Oval Chainrings - Do they Work?
Before anybody shouts, I know these rings are not necessarily 'oval' but you know what I mean, rings that change their diameter and so give different gear ratios at different crank arm positions. Either way 'Oval' shaped chainrings have been growing in popularity again in recent years and the long lingering shots of Chris Froome's chain bobbing up and down on the Tour de France can only reinforce this. I say 'again' because many of us are old enough to remember Shimano's Biopace chainsets of the 80's early 90's which were of a similar shape. Osymmetric and Rotor sell rings in the UK, as do Highpath Engineering who have offered their 'Eggrings' for years.
Biopace rings were fitted to my first proper Dawes road bike, bought when I started to take cycling more seriously. They seemed to fall out of favour though, and with STI shifters coming along a few years later, the upgrade bug got to me and new groupsets or bikes came along with round rings and barely a second thought for the old irregular ones. I used to think that it was an urban myth that Biopace were fitted the 'wrong way round' but sure enough, the largest diameter is approx 90° away from the Osymmetric / Rotor Ring position. I know this because I hoard all the bits I ever take off my bikes and still have a set, having unsuccessfully tried to flog them on ebay a few years ago!
So, always keen to use technology to go faster rather than just train harder, I began to wonder if I should invest in a set of oval rings. I also needed to form a professional opinion as many customers were either coming in with them, or asking me about them. As a Bike Fitter I am well aware of how the 'dead zone' as the cranks go through the vertical can affect your performance and comfort on a bike and see how crank length and saddle height effect this region every day. I have a bugbear about manufacturers fitting cranks that are too long on their bikes, so I became very interested in not only the performance of oval rings per se, but also if they could ameliorate the effects of overly long cranks.
An event that sparked this interest was a recent customer who came in with two bikes, a road bike with 170mm cranks and round rings, plus a TT bike with 172.5mm cranks and Rotor Rings. This chap was quite small, with an inseam even shorter than mine (which is unusual), so I was fairly sure that both crank lengths were too long for him. We fitted the roadbike first and it quickly became apparent that this was the case. Even after optimising his saddle height, he still struggled to hold a cadence much above 85rpm, but when trying him on 165mm, you could see the stress flow out of his system as he became 'free', whizzing up to well over 100rpm. Expecting a similar conclusion on the TT bike, I was surprised to see that he was spinning almost as easily on the 172.5s, something I could only attribute to the rings fitted to this bike.
Rotor UK are based just up the road in Stratford upon Avon so it did not take me long to get hold of a suitable set of rings. I did not want to leap to any conclusions without a proper assessment so decided on both a road and lab based test programme spread over a few weeks. A weeks holiday in Dumfries and Galloway provided the ideal opportunity for the road assessments, especially given our proximity to the delightful Mennocks Pass. The intention was to use a combination of both subjective impressions and performance data recorded on my Garmin and Strava.com. I was also keen to get into the lab environment with the capability to measure HR, Power and via my Computrainer the metrics of Pedalling Efficiency, Average Torque Angle and Left Right contribution.
The equipment variables I had available to me were 34t, 36t and 50t round chainring, plus 36t and 50t Rotor Rings, all of which were compatible with a set of both 165mm and 167.5mm cranks. I had been tinkering with the 167.5mm cranks previously, so knew my preferred saddle height for both crank lengths. I was also a little worried about ‘adaptation’ but as I had been tinkering recently, I wasn’t sure I was strongly adapted to anything anyway!
