The following is an update of the paper I presented at the Mars Society convention, Dallas, Texas, 2011. I replaced the baseline design in the original paper with a solid-core nuclear NERVA propulsion technology, done as a minimum-energy transfer Earth to Mars, and the same to return.
That kind of technology resurrection is more consistent with what might be done in the next decade or less. The original baseline in the original paper was gas-core nuclear, very high-speed transit propulsion. This is actually more consistent with what might be done in the next 2-3 decades, given a crash program.
In both cases, the lander propulsion was solid-core NERVA nuclear. Design safety approaches are unchanged from the 2011 paper.
What NASA (or Any Other Governments) Might Really Do About Sending Men to Mars GWJ 4-13-12
I rather think that if we're lucky, there will be two government-funded "exploration" missions to Mars in the next 20 years. If we're really lucky, it'll be a coordinated and sequenced pair of missions from the same government, or consortium of governments.
I am not at all sure these government-funded missions will do any real exploration. Unfortunately, a very-likely product is another Apollo-style “flag-and-footprints” stunt.
If we're not so lucky, it'll be a duplicated first landing "exploration" by two different governments. That raises the probability further of nothing but “flag-and-footprints” stunts instead of real exploration.
If we're not very lucky at all, there'll only be one government mission. Almost certainly it would be nothing but a “flag-and-footprints” stunt.
Most likely, there won't be any government-funded manned missions to Mars at all. Not the way things have been going with NASA since Apollo was cancelled in mid-program back in 1972. We have never left orbit since then. No other government has yet done those things that we actually did so long ago.
What Is Exploration, Anyway?
That does bring up what real “exploration” is all about. Based on the most successful ventures 300-500 years ago, governments funded voyages to find out what there might be of value in the Americas, Indonesia, and Australia, and further, where exactly those resources were located. That set up the possibility of eventually planting colonies to use those resources (although some were successful and some were not). The least successful voyages were just flag-planting land claims only.
Real exploration answers two deceptively-simply questions: “what all is there?”, and “where exactly is it?” The lesson of history is that if you do not answer those questions, there is little point in going. Even with today’s technology, robot probes cannot adequately answer these questions properly, because none have ever drilled deep, and none could effectively handle anything not anticipated.
Based on that historical definition of “exploration”, we did not effectively explore the moon during the Apollo landings. Those were just “flag-and-footprints” with a tow sack or two of surface rocks. Most of what we really know today about potential resources on the moon comes from other probes since.
Mars is a much tougher problem to visit than the moon. It is a very long way there, and it’s a large and extremely-varied world. “One trip-one landing” doesn’t make much sense in view of the difficulty and expense just to make the trip. What if you pick the wrong site? How would you know? The information from the robot probes is good, but it is inherently limited: they report only what they were designed to perceive, not necessarily what is really there.
For Mars, I am recommending two “exploration” trips to answer the two fundamental questions properly. The first trip should visit multiple sites deemed appropriate by all the prior probes, which means it should have multiple landers. Prototypes of in-situ resource utilization (ISRU) equipment should be evaluated at these sites, as well as intense sampling and deep-drilling activities. This is probably best based from orbit around Mars, with the final site selections made from orbit.
The second trip should concentrate on the most promising one or two sites investigated during the first mission. Surface stays during the second trip are much longer, although the transit vehicle still stays in orbit. In addition to even-more-intense sampling and drilling, much more refined versions of the ISRU equipment (based on the first mission results) also get very thoroughly evaluated. A lot of practical information regarding “how to live off the land” gets acquired on this second trip.
The logical next step would be a permanent base of some kind, at the best site, and designed to be as self-supporting as possible. Establishing a base requires transport of much larger masses to Mars, and also down to its surface. That is most likely a set of larger modular transit vehicles, and larger landers. But we won’t really know what we will need, until the exploration gets done in the first 2 missions!
Success with a largely self-supporting base would provide the information and experience to actually establish a self-supporting colony, with a reasonable and realistic expectation that it might thrive. You simply cannot do that job on the very first mission. The required knowledge and experience are completely missing.
Proper Roles of Government and Private Business
Generally speaking from 500 years' worth of history, a real government-funded "exploration" mission was the enabling prerequisite for a privately-funded trip (or trips) that established a base or colony. Good examples would be the Dutch and British East India Companies, and the British colonies planted in eastern North America after Jamestown. Spain planted colonies in Mexico and South America, but most of them did not prosper as well as the Dutch and British efforts. The closest things to viable French colonies were Quebec and New Orleans.
