MTB Tires for Gravel? A Test

The idea of running cross country race mountain bike tires has been getting a lot of attention lately. Rolling resistance testing in the lab and wind tunnel testing suggests the fastest MTB tires may be faster than gravel tires even on smooth surfaces, and anecdotal reports suggest wider tires and lower pressures are generally faster on rough surfaces. Many are now saying the only reason not to run fast MTB tires for a gravel race is if your frame can’t fit them with sufficient mud clearance for the conditions. I’ve definitely felt that wider is faster in the past, and have had some great results with my old Schwalbe Furious Fred MTB tires (which blew up to 47mm), though they were paper thin and punctured very easily after a few hundred miles of wear (and they are no longer available). Therefore, I decided to run a timed, real world test comparing my current favorite gravel tire, the Tufo Gravel Thundero 44mm, to one of the fastest MTB tires (and an emerging choice for gravel), the Continental Race King Protection 2.2.

Test Setup

For the test I picked a flat, straight, 2 mile stretch of gravel, and a nearby flat, straight, 2 mile stretch of pavement. I placed traffic cones near the end of each stretch, then ran each stretch in each direction twice with each set of tires while holding approximately 250 Watts. I attempted to keep a steady position on the bike, with my hands on the hoods, and I tried to ride the same line for each run. I used the lap function on my Garmin to time each run, pressing the lap button when I passed each cone. There was an approximately 10 mph wind, which did not change during the test, and which was mostly aligned with the direction of the stretches. I accelerated to 15 mph before the first cone for the headwind runs, and 20 mph before the first cone for the tailwind runs. I first performed two sets of runs (tailwind-headwind-tailwind-headwind) on gravel with the Tufos, then did the same on the paved section, then returned to the car, installed the Contis, then repeated the same tests with them. On the rim the Tufos meaured 43.7 mm (casing) and 45.4 mm (tread) on the front and 44.0 mm (casing) and 35.7 mm (tread) on the rear. The Contis measured 53.5 mm front and 53.9 mm rear.

My gravel bike is a Chinese carbon 29er hardtail frame with a Niner rigid carbon fork, drop bars, and GRX. The wheelset uses 46mm deep, 21mm internal Light Bicycle rims. I ran both sets of tires on the same wheels (I brought a charger pump and sealant with me in the car). The Tufos had just been used at Unbound and had gotten a little chewed up there (they looked new prior to Unbound). Here is the front before the test:

And the rear:

The Contis were brand new. For the sake of completeness, here is a picture of the rear Conti after the test:

Here is the complete bike with both sets of tires fitted:

I filled the Tufos to 25 psi front and 31 psi rear. This is my normal gravel tire pressure I arrived at through trial and error. I filled the Contis to 21 psi front and 24 psi rear, which is based on my normal pressures for singletrack MTB.

Issues

While these pressures for the Contis are similar to what one would run to get the same absolute distance of tire sag as the Tufos with my usual pressures, to get the same tire sag as a percentage of the tire size one would run around 17 psi front and 21 psi rear. I may experiment with lower pressures in the future.

After performing the first set of four runs on gravel with the Contis, I noticed that the leftmost row of tire whiskers was slightly contacting the chainstay. While I did not suspect this made a measurable difference in performance, and spinning the tire by hand did not reveal and drag, I pulled off these whiskers before the paved runs and ran a third test on the headwind gravel section instead of a second headwind paved test to make sure the whiskers did not substantially affect the results. As you can see below, the result without the whiskers is consistent with the tests with them.

Unfortunately for me and this test, a farmer had set up a circular irrigation system to spray a few yards past his field, spraying water onto the gravel road when oriented directly West. This caused a wet section to form in the middle of the Northern half of the gravel section during the test, getting bigger throughout. While this did not seem to make much difference, I used Strava segments to look at the speed on only the Southern half of the gravel course to make sure the water didn’t affect the results. While the Strava data is not as accurate as the manual timings, it did confirm the water was not a problem, and also provided more interesting results, since the gravel in the Southern half was substantially more rough and loose than in the Northern half. Also, the average wattage for this section still ended up consistent across runs.

For the tests with the Tufos, the second run in each case was slower than the first. I’m not sure if there was a reason for this or if it was just luck of the draw, since the power was held constant and any changes in wind speed would help one direction and hurt the other. However, in each case, the fastest run on the Contis was (spoiler!) slower than the slowest run on the Tufos, so I think the conclusions stand.

Results

I found the Continental Race Kings to be 4.5% slower on gravel than the Tufo Gravel Thunderos, and 1.8% slower on pavement. Here are the raw numbers:

Tufo Gravel Thundero 44
Test caseDistance (mi)Run 1Run 2Run 3Avg. speed (mph)Watts 1Watts 2Watts 3
Gravel headwind1.806:56.807:03.1515.43250250
Gravel tailwind1.804:41.354:47.7322.77250250
Gravel headwind (Southern half)0.863:273:2814.92259256
Gravel tailwind (Southern half)0.842:152:1522.40236234
Paved headwind1.866:20.436:27.8817.43250251
Paved tailwind1.864:32.704:36.8824.37251249
Continental Race King Protection 2.2% difference from Tufo
Gravel headwind1.807:15.967:19.167:19.0414.79-4.1250250252
Gravel tailwind1.805:00.504:58.6321.63-5.0253253
Gravel headwind (Southern half)0.863:423:3514.25-4.5259260259
Gravel tailwind (Southern half)0.842:292:2320.71-7.5230236
Paved headwind1.866:29.0517.21-1.3251
Paved tailwind1.864:43.524:40.5623.74-2.6251251

Conclusions

Clearly we can see that the Race Kings are slower in all cases, and get worse as the surface gets rougher. Not only is it slower by a wider margin on gravel than on pavement, but it is slower by an even wider margin on the rougher Southern half of the gravel segment. This is not what I expected, as conventional wisdom would indicate that the bigger tire at a lower pressure should perform better as the surface gets rougher. The margin is also greater on the tailwind runs, indicating that the difference in performance is primarily due to rolling resistance and not aero drag, which appears to validate the wind tunnel tests that show a favorable result for the Race Kings.

I suspect what is going on here has to do with the suppleness of the tire casings. The Tufos use a fairly supple casing, while the Contis are somewhat stiffer due to being the “Protection” version instead of the lighter “RaceSport”. The supple casing deforms more easily over the rocks while transmitting less shock to the rest of the bike, minimizing the energy loss due to vibration and bouncing. The tradeoff of course is that a supple casing is typically more prone to punctures (for what it’s worth the Tufos made it through Unbound 200 without a flat). The advantage of a supple casing is difficult to quantify in lab tests, since the wheel is rigidly mounted to the test stand, and therefore does not vibrate or bounce except at high frequency, even if the roller drum is textured.

At some point I might do a similar test with different pressures. Other interesting follow ons could be to test the Tufos in different sizes (36, 40, 48), test the HD (Heavy Duty) version of the Tufo, or test the RaceSport version of the Conti.