SpaceX Dragon 2 Updates and Discussion - Thread 3

[fast]

Going by the mantra "the best part is no part" and the conviction of the worst error an engineer can make is "try to perfect something that is not needed" recently expressed by Mr. Musk, I believe the simplest solution is to revert to the tried and tested three chute design and dump the hipergolic fuels prior to landing to get rid of the excess mass.

Probably things would be much more simple if after dropping propulsive landing option, superdracos and its tanks were moved to the trunk. Or only tanks...

[woods170]

Going by the mantra "the best part is no part" and the conviction of the worst error an engineer can make is "try to perfect something that is not needed" recently expressed by Mr. Musk, I believe the simplest solution is to revert to the tried and tested three chute design and dump the hipergolic fuels prior to landing to get rid of the excess mass.

Probably things would be much more simple if after dropping propulsive landing option, superdracos and its tanks were moved to the trunk. Or only tanks...

NASA accepted SpaceX's proposal voor CCP which had the LAS integrated into the capsule. NASA knew what it was getting into from Day 1. If they didn't want the LAS integral to the capsule they should have not picked SpaceX's proposal.

Yet they did.

[NewSpaceIsFun]

Based on Bridenstine's recent comments on Dragon certification/qualification (LAS, Parachutes), what's the current odds on SH+SS taking someone to orbit and returning before Dragon takes a crew up? Maybe I've missed it, but has SpaceX mentioned sub-orbital crew launches prior to full orbital testing?

[ValmirGP]

Going by the mantra "the best part is no part" and the conviction of the worst error an engineer can make is "try to perfect something that is not needed" recently expressed by Mr. Musk, I believe the simplest solution is to revert to the tried and tested three chute design and dump the hipergolic fuels prior to landing to get rid of the excess mass.

Then you have awesome amounts of fresh hypergol contamination on the capsule's surface just prior to landing. Without taking into account the non-negligible risk that system doesn't work (or works at the wrong time) and you end up overweight under 3 chutes.

Those awesome amounts of hypergol contamination would be highly dependent on how you discard it in orbit, I believe. Also, the capsule will proceed to have a fresh plasma bath followed by a good splash of salt water prior to anyone touching it. I still think it would be a faster/simpler path.

[thirtyone]

What failure in the parachute system was there during DM-1?

I spent some time looking into parachute design earlier (maybe in another thread). Basically, highly reefed, multi-parachute designs on this scale have never been tested before, and industry-wide assumptions about modeling parachutes were incorrect. Orion, Starliner, and Crew Dragon are *all* potentially affected by this, because they use multiple highly reefed parachutes. A recent GAO report notes that both programs are plagued by parachute issues - I suspect Boeing has been more capable of keeping their problems under wraps.

Remember, guaranteeing a 1 in 270 chance of failure without actually testing the spacecraft 270 times means a lot of accurate modeling and assumptions are necessary. Even if a parachute *does not* fail during all testing, if one of those assumptions is incorrect, the safety of the craft is entirely at risk.

There is a magic multiplier for the fact that the load on each parachute rope / element is not perfectly distributed - and a factor of 1.1 was considered sufficient for all previous designs. Evidently that number can be up to ~1.9 in these cases - easily blowing through any safety margin originally put into the design. There are rumors that during a parachute-out test, the Crew Dragon parachute system failed entirely - which would generally point towards the fact that some part of the modeling is incorrect. I believe that is how they discovered that this industry-wide standard factor was outright invalid.

