You are not logged in.
Update:
I looked at Zubrin's Sabatier paper. Very intriguing process. I did notice that the pressures in the system used gas feeds at 50 to 75 psi (3+ to 5+ bars), and the reactor was at 0.8 bar (Denver ambient). How to compress on Mars was unaddressed, something very worrisome considering that compressor weight is proportional to inlet density and something like 10 stages will be needed at minimum.
And, I rather suspect the conversion in the Sabatier reactor works better and better as reactor pressure increases. Chemistry is usually density dependent (another word for mass/volume concentrations in your Arrhenius-type overall reaction rate equations or empirical models). This atmospheric gas compression issue is a critical design issue for using technologies like this on Mars. That near-vacuum of an atmosphere is a real impediment, for a lot of things.
GW
Hi Josh:
Pressure ratio is not much of a compression problem. You just stack up enough stages, like you said. The problem is the mass throughput that you're compressing, which depends upon the first stage inlet density, which on Mars is about 0.7% of what we are used to. This is a governing factor for sizing machinery. Any air compressor on Mars will be physically enormous and very heavy in comparison to the total mass it processes in any one run.
Avoiding that dilemma takes innovation, which is why I suggested the self compression effect of phase change with heated, confined solid resources. There's a lot of large ice deposits all over Mars, usually from +/- 40 latitude to the poles, as I understand it. And, there's lots of ice and dry ice deposits at the poles. That tells me where the first ISRU-dependent landings ought to be made.
But, not every site will have the water. It's not evenly distributed in minable quantities. Based on mining experiences over the centuries here, no one should expect water on Mars to be evenly distributed. You have to locate the buried glacier, and land your base or colony next to it. No different than here.
See you and all the guys next week.
GW
Hi Bob:
Once you accelerate an airbreather to very high speed, you are pretty close to thrust = drag, leaving no way to pull-up without decelerating. The force balance just isn't there. That's why I suggested the very easy-to-implement parallel-burn airbreather-rocket approach for the pull-up staging transient, and it is a short transient. If you are using a ramjet for your airbreather, you already have rocket available, which has been the most practical method of accelerating a supersonic ramjet design to its takeover speed.
Missile-style integral rocket-ramjet is not very practical for a reusable design, and neither is a staged-off booster (way too bloody big, to be inexpensive). That suggests parallel-burn, possibly using a liquid rocket engine that uses the same fuel as the ramjet. That's usually a kerosene or a kerosene-like synthetic. So, you're looking at kerosene-LOX rocket engines packaged somewhere in an airframe with one or more kerosene ramjet ducts, and a set of kerosene and LOX tanks. This is all very doable, well-proven, off-the-shelf stuff. No gravy-train technology-development programs here. Those almost never lead to real flying machines.
If you did do an integral solid booster for the ramjet, then you would need to carry along some sort of JATO bottles to get the extra thrust for the pull-up transient. I don't really recommend that design approach, as it has too many one-shot components for a reusable/inexpensive design. The common-fuel all-liquid approach seems to me to have more potential.
The rocket and ramjet might use liquid methane instead of kerosene. This gets you a higher-performing booster for a little less propellant weight. The ramjet can burn it, although vaporization from the cryogenic requires more care in design than vaporization from a room-temperature liquid. At least, the inlet air approaching the combustor is hot (that helps a lot), especially at Mach 5+ as we approach staging. You don't get much of that effect at all in a subsonic/transonic pitot-type design. But you don't get much speed either: about Mach 2 tops.
That being said, I'd recommend a supersonic inlet design with external compression features, a C-D nozzle, a dump-stabilized combustor, and a takeover Mach in the 1.6-to-2 range. I'd also recommend a minimal variable-geometry feature on the inlet to maintain shock-on-lip throughout the flight Mach range. That sort of thing has a top speed potential in the Mach 5-to-6 range, more dependent on vehicle drag than on ramjet design. I'd operate it "full rich" as an accelerator-thruster all the way up, and only lean it back for a Mach 2-ish lean-burn cruise, back to base after staging. I'd also retain just enough rocket propellants on board to support at least one go-around for what is otherwise a dead-stick glide landing.
How to carry the second stage is the hardest part of such a design. But the same basic idea, operated single stage as an edge-of-the-atmosphere skip-glider at Mach 4+ to maybe Mach 6, becomes a possibility for a transoceanic-range replacement for the SR-71. The "air turboramjets" that pushed the old SR-71 had max 25% air bypass to the afterburner duct, and that came from compressor stage 3 or 4, not the supersonic inlet duct. It was limited to about Mach 3.5-3.8 max, because of hot-air temperature limitations on the turbomachinery that the ramjet does not face.
GW
I've been having serious difficulties signing-in, to discuss things here in the forums. So, I put a posting on "exrocketman" that addresses some of the issues with re-entry dynamics and heat protection. It's titled "Entry Issues", and right now it's the latest thing posted. A lot of that posting might be germane to the discussion here, especially as regards shipping cargo from the moon to Earth.
