totm nov 2023 SpaceX Discussion Thread

[Jacke]

The whole reason there's so much telemetry transmitted from rockets is so as much as possible is learned from every test, even if the rocket ends up destroyed in one way or another.

Much can be learned from failures.  However, there are reasonable and unreasonable failures.  It's important to be sure things are not just setting up unreasonable failures.

In both cases are the WW2 Allied Seaborne Raids/Invasions.

Spoiler

In the reasonable cases are the main seaborne invasions starting from the 1943 invasion of Sicily (Operation Husky), through the several invasions onto beaches in Italy (Operations Avalanche, Baytown, and Slapstick), and culminating in the invasion of Normandy, France (Operation Overlord).  All of them had mistakes and failures.  Some of those failures shouldn't have happened.  But overall, it was a learning curve on planning and executing massive seaborne landings on a hostile shore.
https://en.wikipedia.org/wiki/Allied_invasion_of_Sicily
https://en.wikipedia.org/wiki/Allied_invasion_of_Italy
https://en.wikipedia.org/wiki/Normandy_landings

However, there were many more Allied seaborne raids and invasions throughout WW2.  There was a reasonable level of skill shown from the start.  Mistakes and failures happened, but usually not complete failures.

Then there's the primary example of an unreasonable failure.  Where looking at it from the first possible plans to the final disastrous execution showed so many failures.  Going ahead without enough planning, allocation of assets, preparation, and practice.

Operation Jubilee, the raid on Dieppe, 1942 August 19.
https://en.wikipedia.org/wiki/Dieppe_Raid

There's so much stupid talk of "important lessons learned".  Complete crap.  Nothing was learned about seaborne invasions and raids that wasn't known by the Royal Navy and Royal Marines going back to the Revolutionary and Napoleonic Wars.  Skimming that article shows a back-and-forth messy planning process with many forces not willing to allocate resources that were needed.

The only parts of Operation Jubilee that succeeded were the flank attacks by Commando units that weren't directly against the well-defended port.  Troops got onto the shore on the main effort, but couldn't get properly into the town.

I use this as an example because my Regiment, the Calgary Tanks, now the King's Own Calgary Regiment, was there, with Regimental Headquarters and A Squadron.  The Commanding Officer's Churchill tank left its landing craft in too deep water and was lost.  The CO was seen swimming but was never found.  Other Churchill tanks were successfully landed, but most couldn't get over or around the sea wall and their tracks had problems with the rocky beaches.

Eventually, when the forces were evacuated, the remaining Calgary Tanks fought their tanks to cover the withdrawl.  Many of the soldiers then had to surrender and become POWs for the rest of the war.

In the 1980's, I met and talked to many of those men who were POWs.  They personally suffered from the unreasonable failure of Dieppe.

It's important that a test flight that is an unreasonable failure isn't being excused as a reasonable failure.

[mikegarrison]

1 hour ago, tater said:

One feature (problem, lol?) with current Starship testing is that they are optimizing for both the vehicle, and the process of making the vehicle.

This is the default, unless you are intentionally making only a very limited production run.

I personally think SpaceX is overboard with the "build it before we even know what the design will be" thing, but maybe that's just because they are so far out of their previous experience that they feel like the lessons learned for the production process are worth the wasted production.

Traditionally the process for airplanes was to build mockups, then flight test prototypes, then the initial versions of the vehicle, and then start implementing design improvements. And of course you also have to design the tooling, the factories, and all the other parts of the production stream at the same time.

Starting in about the 1990s, companies started trying to eliminate mockups by using CAD models instead, and also tried to eliminate building prototypes and dedicated flight test vehicles. However, this has a hit or miss record, really. It worked quite well for the 777. But there have been some spectacular failures of this technique, most notably the 787 and A380, both of which had huge delays and cost overruns stemming from trying to cut too much out of the prototyping process.

[tater]

2 hours ago, mikegarrison said:

I personally think SpaceX is overboard with the "build it before we even know what the design will be" thing, but maybe that's just because they are so far out of their previous experience that they feel like the lessons learned for the production process are worth the wasted production.

Given the diameter and engines, I suppose getting production up makes sense—they can always alter the steel (thickness, type, etc), height of vehicle, etc. My guess is that having employed people full time, making stuff they throw away only wastes the steel, as they are paying workers regardless.

Guess we'll know soon enough how it turns out, lol

[darthgently]

Re-entry, re-entry, wherefore art thou re-entry?

