Starship is Go...

GW Johnson · 2024-03-15 15:53:56

Quaoar post1750: I honestly don't know for sure, but they've had tiles coming off even in the final suborbital flight tests with Starship alone. It's a recurring problem. All I know is I saw off-attitude effects in the plasma sheath, then lots of tiles coming off, then loss of signal. I just speculated that tile loss led to loss of the forward starboard flap, leading to a tumble and immediate breakup.

Calliban post 1744: the tile thing is an artifact of the space shuttle days. They glued them on, but always lost some. Apparently the airframe structure was tolerant of minor burn-throughs where a tile was lost here and there, something that may (or may not) be true of Starship. Columbia's loss was a hole punched in a carbon-carbon wing leading edge piece, not a ceramic tile loss at all.

In the Apollo days, the heat shield was a hex structure bonded directly to a bottom plate installed as part of the capsule airframe. Something resembling a fancified caulking gun was used to manually fill each cell in that hex with the Avcoat ablative (and micro-balloon- loaded) polymer material. The result was a monolithic heat shield made of a composite ablative material: Avcoat whose char was reinforced by the hex embedded in it. I think the hex was closer to a fiberglass type of material than a paper, but I'm not sure exactly what it was.

Its predecessors were tried on Gemini and Mercury. Different polymers, different hex reinforcement, but the same idea, and the same result: a monolithic heat shield very positively attached to its base plate. In Mercury's case, the heat shield and its base plate dropped down on a sot of impact absorbing landing bag. That was the reason John Glenn came after 3 orbits instead of the planned 7: they thought the landing bag had tried to deploy on-orbit. They had him re-enter with his retrorocket pack still in place, hoping the straps would hold the heat shield in place.

Some of the early unmanned suborbital test flights of Mercury used a solid chunk of beryllium metal as the heat shield. But that never flew orbital. You'll notice they never did that landing bag thing again after Mercury. Too obvious a failure mode, in hindsight.

The first example of Orion that flew not-as-a-part-of-Artemis, on an Atlas-5 I think it was, used exactly the Apollo heat shield scaled up. It had over 380,000 cells to be hand-gunned with Avcoat. That cost a lot of manual labor and money, so for Artemis-1 (Orion's second flight), they tried cast tiles of Avcoat, bonded to the capsule airframe, with gap sealer between them. There was no hex reinforcement.

It eroded significantly faster than expected, and sloughed-off big chunks of char in erratically-occurring locations, further thinning the post-entry remains of the heat shield in those locations, which in turn threatens point burn-throughs. Since there was no hex to reinforce the char from coming loose, I find that outcome entirely unsurprising. But it seems to have surprised NASA and its "old space" contractors who built the thing.

The problem for Artemis 2's delay is not what is being claimed in recent days, it is that they had already built and installed an Artemis-1 heat shield on the Artemis-2 capsule, and they don't want to spend yet more resources and time to take that off and replace it with the "right" design. Most importantly, they don't want to admit they screwed up and installed what turned out to be a bad design, before even testing it.

If Artemis-2 was another unmanned shot, I'd say make the bonded tiles with hex reinforcement, and try that. It should easy enough to use a press tool to push the Avcoat through the hex cells, then put that loaded hex into a tile mold and trowel it up neat. But since Artemis-2 is a manned shot, they simply must replace the faulty design with what proved successful on the first Orion flight: the Apollo gunned hex monolithic heat shield. There is no other ethical course. Not to do so is exactly the same ignorant, arrogant management mistake that killed two shuttle crews. It is becoming rather clear that NASA management did not learn the lessons of those two lost crews.

GW

Last edited by GW Johnson (2024-03-15 16:05:36)

RobertDyck · 2024-03-15 17:07:32

Starship tiles and the pins that mount them.
ZlBSm.jpg&opi=89978449&sa=U&ved=0ahUKEwiSyK7isPeEAxW8FjQIHZT-DgkQ5hMIBQ&usg=AOvVaw3vTQLJMY-pedc2wXCC9xyg

RobertDyck · 2024-03-15 17:22:28

GW Johnson wrote:

You'll notice they never did that landing bag thing again after Mercury. Too obvious a failure mode, in hindsight.