In my ignorance, I would guess the most practical source of oxygen on Mars to be mined water. One would use the vapor pressure rise upon heating to self-compress a batch of confined ice to usable pressures (near 1 atm). Then just do solar PV electrolysis. It's a whole lot easier to compress the hydrogen and oxygen from 1 atm into 2000 psig bottles than it is from 6 mbar.
What one does with the hydrogen is not well understood by me. It should be possible to make methane from it and the CO2, but the source of the CO2 makes a big difference to practicality. Compression from 6 mbar is a real practical problem, while dry ice can only be mined at the poles (same self-compression mechanism as water ice).
GW
My problem is most definitely a software incompatibility problem with Windows 8 on my new laptop. I just logged in easily on the first attempt, using my wife's older machine, which I think is Windows 7, or maybe one step earlier than that.
This is Microsoft trying rather unsuccessfully to make their devices resemble Apple devices in their screen appearance and operation. I am not very happy with them, as this log-in glitch is not the only trouble I have had.
GW Johnson
Login is still unreliable for me in this new Windows-8 laptop. It took over a dozen attempts this time. I've selected log-me-in-everytime, let's see if that works better.
GW Johnson
A couple of years ago, I visited XCOR in Mojave, CA. They are looking at a future orbital craft. The idea is 2-stage HTO. First stage could be a rocket, but is a lot smaller if an airbreather. They wanted to look at ramjet, which is why they were interested in me. I am a full-capability ramjet expert.
I don't have a copy of their internal report, but I do have the verbal ranking of the important issues at staging. Most important is velocity, the faster the better. Second is path angle, as near to 40 degrees as possible (which relieves the second stage of the hardware required to pull up by lift or thrust, it just flies a gravity-drag ballistic turn). Third is altitude at staging: the higher the better.
Of the airbreathers, ramjet offers the highest staging velocity with something that can be reliably applied right now. A clean, low-drag vehicle design can reach Mach 5 fairly easily, even Mach 6. ASALM-PTV accidentally reached Mach 6 back in 1980, and it was a very clean, low-drag wingless dart shape.
None of the airbreathers has much frontal thrust density above about 60,000 feet, simply because densities and ambient pressures are so low. Compression ratio is limited to what the inlet can do at that Mach. So thrust depends upon incoming pressure, but weight does not. There is no way around these physics. That is why to pull up sharply at staging in the thin air requires more thrust than the airbreather can deliver, or else the vehicle decelerates sharply as you pull up. The most practical way to achieve pull-up thrust is to fire up some rockets in addition to your airbreather; i.e.; parallel burn, not combined-cycle.
Since altitude is the weakest of three effects at staging, one can stage at 60,000 feet instead of 100+ thousand feet, and get way-to-hell-and-gone better thrust results from the airbreather. This means both velocity and pull-up angle can be fully achievable. Plus, at only 60,000 feet, the time and range to accelerate (on the airbreather alone) to max velocity are a whole lot shorter than those at the higher altitudes in the too-thin air. This can have an overwhelming impact on your first stage's design size and weight.
All in all, this suggests to me a parallel-burn rocket and ramjet first stage, boosting to takeover on rocket, flying to staging on ramjet. Staging is in the vicinity of M5 to M6 at somewhere near 60,000 feet, pulling up to somewhere right around 40 degrees at release. The necessarily-supersonic inlet system for stage speeds that high will have a min takeover speed around M1.6 or so, maybe as high as 2.
Any variable geometry ought to be on the inlet, and designed to maintain shock-on-lip, similar to that on the SR-71, but controlled to a different objective (turbine inlets are operated completely differently than ramjet inlets, even though both are made with exactly the same components). Any other variable geometry is just going to be too heavy, and too likely to be unreliable. That's the flying state-of-the-art.
GW
Hi Bob:
This is interesting stuff. 3-stage design, taking off on turbojet. Looks to be VTO, but must arc over and fly more of a HTO flight profile. Another patent for NASA, I guess. But, there's a fundamental problem with Mach 4 at 140-150 thousand feet with an airbreather of any kind.
This is not a lot different from some stuff I did and posted a couple of years ago over at "exrocketman". In the best form I found, I went two-stage for HTO, using rockets to take off and reach ram takeover speed, which for a supersonic inlet design is around Mach 1.5 to 1.6 min. You climb at near-takeover speed on ramjet, then pull-over and accelerate to high speed (Mach 5 to 6, actually). Then you must parallel-burn both rocket and ramjet to pull up sharply to release the second stage.
The first stage cruises back to base on ramjet at near-takeover speed, and glides to landing with reserve rocket propellant on board for go-around capability. The second stage is a straight rocket pod or plane. There are very severe frontal thrust density problems with ramjet above about 60 or 70 thousand feet. You can't climb or accelerate on airbreather alone, which is why combined-cycle airbreathing engines make little sense to me for launch applications.
