0

June 21, 2016

'No single adjustment to a bicycle will affect comfort, handling and confidence as much as tire pressure.'

Make that 'Harsh' Aero bike Ride like 'Plush' Comfort Bike: Change your pressure!

That's a bold statement, but one that has proven true over and over again.  One that has driven professional athletes and mechanics to using high-tech pressure measuring equipment, high-accuracy gauges on pumps and even tire pressure spreadsheets and logs covering course and conditions as well as tire choice and pressures.  For this part in the series we will look at the 'Why' behind the bold statement.

First of all, every component of a bicycle is essentially a spring.  Even things that seem very rigid will deflect at some load.  That seat post seems quite rigid, but load it and it will bend slightly.  Engineers take the amount of load applied and divide it by the amount of deflection to get the 'spring rate' of the item in question.  This is what we did with tires in Part 2, we loaded them up, divided by deflection and the result was the chart below:

Vertical stiffness model for Surfaces: Stiffness given in N/mm

Technical Note about N/mm:  1 Newton (1N) is

The basic argument for tire pressure being the most important adjustment is that the tire is the softest spring in the entire bicycle system, and when springs are added together in system like this (what engineers call springs in series) the softest spring dominates.  Springs in Series add up like this:

Let's look at an example here using simple numbers, you can see that the lowest spring rate completely dominates the system:

The result here feels counterintuitive, the sum total of the springs is LESS than that of the weakest spring, so I like to think of it this way.  If we had our 10N/mm spring in series with a spring of infinite stiffness, we would be left with a spring rate of 10N/mm, since any other spring we put in series with the 10N/mm spring has a rate far below infinity, then the system will be less stiff.

When we talk about comfort in terms of cycling, the system from the ground up is made of springs in this order: tire, rim, nipples, spokes, hubshell, bearings, axle, frame, seat post, saddle.  For simplicity, we generally measure wheels as built and the frames as frame and seatpost.  We also have an amazing resource in Tour Magazin in Germany who measures frames for vertical stiffness from the seat post rails to the dropouts.  This gives us a single 'frame-seatspost' stiffness number to work with.  So for our purposes we will use this equation where K is the spring rate variable:

To the engineer, all of this looks something like this (and yes, I know that the 'rear' wheel has radial spokes, it's just for visualization! ;-)

CAD Visualization of Springs in Series Model of Bicycle/Wheel/Tire System

This helps visualize the differences in spring rates where a wheel can be 5-10 times more stiff than the frame/seatpost which will be 2-5 times stiffer than the inflated tire.  In all non-suspension bicycle systems, the tires will be the spring dominating the system over bumps and other rough surface features.

Putting all of it together: Effect of Different Frames

To put all of this together, we've created a small chart based on our tire data, the wheel stiffness data from the Road to Roubaix Story and some frame stiffness data based on actual data taken from Tour Magazin out of Germany.  The Tour stiffness data measure from the seat rails to the rear dropouts, so gives us an excellent model for the ride stiffness of a bicycle as it would be sold to you.

We modeled using 3 different frame stiffness, 250N/mm, 200N/mm and 150N/mm.  These are generally representative of what we see in the Tour Magazin data for frame vertical stiffness, though there are outliers worth noting.  We highly recommend looking at Tour Magazin data as there is a very aero bike with 125N/mm stiffness and another 'Comfort' looking model at over 300N/mm stiffness which goes to show that the design details of a specific frame can certainly matter more than the look of the frame!

For the sake of this study, we have broken all of the tested frames into 3 representative ranges which fall in the Stiffest Third, Middle Third, and Softest Third of the complete data, we called these ranges 'AeroRace', 'Race' and 'Comfort' to simplify things.    The actual data from the last few years spans 100N/mm-350N/mm for road frames.

The chart below uses our spring equation with actual measured data from our tire study, wheel study and frame/seatpost study, the resulting data represents the system stiffness from seat post rails to ground.

