If I were to build my own Lemond Revolution style turbo trainer (link below) could I use the air speed in meters/second^2 the rider is moving into as a good indicator of what type and size of fan to use?

enter image description here

Or put the other way around could I use the indicated airflow that is moving through the fan at a range of revolutions per minute to provide similar resistance to what is found outside, resistance increasing with the cube of speed?

Would this information be enough to begin building, experimenting and testing fans for a diy turbo trainer?

If the fan has an aero map which states its lb/min in airflow at a given rpm (which is within normal cadence of a geared bicycle) is this a good indicator of potential resistance that is will provide (once converted from statute to metric)?

lemond revolution resistance advert

Lemond revolution trainer link: http://greglemond.com/#!/revolution

  • 1
    You might also need to consider the inertial load of the trainer, as that affects the acceleration component of the equation, as well as what to do to simulate variable gradients. IOW attempting to realistically simulate road conditions is actually pretty hard to do, mainly because it's so variable. I suppose my question is what are you actually trying to achieve? Aug 30, 2014 at 0:57
  • 1
    The resistance of the fan is due to the acceleration of the mass of air moving through it in a unit time. (And, to @AlexSimmons, it's very hard to realistically model the inertial load -- one needs a large flywheel for that.) Aug 30, 2014 at 2:44
  • 1
    @DanielRHicks, Alex is familiar with the inertial load. Here is his home-built resistance trainer and its flywheel. User95786, here are some estimates of the drag parameters (the "virtual" CdA and Crr) for the Lemond Revolution. Perhaps that will help you in building your own.
    – R. Chung
    Aug 30, 2014 at 3:08
  • 1
    Be aware that some fan protective covers can reduce air resistance by generating vortices inside the cover. When I placed a cover over my trainer's flywheel, I had to add extra air blades as the semi-enclosed fan would generate a strong swirl within the cover. Trainers that move sufficient air to provide such a road like resistance curve are usually hideously noisy, which may or may not be a problem. Watching video or listening to music can require earphones at high volumes. In any case, it requires a large fan, or one that rotates very quickly via gearing. Aug 30, 2014 at 6:14
  • 1
    Thanks for the comments and links all, my intention was to build a stable platform on bosch/rexworth extrusions which I have easy access to at the r'd department at work. Then gear up from a custom built hub and fly wheel at a 5:1 ratio to a large diameter compressor turbo with a Similar aero map to this (low rpm's): turbobygarrett.com/turbobygarrett/sites/default/files/…
    – user95786
    Aug 30, 2014 at 8:31

1 Answer 1


The two main forces that are provide resistance when biking are the rolling resistance and the air resistance. The rolling resistance though becomes a small fraction of the resistance once you achieve a constant rate of biking. Here's an article illustrating this idea.

So analyzing exclusively the wind force while riding will cover a large part of the picture but leave some of it out. The best way to mimic the force of riding a bicycle is to match the kinetic energy (or moment of inertia).

You can achieve this result in many ways (mathematically indefinite), as you can either have a huge fan that will spin very slowly, a small fan that spins very quickly, or anywhere in-between.

So any type of fan will work, you just need to manipulate the gear ratio between the crank and the fan so that it matches the energy of the rider. Butt I'm a little confused on the site saying air speed is in m/(s^2) since speed is m/s, and would not use that as the metric for the rider's resistance since it doesn't factor in all resistances felt while riding a normal bike.

This all being said, if you have access to an electric motor you could tie into the fan (with controllable output), you could output wattage of an average rider (or your wattage), see how fast it spins, then using gear ratio calculations between the crank and the fan, see what ratio you'd need to achieve this speed. Any small errors in calculations could be compensated by the ability to change gears so they wouldn't be noticeable.

  • 1
    The x-axis on the graph above isn't m/(s^2), it's (m/s)^2. We can tell this because the y-axis is in Newtons.
    – R. Chung
    Sep 1, 2016 at 2:41

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.