totm nov 2023 SpaceX Discussion Thread

[Elthy]

The tiles of the Spaceshuttle would be way more managable with new technology. AI assisted automated inspection after every flight, automated fabrication of tiles for replacement, even odd shaped ones. I sure SS is designed with such stuff in mind...

[RealKerbal3x]

6 minutes ago, RCgothic said:

Starship's stainless hull can mostly survive without them.

This is the big difference. I feel that in this respect SpaceX is trying to learn from the Shuttle's utter disaster of a TPS by introducing more redundancy. There was one Shuttle flight (I think it was STS-41G) where the heatshield sustained severe damage, and the mission was only saved because the underlying aluminium skin was thicker in that area to accommodate the Ku-band antenna. If your spacecraft survives only by sheer dumb luck, then that's bad engineering.

[Guest]

1 hour ago, tater said:

Huh? The Shuttle tiles were each unique. These tiles being tested are the same. They appear to snap on to the surface (I'm thinking my guess was pretty close to right). The target here is operational reuse, so if they had to spend X thousand man-hours per flight recaulking the tiles, etc... SS has failed. The crtical inflection point is cost. If they have some refurb that results in increased costs, it's fine as long as it's still cheaper to operate than other LVs. SS will be a test vehicle for a while, though as they work this out.

Yes, the hull tiles may be the same, but what about the nosecone ones? A surface curved in such a way is much harder to cover with non-unique tiles. The mounting system might indeed make a difference, but if the tiles are as fragile as Shuttle ones, an inspection may still be needed. It wasn't only about the glue. 

They do have the Shuttle as a cautionary tale, so I suppose it's reasonable to expect that they have some way of not making the heatshield ridiculously expensive in practice. However, whatever approach they pick might have problems of its own. Paper designs always look better than real ones, including for the Shuttle (if they picked anything else, we'd be wondering whether this cool side-mounted arrangement could've helped it avoid whatever issues that one developed).

1 hour ago, tater said:

The first SLS will have cost (minus Orion), what, 30 billion? Each subsequent one for a while will cost 2-3 B? How many flights to amortize the dev cost differential? 10? If something like that, I'd expect you're right, maybe 20 times cheaper expendable (300M$) than SLS? Note that I'm talking SpaceX cost here, not what they'd charge, given the cost of SLS, SpaceX would be foolish sans competition to not jack the price up to whatever level makes SLS look like a poor choice, but is still the most  they can get. Charging half or 1/3 of an SLS flight would of course be a steal (more capability, less cost).

First of all, money spent on SLS is sunk cost. Development money doesn't come into play here, because it's already been spent. We can take 2-3B cost per launch for SLS versus whatever SpaceX can afford to charge for Starship. We also assume that Starship will be able to be human rated with minimum fuss, and that its payload capacity will live up to Musk's promises, neither of which is a certainty. 

7 minutes ago, RCgothic said:

Raptor is a flown engine.

Only the SL version, and even then, knowing SpaceX they'll continue working on the design while Starship is being built. Well, we know that methane engines work, and that Raptor is probably good, but that leaves the rest of the ship. As an orbiter, Starship is far more than just a tank with an engine. About the heatshield, for example, we know that it doesn't fall of the rocket, and not much else.

[tater]

4 minutes ago, Dragon01 said:

Yes, the hull tiles may be the same, but what about the nosecone ones? A surface curved in such a way is much harder to cover with non-unique tiles. The mounting system might indeed make a difference, but if the tiles are as fragile as Shuttle ones, an inspection may still be needed. It wasn't only about the glue. 

Well, if SpaceX decides to put a large part on top of Starship that rains down destructive debris on the TPS they might have a problem.

I suppose they might hit a seagull on the way up now and again. Wonder what the max alt of birds is vs SS velocity profile? They'd probably need to hit a bird near Max Q for it to be a real damage risk to the TPS.

 

4 minutes ago, Dragon01 said:

First of all, money spent on SLS is sunk cost. Development money doesn't come into play here, because it's already been spent. We can take 2-3B cost per launch for SLS versus whatever SpaceX can afford to charge for Starship. We also assume that Starship will be able to be human rated with minimum fuss, and that its payload capacity will live up to Musk's promises, neither of which is a certainty. 

Yeah, I will look at total program costs over some number of flights, and divide by flights.

I have yet to assume crew rating on SS at all, I tend to consider it exclusively as a cargo vehicle, I'll believe people on it when I see that happen.

The payload capacity is just math. Different people can do the math, and have, and they get similar sorts of numbers.

