NASA's NTAC study: brake a Mars lander on reheated Martian air, then reuse the reactor on the surface

What the work claims

This is a conceptual feasibility analysis, not a test report: a NASA Technical Memorandum whose own section titles call the results "Rough Estimations." The authors, from NASA Langley, Old Dominion University, and Tamer Space LLC, propose that a Nuclear Thermionic Avalanche Cell, a power conversion device patented by NASA in 2019, could heat ingested Martian atmosphere to a maximum of 2800 K during the supersonic entry corridor from 100 km down to 10 km altitude, adding an estimated 41% of average thrust to the retropropulsion budget, or an equivalent saving of Earth-carried propellant. After touchdown, the NTAC unit would detach and serve as the landed payload's primary power source for surface operations.¹

The load-bearing numbers, all from the memorandum's Tables 3 and 4: peak heating power of about 7 MW to hold ingested gas at 2800 K through the densest part of the corridor; a 7-MW-rated NTAC sized at 3,045 kg total, 1.684 m in diameter and 0.259 m tall, with a system alpha of 0.40 kg/kW; and a total specific impulse for the combined flow rising from about 259 seconds at 10 km altitude to about 792 seconds at 100 km. The memorandum also floats, but explicitly does not analyze, a 10% to 20% propellant saving for launch ascent from the Moon or Mars by reheating rocket exhaust.¹

How it works

The physics chain starts with the entry vehicle itself. As the vehicle descends through the rarefied Martian upper atmosphere (96% CO2), a bow shock forms on the forebody and compresses the free-stream gas to a density the memorandum takes as roughly ten times ambient, based on cited Mars entry ground-testing data. The concept puts a 60 cm-diameter ingestion port (0.2826 m²) at the center of the stagnation point, where the shock layer collapses into the port. The ingested mixture is assumed to be 90% free-stream gas and 10% shock-layer gas, with shock temperatures from Mars Exploration Rover entry data running from about 18,200 K at 100 km down to about 1,200 K at 10 km; at the assumed 9-to-1 ratio the inlet mixture reaches roughly 570 K to 2,160 K depending on altitude.¹

That pre-heated mixture is the propellant. The port splits into curved channels lined with heating elements powered by the NTAC, driving the gas to the 2800 K ceiling set by refractory heater and channel materials; the memorandum suggests porous refractory metal structures heated inductively. The hot gas expands through four or six nozzles placed near the forebody's trailing edge, a placement the authors cite supersonic-retropropulsion literature for as the most stable configuration. Thrust control comes free: each channel's flow share and heating rate can be modulated independently, giving differential thrust for attitude control. Two effects add: thermal expansion itself (about 220 seconds of specific impulse from the heated gas at most altitudes) and a pressure-difference contribution, because nose stagnation pressure exceeds the lower downstream static pressure where the nozzles sit. The pressure-driven flow dominates at high altitude, the thermal flow below about 25 km.¹

The power plant is the second half of the idea. NTAC, per the cited NASA patent, is a nuclear thermionic avalanche cell stacked in tandem with a thermoelectric (MJ-TE) stage. The memorandum selects a 7-MW-rated unit because its heating demand peaks near 20 km altitude: the text states a maximum requirement of 7.8 MW, while its own Table 3 lists 6,987 kW at 20 km, a discrepancy the paper does not reconcile. The selected unit's Table 4 line claims 8.607 MW of actual output (7.259 MW NTAC plus 1.348 MJ-TE, about 23% above rating), 3,045 kg total mass, and a system alpha of 0.40 kg/kW, with the NTAC core alone at 2.5 kW/kg. These are patent-derived paper sizes, not measured hardware.^{1 2}

Steelman: the best case for this concept

The strongest honest case is mass accounting. Every kilogram of retropropulsion propellant carried to Mars must first be landed on Mars, and the memorandum's core claim is that the atmosphere itself can substitute for part of that propellant: heat ingested CO2 the vehicle is already plowing through instead of decelerating only with tanks brought from Earth. A 41% average thrust contribution from reheated atmospheric gas, if it survived a real design cycle, would mean a smaller propellant fraction for the same entry mass or a steeper, more forgiving entry corridor. The dual-use framing sharpens this: the unit that provides the 7 MW during powered descent then detaches, per the cut line in the memorandum's Figure 3, and becomes a multi-year surface power source. Under that reading the reactor is not parasitic mass carried through entry; it is the one subsystem that pays for its own landing.¹

