(see also 1-11-2011 post for an update to this study)
A manned mission to Mars was investigated for feasibility. The objective was significant exploration, not a single Apollo—style “stunt” landing. It seemed insane to go to all the trouble and expense of sending men to Mars, and not visit several different sites. It also seemed insane to launch so much equipment and not try to reuse it. This consideration eliminates very small vehicle designs.
Earth orbit rendezvous and Mars orbit rendezvous were combined into a single mission design to save weight. Unmanned assets were sent by Hohmann transfer ahead of a fast-trip manned vehicle to save weight. Self-rescue or escape capability was designed-into every mission phase as much as possible. The landers and the manned vehicle were designed as reusable single-stage items.
The unmanned assets were sent single-stage one-way to Mars orbit using the lander propulsion to save weight. Lander and propellant tank assets were left in Mars orbit for refueling and re-use by subsequent missions. The manned vehicle was recovered in Earth orbit for refueling and reuse.
The manned mission time was under 1 year, eliminating the need for voluminous and heavy artificial gravity by spin (because the only requirements currently understood are 1 gee at 4 rpm). The flight deck in the habitat module was assumed to be radiation-shielded against solar flares by water and wastewater tanks, plus a little steel plate. Crew size was 6. Every component was small enough to be launched by a Falcon-9-heavy booster, or smaller. But, the interior volume is comparable to the old Skylab station, thus promising effective alleviation of long-confinement psychological issues.
All assumed propulsion was nuclear thermal rocket (NTR). The lander engines were assumed to be slight updates to the old NERVA solid core technology last tested in 1973. The manned vehicle engine was assumed to be a radiator-cooled gas core NTR, approximating a design that came within about 2 years of first article test in 1972, when all such work was stopped. This old design’s Isp was 6000 sec.
Only launch costs were estimated, as near $8 billion in today’s dollars for 16 landings during the one mission. A trade study evaluated cost reductions available for reducing landings per mission. Programmatic costs, and technology development and hardware production costs, were not estimated. It is thought that getting the gas core NTR technology ready would be the pacing schedule item for such a project. These estimates are only rough hand-calculations done by pencil and paper. They are just good enough to demonstrate feasibility, and to serve as a startpoint for more sophisticated analyses.
Basic Mission Design Approaches, Constraints, And Assumptions:
Fig. 1 (below) shows a list of the fundamental considerations. Top-of-the-list is to design-into every aspect a “way out”, meaning escape or self rescue, if in any way possible. This is the very essence of “man-rating”.
The next two items relate to flying the manned portion fast, in order to cut travel times under a year and allow deletion of the need for artificial gravity, in accordance with experience obtained on the International Space Station (ISS). Mission times exceeding a year require artificial gravity, because we have nothing but indirect (surrogate) data such as bed-rest experiments to suggest otherwise. It is unethical to risk the lives and health of astronauts on indirect data, even if they are willing. Artificial gravity by spin has to be designed for one full gee at no more than about 4 rpm, based on what we currently know by direct experiment. This leads to large, heavy, and costly vehicle designs. Choosing instead more advanced propulsion for the fast trip is very likely the easier, surer, course.
The fourth item says to do both Earth orbit rendezvous and Mars orbit rendezvous in order to save weight, plus obtain further weight savings by sending unmanned vehicles on min energy trajectories. The lander vehicles and the propellant supply supporting their operation can be sent ahead of the manned vehicle, which would then rendezvous with these supplies at Mars. For crew safety, the propellant supply for the return voyage cannot be sent this way, even though it would save considerable weight to do so. This is because the crew would be stranded, and would die, if rendezvous at Mars failed for any reason. If the manned vehicle has sufficient propellant on board to return, this outcome is avoided.
It makes little sense to go to the trouble of sending men to Mars, and not do some serious exploration. This mission study is based on doing 16 separate landings at widely-separated sites, of up to a week each, while the manned vehicle is there. This increases enormously the information return from the mission, a sort of “shotgun-pattern” planetary survey. It might even be possible to begin planting “prospecting” bases on the next mission, instead of further initial exploration.
