If the pressure-fed versus pump-fed question is to be debated, then mass budgets need to be compared. Descending down one layer into what determines the mass budget, one key question is how heavy are the tanks relative to the propellant they contain? Is there stored gas on board to push the propellants out, and how heavy are the extra tanks (or extra tank volume) for storing the gas?

Regarding the earlier suggestion to fill it up with Mars atmosphere then leave pumps on the ground, note that the Mars atmosphere is carbon dioxide, which is 11 times as heavy as helium. A good rule of thumb is that the best aerospace pressure vessels flying today can contain 1/4 of their own weight in helium. Or to put that a different way, the helium needed weighs only 1/4 of the vessel. If carbon dioxide were to be used instead, it would weigh more than the propellant tanks and gas storage vessel(s) combined. What does this look like on the mass budget?

Back in 2001, NASA funded 3 companies to go and design Mars ascent vehicles. The results, made public by AIAA paper number 2002-4318, were:
Lockheed-Martin, 268 kg solid rocket
TRW, 254 kg gel propellant rocket
Boeing, 400 kg pressure-fed liquid bipropellant (hypergolic)
These numbers don't count the mass of essential hardware that remains on Mars, just to support the launch.

The pressure-fed liquid was the heaviest option, and in recent years, all three have been considered too heavy to do the MSR mission within the Mars science budget. All this leads to the conclusion that it's really a nitty-gritty technology problem that is unlikely to be solved in the usual way by engineering design studies.

Before I forget to mention them again, here are two fun-to-read articles that offer an inside view of the people at NASA as they have worked to figure out the MAV. I'm going to mark this spot with a Mars so these refs can be found easily scrolling through here.


Reichhardt, "The One-Pound Problem," Air & Space (the Smithsonian's magazine), October-November 1999, page 50. I have read it online at http://www.AirAndSpaceMagazine.com/ASM/Mag.../1999/topp.html but if you can find the original in a library there are some nice artist concepts too. That article was written just before two Mars-bound spacecraft were lost, and years later the conclusions in that article turned out not to be the final answer.

The other is a two-part personal perspective by Dr. Mark Adler of JPL, on the Planetary Society blog in September 2006, http://www.planetary.org/blog/article/00000701/ and Part Two the next day at http://www.planetary.org/blog/article/00000703/

John W.

Maybe launch the MAV from a high-altitude balloon above earth, to simulate the atmospheric density on Mars. It wouldn't reach orbit, but the trajectory would tell how well it worked.

NASA has considered a bare-bones "scoop of dirt" mission. It is a tough decision to spend all that money without sending science instruments and/or a rover and/or a rock drill.

[...]

I can't imagine how you would land on Mars at all for a Discovery budge - let alone with any form of sample return ability. The only thing in that scale you could do is the SCIM proposal that is forever turned down in Mars Scout AO's

Doug

Apparently what you do is use leftover hardware from a previous, cancelled Mars landed mission. Phoenix plans to land on Mars on essentially a Discovery budget.

But yes, just getting MSR halfway back (e.g. to low Mars orbit) is not even in the ballpark of Discovery/Scout budgets.

And wouldn't you know it, I bet there isn't any leftover Martian orbital ascent hardware lying around anywhere, is there?.

I dunno, Dan -- the last MSR concept I saw (back in the late '90s) used some leftover, off-the-shelf solid-fuel military missile as its basis for an ascent vehicle. I bet there are at least two or three of them left that haven't been fired in anger yet...

-the other Doug

I thought it was planning to land on ice deposits

It's a fair point - but if you take the '01 hardware costs, the scout budget, and the little-bit-extra - Phoenix isn't something you could do from the ground up for a discovery budget. If you got to a genuine 'build to print' state, maybe it'd work?


Doug

Well, don't want to start any debates but remember:

The very first unmanned (Lunar) sample return mission:
http://en.wikipedia.org/wiki/Luna_16

basic MSR JPL-weblink
http://mars.jpl.nasa.gov/missions/samplereturns.html

Speaking of left over equipment

That thing looks like it was made from an old oil drum and surplus scuba gear.

Indeed an awkard looking spacecraft and this 'thing' had a weight of almost 2000 kilograms with the ascent stage over 500 kilograms... and the upper 'ball' (in fact the sample re-entry capsule) had a diameter of 30 centimeters... of course the whole thing 'only' had to escape the Moon's gravity
When we compare the 1971 Mars 2 or 3 spacecraft configuration, it looks almost identical to the Luna 9 or 13 surface capsules. So probably a Soviet-Russian MSR craft might have looked like the 'ugly' Luna 16?

