totm nov 2023 SpaceX Discussion Thread

[kerbiloid]

8 hours ago, cubinator said:

Definitely not a good option when the landing is happening on Mars, where there is no ocean to divert to...

Diverting to ground is even worse than to water.

Edited April 17, 2020 by kerbiloid

[magnemoe]

8 hours ago, sevenperforce said:

Water tower was able to hover because it had a single engine. SN4 and all the rest of its line have three engines in a triad configuration so there's no way to hover on just one central engine.

Except gimbal the engine a bit, to compensate for the off center trust. The reason for three engines in the center is to give engine out capability. One option is to land on two engines or only one, last who would give it low enough trust to hover. 
Downside with this is engine out as it takes some time to start another engine and if you are 5 meter up you might not have enough time. 

[Hannu2]

10 minutes ago, magnemoe said:

Except gimbal the engine a bit, to compensate for the off center trust. The reason for three engines in the center is to give engine out capability. One option is to land on two engines or only one, last who would give it low enough trust to hover. 
Downside with this is engine out as it takes some time to start another engine and if you are 5 meter up you might not have enough time. 

Gimbal can compensate torque if the thrust vector points to the center of gravity. But it causes lateral force, which makes killing of lateral velocity impossible, if the rocket is not tilted so that working engine is under center of gravity. Landing will be asymmetric and cause excessive load to lander legs.

Probably they can save the rocket in emergency situation, but it is hard to believe that they could use asymmetric landing as normal operation and reuse the rocket after that.

[sevenperforce]

4 hours ago, Hannu2 said:

4 hours ago, magnemoe said:

Quote

Water tower was able to hover because it had a single engine. SN4 and all the rest of its line have three engines in a triad configuration so there's no way to hover on just one central engine.

Except gimbal the engine a bit, to compensate for the off center trust. The reason for three engines in the center is to give engine out capability. One option is to land on two engines or only one, last who would give it low enough trust to hover. 
Downside with this is engine out as it takes some time to start another engine and if you are 5 meter up you might not have enough time.

Gimbal can compensate torque if the thrust vector points to the center of gravity. But it causes lateral force, which makes killing of lateral velocity impossible, if the rocket is not tilted so that working engine is under center of gravity. Landing will be asymmetric and cause excessive load to lander legs.

Probably they can save the rocket in emergency situation, but it is hard to believe that they could use asymmetric landing as normal operation and reuse the rocket after that.

I don't recall if it was Twitter or elsewhere, but Elon said that they would typically hoverslam for this reason. They can use asymmetric landing for emergencies but that's it. 

EDIT: 

 

[StrandedonEarth]

5 hours ago, magnemoe said:

Downside with this is engine out as it takes some time to start another engine and if you are 5 meter up you might not have enough time. 

Elon tweeted a while back that they can emergency-start a raptor in less than half a second, IIRC. I want to say the quoted time was under a tenth, but I’m probably wrong 

[tater]

 

 

 

[Flavio hc16]

1 hour ago, tater said:

 

Am I the only one who is curious why they did this static test so much before the normal flight? Are they maybe diluting the times so that less people are around at the same time?

1 hour ago, tater said:

 

That's great, it will be a nice birthday gift for me ( 29th of may). I think that they choose the 27th so that the launch window is at a time where the coverage of the media will be maximised ( USA ofc and EU, but it should be a great time window also for China and our friends down under)

[KerballingSmasher]

On 4/15/2020 at 7:27 PM, kerbiloid said:

21 more skipped?

No, I think it just means that it's the 26th engine... 

[tater]

 

[sevenperforce]

Just now, tater said:

 

"Static aero" suggests that the baseline configuration is changing. We should start speculating.

Unfurled dragon wings?

Fixed wings with dihedral and edge flaps?

Something that would make it controllable via engine gimbal even if the actuators seized up?

[tater]

 

[Flavio hc16]

 

2 minutes ago, tater said:

 

sunny, with high probability of tweetstorm.

I'm very curious of what they are coming up with, we will see probably with sn 5-6, so 1-2 months away

 

[sevenperforce]

30 minutes ago, Flavio hc16 said:

sunny, with high probability of tweetstorm.

I'm very curious of what they are coming up with, we will see probably with sn 5-6, so 1-2 months away

"The best part is no part, the best process is no process, the best system is no system."

Even though the four-flap approach approximates the four limbs of a skydiver well, I've always thought it seemed failure-prone. Curious to see what exactly they can eliminate. If they lose the nose flap actuators altogether and make the forward fins smaller, will they save enough weight to reserve propellant for heavy RCS use during descent?

"Static aero" sounds like passive aerodynamic stability. I wonder if we see smaller, fixed forward fins and dihedral rear fins with the medial region fixed and the lateral region actuated. Half as many failure points. Add a bunch of the hot-gas thrusters and you might have something workable. Presumably the idea is that if the rear flaps were to lock up, you might go into a slow tailspin and miss the center of your landing zone, but you could control the spin with hot-gas thrusters and use engine gimbal to correct before landing nearby.

[sevenperforce]

Here's what I'm thinking

Fixed forward fins parallel to the midline. Rear wings offset ventrally from the midline but tilted back in a dihedral; fixed but with actuating control surfaces in the center. 

Freeze the two control surfaces, and the whole thing falls like a skydiver with passive aerodynamic stability and a small amount of body lift. Pull both of the control surfaces back (dorsal) and pitch up; push both forward to pitch down. Differential rotation rolls left or right. You have a little bit of yaw control by using pitch and roll together, but RCS provides primary reaction control. 

If one of the control surfaces fails, lock the other one in the same position and use RCS to compensate until you are low enough to perform the kick-flip.