This multiple-landing plan is subject to some safety constraints, as the last item indicates. The concept explored in this study is that 3 of the crew visit the surface, monitored from orbit by the other 3 crew. There must be at least one ready lander in orbit, in order to perform a rescue landing, if need be. This rule would terminate the mission upon a lander becoming unserviceable, unless at least three landers are sent to Mars. To minimize crew risks, any rescue landing is piloted by a single crewperson.
To make such a plan work, the landers have to be reusable, so that only three need be sent. To keep from sending lower stages, the landers must be single stage. That means they must be nuclear. As shown in fig. 2, the idea is to separate the lander propellant supply into three parts corresponding to the three landers, and send these to Mars using the lander engines themselves. These assets would remain in orbit at Mars to be refueled and reused by subsequent missions.
Fig. 1 -- Basic Mission Criteria
Fig. 2 – Sending Landers and Lander Propellant Unmanned One-Way to Mars
Fig.3 shows the thinking behind the design of the fast-trip manned vehicle. There are serious considerations for radiation sheltering during solar flare events. In the habitat module, there must be a space surrounded by water and wastewater tanks, and a little steel plate, in which the crew of 6 could shelter during a radiation storm. Prudence dictates that this be the vehicle’s command deck as well, so that critical mission maneuvers can be conducted, storm notwithstanding. The mission should take no more than about 9 months. The vehicle should be stocked with over a year’s supplies, “just in case”.
One of the requirements often soft-pedaled or ignored is volume of space available per crewmember. The psychological impact of prolonged confinement in tight spaces is a very real danger, one that can be confirmed by any prisoner who spent time in solitary. Most crew habitat designs I have seen provide about the same space as was in the Apollo capsule, which is about like a modest bedroom closet per man. That is simply not enough. The space should be comparable to that available to a family of 4 in a small (1200 square foot) house, ideally. The old Skylab space station, at 90 tons, comes pretty close to the size of habitat that is needed. It provides the baseline for this study.
Fig. 3 also shows two main engines, needed for redundancy, and a round trip propellant supply, needed in case rendezvous fails at Mars. There are two crew return capsules, each large enough to carry all 6 crew, but twinned for redundancy. These need to carried along, in case maneuver propulsion fails on the return voyage. In that event, a free return must be attempted in the capsules as the vehicle flies by Earth. These capsules must have enough delta-vee capability to “hit” an acceptable reentry corridor. That means they probably need a small propulsive service module or supply.
Fig. 3 – Safety and Design Considerations for the Manned Vehicle
Fig. 4 shows some practical launch vehicle constraints for assembly of the Mars mission in Earth orbit. Many different items could be selected, these simply correspond to a family of very cost-effective launch rockets. Data were taken directly from the Spacex website, as it exists at the time of this writing. The biggest impact is the size of objects to be launched. Payload mass is more important than fitting within the “factory stock” payload shroud. A lot of the Mars mission components could in fact ride “naked” on top of the launch rocket.
Fig. 4 – Launch Rocket Constraints on Mars Mission Component Designs
Rough Mission Delta-Vee Estimates:
Two scenarios needed investigation: a basic min-energy Hohmann transfer ellipse for the unmanned vehicles, and an “almost straight-line shot” fast-trip high-energy trajectory for the manned vehicle. Of these, determining a realistic delta-vee requirement for the fast trip is actually easier. One simply divides a representative straight-line path length (in this case about 100 million kilometers (km) by a tolerable trip time (for this analysis about 75 days). Assuming impulsive delta-vee events at each end of the trip, one obtains a nearly square-wave velocity trace vs range, because at these speeds, the sun’s gravitational deceleration or acceleration effects are negligible, as is path curvature. See fig. 5.
Fig. 5 – Fast-Trip Scenario Approximation
The average velocity over this trace is very nearly the delta-vee value required to start it, and also to end it. Thus, twice the average velocity is pretty close to the one-way delta-vee requirement for the trip. For a two-way trip, one doubles this again, to about 4 times the average velocity. For these numbers, the two-way fast trip delta-vee requirement is a very demanding 61.72 km/sec. For practical single-stage vehicles, this corresponds to specific impulse (Isp) requirements closer to 6000-7000 sec than the 900-1000 sec of a NERVA-type solid core NTR. Hence the selection of gas-core NTR technology with a waste heat radiator to effect engine cooling.
