What Causes Rolling Resistance
To go faster, the first step is to identify what exactly is slowing us down. When looking at our tires we are looking at casing losses and surface impedance. Casing losses is the amount of energy that you are losing when your tire is deflected and loses heat. The other major factor sapping energy from the system is impedance losses.
The best way to look at impedance is the amount the system is moved up or down based on a bump. For example, on a perfectly rigid wheel like something found on an in-line skate or skateboard if you hit a 5mm bump, the entire system is lifted 5mm off the ground.
When you move to a pneumatic tire like we ride on bicycles the bump is absorbed largely by the tire and the loss is felt in the heat of the tire deflection like we just talked about. The better the tire and the lower the pressure, the more energy is absorbed by the tire.
We have long been told that higher pressures are faster on the road. We thought that to be true and even “tested” it to be true. This is because all of the testing at that point had been done on roller drums which are extremely smooth surfaces. When the surface is extremely smooth like on a roller or a wooden velodrome casing losses account for the vast majority of rolling resistance which is why you would want to run pressures exceeding 140psi.
When real world testing started to be done by Tom Anhalt, he found that there was a breakpoint pressure. This breakpoint is where casing losses are no longer the leading contributor, but surface impedance becomes the driving force to slow you down. Below you can see his test that follows the same curve as the roller test closely on this “good” pavement surface right up until it doesn’t. The losses in the system shoot back up because it is now the surface impedance that takes over. The rougher the surface, the smaller the tire, the lower that breakpoint pressure is going to be.
These graphs lead us to the question of what pressure should we run. As with almost all marginal gains questions the answer is, it depends. What surface are you riding on, what size are your tires, how much does your system weigh, what kind of bike are you riding; all of these play a factor in the optimal tire pressure for the given conditions.
What we are calculating
Simply put, we are calculating the break point pressure. How high can we inflate the tires before surface impedance takes over and begins to raise the rolling resistance again? Across every test we have seen tires get faster as you increase the pressure right up until they don’t. The better the tire the less steep that curve is which means you aren’t going to lose as much by running a few psi too low or too high, but the same idea is the same from 4” Fat Bike tires all the way down to your 19mm onion skin track tubulars.
We don't make wheels or tires so we aren't limited by our own products in the recommendations we provide. This is most likely why some other calculators out there will have a different rating than we do. This is important to keep in mind when comparing calculators and to follow any limitations especially on the high side of pressure ratings. Don't run 80psi on rims only rated for 60psi. On this topic, when tire manufacturers print a pressure on the sidewall of the tire, it doesn't mean you can't go lower than that.
Clincher vs. Tubeless vs. Tubular
So, if other calculators say different pressures are better, why doesn’t ours? This goes back to what we are solving for, breakpoint pressure. We are calculating at what point impedance losses (the surface you are riding on) will become more impactful than the casing losses (deflection of your tire). System weight changes that, riding surface changes that, air pressure and volume (tire size) changes that, there is even some evidence that body fat % would impact it, but one thing that doesn’t change that calculation is style of tire.
Then How Does Tire Choice Impact Tire Pressure?
Just because the breakpoint pressure calculation is the same between a tubular, clincher, or tubeless set up doesn’t mean all three don’t have their place. The tire casing for example does make a difference in the calculation so a high-quality tire will move the breakpoint higher because the casing is more efficient. You might also have something at one of the extreme ends like a wooden velodrome. A good place to start for a high quality 20mm tire on a wooden velodrome is 187psi. The math is the same for tubular, clincher, and tubeless set ups, but only a tubular tire is going to be capable of handling that pressure without ruining the rim. A wide gravel tire is likely to give you a pressure recommendation on a rough surface that would put you at risk of pinch plats with a standard clincher tire. This means you can go tubeless to eliminate that risk or run a less efficient pressure that is higher because its faster than being stuck on the side of the road fixing a flat.
Using the Calculator as a Starting Point
When the calculator gives you a recommendation, use it as just that. It is the place where your breakpoint pressure is likely to be. There are currently 10 different surfaces offered on our calculator and not all pavement, gravel, singletrack, or even wooden velodrome is created equal. This means the optimal pressure could vary.
