Author Topic: Landing and takeoff on unprepared surfaces  (Read 11967 times)

Offline IainMcClatchie

  • Full Member
  • ***
  • Posts: 336
  • San Francisco Bay Area
  • Liked: 196
  • Likes Given: 314
Landing and takeoff on unprepared surfaces
« on: 09/24/2016 12:06 AM »
Return from Mars is going to require not just a soft landing on an unprepared surface, but liftoff from that same surface.  Apollo managed this by using the lower stage of the LEM as the launch platform for the upper stage of the LEM.  If SpaceX means to reuse the entire vehicle, then the entire unit must lift off from the surface that it landed on.

There appear to be a number of challenges to an unprepared liftoff that I haven't seen discussion about.
  • Liftoff might require more thrust than landing, as the vehicle will presumably take on a full load of fuel while on the surface.
  • Mars surface to Mars->Earth transfer requires at least 6300 m/s delta-V, and at least 1000 m/s more would be helpful to get the transfer time down from 6 months into the 3-4 month range.  Two stages would reduce the size of the vehicle.
  • Helium pressurant will not be available.  Propellants will have to be autogenously pressurized.
  • Will electrical power be available?  It would seem that any ground station capable of manufacturing hundreds of tons of propellant will have electricity, but it also seems that expendable umbilicals are unlikely.
  • Debris mitigation.  It seems having the engines on the bottom will not be practical, and so side-mounted engines, pointing away at 45 degrees or more from the vehicle axis, will be required.  Large cosine losses when producing thousands of m/s of delta-V seem undesirable, so that the engines will have to gimbal through large ranges.

  • Offline Hobbes-22

    • Full Member
    • ****
    • Posts: 434
    • Acme Engineering
      • Acme Engineering
    • Liked: 148
    • Likes Given: 95
    Re: Landing and takeoff on unprepared surfaces
    « Reply #1 on: 09/24/2016 10:55 AM »
    A takeoff on an unprepared surface would blast any nearby settlement with debris. You'd need at least a berm (hardened, not loose rubble) between the rocket and any structure.

    Given that debris from Apollo ascent stage takeoff could end up hundreds of miles away, I expect it wouldn't be practical to just move the launch site away from the settlement (you'd end up so far away that travel becomes a burden).

    Offline biosehnsucht

    • Full Member
    • ***
    • Posts: 305
    • Liked: 103
    • Likes Given: 202
    Re: Landing and takeoff on unprepared surfaces
    « Reply #2 on: 09/24/2016 08:19 PM »
    Perhaps they can locate natural berms for landing zones such that they can land on one side of a hill and then around the other side of the hill "set up camp"?

    You could also construct them with bulldozers built for mars, which could even be essentially rovers of a sort (as in, you could do this with your first robotic mission), but taking advantage of existing hills might be easier.

    Eventually i'm sure they'd end up with numerous prepared surfaces for future vehicles, but it would probably be some time before that could be constructed ...

    Offline M.E.T.

    • Full Member
    • ****
    • Posts: 579
    • Liked: 255
    • Likes Given: 18
    Re: Landing and takeoff on unprepared surfaces
    « Reply #3 on: 09/26/2016 09:22 AM »
    This is something which has bothered me, particularly after the recent SpaceX pad failure. If launching rockets is so incredibly risky, even under the extremely controlled conditions at an Earth based launch pad, what would the likelihood of a successful return launch straight from the unprepared Mars surface be? Well below 50%?

    Offline Alastor

    • Full Member
    • **
    • Posts: 272
    • Liked: 148
    • Likes Given: 330
    Re: Landing and takeoff on unprepared surfaces
    « Reply #4 on: 09/26/2016 09:38 AM »
    Remember that there will be a few unmanned missions before people go to mars and there is any kind settlement involved.

    We can assume some kind of timeline like that :
     - First few missions will be one way unmanned and contain robotized equipment tasked with preparing a landing and take off site and possibly facilities for refueling, energy etc.
     - Then we will see some unmanned missions that will take off. It is possible that some of the equipment to enable taking off (fuel production, this kind of things) will arrive as late as that.
     - Late unmanned missions will bring along habitat equipment.
     - Only once it has been determined that the risks of bringing humans along is reasonable enough will the first manned flights take place.

    So I believe taking off on an unprepared surface is a non issue when it comes to MCT.

    Offline M.E.T.

    • Full Member
    • ****
    • Posts: 579
    • Liked: 255
    • Likes Given: 18
    Re: Landing and takeoff on unprepared surfaces
    « Reply #5 on: 09/26/2016 09:43 AM »
    I suspect, though, that it will  be some time before you have the type of infrastructure in place to move a BFR around on the surface. So, either the surface will have to be prepared underneath the BFR AFTER it has landed, or else the BFR will need to land on the exact spot that has been prepared for it to take off from again.

    Which will require a level of precision not seen before in Mars landings. Lets say you land just 50m off target. There won't be a crane system or any other mechanism by which to reposition the rocket for take-off.