For the unmanned vehicles making a one-way trip by Hohmann ellipse transfer, the estimate is made by classical orbital mechanics methods. Calculations were made for the average orbital velocities of Earth and Mars around the sun. An ellipse was fitted between the average distances of Earth and Mars from the sun, and its perihelion and apohelion velocities calculated. Escape velocities were calculated for Earth and Mars, and circular orbit velocities calculated for a low Earth orbit (LEO) altitude of 300 km, and for a low Mars orbit (LMO) altitude of 200 km. At Earth, the delta-vee to escape on a trajectory to Mars was estimated as the difference between escape and circular velocities, added to the difference between transfer perihelion and Earth orbital velocities. The delta-vee to capture at Mars was calculated as the difference between apohelion and Mars orbital velocities, added to the difference between Mars escape and circular orbit velocities. This is probably over-conservative. See fig. 6.
Fig. 6 – Rough Estimate for Hohmann Delta-Vee Requirements
The planes of the transfer ellipse and the straight-line “shot” are more or less in the plane of the ecliptic. Thus the plane of LMO achieved this way will be inclined relative to Mars’s equator, by around 25 degrees. The landers must have some amount of plane change capability, in addition to the delta-vee necessary to land without aerobraking (in the extremely thin “air”). The surface circular orbit velocity is larger than that at 200 km, and makes a good rough estimate of minimum delta-vee. Factoring this value up by about 1.10, accounts roughly for gravity and drag losses. The plane change requirement is figured from an isosceles triangle on the surface circular orbit velocity. These velocity increments are summed for the one-way delta-vee requirement, and doubled for the two-way trip. A maximum plane change requirement of 30 degrees was assumed arbitrarily. By judicious choice from an inclined orbit, this capability brings the majority of Mars’s surface within reach of the landers. See fig. 7 below.
This 11.52 km/s value is the maximum. Not all sites require a 30 degree plane change. The minimum is no plane change at all, for the much smaller two-way delta-vee requirement of about 7.84 km/sec. Such missions have a substantially-smaller propellant “burn”. Such propellant savings, plus the very capable nuclear engine, would very likely make an orbital mission to Phobos possible during this same exploration mission, using the same equipment.
Fig. 7 – Rough-Estimate of Lander Delta-Vee Requirement
The basic layout of the lander is propellant tank-as-airframe, topped by some sort of command cabin big enough for 3, and equipped with long landing legs disposed around the nuclear rocket engine. The width of the footprint should be comparable to the overall length of the vehicle for stability, so this vehicle is rather “squat” in its proportions. There should be some sort of deployable crane arm to provide a hoist to the surface. There should also be some sort of deployable solar panels to augment fuel cell electrical power. Given a heavy solid core engine and extensive landing leg structures, an inert fraction of 20% seems reasonable to assume. Combined with a propellant fraction of 70% and a payload fraction of 10%, the mass ratio is compatible with the solid core NTR Isp of 1000 sec, and the max plane change round-trip delta-vee requirement of 11.52 km/sec.
The payload comprises the crew of 3 with suits, 2 weeks of air, food, water, and fuel cell reactants (conservative for a maximum 1 week mission in case of trouble), a 3-man rover car, an inflatable Quonset hut with camping and cooking gear, and half a metric ton of scientific equipment, to include a small drill rig. Assuming the command cabin structure itself to be part of the payload, I rough-estimated 6 metric tons for payload. Thus the whole lander fully-fueled is 60 tons, with a propellant weight of 42 tons, and a dry-tank weight of 18 tons. I assumed half a ton of waste was left behind at takeoff, in making propellant usage calculations. See fig. 8. Note that an empty lander is within the payload weight to LEO of a Spacex Falcon-9 booster, although not within payload shroud dimensions.
Fig. 8 – Rough-Out Lander Design
This lander design with a 180-200 KN thrust solid core NTR engine does not have quite enough thrust to leave the surface of Mars fully fueled, but can easily take off partly-fueled, after landing from orbit. A slightly higher thrust specification would make fully fueled surface takeoff possible, but at the cost of a slightly heavier and larger engine. This is not really a necessary requirement for explorations conducted from LMO. Performance is compared to requirements in fig. 9 below.
