It’s well understood and accepted that serious (racing) time trialists and triatheletes spend lots of time thinking about aerodynamics. When you’re outside the shelter of a pack of riders, the wind is your enemy. Its the IFB – Invisible Force Bastard – making your life hard.
Concerns about aerodynamics should be of concern to randonneurs too, but I get the distinct impression that few care about it. I’d tend to think that any decisions about gear and clothing should consider the aero drag effects of those choices. After all, aero drag is the single biggest consumer of energy, even at modest speeds and courses with lots of climbing. And what rando wouldn’t like to pick up some speed and energy efficiency if they can?
Some folks throw up their hands and say that they don’t know anything about that stuff, how can I know what the aero impacts of equipment choices are going to be. After all, nobody publishes good, verified data on the aero impact of things like handlebar bags, fenders, wider tires, and other items of interest to randos.
Luckily, this can be a DIY affair. And, when using the right methods, you don’t need to understand anything at all about aerodynamics. If you execute the right method with reasonable care, you can still get the results you want. Physics happens, whether we understand it or not – the key is ‘capturing’ what happened with appropriate measurements, then using the data in a robust way.
Robert Chung has shown the way and has developed methods to measure aerodynamic drag factors (CdA) as well as coefficients of rolling resistance of tires (Crr), another very important consumer of power of the long distance cyclist.
This paper explains it all: http://anonymous.coward.free.fr/wattage/cda/indirect-cda.pdf
(A tip: skip to page 57 and start reading there. The extra text helps.)
Now, of course, most of the full detail is in this paper, but here is a distillation of the main points bout Robert Chung’s Virtual Elevation (VE) methods:
1. The VE method is supremely practical because it doesn’t require holding much of anything constant – classic methods require constant speed, slope, and power. Bah – that’s not the real world. (OR they require very controlled apparatus like windtunnels. Also not reality for most people.) The VE method allows for variability in things like speed, power, and slope. Not only does it allow it, it craves it in order to really tease out CdA and Crr.
2. By metering and logging speed and power on a second by second basis, you very effectively “deal” with the lack of uniformity in those two variables. The non-uniform slope of real world test venues is dealt with by ultimately converting everything that consumes power (gravity, acceleration, rolling resistance, and aerodynamic drag) into a virtual elevation plot. By ridng the same course repeatedly at different levels of speed and effort, you can plot what should be the exact same course profile repeatedly, and you manipulate the values of CdA and Crr used to create the plot till each of the lap profiles looks the same (on the plot). It’s a graphical solution. By ‘looks the same’, I mean the hills and valleys are all lining up horizontally at the same elevations, and the overall shapes look the same.
3. The major variables that are not metered by this method are ambient winds and braking. Those things DO need to be carefully controlled as much as possible. The smaller the things you’re trying to detect, the more fussy those things are. When you get to the finer points, even passing cars can mess with the data a bit, so a traffic free, windless road that you can do circuits on without braking are ideal. The road wants to have some amount of ups and downs on it too, not be perfectly flat. This gives you some “features” on the VE plot that help with the graphical solution.
4. This same graphical approach that allows for us to solve for CdA and Crr also reveals imperfections in the data that were due to wind and braking. Purportedly, with a little practice looking at these plots, wind and braking show up graphically on the plots. Braking and headwinds show up as uphills (or increase in slope), and tailwinds show up at downhills (or decrease in slope). In a sense, if you can learn to visually spot these flaws and learn to ignore them, they won’t have much effect on the accuracy of the graphical solution.
5. I finally have come to the most important realization of all: If someone wants to know is this more aero than that, etc., there is often no need to wonder about it. It can be measured. With some really basic stuff (a garmin with an auxillary speed sensor) or some slightly more advanced stuff (like a power tap), these questions can be answered with good accuracy.
Wouldn’t we all be better off to have actual, reliable answers to these kinds of questions? Or, alternatively, we can search for the answer by way of silly, endless thought experiments in online forums.
Personally, I’d like to get actual, reliable answers…then get into silly online debates about something else instead.
Stay tuned – there’s more to come on VE test results of mine…. and how to determine your CdA and Crr from a coastdown using nothing more than a Garmin bike computer with an auxillary speed sensor…..