Nominally, the kick-flip is initiated by folding both control surfaces back and firing RCS hard. If you lose RCS, you can use the control surfaces + engine gimbal to do the same thing; if you lose control surfaces, you can use RCS + engine gimbal to do the same thing. 

[sevenperforce]

Note: if you like this idea and want Elon to comment on it, retweet it to him here:

 

[tater]

 

[tater]

 

[Silavite]

I had the idea of using a single set of control surfaces which do not rotate, but only translate. Here's a crude MS Paint approximation:

Benefits:

• Fewer control surfaces required in total could help to reduce mass and potential points of failure.
• The fact that these surfaces move perpendicular to the velocity vector (in a skydiver configuration) rather than directly against it means that less powerful and thus less massive actuators would be required.

Drawbacks:

• A big slot on the side of the spacecraft for the necessary translation would likely be... undesirable during reentry.
• Even if the actuators could be smaller, they're delivering power to a moving surface. Thus, the actuator would have to move with the control surfaces, or some kind of system would be needed to deliver the force. (Or maybe you could try using some kind of linear motor.)

Comments? I'm sure there are more flaws which I haven't thought of yet.

[Wjolcz]

43 minutes ago, Silavite said:

I had the idea of using a single set of control surfaces which do not rotate, but only translate. Here's a crude MS Paint approximation:

Benefits:

• Fewer control surfaces required in total could help to reduce mass and potential points of failure.
• The fact that these surfaces move perpendicular to the velocity vector (in a skydiver configuration) rather than directly against it means that less powerful and thus less massive actuators would be required.

Drawbacks:

• A big slot on the side of the spacecraft for the necessary translation would likely be... undesirable during reentry.
• Even if the actuators could be smaller, they're delivering power to a moving surface. Thus, the actuator would have to move with the control surfaces, or some kind of system would be needed to deliver the force. (Or maybe you could try using some kind of linear motor.)

Comments? I'm sure there are more flaws which I haven't thought of yet.

I'm thinking about something like this except as a deployable tail/flaps thing. Basically the part the Shuttle had to protect the engines during reentry (edit 3: apparently it's called a body flap) except it/they (multiple for roll control?) would be a part of the heat shield extending back. Basically a second layer of heat shield when retracted.

It would extend and perhaps even actuate to move the centre of pressure back.

Edit: I guess a better example would be the P-38's Fowler flaps which did just that minus the roll actuatiom.

Edit 2: the problem is I don't know how long such flaps would have to be to work.

Edited April 18, 2020 by Wjolcz

[mikegarrison]

1 hour ago, Wjolcz said:

Edit: I guess a better example would be the P-38's Fowler flaps which did just that minus the roll actuatiom.

Are you talking about dive flaps? Fowler flaps are the flaps that extend back and down from the trailing edges of many airplane wings. Fowler flaps are not suited to be deployed at high speed or else they would break off. They are only used for high lift during takeoff and landing.

[sevenperforce]

1 hour ago, Wjolcz said:

I'm thinking about something like this except as a deployable tail/flaps thing. Basically the part the Shuttle had to protect the engines during reentry (edit 3: apparently it's called a body flap) except it/they (multiple for roll control?) would be a part of the heat shield extending back. Basically a second layer of heat shield when retracted.

It would extend and perhaps even actuate to move the centre of pressure back.

Edit: I guess a better example would be the P-38's Fowler flaps which did just that minus the roll actuatiom.

Edit 2: the problem is I don't know how long such flaps would have to be to work.

This was the original BFR design, actually.

[Wjolcz]

6 hours ago, mikegarrison said:

Are you talking about dive flaps? Fowler flaps are the flaps that extend back and down from the trailing edges of many airplane wings. Fowler flaps are not suited to be deployed at high speed or else they would break off. They are only used for high lift during takeoff and landing.

I was more refering to the fact that they move back when deployed.

6 hours ago, sevenperforce said:

This was the original BFR design, actually.

I forgot lol.

Edited April 18, 2020 by Wjolcz

[RealKerbal3x]

14 hours ago, Flavio hc16 said:

Am I the only one who is curious why they did this static test so much before the normal flight? Are they maybe diluting the times so that less people are around at the same time?

This was a static fire for the upcoming Starlink mission, not DM-2.

[Flavio hc16]

1 hour ago, RealKerbal3x said:

This was a static fire for the upcoming Starlink mission, not DM-2.

i know, and still, 1 week seems a lot to me,usually they do it 2-3 days before, especially with starlink satellites

[sevenperforce]

15 hours ago, Silavite said:

I had the idea of using a single set of control surfaces which do not rotate, but only translate. Here's a crude MS Paint approximation:

Comments? I'm sure there are more flaws which I haven't thought of yet.

Not a bad idea. I am unsure of whether it would have enough control authority to provide fine adjustments. More importantly it would have little or no roll authority.

So I think we will still see flappy fins. How many, and where, is up in the air. But broadly similar.

The Apollo CM positioned heavier equipment on one side, producing a positive-lift aero stability design:

Spoiler

The CM could then use a handful of roll thrusters to turn the lift vector left or right without having to push against the airstream; they were rotating around a passively-stable axis. This was super important.

SpaceX will presumably draw on prior experience, right? And while Dragon 1 used the same approach as the Apollo CM, Dragon 2 uses a ballast sled which can translate on rails between the capsule and the heat shield, thus changing the CoM in real-time. I wonder. What if Starship uses its landing props, pumped forward or aft between discrete header tanks, to change the CoM in real time and control pitch? Then the forward fins could be smaller, and fixed, and roll control would be achieved by feathering smaller flaps on large, fixed rear fins.

We know what the header tank in the nose looks like. Could the new structure in the aft tank be the corresponding balance tank?

Spoiler