Roughing out this vehicle is a supremely important prerequisite for the rest of the mission and vehicles because it is a major payload item, as is the crew habitat module. It should be noted that it is specifically the choice of nuclear propulsion that makes a single-stage lander possible. The delta-vee requirements for a single stage lander are simply out of the practical range of mass ratios for chemical propulsion. Without a reusable single stage lander, the mission exploration return is very much diminished: we are more-or-less back to a very few Apollo-style “stunt” landings. With chemical propulsion, a staged lander may only be used once (although an upper stage might be reused with a new lower stage). The number of landing sites is then no more than the number of landers carried to Mars, and this has a far greater effect on vehicle weights, mission complexity, and costs.
For the design and mission selected here, the average lander mission consumes some 37.56 metric tons of liquid hydrogen (LH2) nuclear rocket propellant. The plan for 16 such landings during the course of the mission then requires the delivery of some 600.89 tons of propellant as payload to Mars to support lander operations. That delivery requires even more propellant for the Hohmann transfer, even with using the lander propulsion as the propulsion sending these vehicles to Mars in order to save weight, launch costs, and complexity. See the unmanned vehicle rough sizing below.
For safety and self-rescue purposes, lander operations are envisioned as a series of sequential single landings, each with a crew of 3, while the other 3 stay in orbit to monitor progress and provide rescue capability with another ready lander. Thus at least 2 landers are required, and unless there is a third, the mission ends if one is rendered inoperative for any reason. That is why this mission plan sends three landers to Mars. Rescue is envisioned as risking only one crew as pilot in the rescue lander.
Fig. 9 – Rough Lander Performance Estimates
Crew Habitat Rough-Out:
There are three fundamental crew survival issues that must be addressed for any mission involving months of travel beyond Earth’s Van Allen Belts. These are (1) radiation protection (solar flares and cosmic rays), (2) protection from medical deterioration due to microgravity, and (3) sufficient habitat volume to stave off the psychological effects of prolonged confinement.
This mission is nominally 9 months, and certainly under 1 year in total duration. The dose of accumulated cosmic radiation is minor. The probability of a solar flare event is quite high. Therefore, there must be a zone inside the habitat shielded by water and wastewater tanks, and a little steel plate, which can support 6 crew temporarily during the event. Safety demands that this shelter also be the ship’s command deck, so that critical mission maneuvers may be flown, radiation storm or not.
The under-1-year mission time is within the realm of experience we have with microgravity exposures on the International Space Station (ISS). Therefore, this design need not provide artificial gravity by spin. The ISS exercise regimens will be adequate. This is very important, because provision of artificial gravity greatly adds to the habitat and vehicle size, weight, and complexity. The design requirements for such artificial gravity are still poorly understood, since the direct experimental work has never been done. We have only imperfect, indirect evidence from surrogate studies, such as bed rest. It is unethical to subject a crew to serious life and health risks, based on no better evidence than that. Therefore, the best design criteria we have are to provide one full gee at no more than 4 rpm. Any slower trajectory pushing total mission time beyond 1 year must deal with this design issue.
The habitat volume per crew issue is something ignored in design studies such as “Transhab”. Most of these designs would confine a crew in a space per person not much bigger than a typical bedroom closet. This is very likely to be psychologically very unhealthy, as any prisoner who has served time in cramped solitary confinement can testify. A design volume per person more like that in a lower middle class home would be far preferable. Units this large would resemble the old Skylab station in dimensions, and would be very difficult to launch. But, such a habitat could be assembled from smaller modules. It would be a part of the payload of the manned vehicle, the crew return capsules being the other part.
This study’s design is comprised of three modules, each 32 metric tons, docked in LEO to form a 96 ton habitat, stocked with substantially more than a year’s supply of food, water, oxygen, and other supplies. Such a habitat is very close to the mass of the old Skylab, but has a longer, narrower form factor. One of these modules would contain the radiation-shielded command deck. Each of these modules is within the near-term projected payload capability to LEO of the Spacex Falcon-9-heavy launch vehicle. Again, payload shroud constraints may be violated. See fig. 10.
