You are not logged in.
I have since learned about the RL-10B-2 with the extendible nozzle. This nozzle was only used to shorten the engine to fit within the interstage. It was never fired except with the extended bell in place as a vacuum engine.
Most modern vehicles have longer interstages, so the extendible bell was not added to the succeeding RL-10C series. It wasn't needed. No one uses the RL-10 for anything today but an upper stage vacuum engine, and it is very good for that. The olds short-bell RL-10A series sea level engines went away a long time ago.
The other LOX-LH2 engine is the RS-25 series of ex-shuttle engines. These are closer to a "compromise" design that can be fired at sea level, but not really a "sea level design". They are larger thrust level designs than the RL-10 series.
GW
I've already answered that question using the spreadsheets and other "orbits+" course materials. Yes, an SSTO can be built, and if LOX-LH2-powered, it can pretty much equal the performance of a TSTO based on LOX-RP1 in the booster and LOX-LH2 in the upper stage. That answer is already either posted or referenced here on these forums. And I have had it posted for some time on "exrocketman".
That was for all-expendables, by the way. Reusable is quite the different story. It is easy to reuse lower stages. It is very hard to reuse upper stages, and even harder to reuse an SSTO, because it must survive by the same means as the so-far-mythical reusable upper stages. That is neither lightweight nor cheap, no matter how one attempts to do it.
GW
It's not just foreigners. American citizens have been detained just for+ the color of their skin. I suspect they have tried to deport more than one. Pretty soon it won't be skin color, it will be for opposing Trump. Which means me. And I fear for my wife, who was born in Japan, but made a naturalized citizen as an infant. The parallel to the Nazification of Germany in the 1930's is just too eerily close!
I think you are right, Rob. Do not risk coming here. Not until and unless we can get rid of this dictator, and all the minions who support him and keep him in power.
GW
86 47
What you put into the rocket equation is a dV that has been increased to cover gravity and drag loss effects. I have been calling that the mass ratio-effective dV. Those corrections can be rather substantial for launch to LEO.
The other thing the rocket equation requires is a realistic Isp from which an effective Vex can be determined. That is the other reality check: picking an Isp out of published comparison tables is too inaccurate to serve. You must actually do the ballistics of the engine to get a reliable figure.
One of the things the rocket equation completely ignores is thrust versus weight. Your results cannot be realistic until and unless you investigate that, with realistic criteria by which to judge. The normal Earth launch gravity loss is around 5% of surface circular orbit speed, but that is only true if your net upward acceleration exceeds half a gee, meaning takeoff thrust to weight exceeds 1.5! Fail to achieve that, and your gravity loss correction could easily double or triple! Those loss figures are just empirical correlations, by the way.
The other thing the rocket equation completely ignores is whether the engines actually fit behind the stage without sticking out laterally into the stage slipstream. If they fit, then for an aerodynamically "clean" shape of L/D 6 to 10, the drag loss is typically about 5% of surface circular orbit speed. If they don't fit, the extra drag can easily double or triple your drag loss. The same is true if your vehicle L/D falls outside the 6 to 10 range, or the shape has bulges, protuberances, or large changes in diameter.
I know it is complicated. But you have to investigate such items to get reliable results! THAT is the lesson of history here! And it is why I do not use the little on-line calculator offerings, as typically, these crucial items are NOT included in those models.
GW
Every one of those things except the solar power calculator, and more (like entry descent and landing), is in the orbits+ course offering and materials.
GW
That patent was for adapting a simple showerhead injector whose ports were the gas generator throat for a fixed flow design, to the presence of a throttle valve upstream as the gas generator throat in a variable-flow system capable of arbitrary fuel flow commands.
I had to maintain choked flow at the ports of the injector while avoiding flow acceleration on the way to the next ports. The fixed-flow form had a very subsonic Mach number in the injector. For the variable flow form I had to step the ID at each set of ports to maintain a constant high internal Mach near 0.8 to 0.9. The port sizes had to be large enough not to unchoke the throttle valve at its max area, but also small enough to still choke at the min valve area. And it had to do this despite the intensely 3-D supersonic expansions and shock-downs just downstream of the throttle valve throat.
That throttle worked as a side-inserted pintle into a throat blast tube. I played a key role in making that throttle valve item work, too, to include working out the real-world ballistics of a solid propellant device with such a variable throat, which showed why linearized throttle control logic systems always failed (those motors invariably blew up). Those ballistics were unbelievably non-linear in nature. Once we had the nonlinear adaptive gain determined, we never lost another motor!
My injector patent was assigned to Hercules Aerospace, where I worked at the time in their McGregor plant (it had been Rocketdyne Solid Rocket Division when I first went to work there). This thing was verified extensively in live fire ground tests and was ready to fly when the McGregor plant was closed. The JV partner ARC put it into their Coyote drone for the Navy. It was originally intended for a ramjet upgrade to the AIM-120 AMRAAM. USAF never fielded that upgrade (and there is a rather sordid story behind that failure), although the Europeans have fielded their version of the very same idea: Meteor.