For VTO, the ramjet assist is a strap-on, on a rocket core, intended to add a little thrust to the mix, just at much higher Isp. But, it's a simple pitot inlet, since most VTO designs leave the sensible atmosphere at about Mach 2-ish at around 80 thousand feet. Pitot inlets have subsonic ram takeover speeds. It's going to be a small effect, if it is worth it all. I'm not yet sure whether that scenario is really worthwhile. But I do know roughly what the fly-back strap-on pod ought to look like.
The problem is low frontal thrust density in the thin air above 60 thousand feet. All airbreathers are afflicted by that. Not even flying super fast with scramjet overcomes it. You either stay low and lose all your impulse advantage in drag, or you have to burn rocket and airbreather in parallel to achieve enough frontal thrust to climb/accelerate in the thin air. I just don't see any way around that dilemma. I see the potential for a lot of "gravy train" R&D programs, but I don't see much potential for anything we might actually fly.
The UK Skylon engine faces the same problem. They're pretty much done the airbreather by 80 thousand feet, gone to rocket only mode. For them, that's about Mach 4. Surprise, surprise!
GW
Why not do it this way? (Which was the gist of my paper at the 2011 Mars Society convention in Dallas.)
Do a first-trip voyage and make landings at many sites in the one trip. Drill deep (a km?) to sample subsurface like we do here at each landing site. Leave behind some gear and a transponder at each site. Each site's ground truth will differ from remote sensing (they always do), and each site will differ considerably from "average Mars" (that's true here, too). You try out multiple ISRU devices at each site, but you don't bet your lives on it.
Second trip a few years later focuses on the best one or two sites to plant a permanent base of some kind. You already know exactly what's there and have been able to plan exactly how to use it. The transponder lets you do a precision landing at that site. You get to use or recycle whatever gear was left there by the first expedition (and you know exactly what it was, which really helps). Betting lives on ISRU is a safe bet this trip, because you did it before, and you already know what works and what doesn't at this site.
The base then experiments with setting up real infrastructure for self support. Heavy construction will be involved, which tells you what kind of expedition this is, and what kind of landers you have to have. Eventually, that base or bases will become a colony. (By about the third trip.)
This kind of mission planning is completely independent of the mission architecture or equipment you actually use to make the trips and the landings. It's very important to think this through long-term, because it is so difficult to send men to Mars without killing them.
You have to deal with two different radiation killers (GCR and X-class solar flares), microgravity disease (plan on doing artificial gravity in transit), having enough space to stay sane (2 weeks in a capsule is the demonstrated max in such quarters, per Gemini 7, and confirmed in the Apollo moon flights), having food that stays edible for a 2.5-year trip (we're gonna need frozen food, despite the weight and volume, because "astronaut food" lasts 12-to-18-months max), a supple space suit (what's the point of going if you cannot actually do anything outside?), and you have to do all this packed-supplies (because we still cannot do a closed recycling ecology).
That last point (packed supplies vs closed ecology) is fundamentally why you cannot set up a viable colony on the first trip, unless you supply its every need from Earth at enormous and unsustainable cost! A self-supporting base or colony will necessarily have a closed ecology. We can't even do that here. Plus, it'll be different there, and by far. That's something the first base must experiment with. It may take decades to get it right, and the "time constant" for detecting success-vs-failure is measured in years.
I'd assess it this way: we have the technologies (maybe not all the hardware -- supple suit !!!!) in-hand to visit (and we have had them since about the 1990's), but we do not have those technologies required to stay long-term. Not yet. Few folks seem to be effectively working on them.
GW
Aside from the money-driven politics, there is another problem they had on X-33. Whether metal or composite, a conformal-shaped tank makes sense only if it is unpressurized.
No sane engineer would propose even a moderate-pressure pressurized vessel of any kind that wasn't cylindrical or spherical. When you store cryogenics, some pressure is usually involved, or else boiloff losses are very high. But apparently a lot of bureaucrats and managers were willing to propose it.
It's really hard to get such a tank to survive realistic loads, especially conformal ones. Further, why they tried that honeycomb thing beats me. Especially as easy as hydrogen penetrates right through so many materials. That was a disaster waiting to happen.
There is a limited class of conformal geometries you can approximate with cylindrical components. The structural soundness looks good, but the manufacture is very difficult. I once worked out a solid rocket motor case that resembled an air mattress, for a conformal situation. That's a very high-pressure vessel, by the way. It would have worked, and was only a mild nightmare to build. Though, once they saw it, nobody wanted to risk it anymore.
GW
Russel:
You know what? What you describe and what I describe are not so very far apart after all. You are deorbiting into the same surface-grazing orbit that I use. You just seem to be burning hard for retro propulsion near 100 km. It also sounds like you are getting benefits from aerobraking, just in the lower Mach range. That kind of thing lies between the two extremes that I was looking at.