Since we are calculating the breakpoint pressure, that means we are focused on that as a means to be faster. There are scenarios where the surface is so different that we might be more focused on something like grip. A muddy cyclocross comes to mind. If we calculate the breakpoint for 180lb total system weight on the category 4 gravel which is the worst surface in our calculator, it gives us 40psi and 41.5psi front and rear for our 33mm cyclocross tires. Anybody who has raced a muddy or dusty cross race can tell you that 40psi would be far too high of pressure. This is because the limiting factor for speed is often grip. That one 200m straight section on the 3k cross course might be best at 40psi but the steep muddy hill or the dusty off-camber requires far lower pressure to stay on the bike.
Cyclocross is the most dramatic variation from the calculator recommendations, but it provides a good starting point to talk about how to use the calculator to test for yourself. If you follow a lot of our advice and are riding on the road, gravel, or mtb the calculator will likely be a close starting point. If you are new to SILCA or marginal gains and are running Gatorskins, you can go ahead and knock a few extra psi right off the bat.
The testing process
If you want to read more about one of the hundreds of tests we have done to build the model for our calculator you can read about that in detail here. The Chung Method is the gold standard for field testing a tire, tire pressure, aero benefits of a wheel, etc. That can be found in great detail here, but we will outline the short of it so you can get started testing your own set up based on the pressure calculator’s recommendations.
The basic principle of the Chung Method is to limit all but one variable and run repeated tests to see which one is the fastest. For tire pressure testing, use the same position on the bike, weight of the system (don’t drink out of the bottle during the ride), start and end at the same elevation, same tire, and same power profile for each run. The only thing you are changing is the pressure.
Take a section of road you want to optimize for and start with the pressure from the calculator. Pick a position on the bike that is very repeatable. TT bikes are great for this, but I find locked out arms are easier to replicate on a road bike than a certain degree of bend. Do a run at this pressure, a few runs at progressively lower pressures, and a few runs at progressively higher pressures. 5 psi is a good increment to start if you are close to the breakpoint to begin with.
In our example below 100psi was the breakpoint. If the surface is the same as what was entered into the calculator and you got 100psi to start with, each of your runs at lower pressures should be progressively slower than the 100psi run, and each run at higher pressures will be progressively slower. This relatively simple method is how we can be certain that 100psi is the absolute fastest tire pressure for the given inputs. When you graph your results, they should look like some variation of the green line. If you don’t see a distinct change in times somewhere in the tested pressures, it means you haven’t yet found the breakpoint and should test with larger pressure differences to find a better starting point.
Consistency in our Findings
We often hear that people don’t want to go through the trouble of testing their specific tires and just want to know if they should err on the side of too high or too low. This testing can take a significant time investment, so the question is certainly reasonable. For once the answer is also straight forward. One finding that has been repeated time and time again through hundreds if not thousands of tests is that it is always better to be too low than too high.
Here you see a graph depicting the wattage losses from one of our tests. The rough milled concrete likely needed more lower pressure data points but as you can see on the new asphalt surface being 10psi below the breakpoint only cost 1w. Being 10psi too high cost 9w. The coarse asphalt followed the same pattern. This test was done with Continental GP 4000 II’s and were some of the fastest tires at the time.
We talked earlier about how quality of tire will amplify the results. What I am trying to say if you are running Gatorskins on your race wheels, 1. Please swap your tires 2. If you aren’t going to swap them at the very least let some extra air out of them because it is going to be a lot more than 9w if you are 10psi too high.
All of this data is great to have but there are some things to keep in mind. One variable we didn’t discuss yet is the pump used in the test. This isn’t a shameless plug to buy a SILCA pump (although you absolutely should) but remember not every pump has the same accuracy. 100 psi on the 15-year-old pump you borrowed from a buddy in the parking lot of your race isn’t the same pressure as the 100psi as your pump at home, and very likely neither are actually 100psi. We have seen new pumps with variations as high as 7-8psi. While that seems high you have to remember it goes both ways so your pump could be 7psi low and your friends 8psi high. So once you put in the couple hours to make sure you are running the fastest tire pressure possible, you throw all hard work out the window because your tires are actually 15psi higher than what you thought they would.
Most pumps are accurate to about 5 or 6% when they are new, and all pumps become less accurate over time. If you are serious about saving that 10 or 15w by running optimal tire pressure, a few hundred dollars on a pump is a great return on investment. If you don’t want to invest in a new pump, that is fine too, just make sure you at least use the same pump every time.
See what we recommend for your set up at the SILCA Tire Pressure Calculator