Author Topic: Pushing the limits of the hoverslam landing  (Read 66792 times)

Offline envy887

  • Senior Member
  • *****
  • Posts: 8150
  • Liked: 6816
  • Likes Given: 2965
Re: Pushing the limits of the hoverslam landing
« Reply #80 on: 05/09/2016 03:23 am »
The numbers are up the thread... 2 engines saves about 40% and three engines about 47% of landing fuel, but sacrifices a lot of control.

According to Musk's tweet, the landing thrust goes down to 40% of a single Merlin. That would be about 16% of a single Raptor.

That's either a super sporty landing, or an incredibly throttleable engine. But S1 Raptor will probably be overexpanded at sea level. Over expanded engines and super deep throttling don't mix well.

Landing a F9 S1 on a Raptor would be quite challenging indeed.




Offline gin455res

  • Full Member
  • ****
  • Posts: 511
  • bristol, uk
  • Liked: 42
  • Likes Given: 73
Re: Pushing the limits of the hoverslam landing
« Reply #81 on: 05/09/2016 05:36 pm »
The numbers are up the thread... 2 engines saves about 40% and three engines about 47% of landing fuel, but sacrifices a lot of control.

According to Musk's tweet, the landing thrust goes down to 40% of a single Merlin. That would be about 16% of a single Raptor.

That's either a super sporty landing, or an incredibly throttleable engine. But S1 Raptor will probably be overexpanded at sea level. Over expanded engines and super deep throttling don't mix well.

Landing a F9 S1 on a Raptor would be quite challenging indeed.





Not sure if you are responding to my question about a falcon-5R? 

As you first mention landing on 2 engines which is approximately 1.8 (the ratio of the  T/W of landing a Falcon-5 on one engine with : the T/W of landing a F9 on one engine), but then mention landing on a raptor. The falcon-5R I'm imagining is a 5 merlin falcon rocket.  The 'R'stands for reusable not Raptor (It would only be able to throttle down to 72% {F9 equivalent} at the end of the landing burn.)

Offline CraigLieb

  • Full Member
  • ****
  • Posts: 1195
  • Dallas Fort Worth
  • Liked: 1349
  • Likes Given: 2408
Re: Pushing the limits of the hoverslam landing
« Reply #82 on: 05/09/2016 06:01 pm »
and if you need a distraction from how pushing the limits of hoverslam landings effect inventory...

New poll how many stages will they have to store by Dec 31st 2016!
http://forum.nasaspaceflight.com/index.php?topic=40267.0

(a bit of shameless self promotion again)
On the ground floor of the National Space Foundation... Colonize Mars!

Offline envy887

  • Senior Member
  • *****
  • Posts: 8150
  • Liked: 6816
  • Likes Given: 2965
Re: Pushing the limits of the hoverslam landing
« Reply #83 on: 05/09/2016 06:25 pm »
The numbers are up the thread... 2 engines saves about 40% and three engines about 47% of landing fuel, but sacrifices a lot of control.

According to Musk's tweet, the landing thrust goes down to 40% of a single Merlin. That would be about 16% of a single Raptor.

That's either a super sporty landing, or an incredibly throttleable engine. But S1 Raptor will probably be overexpanded at sea level. Over expanded engines and super deep throttling don't mix well.

Landing a F9 S1 on a Raptor would be quite challenging indeed.

Not sure if you are responding to my question about a falcon-5R? 

As you first mention landing on 2 engines which is approximately 1.8 (the ratio of the  T/W of landing a Falcon-5 on one engine with : the T/W of landing a F9 on one engine), but then mention landing on a raptor. The falcon-5R I'm imagining is a 5 merlin falcon rocket.  The 'R'stands for reusable not Raptor (It would only be able to throttle down to 72% {F9 equivalent} at the end of the landing burn.)

I did think you were referring to a Raptor... :)

Is there any basis to think the Falcon-5 is ever coming back? My understanding was that all F1 and F5 payloads were permanently moved to F9 launches. F5 LEO payloads maxed at 4000 kg, which can fly in Dragon's trunk. There's no real economic case to invest in a smaller launcher.

Economics aside, it is probably feasible to land the F5 on it's center Merlin, although that would still be a pretty sporty landing. Definitely easier than landing a 5-Raptor Falcon but still certainly pushing the limits of a hoverslam.

Offline speedevil

  • Senior Member
  • *****
  • Posts: 4406
  • Fife
  • Liked: 2762
  • Likes Given: 3369
Re: Pushing the limits of the hoverslam landing
« Reply #84 on: 05/09/2016 07:54 pm »
According to Musk's tweet, the landing thrust goes down to 40% of a single Merlin. That would be about 16% of a single Raptor.

