You are not logged in.
I never said aerocapture would not work. I said it was risky. For a couple of reasons I have never seen addressed. Not by anyone, including Zubrin. I'm the trained hypersonic aerodynamicist and heat transfer guy, not him.
The numbers I ran for you were for an umbrella-type fabric heat shield, no matter what the fabric is made of! Since the mass is uncertain, so are the ballistic coefficient and the results I got. That is inherent in where we are with your "big ship" currently.
The risks are getting enough deceleration in one pass, and the higher heat transfer lower down that is required to get that deceleration. That conclusion does ignore the variability of high altitude Martian atmospheric density, which presents yet another risk of not getting the right amount of deceleration. A third severe risk.
Getting enough gees to slow down in one pass, so far would rule out high-altitude passes where heat transfer is low! Period! For ANYTHING larger than about 1 ton at Mars. Lower down, the heat transfer is drastically higher because of density, which rules out most refractory materials in terms of equilibrium service temperature.
Just because a small probe could aerobrake DOES NOT mean a large craft would not incur exactly the risks I have identified! Within the mass uncertainty, I have ALREADY identified those risks to be quite real at large craft sizes! THERE IS NO WAY AROUND THAT!
You can drastically decrease the difficulty of doing this by going to the slower entry speed associated with Hohmann min energy transfer. But you automatically go to an 8.5 month voyage (instead of a 6 month voyage) when you do that. Entry speeds would be closer to 5.4 km/s than 7.5. Convective heating varies as V^3 and dens^0.5, while plasma radiation heating varies as dens^1.7 and V^12.5. Speed makes one hell of a huge difference. And so does density (meaning altitude).
Bear in mind that convective heating varies as Rn^-0.5, while plasma radiation heating varies as Rn^1. There is tradeoff to be made for optimum Rn. You have to minimize both kinds of heating, to minimize the total heating (their sum). It is quite shape dependent, and it works out quite different at Mars than at Earth!
GW
It's a vector subtraction thing, involving both speed and direction. You have 2-body estimates of the transfer trajectory orbit about the sun, and for Mars (or any planet) about the sun. You need the velocity vectors (speed AND DIRECTION) of both at the point where the orbits cross.
Subtract the planet velocity vector from the velocity vector of the craft on the transfer orbit. The result is the relative velocity vector of the craft with respect to the planet. That velocity vector has a magnitude V, its speed. Which is a measure of its energy "far" from the planet before the planet's gravity pulls the craft toward it.
While there is no direction information, it is possible to estimate quite simply the speed "close" to the planet from that "far" speed. You need the escape velocity of the planet at an appropriate altitude for your problem. The Vnear = (Vfar^2 + Vesc^2)^0.5. This is based on conserved PE + KE, with the change in PE added to the far KE to create the near KE.
GW
Aerocapture with a probe does not take place at entry interface speeds near 7.5 km/s. They would be nearer 5.4 km/s, coming off some slower trajectory more like Hohmann transfer. That makes the dV needed out of the aerocapture pass much lower, at something like 0.5 km/s. Spread over 10 minutes, that's down nearer 0.085 gee average, for a peak gee at periapsis in the 0.212 range. It need not dive so deep to achieve that, and the heating is much lower if you are high up. A wild guess might be near 60 km for the periapsis altitude.
GW
Rob:
I'm using older sources of data and models than you. The ones I use here are the ones in my direct-entry spreadsheet model.
For your problem presume: Rn = 30 m (ballpark), average V at peak heating point about halfway through the pass = (7.5 + 4.85 km/s)/2 = 6.175 km/s (very crude but ballpark). You DO NOT want to underestimate it! It drives the heating numbers!
The old Allen & Eggers, Justus & Braun atmosphere model for Mars: dens = c0 exp(-h/hscale) where C0 = 0.03032 kg/m3 & hscale = 8.757 km is supposedly most valid between 25 and 70 km, with an entry interface altitude of 135 km. It's not too far off outside the "valid" range. It gives us density = 3.267 E-6 kg/m3 at 80 km altitude. To do any better, Justus & Braun recommend the "Mars gram" models, which are seasonally and geographically-dependent.
Allen & Eggers: Q/Aconv-stagn = 1.75E-8 (dens kg/m3 / Rn, m)^0.5 (1000 V km/s)^3 = 1.36 W/cm^2
I added a plasma radiation model from SAE: Q/Arad-stgn = 27.94 (Rn, m) (dens/c0)^1.7 (V km/s / 3.048)^12.5 = 102.62 W/cm^2 ERROR!!! 1.02 W/cm2 n(this came from a 1969 edition of the SAE Applied Aerospace Thermodynamics handbook, which is a substantial-sized book, not a "handbook" at all.) The original Allen & Eggers model had only convection, because warheads are suborbital at Earth, producing just about zero plasma radiation heating.
