Aerolab¶
- Aerolab
- Summary
- Using Aerolab
- Testing Protocols
- Loops
- Out-and-Backs
- Half-Pipes
- Estimating Air Density
- Estimating Rolling Resistance
- Estimating Drivetrain Efficiency
- Background Theory
- History
- Resources
Summary¶
Aerolab is Golden Cheetah's virtual elevation module. Virtual elevation -- also known as The Chung Method -- allows users to estimate two important drag parameters: the coefficient of rolling resistance (Crr) and the aero coefficient of drag times area (CdA). Under certain conditions The Chung Method can provide extremely accurate Crr and CdA estimates.
Using Aerolab¶
Virtual elevation uses successively better guesses of Crr and CdA to determine the closest match to the true elevation profile. You can view this as a video game where you move the Crr and CdA sliders around. When you see an elevation profile that looks like your ride's elevation profile, you win!
When the true elevation profile is not known, you can make use of some standard protocol rides to help us find your numbers. Two popular protocols are loops and out-and-backs. To run an aero field test using these protocols, consult the section on Testing Protocols.
To use Aerolab you'll need:- a power meter
- a scale
- a course where you can run a loop or an out-and-back. It doesn't need to be very long, but it should have a little change in elevation.
#1 Go out and run proper aero field test. Follow the guidelines given in the Testing Protocols, and come back with a nice, shiny new ride file.
#2 Determine the mass of bike + rider and enter this value (in kgs) into the mass slider. Use as accurate a scale as possible for this.
#3 Determine rho, the air density, using one of air density calculators found in the Estimating Air Density section. You'll need to know the barometric pressure, the temperature, your altitude, and the relative humidity. Enter rho using the appropriate slider.
#4 If you're using a crank-based power meter, you'll probably need to estimate eta, your drivetrain efficiency. If you haven't measured it, you can probably use a number between 0.96 and 0.98. To measure it accurately, see the Estimating Drivetrain Efficiency section.
#5 Now start the guessing by estimating the coefficient of rolling resistance of your tires, Crr. This number depends not only on the tires, but also on the inflation pressure, the pavement roughness, and the ambient temperature. If you're using a perfectly-inflated, very fast tire on a glass-smooth surface, the Crr
is probably near 0.350. If you're using knobby tires on rough pavement, you'll be looking at a Crr of 0.800 or more. See the Rolling Resistance section to get a a better estimate. Enter the Crr using the appropriate slider.
#6 Now by varying the CdA slider, try to reproduce the true elevation profile that you ran on. For loops, each loop should come back to the original elevation. For out-and-backs, try to get the out to be a mirror image of the back segment.
When you're happy with the elevation profile, you've found your Crr and CdA. Yay!
Testing Protocols¶
If you don't know your true ride elevation profile, you can use one of the popular aero testing protocols to help identify ride features and symmetries. Loops and out-and-backs are the two most common protocols used.
Keep in mind that virtual elevation works best in windless conditions. Wind adds an unknown dimension that will degrade the accuracy of the Crr and CdA estimates. Gusting wind is the worst. Try to find a day when conditions are calm. The following guidelines help make your field testing safe and accurate:- Keep it safe!! Find safe routes to do your testing. Position a friend at blind intersections, if you have to.
- Hold your position. The slightest shift in position will change your CdA, so hang on tight.
- Hold the line. Use chalk marks or cones, if you have to, to make sure you keep the same line.
- Don't touch the brakes! If you do, you'll be creating a small virtual elevation bump. In some cases, deliberately hitting the brakes to mark a spot might be used to good advantage, however.
- Vary your speed. When doing loops, for example, try 4 loops at different speeds. Starting with the fastest. This will ensure that you can hold the line you've chosen, even on the fastest loop.
- Choose a course with some terrain. Pancake-flat courses will make it impossible to identify VE features. You don't need a lot of elevation change, but try to have at least 10m or so.
- Keep it short. This isn't a century ride. Pick short (eg. 1km) protocol courses. VE doesn't need you to pack a lunch.
Loops¶
When doing loops, you'll probably be throwing out the first lap. For some reason the first lap always seems to be sacrificial. Don't let this stop you from trying to make good use of it, however.
Out-and-Backs¶
Half-Pipes¶
Estimating Air Density¶
Here's a link to a web-based air density calculator that's easy to use.
Estimating Rolling Resistance¶
Estimating Drivetrain Efficiency¶
Background Theory¶
The theoretical basis for virtual elevation starts with Newton's Second Law equation for rider and bike:
F - Crr * m * g * cos(theta) - CdA * 0.5 * rho * (v-vx)^2 - m * g * sin(theta) = m * a,
where
F is the drivetrain force, in Newtons (estimated by power/v) m is the total mass of rider and bike, in kgs g is the gravitational acceleration (9.81 m/s^2) theta is the inclination of the road, in radians rho is the density of air, in kg/m^3 v is the bike speed, in m/s vx is the wind speed axial to the bike's direction of motion, in m/s
By guessing a value for Crr and CdA, we can compute the virtual slope, theta, and integrate it to get a virtual elevation:
ve(x) = ∫ tan( slope(t) ) v(t) dt
ve(x) is compared with a known elevation profile, e(x). By making repeated improvements to the Crr and CdA estimates, ve(x) can be matched to e(x). That, in a nutshell, is the basis of virtual elevation.