-- GW
You can make injection at the throat more feasible by injecting already-hot gas as your throat streams. At comparable densities, the volume blocking effect is maximized. You might get a few percent throat area modulation out of that, but I rather doubt the percentage would be large. Such a concept would need experimental verification in some sort of experimental test. You would also have to face the weight and volume penalties of a device capable of producing a variable hot gas stream on command, of significant mass flow relative to the overall engine massflow.
To the best of my knowledge, nothing like that has ever been attempted. The closest analog would have been injection into the supersonic bell for thrust vectoring effects. It sort-of works, but was not attractive enough for anybody to ever fly it. Gimballing was far more effective for the weight and volume.
GW
There were two variable throat area technologies that I worked on long ago.
One was the rotating "lollipop" for the ASALM variable geometry nozzle option that was not flown. It could have, but the prime did not choose to do that. That is the technology described in the posts just above. It was survivable at ramjet throat conditions, but would not be survivable at rocket throat conditions, where the higher pressure drop would drive larger shear forces that would strip the silica phenolic ablative right off the steel. This also shows up in the fact that monolithic silica phenolic nozzles work well in ramjets, but not in rockets, which needed a hard graphite throat insert instead.
The other was a side-inserted pintle into a throat tube, for the fuel-flow throttle valve of a gas generator-fed ramjet. The fuel-rich solid propellant's effluent stream was the fuel to be burned with air in the combustor. This was a successful technology, and while ours did not fly in a ramjet AMRAAM, something similar to it is flying today in the European "Meteor" missile, which is a very similar gas generator-fed ramjet. Ours did fly in the USN "Coyote" gunnery target drone, after our plant got closed.
It was the relatively cool and reducing environment of the fuel stream that allowed us to make the pintle and the throat tube out of TZM alloy, despite the high solids content and its gritty erosive nature. There was a sort of tower out the side of the throat tube that contained a stack of graphite heat sink rings, which pulled enough heat out of the pintle to allow a gas seal to survive at the cool end of it. This tower's shell was part of the pressure vessel. This thing worked on a transient only, with about a 2 minute capability, at 1700-2500 F gas temperatures, and upstream chamber pressures from 50 to 2200 psia. The cool end of the tower is where the motion servo and measuring gear were mounted. The TZM had a ~4000 F meltpoint, but would oxidize away rapidly if there was any oxygen present.
This was a fuel throttle for massflow control of the solid gas generator burn. It had nothing to do with producing thrust. The smaller the net throat area, the higher the equilibrium pressure, and the higher the flow rate. The flow about the pintle was wildly 3-D, with a wake zone and oblique shocks. The shock-impingement and associated amplified heating upon the throat tube were handled by the TZM, but there was silica phenolic ablative between it and the surrounding pressure shell. This thing led to a showerhead-type fuel injector, which had to be specifically designed for very high subsonic internal flow speeds, in order that the hole sizes would still help direct how much fuel effluent went where. I got the patent for that.
This sort of thing might be at least theoretically used to vary the throat area of a rocket, but the cooling requirements will be enormous. The gas temperatures are well above the meltpoint of TZM, and there will likely be oxygen present at low concentrations, ruling out the use of that alloy. The very high chamber pressures (3000-5000 psia) will stress the thing severely. And there is the wildly-3-D flow field downstream of the pintle. That will really mess up the expansion-for-thrust in the bell. And if the pintle wake does not close before the exit plane, you have the same massive thrust inefficiency problem that I already described for the ASALM lollipop.
The pintle throat was never developed or tested for rocket nozzle application, only the fuel effluent throttle for a fuel-rich solid propellant gas generator (which is a misnomer because of the high solids content in the effluent). The "lollipop" was tested, but only at ramjet conditions. It was developed to the point of readiness for experimental flight test. The pintle was also developed to the point of readiness for experimental flight test, but only at those gas generator conditions, and for a showerhead injector, not a thrust-producing expansion.
Both of these were "one shot" missile designs, not anything to be reusable. These technologies were very difficult development items, even though they were restricted to their less challenging applications.
GW
I have already described how this works or does not work, multiple times on these forums, including in the course materials supplied for the "orbits+" course.
The usual dichotomy is the "sea level" design versus the "vacuum" design. Typical "sea level" designs limit the area expansion ratio at any given chamber pressure to that which produces an expanded exit plane pressure equal to the surrounding atmospheric pressure.