 

 

 

Edited May 30 by darthgently

[tater]

Addressing the heat shield issue which has been discussed here more than once.

[darthgently]

I keep imagining a magical just-in-time aerogel generation system that replenishes an aerogel layer beneath and between the tiles.  Not a hard aerogel, but a "jello" aerogel. Valves at tile vertices and behind tile center (in case of loss of tile) would be simple robust bimetallic temperature sensing valves, no centralized digital control, all independent and decentralized.  Testing on the ground would be a pain sure

Edited May 30 by darthgently

[PakledHostage]

Interesting hint at an interview with EveryDayAstronaut next week in Musk's tweet. Will have to keep an eye out for that.

[Elthy]

Quote

Right now, we are not resilient to loss of a single tile in most places, as the secondary containment material will probably not survive.

Thats not good. Even the shuttle could take a few lost tiles, and hat both a higher ballistic coefficient and an aluminium structure, which has less resistance to heat than the stainless steel. I wonder if the internal pressure of the tanks is to blame, as it could greatly decrease the tolerance for softening of the steel before catastrophic failure...

[PakledHostage]

EveryDayAstronaut (Tim Dodd) has been selected to fly aboard Starship on the DearMoon mission. I wonder how Musk's admission about the heat shield tiles makes him feel? A single point of failure with no abort modes and that (by Musk's own admission) would likely lead to loss of the vehicle is pretty serious.  And while we all understand that space is hard, the criticisms leveled against Shuttle should apply equally to Starship.

[Codraroll]

1 hour ago, Elthy said:

Thats not good. Even the shuttle could take a few lost tiles, and hat both a higher ballistic coefficient and an aluminium structure, which has less resistance to heat than the stainless steel. I wonder if the internal pressure of the tanks is to blame, as it could greatly decrease the tolerance for softening of the steel before catastrophic failure...

 

18 minutes ago, PakledHostage said:

EveryDayAstronaut (Tim Dodd) has been selected to fly aboard Starship on the DearMoon mission. I wonder how Musk's admission about the heat shield tiles makes him feel? A single point of failure with no abort modes and that (by Musk's own admission) would likely lead to loss of the vehicle is pretty serious.  And while we all understand that space is hard, the criticisms leveled against Shuttle should apply equally to Starship.

Valid criticisms if the same solution was used for the final design of Starship, but they are still iterating, and nobody is going to ride on any hardware built today. At the moment, they don't know what the resilience of the thermal tile setup is, so they're testing it. If it's found to be more resilient than the fear, then great. Maybe further testing can certify the current solution for use. If not, the solution has to be changed.

It's not like they find and publicly acknowledge a potentially catastropic design problem during development, then decide not to change anything for the flight hardware. They aren't video game developers.

Edited May 30 by Codraroll

[darthgently]

44 minutes ago, Codraroll said:

It's not like they find and publicly acknowledge a potentially catastropic design problem during development, then decide not to change anything for the flight hardware. They aren't video game developers.

This hits close to home in these forums given recent developments.  I still think given SpaceX's grand vision they'd possibly be interested in acquiring the IP in the name of space engineering education among the youth (old and young) 

[mikegarrison]

3 hours ago, Elthy said:

Even the shuttle could take a few lost tiles

Well, it depended where. Obviously the Shuttle was sadly not able to withstand damage to the reinforced carbon-carbon panels on the leading edge.

[StrandedonEarth]

18 minutes ago, mikegarrison said:

Well, it depended where. Obviously the Shuttle was sadly not able to withstand damage to the reinforced carbon-carbon panels on the leading edge.

 

3 hours ago, Elthy said:

Even the shuttle could take a few lost tiles

The Shuttle could handle a few lost tiles on the upper surfaces, and it could handle damaged tiles on the belly, but losing a single entire belly tile would be seriously risky. Atlantis was nearly lost on STS-27 except for luck. But of course mikegarrison knows all this, I'm sure...

[Deddly]

Shuttle tiles were also a lot smaller, so one missing Starship tile exposes a larger area of unprotected hull.

It really doesn't sound good, considering the previous statements that a few lost tiles should be OK, hence the hexagonal pattern. 

I'm sure there are many ways to get that working, though. With the way they're doing things, they're probably testing cheapest probably-won't-work-but-might-as-well-try options and working their way up to mpre robust solutions. 

[StrandedonEarth]

14 minutes ago, Deddly said:

With the way they're doing things, they're probably testing cheapest probably-won't-work-but-might-as-well-try options

Maybe not the cheapest but the lightest?