Vertical stiffness model for 3 Frame Categories: Stiffness given in N/mm

For these models, we are looking at the system stiffness on an 8mm bump with Zipp 303 wheels, 3 different tire sizes and bicycles from the 3 general classes we defined above.  What is notable about this, is how much the system stiffness is dominated by tire pressure, in fact the difference between the 'Aero Race' frame/seatpost and the 'Comfort' frame/seatpost is LESS THAN 1bar (14.5psi) of pressure!

Clearly the differences between the stiffest third of bikes tested and the most comfortable third of bikes tested is very real, but it just isn't very large in magnitude when you consider that you can go up a tire size and down 1Bar of pressure and be roughly equivalent or even better off.  Again, when optimizing for a specific event or race, the athlete should be maximizing every single advantage within the system for the largest possible total benefit, however, if you only can afford one very expensive race bike, you can take comfort in knowing that some clever air pressure strategy can put you right there with your comfort/endurance bike friends in terms of ride quality.

Now, let's look at the differences within a category.  Below are the equations run on 3 different bicycles within the 'Comfort Race' category.  All of these bikes have been ridden at Flanders and all of them are top sellers from major brands.

Comparison of 3 Comfort/Endurance Race Bike Brands with Same tires and Pressures

These three models are all highly successful and all hotly sold against one another in the market touting the industry standard mantra 'laterally stiff yet vertically compliant.'

While there are differences between them, it equates to about 0.2Bar or 3psi difference in tire pressure.  While this difference is non-zero, it is also less than 1/3 the gauge error and and within the range of repeatability of most bicycle pumps.  Think about that, between these 3 \$10,000 models, the difference in ride quality is equal to the difference you will get if you pumped to the 'same pressure' each day and is likely a fraction of what you would get if you pumped to the 'same pressure' on somebody else's pump.

Editors Note: This was the 'smack in the face' moment for me during the Roubaix project, the realization that the quantitative difference between the bike the team was convinced was 'too stiff' for Roubaix and the bike that was completely designed and optimized for the race, was less than 1bar of air pressure.  Clearly, you take every advantage you can get, and every little bit helps, so you take the most comfortable frame, most comfortable wheel, AND you optimize the pressure.  Yet at the same time, we so often paint these decisions as being black and white and they simply are not.  The reality is the difference between the 'too stiff' bike and the 'just right' bike on a normal day is about the difference you get if you haven't pumped your tires in a few days.

Putting it all Together: The effect of wheels

The effect of wheels and comfort has been long debated, and I count myself as being completely surprised to learn that my own 'magic carpet' Ambrosio Chrono Roubaix wheels weren't actually that compliant.  These effects again point to the strong placebo effects of believing something to be true, and also point to the 'Just noticeable difference' threshold as being greater than most of the the variables we are changing here.  Look at the chart above again and you see that the difference in system stiffness between the 250N/mm and 150N/mm frames with 28mm tires at 8 Bar is right at 10%..which is right on the edge of what people can perceive when blinded to the study variables.

Of the 3 springs in our model, the wheel is by far the stiffest, and therefore has the least effect.  As to our point above, they still make a difference in the model, so optimizing wheel stiffness is worthwhile, but as you see below, the effects are significantly smaller than those of the frame and roughly 1/10th of the effects of tire pressure:

Vertical stiffness model for 3 Wheel Models: Stiffness given in N/mm

Here we have the best possible example of the effects of our Springs in Series model.  The 303 project detailed in the Road to Roubaix story was able to reduce the radial stiffness of the preferred cobble wheel by nearly 50%, it was a major breakthrough in both concept and execution, and yet, from the system perspective, it represents an improvement of just under 2%.

What is 2%, well if you are a pro at Paris-Roubaix, it could be everything!  For the average cyclist reading this post, 2% is something very real that you could completely wipe out by over inflating your tires by a few PSI.

Conclusions and homework

Bringing all of this back to the opening statement and image, we see that the difference in vertical stiffness of the 5 frames in the opening image can be offset by using the 5 different air pressures shown on the gauge.  More interestingly from a SILCA standpoint, is that the difference between 4 of the 5 frames is less than the gauge error of a standard bicycle pump (+/-5%), which has driven us to produce pumps with gauges of significantly higher accuracy.  It is our sincerest hope that you, the cyclist, can use this knowledge to begin doing your own testing and study on air pressure for your bike/wheel/tire system.