There is no such thing as a "vacuum-optimized" design. There is only the large area expansion ratio that will actually fit behind the stage. These "vacuum" designs are entirely driven by that geometry constraint. Area ratio usually ends up somewhere between 50:1 and 200:1. All of them are under-expanded in vacuum, because you simply cannot expand to zero exit plane pressure without an infinite area ratio.
All of them are overexpanded at sea level, since the exit plane expanded pressure is less than the surrounding atmospheric pressure. Many of them, probably most, are so severely overexpanded at sea level that backpressure-induced flow separation occurs in the expansion bell, leading to its overheat destruction within mere seconds.
Towards the lower end of that "vacuum design" expansion ratio range, there are a few designs that are over-expanded at sea level, but not by enough to cause flow separation. Those actually can be fired at sea level, but thrust there is greatly reduced compared to vacuum thrust, and even for the same throat area and nozzle flow rate, thrust is less than a "sea level design". Yet out in vacuum, thrust is better than that of a "sea level design".
If you can accept the relative shortfall in sea level thrust, these somewhat-limited-expansion "compromise designs" will average higher thrust across the altitudes from sea level to space than a "sea level design", but with a thrust out in vacuum less than, but still fairly close to, most "vacuum designs" that could still fit behind the stage. These "compromise designs" will have the highest ascent-averaged specific impulse, but will offer less sea level thrust for the same throat area and nozzle flow rate than a true "sea level" design.
Only if you can tolerate the reduced sea level thrust, this "compromise design" is the better way to go. If you cannot, then you simply must use the traditional "sea level design".
This design dilemma is why two stage to orbit is the better approach, with the first stage using either "sea level" or "compromise" engine designs, but while shouldering a decided-minority of the total delta-vee to orbit. The second stage gets to use the so-called "vacuum design" engines, while also shouldering the great majority of the delta-vee to orbit. Staging is usually at an altitude (150-200,000 feet) far higher than the altitude where backpressure-induced flow separation occurs (usually below 50,000 feet), such that performance is very near vacuum performance, for the entire second stage ascent burn.
The single stage to orbit vehicle must accept the somewhat-lower ascent-averaged specific impulse of the "compromise design", for the entire delta-vee to orbit, but only if the lower sea level thrust can be tolerated. If that cannot be tolerated, then the single stage to orbit design must accept the significantly-lower ascent-averaged specific impulse of the "sea level design", again for the entire delta-vee to orbit.
That is the tradeoff to be made using fixed-geometry nozzles for your rocket engine designs. You either do that fixed-geometry stuff, or you solve the the leak path risks and accept the extra weight of extendible bells for your single-stage to orbit design, something unnecessary entirely, with two-stage to orbit designs.
The only other possibility is free-expansion nozzle designs, except that those are inherently far inferior in specific impulse out in vacuum, due to extreme streamline divergence angle effects. Those would actually be superior only as the first stage nozzles for a two-stage vehicle design, and even then only if the design's streamline divergence angles were not too large at 150-200,000 feet staging altitudes.
GW
ASALM was a ramjet. When you lean down to lower thrust for efficient cruise, the combustor pressure drops a little, which means the inlet pressure drops a little. When that happens, the terminal inlet normal shock lies deeper down the inlet's divergent diffuser section, where it takes place at a higher Mach number and is therefore stronger with a larger total (stagnation) pressure loss. That lowers cruise Isp a little from the higher level that leaning-down provides.
The sketch showing flow "clinging to the walls" is not quite right as it is drawn. There is a "cling to the walls" effect, but there is also a separated wake closure effect that is not at all depicted in the sketch.
This is because of Prandtlt-Meyer expansions and oblique shock compressions and changes of direction. It will over-expand around the obstruction, then where those flows collide on centerline, oblique shocks straighten the flow back out to axial. The more 3-D the obstruction shape, the more complicated the flow pattern.
If you can achieve obstruction wake closure before reaching the bell exit plane, your nozzle kinetic energy efficiency stays high, although not quite as high as a "clean" nozzle with no throat obstruction to create a wake. If you do not achieve closure before reaching the exit plane, your nozzle kinetic energy efficiency drops precipitously!
The low kinetic energy efficiency multiplies the Mach number squared term in the CF equation, which is where most of your thrust comes from. We verified this in flow visualizations, cold flow tests, and hot firing ramjet tests with Shelldyne-H fuel (also known as RJ-5) and heated air.
Injecting fuel near a ramjet throat has no chance of being successful! You might as well just dump it overboard, because it has no chance to burn. The combustor residence time of 2 to 3 msec is just barely enough to get around 95% of the fuel injected in the inlet burnt. From throat to exit plane the residence time is a tiny fraction of a msec.
Besides, the injected fuel streams will have no perceptible effect on the gas flow pattern through the throat, for two reasons: (1) the fuel mass flow is tiny compared to the air mass flow, even if you inject ALL of the fuel there, which you cannot, and (2) the volume of the fuel stream is exceedingly tiny compared to the volume of the air stream, precisely because the density of the liquid is much higher than the gas, and the mass flow of fuel is so small compared to the air.