They did it again. Here is the landing bag for Boeing Starliner.

RobertDyck · 2024-03-15 17:25:17

Ps. Orion unmanned orbital test was launched on Delta IV Heavy.

SpaceNut · 2024-03-15 17:42:18

Shuttle tiles were much smaller so there was less chance of one would cause an overheating problem.

Space x did achieve some things after taking care of 17 crucial upgrades helped SpaceX finally get Starship to space without exploding

tahanson43206 · 2024-03-15 19:43:02

For SpaceNut re #1755 ... thanks for the link to the 17 fixes report...

The forum had a vigorous discussion after the April flight .... here is a pertinent item from the 17 fixes report...

For example, SpaceX updated Super Heavy's engine control algorithms and added new, large slosh baffles to its liquid oxygen tank to reduce propellant from sloshing around too much during the booster's challenging flip maneuver, per SpaceNews.

(th)

Quaoar · 2024-03-16 07:23:51

GW Johnson wrote:

Quaoar post1750: I honestly don't know for sure, but they've had tiles coming off even in the final suborbital flight tests with Starship alone. It's a recurring problem. All I know is I saw off-attitude effects in the plasma sheath, then lots of tiles coming off, then loss of signal. I just speculated that tile loss led to loss of the forward starboard flap, leading to a tumble and immediate breakup.

GW

Thanks GW,
but just a little question: as I've understood seeing the video, they firstly trasferred the propellant from the header tanks to the main tanks and then they performed the deorbit burn. So, when the Starship reeenters, she has all the landing propellant (which is about 20-25 metric tons) in the main tanks, where it has a lot of empty space to move, changing the attitude of the ship (as a sailboat skipper I know how roll and pitch can become nasty when there is a lot of water inside the hull). So why not to direct connect the header tanks to the rockets and keep all the propellant inside the header tanks during the reentery?

Last edited by Quaoar (2024-03-16 07:30:24)

tahanson43206 · 2024-03-16 08:05:34

To save time if someone is looking for the moment just before signal loss for the last time:

https://www.spacex.com/launches/mission … p-flight-3

The final moments are at 1:22 or so into the video.

(th)

GW Johnson · 2024-03-16 11:27:21

Quaoar: I'm unsure what they really did with propellant transfer. If it was me, I'd put the remaining propellant into the header tanks, to isolate it from the about-to-get-hot outer shell, sort of like a Dewar. And, it was my impression from the narrative that the on-orbit burn did not take place, the on-board computer did not like something in the data it was seeing, and called off the burn.

everybody: looking at the video a second time, I pretty sure I saw the vehicle dead broadside with its starboard side into the wind instead of its belly, and I'm pretty sure I saw it tail-first at least twice. It's beginning to look like they did not have attitude control, more than they ever had it.

I saw a lot (!!!) of tiles coming off before I saw any of the bad attitudes. The camera was apparently mounted to the forward port side wing flap. If the forward starboard flap departed the airframe, that camera might not have seen it, depending upon the tumbling attitude. Such a loss could have tumbled the ship irretrievably, which was already experiencing extreme and likely fatal pitch and roll problems, near as I can tell.

There is also the possibility that the flaps simply might not be effective controlling attitude in the extreme low density, despite the very, very hypersonic speed. Shuttle did not use its body flap or aerosurface controls until after peak deceleration gees. Attitude control before that point was by thrusters. Possibly for very good reason.

What was clear, was that the two narrators talking to the people seeing the video had no idea anything was wrong in the images we were seeing! But I certainly saw really bad things happening! The second (!!!) time it went completely tail-first, you could see the plasma brighten, as destroyed engine bay items added hot vapors and particles to the slipstream.

