totm nov 2023 SpaceX Discussion Thread

[cubinator]

31 minutes ago, JoeSchmuckatelli said:

Interesting; I did not know that.  So - they'd have a pipe feeding a something, then another something to measure its work and wires to report on the findings?

Hmm.

Is that a normal part of rocket design -- or something SX did b/c R1 is effectively a prototype (a working prototype, but still a development article)? 

Yep! It's normal to have sensors on some parts of the rocket (for instance the tank and chamber pressure, and the position of the gimbals) but for a newly designed engine like Raptor 1 they would be learning as much as possible about every part of the system by placing sensors in many places.

[JoeSchmuckatelli]

10 minutes ago, cubinator said:

Yep! It's normal to have sensors on some parts of the rocket (for instance the tank and chamber pressure, and the position of the gimbals) but for a newly designed engine like Raptor 1 they would be learning as much as possible about every part of the system by placing sensors in many places.

Yeah - I can see where they would need a certain number of sensors - like any vehicle - to know what's going on and to control/get feedback.  What I wasn't aware of was that it wasn't already pared down to the minimum.  I kind of assumed that they'd have a 'test article' build in the beginning where all the systems were tied in with feedback, and then the production articles would be more streamlined.  Thus when I looked at a R1, I'd see a rocket with all the necessary plumbing to make it work... and think; yep, that's Rocket Science.  Cool.

If what I'm reading from you all is correct, SX basically built a working rocket with all the sensors required to know what's going on with the design... and then kept building every subsequent R1 identically.  Thus, R2 is the 'working' article, where R1 was a bunch of 'test articles'?

(Or did they learn stuff with R1 and not only strip off unnecessary feedback features, but also improve the plumbing?)

[RealKerbal3x]

2 minutes ago, JoeSchmuckatelli said:

If what I'm reading from you all is correct, SX basically built a working rocket with all the sensors required to know what's going on with the design... and then kept building every subsequent R1 identically.  Thus, R2 is the 'working' article, where R1 was a bunch of 'test articles'?

(Or did they learn stuff with R1 and not only strip off unnecessary feedback features, but also improve the plumbing?)

Raptor is still very much an in-development engine. SpaceX didn't build the first Raptor 1, say 'yeah, this is fine' and produced 50 more identical to it. With every Raptor, they've been improving and optimising things, with the goal being to produce the most powerful, efficient, compact and mass-producible engine they can. That's why, up to now, they've needed that rat's nest of sensors, because almost everything about the engine is a moving target.

I find it helpful to think of Raptor 2 as the 'production' version of Raptor. They're still tweaking things and making changes - as Elon said they're still working on increasing thrust - but the engine is enough to get Starship working, and they can focus on producing as many of these as they can, as fast as possible. They're going to need a lot.

[Beccab]

Just now, RealKerbal3x said:

I find it helpful to think of Raptor 2 as the 'production' version of Raptor. They're still tweaking things and making changes - as Elon said they're still working on increasing thrust - but the engine is enough to get Starship working, and they can focus on producing as many of these as they can, as fast as possible. They're going to need a lot.

Adding to this - there is definitely also going to be a Raptor 3, and maybe 4 in a couple years

[tater]

I'm not sure those 2 engines are a fair comparison. Are we sure both are "center" engines—that gimbal and throttle, or might R2 be a "ring" booster engine?

[Beccab]

The beach is now closed and pad is clear for today's cryo

[RCgothic]

 

https://twitter.com/elonmusk/status/1492050352664887315?s=20&t=DB7zfu9rbVJUilFqa5mORA 

 

https://twitter.com/elonmusk/status/1492050572505137161?s=20&t=DB7zfu9rbVJUilFqa5mORA 

 

Urgh, these aren't embedding properly.

 

 

 

Edited February 11, 2022 by RCgothic

[RCgothic]

 

That's better! Now properly embedded.

Basically the remaining wires are going, and aspirationally there'll be a 250tf variant by year end. 

 

Edited February 11, 2022 by RCgothic

[Lukaszenko]

I didn't really understand the part where Elon said they're having problems keeping the combustion chamber from melting. Like...what? That seems kind of critical. How do you have an apparently working and even flying engine where the combustion chamber melts? And then, how do you start mass-producing it before fixing this seemingly not-so-small issue?

 

Edited February 12, 2022 by Lukaszenko
Improvement

[linuxgurugamer]

8 minutes ago, Lukaszenko said:

I didn't really understand the part where Elon said they're having problems keeping the combustion chamber from melting. Like...what? That seems kind of critical. How do you have an apparently working and even flying engine where the combustion chamber melts? And then, how do you start mass-producing it before fixing this seemingly not-so-small issue?