 

4 minutes ago, Dragon01 said:

Only the SL version, and even then, knowing SpaceX they'll continue working on the design while Starship is being built. Well, we know that methane engines work, and that Raptor is probably good, but that leaves the rest of the ship. As an orbiter, Starship is far more than just a tank with an engine. About the heatshield, for example, we know that it doesn't fall of the rocket, and not much else.

Fly, iterate, fly, iterate.

They'll figure it out, or they won't.

[Guest]

Just now, tater said:

The payload capacity is just math. Different people can do the math, and have, and they get similar sorts of numbers.

There's stuff that goes into the math that we do not yet know. Most notably, dry mass of the finished Starship is not known. Some system or another may end up heavier than anticipated, there are also structural considerations. If a part turns out to be too weak during testing, it will have to be strengthened, and the payload goes down.

4 minutes ago, tater said:

I suppose they might hit a seagull on the way up now and again. Wonder what the max alt of birds is vs SS velocity profile? They'd probably need to hit a bird near Max Q for it to be a real damage risk to the TPS.

If it's anything like the Shuttle, a bird pecking on it could cause damage. Shuttle tiles could reportedly be broken by hand, and without much effort, either. In orbit, there's MMOD damage to concern with, not to mention stresses that might induce cracks in the ceramics. Heat, vibrations, dynamic pressure... These things tend to be fragile, and yet have to take quite a beating each launch. Shuttle tiles wore out from that, not from ablation.

10 minutes ago, tater said:

I have yet to assume crew rating on SS at all, I tend to consider it exclusively as a cargo vehicle, I'll believe people on it when I see that happen.

If we're talking a cargo-only Starship, that put it in a different category than SLS. After all, the latter is meant to be human rated from the outset. This is a major consideration, and also a major cost factor. 

[sevenperforce]

1 hour ago, Dragon01 said:

The Starship you're all imagining is a paper rocket. You're comparing it to SLS, which uses flown components (it's basically a Shuttle launch stack with a Delta IV second stage stuck on top).

Sigh. We've been over this many times. Jupiter-DIRECT was a Shuttle launch stack with (or without) a DCSS on top. SLS uses a different core stage (because the ET wasn't strong enough) and different SRBs. Ares 1 was less of a paper rocket than SLS at this point.

1 hour ago, Dragon01 said:

Once Starship uses actual, flown components (read: once it flies), then you will be able to compare them. 

Raptor is an actual, flown component. And it's the whole rocket. Take any rocket in existence, swap the first-stage engines for an appropriate number of Raptors, and the one with Raptors wins in raw payload capacity.

1 hour ago, Dragon01 said:

Yeah, they are hexagonal and Shuttle had square ones. Sure, that might help.

The shuttle didn't have square tiles; it had 21,000 differently-sized tiles that all fit together like an intricate jigsaw puzzle. 95%+ of Starship tiles will be interchangeable.

1 hour ago, Dragon01 said:

Let's face it, in both cases, these are ceramics.

Why can I put my dutch oven in the oven at 550 F but not my coffee mug? Let's face it, both are ceramics.

1 hour ago, Dragon01 said:

Why would Starship tiles not require disassembling and detailed inspection after every flight? 

Because Starship tiles endure less punishment and are not stuck on with glue that degrades.

41 minutes ago, Dragon01 said:

The mounting system might indeed make a difference, but if the tiles are as fragile as Shuttle ones, an inspection may still be needed. It wasn't only about the glue. 

Correct. It was also about foam strikes from being side-slung. Also not an issue.

12 minutes ago, Dragon01 said:

Shuttle tiles could reportedly be broken by hand, and without much effort, either.

Because they were thick and multilayered, to protect the fragile underlying aluminum that Starship doesn't have.

[Guest]

16 minutes ago, sevenperforce said:

Sigh. We've been over this many times. Jupiter-DIRECT was a Shuttle launch stack with (or without) a DCSS on top. SLS uses a different core stage (because the ET wasn't strong enough) and different SRBs. Ares 1 was less of a paper rocket than SLS at this point.

I should remind you that the core stage had already been made, and is sitting in storage pending NASA reopening their facilities. SLS is not exactly Shuttle launch stack, but these changes are incremental. Starship's very concept is unproven, and the only thing remotely close it to was a financial failure.

16 minutes ago, sevenperforce said:

Raptor is an actual, flown component. And it's the whole rocket. Take any rocket in existence, swap the first-stage engines for an appropriate number of Raptors, and the one with Raptors wins in raw payload capacity.

First of all, Raptor is a methalox engine. If you swap any existing rocket's engines for Raptors, it won't go anywhere, because it won't like being fed kerosene (or LH2, for that matter). 

The main reason is so great is that methalox is such a great propellant mix. If you converted it to run on kerosene, performance would not be that good.

