Wind tunnel testing for mortals

You do not need a corporate budget to cut drag. With a power meter, a calm road or track, and a good protocol, you can estimate your CdA, compare setups, and bank real time savings. This guide shows practical, affordable ways to measure aerodynamic drag and turn small tweaks into free watts.

The affordable way to measure drag: what matters and how to measure it

Aerodynamic drag is described by CdA (coefficient of drag times frontal area). Lower CdA means fewer watts at the same speed, or more speed at the same watts. You can measure changes in CdA outdoors using your regular equipment.

What you need (minimum viable kit)

• Power meter with reliable zero-offset.
• Speed source: wheel sensor preferred over GPS for accuracy.
• Barometric altimeter (most head units have one).
• Accurate rider + bike mass (weigh yourself in test kit, with bottles).
• Calm test venue (out-and-back road, quiet loop, or velodrome).
• Temperature reading; note pressure/elevation if possible.

What you can estimate

• CdA: your aero drag figure. Typical road position: 0.28–0.35 m². Aggressive TT: 0.18–0.24 m².
• Crr: rolling resistance coefficient. Good road tyres on smooth tarmac: 0.003–0.005; track: 0.0015–0.0025.

Rule of thumb at sea level: at 40 km/h, a 10 W change is roughly a 0.012 shift in CdA; at 45 km/h it is about 0.008. Many clothing/position tweaks are in the 0.003–0.020 CdA range.

Power ≈ 0.5 · ρ · CdA · v^3 + m · g · Crr · v + m · g · v · grade + inertia
where ρ is air density, v is speed, m is mass, g is 9.81 m/s²

Step-by-step field testing protocols

1) Out-and-back virtual elevation (VE) protocol

This uses your power, speed, and elevation to infer CdA while canceling slope and wind. It is the most accessible option.

1. Choose a straight, quiet road: 1–3 km, minimal traffic, early morning, wind < 2 m/s.
2. Warm up 15–20 minutes. Calibrate power meter. Set tyre pressure. Weigh rider + bike + bottles.
3. Record temperature. Keep clothing constant for a baseline.
4. Ride 4–6 laps of out-and-back at steady speed (e.g., 38–42 km/h) or steady power. Use the lap button for each out and each back. Avoid braking, surges, or drafting.
5. Repeat the set in your alternative configuration (A/B/A if time allows to check drift).

Analysis overview: the VE method adjusts assumed CdA (and Crr if needed) until the reconstructed elevation trace is flat between out and back legs. The CdA that flattens the trace best is your estimate for that run. Compare condition A vs B to see which is faster.

• Do at least 3 matched pairs per change. Look for a consistent CdA gap.
• Treat anything smaller than 0.003–0.005 CdA as noise unless repeated.

2) Track-based lap testing

A velodrome or very calm closed circuit removes traffic and wind variables.

1. Pick a steady speed target (e.g., 45 km/h) or steady power.
2. Warm up, calibrate, and note mass and temperature.
3. Run 6–10 laps in position A, then 6–10 in B, repeat A. Keep line constant.
4. Use the average power at matched lap speeds to infer the lower-drag setup, or model CdA if lap length/elevation is known.

Small differences (0.002–0.004 CdA) can be detected reliably in calm indoor conditions.

3) Coast-down roll tests for Crr

Knowing Crr improves CdA accuracy, especially at lower speeds.

1. Find a gentle, uniform slope or flat road, no wind.
2. From a set speed (e.g., 35 km/h), stop pedaling and coast in the hoods, no braking.
3. Repeat 5–10 times. Use speed decay to estimate Crr (steeper deceleration on flat indicates higher Crr).

Use this to pick tyre/pressure combos, then fix Crr during aero tests to reduce uncertainty.

Data hygiene that makes or breaks results

• Exclude periods with braking, accelerations, or passing vehicles.
• Use the same bottle fill, tools, and clothing between runs unless they are the variable.
• Re-zero your power meter if temperature shifts.
• Prefer wheel speed to GPS; set head unit to 1 s recording.
• Air density: record temperature and elevation. If you test A vs B within 30–60 minutes, density is effectively constant and cancels out.

Coach tip: Test changes that are big enough to matter. Aim to detect at least 0.005 CdA. If you cannot see it clearly in three A/B/A blocks, move on.

Make the gains stick: choices, pitfalls, and next steps

High-value changes to test first

• Position: lower the front end incrementally; narrow elbow pads; shrug practice to reduce shoulder area. Check breathing and power output do not drop.
• Helmet: different shapes suit different back profiles. Test with head held as in race effort.
• Clothing: snug skinsuit or well-fitted jersey; smooth sleeves; properly zipped. Aero socks and shoe covers often test well.
• Bottles and storage: single bottle behind the saddle for TT; on the downtube for road if it fills the frame gap; avoid boxy top-tube bags.
• Wheels and tyres: fast tyres with latex or TPU tubes; match tyre width to rim; keep pressures appropriate to surface.

Common pitfalls

• Wind creep: a “light breeze” becomes a gusty session. If VE plots won’t settle or out vs back are mismatched, reschedule.
• Traffic effects: passing trucks create pressure waves; bin those laps.
• Inconsistent posture: rehearse your race head position and arm width. Video a run to verify.
• Under-fueled tests: slight power drift looks like aero change. Eat, hydrate, and keep effort steady.
• Dirty drivetrain: chain contamination can cost 5–10 W and masquerade as aero loss.

Decide with thresholds

• ≥ 0.010 CdA better: green light, lock it in.
• 0.005–0.009 better: promising; retest on another day to confirm.
• < 0.005: treat as neutral unless there is a secondary benefit (comfort, handling).

Repeat your best configuration over multiple days to confirm it is robust across temperatures and slight winds. Then practice holding that posture at race power so the watts saved show up in your results, not just in your files.

Field testing will not replace a controlled wind tunnel, but it gets you 80–90% of the value for almost no money. Be systematic, chase the big wins first, and you will ride faster at the same FTP without spending a season’s race fees on equipment.