    Offline Alastor

    • Full Member
    • **
    • Posts: 272
    • Liked: 148
    • Likes Given: 330
    Re: Landing and takeoff on unprepared surfaces
    « Reply #6 on: 09/26/2016 11:47 AM »
    If you can master a propulsive Earth landing with a precision down to <10m (which they do), I don't really see why you shouldn't be able to do so on Mars.
    Obviously the first flights will be there to test and possibly refine that, but landing with precision on Mars is currently probably one of the lesser challenges that they have to get under control.

    Online Bob Shaw

    • Full Member
    • ****
    • Posts: 1083
    • Liked: 446
    • Likes Given: 388
    Re: Landing and takeoff on unprepared surfaces
    « Reply #7 on: 09/26/2016 12:21 PM »
    Perhaps they can locate natural berms for landing zones such that they can land on one side of a hill and then around the other side of the hill "set up camp"?

    You could also construct them with bulldozers built for mars, which could even be essentially rovers of a sort (as in, you could do this with your first robotic mission), but taking advantage of existing hills might be easier.


    Just put your base in a crater. Some will be unsuitable, others less so, but imagine landing near Victoria Crater on Meridiani Planum, with thousands of square miles of nice flat land around it. Craters shield you from radiation and nearby impacts (a distinct danger) and in some areas will provide access to ice.

    Online Bob Shaw

    • Full Member
    • ****
    • Posts: 1083
    • Liked: 446
    • Likes Given: 388
    Re: Landing and takeoff on unprepared surfaces
    « Reply #8 on: 09/26/2016 12:23 PM »
    A home from home?

    Offline sghill

    • Full Member
    • ****
    • Posts: 1385
    • United States
    • Liked: 1535
    • Likes Given: 2262
    Re: Landing and takeoff on unprepared surfaces
    « Reply #9 on: 09/26/2016 04:26 PM »
    We have a dedicated thread for this topic already.

    https://forum.nasaspaceflight.com/index.php?topic=37466.0
    Bring the thunder!

    Offline savuporo

    • Senior Member
    • *****
    • Posts: 5155
    • Liked: 985
    • Likes Given: 343
    Re: Landing and takeoff on unprepared surfaces
    « Reply #10 on: 09/26/2016 06:53 PM »
    Obviously the first flights will be there to test and possibly refine that, but landing with precision on Mars is currently probably one of the lesser challenges that they have to get under control.
    You are joking, right ? This is very far from correct.

    http://www.nasa.gov/mission_pages/msl/multimedia/pia16039.html

    And, the future evolution:
    http://mars.nasa.gov/mars2020/mission/technology/entry-descent-landing/
    « Last Edit: 09/26/2016 07:18 PM by savuporo »
    Orion - the first and only manned not-too-deep-space craft

    Offline Zed_Noir

    • Senior Member
    • *****
    • Posts: 2533
    • Canada
    • Liked: 393
    • Likes Given: 592
    Re: Landing and takeoff on unprepared surfaces
    « Reply #11 on: 09/27/2016 01:40 AM »
    Obviously the first flights will be there to test and possibly refine that, but landing with precision on Mars is currently probably one of the lesser challenges that they have to get under control.
    You are joking, right ? This is very far from correct.

    http://www.nasa.gov/mission_pages/msl/multimedia/pia16039.html

    And, the future evolution:
    http://mars.nasa.gov/mars2020/mission/technology/entry-descent-landing/

    All of the previous Martian landers uses a parachute after a ballistic reentry. If SpaceX can hit a barge in the Atlantic Ocean without any misses so far. Then landing on a fix surface point on Mars seem a lot easier by comparison. Will point out that a SpaceX Martian lander will be actively maneuvering towarf a landing point during descend.


    Offline Jim

    • Night Gator
    • Senior Member
    • *****
    • Posts: 32440
    • Cape Canaveral Spaceport
    • Liked: 11187
    • Likes Given: 331
    Re: Landing and takeoff on unprepared surfaces
    « Reply #12 on: 09/27/2016 01:43 AM »
    No, not really.  There is no GPS on mars

    Offline savuporo

    • Senior Member
    • *****
    • Posts: 5155
    • Liked: 985
    • Likes Given: 343
    Re: Landing and takeoff on unprepared surfaces
    « Reply #13 on: 09/27/2016 01:44 AM »
    All of the previous Martian landers uses a parachute after a ballistic reentry. If SpaceX can hit a barge in the Atlantic Ocean without any misses so far. Then landing on a fix surface point on Mars seem a lot easier by comparison. Will point out that a SpaceX Martian lander will be actively maneuvering towarf a landing point during descend.
    And thats got very little to do with why the landing ellipses are as large as they are. No 'actively maneuvering' is possible if you don't know where you are.
    Orion - the first and only manned not-too-deep-space craft

    Offline envy887

    • Senior Member
    • *****
    • Posts: 4714
    • Liked: 2621
    • Likes Given: 1409
    Re: Landing and takeoff on unprepared surfaces
    « Reply #14 on: 09/27/2016 01:56 AM »
    GPS isn't the only possible navigation system. Terrain following or rf beacon homing could allow sub 100m navigation accuracy, perhaps even sub meter accuracy.

