Human Mars lander: the EDL problem above 20 tonnes

How to read this dossier

This dossier connects physical phenomena, sensors, guidance and braking systems while consistently distinguishing three levels: what has already flown on Mars, what is studied for heavy human payloads, and what remains a prospective choice.

Calculation numbers are labelled as teaching examples; historical values or mass classes are tied to NASA sources when used.

Human Mars lander: the EDL problem above 20 tonnes — Functional architecture: details of a real vehicle depend on its mission.

1. Mars EDL: understand entry, descent, landing and the energy to dissipate

Question : Why can only a few minutes separate an interplanetary vehicle from a landed spacecraft?

EDL is a chain of events in which the kinetic energy of a vehicle arriving at several kilometres per second must be dissipated, converted or controlled without losing stability, navigation or landing capability. A robotic mission and a future human vehicle share physical functions, but not necessarily the same technologies or masses.

The first discipline is to separate what physics imposes from what architecture chooses. For énergie, séquence, autonomie, marges, a relationship may be certain while the numerical value still depends on mission, mass, altitude, weather or design margin. This prevents a teaching example from becoming a universal rule.

A second discipline is to follow interfaces. During EDL, thermal conditions affect structure, structure affects mass, mass affects deceleration, navigation affects guidance, guidance consumes aerodynamic or propulsive margin, and landing immediately sets conditions for surface operations. A local value is therefore never isolated.

Useful relation : E_k = 1/2 × m × v²
E sub k equals one half times mass m times speed v squared.

Chronology is also an engineering variable. An event may be correct by itself but dangerous if it occurs too early, too late, with a sensor not yet valid or an actuator outside its envelope. Engineers therefore verify entry conditions for each mode, exit criteria and degraded transitions.

Finally, carrying crew does not change fundamental equations, but it changes risk tolerance, mass, redundancy, abort logic, physiological constraints and qualification. Robotic heritage is therefore experimental evidence, not an automatically transferable human architecture.

2. Ballistic coefficient, lift and entry corridor: choosing how to cross the atmosphere

Question : Why can two vehicles with the same mass decelerate very differently?

Ballistic coefficient relates mass, reference area and drag coefficient. A high value means, all else equal, that a vehicle penetrates farther before losing speed. Lift and bank angle add steering authority. The entry corridor is the set of conditions that avoids skip-out, excessive heating, excessive acceleration and an overly short trajectory.

The first discipline is to separate what physics imposes from what architecture chooses. For balistique, portance, corridor, géométrie, a relationship may be certain while the numerical value still depends on mission, mass, altitude, weather or design margin. This prevents a teaching example from becoming a universal rule.

A second discipline is to follow interfaces. During EDL, thermal conditions affect structure, structure affects mass, mass affects deceleration, navigation affects guidance, guidance consumes aerodynamic or propulsive margin, and landing immediately sets conditions for surface operations. A local value is therefore never isolated.

Useful relation : β = m / (C_D × A) ; L/D = C_L/C_D
beta equals mass m divided by drag coefficient C sub D times area A; L over D is lift-to-drag ratio.

Chronology is also an engineering variable. An event may be correct by itself but dangerous if it occurs too early, too late, with a sensor not yet valid or an actuator outside its envelope. Engineers therefore verify entry conditions for each mode, exit criteria and degraded transitions.

Finally, carrying crew does not change fundamental equations, but it changes risk tolerance, mass, redundancy, abort logic, physiological constraints and qualification. Robotic heritage is therefore experimental evidence, not an automatically transferable human architecture.

3. Supersonic parachutes on Mars: operation, loads and scaling limits

Question : Why can a parachute that works for a probe not simply be scaled to a vehicle of tens of tonnes?

A parachute converts part of kinetic energy into drag through a large textile area. Deployment is a violent dynamic event: inflation, oscillation, line loads, wake interaction and Mach-number sensitivity. On Mars, thin air demands large areas while heavy vehicles increase loads. Parachutes remain a major robotic heritage technology but are not automatically the final solution for heavy human EDL.

The first discipline is to separate what physics imposes from what architecture chooses. For parachute, supersonique, déploiement, échelle, a relationship may be certain while the numerical value still depends on mission, mass, altitude, weather or design margin. This prevents a teaching example from becoming a universal rule.

A second discipline is to follow interfaces. During EDL, thermal conditions affect structure, structure affects mass, mass affects deceleration, navigation affects guidance, guidance consumes aerodynamic or propulsive margin, and landing immediately sets conditions for surface operations. A local value is therefore never isolated.

Useful relation : D = 1/2 × ρ × v² × C_D × A
drag D equals one half times density rho times speed squared times drag coefficient and area.

Chronology is also an engineering variable. An event may be correct by itself but dangerous if it occurs too early, too late, with a sensor not yet valid or an actuator outside its envelope. Engineers therefore verify entry conditions for each mode, exit criteria and degraded transitions.