GW
I was on the team that did the variable geometry nozzle work for ASALM. It was not about altitude compensation. It was for lowering inlet supercritical margin in leaned-back cruise. We did it with a lollipop in the ramjet throat, that turned streamline for the big throat area, and turned broadside for the smaller throat area.
Key to making it function at a good nozzle kinetic energy efficiency is making sure the lollipop wake closes by the time the nozzle exit is reached. That must be true in both positions, but is more difficult to achieve when broadside. Nozzle kinetic energy efficiency with the lollipop done right is about 94-95%, in a nozzle bell that otherwise would be about 98.3%. Screw that flow pattern up and it drops down nearer 80% or worse.
The throat area modulation was fairly low ratio (about 2:1), and only those two positions were allowed. It only worked at the rather low chamber pressures of the ramjet (under 200 psia). Alloy steel covered in silica phenolic ablative. We did a 900 sec burn in ground test, and it still worked at the end of that burn. This technology never flew. The missile prime determined that the higher cruise efficiency with lower supercritical margin was not worth the extra weight and lowered nozzle efficiency (even though the loss was only from 98.3% to 94-95%).
GW
Basically, for things like helicopters and rotor drones, plus chute openings, and small conventional airplanes, if you can design it to work without using atmospheric oxygen and operate at about 110,000 feet here on Earth, it will work on Mars.
There is an issue of how things scale here. Mass increases as dimension cubed, while area increases as dimension squared, all else equal. Weight to be held up by thrust or lift increases as the cube of dimension, faster than any wing or blade areas, which scale up only as dimension squared. That means it is easier to build very small things that fly than it is to build very large things. And THAT is Ingenuity could fly on Mars, when an electric Bell Jet Ranger probably could never be made to fly in air that thin! The Bell has trouble flying at only 30,000 feet.
Kbd512 is right, it's very complicated. There's more than just aerodynamics or propulsion at work here. That scaling effect is VERY real, too!
GW
The advantage of what I suggested is that every one of those vehicles is already well out of experimental flight test and flying operationally, except for Blue Origin's Blue Moon lander. Better to have only one vehicle to develop than many.
GW
Just remember, when in development flight tests, projected operational payload capacity figures to LEO are speculative guesses, spelled "BS". Once this vehicle starts making survivable landings from orbit, we'll have a better idea what that payload might really be. They have a lot of changes to make yet, to solve a lot of fatal problems they still face. And nobody yet knows what those changes really are.
As a first guess toward tanker flights to effect a full Starship refill on orbit, you need the tank capacity, and the mass of propellant deliverable as payload on-orbit. Starship v.1 had 1200 metric ton propellant capacity. If its operational deliverable propellant payload was 200 metric tons, it would take 1200/200 = 6 tanker flights. If its deliverable propellant is 150 tons, it would take 1200/150 = 8 tanker flights. If its deliverable propellant were 100 tons, it would take 1200/100 = 12 tanker flights. Simple as that!
Starship v.2 has 1300-something tons of propellant capacity. Maybe. We'll see how it really changes in order to survive a test flight. It'll take more tanker flights, at the bigger capacity, presuming the same range of payload figures. And to go to the moon or anywhere else outside LEO, you need to do a full refill on-orbit!
as for the alternate lander and mission:
The alternate mission profile makes good sense to me, except that whoever created that illustration still calls out that crazy lunar halo orbit. It is pretty clear there is not going to be a Gateway station, so why retain an orbit that adds 50% or more to the propellant capacity needed of the lander? (***) If instead you do a relatively low orbit, you can carry more lander payload if you need less lander propellant. But beware, your lander is not reusable unless it is 1-stage, and it is refilled on-orbit about the moon.
*** Answer: SLS/Orion using the interim upper stage cannot send an Orion into low lunar orbit and still get back out to come home. Orion's service module is too small for a capsule that size. You do have a bigger dV getting into low lunar orbit compared to that crazy halo with its periapsis so high above the moon. Earlier mistaken decisions do have a habit of coming back later, to bite you in the rear, do they not?
answer to that *** answer: Delete SLS/Orion as too expensive and too incapable. Do it instead with a Dragon atop a Falcon-Heavy, plus another unladen Falcon-9. Dragon's good for 2 weeks life support, all you need for this. Dock the unladen Falcon-9 upper stage (which has the most unused propellant aboard) to the nose of the Dragon, and use that for trans-lunar injection. Use the Falcon-Heavy upper stage for the return to Earth. You should be able to accomplish that from LLO, not that crazy halo.
