HALE Aerodynamics: Low Reynolds Number Testing

A HALE aircraft can have a wingspan longer than a Boeing 747's and still fly in an aerodynamic regime closer to a large model glider than to a jetliner. The reason is altitude, not size: air density at 20 km is well under a tenth of what it is at sea level, and Reynolds number scales directly with density. A big, slow, thin wing at the edge of the stratosphere ends up in the same low-Reynolds-number physics as a small one at sea level — it's just arrived there by a completely different route.

1. The aircraft that makes the point concrete

NASA's Helios Prototype, developed by AeroVironment under NASA's ERAST programme, is the clearest illustration on record. Its official specifications: a 247-foot (75.3 m) wingspan , an 8-foot chord , an aspect ratio of 30.9 to 1 , a wing only 12% of chord thick , and a maximum wing loading of just 0.81 lb per square foot . Cruise speed at low altitude was 19 to 27 mph — the speed of a bicycle. On 13 August 2001 it reached an altitude of 96,863 feet , a world record for sustained horizontal flight by a winged aircraft, with a design objective of operating up to 100,000 feet.

Every one of those numbers — the huge aspect ratio, the paper-thin wing, the near-zero wing loading, the bicycle-speed cruise — is a direct consequence of designing for extreme altitude rather than extreme speed. Even Helios's propellers were purpose-built: two-blade, wide-chord, 79-inch diameter units with a laminar-flow design specifically for efficiency at high altitude .

2. What actually happens to an airfoil at low Reynolds number

At the Reynolds numbers typical of HALE propeller and wing sections — commonly cited in the range of 10⁴ to 10⁶ — the boundary layer behaves fundamentally differently from the high-Re flow most airfoil intuition is built on. Documented low-Reynolds-number wind tunnel testing shows the mechanism clearly: laminar flow separates from the surface under an adverse pressure gradient before it can transition to turbulent flow. The separated shear layer then transitions in mid-air, and if it reattaches downstream, it forms a laminar separation bubble (LSB) — a region of recirculating, largely stagnant flow bounded by separation and reattachment lines that are directly visible in surface oil-flow visualization.

The consequence is not subtle: pressure drag generated over the separation bubble is the leading cause of degraded lift-to-drag ratio at low Reynolds numbers, and it can appear or worsen sharply as Reynolds number drops — exactly the direction altitude pushes a HALE design.

3. Why this demands a specific kind of wind tunnel

Because low-Reynolds-number airfoil behaviour is governed by the laminar boundary layer, test-section turbulence intensity has to be low enough that it doesn't itself trigger early transition — otherwise the tunnel is testing a different, artificially "tripped" boundary layer than the one the aircraft will actually see. Documented practice at facilities purpose-built for this regime achieves turbulence intensity below 0.1%, validated with hot-wire anemometry, together with flow-angle uniformity within about ±0.1° and dynamic-pressure uniformity within ±0.5% across the working section.

The practical implication for a HALE programme: the model has to be sized and the tunnel run at the speed that reproduces the flight Reynolds number , not the flight speed. A model tested "too fast" for its size sits at a higher Reynolds number than the real aircraft will ever see at altitude, and the separation bubble — the exact feature the test exists to characterise — moves or disappears.

4. What a HALE-relevant test campaign needs to capture

Surface flow visualization , not just force data. Oil-flow or equivalent visualization directly shows the laminar separation, transition and reattachment lines that define the bubble — data a balance alone cannot provide.
A genuine low-Reynolds-number sweep , run at the actual altitude-equivalent Reynolds number for the mission profile, not extrapolated from higher-Re data.
Drag measured by wake survey, not balance alone , since profile drag at these Reynolds numbers is small relative to lift and better resolved through momentum-deficit methods in the wake.
Attention to flap and control-surface behaviour at low Re specifically — deflected control surfaces show dramatically increased drag and reduced effectiveness at low Reynolds numbers compared with the same geometry at higher Re, a result that doesn't extrapolate cleanly from higher-speed testing.

5. Five mistakes we see

Sizing the model by convenience rather than by target Reynolds number. A model built to a "nice" scale can land at the wrong Re entirely, testing a different flow regime from the one that matters.
Running the tunnel at a speed the balance likes rather than the Reynolds number the mission needs. At low Re, the physics — not instrumentation comfort — has to set the test point.
Treating tunnel turbulence intensity as a rounding error. At this Reynolds regime, turbulence a fraction of a percent above what's needed can trip the boundary layer early and erase the separation bubble the test was meant to characterise.
Skipping flow visualization because force data "looks clean." A clean-looking polar can still hide a separation bubble in the wrong place; visualization is how you find out.
Extrapolating control-surface data from a higher-Reynolds-number test. Flap and elevator effectiveness at low Re does not scale the way it does at conventional aircraft Reynolds numbers.

6. What this means for your programme

Tell us your target altitude and mission profile , and we'll size a model and set a test Reynolds number that reproduces the actual flow regime your aircraft will fly in — not the one that happens to suit a convenient tunnel speed. Related reading: our hub article on

[Article not found: long-range-fixed-wing-uas-wind-tunnel-testing]

, and

[Article not found: hale-aeroelasticity-gust-response-wind-tunnel-testing]

for the same aircraft class.

Sources and notes

Helios Prototype specifications (wingspan 247 ft, chord 8 ft, aspect ratio 30.9:1, wing thickness 12% of chord, maximum wing loading 0.81 lb/sq ft, cruise speed 19–27 mph, altitude record 96,863 ft on 13 August 2001, design objective of 100,000 ft, laminar-flow propeller design for high-altitude efficiency): NASA Dryden Flight Research Center, "Helios Prototype: The forerunner of 21st century solar-powered 'atmospheric satellites'," FS-2002-08-068 DFRC.
Laminar separation bubble mechanism (laminar separation before transition, free-shear-layer transition, turbulent reattachment, pressure drag as the leading cause of degraded lift-to-drag ratio at low Reynolds numbers), surface oil-flow visualization methodology, and low-Reynolds-number test-section flow-quality standards (turbulence intensity below 0.1%, flow angle within ±0.1°, dynamic pressure uniformity within ±0.5%): Selig, M.S., Deters, R.W., Williamson, G.A., "Wind Tunnel Testing Airfoils at Low Reynolds Numbers," AIAA 2011-875, 49th AIAA Aerospace Sciences Meeting, 2011, tested in the University of Illinois at Urbana-Champaign low-turbulence subsonic wind tunnel across Reynolds numbers 40,000 to 500,000, and validated against NASA Langley Low-Turbulence Pressure Tunnel data.
Reduced flap and control-surface effectiveness with dramatically increased drag at large deflections at low Reynolds number: same source, results on the AG455ct airfoil with a 30%-chord flap.
HALE propeller and wing sections operating in the 10⁴–10⁶ Reynolds number range, and high-altitude aircraft and propellers as an application area for low-Reynolds-number aerodynamics, is broadly consistent with the low-Reynolds-number aerodynamics literature referenced by Selig, Deters and Williamson above.
Model-sizing and test-condition guidance (matching flight Reynolds number rather than flight speed) is TunnelTech engineering practice, consistent with our published article on sizing a wind tunnel for UAV testing.

High-Altitude HALE Aerodynamics: Low Reynolds Numbers and Thin Air