Total stagn heating = conv + rad = just about 104 W/cm^2 ERROR 2.38 !!!, almost all plasma radiation! You need a more conical blunted-cone heat shield shape with a somewhat smaller Rn to reduce that total heating by reducing radiation in favor of some more convection! Typical of Mars. Earth works out the opposite way! It's a tradeoff you must make to find the "right" heat shield shape!
However, at a stagn pt emissivity e = 0.8 (representing a "black" highly-emissive surface), and a Mars-environment sink temperature of 220 K, then with that 104 W/cm^2 ERROR 2.38!!!, I get a thermal re-radiation equilibrium temperature (representing a refractory heat shield with no ablation and no liquid cooling) of 2187 K = 1914 C = 3936 F at the stagnation point ERROR !!! 852 K = 579 C = 1073 F. Tsurf = ((Q/Atot)/(e sigma) + Tsink^4)^.25
ERROR your matrials will do fine at these conditions !!: That kind of surface temperature falls in the ablatives-only range currently! All the alumino-silicate ceramics like Nextel 312 and Nextel 440 suffer a phase change embrittlement at service temperatures down in the 2000 F (1093 C) range. One exposure and upon cooling back below that temperature, they shrink, crack, and become very weak and brittle. They fall to dust if you even just look at them too hard! I have seen this behavior in experiments I myself ran, taking them to just about 3000 F (1649 C). They melt up near 3350 F (1843 C).
There are other refractory ceramics, but almost none are considered ready to apply as entry heat shields, with the exception of the two-layer Tufroc tiles, good to about 3000 F (1649 C). That technology does NOT currently exist in fabrics! Although, the ceramic blankets about the LOFTID inflatable did well at protecting the inflatable in that experimental test. Alumino-silicate fabrics and felt layers wrapped the inflatable, which I presume to have been a silicone-rubberized canvas. The felts provided the low thermal conductivity, while the fabrics physically constrained them in place. I think the windward side was some sort of silicon-carbide fabric atop something similar. All these materials are quite experimental in that context.
ERROR FOLLOWUP: I worked out a peak deceleration gee from a ballistic coefficient presumed to be 500 kg/m^2 and a blockage drag coefficient of 1.5. It turned out to be 0.0127 gee. At a higher Beta of 2500 kg/m^2, it is only 0.00254 gee. You apparently cannot get the deceleration you need at 80 km altitude. You must dive much deeper, which will drive heating up a lot.
I show maybe .245 gees at 40 km, same 6.175 km/s, on that higher B = 2500 kg/m2. But the stagnation point re-radiation temperature is near 4808 K!!! That is because the heating is up, particularly the plasma radiation! About 13 W/cm2 convection, but about 2420 W/cm2 radiation! You need a smaller blunting radius on a more conical shape to get that radiation down.
I put this set of spot check point calculations together as a little spreadsheet, and verified that I got the errors out. I'll send you a copy.
GW
Rob:
I do not have a tool that lets me estimate deceleration and heating during an aerocapture pass. But in direct entry, the peak heating always occurs prior to the peak deceleration, at something like 80+% entry speed, and a significantly-higher altitude. So the stagnation heating will be something "near" direct entry heating.
My cautions regarding aerocapture were two-fold: (1) the Mars atmosphere density, even at entry altitudes, is variable by over a factor of two. I am unsure that anyone knows how to predict it well enough to execute an aerocapture pass. That situation is quite unlike at Earth. (2) I rather suspect that for large objects, the heating rate will be too high to permit more than one exposure to the fabric heat shield material, whether it is carbon or ceramic.
I know about Nexel 440, it is actually quite a similar alumino-silicate fire curtain cloth, to the Nextel 312 that I had a lot of experience with in the mid-1980's. It might go a bit hotter than my 312, but not by much a all, and that service temperature limit has NOTHING to do with meltpoint (near 3250 F), and EVERYTHING to do with a solid phase change that takes place upon heating (near 2250 F for 312, and shuttle tile was limited to 2000 F because of that effect). The material shrinks and embrittles, losing all its strength. It crumbles if you even look at it hard, once cooled back down below that phase change temperature. All those fire curtain materials are good for one, and only one, exposure to a big fire!
