My name is Krzysztof Wierzbicki and I have nearly two years of experience in testing and reviewing everything that can improve comfort on the bike (suspension seatposts, suspension stems, tires, wheels, handlebars, saddles, etc). I have created gravelbikes.cc site to finding the most comfortable bike parts. Now, I want to continue my mission but on a much more professional level. Based on my experience and knowledge I have created a professional bikelab that not only will professionally measure the efficacy of comfort improving bike parts but also allow the creation of objective ranking of those parts so you can really know what is best and what to buy if comfort is what you are after.
I started my testing using a vibration measuring app on my phone. I have carefully selected two test routes (one in the forest with big bumps where I rode at 15 km/h and one on a gravel road with high-frequency chatter thanks to a speed of 35 km/h). In each case, I have been using a phone mounted either at the level of my wrist (measuring front end vibrations) or at my lower back (measuring rear end vibrations). This approach worked quite well but it had two main drawbacks that I really wanted to address. The first was that measuring in the real-world meant that I can’t create a proper ranking of tested bike parts (because of the changes in weather, humidity, temperature, and the road itself I could only draw conclusions based on comparison tests made on the same day). The second thing was that I was measuring vibrations on my body and this meant that, especially at the rear, I registered fewer vibrations than I should because of the dampening effect of my body itself (of course in the winter it was also the case of added layers of clothes).
To address both of those issues I needed an indoor lab. But It was not an easy thing to do mainly because I really wanted to maintain as much as possible of the real-world experience and thus, make this lab mimic those conditions as good as possible. This meant a lot of riding and experimenting with different approaches. I started with recording the vibration on my test routes to see how exactly they look like on the graphs. After considering and testing various different approaches I settled with a professional Woodway treadmill that has 60 separated belts made from aluminum (with rubber on top). This allowed me to mount different obstacles on each belt.
So I begin testing different variants of obstacles to mimic as much as possible real-world scenarios. Having those vibrations graphs I knew what I was after but getting there was no easy task. First I find out that riding even on a good quality tarmac generates constant vibrations at a certain level. So to achieve this I had to install on most of those belts a 5 mm high and 30 mm wide wooden planks. Then, I needed something to stimulate roots and potholes. To achieve that I used wooden and aluminum bars with 10 mm and 18 mm in height. I arranged them in a certain way to mimic various scenarios: a single hit on a 10 mm obstacle, a single hit on an 18 mm obstacle (both wooden and aluminum), and series of obstacles that come one after another (again with different scenarios using 10 mm and 18 mm bars placed in various distances). Then I needed to set up a proper speed. My treadmill offers speeds from 0.1 to 15 miles per hour (0.1-mile steps) so I had plenty of options but after hours of experimenting I settled with 5 miles, 7,5 miles, and 10 miles. The first speed is the closest I could get to my 15 km/h ride in the wood (in terms of vibrations graphs), the last is as close as possible to a 35 km/h ride on a gravel road, and the one in the middle is somehow an average performance.
Putting bike on a treadmill
Yes, I use a flat treadmill to be sure that wheel is behaving in the same way as in the real world but of course putting bike on a treadmill with both wheels on it is simply too dangerous so I had to find the next best thing. Again, after many experiments, I settled with the Cycleops Tempo Fluid trainer and two wooden pallets. Depending on what I want to test (front or rear) I put the front or rear wheel into the trainer which is mounted on wooden pallets. This construction offers a secure hold in terms of forward / backward movement but at the same time, almost free movement side to side and totally free movement up and down (which is the movement I want to measure).
Mimicking the real rider
I started with me sitting on a bike but I quickly discovered that even slight changes in my position on the bike resulted in changes in the measured vibrations. So I had to abandon this approach and try to simulate the bike with the rider by putting appropriate weight on the bike. To achieve that I needed to weigh the front and rear of the bike with me on it and then putting an appropriate load on 3 crucial places: the handlebar, the saddle, and the frame itself (because even when you are riding no-hands, there is still a lot of weight on the front wheel that your frame is transmitting). This sounds quite easy, but believe me, that it was definitely far from easy to achieve a stable bike with proper load (that I can freely change to mimic different riders weight or riding scenarios) moving in a straight line at speeds of 10 or even 15 miles per hour on a treadmill with a lot of obstacles reaching 18 mm in height. The key point is that there is always a load in those three places no matter if I measure front or back-end vibrations (because I always want the bike to behave like the rider is on it).
The measuring itself
Up to this point, my measuring device was a phone with a vibration measuring app. But, for this project, I finally moved to something more professional. I have bought an Arduino Nano 33 BLE external Bluetooth accelerometer that measures the vibrations with a refresh rate of 478 Hz. It took quite some time and a lot of help from Dmitriy, the author of the Vibration Analysis app (big thanks to him!), to properly set up and calibrate this device. But it was worth it and now I can look at my bike riding on a treadmill and use my phone to measure vibrations the way I want. Also, this app allows me not only to get the average vibration over course of 24 seconds of measurement (using ms/2 scale) but also the graphs showing the frequencies at which vibrations are mostly occurring (which can be very useful when comparing the efficacy of different comfort improving solutions).
When measuring front-end vibrations, I put the sensor in the middle of the handlebar, directly below the stem. When I measure back-end vibrations, I put the sensor directly below the saddle. And like in my real-world testing, I do at least 3 measurements (each lasting 24 seconds) and then I calculate the average result.
I know that one picture can be worth thousands of words so to give you more visual testing representation I will also record everything in slow motion (240 frames per second) and publish the video so you can really see how tested products performed not only on the graphs but in real life.
Why I want to do all of this?
Yes, you may still wonder why all of the hassle? The answer is very simple. I really want to create a proper ranking of all comfort improving parts and tell you objectively what is working and how well it is working. So you can make an informed purchase decision and enjoy riding on your bike comfortably.