My baseline mission concept was an orbit-to-orbit transport LEO to LMO and back, with at least the habitat and engine modules reusable. The landers (whatever they were) could push their own propellant supply to Mars one-way as separate vehicles, everybody to park in LMO. I was looking at re-using as much as I could, and staging multiple landings at different sites from LMO in the one trip to Mars. It's a very old concept, dating to the 50's and 60's in one form or another.
I hate to discard so much as a chute. Haven't yet figured out how to make a chemical lander one-stage/reusable except with propellant ISRU on Mars, and I'm not sure I trust that enough to bet lives on it. Not yet. But it really needs to be done. I see no point at all to the one trip/one-landing concept. If one is going to go to all the trouble to send men to Mars, then one really ought to explore a whole lot while there. That means a bunch of landings all over the planet. That kind of information return (a real planetary survey of ground truth) is what must be in-hand to plan viable bases and colonies.
You seem to know something about staging out of high Mars orbit vs LMO. Some sort of delt-vee reduction. I think you had in mind shuttling the landers back and forth. There's a cost there for shuttling, but the savings with transit vehicle might overcome that.
As for starting engines in SRP, a lot depends upon propellant selection. If they're hypergolic, it'd be hard to see how they wouldn't ignite, no matter the slipstream. Isp figures look just about the same for kerolox and MMH-NTO at 300-310 or so. methane-lox with reasonable non-vacuum bells would be similar, I think. But MMH-NTO is hypergolic, and storable without cryogenic considerations for very long times. I've seen no ISRU proposals to make MMH-NTO on Mars, though.
GW
Josh:
New computer is Windows 8, which I despise. No way around that, though. It uses windows explorer as its default, although I start from google's page.
Symptoms: I log in normally, and it says I did so successfully with the little brief message. But when I look top-of-page, it says login, not logout, and there is no box in which to enter text when I look at any given thread.
It took 6 tries to get in long enough to post this. 5 failures out of 6 attempts is not good. And that's with 3 browser restarts along the way. And, I cleaned out all the cookies and temp files before I started.
GW
20 cycles isn't reusable. This is essentially one-shot technology. But it is a dramatic weight-saver. That's good.
GW
Is anybody else having difficulty getting logged in in that last couple of days? Takes multiple tries, if successful at all.
GW Johnson
Hi Louis:
This one has been awfully quiet. But, I agree with you and Buzz.
Someone needs to go, and find out what it is really like. Then we are armed with the knowledge required to found a viable colony. Can't really do that on the first trip.
First trip colonies have a bad track record: Roanoke, then almost-Jamestown. Must explore properly, then colonize.
GW
OK, I don’t have any sort of trajectory code for landings. I have to do this as bounding or feasibility calculations. This just me sitting on the front porch with pencil-and-paper, and maybe a calculator.
I know that my approximate entry analysis is just that: approximate. Of course the tail end of the trajectory bends downward. With a capsule flown off-angle in pitch, you can generate a bit of aero lift, and offset that. We did exactly that with Gemini and Apollo. L/D = 0.1 is about all you can get, but it’s enough.
So, my way-oversimplified, straight-line 2-D Cartesian, entry model is not all that bad. It gets you into the ballpark, which is all it has to do. We’re looking at broad trades here, not fine details.
What I found, running systematic variations for aerobrake entry at Mars, is that you want the min credible de-orbit burn, from the min credible low orbit altitude, in order to minimize trajectory angle at entry. If you don’t, you hit before the hypersonics are over.
I also found that entry vehicles around or under 100 kg/sq.m ballistic coefficient tend to slow to M3 at 15-25 km altitudes, admittedly a bit variable due to the inherent variability of the Martian atmosphere. That’s high enough for chutes to have time to work, and it is typical of all the probes sent to Mars until very recently. I’ve been using the average atmosphere data in the Justus and Braun EDL paper for my calculations.
I also found that bigger vehicles with higher ballistic coefficients penetrate deeper before slowing. No surprises there. At 400 kg/sq.m I had about 5 km altitude at local M3. That’s about 2 minutes from impact at the very most. Not time for a chute to deploy, much less work. That’s the “classic EDL dilemma” everybody yammers about, but the way around is clear: no chutes, just thrust-to-landing.
Whether you wait to M3 to fire up the thrust is an unexplored issue. But you will have to use supersonic/hypersonic retro thrust. No way around that.
Actually, the plume stability issue is easily overcome, it’s just that NASA hasn’t looked at this since the development of Mercury, ca. 1960. But they did look at it then, and anyone easily could again. I’ve already written about how that can be resolved, and won’t repeat that discussion here.
That same Justus and Braun EDL paper describes the very old density scale height-based approximation for non-lifting entry estimates. The dynamics they showed in their paper looked pretty realistic, but the heating estimates were quite inconsistent. I chased this analysis back to its originator: H. Julian Allen at NACA in the 1950’s, and it is something I remember seeing some 4 decades ago in graduate engineering school, too.