I know nothing numerical about the other fabric or inflatable heat shield tests, but I know that the LOFTID experiment had a ballistic coefficient near 25 kg/m2 at 6 m diameter. Square-cube scaling, presuming the same effective density in the core, and the same areal density of the inflatable, says that ballistic coefficient should scale more-or-less proportional to diameter. 10 time the size is 10 times the ballistic coefficient, etc. Your big ship is certainly more than 10 times larger, probably more like a 100 times larger, and with a configuration that is dissimilar to a core riding an inflatable or an umbrella. That throws the density estimates off, so all bets are off regarding a good answer here. But you might be looking at a ballistic coefficient somewhere in the 250 to 2500 kg/m2 range.
Bear in mind that higher ballistic coefficient correlates to higher heating, for direct entries. I have no reason to believe that is not also true for aerocapture passes, just a different correlation. But I cannot get you a number because I do not have the right tool for that. My spreadsheet is for direct entries only, using the simplified dynamics used by H. Julian Allen and A. J. Eggers to estimate re-entry trajectories for ballistic missile warheads back in the early 1950's.
If we had a reliable stagnation heating number, that can be used to estimate a thermal re-radiation surface temperature at the stagnation point, using the Boltzmann radiation equation. Results with that vary noticeably strongly with surface thermal emissivity, since that can vary by a factor of 4 to 10, depending upon what you assume. Surface temperatures are not nearly so sensitive to the actual heating estimate, varying to the 0.25 power of the heating rate.
That's about all I can tell you.
GW
I see reports that static fires of the two Flight 14 stages have taken place. The launch is supposedly targeted for mid-September, and will be fully orbital. To include release of some operational Starlinks. Tower catches were not mentioned in the latest story that I saw.
GW
From what I have read, bone loss and muscle loss were long suspected and long known. They are treatable up to a point with exercise and diets, but not entirely curable. That is why most orbital missions to the ISS or any of the prior space stations were usually 6-9 months long, max. After that, bone and muscle changes start to have permanent effects. Here "muscle" includes the heart.
It is the other things that were more of a surprise, and most of them are still not very treatable, many not at all. Weightlessness affects the immune system, it affects the genetics (the latest surprise, from the astronaut twins), and it affects vision. The longer the exposure, the more likely the effects are at least partly permanent. The genetic changes are permanent!
As far as I know, nobody has yet looked at how reproductive processes are affected by weightlessness. Certainly not in humans or any other primates. We already know it has some weird effects on plant growth.
And NOBODY has ever looked at reduced gee, to see "how much is enough?" That takes a spinning space station. Not necessarily a wheel shape, but spinning for spin gravity.
Given what we do and do not know, the current "prescription" is spin gravity near 1 full gee during long trips, and keep the weightlessness down to a very few weeks to a month or less, depending upon how fit you need to be, to endure entry gees at end of mission. From low circular, that's 4-6 gees, OK for crews weakened by months of weightlessness. For returns from cislunar space, that's 9-11 gees, and those crews were fully fit, not weakened by long weightlessness exposures.
For Hohmann transfer direct entry speeds returning from Mars, it'll be nearer 15-20 gees. For fast trajectory direct entry speeds returning from Mars, it'll be nearer 25-30 gees. NONE of those gee levels for faster returns than cislunar would currently be considered acceptable! It is known from the rocket sled tests that human bodies will actually come apart at about 40-45 gees.
What that really says is that crews returning from Mars need to stop in low orbit, then come home from there! And we DO NOT HAVE the infrastructure for that in place, yet! It takes space stations and space tugs and elliptic capture and departure, and on-orbit refueling without ullage thrust, even with cryogenics.
There's also the Van Allen belts to consider. That's not a problem at Mars or the moon, but it surely is here at Earth! You have to cross them quickly, to avoid a large (likely fatal) dose of radiation! That rules out manned trips with the slow spiraling departures and arrivals inherent with electric propulsion!
Cosmic rays are a relative non-issue. It is large solar flare bursts that are the killer. The shielding for them would look like shielding against Van Allen belt radiation, but you still need to limit the exposure to about a day or two. That still rules out manned spirals through the Van Allen belts, even with solar flare radiation shielding. That's the 15-20 g/sq.cm stuff.
GW
From reports I have seen, the Katalyst reboost vehicle failed by tumbling un-recoverably, despite multiple attempts to regain control. The rescue mission was cancelled, and the SWIFT satellite is now doomed to re-entry.
GW
I see in today's AIAA "Daily Launch" email newsletter two things of interest regarding Starship.
(1) Flight 14 will not feature an attempted tower catch of the upper stage Starship, that has been put off for "a few months" according to Musk. Flight 14 will go fully orbital, though. Reason for no tower catch attempt was not made public, but in the article, suspicion falls on not getting FAA approval to do an orbital flight and a tower catch attempt all in one go. The bugaboo may have been the need to maneuver hard during descent in order not to overfly Brownsville at low altitude.