I don't think he quite said that.
That's 260kN or so, and with a stage weighing 20 tons, a residual accelleration after gravity of only 3m/s.
The rocket would go the last rocket-height in ~6 seconds.
That seems rather too slow.
I'd think that they'd be up near the top of the thrust range - 80% perhaps, not 40%.
If you're at 80%, you can throttle up 20 and down 40%. If you're at 40%, you can't throttle down at all.

40% might be desirable to get a human-scale timeline for landing - but it is actively bad to slow down that far IMO.
The problem is the slower you are decelerating, the more time you have for wind to blow the stage sideways with limited time to correct, and no authority from the grid fins at all.

Offline envy887

  • Senior Member
  • *****
  • Posts: 8150
  • Liked: 6816
  • Likes Given: 2965
Re: Pushing the limits of the hoverslam landing
« Reply #85 on: 05/09/2016 08:12 pm »
According to Musk's tweet, the landing thrust goes down to 40% of a single Merlin. That would be about 16% of a single Raptor.

I don't think he quite said that.
That's 260kN or so, and with a stage weighing 20 tons, a residual accelleration after gravity of only 3m/s.
The rocket would go the last rocket-height in ~6 seconds.
That seems rather too slow.
I'd think that they'd be up near the top of the thrust range - 80% perhaps, not 40%.
If you're at 80%, you can throttle up 20 and down 40%. If you're at 40%, you can't throttle down at all.

40% might be desirable to get a human-scale timeline for landing - but it is actively bad to slow down that far IMO.
The problem is the slower you are decelerating, the more time you have for wind to blow the stage sideways with limited time to correct, and no authority from the grid fins at all.

It seems pretty clear that he is actually saying that: https://twitter.com/elonmusk/status/728753234811060224 Context before and after the 40% part is obviously referencing a 3-engine landing burn.

How long it runs at 40% is not obvious; it may only be throttled that low for >1 second before touchdown.

Online sanman

  • Senior Member
  • *****
  • Posts: 6014
  • Liked: 1327
  • Likes Given: 8
Re: Pushing the limits of the hoverslam landing
« Reply #86 on: 05/11/2016 07:02 pm »
The numbers are up the thread... 2 engines saves about 40% and three engines about 47% of landing fuel, but sacrifices a lot of control.

How do they do 2 engines? Both outer engines and no central gimbaled engine? Or is it the gimbaled central engine plus one of the outer ones? (sounds asymmetric)

For 3 engines, is it the central gimbaled one and 2 outer engines? Or are all 3 outer engines?
« Last Edit: 05/11/2016 07:03 pm by sanman »

Online spacenut

  • Senior Member
  • *****
  • Posts: 5190
  • East Alabama
  • Liked: 2591
  • Likes Given: 2903
Re: Pushing the limits of the hoverslam landing
« Reply #87 on: 05/11/2016 07:17 pm »
I'm not sure, but I think all the engines are gimbaled.  At the very least 4 outer engines would be. 

Offline sevenperforce

  • Full Member
  • ****
  • Posts: 1474
  • Liked: 969
  • Likes Given: 599
Re: Pushing the limits of the hoverslam landing
« Reply #88 on: 05/11/2016 07:17 pm »
The numbers are up the thread... 2 engines saves about 40% and three engines about 47% of landing fuel, but sacrifices a lot of control.

How do they do 2 engines? Both outer engines and no central gimbaled engine? Or is it the gimbaled central engine plus one of the outer ones? (sounds asymmetric)

For 3 engines, is it the central gimbaled one and 2 outer engines? Or are all 3 outer engines?
Only three engines have plumbing for TEA/TEB reignition, and all three receive a reignition pulse regardless of which ones are lit. They are the central engine and two opposite outer engines.

IIRC all the engines are gimbaled, though the center might have more freedom than the outer ones.

Offline mvpel

  • Full Member
  • ****
  • Posts: 1125
  • New Hampshire
  • Liked: 1303
  • Likes Given: 1685
Re: Pushing the limits of the hoverslam landing
« Reply #89 on: 05/11/2016 10:33 pm »


You can see the TVC pistons in the two engines in the forefront of this picture, and a hint of the top of one in the rightmost engine, which is indicative that all engines are vectorable.
"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 Kabloona

  • Senior Member
  • *****
  • Posts: 4847
  • Velocitas Eradico
  • Fortress of Solitude
  • Liked: 3432
  • Likes Given: 741
Re: Pushing the limits of the hoverslam landing
« Reply #90 on: 05/11/2016 11:27 pm »
Quote
It seems pretty clear that he is actually saying that: https://twitter.com/elonmusk/status/728753234811060224 Context before and after the 40% part is obviously referencing a 3-engine landing burn.

