NASA ARC Arc Jet Complex

NASA. Public domain (NASA / US government work).

The Arc Jet Complex at NASA ARC is the agency’s principal thermal protection system test installation: four named arc jet facilities across seven test bays in two laboratory buildings, drawing on shared direct current power supplies, a steam ejector vacuum system, deionized cooling water, high pressure gas and common controls, and used to simulate the heating rates of Earth or planetary hypersonic entry so that thermal protection systems and materials can be selected, validated and qualified [1][2].

Laboratory data

Parameter	Value
Operator	NASA ARC, Thermophysics Facilities Branch [1][4][6]
Location	Moffett Field, California, United States [1]
Commissioned	In thermal protection service since the 1960s; modernization study 2022 [1][2]
Type	Arc-heated hypersonic entry simulation for thermal protection systems [1]
Floor area	Not published; seven test bays in buildings N-234 and N-238
Capabilities	IHF, AHF, PTF, TFD, LEAF, mARC II [1][2][6]
Simulant or terrain	Not applicable; the working fluid is an arc-heated gas stream
Instrumentation	Water-cooled copper calibration plates with Gardon gauges [5]
Ground truth	Calibration plate run of the test article geometry ahead of the material sample
Fidelity limits	Monochromatic radiant source; no single leg spans the envelope [1][5]
Access	Not published as a schedule; modeled mean throughput 483 runs per year
Cited by	Not applicable

The scale of the supporting plant is the reason the complex exists as one installation rather than four. The facilities share two direct current power supplies, a steam ejector vacuum system, a deionized water cooling system, high pressure gas systems and controls, and the magnitude and capacity of that common equipment is the stated reason the complex is unique in aerospace testing. The steam vacuum system is a five-stage steam ejector pumping network about 300 ft long and 100 ft across at its widest, and it is what allows high-altitude flight conditions to be matched with test articles of relatively large size [2]. Cooling water is deionized because the arc heaters require coolant of very low electrical conductivity [2]. Building N-234 holds the Aerodynamic Heating Facility, on a 20 MW power supply and comprising three heaters, AHF Constricted, AHF Huels and TP3, and the Turbulent Flow Duct, also on a 20 MW supply. Building N-238 holds the Interaction Heating Facility on a 60 MW supply, which carries the highest traffic because of the flow rates and enthalpy it provides, and the Panel Test Facility on a 20 MW supply [2]. The Thermophysics Facilities Branch’s own test planning guide states the bookable envelope across the arc jet streams as input power 10 to 60 MW, bulk enthalpy 1,500 to 12,000 Btu/lbm and surface pressure 0.01 to 75 kPa, and fixes the customer-facing process of model size limits, instrumentation channel counts and scheduling that a campaign has to clear before a run [6]. Two heater families are run across the complex: a segmented, constricted heater reaching 28 MJ/kg bulk enthalpy at pressures to 1 MPa with electrode-material contamination below 10 ppm, and a Huels, vortex-stabilized heater reaching 1.7 MPa stagnation pressure but limited to about 9 MJ/kg and a higher contamination level [6].

Capabilities

IHF

Dominic Hart. Public domain (NASA / US government work).

Parameter	Value
Working volume	Conical nozzles 152 mm to over 1 m exit diameter [1]
Test article limits	Flat panels to 610 x 610 mm on the semi-elliptical nozzle
Vacuum	Stagnation pressures 0.01 to over 1 atm, on the steam ejector system [1][2]
Temperature	7 to 47 MJ/kg bulk enthalpy; heat flux 5 to over 6000 kW/m2 [1]
Illumination	Radiant heating only in combination with LEAF [5]
Slope	Not applicable; article inclination set by the sting or panel holder
Gravity offload	Not applicable
Instrumentation	Gardon gauge calibration plate of matching geometry

The Interaction Heating Facility runs a 60 MW Ames-designed constrictor arc heater at 1 to 9 atm arc pressure, and it is the highest-traffic leg of the complex because of the combination of flow rate and enthalpy it delivers [1][2]. Its power supply can deliver 75 MW for 30 minutes or 150 MW for 15 seconds. Interchangeable conical nozzles set the stream size, and a semi-elliptical nozzle converts it into a panel facility.

AHF

Parameter	Value
Working volume	Conical nozzles 76 to 914 mm exit diameter
Test article limits	Samples to 203 mm diameter, or a 660 x 660 mm wedge
Vacuum	Chamber pressures 0.005 to 0.125 atm
Temperature	11 to 33 MJ/kg constricted; 3.5 to 9.5 MJ/kg on the Huels heater
Instrumentation	Five-arm fully programmable insertion system, up to five samples per run

Source: [1].

The Aerodynamic Heating Facility can run either a 20 MW constricted heater at 1 to 9 atm or a Hüls heater at 1 to 40 atm, which is the highest arc pressure in the complex and the reason the leg exists alongside IHF [1]. The five-arm insertion system is what makes it the throughput facility for materials screening: five articles can be exposed in a single run rather than one.