I corrected the errors in the heating estimates in my spreadsheet version, and went to timeline-integrated heat totals, instead of the old closed-form estimates. My oversimplified-estimate data for Apollo returning from the moon look just about like the “real McCoy”, and not just for the heating, but also for the end-of-hypersonics location where drogue chute deployment could begin.
So, my model ain’t too bad. I think you can trust my numbers for aerobrake-to-rocket-final-descent. I see no problems up to at least 60 ton sizes, which means smaller vehicles are even easier. 0.05 km/s to deorbit plus about 1.4 for terminal descent. Total delta-vee to land: about 1.4 to 1.5 km/s.
What happens as you make the de-orbit burn bigger and bigger is two-fold: velocity at atmospheric interface is lower and lower, but trajectory angle is very much higher and higher. The angle increases faster than the velocity decreases.
For even very low ballistic coefficients, you whack the surface of Mars hypersonically, if that entry angle is steeper than about 2-3 degrees or so. That’s physics. The only way around this is to keep on retro thrusting all through entry. That’s what we’re discussing in this thread.
The ultimate limiting case is completely killing all the orbital velocity in your deorbit burn. Then you fall, from rest at orbit altitude, vertically downward. Vertical velocity builds due to gravity. Ignoring air drag, and using constant gravity, from 200 km at 0.38 gee, velocity at impact is near 1.22 km/s, which you have to kill with a last-ditch, last-second burn of that same magnitude. Total: 4.87 km/s delta-vee required for the vertical free-fall descent.
That’s a minimum “credible” figure for the delta-vee required beyond the 3.65 km/s for de-orbit at 200 km. There’s no terminal maneuver kitty in that. By the way, that’s also about M5 just as you begin your last-ditch burn. Aeroheating is getting very severe at Mach numbers like that, by the way.
The other extreme “option” is to let gravity accelerate you downward to a tolerable speed, say around 500 m/s, then burn at thrust equals weight, to descend at constant speed. That’s about M2 deep in the atmosphere, and it corresponds to a 400 sec (6.7 minute) descent from 200 km. Aeroheating is significant but not severe. Exposed metal works, even aluminum. But not any faster than that.
]
But, you have to burn for 400 sec to do that! Assume an average vehicle mass of 8 metric tons during that burn. 8000 kg. On Mars at 0.38 gee, that’s about 29.8 KN of thrust required to balance the pull of gravity. That’s almost 12,000 KN-s of total impulse, which at 300 s Isp (typical of kerolox, MMH-NTO, and methane-lox), is nearly 40 tons of propellant, far more than the spacecraft masses!
So, I have to conclude that this kind of all-low-speed controlled descent is simply not feasible. Even LH2-lox would make no difference to this outcome, Isp is around 450-or-so, for 26 tons of propellants. You’re not going to do it with any imaginable chemical propulsion. Not even NERVA at 900-1000 s Isp could do this.
Therefore, something close to free-fall from a maximum de-orbit burn is what you have to propose for a non-aerobraking descent. But, here’s the thing: if you have to protect from Mach 5 aeroheating anyway, you might as well protect from entry-speed aeroheating, and just do aerobraking. It’s exactly the same kind of protection.
Comparing the two extremes, aerobraking has less delta-vee required, less total propellant, and a smaller, lighter vehicle at entry interface. Meaning less to ship from Earth, which is lower cost.
Now, that’s not to say there might not be an optimum somewhere between these two extremes. As I already said, that trade is unexplored.
The safety architecture doesn’t enter into that decision. That comes later, and applies to both aerobrake entries and non-aerobrake entries, or anything in-between. I really like the idea of having multiple vehicles that could be refueled, adapted, and used as redundant ascent vehicles. How they got there makes no difference.
Finally, even at only Mach 5, beware of shock-impingement heating on adjacent nacelles and on the struts connecting them. Structures like that don’t survive well, even with ablative protection, above about Mach 3.5-4 here on Earth. There is a reason why all entry spacecraft to date have been single-body capsules or very “clean” winged designs. I really don’t think engines on struts around the periphery of a heat shield is going to work.
GW
Josh:
Go take a look at XCOR Aerospace's web page. Select "our products" and go look at their rocket engines and their piston propellant pumps.
They're still down in the 5000 lbth (and smaller) class, but this is what long-life reusable liquid rocket engines look like. When I was there 3 years ago, I got to talk with their head and their chief engineer about these devices. The engines they flew in the rocker racer aircraft have an estimated time-between-major-overhaul exceeding piston engines.
These engines have failure containment built-in, too. Thrust/weight won't look anywhere near as high as the one-shot designs everybody is used to seeing, but Isp is every bit as good. The extra weight is what you have to pay to achieve long life and inherent safety.