16 minutes ago, sevenperforce said:

The shuttle didn't have square tiles; it had 21,000 differently-sized tiles that all fit together like an intricate jigsaw puzzle. 95%+ of Starship tiles will be interchangeable.

How do you cover a rounded nosecone with interchangeable tiles? Here's a hint: you don't. It will need a lot more than 5% unique tiles.

16 minutes ago, sevenperforce said:

Why can I put my dutch oven in the oven at 550 F but not my coffee mug? Let's face it, both are ceramics.

A ceramic coffee mug should be fine at 300 C, if that's what you're asking, at least if the heating is even. Yes, they are both ceramics. It's not particularly high temperature for this class of materials. Whatever it's painted with may have trouble if it's a cheap mug.

Ceramics are diverse, but have a few things in common. For one, they are brittle. They're also relatively heavy. Heatshield tiles need to be lightweight, so they sacrifice mechanical strength to bring mass down. There's no way around it, except using a different type of material.

Edited April 30, 2020 by Guest

[tater]

30 minutes ago, Dragon01 said:

There's stuff that goes into the math that we do not yet know. Most notably, dry mass of the finished Starship is not known. Some system or another may end up heavier than anticipated, there are also structural considerations. If a part turns out to be too weak during testing, it will have to be strengthened, and the payload goes down.

Absolutely. The dry mass of SS is a moving target.

They have a good handle on the prop reserve required for EDL, and they have a ballpark of what the lower limit constraint might be. The actual final mass will be somewhere in between.

The target is 150t with an improved SS later down the line, so they have quite a bit of room on mass. If they want a 100t dry mass for 150t payload, then if it comes in at 200t, they only get 50t to LEO.

 

Quote

If it's anything like the Shuttle, a bird pecking on it could cause damage. Shuttle tiles could reportedly be broken by hand, and without much effort, either. In orbit, there's MMOD damage to concern with, not to mention stresses that might induce cracks in the ceramics. Heat, vibrations, dynamic pressure... These things tend to be fragile, and yet have to take quite a beating each launch. Shuttle tiles wore out from that, not from ablation.

TPS is non-trivial, which is why reusable orbital stages are hard. Their design seems better than Shuttle's.

 

Quote

If we're talking a cargo-only Starship, that put it in a different category than SLS. After all, the latter is meant to be human rated from the outset. This is a major consideration, and also a major cost factor. 

SLS never should have been crew rated in the first place. SLS is saddled with Constellation legacy. Orion is the Constellation MPCV, and it has actually bloated from that design. It's bizarre, because the whole point of the MPCV was that it was NOT to sit on top of Ares V (the analog to SLS).

In a world with SLS flying, and SS flying as a cargo LV, any lunar or Mars architecture not using SS—just as a cargo delivery to LEO, forgetting all other capability—would be idiotic. The only role for SLS would then be to deliver crew on Orion (because it can do reentry from the Moon or Mars return)—and that delivery would be to LEO (meaning no need for SLS at all).

 

Edited April 30, 2020 by tater

[Ultimate Steve]

We are talking about Starships tps like it is the final design. I can almost guarantee you that it isn't.

[Shpaget]

1 hour ago, Dragon01 said:

Again, how do you know the projected costs are accurate (hint: they never are)? It doesn't matter, expandable or not. How do you know that, when you account for everything, Starship's costs do not end up ballooning, and that payload capacity won't end up dropping due to unforeseen weight increases? 

No one expects SS projected costs to be dead on, however Elon Musk has an entire team of experienced engineers and accountants that came up with a number. They did not pull the $2M out of their posterior. They came up with the number by doing a thorough analysis. Will the actual final number be higher? Quite possibly, but they have so much room that it does not matter.  If that number turns out to be wrong by an entire order of magnitude, they are still two orders of magnitude cheaper than projected SLS cost (which is also subject to cost ballooning, as we have already seen). Even if SS costs SpaceX $200M to launch (off by two orders of magnitude) they can charge a cool billion and still be an entire billion cheaper than optimistic SLS prediction, while pocketing $800M. The $2M number is important if they actually want an Earth to Earth passenger service (replace airlines), but for cargo to space, the reasonable price per launch customers are willing and able to pay can be much higher.

[sevenperforce]

2 minutes ago, Dragon01 said:

First of all, Raptor is a methalox engine. If you swap any existing rocket's engines for Raptors, it won't go anywhere, because it won't like being fed kerosene (or LH2, for that matter). 

The main reason is so great is that methalox is such a great propellant mix.

Then why isn't it already being flown on legacy rockets?

2 minutes ago, Dragon01 said:

How do you cover a rounded nosecone with interchangeable tiles? Here's a hint: you don't. It will need a lot more than 95% unique tiles.