I don't know what all the small white flecks were. Some sort of ice particles, they appeared to be. But there sure were a lot of them!

GW

Last edited by GW Johnson (2024-03-16 11:32:41)

RobertDyck · 2024-03-16 13:31:25

White particles could be one of two things. Either ice, or pieces of the underside of tiles. Would thermal expansion cause pins to break off pieces of tile? That would break the tiles loose. The picture I found shows 3 pins for each tile.

Oldfart1939 · 2024-03-16 14:54:24

I just watched this very good video on YouTube that seems to address many of the shortcomings of Starship during the reentry:
https://www.youtube.com/watch?v=Oxb1ySJ … WL&index=2
This is very engineering oriented and the conclusions are very similar to those of GW's.
The commenter stated that the issues seemed to be related to control of the reentry attitude. SpaceX already stated that the vehicle's rolling prevented the in-flight relight of a single Raptor. This video made some good suggestions regarding the use of Inconel in the structure.
This commenter also suggested that the cold gas thrusters are inadequate for control of attitude and should be be upgraded to hypergolics.

Last edited by Oldfart1939 (2024-03-16 15:01:39)

RobertDyck · 2024-03-16 18:59:00

The problem with Inconel is cost. The video didn't just suggest a heat shield of Inconel, but building the whole structure of it. That would dramatically increase cost. And I would have to look up performance at cryogenic temperatures. As GW has told us, 300 series stainless steel gets very strong at cryogenic temperatures. At temperature of densified LOX & LCH4 it's very strong, producing a strength to weight ratio close to carbon fibre composite. How does Inconel perform?

SpaceX has used ideas proposed by other engineers, but SpaceX is turning paper proposals into reality. One proposal was a cylindrical stage with tapered nose with heat shield tiles on one side. I didn't think that was the best design, but Elon is running with it. As for the flaps, I never thought they would have aerodynamic authority required, but suborbital tests got it to work.

Alternativea are a dramatic redesign. One option is to give the upper stage a lifting body shape based on HL-20 and Dream Chaser.

Another is DC-XA, which is a narrow cone with wide base, and heat shield on the bottom. It's bottom heavy and designed to re-enter ass-end first. DC-X and DC-XA had engines recessed with lip of the exhaust cone flush with the bottom heat shield. It also has landing legs that retracted, and extended for landing. Easy for the test article, but an actual orbital vehicle with real heat shield may not be able to land on feet of tile material. Cover door like aircraft landing gear? Or Inconel landing feet?

NASA identified Inconel-617 as metal heat shield material. Alloy:
Nickel 44.5% minimum
Chromium 20.0-24.0%
Cobalt 10.0-15.0%
Molybdenum 8.0-10.0%
Aluminum 0.8-1.5%
Carbon 0.05-0.15%
Others have a max, treated as impurities.

Oldfart1939 · 2024-03-16 21:48:39

Rob-
Inconel at the present usage is would be pretty spendy, but given that in is not being "expended" routinely would make the cost a lot less--if it actually works. I took materials and processes in my university courses, and it was what came to mind almost immediately when the presenter began talking about other alloys. The composition is not from really exotic metals, but is expensive based on the supply-demand curve. My concern would be it's ability to be easily worked and rings fabricated from it?

Last edited by Oldfart1939 (2024-03-17 11:16:28)

Oldfart1939 · 2024-03-17 11:22:37

Regarding the Inconel alloys--they have already been used by North American in the construction of the X-15 rocket plane over 50 years ago. There are Inconel alloys available in powder form for 3D printing of intricate shapes, but my recollection from my college days is they are very difficult to machine; they work-harden and use up milling bits at a prodigious rate. This came from a conversation years ago with a (late) friend who had worked for Pratt-Whitney as a machinist.

kbd512 · 2024-03-17 11:27:10

Stainless steel never approaches the tensile strength of carbon fiber, regardless of temperature. I keep hearing this nonsense spouted off as if it were a fact. It is objectively and provably false.