(2) What Void said about acquiring land in Louisiana for more Starship launch sites turns out to be true. The governor is scheduled to make an announcement about this, 8-25-2026. The site would be near Pelican Island, in the protected wildlife zone there. There is strong opposition to this from local environmental groups, because of the dangers and disruptions the wildlife would have to endure.
If I were Musk, I'd be inquiring whether such a facility might be added to Guantanamo in Cuba.
GW
What I have come up with so far is a fraction over 9 m diameter, for a capsule that delivers 40 metric tons of cargo to the surface, direct off the interplanetary trajectory, and not drop any debris whatsoever. The loaded vehicle would be near 50-60 metric tons. This thing would almost fit a Starship cargo vehicle, fully assembled and loaded, for the trip to LEO.
GW
The density vs altitude trend at Mars is similar in shape but not values. Max heating and max deceleration simply occur at lower altitudes on Mars, closer to the surface. As it turns out, the plasma radiation heating gets quite significant at "typical" Mars entry conditions of around 7 km/s off the interplanetary trajectory, when it is rather insignificant at similar-in-speed 8 km/s Earth orbital entry conditions. It does become quite significant at 11 km/s Earth entry conditions, worse even faster.
Blunter heat shields of the spherical-segment type minimize convective heating but maximize radiation heating. That is why the historical capsules have a spherical segment heat shield shape. Those typically have "nose radius" (radius of curvature) equal to or greater than the heat shield diameter. Mercury, Gemini, Apollo, and now Orion and Dragon, all have that heat shield shape. Blockage area hypersonic drag coefficients are all near 1.4 for those shapes. Bear in mind that larger nose radii reduce convective heating but enhance radiation heating! That’s just the correlation equations for stagnation convective heating and stagnation radiation heating. The reverse is also true: smaller nose radii enhance convective heating, but reduce radiation heating. Those effects are fairly dramatic, but not as dramatic as speed-at-entry.
With the higher plasma radiation contributions, entry at Mars was different. All the heat shields used there so far had blunted-cone shapes. These were pretty near 20 degrees off flat plates (70 degree half-angle cones), and the nose radii were somewhere in the vicinity of ~10% of the diameter. The backshells were heat protected, too, with an entry vehicle shape quite squat compared to the historical Earth entry capsules. These shapes have higher hypersonic blockage drag coefficients, nearer 1.7, which leads to lower ballistic coefficients, also necessary at Mars with its thin atmosphere.
As Earth entry speeds exceed 11 km/s, it may prove necessary to modify the heat shield shape to something resembling more a blunted cone than a spherical segment, because of the quite-significant radiation heating effects. Such entries would be free returns coming back from Mars, direct off the interplanetary trajectory. Such would be a bit over 12.2 km/s off a Hohmann transfer ellipse, and nearer 13 km/s for a faster transfer ellipse with a 1-way travel time in the 4-7 month range.
Bear in mind that higher ballistic coefficient correlates with higher peak heating and higher peak gees, experienced deeper down in the atmosphere, so that the end-of-hypersonics (at Mach 3) occurs at a lower altitude. On Earth, that altitude is still way up high, above 30 km altitude, even for ballistic coefficients as high as 500 kg./sq.m. On Mars, it is quite low, ~10 km even at only 200 kg/sq.m. All the probes so far have been at or under 100 kg/sq.m.
I think that the LOFTID engineers will tell you the same things I did. For only 8 km/s, all they needed was a spherical-segment blunt shape. Even at the lower drag coefficient, with its higher ballistic coefficient, that was enough to get them an order-of-magnitude lower ballistic coefficient with the inflatable, than that of historical capsules. Plasma radiation heating was not a significant problem at that speed. But wake zone heating was, so they had to put insulating blankets over their reinforced-polymer inflatables.
Thos engineers get to run codes for that and get more precise answers. I just run the ancient by-hand models in a spreadsheet. My answers are not as precise, but they are well within the ballpark, and usually fairly close to actual test numbers. Just as they were in the 1950's and 1960’s.
My "spread" of the stagnation convective and radiation heating rates, around the vehicle away from stagnation, is rather crude, but it is still within the ballpark somewhere. Even at very low heating with very low ballistic coefficients, cooling by thermal re-radiation requires surface emission temperatures well above what any reinforced polymer inflatable structures could possibly withstand. All those things fail utterly and start charring between only 250 and 350 F. Of course, the inflatable would burst and deflate if that happened!