The point is, for the last 500 years, private business ventures did not typically invest in the initial exploration, but they did invest in trying to set up profitable colonies based on resources found by the government-funded voyages. This illustrates just how critically important it was to know what resources existed, and where they were located, before planting that base. You have to know exactly what to bring and exactly where to site your base! Fundamentally, it will be no different voyaging in space.
The business profile exception today is Elon Musk and his Spacex. He seems genuinely interested in actually exploring, as well as colonizing. But, Musk lacks a manned landing vehicle suitable for Mars. His Falcon-9 and Falcon-Heavy rockets, and his manned Dragon capsule, plus an inflatable “spacehab” module (perhaps from Bigelow Aerospace), is pretty much all we really need to get men to Mars orbit in a minimum-design mission. But, actually to land on the surface? That is a difficult proposition, for which no viable hardware or well-defined plans yet exist.
I consider it possible, but rather unlikely, that Musk might actually beat NASA to put men on Mars for exploration purposes. I consider it unlikely precisely because of the lander problem. One-way suicide missions are not the answer, although a descent stage with a Spacex Dragon capsule might just be able to pull one of those one-way suicide trips off.
The Transit Problem
With chemical propulsion, and the Spacex rockets I named just above (plus any others at 25 tons payload capacity or more), it is fairly easy to stack up in Earth orbit enough propellant modules, some engines, a spacehab module (or set of modules), two or three crew return Dragon capsules, and maybe a lander or two. This vehicle makes a near-minimum energy trip at 7 to 9 months one way, and the same time to return, when the orbits are “right”, after a long stay at Mars.
It’s roughly 3 years for the entire trip. You end up throwing away all the propellant modules and perhaps all but the crew return capsules. But you don't need a gigantic launch rocket to do this! In point of fact, a rocket big enough to launch all of that in one single shot simply cannot be built with 21st century technology. We simply cannot use the Apollo one launch-one trip model.
But, NASA is now beginning the development of a new gigantic launch rocket, one about the same size as the old Saturn-5 moon rocket. That gigantic rocket development (the new “SLS” mandated by Congress to fund jobs that replace lost shuttle work in certain Congressional districts) is now why NASA is starved for money everywhere else. (This is not the first time that Congressional dictates have completely derailed anything that made any sense at NASA, but that’s a different topic.)
Technologically, a gigantic launch rocket is just not necessary anymore. What we did in Apollo with one launch-one trip is no longer the correct model. Once you're at or above 25-ton shuttle payload sizes, then effectively, only cost per unit payload mass delivered to Earth orbit matters anymore. 25 ton payload size is exactly how we built the ISS. We're there already, with Falcon-Heavy at 53 tons, and with the Atlas-V - 551 and -552 configurations at 25 tons. The Mars transit vehicle built with them is modular, assembled by docking in Earth orbit, just like the ISS.
With solid core nuclear rocket propulsion, you can use exactly the same modular slowboat transit vehicle design, except that (1) it’ll be significantly smaller as-assembled, (2) you can keep most or perhaps all of it, and (3) most importantly you can use it again for the second mission. If you really could use it again, then why not do so, as expensive as this stuff is? Why launch all that stuff twice for the two exploration missions, and then throw it all away both times? That's utterly stupid.
Here’s exactly how we could do it: resurrect the old NERVA solid core nuclear rocket, that did everything but actually fly, before it was abruptly cancelled back in 1973. Resurrecting NERVA for the Mars missions is way-to-hell-and-gone far more important than developing another gigantic moon rocket from old shuttle technology. Mars transit propulsion is what NERVA was originally for. NASA has the wrong “flagship” expensive rocket project!
If with nuclear rocketry you manage to recover most or all of your transit vehicle, and you use it and recover it on both missions, then you still have an asset left over in Earth orbit that could be used yet again. That might be for planting the permanent bases that could become a nascent colony. Again, why launch everything from Earth, again and again, if you do not have to? That would be stupid, stupid, stupid, stupid……..
Crew Survival Issues: Radiation, Microgravity Disease, and “A Way Out” for Transit
Radiation: there are two distinctly-different types, cosmic rays and solar flares. Cosmic rays are a very slow drizzle of extremely-high energy particles, against which we have no effective shielding. The cosmic ray background dose varies with the sunspot cycle between 22 and 60 REM/year, while our astronauts are currently allowed up to 50 REM/year, with an additional accumulated career limitation. 50 vs 60 REM is close enough that, even at worst, I'd say we can get them there and back without killing them. But, one 2- or 3-year round trip is pretty much a career limit. No second trips.
