagne gravel“ or #onRoadoffRoadRepeat – all the advantages are once again with the gravel bike. Therefore, on its intended territory, which I have often alluded to in my „the perfect gravel bike“ articles, the gravel bike is the ideal choice (no surprise!). However, with the current trend towards almost 100% off-road or rougher gravel riding, such a gravel bike no longer makes sense. Not in terms of comfort or ride safety, and as I have shown here, not in terms of the overall package of aerodynamics and rolling resistance. In those cases, all the advantages lie with the (XC) MTB, and depending on your preference or ridden trails, it may even be equipped with full suspension and/or a drop-bar.
With that, I wish you a lot of joy in being outdoors and on your bike – however it may look and be configured. And I look forward to your comments.
To finish it off, I have prepared a little bonus for you. If you want to follow my ideas and recommendations to be as performance-oriented as possible or if you don’t want your conversations to be interrupted by your huffing and puffing while riding side by side and chatting, while your conversation partner effortlessly breathes through their nose, or if you simply want to be the first to reach the cake.
Recommendations
- Take a hardtail MTB and convert it to a drop-bar setup. Choose a standard drop bar or one with only moderate flare (you can find some suggestions here.)
- Every cable or hose that can be eliminated (electronic shifting) or integrated is advantageous. For example, use Sram AXS wireless and route the rear brake hose directly into the steerer tube if the frame allows it. It’s not only more aero and looks more pleasing it’s also more convenient for use with bags when bikepacking.
- Pay attention to tires with low rolling resistance. You can find suitable options among MTB tires, and you won’t compromise much (if at all) in terms of robustness or grip, unlike you would have to do when choosing some top performers in the gravel tire rankings.
- For „champagne“ gravel, a gravel bike with a well-rolling, moderately wide gravel tire and matching rims is the faster alternative, without sacrificing any safety or comfort. In this case, it is advisable to choose e.g. a Schwalbe G-One RS instead of a G-One R (staying within the tire models previously mentioned). If the G-One RS is not considered sufficient (due to its less pronounced center tread), chances are that a drop-bar MTB with a tire like the Continental RaceKing will be the overall faster and safer choice for that course anyways.
- Have fun. This should always be the case, whether with gravel bikes or MTBs. You can also have fun with tires that don’t rank at the top of the BRR lists. Therefore, if your tire is not among the top performers in rolling resistance tests, it shouldn’t hinder your enjoyment of cycling. However, if the terrain becomes too rough for the chosen bike and tires, the fun might be diminished. And if you’re on a long ride with someone who you believe is not as fit as you but they’re able to breathe through their nose while your sentences during conversation and side-by-side riding on flat terrain come out more fragmented, then it might be the exact 24 watts difference that the tires make. And that might not be as enjoyable after all. ;-)
Torsten Frank . : : . tfrank.de - Das Blog 
Thanks for the interesting post! I am currently thinking about building a versatile bike for ultra distance gravel and road cycling. What would you estimate is the aerodynamic penalty of running a 32mm tire in a frame that has clearance for up to 50 mm tires? I’m specifically looking at the Carbonda CFR696, I would stick with the rigid fork.
Hey Andi,
so you wondering if you get an aero disadvantage if you are running the mentioned bike with smaller than the usual wide gravel tires it has clearance for?
I’d say: that’s an answer only a very detailed analysis (like test runs the same way I did here or a wind tunnel) could provide. And the answer could be: it’s a wash, it’s slightly better and most probably it’s not gonna be worse.
For a first: I’m pretty sure that bike / frame hasn’t been designed with any aero optimizing involved. And certainly not in regard to the used maximum tire width.
Yes: for high end Time Trial and Triathlon rigs these considerations: tire width and rim width/height are very much taken into account. Not with normal road and gravel bikes, though.
That being said: I’d wager the result won’t be a penalty but a net gain purely by virtue of the smaller tire width. And if you accompany this 32 tire by a aerodynamically effective wheel it’s probably an aerodynamic win in every regard.
Thanks for your quick reply, very much appreciated. I realized that the second part of my question got lost. What I wanted to ask is:
What would you estimate is the aerodynamic penalty of running a 32mm tire in a frame that has clearance for up to 50 mm tires COMPARED to the same tyre in a frame designed for smaller tyres (let’s say with max clearance 32mm)?