Finally, carrying crew does not change fundamental equations, but it changes risk tolerance, mass, redundancy, abort logic, physiological constraints and qualification. Robotic heritage is therefore experimental evidence, not an automatically transferable human architecture.

4. Supersonic retropropulsion: braking a heavy lander with engines

Question : Why ignite engines while the vehicle is still supersonic?

Supersonic retropropulsion uses thrust opposing motion before the vehicle becomes subsonic. It becomes particularly attractive when mass exceeds what a conventional parachute can reasonably handle. The problem is not only to make enough thrust: engines must restart, throttle, gimbal, receive propellant and remain controllable amid complex jet-supersonic-flow interaction.

The first discipline is to separate what physics imposes from what architecture chooses. For rétropropulsion, poussée, ergols, moteurs, a relationship may be certain while the numerical value still depends on mission, mass, altitude, weather or design margin. This prevents a teaching example from becoming a universal rule.

A second discipline is to follow interfaces. During EDL, thermal conditions affect structure, structure affects mass, mass affects deceleration, navigation affects guidance, guidance consumes aerodynamic or propulsive margin, and landing immediately sets conditions for surface operations. A local value is therefore never isolated.

Useful relation : T ≈ m × (g_Mars + a) ; ṁ = T/(I_sp × g_0)
thrust T is approximately mass times Mars gravity plus desired upward deceleration; mass flow m dot equals T divided by Isp times g zero.

Chronology is also an engineering variable. An event may be correct by itself but dangerous if it occurs too early, too late, with a sensor not yet valid or an actuator outside its envelope. Engineers therefore verify entry conditions for each mode, exit criteria and degraded transitions.

Finally, carrying crew does not change fundamental equations, but it changes risk tolerance, mass, redundancy, abort logic, physiological constraints and qualification. Robotic heritage is therefore experimental evidence, not an automatically transferable human architecture.

5. Heavy human EDL: why moving from 1 tonne to more than 20 tonnes changes the architecture

Question : What stops being a simple extrapolation when landed mass increases by more than a factor of twenty?

Robotic Mars heritage demonstrates remarkable technologies, but human missions impose a different mass class. NASA EDL syntheses compare Mars robotic heritage, around the one-tonne landed class for Mars 2020, with human objectives around 20 tonnes and above, including studied concepts in the 26 to 36 tonne landed class. This is not one official final architecture: it is a problem scale motivating new entry systems, guidance, propulsion, surface sensing and test campaigns.

The first discipline is to separate what physics imposes from what architecture chooses. For humain, échelle, masse, architecture, a relationship may be certain while the numerical value still depends on mission, mass, altitude, weather or design margin. This prevents a teaching example from becoming a universal rule.

A second discipline is to follow interfaces. During EDL, thermal conditions affect structure, structure affects mass, mass affects deceleration, navigation affects guidance, guidance consumes aerodynamic or propulsive margin, and landing immediately sets conditions for surface operations. A local value is therefore never isolated.

Useful relation : facteur = valeur_2 / valeur_1 ; A ∝ diamètre²
scale factor is the ratio of two values; for similar shapes, characteristic area scales with diameter squared.

Chronology is also an engineering variable. An event may be correct by itself but dangerous if it occurs too early, too late, with a sensor not yet valid or an actuator outside its envelope. Engineers therefore verify entry conditions for each mode, exit criteria and degraded transitions.

Finally, carrying crew does not change fundamental equations, but it changes risk tolerance, mass, redundancy, abort logic, physiological constraints and qualification. Robotic heritage is therefore experimental evidence, not an automatically transferable human architecture.

6. Sequence matters as much as each technology

EDL is not the independent sum of a heat shield, parachute, radar and engines. Each phase hands the next one a state of velocity, altitude, attitude, position and health. A phase that technically succeeds while spending too much margin can still cause later failure.

This point must be connected to mass, power, data, time, propellant and risk budgets. A decision improving one budget may worsen another; justification must therefore be preserved as a traceable system decision.

7. Uncertainty and dispersion

A robust architecture explicitly handles atmospheric-density distributions, navigation errors, mass variation, actuator performance and delays. The nominal trajectory is tested across dispersed cases to show thermal, structural and propulsive limits remain satisfied.

This point must be connected to mass, power, data, time, propellant and risk budgets. A decision improving one budget may worsen another; justification must therefore be preserved as a traceable system decision.

8. Earth-Mars distance forces autonomy

EDL unfolds far too quickly to be piloted from Earth. Critical decisions must be onboard: event detection, state estimation, guidance, mode selection, anomaly management and safe continuation of the sequence.