Might be worth considering & contrasting US & old Soviet-era (SE) design approaches when thinking about MSR. From what I gather, most SE flight hardware was very rugged, implying that functional modules were optimized for their specific performance, and holistic system interfaces/dependencies were minimized in order to reduce risks. The US approach was almost diametrically opposite, relying instead on a robust C&DH capability to adaptively sequence critical events, which in turn allowed more trade space with respect to subsystem performance margins.

Wonder if there just might be a truly optimal middle ground, here...

I was recently fascinated to learn how the Soviet Luna (16, 20, & 24) achieved its ascent and earth return. Yes, it was functionally very simple engineering, tailored to the particular physical situation. The moon's small size (compared to Mars) permitted a direct return. Not going into lunar orbit meant no circularization (orbit insertion) burn, and the fact that the target (earth) was gravitationally large and nearby meant no midcourse corrections either. No need for any engine restarts or staging. A single propulsive burn from the 1-stage ascent vehicle was simply timed (both moment of launch relative to the calendar, and burn duration).

Guidance consisted of flying a vertical trajectory off the moon. The vernier engines were controlled by a local vertical sensor, a pendulum! Site selection was limited to the east side of the moon, where a vertical ascent reduced the geocentric velocity compared to the moon's, so it was effectively just a deorbit burn with respect to the earth. Velocity would have been less than lunar escape velocity, since the earth was sitting there pulling it home. The return stage had a transmitter that could be switched on and off by commands from earth, and the resulting signals received on earth were used to predict the landing point accurately enough to go out and find it.

All this is explained in a paper by Boris Girshovich, presented at the National Space Society's 26th International Space Development Conference, Dallas Texas 2007May25-28. See isdc.nss.org/2007/index.html.

My notes from reading the above paper say that the earth entry capsule was 3 feet in diameter, while the above posting from PhilCo126 a couple days ago says 30 cm. I suspect both numbers are from memory or word of mouth, so does anyone have any solid references to cite here? Has anyone been to Russia where the capsule is presumably in a museum somewhere?

Mars ascent is MUCH harder to do, considering the need for a smaller size, higher delta velocity, double the thrust-to-weight, and more complicated navigation to orbit.

John W.

Fascinating & ingenious; really doing more with less. Thanks, John!

Maybe the way to approach this is to design a mission based on what we know we can do, rather then what we hope to be able to develop. (This is not meant to limit technological advance, but instead to constrain the problem). For example, if we assume direct EDL and payload DTE return, this simplifies the mission requirements considerably in some ways but possibly complicates them in others (one being the ability to meet really tight launch windows from Mars to Earth).

Just throwin' that out there...

Couldn't resist running a quick & dirty trajectory simulation to compare a MAV with military missiles. First verified the flight of a 100-kg MAV by running the trajectory simulation to reach a 500-km circular Mars orbit using about 4150 m/s delta velocity (results match a previous case in the Journal of Spacecraft & Rockets, Nov 2005 p. 1041). Then simulated flight of the same vehicle starting on earth. Had to increase thrust 50 percent so it exceeds earth weight of the vehicle. To reduce the effect of the thick atmosphere on such a tiny vehicle, moved the launch site to 10 km altitude (32,800 feet). The latter seems fair for comparison to air-to-air missiles, which might launch at such an alitude.

The simulation result indicates that a 100-kg Mars ascent vehicle launched 10 km above earth can go more than 500 km downrange. Now, what military missile in this size class has such a capability?

Take a look at www.designation-systems.net, and click Directory of US Military Rockets and Missiles. Note the extensive list available in the selection box. As an example, the latest Sidewinder (AIM-9) is said to have a mass just under 100 kg, but its range is said to be only tens of kilometers. A Navy Standard Missile (RIM-156B or RIM-161) has roughly the same reach as a MAV, but it weighs about 1.5 metric tons. Based on these 2 examples, military missiles appear to offer only a tenth the distance relative to mass, compared to what Mars ascent needs.

While it might be possible to push solid rocket technology toward sufficiently less inert mass to make a solid-propelled MAV, there is no indication that anything off the shelf is capable. If performance details for military missiles and their rocket motors could all be public, there would probably be a more widespread appreciation of just how much harder it is to make a MAV.

John W.