Austenitic Steels: Mechanical Properties at Cryogenic Temperatures

From the article:

Since we discussed the maximum service temperatures of common austenitic steels in Engineering Bulletin #106, we’ll now look at how mechanical properties of austenitic steels are influenced by cryogenic temperatures and what types of stainless steel alloys are best suited for low temperature applications.
During World War II, experience with the brittle fracture of steel ships caused engineers to look closely at what happens to metals in cold weather. They found that though many metals have good “room-temperature” characteristics, they do not necessarily maintain those characteristics at low temperatures.
For example, Ferritic (405, 409, 430), Martensitic (403, 410, 414, 416) and Duplex stainless steels (329, 2205) tend to become brittle as the temperature is reduced. Fracture can occur, sometimes with catastrophic results. While stretching or bulging may serve as an indicator of impending plastic failure, such signs are absent in the case of these metals. Therefore alloys for low-temperature service are those that retain suitable properties such as yield, tensile strength and ductility.
The austenitic stainless steels such as 304 and 316 retain these engineering properties at cryogenic temperatures and can be classified as ‘cryogenic steels.’ They are commonly used in arctic locations and in the handling and storage of liquid gases such as liquid nitrogen and liquid helium. Liquid helium is the coldest material known with a boiling point of -452°F (-269 °C).
The table below shows mechanical properties of stainless steels at low temperatures. Elongation is an indication of their good ductility. There is an increase in tensile and yield strengths as the temperature decreases as well.
Disclaimer: The info presented here has been compiled from sources believed to be reliable, including the American Society of Materials Specialty Handbook on Stainless Steels. No guarantee is implied or expressly stated here and the data given is intended as a guide only.

Scroll down and look at the section entitled "Mechanical Properties of 304, 321 and 316 Stainless Steels at Cryogenic Temperatures." Look at the yield strength, not the tensile strength. Tensile strength, also known as "Ultimate Tensile Strength", is the point at which complete failure occurs. There are no austenitic stainless steel alloys which approach 100ksi of yield strength at -196C.

Toray standard modulus carbon fiber yields / fails at 415ksi. Their T700 high modulus carbon fiber yields / fails at 710ksi. T800 yields / fails around 852ksi. Some of the strongest steels available, none of which are suitable for cryogenic temperatures, yield between 350ksi and 400ksi. There is no metal alloy that I'm aware of that yields at 700ksi. I'm just shy of absolutely certain that no metal alloys yield at 852ksi, and any that did would be unsuitable for propellant tanks subjected to both tensile and compression loads, never mind cryogenic temperatures.

High Manganese steel has a higher yield and tensile strength than almost any kind of stainless. It does worse on a Charpy V-notch test than stainless at cryogenic temperatures, but it is much stronger and much cheaper, and still fairly ductile at LOX /LCH4 temperatures. Eglin ES-1 steel retains 95% of its room temperature strength at temperatures up to 500C, but yields at over 190ksi at 500C. Cryogenic performance was never a consideration for ES-1, though. There was a similar steel to ES-1, with a small amount of Tungsten added to it, with interesting low and high temperature performance, and even more yield strength.

High levels of Manganese, as well as Nickel and Chromium, prevent the austenitic crystalline structure of stainless from transitioning to martensite at cryogenic temperatures. The very same property which makes Mangalloy and stainless tough and ductile, also prevents them from becoming stronger and harder. Mangalloy is annealed (made softer) when rapidly cooled to cryogenic temperatures, or rapidly heated, to the point that rapid cryogenic "quenching" has the opposite effect that it does on most types of steel.

Martensitic stainless alloys, which do have much higher yield strengths, are also NOT DUCTILE at cryogenic temperatures. Most steel alloys, or at least the ones that I'm familiar with, display greater tensile strengths at cryogenic temperatures, because their crystalline structure transitions from austenite to martensite. The problem is that their ductility all but disappears when that happens.