Use the extra tank volume left over in the lander that was designed for the higher 2-way dV, for getting into LLO instead of the halo orbit. Then you can still carry more payload than you previously thought to the moon, because the propellant you still have is more than enough for the lower dV requirement. You set your LLO altitude just high enough to make this work, but nowhere near as high as that 3000 km periapsis of that crazy halo.
BTW, that crazy halo was unstable: it had a 70,000 km apoapsis, when the stability limit Hill sphere radius was only 60,000 km. That's what drove periapsis speed to only a snit lower than lunar escape. In turn driving up the 2-way dV requirement to land by more than 50%.
GW
I took a look around at several sites regarding the RL-10B-2 with the extendible exit cone. I did find it was only used as an upper stage vacuum engine. The stowed nozzle allows it to fit in a shorter interstage space, but it is not fired that way! The ONLY data for it are in the extended vacuum configuration. That's not to say a 2-piece extendible nozzle could not be designed and developed, but the RL-10B-2 is NOT one of those!
Most applications have more interstage length available, which is why the other variants, including the newer RL-10C's, have fixed geometry vacuum nozzles. It's only used for upper stage service as a vacuum engine these days. The low-expansion RL-10A's are all long-retired.
GW
1.2 m dia is 0.6 m radius. Tip V = R w, where w is the rotation rate in radians/s. At 2900 rpm, w = (2900 rev/min) * (2 pi rad/rev)/(60 sec/min) = 303.7 rad/s. Tip V = R w = 182.2 m/s. If the speed of sound c is about 240 m/s on Mars, then tip Mach M = tip V/c = 0.76.
GW
Rocket nozzle thrust is F = CF At Pc, where Pc is understood to be the chamber pressure before flow starts necking down into the nozzle throat. Further, it is presumed that area contraction ratio is near 10, so that static pressure and total pressure at the Pc station are indistinguishable.
Here is the most convenient form of the CF equation: CF = (Ae/At)(Pe/Pc)(1 + g nKE Me^2) - (Pa/Pc)(Ae/At), in which Ae/At is the nozzle expansion ratio, Pa is the ambient backpressure, g is the specific heat ratio, nKE is the nozzle kinetic energy efficiency, and Pe is the expanded pressure in an unseparated nozzle, even if it is over-expanded.
There are two terms in CF: the first one (which is actually the vacuum CF) depends ONLY upon the expansion ratio Ae/At, which sets Me and Pe/Pc, for any given g, plus a value for nKE matching the detailed nozzle geometry. Me usually has to be determined iteratively from Ae/At. Once you know Me, you know everything: the compressible formulas are given in terms of Me and g, for Pe/Pc and for Ae/At.
The second term also depends upon Ae/At, but explicitly depends upon the actual values of Pc as well as Pa. This is the backpressure correction term for the retarding effects of being immersed in an atmosphere at some Pa. It is exactly zero in vacuum, because in vacuum Pa = 0 by definition.
The old standard sea level designs always used 1 atm for Pa, and set the expansion ratio Ae/At such that Pe = Pa. The backpressure term is significant: there is always less thrust at sea level, than out in vacuum, all else being equal. These Pe = Pa sea level nozzles have all the thrust you are going to get at sea level, but only somewhat more in vacuum, because the Pe = Pa design limits your expansion ratio rather strongly. It is the Pe = Pa design approach that makes Isp sensitive to Pc, via that backpressure term.
You can use any of a variety of larger expansion ratios at sea level, as long as you do not separate the bell due to excessive Pa compared to Pe. When you do that, the thrust you get at sea level reduces below the old-time sea level design’s thrust, but the vacuum thrust you get with it is larger, because the expansion ratio is larger. That increases the average ascent Isp of the nozzle, flying from sea level out into vacuum, but the cost for that is the reduced thrust at sea level, just at the time when you are the very heaviest at launch, trying to fly straight upward against gravity!
The dilemma is that you are always short of thrust at launch, limited by how many engines of a given thrust will actually fit behind the stage. I cannot emphasize that enough! It is why just “doing the rocket equation” will ALWAYS lead you astray when you do actual design sizing work.
The limit for separation is determined by any of a number of empirical (!!!!) correlations over the years. The one I like is very simple to use, and determines Psep/Pc = (1.5*(Pe/Pc))^0.8333. At any given design with a specific Pc value, you then know the value of Psep. If Pa > Psep, the bell separates (and is usually destroyed in a matter of a few to only several seconds). As long as Pa < Psep, there is no separation, although the backpressure term at sea level does strongly reduce your CF because of the larger Ae/At, and thus your thrust.
The trend in recent years toward higher Pc values is easily explained looking at the thrust equation F = CF At Pc. At the same expansion ratio and CF, higher Pc lets you use a smaller At for the same thrust F! Smaller At is smaller Ae and smaller engine length. More of these higher-Pc engines will fit behind the stage, getting you more thrust at sea level launch when you need it the most.
