Compute wall-interference (blockage) corrections for a model inside a wind tunnel using Maskell and Mercker formulas. As you vary test section, frontal area, CD and tunnel type, blockage ratio, ε, corrected velocity and corrected CD update in real time — so you can judge whether a result can be reported without further correction.
Parameters
Tunnel type
Wall condition sets the tcf multiplier
Test type
Switches acceptable BR threshold
Test section height H
m
Test section width W
m
Model frontal area A
m²
Drag coefficient CD
Sedan 0.25–0.35, truck 0.6–0.9
Wind speed V∞
m/s
Results
—
Test section area (m²)
—
Blockage ratio (%)
—
Total correction ε
—
Corrected velocity (m/s)
—
Corrected CD
—
Status
—
Test section & model blockage view
The model (blue/coloured ellipse) sits in the test section. Streamlines (white) compress around it and the wake (red) is squeezed against the walls. The bar below is the blockage ratio (green → orange → red).
Corrected CD uses dynamic pressure at ρ = 1.225 kg/m³ standard air. Open-jet tunnels give negative ε (tcf = −0.5), adaptive walls give tcf = 0.05 (smallest).
A wind tunnel just blows air at a model, right? Why do we need a “correction” at all — isn't it already representative of the real free stream?
🎓
Good catch — it isn't quite the same. Out on the road there's infinite air around the car, but in a tunnel the walls are always close by. When the model narrows the flow path, the air can't fully go around it, so it accelerates locally. The wake behind the model also gets squeezed by the walls, which drops its pressure. We call this wall-interference “blockage,” and without correction your measured CD can be 5–10% higher than the true free-stream value.
🙋
5–10%! And Maskell and Mercker correct for that? When I bump up the “model frontal area” on the left, the blockage ratio and ε both jump.
🎓
Exactly. Blockage ratio = frontal area / test-section area is the dominant parameter. Maskell (1965) modelled the wake-on-wall pressure drop with ε_w = 0.96·CD·A/C. Mercker (1980s) added solid blockage and packaged the whole thing for automotive testing. This tool sums both for bluff bodies, uses only solid blockage for airfoils, and counts the wake at half weight for lifting bodies — a practical compromise.
🙋
When I switch the tunnel type to “open jet,” ε goes negative. So the effective wind actually slows down?
🎓
Sharp eye. An open-jet section has free boundaries instead of walls, so the air the model pushes can escape outward. The flow doesn't speed up; instead pressure leaks out and the effective stream slows. That's the tcf = −0.5 sign flip. Slotted walls sit between closed and open (almost zero), and adaptive walls deform their shape to follow the streamlines so the correction nearly vanishes. F1 tunnels and the Mercedes Aerodynamic Center use adaptive walls plus moving belts to keep corrections under 1%.
🙋
So as long as I correct, any blockage ratio is fine?
🎓
That's the trap. Both Maskell and Mercker are linear approximations valid only up to BR ≈ 5–10%. Above ~10% the correction itself drifts, so the tool reports three levels — Acceptable / Correction required / Excessive. Once you hit “Excessive,” scaling the model down or using a bigger tunnel is more honest than trusting the formula. The defaults here (2.5 × 3.5 m tunnel, A = 0.4 m²) give BR = 4.57%, right in the sweet spot where Maskell/Mercker is trustworthy.
🙋
Got it. So if a prototype run comes in at BR = 8%, I shouldn't just publish the corrected CD — I should also flag it as “outside the linear-correction band.”
🎓
Right. Test reports should always carry the trio: BR, correction method and corrected CD. F1 teams like Sauber, Williams and Honda Racing use 50–60% scale models to keep BR under 2%. Even Tokyo Sky Tree was tested at 1/300–1/500 scale to land BR at 1–3%. The hard truth is that most of the “blockage problem” is won or lost at the BR-budget stage, before you even apply Maskell.
FAQ
When a model is placed in a wind tunnel test section, the model itself constricts the flow path and accelerates the surrounding stream (solid blockage), while the wake region is squeezed by the walls and loses pressure (wake blockage). The flow then differs from an infinite free-stream condition. Blockage correction estimates that difference as a coefficient ε and converts measured velocity and CD back to free-stream values. This tool combines Maskell (1965) wake correction with Mercker drag-blockage correction to give V_corr = V∞(1+ε) and CD_corr = CD/(1+2ε).
For bluff bodies (cars, buildings) the typical limit is 5% (model frontal area / test section area). Streamlined airfoil bodies tolerate up to 10% because their wakes are smaller. Beyond that the correction formula itself loses accuracy, so the practical fix is to scale down the model or move to slotted / adaptive wall tunnels. This tool switches the threshold by test type and labels the result as Acceptable / Correction required / Excessive.