 

Well, that's why they are still in development.  They aren't flying real missions, the only flights have been test flights, which tests an unimaginable number of things.  Quite possible that the flights they did weren't at full throttle

[RCgothic]

36 minutes ago, Lukaszenko said:

I didn't really understand the part where Elon said they're having problems keeping the combustion chamber from melting. Like...what? That seems kind of critical. How do you have an apparently working and even flying engine where the combustion chamber melts? And then, how do you start mass-producing it before fixing this seemingly not-so-small issue?

 

An engine that melts at 340bar and 250tf thrust will be fine at 320 bar and 230tf thrust. And in the mean time the collected data from multiple firings will help them progress a solution.

[mikegarrison]

42 minutes ago, Lukaszenko said:

I didn't really understand the part where Elon said they're having problems keeping the combustion chamber from melting. Like...what? That seems kind of critical. How do you have an apparently working and even flying engine where the combustion chamber melts? And then, how do you start mass-producing it before fixing this seemingly not-so-small issue?

Combustion is generally understood to be a "black art", although modern computational tools are changing that somewhat.

Anyway, if it were easy to do, I can safely say my career path would have been very different.

You need to get the flame to start burning, get it to keep burning, manage the heat transfer to the combustor walls, avoid acoustic instability ("screech"), minimize the amount of cooling you can get away with (because that is always parasitic on the ultimate power output), etc.

Typically the problem is going to be that there are "hot spots". This is because the fuel and oxygen are injected in discrete locations, not magically teleported into the combustion chamber in a perfect premix.

Also, while the main engine does not have this worry, the upstream turbopumps have to not only keep the combustor from melting down but also have to worry about keeping the turbine alive.

Edited February 12, 2022 by mikegarrison

[JoeSchmuckatelli]

I was going to suggest metal foam, but then Mike said that you need to minimize cooling - so I guess it really is a dark art. 

Ogun, show us the way! 

[Silavite]

59 minutes ago, mikegarrison said:

Combustion is generally understood to be a "black art", although modern computational tools are changing that somewhat.

Anyway, if it were easy to do, I can safely say my career path would have been very different.

You need to get the flame to start burning, get it to keep burning, manage the heat transfer to the combustor walls, avoid acoustic instability ("screech"), minimize the amount of cooling you can get away with (because that is always parasitic on the ultimate power output), etc.

Typically the problem is going to be that there are "hot spots". This is because the fuel and oxygen are injected in discrete locations, not magically teleported into the combustion chamber in a perfect premix.

Also, while the main engine does not have this worry, the upstream turbopumps have to not only keep the combustor from melting down but also have to worry about keeping the turbine alive.

54 minutes ago, JoeSchmuckatelli said:

I was going to suggest metal foam, but then Mike said that you need to minimize cooling - so I guess it really is a dark art. 

Ogun, show us the way! 

To clarify this point about minimizing cooling, it is specifically film/dump cooling that should be minimized. Film cooling acts to prevent heat transfer to the chamber walls by injecting a layer of (usually) fuel around the chamber periphery. This creates a region of very rich, off-nominal mixture ratio fluid near the wall that does not burn and serves to protect the chamber wall from the extremely hot "core" combustion. This has a negative effect on performance because you're intentionally creating an area of maldistributed (from a combustion efficiency perspective) fuel which does not completely burn. This paper partially discusses the deleterious effects of film cooling and how designers might try to use less of it.

I'm unsure about exactly what effects ablative cooling have on performance. The only things I can say with much certainty are that the engine's thrust will increase (due to the increased mass flow rate from the increase in throat diameter as the throat material ablates away), and that the engine's vacuum exhaust velocity will decrease (as a consequence of decreased expansion ratio, which is in turn a consequence of increasing throat diameter).

Regenerative cooling of the combustion chamber has basically no effect on performance. The energy lost through the chamber liner into the coolant ends up going back into the chamber as somewhat-warmer-than-ambient propellant (indeed, expander cycle engines take this concept to the extreme), so very little heat actually escapes the engine system in regenerative cooling. Regenerative cooling of the supersonic flow in the nozzle extension can actually have a positive effect on performance (due to the decrease in entropy), though the effect is quite small (on the order of 10 m/s exhaust velocity) for all but the highest of expansion ratios. See here for more reading about the effects of regenerative cooling on performance.

[mikegarrison]

Anything ablative is pretty much exactly opposite their stated intention of rapid reuse.

[RCgothic]

Raptor doesn't use ablative cooling.

Other than being non-reusable, Ablative cooling is generally negative on performance. The extra mass is heavy and as the nozzle throat erodes the engine loses expansion ratio.

[sevenperforce]

On 2/10/2022 at 3:22 AM, kerbiloid said:

On the first stage it tells that the required engine number was underestimated (possibly because the available engine thrust was overestimated, compared to the actual).

To the contrary, Raptor 2 has thrust levels significantly higher than what was estimated when the booster diameter was set.