This technology looks like it is scalable to larger sizes, to me. It is what they will fly in Lynx.
GW
When I ran my Mars lander studies, I was able to achieve 5 km altitude at local Mach 3, simply by shallow-angle entry from low Mars orbit. The required delta vee for ascent (complete with gravity and drag loss estimates) is about 3.7 km/s to 200 km. That's the absolute minimum for powered landing flown as ascent-in-reverse, but you'll actually need around another km/s more, to cover all the real-world maneuvering and hovering.
To aerobrake from LMO, with my rather large vehicles (at 400 kg/sq.m ballistic coefficient, some 60 ton size at entry), I needed about 50 m/s de-orbit burn, and hit the "air" at about 1.6 degree down angle. Local Mach 3 at 5 km altitudes is 0.7 km/s velocity. So, I braked by 3 km/s without using but 50 m/s worth of propellant. From that point, theoretically I need only 0.7 km/s to land, but the actual real-world maneuver delta vee was closer to 1.4 km/s.
Still, mass ratio for 1.4 or 1.5 km/s delta vee, or mass ratio for at least 3.7 and very likely 4.7 km/s delta vee, just to land? You decide which one is the bigger, more expensive thing to launch and bring from Earth. Bear in mind, the ascent will require about 3.7 km/s, no matter how you choose to do it. That propellant all has to come from somewhere.
The aerobrake entry velocity is 3.7 km/s, which is a lot of local Mach numbers, but also a lot less energetic than from LEO (11 km/s). Peak heating rates are proportional to entry velocity cubed (not squared). Mars is a lot less demanding. PICA-X ablatives are a nice choice at 0.27 sp. gr. and about 1.6 inches thick, for one-shot vehicles. My ceramic heat shield study showed that black-surfaced alumino-silicates would be fine, even for direct entry at 5.6 km/s, as long as the angle was shallow. Skin temperatures at the stagnation point fell well under the 1290 C phase-change limit for shrinkage-cracking in alumino-silicates.
My "stuff" would be reusable, damage-tolerant, and easily repairable for fully-reusable vehicles. Think under an inch thick at densities near 0.03. Like shuttle tile, but a whole lot tougher, and buildable in large bolt-on panels. Two-component ceramic composite, laid up a lot like fiberglass is.
There's no need to try inflatable heat shields at Mars (plenty of use here from LEO though). You just go from hypersonics to rocket landing. There's a lot of tradeoffs yet to be done to select such a flight path, but the feasibility is there right now, and with very large capsule-like vehicles massing many tons.
GW
Oops, I forgot. My oddball ceramic heat shield is retained (two ways) upon backplates that are part of the vehicle shell assembly. The idea is to bolt-on directly to the interior structure, big extended panels of shell plating that have integral heat protection. They'd come off the same way for interior access.
GW
Hi Russell:
I think we're dancing around the extremes of a trade study here. It is quite possible to maximize aerodeceleration hypersonically at Mars, and then cope with a low altitude when you come out of hypersonics. In my studies, I just went to a retro-thrust powered landing, since for 0.7 km/s (local Mach 3) at 5 km altitude, even at low path angle you are a mere two minutes from impact.
In your study, retro thrust is used throughout the entry, which is going to reduce peak heating and total heat to absorb, however it is resisted. This will also increase the altitude at which one comes out of the hypersonics. That makes chutes and ballutes feasible, and therefore a thing you could consider. Although, I don't recommend a chute or ballute simultaneous with retro thrust, except for last-second touchdown of extremely heavy loads. It's too easy for flight control to be upset by wind gusts laterally.
The trade study would be when to start retro thrust during hypersonic entry. You say at the beginning of the hypersonics, I say at end of hypersonics, maybe there is an optimum in between! The idea here is to land the greatest tonnage of payload for the least tonnage of propellant. That would be true whether or not you plan to make return propellant while on the surface of Mars.
To my knowledge, the "when-during-hypersonics-do-we-start-retro-thrust?" trade study has never been run by anyone. People have shied away from concepts like that because of (what I consider to be relatively groundless) fears over firing through holes in heat shields and stability of retro plumes versus vehicle attitude control.
There is something very serious to consider about hypersonic flight conditions and vehicle configurations: extreme shock-impingement heating above about local Mach 5 or 6 speeds. This is a demonstrated risk from the old X-15 program. You must fly a very clean shape, not an assembly of nacelles, because each nacelle sheds a shock system that impacts the other nacelles, and the structures connecting them.
The X-15 flight with the scramjet test article replacing the ventral fin was the flight that reached Mach 6.67. The scram nacelle bow wave impinging upon the adjacent undersurface of the fuselage nearly cut the tail section off the bird in a matter of seconds, once about Mach 6 got exceeded. And don't forget that the skin was exotic refractory Inconel-X.