    Offline savuporo

    • Senior Member
    • *****
    • Posts: 5155
    • Liked: 985
    • Likes Given: 343
    Re: Landing and takeoff on unprepared surfaces
    « Reply #15 on: 09/27/2016 02:11 AM »
    GPS isn't the only possible navigation system. Terrain following or rf beacon homing could allow sub 100m navigation accuracy, perhaps even sub meter accuracy.
    Read the links above ( and we have other threads about this here ).
    There are many challenges implementing better position estimation systems, and unfortunately the biggest error accumulation happens in hypersonic phase of flight where the lander is basically blind and deaf.
    Orion - the first and only manned not-too-deep-space craft

    Offline biosehnsucht

    • Full Member
    • ***
    • Posts: 305
    • Liked: 103
    • Likes Given: 202
    Re: Landing and takeoff on unprepared surfaces
    « Reply #16 on: 09/27/2016 03:50 AM »
    I'd be surprised if they didn't have some sort of quasi-GPS system in place via combined comms and navigation satellites pre-positioned before MCT gets there. Especially if you know where you want to land well in advance, you don't need global coverage just yet. They could potentially be sent as specialized trunk payloads on RD missions or sent similar to RD missions but as dedicated payloads. They're going to want if not need high bandwidth comms from Mars to Earth anyways, they've spoken of using laser comms, and you'd do that orbit-to-orbit rather than surface-to-surface. So just use the same assets for a rudimentary positioning system.

    I'm not sure what the minimum number necessary would be, but if you already have a ballpark idea of where you are, then you probably don't need as many sats to have effective coverage. Enough areosychronous (or whatever the term would be) satellites to get basic triangulation from their known positions is probably enough, when combined with other data (i.e., you probably know whether you're in the north or southern hemisphere, so as long as you're not landing on the equator that's one less thing you need to determine, meaning one less "dimension" of satellite navigation coverage you need). Throw in star / planet tracking prior to entry and inertial during, and even 2 satellites along the equator in view are probably enough to get accuracy within dozens of meters (at least of positional knowledge, landing accuracy might not be that good due to limits of control authority etc), maybe even approaching old civilian grade GPS accuracy.

    Offline savuporo

    • Senior Member
    • *****
    • Posts: 5155
    • Liked: 985
    • Likes Given: 343
    Re: Landing and takeoff on unprepared surfaces
    « Reply #17 on: 09/27/2016 04:05 AM »
    I'd be surprised if they didn't have some sort of quasi-GPS system in place via combined comms and navigation satellites pre-positioned before MCT gets there..
    According to people who have career in Martian EDL solutions, the biggest immediate improvement for reducing the landing ellipse sizes would be a local weather station or two.
    Orion - the first and only manned not-too-deep-space craft

    Offline CuddlyRocket

    Re: Landing and takeoff on unprepared surfaces
    « Reply #18 on: 09/27/2016 05:34 AM »
    GPS isn't the only possible navigation system. Terrain following or rf beacon homing could allow sub 100m navigation accuracy, perhaps even sub meter accuracy.

    You don't need accuracy, you need precision. Whatever your navigation system, you set it up, decide where you want to land, measure that spot's position according to that system, and then have your lander aim for that position using the same system. As long as your system gives the same answer every time, it doesn't matter how accurate it is! It's exactly how the F9 first stages find the barge.

    Offline Stan-1967

    • Full Member
    • ****
    • Posts: 544
    • Denver, Colorado
    • Liked: 286
    • Likes Given: 198
    Re: Landing and takeoff on unprepared surfaces
    « Reply #19 on: 09/27/2016 06:11 AM »
    GPS isn't the only possible navigation system. Terrain following or rf beacon homing could allow sub 100m navigation accuracy, perhaps even sub meter accuracy.

    You don't need accuracy, you need precision. Whatever your navigation system, you set it up, decide where you want to land, measure that spot's position according to that system, and then have your lander aim for that position using the same system. As long as your system gives the same answer every time, it doesn't matter how accurate it is! It's exactly how the F9 first stages find the barge.

    I think F9 is a very imperfect comparison for Mars EDL.  If you are parsing the meaning of accuracy vs. precision, the way I think you are, it is incorrect to say you do not need accuracy.   Accuracy is needed before precision, because if you miss your Mars entry interface by a few dozen seconds due to atmospheric variation that day, & your spaceship is going around 12,000 mph, you are now around a hundred miles uprange or downrange.   There is no guarantee that your level of precision in the guidance algorithms will be able to compensate for the range error if your control authority given by aero surfaces and propulsion can't get to a solution.

    With F9 these variables are very well characterized by now, and other than wind speed, and variations in the rocket performance, there is not much that is an unknown.   Mars EDL is likely to be different for every landing, so as said before, weather stations both on the ground and perhaps even some method to measure the localized atmosphere at entry interface is going to become the priority.   Building the GPS & timing infrastructure is probably the smaller problem.



    Offline biosehnsucht

    • Full Member
    • ***
    • Posts: 305
    • Liked: 103
    • Likes Given: 202
    Re: Landing and takeoff on unprepared surfaces
    « Reply #20 on: 09/27/2016 06:49 AM »
    Is it upper atmospheric or just atmospheric in general they need? I.e. do they need weather satellites or weather stations. It should be "trivial" (as far as potential RD payloads go) to send basic air temp/pressure/speed/direction sensors on all RD missions and thus litter the interesting areas with weather stations, but I suspect it's the upper atmosphere that will matter the most (since lower atmosphere you'll have active terminal guidance for).