Solar flare radiation is a flood of much-lower energy particles ejected irregularly from the sun. Against this danger we do have effective shielding, although we have never before flown with it. Shielding just requires some water and wastewater tanks around the crew as a suitable storm shelter. Myself, I'd make it the vehicle's flight deck, so that critical maneuvers can be flown, no matter what the sun does.
Microgravity disease is a host of skeletal and circulatory ailments that become ever-more permanent (and potentially lethal) in their effect, beyond roughly a year in zero-gee weightlessness. We do not know what level of partial gravity might be therapeutic, because those experiments have never been done in Earth orbit with a spinning habitat or station. It is unethical to risk lives on bed-rest as a surrogate for partial gee, precisely because it is only a surrogate, not actual experience.
So, at this time, we have to assume that one full gee is required for voyages lasting more than one year, which limitation a slowboat round-trip Mars mission certainly exceeds. But, there is actually an easy way to provide one full gee of artificial gravity with a vehicle assembled from modules docked in Earth orbit. It does not have to be gigantic or complicated.
Just build your vehicle as a long stack, put the habitat on one end and the engines on the other. Spin it end-over-end at no faster than the fuzzy 4 rpm limit we already know. 4 rpm at 56m "radius" is one full gee. The transit ship should stack up way longer than 112 m long, anyway. This also solves a world of problems with weightless cooking, bathing, going to the toilet, etc. Just de-spin for maneuvers or docking operations, then re-spin for coasting cruise.
Consider sending the landers and landing supplies, and all the landing propellants, as separate vehicles also assembled out of modules in Earth orbit. Just send it (them) one-way. Have it all waiting in Mars orbit for the crew to use when they get there.
But, we must send the men with enough propellant to get them home, in case rendezvous fails in Mars orbit! This is a “suspenders-and-belt” crew survival issue: providing them “a way out”. Dead crews are more expensive than anything else at all. Just ask NASA.
The Lander Issue
Myself, I'd solve the lander problem before I designed anything else. Those are essentially dead-head payload until used, and they’re not small items. The more landings you make, the more landers you send, the more you have to launch from Earth, the more it all costs, and the less likely it'll ever get done.
But, exploration is not "flag-and-footprints". What we did in Apollo with one trip-one landing was not really exploration at all. Don’t think like that. Too many still do. You have to think multiple landings in the one trip.
Mars is tough to land on. There’s too little air to assist properly in decelerating to touchdown, and too much to ignore for simple rocket braking and rocket ascent, as we did on the moon. There’s too much gravity to do it in a two-stage lander with chemical propulsion: about twice that on the moon. Deliverable payload for the lander size starts looking very ridiculously small very quickly, as you consider 3, even 4 stages with the rocket equation.
Deliverable payload to the surface is more-or-less fixed by mission requirements. The resulting lander is very large, or else you send a great number of them to exactly the same site, or both. That’s a lot of very expensive weight to send all the way to Mars, and then just throw away!
All of these problems have given the nations of Earth a rather poor track record landing robot probes on Mars, until lately. And there has been no sample return yet. It’s a very difficult design proposition to go one-way, much less two-way.
If there were a feasible single-stage re-usable lander, then you don't have to send very many of them to Mars, yet you could still make a lot of landings all over Mars on that first mission, refueling them from propellant supplies left in orbit. Chemical simply can't do that job. NERVA could. And NERVA could also push all the landing stuff to Mars, too.
There’s also the out-of-plane orbit situation to consider, as well. The inclined orbits we use here on Earth could be compatible with transit to Mars, if departure were timed correctly. That puts the transit vehicle(s) in an inclined orbit around Mars, probably near 20-25 degrees to its equator. The best landing sites could be up to 30 (or more!) degrees out of that orbital plane, which drastically increases the requirement on lander propulsion. This makes chemical propulsion look even more infeasible.
But, with a resurrected NERVA solid core nuclear engine capability, there is enough performance available to support a nice payload fraction (10%) with a very robust structural fraction (20%) at enough propellant fraction (70%) to support a two-way full-load trip at up to 30 degrees out of plane. That’s with no aerobraking assist during descent. You just “gas it up and go again, and again, and again…” using propellants left in orbit to refuel. The potential of NERVA is only just slightly beyond what was already demonstrated 4 decades ago: Isp about 1000 sec, and engine thrust/weight about 4.
