Reusable Chemical Mars Landing Boats Are Feasible

I was initially quite skeptical that a single-stage chemical lander could be feasible for one-shot use, much less reuse-for-multiple trips,  with refueling.  But,  after some investigation with bounding analyses,  I have changed my mind.  These things are very feasible.  So,  we need not incur the difficulties of solid core nuclear propulsion,  in order to have a very practical ferry capability between surface and orbit at Mars.

Propellant Options

I do not know what propellants might really turn out to be manufacturable on Mars.  Right now,  I doubt anyone else really does,  either.  So,  I picked four combinations that might be “typical” of things others are considering.  The exact species were simply what was available in some old references on my bookshelf.  The four,  with rationale,  are:

Liquid oxygen – rocket-grade kerosene,  LOX-RP1,  well-developed technology

Nitrogen tetroxide – unsymmetrical dimethyl hydrazine,  NTO-UDMH,  well-developed, fully storable

Liquid oxygen – liquid methane,  LOX-CH4,  new,  easily-manufacturable?

Liquid oxygen – liquid hydrogen,  LOX-LH2,  highest-performing,  from ice?

Rocket Ballistics and Design

The data I had on-hand in my library allowed me to determine an estimate of rocket characteristic velocity (c*) as a power function of chamber pressure for all four selections.  I was able to scale the c* to an easily-achieved design goal of 500 psia (3.447 Mbar) for a medium-to-small rocket engine.  Expanding from that,  to 6 mbar backpressure,  gave me expansion ratio and ideal thrust coefficient,  read off a standard chart.  I corrected that by an appropriate nozzle efficiency reflecting realistic half-angles,  for single-engine uninstalled engine performance and gross geometry, derivable from a thrust specification. 

Once sized by a thrust requirement,  the throat diameter is known,  which with the expansion ratio and half angle sizes-out the bell dimensions.  I picked an arbitrary 10:1 area contraction ratio chamber-to-throat,  and sized chamber length off some empirical L* values.  Thus “reasonable” engine dimensions can be estimated,  once a thrust requirement is set.  The point is “getting into the ballpark”.

The Payload

I did my study around the concept of a fixed payload to be landed.  For an early mission,  this might typically be 3 men with a month’s supplies.  I included a guessed allowance for a rover car with a drill rig on it,  and allowances for surface exploration and experimentation gear,  plus a small inflatable “pup tent” in case they were too far from the lander to return.  I got 3.191 metric tons to cover this,  but that’s just a guess.  So also is 60+ cubic meters to contain it all,  plus some “living space” inside the lander. 

Flight Paths

These vehicles will have ballistic coefficients far larger than any of the lander vehicles so far delivered to Mars.  Investigations I have run came to the same conclusions as others:  such vehicles can aerobrake successfully,  but will come out of hypersonics at too low an altitude for chutes or ballutes to deploy,  much less do any good,  in the thin “air” of Mars.  Therefore,  hypersonic/supersonic retro-propulsion is going to be mandatory.  The vehicle simple rocket-brakes directly to touchdown,  once the entry hypersonics are over,  or at least mostly-over.  That kind of descent is illustrated in Figure 1 (below). 

These estimates are based upon a surface-grazing transfer ellipse from a 200 km altitude orbit (low Mars orbit,  LMO).  The de-orbit burn requires only a 50 m/s “delta-vee”,  something that attitude thrusters can provide.  LMO speed is 3.455 km/s at 200 km.  Speed at the 140 km interface is 3.646 km/s.  Speed at the local Mach 3 point varies some,  but falls in the 0.7 km/s range.  Allowing for heavy rocket-braking inefficiencies plus final near-hover for touchdown,  a good guess for the effective ideal “delta-vee” for descent is 0.9 km/s,  after getting some 2.9 km/s deceleration from aerobraking. 