    So let's make those comms/nav satellites also weather sats. After all, the weather sats need to communicate, and you can use their comms for navigation as well. If we're building more than one, then producing them is slightly "easier" / "cheaper" per unit, even if it's not really enough to call it mass production. Send one on one synod, make sure it works as intended and if not correct, then send the rest on the next? Or we can send them with SEP out-of-synod since there's no humans on board.

    Mars Climate Orbiter was only 338kg including it's hydrazine maneuvering and attitude thruster system, and it's orbital insertion system (hydrazine + NTO). Let's say we double the mass for some really big solar arrays and such, then double it again to add on newer fancier weather instruments and laser communication gear, that's still less than 2000kg. Looks like F9 claims to 4000kg "to Mars", so hell we don't even need FH to start sending these communications / navigation / weather satellites!

    Alternatively, a more realistically sized satellite might be under 1000kg each, and if they shared the initial insertion motor and used SEP instead of hydrazine for station keeping & etc, you could send perhaps 4 at a time on F9 or eight on FH! You might be able to set up a proper constellation in only a few launches after all... and have global weather, communication, and navigation capabilities. A single launch on F9 might get you 360 coverage from aerostationary orbit, or you can use a few launches and put up a GPS-style constellation covering the entire planet.

    Offline mvpel

    • Full Member
    • ****
    • Posts: 1124
    • New Hampshire
    • Liked: 1295
    • Likes Given: 1686
    Re: Landing and takeoff on unprepared surfaces
    « Reply #21 on: 09/27/2016 06:52 AM »
    The graybeard engineers around here chuckle at the youngsters when they lament the difficulty of autonomous-systems navigation in GPS-denied airspace. Back in the 70s, all airspace was "GPS-denied" airspace. :D


    Sent from my iPhone using Tapatalk
    "Ugly programs are like ugly suspension bridges: they're much more liable to collapse than pretty ones, because the way humans (especially engineer-humans) perceive beauty is intimately related to our ability to process and understand complexity. A language that makes it hard to write elegant code makes it hard to write good code." - Eric S. Raymond

    Offline savuporo

    • Senior Member
    • *****
    • Posts: 5155
    • Liked: 985
    • Likes Given: 343
    Re: Landing and takeoff on unprepared surfaces
    « Reply #22 on: 09/27/2016 07:35 AM »
    The graybeard engineers around here chuckle at the youngsters when they lament the difficulty of autonomous-systems navigation in GPS-denied airspace. Back in the 70s, all airspace was "GPS-denied" airspace. :D
    Yep, now start the runway approach with 11000mph, large initial velocity and position uncertainty, with unknown air density and winds. Wont be able to see the ground most of the time due to plasma sheath, and if you phone the tower, it'll take about half an hour to get a reply.
    Did i miss anything ?

    It would be a tad more reasonable if one were to brake and capture to orbit first.
    Orion - the first and only manned not-too-deep-space craft

    Offline Alastor

    • Full Member
    • **
    • Posts: 272
    • Liked: 148
    • Likes Given: 330
    Re: Landing and takeoff on unprepared surfaces
    « Reply #23 on: 09/27/2016 09:18 AM »
    According to people who have career in Martian EDL solutions, the biggest immediate improvement for reducing the landing ellipse sizes would be a local weather station or two.

    And you don't think that if they are able to send robotized payloads that are able to build a landing pad, they can also probably send a weather station network, a few comms/weather/positioning satellites at the same time ?

    I'm not saying that doing an accurate landing on Mars is something they would be able to do tomorrow. I'm saying that they know how to do it. They have demonstrated that very well on earth. It probably needs quite some logistics and other equipment to be brought there beforehand, but that's what the previous one way missions are for.

    They have much bigger challenges than precision landing on quite a few things that we currently do not master, to be able to sustain people living on Mars and bring them back.
    So to answer your previous question. No, I'm not joking.

    Offline JamesH65

    • Full Member
    • ****
    • Posts: 905
    • Liked: 592
    • Likes Given: 8
    Re: Landing and takeoff on unprepared surfaces
    « Reply #24 on: 09/27/2016 10:15 AM »
    Whatever scheme SpaceX use, they are going to need very accurate landing, because they are going to have to land close to any existing base, where the fuel is. Landing ellipses are out. Not easy, but still easier than a lot of other stuff they need to get right.

    Offline rocx

    • Full Member
    • ***
    • Posts: 385
    • NL
    • Liked: 265
    • Likes Given: 145
    Re: Landing and takeoff on unprepared surfaces
    « Reply #25 on: 09/27/2016 10:33 AM »
    I'd be surprised if they didn't have some sort of quasi-GPS system in place via combined comms and navigation satellites pre-positioned before MCT gets there..
    According to people who have career in Martian EDL solutions, the biggest immediate improvement for reducing the landing ellipse sizes would be a local weather station or two.
    I assume that most of their EDL solutions include a hypersonic parachute and do not involve retropropulsion. With retropropulsion you should be able to actively correct during a long part of the descent. Still leaves the problem of knowing where you are.
    Any day with a rocket landing is a fantastic day.

