Monday, August 10, 2009

Power and Aerodynamics on the Saturday morning Time Trial

Every Saturday morning from mid April through late August at 7 AM, David Harwi organizes a free time trial out and back on West River/Martin Luther King Drive. It is a great opportunity to practice your time trialing skills on a (relatively) closed course. Because the road is fairly flat, the wind is usually light and the other conditions are fairly constant, it is also a great opportunity to gather data. David has been kind enough to provide times for every rider that has done the time trial since 2002, along with the weather conditions and any other notes for the day (e.g. course shortened by 0.45 miles...)

Recently, there has been a lot of discussion on the West River Time Trial yahoo group ( about how best to improve your time. Many have offered excellent advice about training, equipment, pacing strategy and bike fit. All this discussion piqued my interest about how different factors influence your speed on the bike. Of course, I am by no means the first one to investigate this question. People like John Cobb and Steve Hed have spent their careers developing aerodynamic bicycle equipment. Zipp, Cervelo, Trek, Felt and BMC as well as many others have created products with heavy utilization of the wind tunnel. It is no accident that in the last 15 years, Pro Tour time trial speeds have gone up and up, despite more and more restrictions on equipment by the UCI.

I have no desire to either repeat, replace or negate the studies that others have done, but I did have some questions that I was curious to answer.

Question #1: How does power affect speed in a TT?

Answer: If you want to go faster, you can have 2 choices: either produce more power or reduce the forces that slow you down (aerodynamic drag, rolling resistance, gravity, drive train resistance and braking force). How you might go about increasing power output is subject to opinion, and varies greatly from one athlete to the next depending on their strengths, weaknesses and constraints. I won't get into the details here, but I will say that you can read all the books and articles you want but nothing will replace the value of a good coach. With risk of seeming prejudiced, time trialists tend to be very regimented people and often hesitant to give up control. Getting the most value out of a coach means letting go and trusting them, even when you might not completely understand why they are asking you to do certain things. As the old saying goes, an idiot is defined as someone who repeats the same behavior and expects different results. Without a coach, or without trusting the coach, we tend to work most on what we are already best at and least on what we most need to work on to achieve our goals.

So, let's say you hired a coach, everything has went well and your sustainable power has gone up. If nothing else has changed, how much faster can you expect to go? To answer this question, I used Tom Compton's, which allows you to calculate speed at a given power, or power at a given speed. For a flat course with no wind, a 5'10" 156 lb. rider with a 20 lb bike and an "average" aero position (I assumed a cdA of .27) can expect to go about 24.5 mph at 250 watts, which comes out to a time of 20:26 on the 8.35 mile West River Drive TT. If this rider is able to increase his average wattage over that distance by 10% to 275 watts, their average speed will increase to 25.4 mph and their time will decrease to 19:43. Not bad...

Question #2: How much do aerodynamics affect speed in a TT?

Answer: On a flat course, about 85% of the opposing force that riders fight is from aerodynamic drag, so it's pretty significant. There are 2 key components that determine how much aerodynamic drag you will have at a given speed are: 1) Frontal area and 2) Drag coefficient. Frontal area would be altered by the position you hold yourself in and the shape of your body. If you ride in the drops instead of the hoods, or on TT bars instead of the drops, you will reduce this area. For a visual demonstration of this, check out this video I found on YouTube:

Drag coefficient is a little tougher to measure without a wind tunnel, so we have to rely primarily on studies done by others. Luckily there are quite a few companies with a vested interest in ensuring that their products are as "slippery" as possible. Drag coefficient is influenced by riding more aerodynamic wheels, a skinsuit, aero helmet, booties, or time trial frameset. However, one factor that complicates things here is yaw angle, or the angle that the air hits the rider and bike at, which is a combination of the rider's forward velocity and wind velocity in other directions.

Most of the time, wind tunnels are used to study airplanes and automobiles where wind direction is minimal if not negligible. A 5 mph direct crosswind would mean a 0.6 degree yaw angle for a jet moving 500 mph, 5.2 degrees for a car moving 55 mph and 9.6 degrees for a bike moving 30 mph. Different equipment will perform better or worse under different yaw angles. Does anyone remember the final time trial in the 2005 Tour de France, where Lance Armstrong had people out on the course measuring the wind conditions so that he could choose the right wheels? Due to the high crosswinds, Lance chose to ride a Hed 3-Spoke rather than a disc wheel on the rear because of it's better performance in high yaw angles. When choosing aerodynamic equipment and fitting a rider an aerodynamic position, it is important to realize that different conditions may produce different results in terms of what is "optimum". In addition to wind conditions, rider size and shape and course profile may all be influences.

I said earlier that I assumed a cdA of 0.27, which is an "OK" aero position for a 5'10" 156 lb rider. This same rider probably has a cdA of around 0.31 on his road bike, riding in the drops, which would result in a speed of 24.3 mph on a flat course at 275 watts, or a time of 20:37 over 8.35 miles. Getting into a "Really good" aero position might bring his cdA down to 0.23, resulting in an astounding increase in speed to 26.6 mph, or 18:50 over 8.35 miles at the same wattage! For a great article about how aerodynamic changes affect power and speed, check out

Question #3: How does weight affect speed in a TT?

Answer: On a flat TT, weight doesn't affect speed too much... directly. With all other things being equal, our rider would only add about 6 seconds to his time if he gained 20 lbs. or take about 5 seconds off if he lost 20 lbs. So, with all the effort required to lose weight, the risk of bonking, missing workouts or getting sick probably outweigh the benefits for most. It would only take a 3 watt power loss to negate the direct benefit of 20 lbs. of weight loss. There is, however, one complicating factor: a thinner rider is usually more aero than a fatter rider, so losing 20 lbs. may actually help more because of the aerodynamic benefits than the gravitational benefits.

Question #4: How does temperature affect speed in a TT?

One of the things that I was intrigued by about David's records was that he recorded the temperature each day. Temperature affects air density, which affects aerodynamic resistance. This year, the temperatures at 7:00 AM on Saturday morning have been consistently in the mid 60's since early May, but in past years, the temperatures seemed to range from 55 to 75 degrees Fahrenheit. For every 5 degree increase in temperature, air density reduces 0.0007 lb/cubic foot, which means a 0.089 mph speed increase. This may sound minimal, but a 20 degree increase results in a 0.36 mph increase, or almost 30 seconds over 8.35 miles with all other things being equal. Not taken into consideration, though, is the effect of temperature on the human body. Certainly, extremely cold or hot temperatures will result in significant power loss for those not acclimated.

Though not a variable in this model, altitude change has a similar affect. All other things being equal, if West River Drive were at 5000 feet, average speeds would be about 1.6 mph faster, which equates to an average time savings of over a minute over 8.35 miles. It is no accident that the majority of the all time U.S. track records were set at the Colorado Springs Velodrome, which is 6000 feet above sea level.