When researching new equipment choices remember, it ALL matters.  That more comfortable frame really can be a big deal.  Those more comfortable wheels, most definitely are a big deal, and of course, top it all off with optimized tire pressure, because the wrong tire pressure can quite literally undo all of the benefits of the highly engineered equipment you are purchasing or already own.

Our recommendation is to begin keeping a log of your air pressures.  Start where you are today and reduce pressure by 5psi or 0.5Bar and ride it for a few days, then reduce it a bit more, etc.  We have worked with hundreds of athletes, both professional and amateur and find that just a few weeks of keeping a pressure log will begin to completely change the way you think about your pressures and tires.  In many cases, we find athletes deciding that they can race their aero road bike on that course they were planning to buy a more comfortable bike for.  Triathletes, some of whom are now racing at pressures 2Bar (30psi) lower than before, are telling us that they are running better off the bike as they are less fatigued from vibration, and better still, they aren't riding any slower!

Next week in Part 4 we will be looking at tire size and pressure and how they affect Rolling Resistance!  The results will surprise you and will explain how you can lower pressures up to 2 Bar and STILL have Low Rolling Resistance, if not eve LOWER Rolling Resistance!

#### 12 Responses

###### Ron George

July 22, 2018

Sorry , I meant to say ‘latex’ inner tubes above.

###### Ron George

July 22, 2018

Springs in series, they use the same concepts in vehicle suspension design and yes it should be no different than to bicycles.

I’m fairly curious if you’ve done any testing showing the “loss of tire pressure” over the course of say a 2-3 hour ride. We know that latest inner tubes lose tire pressure somewhat faster than Butyl tubes, or atleast this has been so from my experience. A loss of 1-2 psi from the tires, by this model, can affect overall stiffness quite a bit. What would be the advice for that – to keep repeatedly checking pressure over the course of a ride? Not practical in a race scenario. Interested to hear thoughts…

###### Long Toe

December 18, 2017

It’s cool that you as professionals so exactly describe where the pain lies! I also liked the road-to-roubaix blog. What a fantastic eye-opener!! I’m looking forward to read the story about rolling resistance & impedance!!

###### old geezer

August 07, 2017

Re: springs in series, the new awareness of pneumatic rubber-damped suspensions – ie tires – is eye-opening. For decade I pumped my skinny 18-21mm tires over 110 psi, though I weighed between 120 and 130 lb. Filling jarring rough rides kept me berating the bikes I rode. Now, on 25mm tires with 80-85 psi I get a much smoother ride, feel no drag or loss of speed that matters, and control rough, even gravel trails far better, so as to be enjoyable. Folklore about stiff wheels is proven as so much silliness, as actual wheel flex is microscopic – if not, alloy rims would explode from aluminum work flex fatigue and they simply don’t. Digital gauges are exact enough to 0.5 or better; it is less relevant whether the air loss after removing it matters, as you are simply going by final result, and assume the loss is more or less the same each time. If your best comfort/performance zone is found from a measured 84psi, it does not matter if the remaining air is really 82, because you have a repeatable standard in place that is going to replicate consistent results.

###### Lee

April 08, 2017

This post begs the question: assuming I have a perfectly accurate tire pressure guage, how do I know the pressure in the tire after removing the pump head? There is always some amount of air that escapes between removing the pump head and the tire valve closing. Have there been any studies done to determine how much pressure is lost during pump head removal? If the ideal tire pressure for a given bike system is 90 psi, for example, and I pump to an indicated 90 psi, how do I know what psi is actually in the tire after I remove the pump head?