PTF

Parameter	Value
Working volume	Semi-elliptical nozzle on a 20 MW constricted heater at 1 to 9 atm [1]
Test article limits	Panels of about 355 x 355 mm, surface inclination -4 to +8 degrees
Vacuum	Surface pressures 66 to 4700 Pa
Temperature	7 to 35 MJ/kg; cold wall heat flux 6 to 340 kW/m2
Instrumentation	Runs of up to 30 minutes; Gardon gauge calibration plate [1][5]

The Panel Test Facility produces a flat-panel environment rather than a stagnation-point one. The same panel environment can be recreated in AHF or IHF with a semi-elliptical nozzle of the same family, so PTF is a routing choice rather than a unique capability, and its long run duration is what distinguishes it in practice [1][2].

TFD

Parameter	Value
Working volume	Supersonic rectangular duct, 2 by 9 in nozzle, on a 12 MW Huels heater [1][2]
Test article limits	Models 203 mm high by up to 508 mm long [1]
Vacuum	Surface pressures 0.02 to 0.15 atm
Temperature	3 to 9 MJ/kg; cold wall heat flux 20 to 700 kW/m2
Instrumentation	Shear stress 5 to 70 kg/m2 across the model surface

The Turbulent Flow Duct is the shear facility: instead of a free jet onto a stagnation point, it drives turbulent flow along a flat surface, which is the loading that erodes tile gaps, seams and coatings. It is the only leg for which surface shear stress is a published test parameter [1].

LEAF

Parameter	Value
Working volume	Four beams routed into the IHF test box
Test article limits	152 x 152 mm wedge, or 432 x 432 mm panel in the spread configuration
Illumination	Four 50 kW continuous-wave ytterbium fiber lasers, 1070 +/- 4 nm
Temperature	Radiative flux to 375 W/cm2 on a wedge; 551 W/cm2 combined with IHF
Instrumentation	Six Gardon gauges each on radiative and convective calibration plates

Source: [5].

The Laser Enhanced Arc jet Facility adds what an arc jet cannot produce. Uniform radiative flux is delivered to within plus or minus 6 percent across the article [5]. In the wedge configuration the system has provided up to 375 W/cm2 on a 152 by 152 mm article, and combining that with the 160 W/cm2 of convective heating IHF produces on a wedge behind a 9 in nozzle gives a combined 551 W/cm2; the panel configuration spreads the same four beams over a 432 by 432 mm square and is designed for 100 W/cm2 [5]. That separates radiative from convective heating in a way the plasma stream alone does not, and it is a requirement for EM-2 certification of the Orion thermal protection system.

mARC II

Parameter	Value
Working volume	30 kW segmented constricted heater on a Hypertherm MAX200 torch cathode [3][9]
Test article limits	Small material and instrument samples; scale not published
Vacuum	Base pressure 4 Pa after the vacuum system upgrade, versus 53 Pa before it [7]
Temperature	Bulk enthalpy 4 to 14 MJ/kg by energy balance or sonic-flow correlation [3][7][9]
Simulant or terrain	Working gas air, nitrogen or carbon dioxide
Instrumentation	Water-cooled Gardon gauge on a sweep arm; optical emission spectroscopy of the freestream [3][8]

Source: [3][7][8][9][10].

mARC II is the small-scale counterpart, built because operational complexity accompanies the high-power facilities and some programs are better suited to a smaller scale, whether for budget reasons or to test many materials or instruments quickly. Strongback, constrictor disks, anode and nozzle are oxygen-free high conductivity copper [3]. The primary distilled water loop cools both arc heater and chamber walls at about 150 L/min, and a dedicated 45 L high-pressure loop cools sensors at up to 5.7 L/min during insertion into the jet; gas supply runs through Sage Prime SIP-030-DC24 thermal mass flowmeters.

The facility’s measured envelope has moved in both directions since the initial characterization. Moving the Gardon gauge downstream to 52 and 68 mm from the nozzle exit extended the cold-wall heat flux floor down to 200 W/cm2, below the 800 to 1900 W/cm2 range of the first-generation mARC, at the cost of a systematic instrumentation defect in which measured arc current reads 0.923 times the set current [9]. A later vacuum system upgrade, a booster pump, water-cooled diffuser and heat exchanger, cut base pressure roughly 13-fold and let the nozzle run underexpanded rather than back-pressure-limited, which lowered the achievable stagnation heat flux by about a factor of four at the same current and flow settings and produced the lowest fluxes measured in the facility to date, 26 to 81 W/cm2 [7]. That result reframes the earlier 200 W/cm2 floor as an artifact of chamber back pressure rather than a limit of the heater. A laminar Navier-Stokes simulation of the same runs matched measured heat flux to within 3 to 15 percent at low arc current but diverged to 49 percent at the highest current tested, for reasons the authors could not isolate [7].