The great thing is you can actually see whether or not a parachute system is susceptible to this particular asymmetric loading. Basically, while the main chutes are opening, if one or more of them does not open with a mostly circular cross section, the loading on that parachute's lines may be extremely asymmetric. If you look at Starliner's original parachute deployment videos, you'll note that one of the chutes tends to open in a very obvious oval, so I suspect that even if their testing was successful, they are at risk of blowing through their design safety margins as well. We can't really know unless we actually look at the numbers. The funny thing is it looks like someone deliberately cut footage of the reefing stages in Boeing's more recent Starliner parachute video. I'm sure it was just to keep the video short, but part of me can't help but wonder...

tl;dr No one could have realized that multi-chute systems were this difficult. The safety of all three craft was in fact compromised to some degree by incorrect modeling, and these failures on Crew Dragon helped everyone realize this. We can't really predict if the compromise was enough to bring any of craft over the line without exact numbers. Fixes probably involve increasing line strength (which mean more weight)

[eeergo]

Going by the mantra "the best part is no part" and the conviction of the worst error an engineer can make is "try to perfect something that is not needed" recently expressed by Mr. Musk, I believe the simplest solution is to revert to the tried and tested three chute design and dump the hipergolic fuels prior to landing to get rid of the excess mass.

Then you have awesome amounts of fresh hypergol contamination on the capsule's surface just prior to landing. Without taking into account the non-negligible risk that system doesn't work (or works at the wrong time) and you end up overweight under 3 chutes.

Those awesome amounts of hypergol contamination would be highly dependent on how you discard it in orbit, I believe. Also, the capsule will proceed to have a fresh plasma bath followed by a good splash of salt water prior to anyone touching it. I still think it would be a faster/simpler path.

Ah, I read the "just prior to landing" as JUST prior to landing, i.e. after reentry.

I'm not sure how feasible it is to vent that large amount of propellants with the current system, or conversely adapt it to be able to do it - it would be a necessary operation on every flight - and as mentioned in the second part of my reply, would *absolutely* have to work or landing won't be possible, and only when needed. And still, residues coming off the vented tanks once the crew has to get out can be *very* damaging (see ASTP Apollo capsule), no need to touch it.

Probably still better off trying to iron out the issues with the chutes, which has the plus side of learning something new and useful (see the great post above this one).

[Comga]

How many parachutes does Soyuz have? How were they modeled? Oh wait...

Crew Dragon is several times heavier than the Soyuz descent module.

And Soyuz has retro-rockets to soften the higher speed landing from using one chute.
If only Dragon had some rockets that could slow the terminal descent.....

As to dumping the fuel, much of it could be burnt off using the Dracos in a manner that moves the landing point without changing the reentry velocity.  This could be done with a few Dracos perpendicular to the velocity vector or by firing opposing Dracos.  If this was done before reentry the capsule should be pretty well decontaminated after reentry. 

How long would the Dracos have to fire to dump all of the fuel allotted to the Super Dracos for the launch abort?

[woods170]

How many parachutes does Soyuz have? How were they modeled? Oh wait...

Crew Dragon is several times heavier than the Soyuz descent module.

And Soyuz has retro-rockets to soften the higher speed landing from using one chute.
If only Dragon had some rockets that could slow the terminal descent.....

As to dumping the fuel, much of it could be burnt off using the Dracos in a manner that moves the landing point without changing the reentry velocity.  This could be done with a few Dracos perpendicular to the velocity vector or by firing opposing Dracos.  If this was done before reentry the capsule should be pretty well decontaminated after reentry. 

How long would the Dracos have to fire to dump all of the fuel allotted to the Super Dracos for the launch abort?

Not feasible on short notice. NASA raised a Mount Everest sized pile of requirements for any attempt to fire the SuperDracos during landing. And those apply even to "just" burning off the fuel.

[ValmirGP]

Going by the mantra "the best part is no part" and the conviction of the worst error an engineer can make is "try to perfect something that is not needed" recently expressed by Mr. Musk, I believe the simplest solution is to revert to the tried and tested three chute design and dump the hipergolic fuels prior to landing to get rid of the excess mass.

Then you have awesome amounts of fresh hypergol contamination on the capsule's surface just prior to landing. Without taking into account the non-negligible risk that system doesn't work (or works at the wrong time) and you end up overweight under 3 chutes.