I think you meant 5%.

But no matter. Starship has a diameter of 9m; its cylindrical section is 37.1m and its ogive section is 11.7m. The heat-shielded portions of the fins are roughly half the total body shield area. So the shielded surface area of the (roughly conical) nosecone is 11.3% of the total shield. More than half of that can use the same hexagonal tiles. QED.

2 minutes ago, Dragon01 said:

A ceramic coffee mug should be fine at 300 C, if that's what you're asking, at least if the heating is even.

Heating is not even if I am, say, braising a roast.

2 minutes ago, Dragon01 said:

Ceramics are diverse, but have a few things in common. For one, they are brittle. They're also relatively heavy. Heatshield tiles need to be lightweight, so they sacrifice mechanical strength to bring mass down. There's no way around it, except using a different type of material.

Or using a really thin tile...which you can do if your backshell can take temps up to 1000 K.

[Ultimate Steve]

Eric Berger is saying on Twitter that SpaceX bid Starship for the Artemis lander program. If true, and I hope I'm wrong, that is a surefire way to not be chosen.

[IncongruousGoat]

3 minutes ago, Shpaget said:

If that number turns out to be wrong by an entire order of magnitude, they are still two orders of magnitude cheaper than projected SLS cost (which is also subject to cost ballooning, as we have already seen).

If that number turns out to be wrong by an order of magnitude, they're still cheaper than nearly every other launch vehicle flying today, in cost per launch as well as per-ton. So yeah, assuming that the $2M number comes from somewhere real, there's a lot of margin for error that'll leave Starship as a viable source of income for the company.

[RCgothic]

SpaceX selected. They bid Starship.

Boeing not selected.

 

Edited April 30, 2020 by RCgothic

[Ultimate Steve]

Well that was unexpected!

Edited April 30, 2020 by Ultimate Steve

[IncongruousGoat]

Wait, seriously? They bid Starship as a crew lander and got selected? How in the heck did that happen?

Well, all I can say is that I did not see that one coming.

[sevenperforce]

2 minutes ago, RCgothic said:

SpaceX selected. Boeing not.

 

BESPOKE MOON LANDING STARSHIP

LANDER WITH SIDE-MOUNTED ENGINES

Super cool.

Just now, IncongruousGoat said:

Wait, seriously? They bid Starship as a crew lander and got selected? How in the heck did that happen?

Well, all I can say is that I did not see that one coming.

Might not be a crew lander.

Wait, I'm wrong.

 

Edited April 30, 2020 by sevenperforce

[sh1pman]

What did they do to Starship? It doesn’t look like it’s going to return to Earth.

[sevenperforce]

Three stages!

My guess is that the landing stage provides transfer from LOP-G all the way to the lunar surface, the ascent stage gets back to LLO, and then the reusable crew capsule goes back to the LOP-G by itself. Pretty much the exact design I suggested a few months ago, although probably with an extra engine.

Just now, sh1pman said:

What did they do to Starship? It doesn’t look like it’s going to return to Earth.

Dedicated moon landing version.

[tater]

Who would want Starship for anything?

All 3 are good designs, actually. Of the 2 more traditional designs, I like Dynetics.

 

[sh1pman]

On 4/29/2020 at 5:46 AM, SOXBLOX said:

What I mean is that there are no missions for Starship to fly.

Well, that didn’t age well, did it?...

[sevenperforce]

"Starship is a fully reusable launch and landing system designed for travel to the Moon, Mars, and other destinations. The system leans on the company’s tested Raptor engines and flight heritage of the Falcon and Dragon vehicles. Starship includes a spacious cabin and two airlocks for astronaut moonwalks. 

Several Starships serve distinct purposes in enabling human landing missions, each based on the common Starship design. A propellant storage Starship will park in low-Earth orbit to be supplied by a tanker Starship. The human-rated Starship will launch to the storage unit in Earth orbit, fuel up, and continue to lunar orbit."

I spy oval-shaped engine exhaust ports on the sides. I wonder if those are SuperDracos. It would avoid the "digging up the ground" landing issue. Use those engines for touchdown and liftoff; use vac raptors for the major dV budget.

[sh1pman]

 

[RCgothic]

Incredible validation for Starship right here. NASA thinks it will work. Including orbital propellant transfer.

And that's me out of likes for today!

 

[Ultimate Steve]

1 minute ago, sh1pman said:

 

 

3 minutes ago, sevenperforce said:

 

I spy oval-shaped engine exhaust ports on the sides. I wonder if those are SuperDracos. It would avoid the "digging up the ground" landing issue. Use those engines for touchdown and liftoff; use vac raptors for the major dV budget.

Well assuming that picture is accurate, there's your answer!