Fig. 10 – Three-Piece Assembled Habitat Module
Crew Return Capsules:
For safety purposes, it is critical that these be carried with the habitat on the entire round-trip mission. This is because the vehicle propulsion might fail on the return trip, leaving no way to slow for capture. In that event, a crew return capsule capable of making a free return reentry at speeds very significantly higher than Earth escape speed offers the only avenue of crew escape. Each capsule should be capable of carrying the entire crew of 6, and there should be two such craft for redundancy. Some amount of service module propulsion is required to effect a proper reentry angle for survival.
Such a capsule already exists in its initial form as the Spacex Dragon. Dragon has crew capacity up to 7, and a heat shield rated for free Mars return. It fits a Falcon-9 launcher, although fitments need to be changed to accommodate some extra propulsion. I simply guessed this add-on propulsion module at 2 metric tons each. This plus the empty Dragon should be in the vicinity of 22 tons. See fig. 11 below.
Fig. 11 – Modified “Dragon” as the Crew Return Capsule, Two Required
Common Propellant Tank Module:
This is a more sophisticated design item than it first appears. There is a need to store LH2 for months at a time in zero-gee conditions. That requires what amounts to a double-shell tank, essentially a Dewar, with protection from solar thermal radiation, and at least a little meteor protection. As stackable modules, these require substantial structural strength. There is some sort of cryo-cooler (or a suitable equivalent) equipment required, plus the solar power to run that. There is considerable interconnect piping to meld these modules into an integrated propellant supply, plus a kit of extra pipe lengths and fittings to make those interconnections. It will be a substantial design challenge to achieve this in a 10% inert weight budget. Loaded tank size is set by the projected Falcon-9-heavy deliverable LEO payload weight of 32 metric tons. See Fig. 12 below.
Unmanned Vehicle Rough Sizing:
A part of the payload for this vehicle is an empty lander (18 metric tons), whose engine is also the unmanned vehicle propulsion. Inert weights of 3.2 tons per tank module add to this payload to comprise the dry-tank “burnout” weight. Loaded weights of 32 tons per tank module add to this payload weight to comprise the departure “ignition” weight. Thus vehicle mass ratio and delta-vee capability is a function of the number of propellant modules in the vehicle stack. Enough untapped modules need to arrive at Mars to support the mission’s lander operations. One third of that requirement (rounded up to the next largest number of tanks) is carried by each of the three unmanned vehicles, each with a lander. Those untapped tank modules are the remainder of the vehicle payload. See fig. 13 for a pictorial, and fig. 14 for estimated vehicle performance on its one-way mission.
Remember, after the mission concludes, the landers and empty tank assets are left docked in LMO. Subsequent missions need only bring more propellants, and reuse the landers, up to the lander engine lifetimes. Empty tank assets could be cannibalized for other purposes in future missions.
One other note: the same basic vehicle design is suitable for a variety of inner solar system missions. One simply stacks up enough common tank modules to meet the mission velocity requirements.
Fig. 12 – Common Tank Module
Fig. 13 – Unmanned Vehicle Stack (One of Three)
Fig. 14 – Rough Estimates of Unmanned Vehicle Performance
Manned Vehicle Rough-Sizing:
The design approach for this vehicle is very similar to the unmanned vehicles, only the payload and propulsion is different. The payload comprises the three-piece habitat module, plus two crew return capsules. To the radiator assembly (30 tons) and twin gas core NTR engines (half ton each of two for redundancy), one adds 3.2 tons per tank module for the inert weight total. This plus payload is the dry-tank “burnout” weight. To this, one adds 28.8 tons of propellant per tank module, to arrive at the departure “ignition” weight. As with the unmanned vehicles, mass ratio and delta-vee is a function of the number of tank modules in the stack. See fig. 15 below.