The second honest strength is specific power at scale. Table 4's sizing curve runs from 1 MW (0.63 kg/kW) to 15 MW (0.29 kg/kW) with the NTAC core reaching 3.45 kW/kg, and the pancake geometry, under 2 m in diameter and about 0.26 m thick, fits inside an entry vehicle's forward compartment rather than demanding a structural redesign. For a field that has watched surface fission studies repeatedly stall on radiator and shielding mass, a compact converter claiming kilowatts per kilogram is a serious design point, with the caveat that it is a claimed, patented design point and nothing in this memorandum measures it.¹

Where a skeptic should push

The most load-bearing assumption is the inlet flow itself. The entire thrust estimate rests on a 90/10 free-stream-to-shock-layer mixture ratio and the roughly tenfold shock-density jump, both imported from studies of blunted-cone entry flows that did not have a hole at the stagnation point. Open the port and the shock structure changes: the memorandum assumes the shock layer "collapses" into the port, but no test data, CFD run, or reference in the paper demonstrates that behavior for this configuration. If the ingested gas arrives colder or slower than assumed, the heating power climbs above the 7 MW already sized to the edge of the selected unit, and the 41% thrust gain shrinks accordingly. This is the classic pattern of a feasibility estimate whose critical fluid mechanic is asserted, not shown.¹

Second, the arithmetic does not fully reconcile, which matters for a mass-critical concept. The text's 7.8 MW peak does not match Table 3's 6,987 kW at 20 km; the Table 4 component masses (763 kg NTAC, 622 kg shielding, 485 kg thermoelectric for the 7-MW unit) sum to 1,870 kg against a stated 3,045 kg total, with the roughly 1.2 tonne difference never itemized; and the memorandum's conclusion concedes that "further study is required to quantify the effect of added mass vs. the increase in thrust," which is the study's real headline hiding in its last section. Third, on maturity: NTAC is a 2019 patent with paper-level sizing. No NTAC hardware appears in this memorandum, no test environment is named, and nothing here is bench, vacuum, or flight demonstration. The 10% to 20% ascent propellant saving is flagged by the authors themselves as not analyzed and should be ignored entirely.¹

What NTAC means for power in space

Non-obvious implication: this concept couples the two mass trades that planetary power architects normally run separately. The EDL mass budget and the surface power budget become the same number, because the power plant is landed by the maneuver it powers and then inherited by the payload. The memorandum's own framing, EDL assist first and excavators or 3D printers later, implies that the right way to evaluate any high-specific-power surface reactor is to subtract the landing-system mass it displaces, a subtraction that makes compact nuclear look better against solar-plus-storage than surface-only comparisons suggest.¹

Genuine threat: the sizing numbers risk becoming anchoring points in architecture studies that cite them without the caveats. A 0.40 kg/kW system alpha and a 41% thrust gain, both traceable to an unvalidated inlet-flow model and a patent, are exactly the kind of figures that get laundered into "fission-enabled Mars EDL" slide decks. If the inlet physics fails, the fix is not a parameter tweak: without shock-layer ingestion the concept needs stored propellant for the thermal stage, and the mass case collapses back toward conventional supersonic retropropulsion.¹

Opportunity: whatever happens to NTAC itself, the study identifies a testable, high-value experiment: a heated-flow supersonic retropropulsion demonstration using a port-at-stagnation inlet, measuring actual mixture temperature and mass capture against the 90/10 assumption. The supporting pieces are within reach: 2800 K channel heaters are terrestrial technology, Mars-atmosphere arc-jet facilities exist, and the MER-derived trajectory and atmosphere tables in this memorandum's Tables 1 and 2 are ready-made validation inputs. A ground test that confirmed even a fraction of the 41% would change how the community prices atmospheric-ingestion propulsion for crewed Mars landers.¹

The bottom line

Established by this memorandum: a physically coherent concept for atmospheric-ingestion retropropulsion powered by a compact nuclear conversion device, with a complete set of estimated performance tables and an honest statement that the added-mass trade remains unquantified. Not established: any NTAC hardware performance, any measured inlet-flow behavior, or any verified system mass; everything quantitative here is analysis built on cited entry data and a patented design, at concept-study maturity with no bench demonstration described. Confidence is moderate that reheating ingested CO2 can produce real thrust, since each physical link is individually plausible; low that the 41% figure survives inlet validation; and very low for the ascent-mode claims the authors themselves declined to analyze. What would confirm the concept is a supersonic wind-tunnel or arc-jet test of a stagnation-port inlet measuring mixture temperature and capture ratio; what would break it is that test showing shock-layer ingestion does not occur, which would leave a 3-tonne reactor to justify as surface power alone.