Closed tunnels fully constrain the flow, so ε is positive (effective velocity goes up). Open-jet tunnels let pressure escape outward, so ε is negative (effective velocity goes down) and is roughly half of the closed-wall value with the opposite sign. Slotted walls give a small ε, and adaptive walls deform the wall shape to follow streamlines, suppressing the correction to roughly 1/20. The tunnel class selector in this tool switches the tcf multiplier automatically.
Maskell (1965) gives the wake blockage ε_w = θ·CD·A/C analytically and works for bluff bodies with strong separated flow (flat plates, car frontal projections). Mercker (1980s–) extends it into a combined correction (solid + wake + buoyancy gradients + horizontal buoyancy) and is the standard at Audi, BMW, Volvo and similar European car makers. This tool blends them: bluff = Maskell + Mercker combined, airfoil = solid blockage only, lifting body = wake counted at half weight.
Real-world applications
Automotive aerodynamics: The dominant use case. BMW, Audi, Toyota, Honda, Sauber and others apply Mercker as a standard correction and combine it with a moving belt (rolling road) to discuss differences as small as ΔCd = 0.003 on 0.20–0.30 baselines. The bluff-body mode in this tool reproduces exactly that automotive correction flow.
Aircraft & airfoil testing: Wings, airfoil sections and full-aircraft models have small wakes, so solid blockage dominates. The airfoil mode here counts ε_solid twice (top/bottom symmetric image effect). NASA, JAXA and ONERA transonic tunnels use slotted walls precisely to make the correction smaller in the first place.
Buildings & bridges: Tokyo Sky Tree, Tokyo Gate Bridge and the new National Stadium were tested in atmospheric boundary-layer tunnels at 1/300–1/500 scale, keeping BR at 1–3% and reproducing the ABL with roughness blocks on the floor. In structural-wind engineering the accuracy of Cf feeds directly into safety factors, so blockage correction directly affects design margins.
Wind turbines, trains and sports gear: Wind-turbine blade rigs, high-speed train nose shapes (Shinkansen / ICE / TGV), bicycles, helmets and ski-jump suits all require blockage correction. Train tests are particularly long-body and pick up tunnel longitudinal constraint, so full Mercker corrections including horizontal buoyancy are routine. Cross-checking against CFD (OpenFOAM, STAR-CCM+) is a common verification step.
Common pitfalls
The first big trap is the assumption that “once corrected, any BR is trustworthy.” Maskell and Mercker are linear expansions valid for small BR (≤ 5–10%); above BR ≈ 15% the correction itself departs from reality. The right first move is to physically lower BR — scale the model down or use a larger tunnel — not just to apply a bigger correction. F1 and automotive testing use scale models precisely to keep the correction small in the first place, not just to make it more reliable.
The second trap is treating open-jet tunnels as “correction-free.” ε is smaller in open jets, but it is not zero — it flips sign. Apply the correction in the wrong direction and you will under-estimate CD instead of over-estimating it. Switch tunnelClass from closed to open in this tool and watch ε change sign. Always log the tunnel type and the tcf you used, and keep both raw and corrected numbers in the report.
The third trap is the belief that only CD needs correcting. Blockage actually affects lift CL, pitching moment Cm and the entire pressure distribution Cp. On lifting bodies (aircraft wings, F1 rear wings) wake blockage feeds into CL correction too — that is why the lifting mode here counts wake at half weight. Full corrections are spelled out in Mercker's papers and Chapter 10 of Barlow-Rae-Pope, “Low-Speed Wind Tunnel Testing.” In practice, use this simple tool to confirm that BR is in the linear band; if it isn't, move on to full corrections or coupled CFD.
How to Use
Enter tunnel cross-section dimensions (height and width in meters) to calculate test section area.
Input model frontal area in m² and measured drag coefficient from pressure measurements.
Select blockage correction method (Maskell or Mercker) based on your tunnel geometry and Reynolds number regime.
The simulator calculates blockage ratio (%), interference factor ε, and applies corrections to velocity and CD automatically.
Review corrected drag coefficient and velocity values for freestream reference conditions.
Worked Example
Automotive model testing in closed-circuit tunnel: test section 2.0 m × 1.5 m (area = 3.0 m²), model frontal area = 0.42 m². Measured CD = 0.285 at blockage ratio 14%. Using Maskell correction, interference factor ε ≈ 0.042, yielding corrected CD = 0.298 and freestream velocity correction of +2.8% at nominal 40 m/s test speed (corrected = 41.1 m/s). This accounts for wall-induced static pressure rise reducing dynamic pressure at model location.
Practical Notes
Blockage ratio >15% requires correction; ratios >25% demand slotted-wall or adaptive-wall tunnels to suppress interference effects.
Maskell formula suits subsonic incompressible flow (M < 0.3); Mercker for higher speeds and compressibility effects in transonic regimes.
Automotive and aerospace models need separate frontal area measurements (plan view projection perpendicular to freestream) to avoid underestimating blockage penalties.
Corrected CD values should match published reference drag data (e.g., SAE J1594 standard bodies) within ±0.005 CD for method validation.