Unfortunately, as anybody who has tried to do anything similar will know, the vagaries of our British weather far outweigh the differences between these components. If you were to believe my Strava data, the 36t round rings were far superior to anything else because I climbed so much faster on them. In fact, the tailwind was so strong that day that when I turned round to come back down the 5% hill, I had to pedal quite hard to make any sort of progress at all! So although cycling in Scotland had been fun, it was inconclusive and I needed to get a few more miles in back home. Fortunately, we then ran into a more settled spell and I could get in three assessments under very similar conditions. I chose a 38mile local route incorporating a couple of tough climbs (Burton Dassett, Edge Hill). The specs and average speeds are as follows, with similar average and maximum heart rates.
|36t round, 50t round, 165mm cranks||16.6mph|
|36t Rotor, 50t Rotor, 165mm cranks||16.8mph|
|36t Rotor, 50t Rotor, 167.5mm cranks||16.2mph|
This is by no means a statistical sample, but the average speeds tended to support the subjective impressions. I had decided recently that I preferred 165mm to 167.5mm cranks and the Rotor Rings had not changed this conclusion. This is a shame as I really wanted them to work, but, as with round rings, the longer cranks just made my legs feel heavy and everything seemed much harder work. When on the 165mm cranks, if anything the subjective impression with Rotor Rings was better than the average speed increase suggests. It is not easy to describe the sensation, but it feels like you are getting the lower gearing benefit of a 34t chainring coupled with the performance of a 36t. More significantly, two out of three of my fastest climbs on Edge Hill since my Strava records began have been on Rotor Rings! The one downside is when climbing out of the saddle, as your foot seems to accelerate quickly around the bottom of the stroke in a slightly jerky fashion. I imagine this needs a few rides to get used to.
As a Bike Fitter, I am quite keen on making very quick changes to the bike for people to subjectively assess. The quicker you can make a change, the easier it is for people to notice the difference. This philosophy is also applicable to objective measurements as stopping to change components, even for only 10 minutes can cool people down with no guarantee they will stabilise at the same condition on restarting. So what better way to affect a quick change than to fit both the same size rotor and round ring side by side and use the front derailleur to flick between them. This only works with the small ring, can be a bit tricky to set up, but is possible. The two big rings can also be fitted alongside each other, but you have to take the front mech off completely and move the chain across manually.
Although the weather was no longer an issue, I then worried if the inertia properties of the turbo trainer would compromise the result. As there is no translational inertia of the bike and rider, but only rotational inertia of the wheel, trainer and cyclists legs, how would this effect the changing gear ratio and by implication wheel or foot speeds? If you worry about things too much though you can end up not doing anything. So I thought I’d crack on and see what happens.
The simplest way to assess the different rings was to measure heart rate and power, correlate to perceived effort and keep switching back and forth between round and rotor rings every few minutes or so at a range of different intensities. Somewhat disappointingly, there was very little difference seen on either the 36t or 50t rings when just considering HR and power, but the Rotor Rings always felt a little easier. I am also able to look at other metrics provided by the Computrainer Turbo and Spinscan software. The brake unit in this turbo has a fine enough resolution to measure torque every 15 degrees of crank rotation and plot this against crank angle. This ‘Polar Plot’ can be very informative, showing any significant asymmetries and highlighting the extent of the torque drop off as the cranks go through the vertical.
An idealised polar plot is something like that shown here.
The definitions of the relevant metrics are as follows.
Spinscan, also known as pedalling efficiency.This takes the average torque and divides it by the maximum torque. If you were to pedal in perfect circles with a consistent torque, this number would be 100% but as a rule, I think anything over 70% is fine.
Average Torque Angle. (L.ATA) The angle of the crank where the torque measured at the brake unit is at its highest. Ideally this would be the same each side and between 90° and 110°.
%Watts – Contribution to total power from either the right or left down-stroke. This is sometimes misread as a right / left leg contribution, but as the upstroke leg may also be pulling round the bottom, pulling up, or pushing forwards, it will also be making a contribution to the down-stroke.
Thinking about how Rotor Rings are supposed to work, with a smaller diameter as the cranks go through the vertical, larger on the down-stroke, you could predict that the Spinscan numbers could appear to be worse. As the diameter reduces, the foot will need to speed up, but as the feet, legs, shoes, cranks and pedals have their own inertia, there will be less of the applied muscle force available to turn the cranks as it is being used to accelerate the feet. Likewise, as the diameter increases. the foot will slow, with the change in inertia creating a positive impulse to the pedals.