The ascent is easier to rough-estimate.  The total velocity to be achieved is LMO speed,  with empirical “kitties” added to cover gravity losses and drag losses.  For rough-estimating clean rockets here on Earth,  5% of target velocity is adequate for each of both kitties.  The losses are less at Mars because of the weaker gravity and thinner “air”.  I simply ratioed-down the 5% figures by the surface ratios of gravity and density,  to a combined-loss “kitty” of 1.94%.  The ascent ideal “delta-vee” figure is then at most about 3.6 km/s,  as given in Figure 2 below. 

Those “delta-vees” add (for no reduction in payload upon ascent) to a total two-way “delta-vee” capability of about 4.5 km/sec.  For estimated installed specific impulse performances,  excepting hydrogen,  the resulting propellant fractions then fall near 75%. 

The Vehicles

Each propellant combination leads to a different vehicle size-out.   My concepts are based around a guessed constant 20% inert weight,  to cover both structural and landing equipment items,  but hopefully with enough robustness to provide a significant service life in re-use.  Once a propellant fraction has been determined,  it and the inert are deducted from one,  determining the payload fraction.  The known payload weight then sets the entire weight statement.  The key is thus determining the propellant fraction,  from performance parameters and the total “delta-vee” requirement. 

My vehicle layout concept is based on historical US capsule shapes:  blunt heat shield with a more-or-less conical afterbody.  Heat shield shape is spherical-segment,  with a radius of curvature equal to the diameter.  I used a conical afterbody in the 20-to-30 degree half-angle range,  with a cylindrical extension,  since cylindrical tanks and pressure vessels are easier to build than conical ones.  The flight control station is at the end of the cylindrical section,  for really good visibility. 

The structure would be a deck frame with attached heat shield,  and four extendible landing legs.  Height-to-stance width is near one,  for good stability.  In the center is a cylindrical sealed compartment containing 4 canted engines (I chose 10 degrees arbitrarily) to “enforce” retro plume stability during descent.  With a sealed engine compartment,  there can be no throughflow through wide-open ports in the heat shield for the engines.  That eliminates the need for port covers,  and for swapping ends during hypersonic flight. 

A small makeup-massflow/coolant-flow might be needed to balance the volume-filling transient as the vehicle descends,  in order to prevent intrusion of any entry plasma through the open ports.  That was not analyzed here.  But,  it should be noted that this method of preventing throughflow (with a sealed engine compartment) should work equally well,  whether the engines are firing,  or not!

I arbitrarily picked an installed thrust sizing such that the vehicle accelerates at two standard gees,  at its maximum ignition weight.  I also assumed 4 engines in the cluster.  Correcting for cant angle and four engines,  produces the individual engine size-out,  and thus the necessary engine compartment dimensions,  and all the performance parameters. 

The required quantity of propellants and their densities sizes the tankage volume to stack on top of the engine compartment.  The 45-degree cone,  on top of all of that stack,  is the crew flight station.  The conical segment around the engine compartment is the crew living space plus cargo volume plus compartmentalization for airlock purposes:  60+ cubic meters.  I assumed that some of the conical shell panels are hinged at the deck line,  so that they could also be used as unload ramps.  Only the LOX-LH2 vehicle sized-out such that the conical segment had to extend partway up the cylindrical propellant tankage section,  a consequence of its lower mass (but similar overall volume) relative to the other three.  That vehicle had the smallest engine compartment,  by far. 

Figure 3 below shows rough vehicle layout and dimensions,  all four being roughly the same overall shapes to within a fraction of a meter.  The weight statements are quite different,  as a function of propellant selection,  as shown in Figure 4.  Note than none of these vehicles could ride assembled to Earth orbit inside existing payload shrouds,  which are around 5 m diameter max. 

These things will have to be assembled on-orbit in low Earth orbit (LEO) from docked and assembled components,  and then sent to Mars.  I would suggest each lander push its at-Mars propellant supply to Mars as an unmanned cluster vehicle,  via min-energy trajectory.  The quantity of propellants each lander pushes to Mars then depends upon the scope of planned on-orbit-based operations.  Determining that is out of scope here.