And, yes, Bob, you can use my spreadsheet to model that extendible-bell RL-10 variant. Run the sea level design and determine the flow rate and throat area. Then re-run the problem with the bigger expansion as a vacuum design, and adjust your vacuum thrust sizing value until you get exactly the same flow rate and throat area as the sea level design. Pick an altitude to do the switch-over. Use the sea level's performance data vs altitude up to that point, then use the vacuum's performance data vs altitude from that point on up to vacuum. Then combine those data into one plot. Do the averaging of Isp across altitude with the same two sets of data, but combined into one set at the switch-over altitude, and that’s a pretty good approximation to the ascent-averaged Isp you would see.
No, it is complicated, you simply don't do this with a single or simple calculation. It takes lots of calculations, which is exactly what the spreadsheet does for you. But you can do it.
GW
The weight of "air" above an area is literally the atmospheric pressure. On Mars that varies with elevation, but seems to be near 6 millibars on the kinds of plains where the Viking landers landed. That's 0.006 bars of pressure, when a bar = 100,000 N/sq.m. So at any location where the pressure is 6 millibar, there are 600 N weight of air above 1 sq.m. At .38 gees and 1 gee = 9.80667 m/s^2, using W = mg, I calculate that to be 23.5 kg mass of "air" above that same sq.m.
Compare that to Earth at 1.01325 bars sea level pressure and 1 full gee of gravity. That's 101,325 N weight of air above 1 sq.m, corresponding to 10,332 kg of air above that same sq.m.
GW
From AIAA’s “Daily Launch” email newsletter for Monday, 6-23-2025:
------
SpaceX traces Starship test-stand explosion to failure of pressurized nitrogen tank
By Mike Wall published 3 days ago (on Space.com)
"Initial analysis indicates the potential failure of a pressurized tank known as a COPV."
SpaceX thinks it knows why its newest Starship spacecraft went boom this week.
The 171-foot-tall (52-meter-tall) vehicle exploded on a test stand at SpaceX's Starbase site late Wednesday night (June 18) as the company was preparing to ignite its six Raptor engines in a "static fire" trial.
A day later, SpaceX narrowed in on a likely cause.
"Initial analysis indicates the potential failure of a pressurized tank known as a COPV, or composite overwrapped pressure vessel, containing gaseous nitrogen in Starship's nosecone area, but the full data review is ongoing," the company wrote in an update on Thursday (June 19).
"There is no commonality between the COPVs used on Starship and SpaceX's Falcon rockets," the company added. So, launches of the workhorse Falcon 9, which has already flown 75 times in 2025, should not be affected.
The Starship explosion did not cause any reported injuries; all SpaceX personnel at Starbase are safe, according to the update. People living around the site, which is near the border city of Brownsville, shouldn't be worried about contamination from the incident, SpaceX said.
"Previous independent tests conducted on materials inside Starship, including toxicity analyses, confirm they pose no chemical, biological, or toxicological risks," the company wrote. "SpaceX is coordinating with local, state, and federal agencies, as appropriate, on matters concerning environmental and safety impacts."
That said, the explosion did damage the area around the test stand, which is at Starbase's Massey site (not the orbital launch mount area, from which Starship lifts off).
"The explosion ignited several fires at the test site which remains clear of personnel and will be assessed once it has been determined to be safe to approach," SpaceX wrote in the update. "Individuals should not attempt to approach the area while safing operations continue."
Wednesday night's explosion occurred during preparations for Starship's 10th flight test, which SpaceX had hoped to launch by the end of the month. (Static fires are common prelaunch tests, performed to ensure that engines are ready to fly.) That timeline will now shift to the right, though it's not clear at the moment by how much.
The incident was the latest in a series of setbacks for Starship upper stages. SpaceX lost the vehicle — also known as Ship — on the last three Starship flight tests, which launched in January, March and May of this year.
Starship's first stage, called Super Heavy, has a better track record of late. For example, on Flight 7 and Flight 8, the huge booster successfully returned to Starbase, where it was caught by the launch tower's "chopstick" arms as planned.
------
My take: if the description “in the nosecone” for the location of the COPV is correct, then it is located very close to the oxygen header tank (as the version 1 with 1200 metric tons propellant capacity was laid out), which is also in the nose of the vehicle, ahead of the “cargo bay” area. Such a COPV explosion would easily rupture that oxygen header tank. Compressed gases drive great explosive violence (with shrapnel) when such vessels burst.
There would seem to be an oxygen header tank transfer piping line down the windward “belly” of the cargo bay section, based on descriptions I have read. In the explosion slow-motion video, the cargo bay splits open through its heat shield, right where that transfer line supposedly is, with gush of something white (not fire) bursting through, followed immediately by an explosion engulfing about the top half of the vehicle, and a second or so later by a second explosion seemingly centered lower down.