The re-radiation temperatures are very sensitive to thermal emissivity estimates, even at those low heating rates, because of the temperature-to-fourth-power dependence of the thermal radiation equation. Making those wake zone structures "black" helps, but probably not enough. Certainly not enough for higher-than-8 km/s entry speeds at Earth, and also certainly not for Mars, where radiation is a bigger contribution, because the wake zone radiation heating is a bigger fraction of stagnation radiation heating, than that of convection. For wake zones, I use stagnation/3 for radiation, and stagnation/10 for convection. The “stagnation values” are different for convection and radiation. They are not the totals! You spread them as separate items. But at any one location, you sum the two up to a total, to do the thermal re-radiation calculation.
GW
Void:
I do not understand from your post whether you are referring to NASA or SpaceX.
Multiple launch pads only does you any good if you have rocket vehicles to put upon them, and they are physically compatible with those launchers (most are not). So my guess (which is only a guess) is that you refer to SpaceX. They have the highest demonstrated launch rate.
The only sites currently active for orbital launch pads in the US are Cape Canaveral Florida (several pads of different types), Brownsville Texas (maybe two pads, once both are converted to launch version 3 Starship/Superheavy, but right now there is only 1), Wallops Island (a single handful of pads compatible with different types of smaller vehicles), and Vandenberg AFB in California (a handful of pads for different launchers, but ONLY suitable for launching polar or retrograde).
As far as I know, there is NOTHING in Louisiana! Or any other state, to the best of my knowledge. Certainly not for orbital operations! Suborbital is different: there are a few others. But those are NOT considered useful, by anyone, for orbital launches of anything! Van Horn Texas and the airport at Mojave California come to mind. There are a single handful of others, but no one has flown anything from them yet, not even suborbitally.
There is air launch to orbit, by a couple of contractors. But NONE of those vehicles fly manned, nor could they. They are too small.
NASA has never even come close to 1 launch per day, but SpaceX has. They have, but only with Falcon-9/Falcon-Heavy. Not Starship/Superheavy! Starship/Superheavy is still deep in experimental flight test, and will be, for a while yet. It is NOWHERE NEAR ready to fly with crews!
NASA's closest approach was about a week between its launch of Gemini 7, and their second attempt to launch Gemini-6, a few weeks after a launch attempt abort with Gemini-6. They were hoping to launch shuttles to rescue stranded shuttle crews within about 1 month. Such WAS NEVER demonstrated!
Sorry! Reality bites! But I try to tell the actual truth, as best I can see it! Based on the actual history of what was actually done.
GW
If you look at about 700 F as quoted for the inflatable membrane interface, plus about 590 F drop through its insulation layer, also as quoted (1000 C), you get a backside surface temperature for the inflatable heat shield of about 1290 F. Which is NOT VERY FAR AT ALL from what I estimated with my crude little by-hand entry modeling technique (about 1300 F)!
Which proves my other point that even wake zone backside surfaces need heat protection, at entry speeds as low as 8 km/s, if they are not 2000-2500 F-capable superalloys.
The front surface quotation was 1400-1600 C (2550 F to 2910 F) is above my little estimate, as was the 9 gees. Silicon Carbide is a refractory re-radiator, not an ablative, so my surmise that their entry averaged a little steeper than the 2 degrees I analyzed, is therefore confirmed indeed! I could probably very closely duplicate their numbers at something in the 3 to 5 degree average angle range, but I am NOT going to waste my time doing it!
My main point is that going to very low ballistic coefficients, which LOFTID successfully did, does NOT relieve you of heat protection requirements during entry. Their 25 kg/sq.m corresponded to around 40 W/sq.cm, not far from what I estimated, and far below the ~200-ish W/sq.cm you get at ballistic coefficients nearer 300 kg/sq.cm.
But the re-radiation temperatures are still quite high! That's the temperature to 4th power variation of the Boltzmann equation for thermal radiation, inverted to yield a temperature that produces a certain re-radiation rate. It means for a factor-12 reduction in heating rate to be re-radiated, your surface re-radiation absolute temperature is crudely only 12^0.25 = 1.86 times smaller!
There is simply no way with any materials technologies that we have, including these new inflatable heat shields, that we are going to be able to delete the need for heat protection, by going to a very low ballistic coefficient. Thermal re-radiation physics and the well-established correlations for entry heating, simply will NOT allow that!
GW
W have yet to see an accurate and complete definition of what the Starship heat shield really is. I am only guessing they are tired of outsiders second-guessing or reverse-engineering what they do.
The tiles are said to be "ceramic", whatever that really means. They are intended not to ablate, but to cool by re-radiation of heat from the hot surface. It does NOT mean the low-density, low-strength alumino-silicate tiles NASA used on the Space Shuttle, or that USAF uses on the X-37B. I think it probable they are a variant on the outer tough layer of the so-called "Tufroc" two-layer tiles used on the X-37B nose cap and wing leading edges. But I do not know, and SpaceX does not say.