In this particular design, all the propellant required for the two-way trip is included in the vehicle. It would save weight to send the return trip propellant as a Hohmann transfer unmanned package, but, this incurs the risk that the manned vehicle might not be able to rendezvous with the unmanned fleet. In that event, the crew would be stranded, and would die. Abort scenarios where capture is avoided at Mars for an immediate return home would also be impossible. From a safety standpoint, it is simply more prudent to fuel the vehicle to be able to return home independently of all the other mission components.
Fig. 15 – Manned Vehicle Design for the Fast Trip
The Missing Technologies: Solid and Gas Core Nuclear Rockets
The key element to the fast-trip design of the manned vehicle is, of course, its engine. This is presumed to be a gas core open-cycle version of the basic nuclear thermal rocket. Unlike the NERVA-type solid core engine design of the lander, the gas core machine was never tested as a rocket engine. However, it did undergo component bench tests verifying containment of the uranium relative to the hydrogen, at about 1000:1 hydrogen:uranium flow rate ratio. It also underwent bench tests verifying controlled gas phase nuclear fission. The gas core engine was about 2 years away from a first-article rocket test when the program was shut down in 1972. The mission plans at that time allowed about 15 years to test and perfect the design, before potentially employing it on a manned Mars mission then scheduled for 1987. Most of this history is forgotten today.
The open cycle gas core NTR was thought to be adequately cooled by regenerative cooling, up to power levels corresponding to Isp around 2000-2500 sec. Above that power level, regenerative cooling was known to be inadequate. This necessitated use of a high-temperature radiator to cool the engine, whose characteristics are still guesswork. It was also thought there was a power limit above which the engine would vaporize itself due to propellant transparency to all radiation, up around 10,000 sec Isp. The planned Mars engine for 1987 was an Isp = 6000 sec design, well under that poorly-understood upper limit. That same projected design is assumed for this mission study. See fig. 16.
Fig. 16 – The Radiator-Cooled Open-Cycle Gas Core Nuclear Thermal Rocket Engine
Assuming that the developed engine and radiator system have characteristics even close to what I used for this study, then the performance of the vehicle powered by it can be calculated with at least some confidence. The results of have 6000 sec of Isp available is astounding, as given in fig. 17. Comparing this plot to the performance plot for the unmanned vehicle (of crudely similar size), one can see the difference in the delta-vee levels achievable: several tens vs only several km/sec.
Fig. 17 – Rough-Estimated Performance of the Gas-Core Manned Vehicle
As a comparison, fig. 18 below illustrates the updated NERVA engine used in the lander. That technology was substantially mature, and this shows in the quoted data in the figure. It should be noted that the lander design as worked out uses one engine, not a redundant two or three. The problem is one of thrust against gravity, and scalability of the nuclear design. The size used herein is not all that far from the original NERVA. There is some question whether a much smaller engine could even be made to go critical and produce power. Engine-out under gravity means that the remaining engines must throttle-up thrust levels to compensate for the lost engine. It might actually be easier to simply redesign the basic engine to be more reliable. This is an issue needing investigation before any designs can be finalized.
Fig. 18 -- Solid Core Lander Engine Based on NERVA Technology
Mission Information Return vs Mission Cost:
The direct launch costs are “retail”, based on number of payloads times the cost for the appropriate launcher. The data were obtained from the Spacex website, including projected costs for the yet-untested Falcon-9-heavy vehicle, and the Falcon-9 vehicle currently in flight test. On this basis, the total launch cost to LEO for 3 unmanned and one manned vehicle is right at $8 billion. That “buys” 16 landings, each up to a week long, at 16 separate and widely-dispersed sites on Mars, up to 30 degrees worth of plane change from the ecliptic. It also buys hardware that can be used again on subsequent missions, and other missions in the inner solar system.
There are hardware development and production costs to be considered, and programmatic costs. The habitat modules, the common propellant tank module, the modified “Dragon” crew return capsules, and the lander, are all items needing a “normal” amount of development, in the aggregate perhaps totaling around a billion dollars. The updated NERVA engine for the lander would actually require very little development. On the other hand, the gas core engine would be a very serious development item. Taken together, those two engines might total around a billion or two dollars. That is a wild guess predicated upon these projects being done by lean, efficient organizations. For “business-as-usual” with large, inefficient organizations, one should probably double or triple those estimates. Thus, the lower bound “wild guess” is then about $11B for 16 landings on Mars, all in one trip. See fig. 19.