Rough Performance Estimates

These vehicles could be operated as orbit ferries in either of two ways:  (1) on-orbit basing,  meaning refueling from supplies in LMO,  and (2) surface-basing,  meaning refueling from supplies manufactured on the surface.  In the first case,  entry is at very nearly maximum mass,  followed by ascent at reduced mass.  In the second case,  ascent at maximum mass is followed by entry at much-reduced mass.  These two cases very much affect the entry ballistics,  resulting in much-different final rocket-braking needs. 

On-orbit basing is the more stringent descent case.  My analyses indicated rocket-braking requirements in the 3-to-5 gee range,  if braking was delayed to the local Mach 3 point.  This violated considerably the two-gee engine sizing assumption.  Indicated “hover time” at 200 m was barely adequate at 55 sec to cover the touchdown.  This outcome simply indicates that rocket braking must start earlier,  closer to the entry max deceleration gees point,  which is at substantially-hypersonic speeds.  There is no reason this could not be done,  once the hypersonic/supersonic retro-propulsion approach is adopted at all.  I did not re-analyze this change in detail,  having already established the basic feasibility. 

With surface-basing,  the descent rocket braking requirements (waiting to local Mach 3) all fell within the two-gee intended design.  The 200 m hover times were at least twice those of the other case,  for all four propellant options. 

For both cases and all four propellant options,  entry peak deceleration gees fell in the 0.71-0.73 range.  It is the 2-gee rocket braking that is the highest gees the crew must withstand during descent.  Accelerations upon ascent were not analyzed,  but should be comparable to 2 gees,  assuming throttleable engines.  Clearly,  the crew should arrive at Mars fully physically-fit,  and able to endure a few minutes at 2 gees,  seated,  with full human functionality. 

This implies that their transit vehicle should provide artificial gravity (by spin) at 1 full (Earth-normal) gee.  That is an issue for manned transit vehicle design,  out of scope here,  but very definitely needing attention called to it.   

Employment of the Landing Boats

These craft could be employed flying multiple missions,  either on-orbit based,  or surface-based.  It is easy to imagine a first manned exploration mission to Mars,  where the landing boats are operated on-orbit,  using propellant supplies brought from Earth.  At the end of the mission,  the landers and any remaining propellants would be left in Mars orbit “for subsequent missions to use”. 

A subsequent mission with the objective of establishing a more-or-less permanent base or outpost,  might use them differently.  The initial landings,  to ferry vehicles and equipment down,  might use on-orbit propellant,  with subsequent flights using propellants manufactured on the surface.  Once that capability exists,  and we are interested in only one (or maybe two) specific outpost sites,  the landers would be operated in surface-based mode,  probably for the remainder of their service lives. 

A variation on that scenario would be to accomplish both objectives in one manned mission to Mars,  something feasible because the stay time at Mars until the orbits are “right” for the return is over a year long.  Initial on-orbit based explorations are done,  until a site is identified where large quantities of propellants can be manufactured fairly rapidly.  Then the vehicles and equipment are all transferred to that site,  for surface-based operation,  from what will become a more-or-less permanent outpost.  This presupposes that propellant can be manufactured in 20-ton lots on a timescale of a few weeks,  on that first manned trip.  If that is not true,  we are inevitably reduced to the first scenario. 

I would suggest that we plan for the two-objectives-in-one-mission scenario,  but with sufficient on-orbit propellants to fully support the fallback position. 

About the Heat Shield

The entry analyses yielded a worst-case peak stagnation heating rate of roughly 5.5 W/sq.cm.  This is low enough to allow the use of black-surfaced low-density ceramic heat shield materials on the windward surfaces,  even at the stagnation point.  White-surfaced low-density ceramics can be used on all lateral and leeside surfaces. 

However,  because these vehicles land on natural regolith surfaces,  there is the dead certainty of suffering dirt and stone impacts to the heat shield,  due to rocket back-blast effects at takeoff and touchdown.  Low density ceramics far less fragile (and far less labor-intensive to maintain) than the well-known Shuttle tile are thus demanded. 