    Offline LouScheffer

    • Full Member
    • ****
    • Posts: 1910
    • Liked: 2501
    • Likes Given: 279
    Re: Landing and takeoff on unprepared surfaces
    « Reply #26 on: 09/27/2016 01:24 PM »
    About SpaceX barge landings showing accurate Mars landings are posssible...
    No, not really.  There is no GPS on mars
    Apollo achieved quite accurate (10s of meters) landings on the moon without GPS.   They used a combination of a reasonably accurate initial trajectory, corrections based on looking out the window and accurate maps, and sufficient control authority.

    This should work fine for Mars, I'd think.  Initial nav can be differential with respect to the satellites already around Mars - these have orbits known within a few meters.  There are plenty of features of all sizes for visual navigation, and we have good maps for the whole surface, and can generate truly excellent maps (sub meter resolution) for any potential landing spots.

    Finally, most of the time where a super accurate landing is needed, it's because you want land close to a previous mission.  That mission can have a beacon for this exact purpose.

    So while landing in an exact spot is certainly not trivial, if you've got the control authority, it should be fairly straightforwardly possible.


    Offline savuporo

    • Senior Member
    • *****
    • Posts: 5155
    • Liked: 985
    • Likes Given: 343
    Re: Landing and takeoff on unprepared surfaces
    « Reply #27 on: 09/27/2016 02:22 PM »
    I assume that most of their EDL solutions include a hypersonic parachute and do not involve retropropulsion. With retropropulsion you should be able to actively correct during a long part of the descent. Still leaves the problem of knowing where you are.

    You assume wrong. The current landing systems fly a hypersonic lift trajectory with active bank angle modulation, i.e. active correction. Again, this is not the reason why the error accumulation is as large as it is.
    Orion - the first and only manned not-too-deep-space craft

    Offline savuporo

    • Senior Member
    • *****
    • Posts: 5155
    • Liked: 985
    • Likes Given: 343
    Re: Landing and takeoff on unprepared surfaces
    « Reply #28 on: 09/27/2016 02:23 PM »
    Apollo achieved quite accurate (10s of meters) landings on the moon without GPS.   They used a combination of a reasonably accurate initial trajectory, corrections based on looking out the window and accurate maps, and sufficient control authority.

    This should work fine for Mars, I'd think... 

    Moon has no atmosphere, making this a completely different, and a whole lot easier problem

    EDIT: Oh, and it also happens to be slightly closer, making the initial position and velocity estimate error from tracking a few orders of magnitude smaller.
    « Last Edit: 09/27/2016 03:02 PM by savuporo »
    Orion - the first and only manned not-too-deep-space craft

    Offline Stan-1967

    • Full Member
    • ****
    • Posts: 544
    • Denver, Colorado
    • Liked: 286
    • Likes Given: 198
    Re: Landing and takeoff on unprepared surfaces
    « Reply #29 on: 09/27/2016 02:34 PM »
    Apollo achieved quite accurate (10s of meters) landings on the moon without GPS.   They used a combination of a reasonably accurate initial trajectory, corrections based on looking out the window and accurate maps, and sufficient control authority.

    This should work fine for Mars, I'd think... 

    Moon has no atmosphere, making this a completely different, and a whole lot easier problem

    You beat me to the "moon analogy" limitations, but as I go through Mars EDL sequencing for an approach like the Skycrane that Curiosity used, and then think of how it might be done with Retropropulsion, I'm really searching for more information on modeling the approach.   
    1.  When does the retropropulsion burn start?  outside or inside the mars atmosphere?
    2.  Will there be multiple burns like with current F9 similar to the out of atmosphere burns to kill velocity prior to entry?  Does this allow for a steeper entry angle, and increased accuracy?
    3.  How much of the DV budget for landing comes from using atmospheric drag vs. propulsion?
    4.  How much does the timing matter when that DV is delivered from drag?

    Edit:
    Where I am going with this, is that if your landing solution starts outside the atmosphere with a propulsive maneuver as the craft enters the atmosphere, your model of the atmosphere & corresponding DV change due to drag, can be sensed by accelerometers and compensated for with throttle response.    The point being you set boundaries on the DV shed during each timed portion of the approach, which minimizes the uncertainty going forward in time.  If your entry profile relies on first shedding some initial DV by entry drag, that uncertainty gives a big error in the landing ellipse for a ballistic trajectory. 
    « Last Edit: 09/27/2016 03:03 PM by Stan-1967 »

    Offline savuporo

    • Senior Member
    • *****
    • Posts: 5155
    • Liked: 985
    • Likes Given: 343
    Re: Landing and takeoff on unprepared surfaces
    « Reply #30 on: 09/27/2016 03:46 PM »
    Edit:
    Where I am going with this, is that if your landing solution starts outside the atmosphere with a propulsive maneuver as the craft enters the atmosphere, your model of the atmosphere & corresponding DV change due to drag, can be sensed by accelerometers and compensated for with throttle response.    The point being you set boundaries on the DV shed during each timed portion of the approach, which minimizes the uncertainty going forward in time.  If your entry profile relies on first shedding some initial DV by entry drag, that uncertainty gives a big error in the landing ellipse for a ballistic trajectory. 

