Why fans and passive heatsinks lose cooling capacity at altitude, how DBD plasma actuators behave at low pressure, and what to test. Talk to our engineers.
TL;DR. Air-cooled avionics lose cooling capacity as altitude rises because the air they rely on gets thinner. Fans keep moving roughly the same volume of air but much less mass, and natural convection weakens too. DBD plasma actuators respond to pressure differently: published experiments show their thrust first rises as pressure drops, then falls, which makes them an interesting candidate for low-pressure bays, provided the drive is tuned and the design is qualified for the target altitude.
On the ground, electronics cooling is mostly a heat flux and space problem. In the air, it is also a density problem. In the U.S. Standard Atmosphere, air density falls from about 1.225 kg/m³ at sea level to roughly 0.31 kg/m³ at 12,000 m, about a quarter of the sea-level value [1].
Convective cooling depends on the mass of air that passes over a hot surface, not just its volume. Unpressurised equipment bays, pods, UAV payload bays and high-altitude platforms all see some version of this. Even in pressurised aircraft, cabin altitude is typically held well above sea level, and qualification standards such as RTCA DO-160G require equipment to be tested across altitude and decompression conditions, not just at ground level [2].
The practical consequence: a box that runs comfortably on the bench can run hot at cruise.
An axial fan at fixed speed delivers roughly a fixed volumetric flow. As density drops, the mass flow it delivers drops with it. Less mass flow means less heat carried away per unit time and a lower convective heat transfer coefficient on the heatsink fins. The fan curve's static pressure capability also scales with density, so a fan that just met the system impedance at sea level may fall short at altitude.
Designers usually compensate in one of three ways:
Each option has a mass and reliability cost, and mass is expensive in anything that flies. Bearing wear also matters more in avionics than in consumer products because maintenance access is limited and unscheduled removals are costly. For a deeper comparison of fan failure modes, see Ionic Wind vs Fans for Electronics Cooling.
Natural convection is driven by buoyancy, and buoyancy-driven flow also weakens as density falls. Heatsinks sized for natural convection on the ground lose margin at altitude. Radiation does not depend on air density, which is why high-altitude and space hardware leans on emissive coatings and radiators, but radiation alone is rarely enough for dense electronics at moderate temperatures. Our piece on active vs passive cooling limits covers where heatsinks stop being enough even at sea level.
A DBD plasma actuator has no moving parts. A high-voltage AC signal across two electrodes separated by a dielectric ionises a thin layer of air, and the ions transfer momentum to neutral air molecules, producing a wall jet along the surface. That wall jet thins the thermal boundary layer directly at the hot surface.
Pressure changes the discharge itself, and this is where the behaviour departs from a fan:
Two cautions follow from the same data. First, power rises as pressure falls, so efficiency at altitude must be measured rather than assumed. Second, at very low pressures thrust collapses, so there is an operating envelope. The design levers are dielectric thickness, drive voltage and frequency, which means the actuator can be tuned for a target altitude band rather than just a sea-level point. For more on how dielectric choice affects this, see dielectric materials for plasma actuators.
| Factor | Axial fan | DBD plasma actuator |
|---|---|---|
| Response to falling pressure | Mass flow falls roughly with density at fixed speed | Thrust rises then falls; peak position depends on dielectric and drive [3][4] |
| Moving parts | Rotor and bearings | None |
| Thickness | Limited by blade and hub geometry | Thin electrode and dielectric stack |
| Vibration | Inherent from rotation | No mechanical vibration |
| Main tuning levers at altitude | Speed, fan size | Voltage, frequency, dielectric thickness |
| Open questions | Well characterised | Efficiency vs pressure, long-term dielectric ageing, EMI from the high-voltage supply |
This is not a claim that DBD actuators replace fans across avionics. Fans move far more air and remain the right answer for high-load racks. The opportunity is narrower: thin, sealed or vibration-sensitive modules where a fan cannot fit, and where the altitude derating of a fan is the reason the thermal design closes badly.
If you are evaluating any active cooling method for flight hardware, the data you need is specific to your box:
YPlasma builds DBD plasma actuators for electronics thermal management. If you have an avionics or UAV module where fan derating at altitude is the constraint, we can help scope a low-pressure test. Contact our engineering team.