There might be one:  my oddball experimental ceramic-ceramic composite appears to have the necessary toughness,  with no “show-stoppers” anticipated to completing its development.  This material is described in the posting dated 3-18-13 and titled “Low-Density Non-Ablative Ceramic Heat Shields”.  My recent well-received paper at the 16^th Annual International Mars Society convention (in Boulder,  CO) also covered this same proposed material.   

Making the Propellant Selection

That topic is beyond scope here.  I do not know what propellant combinations might actually prove practical to manufacture on Mars.  I doubt anyone else really knows yet,  although many might claim to know.  That answer needs to be found first,  so that the landers we design and send,  on that first manned mission,  are compatible with what we can actually produce there.  That way,  they can serve for quite a while.  Making that selection is the fundamental pacing item for picking a lander design approach,  and then making it a flight-ready vehicle. 

Conclusions

1.       The first priority is to decide “for sure” what propellant selection could “best” be manufactured on Mars.

2.       Next,  the “landing boat” design given herein,  matching that propellant selection,  should serve as the design start-point for an actual “landing boat” design.  This design should be built,  tested,  developed,  and readied for operational use. 

3.       The manned Mars mission,  or sequence of missions,  should employ this “landing boat” in the relevant role (or roles),  and these craft should be left there for future uses. 

4.       This “landing boat” should be designed with the maximum possible expected service lifetime,  far beyond the needs of one mission,  so that it may serve subsequent missions or roles with local refueling,  for as long as possible. 

5.       There is no reason this “landing boat” development could not be started right now,  and thoroughly tested in low Earth orbit,  just as the lunar lander was.  The pacing item is (again) deciding which propellant selection could actually be manufactured upon Mars. 

 

Figure 1 – Descent Trajectory Assumptions and Requirements

Figure 2 – Ascent Trajectory Assumptions and Requirements

Figure 3 – Rough Overall Vehicle Dimensions,  Without Internal Layout Details

Figure 4 – Weight Statements and Performance Parameters by Propellant Selection







Update 9-9-13:  Some Further Thoughts About Selecting the Right Propellants

I really don't think kerosene is something we could practically manufacture on Mars,  but it might be somewhat representative of a hydrocarbon heavier than methane,  that we might dream up a process for.  It is a very well-known technology.  But,  I'd bet we can find a way to ignite or keep-unfrozen any of these choices,  though. 

I really don't think NTO or any of the hydrazines might actually be practically manufactured on Mars,  without a source of fixed nitrogen.  That's a huge obstacle there,  as far as I know.  But,  we already know those propellants can be easily stored,  and we have had engines that re-light multiple times in vacuum with them,  for decades now.  That's pretty much the technology of the shuttle OMS maneuver engine pods. 

I suspect LOX-LH2 would actually be the "easiest" to manufacture on Mars,  using mined ice and electrolysis as the basis.  LOX is not too much trouble to liquefy and store;  LH2 is much trickier to do,  with the ortho vs para form problem perhaps now the easiest problem of several to resolve. 

On Mars,  the truly fundamental problem is "where is the ice deposit big enough to be worth mining?"  We now know for sure that Mars has lots of water still (in the scientific sense),  the trouble is that it's just not "everywhere".  The kind of ice lenses Phoenix found near the pole is not the kind of deposit that supports practical mining and manufacture.  What we need is a buried glacier 10+ meters thick and many,  many km in lateral extent. 

BTW,  it'll be subliming as we dig it out.  Every mine hole has to be regolith-buried when not in use.  There will be one whale of a lot of regolith-moving operations involved in this activity.  The machines will look like heavy mining and road-building equipment.  That takes a big lander,  even if shipped in small pieces and assembled on site.  These things will not be carried by a series of Apollo-like dinky-little landing modules.  No way.  We need real "landing boats" of very significant size. 

They're not gonna fit existing payload shrouds for launch to LEO for this mission.  Something else to think about. 