Oldfart1939 · 2024-03-17 12:02:29

Agreed about the stainless steels you mentioned, but the Inconels are principally Nickel alloys, since Nickel and Chromium together constitute 65-70% of the metals present, and Molybdenum is also a major component ~10%.

GW Johnson · 2024-03-17 12:33:39

The specific alloy used on the X-15 was then known as Inconel-X, and is now listed in Mil Handbook 5 as Inconel X-750. It was intended as a high-temperature material, although one of the Inconels (617?) is also known for decent properties cryogenically. On the X-15, Inconel was used for skin panels and the structures they were attached to. I do not know what the actual propellant tanks were made of, but it used initially LOX and ethanol on the initial small 2-engine cluster of the same 4-chambered engines used in the X-1. When they went to the "big engine" (the XLR-99), that one used LOX and liquid ammonia, which ammonia is both very cold and stored at significant pressure. The X-15A-2 (variant "A", ship number 2) was the one with the twin drop tanks containing additional propellants to reach higher mission energies. You drop those before going hypersonic, which is why shock impingement damage was not seen until the flight with the scramjet article mounted to the ventral fin stub, which peaked at Mach 6.7 and right at 100,000 feet.

GW

Last edited by GW Johnson (2024-03-17 12:35:00)

RobertDyck · 2024-03-17 12:42:39

kbd512,
Now compare yield strength to weight ratio of 304L stainless steel at -206°C to carbon fibre composite at the same temperature.

kbd512 · 2024-03-17 14:52:13

RobertDyck,

Sure, let's do that:

Cryogenic performances of T700 and T800 carbon fibre-epoxy laminates

The temperature dependence of thermal expansion, thermal conductivity and mechanical properties of T700 carbon fibre (T700 CFs) /epoxy composite and T800 CF/epoxy composite were investigated. The mechanical and thermal properties of the unidirectional composite material laminates (0°/90°) at low temperature were studied. The results show that comparing the composite material T700 CFs with T800 CFs, the thermal expansion and thermal conductivity performances of T800 CFs (0°/90°) are all smaller than those of T700 CFs. Typically, the coefficient of thermal expansion (CTE) of T800 CFs in 0° is very low in the temperature range of 120-300K, which reaches as low as -0.4×10-6 K-1. The value of thermal conductivity of this material at 0° is about 3.2 W.(m.K)-1 at room temperature. Tensile and compression tests indicate that the tensile strength of T800 CFs in 0° direction at 77K reaches 2310 MPa, while the compressive strength is about 852 MPa. This composite material may possibly be exploited to design the critical components for practical applications such as hydrogen storage tanks.

77 Kelvin = -196.15 Celsius
2310MPa at 77K = 334.95ksi

Per the link from my Post #1765, 304 stainless yield strength at -196C is 39ksi, and ultimate tensile strength is 221ksi.

When you apply a constant load to a material, if you exceed its yield strength, then it will continue to deform like plastic, until it's been pulled apart. The reason the YS and UTS of stainless are so far apart from each other at cryogenic temperatures, is that stainless will work harden / "become stronger" while it's in the process of failing / "being pulled apart". All structural metals do this to one degree or another, including things like brass and Copper. There may be some weird alloy like Titanium shape-memory alloy that doesn't, but you can't use that to make cryogen tanks / rocket stages. I think what you're doing is confusing yield strength with the short-hand for UTS / "ultimate tensile strength", which is commonly written as "tensile strength". Most materials will yield / begin to fail before they reach UTS. Certain kinds of materials like glass and carbon fibers have very little yield / plastic deformation before they fail. This is the "glass rod" quality GW and I have spoken about before.

Anyway...

At no point in time does stainless approach the yield / tensile strength of T700 or T800 carbon fiber used for Hydrogen storage tanks. Maybe if we're talking about low-cost / low-quality carbon fiber not suitable for Hydrogen storage or aerospace fabrication, then you have something similar in strength at cryogenic temperatures.