Those awesome amounts of hypergol contamination would be highly dependent on how you discard it in orbit, I believe. Also, the capsule will proceed to have a fresh plasma bath followed by a good splash of salt water prior to anyone touching it. I still think it would be a faster/simpler path.

Ah, I read the "just prior to landing" as JUST prior to landing, i.e. after reentry.

I'm not sure how feasible it is to vent that large amount of propellants with the current system, or conversely adapt it to be able to do it - it would be a necessary operation on every flight - and as mentioned in the second part of my reply, would *absolutely* have to work or landing won't be possible, and only when needed. And still, residues coming off the vented tanks once the crew has to get out can be *very* damaging (see ASTP Apollo capsule), no need to touch it.

Probably still better off trying to iron out the issues with the chutes, which has the plus side of learning something new and useful (see the great post above this one).

Sorry if I have not made myself clear. English is not my primary idiom (though I am sure is not the one I am worst at).
I agree with you that it's better to iron out the issues and retain a possible fail safe of firing the SuperDracos at the very last moment if all else goes wrong.
My comment was in relation to the issue of delays in CCP. If this issue is to become a scapegoat to delay Crew Dragon, go to the faster solution instead.
You could possibly even do both things.

[speedevil]

I'm not sure how feasible it is to vent that large amount of propellants with the current system, or conversely adapt it to be able to do it - it would be a necessary operation on every flight - and as mentioned in the second part of my reply, would *absolutely* have to work or landing won't be possible, and only when needed. And still, residues coming off the vented tanks once the crew has to get out can be *very* damaging (see ASTP Apollo capsule), no need to touch it.

If the concern is overweight landing - and the superdracos are not used for OMS (That would be insane), there are plenty of pairs of draco engines that can be fired without net propulsive effect, for no longer per mission than is normally done for the nominal rendevous/deorbit burns, and sum up to the whole amount of propellant an abort would normally use.

[Comga]

How many parachutes does Soyuz have? How were they modeled? Oh wait...

Crew Dragon is several times heavier than the Soyuz descent module.

And Soyuz has retro-rockets to soften the higher speed landing from using one chute.
If only Dragon had some rockets that could slow the terminal descent.....

As to dumping the fuel, much of it could be burnt off using the Dracos in a manner that moves the landing point without changing the reentry velocity.  This could be done with a few Dracos perpendicular to the velocity vector or by firing opposing Dracos.  If this was done before reentry the capsule should be pretty well decontaminated after reentry. 

How long would the Dracos have to fire to dump all of the fuel allotted to the Super Dracos for the launch abort?

Not feasible on short notice. NASA raised a Mount Everest sized pile of requirements for any attempt to fire the SuperDracos during landing. And those apply even to "just" burning off the fuel.

I think you missed my point
There was no suggestion for firing the Super Dracos, not during descent, not on orbit.
The question was could SpaceX burn off the bulk of the propellants using the regular Draco engines, after departing the ISS?  This could be before, after, or as part of the deobit burn.

[woods170]

How many parachutes does Soyuz have? How were they modeled? Oh wait...

Crew Dragon is several times heavier than the Soyuz descent module.

And Soyuz has retro-rockets to soften the higher speed landing from using one chute.
If only Dragon had some rockets that could slow the terminal descent.....

As to dumping the fuel, much of it could be burnt off using the Dracos in a manner that moves the landing point without changing the reentry velocity.  This could be done with a few Dracos perpendicular to the velocity vector or by firing opposing Dracos.  If this was done before reentry the capsule should be pretty well decontaminated after reentry. 

How long would the Dracos have to fire to dump all of the fuel allotted to the Super Dracos for the launch abort?

Not feasible on short notice. NASA raised a Mount Everest sized pile of requirements for any attempt to fire the SuperDracos during landing. And those apply even to "just" burning off the fuel.