Josh:
There's rocket engines, and there's complicated rocket engines. Hydrogen-oxygen things tend to be quite complicated and therefore expendable one-shot items. Its only advantage is 450+ s Isp. Methane-oxygen might be complicated, or maybe not. It is 300-class Isp.
XCOR has the habit of making very reliable, very long-life reusable engines without much of an Isp penalty. They've certainly done it with kerosene-oxygen (300-class Isp), and I know they're working on methane-oxygen. Their engines don't look much like those NASA has always procured heretofore. They look like something you could safely put in an airplane. Some of them use piston pumps, not turbopumps.
The easiest, simplest, most reliable engines of moderate thrust that I know of are the hypergolic-ignition units that use hydrazine-nitrogen tetroxide (also 300-class Isp). Could be any of the hydrazines, but MMH seems favored. These can range from 1-pound thrusters all the way up to the huge OMS units on the shuttle. Pretty reliable technology, we've been using it since the late 50's. Something like it powered the old Titan-II’s that launched Gemini.
Unfortunately, MMH-NTO doesn't yet sound like something we might make in-situ on Mars. But it is quite dense and easily shipped, and extremely easily-storable for years at a time. I dunno, but MMH-NTO sounds like an ideal Mars lander/ascent vehicle fuel. Until we figure out how to make it there, we'll have to bite the bullet and ship it from Earth.
For the transit vehicle, cryo boiloff problems can be solved with only a modest loss for using LH2-LOX (450-class Isp) or LH2-nuclear (750-1100+ class Isp as solid core, depending upon details). But I doubt cryo tankage like that could also be successfully adapted to resist the heat and air loads of aerocapture at Mars. Too many conflicting requirements for the designs.
MMH-NTO (300-class Isp) is a serious mass ratio penalty for transit, but could easily be protected for aerocapture at Mars. There's a lot off tradeoffs there. Including LH2-LOX to Mars, aerocapture, and MMH-NTO for the return. Haven't seen anybody look at that yet.
I'm not yet sure at all how the ISRU propellant restrictions feed back into the lander design, much less aerocapture vs transit vehicle propellant restrictions. My point is that there are very serious restrictions on your mission design imposed by these choices. Your choices are not free.
If I was doing it without aerocapture or ISRU, I'd used LH2-nuclear for the transit vehicle, and MMH-NTO for the landers. I'd stage the landers out of LMO and refuel them there for reuse with propellant I brought along. I build this from 20-50 ton modules launched to LEO, with rockets we already have by next year, and docked there. I’d spend my money on the mission that way, instead of developing a giant rocket that can launch 100-ton modules, but that nobody else has any use for, yet.
There’s a brute-force baseline, with the best stuff we have. Now, try lowering the launched mass by aerocapture and ISRU, while staying within the propellant-choice restrictions they impose on your designs. That’s the way to reduce mission costs.
GW
Myself, I am not cognizant of all the issues and numbers being discussed here. But I do know something about entry.
The peak heating rate per unit area (W/sq.cm) at the stagnation point is proportional to velocity-at-entry-interface cubed. Not squared, cubed. Entry from LEO here at Earth is pretty close to 11 km/sec. From Mars LMO, it is only about 3.7 km/sec. Even for direct interplanetary transfers, the entry interface velocity at Mars is only in the neighborhood of 5.6 km/sec. Entry at Mars is less demanding than from LEO by an order, to orders, of magnitude, in terms of peak heating.
Back in the mid-1950's a fellow named Julian Allen at NACA did the reentry stuff for warheads here on Earth. He and those working with him looked at heat sinks, ablatives, and (non-ablative) "refractories" for that mission. All would work, but heat sinking was by far the heaviest of the three. Back then, ablatives (in spite of their densities back then, such as silica and carbon phenolic, which sink very quickly in water) were a bit lighter weight than the refractory solutions, which depended upon high-density things like graphite and tungsten.
That's why ablatives were selected for Mercury. The refractories were to be tested on the X-20 Dyna-Soar, which got cancelled. This early Allen stuff didn't get published in the open until the mid-1960's, when it was finally declassified. You can find it on the web now, I did.
Since then, there's been two changes: (1) ablatives got lighter, and (2) refractories got lighter. Heat sinks never did, since the thermal capacitance per unit mass is density*heat capacity, which maximizes for high density and high heat capacity, and both of those things always correlate positively for all known substances.
Refractories got lighter with shuttle tile ceramics, but these could not be used near stagnation regions due to temperature limitations of no more than 2350 F. That's why the shuttle nose cap and aerosurface leading edges were carbon-carbon composite ablatives. These are very heavy, somewhat fragile, and require replacement every few flights. The tiles themselves were very, very fragile, and very, very, very labor-intensive to maintain.
More recently, ablatives also got lighter with the lower-density PICA and PICA-X materials. You can fly these a few times from LEO (presumably several to many times from LMO), and they are far less labor-intensive to install and maintain than refractory shuttle tile. I did finally find a published density for PICA-X at 0.27 g/cc. The panels on Dragon look to be about 1.5 to 2 inches thick, for 2-3 flights LEO.