Airliners use T700 and T800 fibers in the construction of their wing boxes and skins. T700 fiber was used for the wing skins of aircraft like the F-22 and F-35. F-22 Titanium wing spars were used to decrease cost, not because they were stronger than CFRP for a given weight. Lockheed specifically and deliberately, and with full government knowledge, made the F-22 heavier than it needed to be, purely because machined Titanium forgings were much cheaper than high-modulus CFRP back then, so every 3rd or 4th spar in its wings is made from Titanium instead of CFRP to keep the cost down. I can't recall if T800 fiber existed back then. F-35 had some forged Aluminum wing spars for the same reason. It provided the demanded performance at the demanded cost. Both airframes could be made 10% to 30% lighter if Lockheed "flipped the bird" to airframe cost.

Stainless has an advantage over CFRP at high temperatures only, not cryogenic temperatures, and not room temperature. That is a function of the epoxy resin, however, not the fiber itself. That's why acceptable service temperatures for Reinforced Carbon-Carbon composites are so much higher than for stainless. The "glue" that holds RCC together, unlike conventional CFRP composites, is a ceramic. Nobody uses stainless for brakes. Race cars and heavy aircraft do use RCC for brake rotors and pads.

I hope this explanation of what's going on here helps. I think SpaceX made the right choice in choosing stainless over CFRP for reasons of durability at high temperatures and much lower fabrication costs, but it would be a mistake to think that stainless is anywhere near as strong as CFRP, except at elevated temperatures, because it's objectively not. Stainless steel has the same yield strength as dirt cheap A36 structural steel, but also maintains the toughness of A36 at cryogenic temperatures and greatly elevated temperatures. Stainless is objectively weaker than 2219 Aluminum alloys, in terms of yield strength, at room temperature and cryogenic temperatures. Elevated temperatures are where stainless really shines. 2024 or 2219 or 5083 and 304L are about the same cost, but 304L is easier to weld, about as easy to form, and easier to use in construction because fatigue is much easier to estimate with 304 vs almost any kind of Aluminum alloy or CFRP, where you either need very expensive equipment to empirically make that evaluation or highly sophisticated software analyses. In general, steel is very forgiving to work with. Aluminum is much less forgiving. CFRP is completely unforgiving of manufacturing error and variation.

Edit:
I should've stated that 304 maintains the yield strength and toughness of A36 at greatly elevated temperatures. At cryogenic temperatures, A36 will have a dramatically greater yield strength than 304, but it will also become too brittle, meaning near-zero impact force will shatter A36, whilst the face-centered cubic crystalline structure of 304 will maintain ductility because it will remain in its austenite phase, because it will not change from an austenitic to a martensitic crystalline structure, which is what happens to A36 and most other low and high alloy steels that don't include enough Nickel and Chromium or Manganese.

Last edited by kbd512 (2024-03-17 15:11:51)

kbd512 · 2024-03-17 15:30:33

Now that we're all aware of what stainless, super alloys, GFRP / CFRP / CNTRP, RCC, and other materials bring to the table, the real question is how to get the heat shield cost down to something reasonable, and how to mechanically fasten the ablative tiles or fabrics to the vehicle.

Metals that don't phase change at very low or moderately high temperatures are typically easier to work with than any kind of Aluminum or composite. They will still be a pain to bend and weld, but we have the tech and institutionalized knowledge to work with them. They are only stronger than composites at elevated temperatures, where the epoxy resins of composites begin to fail in a dramatic way.

Composites are, to this very day, more akin to arts and crafts than metal working. They are stronger from deeply cryogenic temperatures to temperatures where Aluminum starts to get soft. After that, they're completely unusable without protection that keeps them at much lower temperatures than off-the-chart reentry temps.