We have orbital observations of where some such buried glaciers might be (emphasis on "might"),  but we have absolutely no ground truth about it.  I have never seen a robot probe design capable of determining that kind of ground truth,  either.  So,  if we are going to plan on making LOX-LH2 to return,  where do we land? 

Tough question.  We have to be close enough to walk to the ice,  or it ain't gonna work.  We're talking front end loaders,  bulldozers,  and large pressure-vessel process machinery here,  with maybe even some pick-and-shovel work by more than 2 men.  Long range transport is simply out of the question,  that first time up with propellant manufacture. 

As for making methane out of water,  and the CO2 in the "air",  the low inlet densities make all your machinery (whatever it is) look very large and heavy and energy-intensive,  compared to what we are used to here at home,  by about a factor of 14.  My guess is you can make 1's,  not 100's,  of kg per day.  You'll not accumulate enough to return a crew (tens of tons),  not even in a year's stay,  even if it doesn't break down or encounter unexpected problems. 

And you will encounter unexpected problems (lots of emphasis on "will").  Done robotically before the men arrive offers a potential way out,  except that robots-as-we-know-them-today are simply inept at solving unexpected problems.  Put the men there to solve those problems,  and you are right back to the inability to accumulate tons of propellants in time.  Plus,  with LOX-CH4 you still have to solve basically the same water problem as LOX-LH2,  to get the oxygen and the hydrogen. 

So I dunno which one to try.  And I don't yet see much of a path to resolving this in time for a mission in the 2030's,  much less the 2020's we'd all like to see.  NASA has no plans to send the right kind of probes that could locate the propellant-making resources.  I don't see anybody else sending the right kind of probes,  either.

That puts me back to the costly-but-sure-thing concept:  first mission relies on propellants-sent-from-Earth.  Which means it is an LMO-based mission,  sending down multiple ferries to multiple interesting sites,  and emplacing the machinery to experiment with propellant manufacture at the most promising ones after the men return home.  Leave the ferries in a higher Mars orbit,  with whatever propellant is left over,  for the next mission to use.  What's the point of going all that way with men,  and only making one landing?  That's really dumb!

Meanwhile,  we have to guess which propellants might actually be made on Mars most practically,  and build the first-mission ferries to use that.  That way subsequent missions (including planted bases) can refuel and re-use the same ferries.  Right now,  I'd guess LOX-LH2 from ice.  But with an engine compartment big enough to accommodate being refitted with different engines.  And with compartmentalized tankage to accommodate being re-plumbed for different propellants.  That's heavier,  and so is structural robustness necessary for long-term reusability.  My assumptions of inert structural weight 20% are quite likely too low. 

 Update 9-11-13:

I have had some conversations with John Strickland about his designs versus mine.  Coming from very different starting points,  his results for lander vehicles and mine are amazingly similar.  This study of mine is a lot more realistic than should ever be expected of a bounding calculation.  (And that's what it is,  so don't read too much into all the nitty gritty little details).   

Update 9-22-13:

I'm beginning to think,  since all 4 of my vehicle rough-outs turned out to be about the same overall physical size,  that a "good" design might be one based around LOX-LH2,  but with enough space internally to be reconfigured and re-engined for LOX-CH4.  Comments?  Ideas?  Please weigh-in!

My best guess is that this LOX-CH4 combination is the "most likely" in-situ-produced propellant combination,  long term.  Short term,  I really think it might be LOX-LH2 from mined ice,  with nothing but electrolysis and liquefaction.  That is the simplest and most direct combination we have.  It is only restricted by where significant ice is actually buried on Mars. 

Basically,  vehicle size is fundamentally "set" by the payload mass to be landed.  My concepts are for "well-empowered" explorers,  not permanent base builders.  Please weigh-in with payload mass ideas for the follow-on base-building missions.  I know a lot less about that. 

The real question to answer here is:  what are we really going to do with men on Mars?  Explore?  Build bases?  Both sequentially?  Both at once?  The answer makes a huge difference to the mission approach,  architecture,  and component designs.  Overwhelming,  actually.