Optical emission spectroscopy of the freestream, 25 mm downstream of the nozzle exit, finds copper and silver electrode-erosion products and sodium, lithium and potassium impurities in every run tested, the sodium traced to a sodium nitrite additive in the electrode grease rather than to any test sample, which establishes the impurity as facility-inherent [8]. Anode erosion peaks during the unstable-current arc startup transient, so ramping the mass flow up slowly rather than opening the gas valve at once measurably cut startup copper emission [8]. Run duration at the facility is set by the model support hardware rather than the heater: an uncooled sweep-arm feedthrough capped runs at about 450 s, and only at minimum arc power and maximum standoff from the nozzle, until a water-cooled aluminum sleeve over the feedthrough let customer testing, including SpaceX and Blue Origin articles, run to 750 s at higher power and closer to the nozzle [10].

Instrumentation

Heat flux in a panel or wedge test is calibrated by running a water-cooled copper calibration plate of the same geometry as the test article, instrumented with Gardon gauges at fixed locations, ahead of the material samples; in LEAF the radiative and convective calibration plates carry six Gardon gauges each at matching positions, and the convective plate also takes surface pressure [5]. The complex publishes a measurements handbook as controlled branch-level documentation [4].

mARC II is instrumented for enthalpy as well as heat flux. A differential temperature sensor and flowmeter on the primary 750 L distilled water loop provides an estimate of the bulk enthalpy delivered to the plasma by energy balance, which is a facility-level measurement the larger legs obtain differently [3]. That energy-balance estimate generally reads lower than the sonic-flow correlation used for the larger arc heaters, reported for mARC II for the first time alongside the low-flux vacuum upgrade campaign [7]. Erosion and contamination are read from line-of-sight optical emission spectroscopy rather than from electrode mass loss, which biases the derived plasma temperature toward the hot jet centerline [8].

What it does not reproduce

Coverage in one place. No single leg spans the envelope. The four facilities differ in heater, pressure range and enthalpy range, and a panel environment that PTF produces with a semi-elliptical nozzle can also be recreated in AHF or IHF with the same nozzle family, so a campaign is routed to whichever bay reaches its conditions rather than to one large facility [1][2].

The radiation spectrum, in the laser facilities. Shock-layer radiation during entry spans the ultraviolet into the infrared, and there is no light source that can illuminate a wedge or panel at high heat flux across that range. LEAF’s fiber lasers are monochromatic at 1070 nm, which was accepted as the price of reaching the power levels required; a broader-spectrum ground facility awaits practical multi-kilowatt visible and ultraviolet lasers [5].

Chamber pressure at the small scale. mARC II’s as-built vacuum configuration holds the chamber below approximately 3 kPa through a test, which the operators note is higher than optimal, producing a slightly over-expanded jet; modifications to increase system conductance were under consideration [3].

Run duration and throughput. A PTF run can last up to 30 minutes, and exposures across the complex have run from a few minutes to over an hour [1]. Throughput is the binding constraint at the complex level: a Monte Carlo model of the legacy complex, on notional inputs of 250 working days, three test runs per day for complex testing or five for simple, and per-facility availability of 20 to 40 percent, put the mean at 483 test runs a year [2]. At mARC II, run duration is limited by the model support hardware rather than by heater power or flow, and was extended from about 450 s to 750 s by cooling the sweep-arm feedthrough rather than by any change to the arc itself, which is a caution against reading a legacy duration figure as a heater limit anywhere in the complex [10].

Instrumentation artifacts at the small scale. mARC II’s arc current measurement reads a consistent 0.923 of the set current across every condition run, for a cause the operators have not isolated, and its vacuum system state changes the achievable heat flux at a fixed arc setting by a factor of four, so a flux number from this facility is not comparable across campaigns without knowing which vacuum configuration and which current calibration produced it [7][9].

Campaigns run there

Thermal protection system development, 1960s to present. The complex has supported Apollo, Space Shuttle, Viking, Pioneer-Venus, Galileo, Mars Pathfinder, Stardust, X-33, X-34, SHARP-B1 and B2, X-37 and the two Mars Exploration Rovers, in the roles of material selection, validation and qualification [1].

Orion thermal protection, LEAF, 2018 onward. LEAF testing of Orion thermal protection, including the tiled Avcoat heatshield and its compression pads, is a requirement for EM-2 certification of flight readiness, and it is run as a combined radiant and convective exposure in the IHF test box rather than in either mode alone [5].

mARC II characterization, 2020 to 2024. Initial characterization of the upgraded second-generation miniature arc jet reported measured arc current and voltage, column and chamber pressures, mass flow rates and initial column diagnostics, at settings including 40 A and 100 A at 0.25 g/s of air, with the facility undergoing integrated systems testing at the time [3]. Later campaigns pushed the envelope down to 200 W/cm2 by relocating the heat flux gauge [9], then down again to 26 to 81 W/cm2 after a vacuum system upgrade removed a back-pressure limit that the earlier campaigns had mistaken for a heater floor [7]. Optical emission spectroscopy of the freestream, run separately, established that electrode erosion and grease-borne sodium contaminate the flow independent of any test article [8].

Customer testing with an extended run-duration envelope, mARC II. After a water-cooled sleeve was fitted over the model support’s vacuum feedthrough, customer runs for SpaceX and Blue Origin articles reached 750 s at higher power and closer standoff than the uncooled 450 s limit allowed, validated by a support thermocouple that read 160 C against an 87 K more conservative thermal-model prediction [10].