I think you missed my point
There was no suggestion for firing the Super Dracos, not during descent, not on orbit.
The question was could SpaceX burn off the bulk of the propellants using the regular Draco engines, after departing the ISS?  This could be before, after, or as part of the deobit burn.

Given the amount of propellant on-board remaining, after the deorbit burn, it would require very lengthy Draco burns. Substantially more lengthy than the deorbit burn. I don't think the Dracos are qualified for that. And probably neither is the Crew Dragon primary structure.

If burning off the remaining propellant - via the Dracos - had been the easy solution than SpaceX probably would have taken that route already.
The fact that they haven't is telling.

[Nomadd]

 Has the possibility of using the Super Dracos for a chute failure scenario been discounted? It seems like a little software could make all the difference.

[kevinof]

I think it has. This was discussed a lot on NSF over the last year or so - Would require accurate altimeter plus new software, much testing and oh lots and lots of paperwork.

I think at this point, Spacex just wants to get D2 operational and then park development of it.
 

Has the possibility of using the Super Dracos for a chute failure scenario been discounted? It seems like a little software could make all the difference.

[eeergo]

I'm not sure how feasible it is to vent that large amount of propellants with the current system, or conversely adapt it to be able to do it - it would be a necessary operation on every flight - and as mentioned in the second part of my reply, would *absolutely* have to work or landing won't be possible, and only when needed. And still, residues coming off the vented tanks once the crew has to get out can be *very* damaging (see ASTP Apollo capsule), no need to touch it.

If the concern is overweight landing - and the superdracos are not used for OMS (That would be insane), there are plenty of pairs of draco engines that can be fired without net propulsive effect, for no longer per mission than is normally done for the nominal rendevous/deorbit burns, and sum up to the whole amount of propellant an abort would normally use.

I believe the hypergol tanks are not common to Dracos and Super Dracos (although I could be wrong). Either way:
- If they are not common, you cannot do this.
- If they are common, you have to redesign the system in such a way to be damn sure you don't inadvertently vent it in space (rendering the spacecraft unable to maneuver, including deorbit), but still be damn sure it works when it has to (or, being overweight, you cannot land under 3 parachutes). That's certainly not trivial, and probably more complicated / time consuming / expensive than finding out how to solve the 4 parachute problem - which OTOH will have the side effect of advancing the state of the art in parachute design.

[tyrred]

I'm not sure how feasible it is to vent that large amount of propellants with the current system, or conversely adapt it to be able to do it - it would be a necessary operation on every flight - and as mentioned in the second part of my reply, would *absolutely* have to work or landing won't be possible, and only when needed. And still, residues coming off the vented tanks once the crew has to get out can be *very* damaging (see ASTP Apollo capsule), no need to touch it.

If the concern is overweight landing - and the superdracos are not used for OMS (That would be insane), there are plenty of pairs of draco engines that can be fired without net propulsive effect, for no longer per mission than is normally done for the nominal rendevous/deorbit burns, and sum up to the whole amount of propellant an abort would normally use.

I believe the hypergol tanks are not common to Dracos and Super Dracos (although I could be wrong). Either way:
- If they are not common, you cannot do this.
- If they are common, you have to redesign the system in such a way to be damn sure you don't inadvertently vent it in space (rendering the spacecraft unable to maneuver, including deorbit), but still be damn sure it works when it has to (or, being overweight, you cannot land under 3 parachutes). That's certainly not trivial, and probably more complicated / time consuming / expensive than finding out how to solve the 4 parachute problem - which OTOH will have the side effect of advancing the state of the art in parachute design.

Common tanks, from what I gather in muskology. 

Advancing the SOA in parachute design... yes, please!

[woods170]

I think it has. This was discussed a lot on NSF over the last year or so - Would require accurate altimeter plus new software, much testing and oh lots and lots of paperwork.

I think at this point, Spacex just wants to get D2 operational and then park development of it.
 

Has the possibility of using the Super Dracos for a chute failure scenario been discounted? It seems like a little software could make all the difference.