The main propellant tanks below the cargo bay would be the main methane tank forward, with the methane header tank located inside, at the base of that tank, and finally the main oxygen tank, just ahead of the engine bay. The upgraded version 2 has a bigger propellant capacity, but should be laid out similarly.
I would hazard the guess that the COPV explosion and bursting oxygen header tank somehow put a large force on the transfer line, which split open the belly at the cargo bay, allowing liquid (and vapor) oxygen out through that split, as well as releasing a few tons of liquid oxygen to fall down on top of the main methane tank.
My guess is that spilled header oxygen and vented methane vapors are much of the first explosion. Bear in mind that the impact of a few tons of liquid oxygen on the top of the main methane tank would rupture it as well, adding some fuel to that first explosion pulse. That first explosion pulse would massively rupture the main methane tank, and also likely the main oxygen tank below it. That’s the second pulse of the explosion, which was larger and longer, reflecting the larger mass of reactants.
All of that scenario is just an educated guess on my part.
As for the nitrogen tank, said to be a “COPV”, or “composite overwrapped pressure vessel”, maybe that is not the right choice this early in the flight test program. Such a design is a metal shell that is simply too thin to hold the pressure, overwrapped by a yarn or fabric-reinforced composite material, to bring it up to strength at a lighter weight.
Here’s the problem: no composite material has a large plastic (post-yield) strain capability. If the COPV over-pressures for any reason whatsoever, failure will be sudden, without any warning! Maybe a heavier all-metal nitrogen tank, one with much more plastic strain capability, would be a better choice until the other bugs all get worked out. At least you could see it stretch before it explodes. You do not want to fly even experimentally, with too many possible failure modes!
Lots of things look good “on paper”, but there are a lot of other things to worry about, many of which cannot be put on that paper. This is where the “older hands”, with many years of school-of-hard-knocks experiences, can be effectively very much wiser than youngsters fresh out of school. SpaceX has no “old hands” on its staff: they hire no one over about age 40 or 45. There’s no gray heads visible anywhere in that organization.
GW
SpaceX is not anywhere near "as close" as most forums participants seem to think. They have serious vehicle design problems to solve with the upper stage Starship, they still have Raptor engine problems to solve (as in pressure/thrust oscillations at a definite infrasound frequency, hidden in the noise hash, which can, and likely did, excite structural or plumbing organ-pipe modes in the vehicle), they still have not demonstrated full survivability upon entry, and they still have not demonstrated how they will land this Starship thing on Earth, much less the moon and Mars, where there are no steel decks, no concrete pads, and certainly no "chopstick" catch towers. And a critical enabling item, if it is ever to be more than just an LEO transport, is how to refill it on-orbit. They don't have that either!
Sorry, but them's the facts! So says one old retired rocket/ramjet/lots-of-other-things hand.
GW
I thought Scott Manley did a good job in that link, primarily saying "we don't know, must wait and see". He did raise some issues with the header tank piping causing the initial splitting-open on the heat shield side. That slow-motion he has is better than anything else I've seen up to now.
It shows cold gas erupting through the split opening up through the heat shield, splitting the thing open in the cargo bay zone before ever there is any combustion. That's a rather important clue! And I don't know what it means, either. And right now, I doubt even SpaceX's team knows. And rightly so.
"Build it, break it, build another" only works if you don't make very many changes from one build to the next. Too many things changed between versions 1 and 2 of the Starship upper stage. None of the 2's has been at all successful yet, while some of the 1's were. I think that outcome makes clear that too many changes were made jumping from 1 to 2, something only easily recognizable after-the-fact. There's no point to "you should have known better" blame games. No one could know better, especially without any "old hands" on the staff to provide any "hard-knocks" wisdom to a crew of newbies. Newbies being pushed to do it too fast by the head honcho (that being Musk, who as both Bob Clark and I have both previously pointed out, is not a qualified engineer in any sense of that word).
Myself, I would not finish ship 37 (a version 2), I'd build another version 1, just putting the revised heat shield and the current version 2 Raptor engines in it. If that should actually fly all the way to an ocean landing, then you know that the trouble lies in the other version 2 changes, and more likely among those made trying to cut down inert weight. You inevitably lose structural margins when you do that.
We also did not see as many upper stage engine problems with Starship version 1 (unlike with the booster), whether it was Raptor version 1 or 2. Therefore I suspect not so much the engines themselves, as I suspect the feed plumbing in the vehicle bringing propellants to them. Yes, the engines were probably too vulnerable to the unanticipatedly-intense engine bay fires from leaks. Leaks less likely at the power heads of the engines themselves, and more likely in the supply plumbing and maybe the tankage in the vehicle design, independent of which version of the engines was being flown. Just wrap that stuff up against the heat of the leak fires. Find and fix that first, and only then try to reduce inert weights without screwing up that prior gain.