How long it runs at 40% is not obvious; it may only be throttled that low for >1 second before touchdown.

Here is one notional scheme for a 3-engine landing burn.

The final single-engine throttle-down sequence can be a pre-programmed curve with known total impulse based on engine testing. It's important that the thrust-vs-time curve during throttle-down be well-characterized by engine testing on a thrust stand because there's no room for error that close to touchdown. So that curve can be characterized/optimized by experiment on the test stand and then programmed into the autopilot for repeatability.

With the throttle-down curve pre-programmed, the critical variable becomes the time lag between shutdown of the 2 outer engines and initiation of the center engine throttle-down program. This delta-t can be calculated in real time based on altitude, descent speed, and acceleration with the center engine running at 100% throttle (from which stage mass can be derived). Those three variables should be sufficient for the autopilot to determine when to intitiate the throttle-down sequence for a soft landing.
« Last Edit: 05/11/2016 11:44 pm by Kabloona »

Offline dorkmo

  • Full Member
  • ****
  • Posts: 711
  • Liked: 339
  • Likes Given: 848
Re: Pushing the limits of the hoverslam landing
« Reply #91 on: 05/12/2016 06:07 pm »
wonder how much fuel reserve is taken into account by the computer. would the most fuel efficient option be to jump directly from 3 engines at 100% to 1 engine at 40%?

Offline sevenperforce

  • Full Member
  • ****
  • Posts: 1474
  • Liked: 969
  • Likes Given: 599
Re: Pushing the limits of the hoverslam landing
« Reply #92 on: 05/12/2016 06:32 pm »
Quote
It seems pretty clear that he is actually saying that: https://twitter.com/elonmusk/status/728753234811060224 Context before and after the 40% part is obviously referencing a 3-engine landing burn.

How long it runs at 40% is not obvious; it may only be throttled that low for >1 second before touchdown.

Here is one notional scheme for a 3-engine landing burn.

The final single-engine throttle-down sequence can be a pre-programmed curve with known total impulse based on engine testing. It's important that the thrust-vs-time curve during throttle-down be well-characterized by engine testing on a thrust stand because there's no room for error that close to touchdown. So that curve can be characterized/optimized by experiment on the test stand and then programmed into the autopilot for repeatability.

With the throttle-down curve pre-programmed, the critical variable becomes the time lag between shutdown of the 2 outer engines and initiation of the center engine throttle-down program. This delta-t can be calculated in real time based on altitude, descent speed, and acceleration with the center engine running at 100% throttle (from which stage mass can be derived). Those three variables should be sufficient for the autopilot to determine when to intitiate the throttle-down sequence for a soft landing.
One other issue to keep in mind is that the shutoff itself is also a gradual throttledown in thrust, though a more rapid one....


Offline Kabloona

  • Senior Member
  • *****
  • Posts: 4847
  • Velocitas Eradico
  • Fortress of Solitude
  • Liked: 3432
  • Likes Given: 741
Re: Pushing the limits of the hoverslam landing
« Reply #93 on: 05/12/2016 06:34 pm »
wonder how much fuel reserve is taken into account by the computer. would the most fuel efficient option be to jump directly from 3 engines at 100% to 1 engine at 40%?

It's more fuel efficient to burn at high thrust as long as possible, so if you're going to shut down 2 engines, it's better from an efficiency POV to run the center engine at 100% as long as possible before throttling down to 40%.

And the computer doesn't need to know how much propellant remains. It's probably pre-programmed to do an entry burn of specific time duration, and the landing burn is probably pre-programmed to start at a certain altitude/velocity, based on pre-launch Monte Carlo simulations that give them a good idea of how much propellant will be used during the burns. And they would do Monte Carlo runs to calculate the propellant reserves at MECO and verify that the expected reserves at MECO exceed the expected consumption during entry and landing.

For SES-9 the results of those two Monte Carlo simulations were probably quite close, resulting in their public prediction that successful landing was unlikely.

But the flight computer probably just flies the mission assuming it has enough propellant to land safely. And you wouldn't want it to know if it didn't.  ;)

Quote
One other issue to keep in mind is that the shutoff itself is also a gradual throttledown in thrust, though a more rapid one....