I wondered if Computrainer had any view on this but a brief peruse on their forum revealed this response to a previous similar enquiry.
“We’ve never done any research with them, nor do we have any plans on doing so.”
I found this a bit surprising from a company whose product is advertised as helping to reduce the dead-spots in the pedalling action and suspect they may have predicted and are possibly perturbed by the likely outcome. Sure enough, as shown below, the Spinscan numbers and shape of the plot do deteriorate when fitting Rotor Rings. But as stated previously, the power and heart rate did not change and the perceived effort was lower.
You may not be able to read the Spinscan numbers, but they both drop about 7% when the Rotor Rings are fitted. In the cold light of day, if your ‘pedalling efficiency’ reduces when fitting Rotor Rings, then why on earth would you do so. Either Rotor Rings don’t work or ‘Pedalling Efficiency’ is an inappropriate metric. I’m fairly sure that Rotor Rings do work and the only thing wrong with the metric is its name and targets. So don’t worry Computrainer blokes, Spinscan is still a fantastic tool to develop a good pedal stroke, but you need to acknowledge that high numbers will be harder to achieve with Rotor Rings. Calling the metric ‘Spinscan number’ is fine, but ‘pedalling efficiency’ could be misleading.
Thinking again about how ring shape and position affects foot speed. If you really wanted to pedal in perfect circles (and I’m not suggesting you should), you would place the large diameter in line with the crank to take advantage of the foot deceleration and fill in the dead zone as the cranks go through the vertical. The acceleration 90° later would be coincident with high pedal forces, so less of a problem. This would position the high points of the rings around 90° from the Rotor recommendation- Blimey! I’ve just reinvented Biopace!
Objective measurements taken on a turbo trainer, which I appreciate do not include oxygen uptake or blood lactate are inconclusive, showing no significant benefits to fitting Rotor Rings.
Objective measurements on the road, particularly on specific climbs suggest Rotor Rings are better.
The subjective impression of Rotor Rings in most riding conditions is positive.
I have no hesitation in recommending Rotor Rings to my customers, but not as a means to cope with cranks that are too long for them.
I shall keep the Rotor Rings on my bike and continue to assess over the usual range of distances and efforts that constitutes my typical riding.
But why do some people not like them? I can’t help feeling that the issue of crank length is still relevant, so will try to gather more data on the subject.
So what shall I do with my Biopace rings now? That’s easy, I’ve put them as close as possible to the right position (i.e. rotated round 1 spider arm = 72°) and reunited them with an old steel frame I use as a turbo trainer bike.
Since first posting this I’ve been alerted to this fascinating report
Not only does this describe the shape of the far more numerous alternative chainrings than I thought existed (Rasmussen oval, Ogival oval, Ovum ellipse, Polchlopek oval anybody?) and uses some complex mathematics to try to define the advantages of each one. I’m pleased to see the report agrees with my conclusion that Biopace was designed to utilise the inertial properties of the system, and that essentially it does not work. As a rule, though, most other designs do seem to offer some benefit.
DIY Dynamic Bike Fitting
You may be interested in our downloadable DIY guide.
49 Page PDF download describing all the observations, simple measurements and calculations required to do your own Dynamic Bike Fits.
Revision 2 released May 2020.
Contains updated male and female saddle height calculators and new information on male / female leg length proportions. Plus a new section to assist with saddle choice and improving comfort.
British Pound, Euro and US Dollar payments are processed by PayPal and all major credit cards are accepted. You will receive an email confirming your payment and will be automatically redirected to a page where you can download the ebook. The ebook is in PDF format so you will need to have a copy of Adobe Acrobat Reader or a similar PDF reader to read the ebook.