500 bars is about 7250 psi. That's high enough to be quite challenging, but it is something attainable. I'm not sure that anything other than SS304/304L has the plastic strain capability to handle system-type problems at pressures like that. That stuff will stretch almost an order of magnitude further before failing, than any other inert gas container material candidate, that I can think of.
Right now, rectifying a failure mode is way more important than saving inert weight. I think most other "old hands" would agree with me about that.
GW
PS -- and even after saying all of that, I still feel (too low a frequency to hear) some sort of thrust/pressure oscillation in the Raptors they are testing at McGregor, including the new Raptor 3's. They haven't recognized that risk factor yet, much less fixed it. So there still are engine problems to fix, as well.
Rob:
Well, I wondered how long that would take. Not very long.
Kbd512:
You are entitled to your own opinions, but not your own facts. I think you may actually be the one in the echo chamber. We don't get to "vote" on facts. They simply are, or are not.
As for Trump, that was an awfully slim "majority" in the popular vote. The electoral college does not reflect the popular vote very well, because most of the states do "winner take all". Whether that's good or bad, opinions differ. But it is a fact.
And what it means is that there is no wide mandate for all the MAGA nonsense being imposed on the rest of us. There never was. Those who tell you there was a vast majority for a mandate are lying to you. Fact.
And the country seems headed down the very same road taken by Germany in the 1930's with its would-be dictator who quickly became one. Fact.
I honestly see no difference in the behavior of ICE and the behavior of Hitler's Gestapo. Opinion. I understand why the agents want to wear masks and no ID badges. They know they are doing evil, and they fear for the safety of their families because of what they are doing. Opinion, but likely true.
That and the rest of the MAGA nonsense is exactly why some 11 million rose up in loud public protest a few days ago. Fact. Most of it was quite peaceful, despite the ballyhooing of the very few violent events. Fact. (Reporting bad news makes more profit -- fact).
Another historical observation is that when 3-4% of the population of a country starts rising up in loud public protest, that country's regime is usually doomed to fall. Fact. We do seem to be headed that way, too. 11/330 = 3%. Fact.
And getting "flamed" for not being a right-wing extremist, is why I so rarely ever visit this thread. Fact.
The only reason I did this time, was to see what Rob had to say about Ukraine and Russia. I thought he did a good job trying to relate it all, rather factually.
GW
Getting accurate reports about this is not as easy as it should be. Some of the things I read early on said it was Starship stacked atop a Superheavy. I did see two explosions about a second or two apart. The video quality was too low to properly identify what blew up. Now most of the reports say it was Starship only, no booster. That's a change in the reporting in only 1 day.
I guess it doesn't matter. This time it was not an engine failure, because none of them were running. I do note that every Starship flown since they went to version 2 has failed in some way. The only 2 or 3 that actually made it down to the Indian Ocean were version 1. Making the changes clearly has not turned out so well.
Now a tweet has surfaced from Musk saying it was most likely a bursting inert gas container that triggered the event. I suppose that could be true. But it sure is early for someone to be saying anything about cause.
GW
Rob:
Your post 2957 summary is excellent, and the details in post 2956 are even more excellent. And you are quite correct: your assessment, truthful and accurate as it is, is at wide variance with what is claimed on (totally unregulated for truth) right-wing extremist social media, and even Fox (Faux) News. I go there rarely any more, because it nauseates me. But last time I went, I estimated 30% of that nonsense was implanted there by the Russians.
Thank you for putting the correct history into fairly-succinct perspective. I am surprised you haven't been "flamed" for this. Last time I tried something similar, I was "flamed". Which is why I rarely visit this thread at all anymore.
GW
This from today's AIAA "Daily Launch" email newsletter:
-----
article image
CBS News
SpaceX Starship upper stage explodes during ramp-up to expected engine test
A SpaceX Starship upper stage exploded in a spectacular conflagration during ramp-up to an expected engine test firing at the company's Starbase manufacturing facility on the Texas Gulf Coast late Wednesday, destroying the rocket in what appears to be a major setback for the Super Heavy-Starship vehicle Elon Musk says is critical to the company's future. Video from LabPadre, a company that monitors SpaceX activities at Starbase, showed the Starship suddenly exploding in a huge fireball just after 11 p.m. CDT, 10 to 15 minutes before the anticipated engine test firing, sending flaming debris shooting away into the overnight sky from a churning fireball that engulfed the test stand.
-----
It's not clear from this that any of the engines were even running.
GW
PS -- After looking at more reports plus a video of the explosion, I can say for sure that no engines were running. I can say it is likely the Starship upper stage exploded first, followed a second or two later by the booster exploding. There are a lot of possible fault trails, but it would appear on the surface that those that might be responsible for this explosion lead to the vehicle itself, not to its engines.