On 2/10/2022 at 3:22 AM, kerbiloid said:

Otherwise they would just make the pencil a meter wider, and needed no skirt (which is additional air drag).

If you make it wider, then it carries more propellant, which makes it heavier. So you need more engines at the base. So you need a wider rocket. Which then makes it heavier. So you need more engines. And so forth.

Or did you think the Saturn V should have been wider too?

On 2/10/2022 at 3:22 AM, kerbiloid said:

On the upper stage this in turn  is forced by the lower stage diameter (and probably by the fires in almost every test caused by the engines put too close to each other), as no engineer in clear mind would make these cutout willingly.
(Because they both weaken the constructon and complicate the production)

The cutouts in the stiffener ring are attached to secondary stiffener rings wrapped around each engine. 

Despite your contentions about the engines being too close together and resulting in fires, you will note that the three central engines have not changed position. The placement of the vacuum engines is to allow wide gimbal range for the central engines.

[sevenperforce]

On 2/11/2022 at 10:02 AM, Beccab said:

Size comparison of the full stack with other rockets

I love how you included New Glenn right there between the Saturn V and the Starship.

On 2/11/2022 at 10:57 AM, JoeSchmuckatelli said:

How in the world do they go from 1 to 2 with such a change in plumbing/wiring?

If we assume all of that was necessary for 1 to work, it just doesn't look like everything is there on 2.  (Here's where I express my ignorance of rocket parts) - the streamlined thingy at the top of 2 looks like something that could be on 1, but under that giant manifold thing sitting at the top of 1.  IF that manifold thing was necessary - doesn't it come with a lot of plumbing, and might all that be added back to a working 2?

As he said, it was a complete redesign. A lot of it was deletion, and then there was combination/consolidation as well. A lot of the wiring and electronics were consolidated, and a lot of the piping was redundant or used an inefficient path.

[JoeSchmuckatelli]

12 minutes ago, sevenperforce said:

I love how you included New Glenn right there between the Saturn V and the Starship.

8P

 

[tater]

57 minutes ago, sevenperforce said:

I love how you included New Glenn right there between the Saturn V and the Starship.

That's gonna leave a mark.

[CatastrophicFailure]

14 minutes ago, tater said:

That's gonna leave a mark.

Burns more than the real thing... 

[sevenperforce]

23 hours ago, linuxgurugamer said:

Well, that's why they are still in development.  They aren't flying real missions, the only flights have been test flights, which tests an unimaginable number of things.  Quite possible that the flights they did weren't at full throttle

Well, the engines in the test flights were Raptor 1 engines, not Raptor 2s. In those test flights, the Raptor 1s started at full thrust (in fact, they had to detank significantly in order to get a TWR > 1) but throttled down over the course of the flight in order to reach hover at 10 km. But, significantly, that “full thrust” was the full thrust for Raptor 1.

Elon said that they have been able to sustain Raptor 2 firings at 230 tonnes-force, so presumably the Raptor 2 chamber doesn’t melt at that thrust level. The chamber melt presumably happens up around the 247 tonne-force level he referenced. That makes sense; at a higher chamber pressure, the heat transfer between the combustion zone and the chamber wall is going to increase. With film cooling, increased pressure is also going to lead to less effective film cooling because the fuel flow along the boundary will be compressed. But adding more film cooling decreases efficiency (since it’s unburned propellant) and may introduce other issues. It’s possible that there is a physical limit to the amount of film cooling that can be used before the boundary layer flow becomes chaotic and stops protecting the copper liner from heat.

In any event, I think we can be confident that the current Raptor 2 engines they are churning out will be just fine at 230 tonnes-force. SpaceX might be able to run the existing Raptor 2s at higher thrust simply by getting the right mixture and combustion parameters (a software solution), or they may need to do some hardware tweaks on the production line for later Raptor 2s. 

[mikegarrison]

3 hours ago, sevenperforce said:

Despite your contentions about the engines being too close together and resulting in fires, you will note that the three central engines have not changed position. The placement of the vacuum engines is to allow wide gimbal range for the central engines.

The engines being all packed in there are a definite risk item. There is always a possibility of cascade failure if one blows up, that damages the next one, which also blows up, etc.

That's not to say it's the wrong decision. Risks are risks, and some of them need to be taken. But it's also not a ridiculous concern.

[Nothalogh]

On 2/11/2022 at 7:17 AM, RCgothic said:

if Blue Origin were to build a 10 diameter (approx outer diameter of Superheavy outer engine ring) vehicle, they could fit perhaps 12 fixed and 3 or 4 gimbaling engines

Blue Origin would first have to build engines

[tater]

3 hours ago, Nothalogh said:

Blue Origin would first have to build engines

Between you at @sevenperforce BO is down for the count today.