###### Jason Burchell

April 06, 2017

All points clearly accepted. I feel that the research is missing something very crucial.
The speed that a latex tube works vs a butyl, the simple frictional losses found in a softer tire with a then larger footprint and the hysteresis losses that go with that.
It is indeed true that the correct tire pressure separately front and rear have a large impact. My front is always just soft enough to remove the vibration in the bars. The rear will either feel like it hangs up on rough sections like glue or it will crash in to the bumps and take more energy to maintain speed and is relevant to both tires.
Spokes play a major part in the general speed of a wheel. To many people run them to soft and this costs energy.
Rolling weight comes in right up there with spoke tension.
So quite simple really, build a good, light bike with a laterally stiff frame as aero as you can with as tight as you dare on the spokes, use latex tubes and the highest cotton count tires.
Pump the tires as hard as they will and adjust down the front to remove the buzz in the bars, set the rear similar but pay attention to the fact it will need 3-12psi more than the front if all things are equal. There is an exact sweet spot, the
Same rules apply on fast downhill and any other off-road bike. Even road motorbikes, cars, etc. Just a lot easier on a hardtail mtb or road bike.

###### Peter M

April 06, 2017

CraigSJ
I hesitate to comment as I have little experience – but I have come here to learn.
I interpreted the author’s focus on improved pressure gauge accuracy as being related to just how important tire pressure is in the overall ‘comfort’ picture.
I did not interpret the author as suggesting it is dependent in any way in which particular frame one is riding – I think you are both in violent agreement!

###### craigsj

March 09, 2017

Why is author driven to improve the accuracy of a pump gauge because the difference in compliance of frames is of little consequence? That’s absurd.

People either need accurate gauges or they don’t. What frame they are riding doesn’t affect that.

###### Nathan

January 28, 2017

Fascinating posts! Thanks so much for sharing your research on this. I’m confused by how lowering spring rate translates to improved comfort. A lower spring rate means more (vertical?) deflection for a given impact, but why does this make the ride more comfortable? Is spring rate used as a proxy for how much energy the bike system can absorb (as opposed to that energy being transmitted to the rider)?

Nathan

###### Josh at SILCA

July 12, 2016

Alon,
You are correct that the stiffness on flat surfaces is much higher, but on surfaces like that you are not hitting things..those surfaces are not causing discomfort. Most of the things on the road which are going to cause you discomfort are things like road seams, pothole edges, grate edges, sticks, small rocks, chip-seal, pavement cracks…you get the idea. Most of the things you run over that cause discomfort have small radii on the order of 8mm. The CobbleStone radius of 8cm also has a similar effect of higher stiffness than the 8mm radius but the reality of cobble stones is that they are quite rare in the real world and even on that surface, very few of them have a large, domed radius that you are hitting..most of them have much smaller edges and corners and those are what you are hitting. The stiffness of the tire on flat surfaces is mostly critical when looking at the lower limits of tire pressure when trying to prevent bouncing or bobbing due to pedaling motions. However, when considering ‘comfort’ over 85% of the things on the road which you will feel at the handlebars or saddle will have size on the order of 1cm or smaller.
Alister,
Thanks, and yes, the key is to use the placebo effect to your advantage..learn the science and do what actually makes a difference AND then believe in THAT! I’m on 28mm tires on my road bike here in Indiana with rough pavement, chip-seal, pot-holes, etc.. and I’m running them at 68-72psi and for my weight it’s a sweet ride with great grip and still rolls fast. Try it and you’ll love it!
Josh

###### Alon L

July 07, 2016

If I understand correctly, in these comparisons you assume that the road impact is caused by 8mm radius objects (~40N/mm stiffness). This leads to the tire being by far the most important spring in the system. However, if you were to assume flat objects (~200N/mm), this is no longer the case. Why did you choose 8mm radius? I feel that not mentioning this is misleading.
Great series of posts regardless!

###### alister

June 28, 2016

Fascinating post – I am eagerly awaiting the next installment. As a non-practising mechanical engineer I had a faint memory of springs in series from uni 20 years ago, and being in the market for a new bike I was internally debating whether frame and seatpost stiffness techniques were real and/or significant. I was just about to give in and accept there was no way to see through the cloud of marketing hype when I found this post. This post means I add “fitting 28mm” tires to my criteria and maybe ditch the non-aero “too stiff” criteria. Then of course when I run 28mm tires, and combine it with my own placebo effect from accepting your data I will be doubly comfortable and maybe even faster :)