“A Way Out” for Crews Going Down in the Landers
It’s just plain common sense: on the first mission, and probably the second, don’t send everybody down at once, and, always have one working lander “in reserve” for a rescue mission. The crew staying in orbit can do science from orbit as well as monitoring the crew on the surface. If rescue is required, send down the reserve lander with one aboard, to rescue those trapped on the surface.
Employing a rule like that means that if there is no reserve lander available, the mission ends. Probabilities of vehicle failure being what they are, the common sense approach is to send a minimum of three landers, so that the loss of one does not terminate the mission. But, the loss of two establishes a trend, so that mission termination then becomes wise.
The other thing to consider is a multi-engine lander, arranged such that the loss of one engine does not end its viability completely. This also allows slightly-canted thrust on descent, which should address any plume stability problems associated with rocket braking during hypersonic entry conditions. The vehicle itself should be rather short and squat, for stability on landing. So, that’s a short, fat conical design, built tough as an old boot.
Equipment to Take to the Landing Sites
The first thing to consider is a suitable habitat while on the surface, which needs to be pressurized and rigged for cooking, bathing, and all the rest of the life support issues. This thing could be essentially an inflatable: a gas balloon with an airlock, and a packed life support equipment pallet inside. It does need to be anchored against windstorms. Especially if the lander is nuclear, this thing needs to be remotely sited from the lander by around a kilometer.
The second thing is a rover car with a real drill rig mounted on it. You drive where you want, and drill to depths up to a kilometer or so, as needed, to answer the two exploration questions.
The third thing is robot assistants, somewhat similar to the small rovers we sent to Mars recently. Between the men and the robots on site, there is a lot of real exploration that could be done that way. Plus, the signal transmission time-delay is essentially zero. And, a robot in trouble can be rescued by the men in the rover.
By all means take some experimental ISRU gear on the first mission and try it out. It's just not smart to count on it for crew survival on that first mission, because the probability is, it won't quite work right, maybe not at all. That's just the nature of engineering development.
If the first mission really does get done right (and I think that is a low probability, given NASA's track record since 1972), then the second mission could be based on the surface at one, or at most two, sites. This would be instead of the strictly-orbital basing of the first mission, although the return transit vehicle still gets left in orbit.
Some better ISRU machinery prototypes could really get "wrung out" on a second mission like that, but it's still just plain stupid to count on them for crew survival. My experience with engineering development (some 19 straight years in cutting-edge defense work) is that "second time up" still does not work well enough to serve. It really doesn't matter what you are attempting, that's just pretty much a "given" in the real world.
And that's why I'd like to see two properly-sequenced government missions before a corporate visionary takes over, like Musk. (Boeing and Lockheed-Martin, which together are ULA, sure as hell won't offer to take over.) By that time, Musk would have both the lander and the ISRU technology from the government, which would enable him to really succeed at planting a proper base, one that might actually become a nascent colony.
Do it wrong or out of sequence, and that viable colony just plain will not happen in the next century, at least. It'll take that base/colony a significant while (measured in years to decades) to become self-supporting. That's been the history of things for 500 years. That's also why ULA won't plant it: there is no short/near-term profit in doing such a thing, unless the government pays them to do it. And I can pretty much guarantee you that it won't.
NASA is doing the wrong thing developing another Saturn-5-class rocket, when all that is needed is 25+ tons at $2500/lb payload, or less. Falcon 9 is $2500/lb at 10 tons, Atlas-V 551/552 is $2400/lb at 25 tons, and Falcon-Heavy is projected at $800-1000/lb at 53 tons.
To sharply reduce the total tonnage assembled in Earth orbit and sent to Mars, the NERVA solid-core nuclear rocket should be resurrected as soon as is at all possible. This is far more important than a new gigantic launch rocket. This nuclear option supports both smaller orbit-to-orbit transit vehicles, and also single-stage re-usable landers.
Ideally, there should be a sequence of two manned exploration missions to Mars; the first visiting many sites from orbit, the second concentrating upon the best one or two of those sites. These two missions are the experiential prerequisites to any establishment of bases as nascent colonies. They also act as trials of prototype ISRU equipment, something any practical base or colony will require.
Everything about manned missions to Mars should be done in such a way that “a way out” is designed into each and every mission phase. The highest cost of all is a dead crew.
Everything about manned missions to Mars should also be done in such a way that the effects of radiation, weightlessness, and other life support issues are addressed by means of well-known direct-experience data, not assumed-to-be-applicable surrogates. Anything else is unethical.
Such a set of two exploration missions could start within the next 5 years, assuming that NERVA is resurrected in time to support it.