The lower layer of Tufroc is weaker and will not stand as high a temperature, but does provide the low thermal conductivity to keep the backside tile temperature down to something an aluminum substructure can withstand, on the X-37B. Starship has a stainless steel substructure that can withstand much higher tile backside temperatures, which my best guess says that is why they only need a variant of that outer Tufroc layer.
The current Starship tiles do not go out as far around the circumference from the stagnation line, as they used to go. There is a smooth black coat or layer that appears to exit from under the tiles and then go out out as far as the older tiles went. SpaceX has said absolutely nothing about what that layer of material might be! I have heard no one else comment about it, either! But it is black, not the supposed white of the backup underlayer ablative coat, that leaves the white streaks when it ablates.
So none of us know very much at all about this iteration of the SpaceX heat shield design, other than it seems to work for shallow-angle Earth entries at about 8 km/s speeds at entry interface.
Long ago, NASA developed a sort of caulking gun that could put a goo in the hole left by a missing Space Shuttle tile. But, NASA management refused to fly it. It would be unlikely to have been able to patch the hole in Columbia's wing leading edge, without also flying some sort of ceramic or carbon fiber cloth. But Columbia's crew never had that chance, with no repair capabilities flown at all.
They were not even told that some NASA engineers feared there was damage to the wing, nor were they allowed to go and look for such damage. And those NASA management decisions are what killed that crew! Plans called for launching a second shuttle to go rescue a stranded crew, but nobody ordered those launch preparations, or told the Columbia crew to ration their life support supplies until the rescue could be launched. One bad decision after another!
Most of the problems and risks that any crew faces are managerial, not technical! I see that nothing has really changed at NASA (or they would not have flown a flawed heat shield on Artemis-2), and I rather doubt Musk's SpaceX believes in proper risk reduction for manned flight yet.
That may be pessimistic, but it is a realistic assessment!
GW
The consensus of opinions and Musk postings "out there" is that they may fly Flight 14 late this month (August). It would appear to be a full orbital mission, with an attempted tower ship-catch back at Starbase in south Texas, if the FAA approves the flight plan.
As I thought, they are not going to risk the tower with a booster-catch, this time around. They need a soft landing (or preferably two) of the booster before they risk that, and that means they must solve the "booster Raptor relight problem", whatever that really is.
I only suspect it may have something to do with the roughly 2000 F air caught in the engine bay and heat-soaking the engines, during booster entry at approximately Mach 5. It would potentially counter the cryogenic cooldown they seem to need, in order to light them. But I may be wrong about that. Just note that nobody has said anything in public about that problem. They may, or may not, understand it yet.
The ship will be flying over populated areas in Mexico and south Texas during its entry, descent, and landing after an orbital flight, which will be of great concern to the FAA in the flight plan approval process. Not to mention the interruption of a lot of airliner flights to and from Mexico. The post-hypersonics "swoosh" maneuver during descent is how they will attempt to fly around Brownsville, instead of over it, as they approach Starbase.
If Shotwell is successful at getting Musk to keep his loud mouth zipped about this flight, they may actually get this plan approved. We will soon see. He does have a demonstrated track record of pissing off the FAA, going all the way back to first very-sub-orbital ship-only flights. She knows that.
GW
The first heat shield exposure a Starship might see is 7.4 km/s entry at Mars, for which it is probably adequate, as designed for 8 km/s entries at Earth. This despite the higher radiation heating seen at Mars than at Earth. Which is why you need a blunted cone heat shield shape at Mars, not the simple blunt spheroidal segment shape most often used at Earth.
Only if you send the Starship back to Earth does the second entry matter! And that is likely at 12-13 km/s speeds, far higher than anything the current heat shield is designed for. And it is the second entry of two, done without refurbishment of any kind, at Mars. Will the current Starship heat shield support a return from Mars? The short answer is "NO".
GW
Harold:
No, I never forget. That was a bad joint design to begin with. I think we have beat that to death before. You DO NOT DO 2 O-RINGS in a solid joint! You do only 1! If it holds at 20-30 psia, it will hold at 2000-plus psi.
Ask me sometime about the Tarter-Terrier gas generator canister. Same stupid problem. Just embodied differently.
GW
What is available "out there" is NOT silicon, but various forms of silica, the oxide of the mineral! It takes enormous energy to convert a pound of that stuff into the pure silicon necessary for combustion with oxygen! This stuff is not mine-able "for free" out there! It is NOT like coal or oil. Or natural gas. Or any other fossil fuel.