[Terwin]

18 hours ago, StrandedonEarth said:

The Shuttle could handle a few lost tiles on the upper surfaces, and it could handle damaged tiles on the belly, but losing a single entire belly tile would be seriously risky. Atlantis was nearly lost on STS-27 except for luck. But of course mikegarrison knows all this, I'm sure...

I rather expect that this is the working assumption until proven otherwise.  It is safer to assume that even with a tougher skin, Starship will be vulnerable to the same sorts of thermal management failures to which the shuttle was vulnerable.

Remember: Musk wans mars colonies, so even if SS could re-enter fine from LEO or even the moon without any tiles at all, SS reentry is not safe until it is safe for SS to reenter from Mars.

[insert_name]

On 5/30/2024 at 3:34 PM, PakledHostage said:

EveryDayAstronaut (Tim Dodd) has been selected to fly aboard Starship on the DearMoon mission. I wonder how Musk's admission about the heat shield tiles makes him feel? A single point of failure with no abort modes and that (by Musk's own admission) would likely lead to loss of the vehicle is pretty serious.  And while we all understand that space is hard, the criticisms leveled against Shuttle should apply equally to Starship.

Looks like DearMoon isnt happening anymore

https://www.msn.com/en-us/money/topstocks/japanese-billionaire-maezawa-dearmoon-mission-cancels-moon-flyby/ar-BB1nqRRm?ocid=BingNews

[Flavio hc16]

7 minutes ago, insert_name said:

Looks like DearMoon isnt happening anymore

https://www.msn.com/en-us/money/topstocks/japanese-billionaire-maezawa-dearmoon-mission-cancels-moon-flyby/ar-BB1nqRRm?ocid=BingNews

I really think it might be very gutting for him. I feel really sorry for him

[Ultimate Steve]

Damn.

Understandable seeing as there reportedly wasn't even a timeline at this point according to one of MZ's statements, but damn all the same.

Was supposed to be a strong starting point towards more common and ambitious and affordable private space missions. Now it ain't gonna start for some many years longer.

[Ultimate Steve]

It just kinda feels like the future has been crashing down lately. Faster and harder than usual. All the things I was excited for in 2019 are being cancelled. KSP 2 is gone, DearMoon is gone, most Aerospace entry level jobs are gone...

[RCgothic]

I feel so bad for Tim. 

[AckSed]

Balls. Well, that's a kick in the space-teeth.

[darthgently]

On 5/30/2024 at 6:34 PM, PakledHostage said:

EveryDayAstronaut (Tim Dodd) has been selected to fly aboard Starship on the DearMoon mission. I wonder how Musk's admission about the heat shield tiles makes him feel? A single point of failure with no abort modes and that (by Musk's own admission) would likely lead to loss of the vehicle is pretty serious.  And while we all understand that space is hard, the criticisms leveled against Shuttle should apply equally to Starship.

Re-entry with SS hasn't really been iterated on yet.  It is still very early in the game.  I'm extremely hesitant to make predictions of failure. 

A cheap process to make pure carbon aerogel tiles or coating could change everything.  Li batteries really shook things up, for example.  None of us would rely so much on smart phones if they ran on NiCads.  As it is they became enmeshed so deeply into our productivity so quickly it is stunning. 

So I see the re-entry issue as mostly a materials science problem and there is a lot of activity there.  Who knows?

Edited June 1 by darthgently

[PakledHostage]

3 hours ago, darthgently said:

Re-entry with SS hasn't really been iterated on yet.  It is still very early in the game.  I'm extremely hesitant to make predictions of failure. 

Sure. But Musk himself said in his tweet that it's a significant problem that they are working hard to overcome. Stating that doesn't mean "predictions of failure" and betting against SpaceX is probably not a good choice, but you also can't brush it aside as just an engineering problem. At every turn, they are finding that they need to add mass to the stack, which eats directly into the payload. Those straws all add up after a while. They need some breakthroughs.  

[darthgently]

2 hours ago, PakledHostage said:

Sure. But Musk himself said in his tweet that it's a significant problem that they are working hard to overcome. Stating that doesn't mean "predictions of failure" and betting against SpaceX is probably not a good choice, but you also can't brush it aside as just an engineering problem. At every turn, they are finding that they need to add mass to the stack, which eats directly into the payload. Those straws all add up after a while. They need some breakthroughs.  

Understood.  I'm just operating under the rubric that solutions are easier to find if one can refrain from dwelling on why a solution might not exist prior to trying to things