3D printing in metal is also akin to arts and crafts, but we're becoming much more conssitent with the density and precision of the printed parts, to the point that they can offer previously unobtainably low weights.

Standard sheet metals can be and typically are much easier to work with.

Companies like ULA and SpaceX continue to do a lot of work with Aluminum because it offers such a good mix of desirable properties, especially the Copper-based alloys. Aluminum can do a lot for aerospace, but ultimate durability and low maximum service temperatures are its weak points.

Titanium alloys used in aerospace are akin to moderately high strength steels, though very far from the strongest steels, but with less weight, yet they also become silly putty at high temperatures that stainless and super alloys can operate at, also display undesirable characteristics at cryogenic temperatures, and all-around more of a pain to work with. Titanium is also subject to certain kinds of severe chemical attacks, and that is why you see very little Titanium used in shipbuilding, apart from higher costs than Aluminum.

Ceramics and ceramic metals are one area that is relatively new and less well-explored in metallurgy. There are some very promising new ceramic metals available that mix in nano-scale ceramic powders with a base metal, obtaining better mechanical and thermal properties than either high-alloy metals or denser / heavier base metals.

I think reentry temperatures are so high that only low-cost ablatives will maintain feasibility in the go-forward, but we should be open to what ceramic metals can offer.

RobertDyck · 2024-03-17 19:50:48

kbd512,
You really like to hear yourself talk. I didn't say they were the same, I repeated what Elon Musk said, which is they are close at cryogenic temperatures. Now factor in cost of materials. Then add cost of fabrication. Remember stainless steel can be welded outdoors in a yard at Boca Chica, while carbon fibre requires a machine to carefully wind the fibres in a consistent pattern with consistent fibre spacing and consistent fibre tension. Then it requires an oven as large as the tank to cure the composite at controlled temperature. Starship fabrication started with a guy holding an arc welder outside in a yard. They do have machines to weld now, but rings are fabricated in a tent, and assembled into tanks in a high bay with one side open to the outdoors. Total cost of the launch vehicle is not even comparable.

And to make myself clear, I repeat again what Elon said: at cryogenic temperatures (densified LOX and LCH4 temperature, not LH2 temperature) stainless steel 304L is not "that much" different than carbon fibre composite.

::Edit:: Use of carbon fibre composite would also require a re-entry burn for SuperHeavy because carbon fibre couldn't handle entry temperature. And Starship would require much MUCH more heat shield insulation. Stainless steel eliminated that.

Last edited by RobertDyck (2024-03-17 19:57:24)

RobertDyck · 2024-03-17 19:54:37

What you said does tell me to stick to the back-of-the-envelope calculations I did for a Falcon Heavy mission to the Moon. I calculated with assumptions of carbon fibre composite for the crasher stage and propellant tanks of the Dragon trunk, and LOX/LCH4. It would also use the Falcon upper stage as-is. But since numbers barely worked, change to stainless steel wouldn't work.

kbd512 · 2024-03-17 20:47:48

RobertDyck,

If Elon Musk says 10kW is close to 100kW, are you going to repeat that as well?

SpaceNut · 2024-03-20 16:52:08

Video of starship flight
https://youtu.be/8hlXbeQa2ZI?si=ddJvPHweWj_d7Po7

I notice that frost on the booster so must assume same is happening under the tiles for starship.

GW Johnson · 2024-03-22 10:00:37

Fabric or yarn-reinforced composites made with organic polymer binders are porous. Period. There is always void space within the yarns that the matrix cannot fill. They require an impermeable liner to be used as propellant tanks, with that requirement more stringent as the molecular weight of the propellant reduces, simply because smaller molecules fit through smaller leak paths through the material. That adds weight, complexity in manufacturing, cost, and more failure modes to design-around. The X-33 program failed because of that, and because they tried to design to way too low an inert weight fraction.