It will take something like 3-D printing of the fully-purified silicon, to create the microscopic passages through the mineral form that are required to get "efficient" combustion of silicon Si into silica, the silicon dioxide (SO2). That is NOT WITHOUT SERIOUS ENERGY COST!
And I have already opined on the problem of consensed forms not contributing to thrust and impulse from a nozzle. Liquids and solids are incompressible, and contribute NOTHING toward nozzle thrust production. PERIOD!
I simply do NOT see the "advantage" of trying to go this route. If it ain't "nearly free" for the taking, what is the point? And it ain't!
GW
Manned Orbiting Laboratory (or MOL) was a USAF program aimed at a manned spy satellite capability. It was to be launched by a Titan-3 upgrade to the then-existing Titan-2 launcher, and it consisted of a modest house trailer-sized laboratory and a Gemini-B capsule.
Gemini-B was a revised design with shorter life support and electric power duration, and with the seats displaced angularly more at the heads, so that a hatchway through the heat shield would be accessible by the two astronauts. This hatchway led directly into the laboratory, so that the crew could go into it without depressurizing and going outside. The capsule had a month of life powered-down, but only a few days powered-up.
The mission was a month long, and the lab not reused or recovered. This was the first use of telescopic cameras with folded optical paths, to enable car license plates to be read from orbit. (They can recognize human faces from orbit now.) Most of the lab volume was the camera telescope. It was fairly crowded in there for a crew of 2, but not as crowded as the Gemini itself.
This program began before 1963, along with the X-20 Dyna-Soar, also a USAF program. In 1966, the USAF manned space program was terminated in favor of the NASA manned space program. That killed X-20, and MOL was then in limbo, pending unmanned spy satellite development progress. Gemini-B flew successfully with the hatchway cut through its heat shield, in 1969, I think it was. But by that time, the unmanned spy satellite technologies were ready to deploy and proved to be superior in terms of both cost and lifetime on-orbit. That was the end of MOL. Most of the USAF astronauts went to NASA, and became shuttle astronauts, some by way of the X-15.
The Russians had been doing something similar, called Almaz. If memory serves, it flew 3 times under the name Salyut as the initial Russian space stations. But the Russians found most of the same answers we did: unmanned spy satellites had longer on-orbit life and lower costs. Many of theirs were nuclear-powered. Cosmos 954 that crashed in Canada was one of those.
GW
For void re posting 2357: you have to be very careful about risking a burn-through into a propellant tank, for 2 reasons, either of which could be fatal to the vehicle. (1) The pressurization level of the tank may be part-and-parcel of the strength of the tank to resist crushing by the wind-stream pressures. (2) The propellant feed to the engines' turbopumps, and indeed even the boiloff speed of the propellant inside the tank, depends upon maintaining the right amount of tank pressurization.
It would be into the cargo bay, the engine bay, or the aerosurfaces themselves, that they would have risked burn-throughs by leaving off a tile here or there. Those spaces are already unpressurized to begin with.
GW
For RG Clark re: post 2355 --
Bob, I worked out a model from the data in your posting about the LOFTID test, and got something fairly representative of their blunted cone shape and size and ballistic coefficient. And yes, the shape makes a difference: blunted cones are what has been used at Mars for decades to minimize plasma radiation heating, and those drag coefficients are near 1.7. Spherical radius-type blunt heat shields have been used at Earth where plasma radiation is not as severe, and have lower drag coefficients, nearer 1.4. For the same size and mass, that is a big difference in ballistic coefficient, and thus in peak heating, as well as how deep into the atmosphere it penetrates before slowing.
I got numbers more or less similar to those in the article you linked and quoted, except I got nearer only 6 gees at peak deceleration, and that article indicates something over 9 gees. I used an average 2 degrees below horizontal, they may have come in a bit steeper than that, although steeper does drive up heating, too.
The temperature I got for a re-radiating heat shield with no ablation is close to what they seem to say in the article. I must be somewhere in the ballpark with my little model. However, when I spread heating rates around the vehicle, I got surface temperatures seemingly too high for inflatable fabrics to survive. I think they said something about insulating the inflatables with ceramic fabrics. The problem with equilibrium temperature estimates is that they are extraordinarily affected by the thermal emissivity that you use.
My point: the picture a lot of people share, that you only need protection on the windward side, is false. Even the wake zone sees the same super-high effective plasma temperatures that the windward side sees right behind the bow shock. It's just that the effective heat transfer coefficients are drastically lower. That is why Apollo's (and Orion's) lateral sides also had ablative coatings. The higher speed drove that. Bare superalloy skins survived on Mercury and Gemini, but only because entry speeds were lower.