The stuff I came up with is extremely experimental, but it handled as if it were commercial Styrofoam (somewhere near 0.03 g/cc). I used 0.2 inches of it in a ramjet combustor running at almost 4000 F, and it withstood the extremely-violent effects of rich blow-out combustion instability repeatedly, while serving for hours of accumulated burn in dozens of tests. I cannot go that hot for entry, I must avoid shrinkage cracks by staying under 2350 F, just like shuttle tile. but, that's where refractories can take us. I think they are the ultimate winner for entry heat shielding.
As for retro thrust, Dragon's Super Draco's are arranged to fire around the heat shield perimeter at 45 degree cant, just so that they do not have to "solve" the heat shield port problem. They have plenty of thrust, that's the landing system for manned Dragons, even here on Earth. it's far easier on Mars.
What I was talking about was firing a retro engine through an open hole in the heat shield. You can do that, if you seal the engine compartment to stop all gas throughflow through the hole. Don't do this on centerline, the plume won't know which way to flip-flop as it reverses, leading to induced destabilizing forces of sufficient magnitude to tumble the craft. Do this off-centerline with cant angles in the 5-15 degree range, for "perfect" plume stability at little sensible off-angle thrust loss.
The basic no-throughflow idea for holes in the heatshield already worked in 1969 with the Gemini-B test flight before MOL was cancelled.
GW
Hi JoshNH4H:
I'll be at the banquet with my wife. We'll be looking for all of you. Unless she makes me leave it at home, I'll be wearing my straw cowboy hat. Look for the old guy with the slim wife. I'm 63.
GW
Hmmmm.
Entry heating protection at Mars is nowhere near as difficult as at Earth. Peak entry heating varies as entry interface velocity cubed, and it's a whole lot lower at Mars, by factors of 2 to 3.
There's a really good semi-reusable solution with Spacex's PICA-X ablative, which could be flown several times, maybe even a few dozen times, before replacement. And, it's fairly low density at 0.27 g/cc, you need around 2-3 inches of it for "long" life, and to cut off the conductive heat load.
Except for the demonstrated fragility and high maintenance costs, low density ceramics of the shuttle-tile type could be a longer-life solution. It takes about half an inch or so of that stuff to cut off conduction, and be processible for bonding the tiles. Except for damage repairs, the life is theoretically infinite.
I have an oddball low-density ceramic-composite material that I made and tested for a different use over 30 years ago. It has the density of Styrofoam (roughly 0.03 g/cc) and very low conductivity (around 0.035 W/sq.cm). It's a two-component composite laid up sort-of like ordinary fiberglass, and features redundant retention, even in large panels. It proved very tough, very damage-resistant, when I tested it so very long ago. I'll present a paper on it at the convention in Boulder CO this August. Much less than an inch of it would be needed.
That gets you through the hypersonics. That's where aero deceleration is most effective on Mars, completely unlike Earth.
If your entry angle was shallow (say near 1.6 degrees at interface), you come out of hypersonics at local Mach 3. On Mars, that's around 0.7 km/s at around 5 km altitudes, typically (in anything big enough to carry people or tons of cargo). That's too low to deploy a chute at all, much less have it decelerate you any noticeable amount, so why bother? Just go to direct rocket braking-to-touchdown from there.
You will use less fuel in a lower mass vehicle than in any other imaginable scenario, done that way. I already looked at a multitude of approaches. That was the best.
Firing through openings in heat shields is no problem, just seal the engine compartment behind the heat shield so that there is no through-flow through the hole. You might need a tad of coolant gas injection into the sealed compartment to make up for the volume-filling transient as you descend, but it's a very minor massflow.
Plume stability for supersonic retro thrust is an issue often raised, but easily addressed by using multiple nozzles canted off centerline just a few degrees. Spacex knows of this stabilizing effect already, why else would the super Draco thrusters be arranged to fire canted at 45 degrees around the edge of the Dragon heat shield? You don't have to fire at 45 degrees around the edge, you can fire through ports in the heat shield canted at only 5-10 degrees and still get stability.
Landing big stuff on Mars is not really such a difficult thing, but you do have to do something different than we have been doing all these decades. Fear-of-the-new more than actual technical issues is holding us back.
GW
Did any of y'all see the "mushroom building" concept I posted over at "exrocketman" a while back? Given a concrete substitute and a way to make glass in-situ, this sort of thing would work for both habitats and greenhouses, even in locations where burrowing into a hillside wasn't practical. See the article dated 1-26-13.
JoshNH4H: I haven't yet seen a presentation schedule for TMS convention. I dunno when I present during the event, not yet. But I sure do want to meet you, bobunf, midoshi, and anybody else, while I'm there. Do any of you plan to attend the banquet Sat nite?
GW