This point must be connected to mass, power, data, time, propellant and risk budgets. A decision improving one budget may worsen another; justification must therefore be preserved as a traceable system decision.

9. A human landing prepares surface operations

Site, plume effects, dispersion, habitat proximity, emergency access and logistics change the meaning of “landed”. Success is not merely zero vertical speed: the vehicle must be stable, usable and compatible with operations that begin immediately afterward.

This point must be connected to mass, power, data, time, propellant and risk budgets. A decision improving one budget may worsen another; justification must therefore be preserved as a traceable system decision.

Budgets, margins and sizing cases

An EDL dossier cannot be sized from one average speed or a single mass. Several budgets must be tracked together: entry mass, propellant mass, available electrical power, battery energy, computing capacity, data rate, thermal load, acceleration, control authority and time remaining before each event. Each budget has its own margin and calculation convention.

The sizing case is not necessarily the nominal case. A denser atmosphere can increase some aerodynamic and thermal loads, while a thinner atmosphere may reduce braking and force propulsion to take over more work. A lighter vehicle does not automatically make every constraint easier either, because centre of gravity, ballistic coefficient and aerodynamic authority may change.

Good practice therefore publishes not only a target value but also its boundary, dispersion, available margin and the mode consuming that margin. This prevents isolated numbers from creating false precision.

Qualification and test strategy

No single Earth test reproduces Mars gravity, atmospheric composition and density, hypersonic speed, full scale, propulsion plumes and the real surface at the same time. Qualification is therefore a mosaic of evidence: material tests, wind tunnels, arc jets, engine tests, avionics benches, numerical simulation, integrated tests and flight data.

Each piece of evidence must be tied to a requirement. A material surviving a heat flux does not prove complete structural stability; an engine restart on a test stand does not prove that a multi-engine system will restart correctly inside coupled supersonic flow; a camera recognising terrain in a laboratory does not yet prove performance with Mars dust, vibration and lighting.

The test programme must deliberately seek regions where phenomena couple. The farther a test is from the final environment, the more clearly the transfer model and its uncertainty must be stated.

Observability, telemetry and state truth

During EDL, software acts on a state it never knows perfectly. Position, velocity, attitude, altitude, ground-relative speed, engine health and environment are estimated from imperfect sensors. Observability asks whether the available measurements can actually distinguish states that require different decisions.

Coherent telemetry is equally important after flight. Learning from EDL requires reconstruction of chronology, commands, measurements, model deviations and abnormal events. Flight data then improve simulations, margins and later generations.

For a human mission, telemetry must not become dependence on Earth. The vehicle has to complete the sequence even if the interplanetary link is unavailable. Communications support awareness and later analysis, not real-time piloting from Earth.

Human factors and safety criteria

Crew adds requirements that cannot be reduced to adding seats. Acceleration, vibration, noise, body orientation, life-support availability, abort possibilities, fire, loss of pressure and the ability to evacuate after touchdown all change the design problem.

Human-machine interfaces must remain understandable during a phase in which most decisions are automated. Crew need to know what the system is doing, which anomalies require action, and which actions would be dangerous because they interrupt a very fast sequence. Alerts therefore need hierarchy and procedures that are executable in the time available.

Human safety also forces scrutiny of common-cause failure. Two redundant chains sharing the same power source, software or environmental sensor can fail together. Redundancy only has value when independence is sufficient.

Interface with the base, logistics and later missions

Touchdown location immediately affects colony logistics. Long distance increases transfer time for people and cargo; excessive proximity can expose habitats, solar arrays, antennas or vehicles to plumes, regolith ejecta and accident risk. The landing site therefore becomes infrastructure in its own right.

A mature architecture must plan for repetition. Early missions may accept an isolated zone, but a base receiving regular cargo will need approach corridors, exclusion zones, beacons, hazard maps, recovery routes and inspection after each landing. Cumulative effects on the surface must be monitored.

Post-landing use matters as well. A lander later serving as tankage, habitat, power source or structure has different requirements from a discarded stage. Surface mission needs therefore belong in EDL definition from the beginning.

Unknowns, reversible decisions and continued research

A technical Bible must state what is not yet closed. Heavy human EDL performance depends on choices still under study: entry geometry, thermal materials, lift level, engines, propellants, throttling depth, engine count, surface sensors, divert strategy and landing-site organisation.

When an uncertainty is important, the architecture should avoid freezing a solution too early. Interfaces can preserve options through power reserve, compute capacity, data ports, integration volume or mass margin. Flexibility has a cost, but it can prevent complete redesign when a technology changes.

Knowledge advances by progressively closing uncertainty. Each test, robotic mission, engine demonstration or navigation improvement should be tied to the question it reduces. That traceability turns an accumulation of facts into manageable architecture.

Primary NASA sources

Sources document heritage and studies; they do not validate any specific Delta-Sierra concept.