    I get that. But the issue is more of actually having low latency inputs to correct the internal navigation model, rather than lack of control authority. More than anything, accurate weather model, massive ring laser gyros, and at least two surface or orbital radio beacons would help. Both surface and orbital beacons would have their own issues, radio propagation in hypersonic flight and visibility over horizon.
    Reducing the initial navigation velocity and position error would help too - this could be done by doing entry from well tracked orbit, rather than interplanetary cruise.
    Orion - the first and only manned not-too-deep-space craft

    Offline LouScheffer

    • Full Member
    • ****
    • Posts: 1910
    • Liked: 2501
    • Likes Given: 279
    Re: Landing and takeoff on unprepared surfaces
    « Reply #31 on: 09/27/2016 04:09 PM »
    Apollo achieved quite accurate (10s of meters) landings on the moon without GPS.   They used a combination of a reasonably accurate initial trajectory, corrections based on looking out the window and accurate maps, and sufficient control authority.

    This should work fine for Mars, I'd think... 

    Moon has no atmosphere, making this a completely different, and a whole lot easier problem
    Sure, but SpaceX has real, practical experience with pin-point powered landing in an atmosphere.  And the wind forces are much stronger than they will be on Mars.   So this problem reduces to knowing where you are and where you want to go.
    Quote
    EDIT: Oh, and it also happens to be slightly closer, making the initial position and velocity estimate error from tracking a few orders of magnitude smaller.
    This seems intuitively reasonable, but it's wrong.  Tracking has made huge advances since Apollo, and current near-Mars tracking is much better than Apollo era lunar range tracking.   Due to "known deficiencies in the R2 lunar (gravitational) potential model" (from Apollo 11 Lunar Surface Journal), among other factors, Apollo 11 landed 4 miles (6 km) from its target point.  In contrast, the positions of Mars orbiters and landers are known to meter-class precision.  See Planetary ephemeris DE424 for Mars Science Laboratory early cruise navigation, figures 3 and 4. So tracking at Mars today is a few orders of magnitude better than tracking at the Moon was in the time of Apollo.   Of course tracking of the Moon is much better, too - the location of the vehicles with retro-reflectors on the moon is known to about a centimeter(!). 

    So overall, tracking at the Moon now is a few orders of magnitude better than tracking at Mars now, but tracking at Mars now is a few orders of magnitude than tracking at the Moon, at the time of Apollo.

    Offline savuporo

    • Senior Member
    • *****
    • Posts: 5155
    • Liked: 985
    • Likes Given: 343
    Re: Landing and takeoff on unprepared surfaces
    « Reply #32 on: 09/27/2016 04:46 PM »
    In contrast, the positions of Mars orbiters and landers are known to meter-class precision.  ...
    So overall, tracking at the Moon now is a few orders of magnitude better than tracking at Mars now, but tracking at Mars now is a few orders of magnitude than tracking at the Moon, at the time of Apollo.
    Tracking an orbiter or a landed vehicle and tracking an vehicle on approach navigation from interplanetary cruise post-TCM maneuvers are still different problems. Time constraints.
    Of course tracking is whole lot better overall, doesn't change the fact that landing on airless body that is much closer has fundamentally different set of problems.
    Orion - the first and only manned not-too-deep-space craft

    Offline IainMcClatchie

    • Full Member
    • ***
    • Posts: 336
    • San Francisco Bay Area
    • Liked: 196
    • Likes Given: 314
    Re: Landing and takeoff on unprepared surfaces
    « Reply #33 on: 09/28/2016 12:42 AM »
    Elon's presentation shows the MCT landing such that the six vacuum nozzles are less than one nozzle diameter from the Martian surface.  I assume they'd land on the three low-expansion nozzles, and though I'm not sure what the clearance would be there, it still seems very close.

    It doesn't appear they have a plan for debris mitigation.

    The booster uses 7% of its propellant for return, which is 3729 m/s delta-V.

    MCT delta-V returning from Mars with 50 tonnes payload is 8900 m/s.  I'm not sure how this works, as they need 3800 m/s to get off the surface into low Mars orbit, 2500 m/s to get back to Earth, and at least another 3729 m/s for a propulsive Earth landing.  They are 1129 m/s short.  Maybe they are thinking about aerocapture at Earth return, followed by refuelling before landing, but Elon didn't articulate that part at all.

    Offline Lars-J

    • Senior Member
    • *****
    • Posts: 4328
    • California
    • Liked: 3773
    • Likes Given: 2305
    Re: Landing and takeoff on unprepared surfaces
    « Reply #34 on: 09/28/2016 12:44 AM »
    ...and at least another 3729 m/s for a propulsive Earth landing. ...

    Huh? Where in the world did you pull *that* number from!?!?! Did you forget earth's atmosphere?

    Offline envy887

    • Senior Member
    • *****
    • Posts: 4714
    • Liked: 2621
    • Likes Given: 1409
    Re: Landing and takeoff on unprepared surfaces
    « Reply #35 on: 09/28/2016 01:24 AM »
    ...and at least another 3729 m/s for a propulsive Earth landing. ...

    Huh? Where in the world did you pull *that* number from!?!?! Did you forget earth's atmosphere?

