Feeling the pressure... what I learned from a gravel roll down tire test

I recently listened to three podcasts from the engineers at Flo Cycling, covering different aspects of tire pressure and rolling resistance (episodes 20, 21, 22, 23 and 24, two being two part podcasts). A couple of the guests made comments on gravel tire pressure that were something like, “lower your pressure until you just get a rock strike, and then add a couple of psi.” Their point was that people tend to put too much pressure in their tires, and they actually be losing some speed in the process. Generally, lowering air pressure increases rolling resistance. But, on bumpy surfaces the bouncing and vibration creates additional losses (and fatigue), which can more than offset the simple rolling resistance losses observed in a lab. Here is a great summary getting into some of the theory of losses associated with what is sometimes called suspension loss or tire “impedance” associated with bumpy surfaces. Another good article is here covering the basics of tire pressure (and why it’s good to experiment), and minimum pressure requirements here. I’d recommend reading or skimming the articles to get a good background on rolling resistance, or listening to the podcasts referenced above.

I have two sets of wheels and tires for my gravel bike, both set up tubeless. The first set is a set of 700c’s with Maxxis Rambler 120 TPI tires at 40mm and Stan’s Avion wheels, and the second are 650b’s with Terrene Elwood Light tires at 47mm and Hunt Adventure wheels. Worth noting, the Hunt wheels are only 22mm deep, while the Stan’s wheels are 40mm deep (aero advantage for the 700 set), but the 650b set weighs about 140g less than the 700c set (total both wheels). I’ve always wondered if there would be a noticeable difference in speed between the two sets of wheels/tires. From what I’ve read, most people “feel” 700c’s are faster, but I’ve yet to see any tests or data. So, it seemed like good motivation for me to spend some time rolling down and climbing back up a gravel hill, to see if I could come to any conclusions on the differences between the two tire sets, and alternatively… how tire pressure may impact speed as well.

First, let me comment that this certainly isn’t a lab quality test, and doing a roll down test on gravel certainly introduces a lot of variance… you never pick exactly the same line down the hill (meaning gravel will change based on path taken) and as you ride there is the chance that the gravel may change from even being packed down a bit by the bike tires. There is also the chance that winds, temperature, air pressure or even my body position may change, impacting aerodynamics. I also realized going into this that any results would be applicable only to my bike, wheel and tire combination. That said, I thought there may be some interest in the gravel community on what I found as part of this process.

Historically I would run 36 psi front, 40 psi rear on my 700c tires, and more recently tried 34/38. I hadn’t really dug into any numbers, but more or less asked others what they rode and just tried a few different values along the way. In the back of my mind I’ve always expected more pressure to be faster, based on rolling resistance testing done in labs (discounting suspension / impedance more than I should). Going in to this test, I decided to measure the weight on my front and rear wheels, with me on the bike. After averaging my front and rear weights on the bike from five samples each wheel, I and found my weight to be around 190 lbs total (me, bike, clothing, hydration, tools, etc.), with a very close distribution of 40% front and 60% rear. I decided to try to balance the contact patch size front to rear, so I calculated what my front tire should be with my rear around 40 psi. Doing some quick math: 75.9 lbs front / (113.8 lbs rear / 39.6 psi rear) = 26.4 psi front (rounding to 26/40, front / rear). I then calculated what a similar starting point would be for the 650b set, so it would have similar tire tension (see this article on tire tension, and how to set pressure based on different tire widths). Since I was already at relatively low pressure, I decided to drop the average pressure by 5 psi for the second set of tests and another 5 psi for the third (each resulting in a drop of around 4 psi front, 6 psi rear).

I headed to a short but somewhat steep gravel hill, relatively close to home. The gravel was a mix of somewhat fresh (sandy) and somewhat packed… no heavy limestone and no super hard concrete-like surfaces either… just a general representative blend. I decided to do five runs with each wheel set at each pressure, resulting in 30 total runs.


The data is shown below, with the wheel size and air pressure (front/rear) shown in the left column. The results are in inches traveled (down the hill and partially up the other side). The average of the five runs, the percent difference (650b at the highest pressure was my baseline, row 1), as well as the standard deviation and the percent deviation.

Screen Shot 2019-05-16 at 8.47.57 PM.png

My thoughts on the results:

  • None of the differences were large enough (relative to the standard deviation) to draw any statistical significance. In other words, from a statistical point of view, the variances between options could simply be “noise” in the data from run to run.

  • Interestingly, the average of the 650b and 700c wheels at the highest pressure were the same. I double checked my field notes twice, since the averages came up virtually identical. The average distance was slightly better on the 700c’s than the 650b’s (as a whole, including all pressures), but again not to a significant degree. There wasn’t much difference in the 650b performance when varying tire pressure… maybe a little slower (less distance traveled), but the differences were very small. For the 700c’s, they seemed to get faster at lower pressure (more distance traveled), at least in terms of the averages between runs (again somewhat small differences). I suspect the small changes in performance based on pressure are a result of using light casing / more supple tires. As discussed in this article, lighter casings tend to transmit less vibration, lower rolling resistance, and have much smaller changes in performance as pressure changes.

  • From a perception and feel standpoint, there was one notable difference, and it was rear tire traction when climbing back up the hill in a standing position. The 650b’s were a bit better than the 700c’s (but the tread pattern is different too), but both got noticeably better as the pressures got lower. At my highest pressure the 700c’s were difficult to climb in a standing position, with a lot of wheel spin compared to limited spin at the lowest tested value. The other perceived difference was vibration. Maybe a placebo effect, but the bike felt smoother as the pressure dropped, but maybe a little soft or squishy at the lowest values. Part of the perceived difference may also be that calculating to better equate pressure to load resulted in lower front tire pressure than I have traditionally utilized.

My takeaways from this test:

  1. For my situation, at pressures similar to what I’ve run historically, there isn’t a big difference between my 700 and 650 wheel sets / tires. That said, at slightly lower pressures, there may be a slight edge for for the 700’s, and of course the deeper wheels would provide more benefit for windy days. On days with more fresh gravel or mud, I’ll continue to grab the 650’s.

  2. I will be running lower air pressure in both sets going forward. Upon more refection, the static weight balance (40/60) may not be completely realistic, as when you are pedaling, you put pressure on the downward stroke which occurs in front of the bottom bracket (at the distance of the crank arm), which may shift the center of gravity slightly more forward. Likewise, I also use aerobars, which will also shift more weight to the front wheel as well (although the methodology was okay for this test, as I coasted only and rode on the hoods). Using the middle air pressure range and changing to a 45/55 weight distribution, this would result in a pressure of 25/31 psi F/R rear for 700c wheels and 22/26 psi F/R for the 650b wheels.

To make this a bit easier going forward, I created a small spreadsheet to estimate my air pressures based on overall weight, tire width and weight distribution (below). I’ll start with one of my favorite quotes, “all models are wrong, but some are useful”. There are a lot of caveats / notes to this… #1 is you need to make sure you are within your wheel and tire manufacturer’s guidelines, and you maintain enough minimum tire pressure to ride safely (see this article). This of this just as a starting option, and adjust according to feel. I’d start with an adjustment factor of +20%, and work your way down in testing / riding. My comfort point was at 0% adjustment with my specific bike, tires and wheels combination. I’ve listed some notes and links in the spreadsheet as well.

Again, if you use this calculator, make sure you are within any and all manufacturer’s guidelines for your specific bike, wheels and tires. Be safe, smart and have fun experimenting!