Yup, my drawing left out a few details...
« Last Edit: 05/12/2016 06:44 pm by Kabloona »

Offline llanitedave

  • Senior Member
  • *****
  • Posts: 2286
  • Nevada Desert
  • Liked: 1545
  • Likes Given: 2055
Re: Pushing the limits of the hoverslam landing
« Reply #94 on: 05/13/2016 12:00 am »
wonder how much fuel reserve is taken into account by the computer. would the most fuel efficient option be to jump directly from 3 engines at 100% to 1 engine at 40%?

It's more fuel efficient to burn at high thrust as long as possible, so if you're going to shut down 2 engines, it's better from an efficiency POV to run the center engine at 100% as long as possible before throttling down to 40%.

And the computer doesn't need to know how much propellant remains. It's probably pre-programmed to do an entry burn of specific time duration, and the landing burn is probably pre-programmed to start at a certain altitude/velocity, based on pre-launch Monte Carlo simulations that give them a good idea of how much propellant will be used during the burns. And they would do Monte Carlo runs to calculate the propellant reserves at MECO and verify that the expected reserves at MECO exceed the expected consumption during entry and landing.

For SES-9 the results of those two Monte Carlo simulations were probably quite close, resulting in their public prediction that successful landing was unlikely.

But the flight computer probably just flies the mission assuming it has enough propellant to land safely. And you wouldn't want it to know if it didn't.  ;)

Quote
One other issue to keep in mind is that the shutoff itself is also a gradual throttledown in thrust, though a more rapid one....

Yup, my drawing left out a few details...


Shouldn't really have to assume anything.  The system knows how many engines it's running and how they are throttled.  It knows what the turbopump speed and pressure is.  It knows how the engine responds.  It should be able to calculate to a large degree of accuracy what the thrust being produced by those engines is at any time slice.


At the same time, it knows what acceleration the stage is experiencing, and in what directions.  It knows the empty mass of the stage and it knows how much fuel and oxygen has been loaded.  Therefore, it's a straightforward calculation to derive the current mass of the stage, and therefore how much propellant is currently remaining.


How it would use that information, what kind of autonomous decision-making ability it has, I wouldn't know.  At the very least, it can be comparing its actual performance with that of the pre-flight simulations to help refine future plans.  At most, it might be able to make adjustments on the way down to try to optimize its trajectory and touchdown velocity.
"I've just abducted an alien -- now what?"

Offline Kabloona

  • Senior Member
  • *****
  • Posts: 4847
  • Velocitas Eradico
  • Fortress of Solitude
  • Liked: 3432
  • Likes Given: 741
Re: Pushing the limits of the hoverslam landing
« Reply #95 on: 05/13/2016 12:15 am »
Quote
How it would use that information, what kind of autonomous decision-making ability it has, I wouldn't know.  At the very least, it can be comparing its actual performance with that of the pre-flight simulations to help refine future plans.  At most, it might be able to make adjustments on the way down to try to optimize its trajectory and touchdown velocity.

My point was, the autoilot doesn't need to care or know how much propellant is left. The boostback/entry/landing burn parameters are probably pre-programmed in such a way that the propellant usage is pretty much fixed by the flight profile (subject to the usual dispersions of course) which has been run on Monte Carlo simulations, so they know ahead of time how much propellant will be needed at MECO.

All the flight computer is doing is running the boostback/entry/landing burn sequences based on given parameters, and controlling the landing burn throttle-down timing to achieve zero-zero landing. None of that requires knowledge of propellant remaining. Naturally the computer is adjusting throttle based on accelaration and velocity, but that also is a function of stage total mass at any given time and doesn't require derivation of propellant mass remaining.

So I was being flip about the flight computer "assuming" it has enough propellant, but it really doesn't need to know anything about remaining reserves. It simply runs its boostback/entry/landing burns until it either touches down or runs out of propellant. Either way, knowledge of propellant quantity remaining wouldn't change how the autpoilot operates, IMO, so there's no need to try to derive it in real time.
« Last Edit: 05/13/2016 12:23 am by Kabloona »

Offline llanitedave

  • Senior Member
  • *****
  • Posts: 2286
  • Nevada Desert
  • Liked: 1545
  • Likes Given: 2055
Re: Pushing the limits of the hoverslam landing
« Reply #96 on: 05/13/2016 12:20 am »
Quote
How it would use that information, what kind of autonomous decision-making ability it has, I wouldn't know.  At the very least, it can be comparing its actual performance with that of the pre-flight simulations to help refine future plans.  At most, it might be able to make adjustments on the way down to try to optimize its trajectory and touchdown velocity.