At cryogenic temperatures, there's not much problem with the physical mechanical properties of the fabric or yarn if it is carbon. Polymer-type fibers get brittle that cold. The polymer matrix does get brittle cold, even if the reinforcement is carbon. This shows up as a reduction in elongation at failure. It's usually fairly dramatic at cryogenic temperatures.

Most metals get really brittle at cryogenic temperatures. Almost no elongation to failure. The well-known exceptions are the austenitic 300-series stainless steels, particularly 304 and 304L. These have a long history of good service even down to liquid helium temperatures, and are the material of choice for tanks and plumbing in most facilities that handle very cold liquids and gases. Use 304 for cast and forged parts, and 304L for things welded up of plate, sheet, and tube. Do not weld 304, it will crack at the weld!

On the flip side, the same 300-series stainless steels are not very strong, compared to many other alloys. They are not magnetic, and they are not heat-treatable, except to anneal. They do work-harden, but it does not take very much heat to anneal them back to the soft state. That means you design only with soft properties, if there is any heat risk whatsoever! Room temperature and cold strengths vary some from alloy to alloy, but they all start falling off a cliff in terms of tensile strength at about 1200 F material temperature. Some will go hotter without incurring corrosion and scaling, than some others (such as 309 and 310 being free of scaling to 1900 F), but they all have just about the same hot strengths. And all of them go to butter-soft much beyond 1200 F.

There is one Inconel that had decent cold properties, which I saw as tubing and fittings in a photo all ice-covered, carrying very cold liquids. I'm not sure, but I think it was Inconel 617. Most of these have better hot physical mechanical properties than the 300-series stainless steels, but they are a tad heavier, and quite a bit more expensive. Inconels are not steels! But they are treated like steels. Most of them are hard to machine: you go through a lot of tools working them. What was called Inconel-X in the 1950's is now called Inconel X-750 in Mil Handbook 5 today. That was what they made the skins, nose tip, and leading edges from, on the X-15. It was also what the hypersonic wind tunnel was constructed from, at UT Austin, when I was a graduate student there in the early 1970's. That tunnel could run hot air up to 2500 F, back then. It's no longer there today.

GW

Last edited by GW Johnson (2024-03-22 10:11:02)

New Mars Forums

Announcement

#1751 2024-03-15 15:53:56

Re: Starship is Go...

#1752 2024-03-15 17:07:32

Re: Starship is Go...

#1753 2024-03-15 17:22:28

Re: Starship is Go...

#1754 2024-03-15 17:25:17

Re: Starship is Go...

#1755 2024-03-15 17:42:18

Re: Starship is Go...

#1756 2024-03-15 19:43:02

Re: Starship is Go...

#1757 2024-03-16 07:23:51

Re: Starship is Go...

#1758 2024-03-16 08:05:34

Re: Starship is Go...

#1759 2024-03-16 11:27:21

Re: Starship is Go...

#1760 2024-03-16 13:31:25

Re: Starship is Go...

#1761 2024-03-16 14:54:24

Re: Starship is Go...

#1762 2024-03-16 18:59:00

Re: Starship is Go...

#1763 2024-03-16 21:48:39

Re: Starship is Go...

#1764 2024-03-17 11:22:37

Re: Starship is Go...

#1765 2024-03-17 11:27:10

Re: Starship is Go...

#1766 2024-03-17 12:02:29

Re: Starship is Go...

#1767 2024-03-17 12:33:39

Re: Starship is Go...

#1768 2024-03-17 12:42:39

Re: Starship is Go...

#1769 2024-03-17 14:52:13

Re: Starship is Go...

#1770 2024-03-17 15:30:33

Re: Starship is Go...

#1771 2024-03-17 19:50:48

Re: Starship is Go...

#1772 2024-03-17 19:54:37

Re: Starship is Go...

#1773 2024-03-17 20:47:48

Re: Starship is Go...

#1774 2024-03-20 16:52:08

Re: Starship is Go...

#1775 2024-03-22 10:00:37

Re: Starship is Go...

Board footer