The inflatables at the lower speeds will need wake zone heat protection, because the inflatable materials have lower max service temperatures than superalloys. By far.
GW
Update 2026/08/07 ... GW Johnson asked that this be added to post #2358
This post is to display a link to a study done by GW Johnson about the inflatable heat shield concept tested by NASA.
https://www.dropbox.com/scl/fi/tzu7fwhd … tkpq7&dl=0
Attached is the pdf file for the latest version of my entry study for the LOFTID inflatable heat shield, that Bob Clark posted about. This version replaces anything I previously sent, being far more complete. If you decide to post it at all, it should go in my post #2358, in the "Starship is Go" thread, under "human missions".
The AI pointed out exactly the same difficulty that I pointed out in my email to you. The silicon oxide is significantly a condensed species that cannot expand in the nozzle, thus not producing any thrust. The AI used the term "two-phase losses" pointing this out. It's not an either-or thing, but it is a significant loss, which is why the Isp estimated for it is actually lower than solids with sea level nozzles, even AN-oxidized solids.
If you plan on accepting an Isp that low, why nor just use the solid instead, and avoid all that silica particle contamination it leaves behind? About the worst offender in a solid's plume might be carbon soot plus some HCl (hydrochloric acid).
Most solids today are AP-oxidized, might or might not include aluminum, and are bound with HTPB or CTPB (not PBAN, it cracks upon soaking out cold to around 0 F). Isp with SL nozzles is near 250 s, and with a vacuum nozzle, can approach 280. And if you stay away from PBAN binder, they can easily be qualified to Mil Std 210 hot and cold soak, which is 145 F down to -65 F. The older 210 standard when I first entered the industry had -65 F, but the hot soak was higher at 165 F.
Propellant soak temperature affects burn rate (higher hotter, slower colder) which affects both chamber pressure and thrust, and burn time. It does not affect total impulse hardly at all. The higher pressure fully hot is what you size the case to take. In tactical sizes that is usually in the vicinity of 2000 psia. In shuttle booster sizes, it was nearer 1000 psia, because steel strength does not scale up with size.
GW
Rob:
The short answer to your question "could the Starship heat shield survive a return from Mars?", is "NO".
GW
I have been in occasional casual conversation with Dr. Ed Pope at MATECH in California, about whether any of his exotic new materials might have promise as larger-than-tiles heat shield panels, for applications like Starship, or for capsule-type heat shields.
There are a lot of issues to resolve before ANYTHING could be in the least practical for such application. There is low thermal conductivity versus strength to endure the loads, there is densification of the surface (only) to prevent infiltration of plasma into the material, there is differential thermal expansion between the cold side and the hot side, and between the heat shield panel and the substrate behind it, there is the issue of how you attach this panel to the substrate, and how to hang onto it as things expand and loads vary strongly. There is the issue of how to survive the burn-through risk when something cracks or breaks. There is the issue of atmospheric moisture condensing and freezing into the porosity of the material, possibly resulting in steam explosions when the material gets heated.
And if I was a real heat shield designer (I am not), I could probably list a lot more! I know just enough to actually ask some (not all) of the practical questions.
My point in this posting is to expose at least some of the VAST gulf between what might be possible, and what is really practical to do. And that gulf changes with time as materials and technologies change.
GW
The technology and materials and procedures for a heat shield that can make hundreds of LEO entries without significant repair simply DOES NOT EXIST YET!
NO ONE has such a thing, not NASA, not SpaceX, not Boeing, not Lockheed-Martin, not Northrup Grumman, not Blue Origin, not anyone! To criticize SpaceX over that, is rather pointless, until and unless such a better technology is actually ready to apply.
The basic science and perhaps some exotic material candidates do exist, but that "state of the art" is decades away from being ready-to-apply, if it ever gets there at all. That little ugly fact of life is too often forgotten by far too many!
What SpaceX does have, is a heat shield that might fly several times, with some repairs made in between, much like we saw with Shuttle, and much like what happens with X-37B. Their backup underlayer is a good thing, something Shuttle and X-37B do/did not have. It appears to work. They have not flown it twice yet, simply because they have yet to recover a Starship for potential re-flight.
Once they start recovering Starships, we will see their heat shield re-flown, and we will (finally) see how much repair is required between flights. If that is less effort than what was required for Shuttle, then yes, they have made a big improvement! So what if it is not yet "perfect"? Whatever "perfect" really is.
The complainers anger me with their lack of proper perspective, and their lack of acknowledgement of the VAST difference between what might be possible, and what is actually practical to do.
GW