Speculation into what future versions of Dragon will involve

[solartear]

There will probably be at least 3 legs.  I would assume exactly 3.

Musk has said they plan to have future grasshopper have 5 legs for greater stability. I would assume the reasoning also applies to the propulsive landing Dragon. It makes sense to have 5 in case 1 leg fails, or even 2 legs if they're opposite sides.

[Dave G]

There will probably be at least 3 legs.  I would assume exactly 3.

Musk has said they plan to have future grasshopper have 5 legs for greater stability. I would assume the reasoning also applies to the propulsive landing Dragon. It makes sense to have 5 in case 1 leg fails, or even 2 legs if they're opposite sides.

For Dragon, each leg has to somehow fit in with the heat shield, so from that perspective, the fewer legs the better.

Also remember grasshopper is really tall, skinny, and doesn't have much lateral control, so it seems appropriate that the first stage may require more legs.

In any case, the legs are primarily for preventing damage to the vehicle.  For example, if a Dragon leg breaks during landing, then I doubt anyone would be killed, so you wouldn't have loss of crew or loss of mission, and its even possible that Dragon could be refurbished. My point is you probably don't need much in terms of margins or redundancy for the legs.

[mlindner]

My prediction is that the super dracos and landing legs will be external and mounted on pivots, attached just above the heatshield. They will be in the rearward position for launch. For reentry they will pivot forward and be partially covered with PICA-X. Once subsonic, they will pivot down again to use the super dracos to slow descent and then extend the legs. Latest SpaceX simulation shows parachutes almost all the way down. If the super dracos could be explosively separated in case of malfunction the current chutes would be able to land the capsule portion safely.

One of the chief SpaceX design philosophies is, from what I've seen, KISS (keep it simple stupid). SuperDracos won't be deployable, and if they can make the landing legs also non-deployable then I think they will.

[Norm38]

***WARNING:  Electrical Engineer speculating outside his field***

What I'm considering isn't anything short term, if it would even still be called Dragon.  Listening to Musk's talks, he keeps mentioning full resuablity, and getting as close to gas-n-go operation as possible.  Having the F9 uppperstage split into second stage, capsule and trunk doesn't fit the bill.  Then there's the mention of the methane Raptor engine and the "MCT spaceship" it's attached to.

For full reusability, having a second stage that already makes orbit land separately from the capsule is a huge duplication of effort and recovery hardware mass.  Both the second stage and capsule need heat shields, RCS, legs, etc.

If the methane SC engine is sucessful as a Merlin 1D replacement for the upper stage, then can the ISP improvement from 304sec to ~380sec allow the upperstage tank and engine to be integrated with the capsule into one piece?  The empty tank makes the vehicle "fluffier" for less reentry heating, and if structural, gives mounting points for superdracos and other subsystems.

I'm picturing the MCT as a single piece, fully resuable upperstage that starts life as a LEO taxi.  But as the Dragon was designed around maybe being able to land on Mars, the MCT would be designed to be able to land on Mars and then use it's ISRU refueled Raptor engine to take off again.

Anyway, that's my MCT theory.

[Elmar Moelzer]

I have wondered the same thing, Norm. One of the biggest issues I have with this is that the Superdracos would (at least IMHO) not be strong enough for a successful launch abort if the second stage and capsule were to be combined. I might be wrong though and they might just add more to the second stage (which already needs them for propulsive landing), but then how much would be left for them to gain from combining them?

[ciscosdad]

I like the idea of integrating the upper stage with Dragon, but the abort capability is a problem with a full methane/lox stage attached.
Maybe it could be more like the Dual Axis Lunar lander, and the dragon and Upper stage are mated and only separate in a launch abort scenario.
If the shape was more like a sled with crewed section at front and top access for cargo and solar panels and radiator on the bay doors a la shuttle. Big propellent tank and rear facing main engine.
I admit its a bit further down the track than we have been discussing, but maybe it has possibilities.

[Elmar Moelzer]

But if you were to have the capsule detachable from the second stage, even if only for an abort you once again have to double many of the systems and you probably wont end up gaining much (though I may be wrong there).

[ciscosdad]

You only have to separate the crew capsule early in the launch process. At some point it becomes possible to abort with the whole upper stage (Ref Apollo abort modes.)

I would visualise the same basic arrangement as the DTAL but with aerodynamic and heat shielded hull. There may not even be any need for a shield in the separation plane (pure speculation. I am no rocket scientist as is probably obvious).

[robert_d]

But the level of complexity they are willing to accept was already shown to be substantially greater than the current version with the landing legs deploying through the heatshield. Two of Elon's comments struck me - he said the four sets of superdracos would not be exactly perpendicular to each other but oriented slightly to the "windward and leeward" sides. Also that they could develop a minimum 6 G acceleration at any point of abort. I buy the KISS argument in the sense that he does not want to change the design of the pressure vessel at this point - he and NASA are happy with the current version, and he won' give up commonality with the cargo version.  Also he said it "looks cool" For a space geek like Musk, what could be cooler than something that has moving pieces and would look like a Star Wars X-wing?

My prediction is that the super dracos and landing legs will be external and mounted on pivots...

One of the chief SpaceX design philosophies is, from what I've seen, KISS (keep it simple stupid). SuperDracos won't be deployable, and if they can make the landing legs also non-deployable then I think they will.

[meekGee]

You only have to separate the crew capsule early in the launch process. At some point it becomes possible to abort with the whole upper stage (Ref Apollo abort modes.)

I would visualise the same basic arrangement as the DTAL but with aerodynamic and heat shielded hull. There may not even be any need for a shield in the separation plane (pure speculation. I am no rocket scientist as is probably obvious).