My point was, the autoilot doesn't need to care or know how much propellant is left. The boostback/entry/landing burn parameters are probably pre-programmed in such a way that the propellant usage is pretty much fixed by the flight profile (subject to the usual dispersions of course) which has been run on Monte Carlo simulations, so they know ahead of time how much propellant will be needed at MECO.

All the flight computer is doing is running the boostback/entry/landing burn sequences based on given parameters, and controlling the landing burn throttle-down timing to achieve zero-zero landing. None of that requires knowledge of propellant remaining. Naturally the computer is adjusting throttle based on accelaration and velocity, but that also is a function of stage total mass at any given time and doesn't require derivation of propellant mass remaining.

So I was being flip about the flight computer "assuming" it has enough propellant, but it really doesn't need to know anything about remaining reserves. It simply runs its boostback/entry/landing burns until it either touches down or runs out of propellant. Either way, knowledge of propellant reserves wouldn't change its performance, IMHO.


It may be minor, but the total amount of mass remaining would affect its acceleration at a given thrust level.  So either it needs to be able to adjust it parameters on the fly, or the preflight predictions need to be very accurate.  I'm sure they are.
"I've just abducted an alien -- now what?"

Offline Kabloona

  • Senior Member
  • *****
  • Posts: 4847
  • Velocitas Eradico
  • Fortress of Solitude
  • Liked: 3432
  • Likes Given: 741
Re: Pushing the limits of the hoverslam landing
« Reply #97 on: 05/13/2016 12:33 am »
Quote
It may be minor, but the total amount of mass remaining would affect its acceleration at a given thrust level.  So either it needs to be able to adjust it parameters on the fly, or the preflight predictions need to be very accurate.  I'm sure they are.

True, but again the instantaneous acceleration at any given time is based on thrust/total mass. It already knows its acceleration during the 100% thrust portion of the landing burn, from which instantaneous total mass can be derived, and it knows how fast it's burning propellant to subtract from that total mass. So it knows how fast total mass is/will be decreasing based on throttle level.

Again, you *could* derive remaining propellant mass from that, but it's unnecessary.
« Last Edit: 05/13/2016 12:36 am by Kabloona »

Offline CameronD

  • Senior Member
  • *****
  • Posts: 2408
  • Melbourne, Australia
    • Norton Consultants
  • Liked: 890
  • Likes Given: 560
Re: Pushing the limits of the hoverslam landing
« Reply #98 on: 05/13/2016 12:47 am »
Quote
It may be minor, but the total amount of mass remaining would affect its acceleration at a given thrust level.  So either it needs to be able to adjust it parameters on the fly, or the preflight predictions need to be very accurate.  I'm sure they are.

True, but again the instantaneous acceleration at any given time is based on thrust/total mass. It already knows its acceleration during the 100% thrust portion of the landing burn, from which instantaneous total mass can be derived, and it knows how fast it's burning propellant to subtract from that total mass. So it knows how fast total mass is/will be decreasing based on throttle level.

Again, you *could* derive remaining propellant mass from that, but it's unnecessary.

That all sounds great in theory, but does anyone know precisely what they'd be basing those mass figures on?  To know how fast it's burning propellant, does it use mass flowmeters - or something else?  If they use mighty-accurate mass flowmeters, can anyone hazard a guess at a make/model??

Reason I ask is that the results of calculations of this nature are only going to be as accurate as the accuracy/response time of their sensors (input data), so it'd be nice to know what that was to know whether or not it's possible to do it..
« Last Edit: 05/13/2016 12:50 am by CameronD »
With sufficient thrust, pigs fly just fine - however, this is not necessarily a good idea. It is hard to be sure where they are
going to land, and it could be dangerous sitting under them as they fly overhead.

Online Mark K

  • Full Member
  • *
  • Posts: 139
  • Wisconsin
  • Liked: 79
  • Likes Given: 29
Re: Pushing the limits of the hoverslam landing
« Reply #99 on: 05/13/2016 01:36 am »

That all sounds great in theory, but does anyone know precisely what they'd be basing those mass figures on?  To know how fast it's burning propellant, does it use mass flowmeters - or something else?  If they use mighty-accurate mass flowmeters, can anyone hazard a guess at a make/model??

The point being made is that the acceleration measurement, plus the known thrust of engine at the measured throttle allow the system to derive the mass of the booster at any point. Subtract the known non-propellent mass and you know the mass of the remaining fuel plus oxidiser. Whether the system cares or uses that is another issue.


Tags:
 

Advertisement NovaTech
Advertisement Northrop Grumman
Advertisement
Advertisement Margaritaville Beach Resort South Padre Island
Advertisement Brady Kenniston
Advertisement NextSpaceflight
Advertisement Nathan Barker Photography
1