Fig. 19 – Summary of Mission Design Characteristics
Reduced Mission Scope?
It is entirely possible to lower costs by reducing the number of landings under the same mission safety rules. The minimum is three. Only the three unmanned vehicles reduce in size, the manned vehicle is unchanged. But on a dollars-per-landing basis, that would be a very inefficient thing to do. See fig. 20 below for estimated savings from reducing the number of landings, total mission cost, and the prorated per-landing cost, using $11B as the total for a 16-landing mission.
Plus, there is the political effect of “getting much of the initial exploration done” in a single trip with 16 landings, in such a way as to enable future prospecting-base missions, and eventually, a colony. Compare that to the single-landing scenario, which would require more exploration missions before anything else could be done. Each and every one of these follow-up missions could be cancelled.
As a proper exploration strategy, allow me to suggest a “shotgun-pattern” planetary survey with the full 16 landings, perhaps to be followed by a second exploration mission of fewer landings at the most promising sites uncovered by the first mission. These fewer landings in the second mission would stay substantially longer times on the surface, rather similar to those proposed in “Mars Direct”.
Fig. 20 – Cost Trades vs, Number of Landings in a Single Mission
Once the exploration planetary survey is done, we are ready for a different type of mission in which “prospecting” bases are built. This is the type of mission where the in situ resources begin to be utilized, and the first indications are obtained as to what trade commodities there might be, and what the trade economy might be. These are the prerequisites for an actual colony in the future.
Alternatives to Gas Core Nuclear Thermal Rockets:
To do the fast trip manned vehicle with solid core technology requires a throwaway staged vehicle, which is neither cost effective, nor conducive to authorizing follow-on missions. Slowing to Hohmann-transfer speeds puts the total mission well over a year, which requires artificial gravity (and the resulting huge impacts on vehicle weight, size, and complexity, as well as costs). This path is not recommended.
One could spend efforts developing a flightweight nuclear electric power plant in the multi-megawatt range. Then the same manned fast trip could be done with VASIMR, or something very much like it. Developing such a power station is likely about the same risk as producing a gas core NTR engine. There would be a larger inert weight for the manned vehicle, and a much smaller propellant weight, of a different type, with VASIMR. The unmanned vehicles would look the same. This path is a recommended possibility, although there is less commonality with the unmanned vehicles.
Opting instead for nuclear pulse propulsion runs into the odd efficiency scaling that kind of propulsion entails: Isp is higher at larger vehicle masses. At the masses of these exploration vehicles (619 metric tons at departure from LEO for the manned vehicle, 690 tons for each of 3 unmanned vehicles), pulse propulsion Isp resembles no more than gas core NTR Isp, and is maybe not as good. At vehicle masses around 10,000 tons and up, pulse propulsion looks like Isp = 10,000 sec or higher, and this at vehicle accelerations in the 2-4 gee range. These kinds of characteristics are well suited to large scale operations like base-building and planting actual colonies.
A Note on Crew Selection:
This mission is planned around 6 crew members, going to the surface 3 at a time. Since the goal is a science information return, to be obtained with maximum crew safety, I suggest each group of 3 be one pilot/engineer, one geology specialist, and one chemistry/biochemistry specialist. That would be two of each comprising the 6 total. Each should be cross-trained enough to function as a lander pilot for emergencies. Each should be cross-trained enough to support the other science specialties.
The point of this study was to show that a manned mission to Mars is feasible, and safe, with launch rockets available today or within 5 years.
All of the known crew health and survival issues can be addressed in a design that can be assembled from docked modules that fit the presumed launch rockets. Mission times are short enough not to provide artificial gravity. To go further out than Mars will require artificial gravity.
There are two missing propulsion technologies: (1) an update of the old solid-core NTR “NERVA” technology, which could probably be available in under 5 years, and (2) a gas core NTR (or VASIMR equivalent), which will likely require about 10 years to make ready.
These vehicle designs that support a well-planned exploration of Mars are reusable, and could be utilized anywhere in the inner solar system.
(see also the 1-8-2011 post for an update to this study)