    Terminal velocity for ITS spaceship on Earth will be subsonic, requiring only a few hundred m/s for landing.

    The booster is doing boostback and entry burns, which use about 90% of that 3700 m/s. The spaceship does neither.

    Offline Alastor

    • Full Member
    • **
    • Posts: 272
    • Liked: 148
    • Likes Given: 330
    Re: Landing and takeoff on unprepared surfaces
    « Reply #36 on: 09/28/2016 01:16 PM »
    Elon's presentation shows the MCT landing [...] doesn't appear they have a plan for debris mitigation.

    I disagree.
    He didn't talk about debris mitigation, but that doesn't mean that there isn't one.

    My take on their plan is :
     - Red dragon missions are planned for every window starting 2020. They are going to serve as robotized exploration stations but also as meteo stations. (missions mentioned : Identifying resources available, suitable landing sites. But Long term meteo/radar/comms station seems likely). They will probably need multiple of them per window even given the announced missions (2 or 3 red dragons don't seem likely to be sufficient to both identify useful resources and identify suitable landing sites).
     - They announced that first fuel plant will be cargo of first manned fleet, but also that first crew are far from certain to return. So I don't think they expect debris mitigation in place by then, which mean that those ships are there to stay (at least for a long time). They probably will be the ones to build landing sites and habitat, so they will be with construction equipment. Maybe send an empty ship later to bring them back or more likely bring them back on unused seats in returning ships ?

    Now that I think of it, they will need many landing sites to be able to land a fleet. Even worse given that the ships will probably not return on the next window (or else most of them would return empty).

    So in retrospect, maybe it would be more beneficial to use a landing area that is far away from the habitation site and build a transportation system to go from one to the other (Hyperloop anyone ? ;D (just joking here)). That way they would not Need to worry about damaging habitations upon take-off. Only about other ships (that would be protected against micrometeorites anyway, so probably fine with a take-off nearby).

    Just thinking out loud.
    « Last Edit: 09/28/2016 01:18 PM by Alastor »

    Offline Hobbes-22

    • Full Member
    • ****
    • Posts: 434
    • Acme Engineering
      • Acme Engineering
    • Liked: 148
    • Likes Given: 95
    Re: Landing and takeoff on unprepared surfaces
    « Reply #37 on: 09/28/2016 02:22 PM »

    Now that I think of it, they will need many landing sites to be able to land a fleet. Even worse given that the ships will probably not return on the next window (or else most of them would return empty).

    Returning empty would be a lot cheaper than having to build a new one.

    Offline IainMcClatchie

    • Full Member
    • ***
    • Posts: 336
    • San Francisco Bay Area
    • Liked: 196
    • Likes Given: 314
    Re: Landing and takeoff on unprepared surfaces
    « Reply #38 on: 09/30/2016 04:48 AM »
    ...and at least another 3729 m/s for a propulsive Earth landing. ...

    Huh? Where in the world did you pull *that* number from!?!?! Did you forget earth's atmosphere?

    That's the delta-V that the first stage uses to get back to the launch point after stage separation.  I completely forgot about the boostback.  D'oh!

    That said, I'm not sure why the first stage would do an entry burn but the ITS would not.  I guess another way of saying the same thing is: why does the Falcon 9 first stage do an entry burn when Dragon does not?  If the burn is there to protect the engine bells, doesn't the ITS have the same problem?

    Offline Lars-J

    • Senior Member
    • *****
    • Posts: 4328
    • California
    • Liked: 3773
    • Likes Given: 2305
    Re: Landing and takeoff on unprepared surfaces
    « Reply #39 on: 09/30/2016 05:59 AM »
    ...and at least another 3729 m/s for a propulsive Earth landing. ...

    Huh? Where in the world did you pull *that* number from!?!?! Did you forget earth's atmosphere?

    That's the delta-V that the first stage uses to get back to the launch point after stage separation.  I completely forgot about the boostback.  D'oh!

    That said, I'm not sure why the first stage would do an entry burn but the ITS would not.  I guess another way of saying the same thing is: why does the Falcon 9 first stage do an entry burn when Dragon does not?  If the burn is there to protect the engine bells, doesn't the ITS have the same problem?

    The first stage does an entry burn because it lacks sufficient heat protection. They are using propellant (which they have) instead of covering the bottom and whole stage with PICA-X or something similar.

    The ITS already needs to support aero-capture and will have a fairly robust protection on the side during its side entry. Note in the CFD model rendering of the ITS that the "bottom" side as an extended flap that wraps almost halfway around and covers the engine bells. (See image 1 below) This is similar to how the shuttle's main body flap  protected the engines from the worst re-entry heat.
    In image 2 you can see a better view of how the engine nozzles are protected on one side. (the 'down' side during entry)
    « Last Edit: 09/30/2016 06:04 AM by Lars-J »

    Offline IainMcClatchie

    • Full Member
    • ***
    • Posts: 336
    • San Francisco Bay Area
    • Liked: 196
    • Likes Given: 314
    Re: Landing and takeoff on unprepared surfaces
    « Reply #40 on: 09/30/2016 10:13 PM »
    Boostback takes a lot of fuel.