Fully correct analysis. Use of Super Dracos in case of parachute failure runs into exactly the same Mount Everest-sized stack of requirements as use of Super Dracos for propulsive landing.
Hence why it is the path not followed by SpaceX.

Right now SpaceX is doing its very best to get Crew Dragon flying. Once that it is done they can leave it behind them and use the gained knowledge and experience for the next thing (and we all know what that next thing is).

[savantu]

Seems like acrimony gets to a whole new level.

https://www.theatlantic.com/science/archive/2019/10/elon-musk-jim-bridenstine-starship-commercial-crew/599218/

Koren: After the big presentation on Saturday, a reporter asked Elon to respond to your tweet. Elon said, “Did he say Commercial Crew or SLS?” [SLS stands for the Space Launch System, the NASA capsule and rocket that is meant to deliver astronauts to the moon.] What do you make of that?

Bridenstine: Well, I don’t think that’s helpful. Commercial Crew is about getting to low-Earth orbit. We are spending $85 million every time we have to buy a Russian Soyuz seat to get to the International Space Station.

SLS, that’s a whole different mission. SLS is going to the moon. So I don’t know why you would compare the two. But certainly SLS is behind schedule and over cost. We want them to get back on cost and schedule as well. We are holding them accountable just as much as anybody else.

and the timeline for flight is totally open

Koren: Musk said this weekend that SpaceX would be ready to fly NASA astronauts within three to four months. Does that sound realistic to you?

Bridenstine: No. They have redesigned their launch-abort system, and with that redesign, [the system] has to be qualified. We are lucky that the explosion happened … during a test. If that wouldn’t have happened, we would be taking a lot more risk that we would not be aware of right now. But now that we have a new design, it needs to be tested; it needs to be qualified.

And that’s not the hardest problem. The hardest problem is the parachutes. We do not have the margin of safety [that NASA requires] in the parachutes, and that’s going to take probably more time to resolve than the launch-abort system.

[speedevil]

Given the amount of propellant on-board remaining, after the deorbit burn, it would require very lengthy Draco burns. Substantially more lengthy than the deorbit burn. I don't think the Dracos are qualified for that. And probably neither is the Crew Dragon primary structure.

If burning off the remaining propellant - via the Dracos - had been the easy solution than SpaceX probably would have taken that route already.
The fact that they haven't is telling.

Deorbit burn from 408 to 120km is 84m/s.
The superdracos with a max weight capsule and full tanks do around 330m/s.
Assuming only that all dracos are identical, and that they use only one axes to do that deorbit burn, you have at least six times that that total delta-v = ~500m/s.

[woods170]

Given the amount of propellant on-board remaining, after the deorbit burn, it would require very lengthy Draco burns. Substantially more lengthy than the deorbit burn. I don't think the Dracos are qualified for that. And probably neither is the Crew Dragon primary structure.

If burning off the remaining propellant - via the Dracos - had been the easy solution than SpaceX probably would have taken that route already.
The fact that they haven't is telling.

Deorbit burn from 408 to 120km is 84m/s.
The superdracos with a max weight capsule and full tanks do around 330m/s.
Assuming only that all dracos are identical, and that they use only one axes to do that deorbit burn, you have at least six times that that total delta-v = ~500m/s.

Deorbit burn alone lasts roughly 15 minutes on Crew Dragon (going by the numbers for DM-1, as provided by SpaceX in their press kit). So if I understand you correctly Crew Dragon would have to do a single-axis Draco burn lasting 1.5 hours (6 x 15 minutes = 90 minutes = 1.5 hours)  to burn off the excess propellant.
If that burn was initiated as the deorbit burn the Dracos would still be burning for half-an-hour AFTER splashdown. Which makes this whole scenario moot.


If SpaceX is not allowed to use the SuperDracos to burn off the excess propellant (which is the current situation) than the propellant will have to come back down, all the way to splashdown. Which is why that fourth parachute is there.