The other frame I’m looking at is the Rinasclta Granite with 38mm clearance which they say is aero optimized: https://rinascltabike.com/de/rinasclta-granite-all-road-bike-rahmen/
Ah ok – but also for this frame set my answer holds. While they say (and apparently they did so) they feature aero optimized tube shapes and a handlebar nowhere do they state they optimized the frame in regard to work optimally with a special wheelset or tire width.
So the frame might be a bit more slippery in regards to the wind when you compare it to the other bike but with the same wheels and tires. So again: I expect no penalty for going „slimmer“ on the frame with clearance for 50 mm tires. But I do expect the Rinasclta faster nevertheless, because this was designed with aero considerations. How much? Hard to say? There could be a difference for up to maybe 10 watts or so (for 30 km/h) or there maybe a difference from only 0 to 5 watts, depending on how up to par that aero optimisation of the Rinascalta is and how aero without even much trying the other frame is.
Great post! I’m curious how fast you will be on the Exceed with the Schwalbe Pro One 700 (34c) mounted on your Newmen wheels. Is the aero advantage from the narrower tires bigger than advantage of rolling resistance from the fast Conti 2.2 tires?
Hi Hans, great question!
So – from this post here and also from the most current one (the fastest bike for Badlands) we can derive all the numbers. But the foundations are laid in this article and testing.
So lets just assume the Schwalbe Pro One TLE and the Schwalbe G-One RS would be fine tuned with the pressure to be at the exact rolling resistance level (they are close enough). So we can put them both into the same bucket and show not only how a road slick but also a quite fast gravel tire do come out.
That would put them both at 4 Watts per tire, so 8 Watts in total of an disadvantage to the Conti Race King at 30 km/h (29 km/h to be precise).
I figured, that the aero advantage of a fitting 34 to 40 mm tire on the Nextie 45 rim should be around 9 Watts at the same speed compared to the 2.2 inch Conti on MTB wheels.
So – there is your answer already: Yes, the aero advantage of just the tires and wheelset combo is bigger. For probably 1 Watt. At 30 km/h.
But you asked specifically for the Exceed. And with it also comes the front suspension fork. It’s the reason why it has so many more advantages for Gravel and Off-Road setting after all.
I reckon that this fork is „good“ for another 9 Watts of aerodynamic drag at 30 km/h.
Adding to the wheel and tire aerodynamic loss of already 9 Watts we are now at an 18 Watts aerodynamic disadvantage of the Exceed with the fork and the Conti Race King tires compared to the 8 Watts rolling resistance advantage of the Contis. So still 10 Watts overall Loss.
But now I suddenly broadened our range of comparison. I now compare my Exceed with a suspension fork and the Conti Tires on MTB rims against the same Exceed with a sleek rigid fork and narrow Schwalbe ProOne or G-One RS tires on Gravel-Aerowheels. Or if you will (with a bit degree of freedom) against any other Gravelbike with the latter tire and wheelset combo.
Now – since rolling resistance increases linear with speed while aerodynamic drag does so quadratically, there should be a lower speed than 30 km/h at wich the tire advantage of the Contis is still big enough but the aerodynamic drag part has become smaller than that.
So I did the math and that speed is: 20 km/h!
For 20 km/h, the pure rolling resistance part for the Contis is 20.1 Watts, for the ProOne and G-One RS its 25.3 Watts. So still 5.2 Watts advantage for the Contis. But the aerodynamic drag component of Tire/Wheelset and Fork now only amounts to (nearly the same) 5.3 Watts. So they draw even.
E.g. at 17 km/h the advantage of the bigger Conti Tires would be around 1.1 Watt.
That is the pure rolling resistance and aerodynamic part. For a very good, hard surface. Like smooth hardpack or good asphalt. The numbers will vary with rougher surface (overall slower speeds reached, higher impedance losses, more value placed on lower tire pressures and also further advantages due to suspension in regard to the overall impedance losses of the complete system of bike and rider). Another big part of the considerations are grip, traction, cornering speed and for longer races reduced fatigue. Things I alluded to in my most current article „The fastest Bike for Badlands“.
But the pure Watts saved numbers are already impressive before you take all these other things just mentioned into consideration.
The gist is: you have to have a really aerodynamic working tire and wheelset combo (don’t bother with a 45 mm gravel tire on any old aluminium box shaped rim) _and_ go really fast (above 30 km/h on average) to reap the aerodynamic benefit if the gravel or road tire you mount is of average or even of above average rolling resistance quality and thus still a bit slower rolling than a good XC MTB tire.