I really don't like that - it doesn't give an answer if the second stage is failing - you need to return with the capsule from an almost-orbital speed in that case.

[Lars_J]

A merged upper stage and Dragon will only make sense once there is a significant flight history to prove that aborts at launch are only a statistical anomaly and can be ignored.

Otherwise you are back to creating a separating cabin and/or ejection seats, and you forfeit the advantage of a merged unit anyway.

[Norm38]

^^^  Maybe, but abort capability at every single point doesn't really exist.  Say crewed Dragon is a reality, and a capsule is coming down for a propulsive landing.  If one of the superdraco engines were to explode, then what?  Do we double the size of Dragon to add an escape pod?  And watch payload fall to zero?

Single points of failure will always exist with any vehicle.  Personally I'd rather just have a solid vehicle with N+1 engine redundancy rather than fly in a vehicle designed to break apart into little pieces in the event one piece explodes.

Look, I understand the desire for abort modes, but I always stop before the paranoia stage.  Here in Chicagoland a father and daughter just died in a single engine plane crash.  No abort mode for them, and no one is going to start redesigning Cesnas to add ejection seats.  So why the double standard?

Subsystem redundancy?  Fine.  But airframe redundnacy?  Having each spacecraft be several completely independent spacecraft bolted together?  We don't have the mass allowance for that.

[Elmar Moelzer]

I agree with Lars_J and meekGee. As attractive as merging the capsule and the upper stage is from a cost perspective, as unattractive it is from a savety perspective...

[Elmar Moelzer]

If one of the superdraco engines were to explode, then what?  Do we double the size of Dragon to add an escape pod?

I think that they have engine out capability for the dragons on propulsive landing, at least they will have parachutes as a fallback.

[Norm38]

^^^So have engine out capability on an integrated second stage.  If the current Dragon second stage fuel tank were to explode, aborts are probably gone anyway.  The capsule would never clear the shockwave.  Design for reliablity at the subsystem level.  But if the airframe fails, well sorry everyone dies, just like in a plane crash.

The colonization of Mars is not going to have a zero fatality rate.  Can't we accept that and do good design without the paranoia?

[meekGee]

^^^  Maybe, but abort capability at every single point doesn't really exist.  Say crewed Dragon is a reality, and a capsule is coming down for a propulsive landing.  If one of the superdraco engines were to explode, then what?  Do we double the size of Dragon to add an escape pod?  And watch payload fall to zero?

Single points of failure will always exist with any vehicle.  Personally I'd rather just have a solid vehicle with N+1 engine redundancy rather than fly in a vehicle designed to break apart into little pieces in the event one piece explodes.

Look, I understand the desire for abort modes, but I always stop before the paranoia stage.  Here in Chicagoland a father and daughter just died in a single engine plane crash.  No abort mode for them, and no one is going to start redesigning Cesnas to add ejection seats.  So why the double standard?

Subsystem redundancy?  Fine.  But airframe redundnacy?  Having each spacecraft be several completely independent spacecraft bolted together?  We don't have the mass allowance for that.

You are correct, but I think in the near term, a good assumption is that loss-of-thrust on a pump-fed engine is a likely scenario and you need to assume it will happen.  At a second level, a RUD, and loss of thrust on pressure fed engines.

IIRC, on Dragon, the SD's are paired, and they might even carry a parachute.

But yeah, people will die and we need to accept that.

[Elmar Moelzer]

^^^So have engine out capability on an integrated second stage.  If the current Dragon second stage fuel tank were to explode, aborts are probably gone anyway.  The capsule would never clear the shockwave.  Design for reliablity at the subsystem level.  But if the airframe fails, well sorry everyone dies, just like in a plane crash.

Lots of assumptions there. Got data to back them up?

The colonization of Mars is not going to have a zero fatality rate.  Can't we accept that and do good design without the paranoia?

I fully agree with that notion.  If you dont like risk, dont do risky stuff like going to space and mars. Unfortunately nowadays our society has become unwilling to take risks, even to accept others taking risks.

[Lars_J]

Accepting risks does not mean one has to be foolish about it.

In fact, I would argue that a robust abort scenarios allow you to take *more* risks - or at least manage them better.

[Norm38]

Lots of assumptions there. Got data to back them up?

This isn't my field, so no.  I would love to be proven wrong and see a simulation showing the Dragon capsule out-accelerating and surviving the explosion of the second stage tank without warning.  I just don't see it happening, and I consider that a single point of failure.

As with the super dracos.  I didn't say if an SD failed, I said exploded, as in ripped out a huge chunck of the hull.  And for sake of argument, at an altitude too low for parachutes.  Again, single point of failure.

Unfortunately nowadays our society has become unwilling to take risks, even to accept others taking risks.

That's not entirely true.  People still go mountain climbing and fly single engine planes, and no one much cares.  People are allowed to be as reckless as they want with their lives and their money.  But when government funded spaceflight fails, what happens is taxpayers say "Hey, the gov't could have wasted that money on ME instead."  Unfortunately that part is not soon likely to change.

Accepting risks does not mean one has to be foolish about it.

Where am I being foolish?  There will still be plenty of abort modes and redundancies, and still way more than offered to 747 passengers.  But a higher level of integration, and a more robust vehicle overall may be worth the loss of an abort mode that may not even really work.  Especially if it lowers costs, opens markets and increases our capabilities.

747s don't fly at 10,000ft with every passenger in a parachute ready to jump.  There are tradeoffs that have to be made sometimes.

[Elmar Moelzer]

I said exploded, as in ripped out a huge chunck of the hull

Can this even happen?
We saw a first stage engine RUD during the last F9 mission and it did not result in any major damage to the stage. I would be surprised if the superdracos were not equally shielded from the rest of the craft to prevent something like that.