    If the staging velocity is something like 1800 m/s (CRS-8), then it's not surprising they've reserved 3729 m/s of delta-V for getting the booster back to the launch site through boostback.

    If I assume 729 m/s delta-V is for landing, then 3000 m/s of delta-V for boostback is 410 tonnes of fuel.

    PICA-X is pretty lightweight.  You could put a very thick PICA-X shield over most of the booster for way, way less than 410 tonnes.  The thing's dry weight is just 275 tonnes as it is.  Actually, a PICA-X shield just on the bottom might do fine -- 61 kPa of dynamic pressure at Cd=1.0 is enough to support the vehicle, so dynamic pressures around two atmospheres would be plenty for aerobraking.

    My guess is that a booster which used a barge landing could loft the same second stage to orbit with maybe 5 or 10 fewer engines.  I think SpaceX is optimizing for a very fast launch cadence.  They're assuming that if it took a week to get the booster back, they'd need twice as many of them, and it's cheaper to make each one a bit bigger.  I suppose the big booster also makes it possible to get a bigger max payload up if you are willing to wait for the next launch and have a barge handy.

    It's a big vision and a big bet.

    Online Bob Shaw

    • Full Member
    • ****
    • Posts: 1083
    • Liked: 446
    • Likes Given: 388
    Re: Landing and takeoff on unprepared surfaces
    « Reply #41 on: 09/30/2016 10:35 PM »
    Due to "known deficiencies in the R2 lunar (gravitational) potential model" (from Apollo 11 Lunar Surface Journal), among other factors, Apollo 11 landed 4 miles (6 km) from its target point

    There has been a further analysis of the Apollo 11 data, which shows that procedural errors in the way that the LM undocked (specifically, propulsive effects from residual atmosphere in the CM/LM Docking Tunnel) played as big a part in the landing site drift as anything else. Procedures were changed in later missions, as well as improvements in knowledge of the Lunar geoid and mascon distribution which helped navigational accuracy too.
    « Last Edit: 09/30/2016 10:36 PM by Bob Shaw »

    Offline Kansan52

    • Full Member
    • ****
    • Posts: 1150
    • Hutchinson, KS
    • Liked: 359
    • Likes Given: 411
    Re: Landing and takeoff on unprepared surfaces
    « Reply #42 on: 09/30/2016 10:56 PM »
    Bob, could you point me to the material the talks about residual atmosphere between the CM and LM? I heard the story about that 'gust' but have never seen the original material. Would love to read that.

    Offline meekGee

    • Senior Member
    • *****
    • Posts: 8669
    • N. California
    • Liked: 4840
    • Likes Given: 893
    Re: Landing and takeoff on unprepared surfaces
    « Reply #43 on: 10/01/2016 03:34 PM »
    GPS isn't the only possible navigation system. Terrain following or rf beacon homing could allow sub 100m navigation accuracy, perhaps even sub meter accuracy.
    Read the links above ( and we have other threads about this here ).
    There are many challenges implementing better position estimation systems, and unfortunately the biggest error accumulation happens in hypersonic phase of flight where the lander is basically blind and deaf.
    INS can take care of that phase of flight.

    SoaceX will have a GPS like system, either ground based or satellite based.

    There will be a lot of surface vehicle activity as part of normal life there, and these vehicles must know how to navigate.
    ABCD - Always Be Counting Down

    Offline LouScheffer

    • Full Member
    • ****
    • Posts: 1910
    • Liked: 2501
    • Likes Given: 279
    Re: Landing and takeoff on unprepared surfaces
    « Reply #44 on: 10/01/2016 08:50 PM »
    Bob, could you point me to the material the talks about residual atmosphere between the CM and LM? I heard the story about that 'gust' but have never seen the original material. Would love to read that.

    Here is some more from Apollo 11 Lunar surface journal that might at least point to the original:

    Quote
    [Journal Contributor Ron Wells calls attention to discussions in Gene Kranz's Failure is Not an Option which attribute the positional error at PDI to residual pressure in the tunnel between the LM and CSM at undocking at 100:12:00. In a 2002 e-mail, Kranz elaborates, "Floyd's note was correct on the velocity-induced position error at the start of the descent. There were several interrelated navigation problems, i.e. known deficiencies in the R2 lunar (gravitational) potential model, down-track (along the flight path toward the landing site) and cross-track (left or right, perpendicular to the flight path) propagation errors (errors that start out small but become larger as the flight proceeds), and errors induced by maneuvering of the spacecraft. The principal error induced by maneuvering of the spacecraft was, however, the incomplete vent of the tunnel propagated over one orbit after separation. We made a change in all future missions to get a MCC go-nogo on tunnel delta P before giving the crew a Go to undock. Page 82 of the Apollo 11 post mission report says ....'because of uncoupled attitude maneuvers such as hot fire tests, undocking impulse, station keeping, sublimator operation and possible tunnel and cabin venting. The net effect of these perturbations was a sizeable down-range miss.' To my recollection, the trajectory reconstruction determined that with the exception of the tunnel venting, most of the other perturbations were essentially self canceling. Further the post mission review indicated that the delta P gauge was too gross, the markings misleading and the tunnel had to be vented earlier in the timeline and the valve left in the tunnel vent position rather than returned to off."]

    Tags: