Kudos to SpaceX for Flight 12

Update 5-28-2026:  I corrected an error in Figure 1.  "Belly" should have been "dorsal".

------   

I have watched the SpaceX website video a couple of times now,  and seen some other things from other sources.  This test flight was mostly successful,  especially the performance of the heat shield.  I saw nothing to indicate any hinge line burn-throughs at the aft flaps,  seen in some of the earlier test flights.  And the heat shield looked to be in still-usable condition at the time of the test flight splashdown. 

This flight achieved most,  but not all,  of the intended objectives.  The two main shortfalls were the Superheavy booster boost-back,  and the engine relight test aboard Starship in space.  Plus,  Starship flew its mission with one engine out.  The Superheavy was prevented from doing its boost-back by the loss of almost all its engines.  There was only one still working at its downrange splashdown,  which appeared to hit at about the speed of sound.

The proximity in time of that engine-out on the second stage Starship and all the lost engines on the Superheavy first stage booster,  to the hot staging event itself,  raises the possibility these problems are related somehow to that hot stage event.  This author is not an insider to SpaceX,  so he does not know that to be true,  but the close timing is very suspicious.  It is a good place to start looking. 

Previous flights of the Version 2 configurations that included hot staging,  did exhibit some degree of upper stage rocket blast damage to the grid fins on the lower stage.  Those boosters had 4 grid fins,  equally-spaced around the circumference.  The chance was pretty high that one grid fin might see some rocket blast during the hot stage event.  Unlike the propellant tank walls,  these grid fins were not cooled by contact with cold propellant vapors.  Rocket blast damage happens very rapidly to uncooled structures.

For version 3,  there were changes to both stages.  Both were fitted with the new,  higher-pressure Raptor-3 engines,  including 3 vacuum Raptors in the second stage Starship.  The first stage Superheavy had 3 grid fins 90 degrees apart,  each about 50% larger than before,  but not equally spaced!  The “missing fin” spot on the Superheavy booster lined up with the mid-line of the belly heat shield on the upper stage Starship,  in pre-launch images. 

Hot staging involves starting the upper stage engines while still in contact with the lower booster stage,  which in turn is still slightly-thrusted to keep the propellants down in the aftermost ends of the tanks where the pump suctions are located.  The upper stage has to accelerate away from the lower stage at a higher acceleration than the lower stage can achieve,  even without the heavy upper stage still attached.  If not achieved this way,  they will collide catastrophically. 

The lower stage acceleration still has to be enough to keep the propellants settled in the tanks,  so that the lower stage engine pumps can maintain a suction on only liquids.  Drawing vapor into the pumps of an operating rocket engine is also very catastrophic.  

That complicated thrust balance is why something like only 3 to 5 engines,  out of 33 on the booster,  are all that are used during hot staging.  These are also strongly thrust-vectored to help produce the flipping action that points the stage back toward the launch site for the boost-back event.

The rocket blast from the upper stage hits a piece of armor atop the forward tank of the lower stage,  otherwise that rocket blast would likely burn holes into that propellant tank,  another quite likely-catastrophic event. 

The pushback from that upper stage rocket blast cannot exceed the thrust of the few lower stage engines used during the staging event,  or the lower stage gets accelerated net rearward,  with the propellants moving to the forward ends of the tanks!  That sends vapors instead of liquids to the lower stage engines,  which is likely catastrophic,  in that their turbopumps likely will explode,  the same way an aircraft jet engine comes apart if it swallows too big a bird,  or swallows something truly hard of any significant size at all.

This pushback can even be directed more to one side than the other,  by varying the open areas of the interstage,  circumferentially.  That way the upper stage rocket blast pushback force also helps the vectored booster thrust to flip the booster around toward the launch site faster.  But not too fast! 

If the flip-around spin rate is too high,  the propellant in the booster forwardmost tank (in this case liquid methane),  gets flung forward,  letting the booster engine turbopumps suck vapor methane instead of liquid.  Leading again to turbopump explosions.  Catastrophic!

In other words,  if you do not do hot staging “right”,  within some rather narrow limits,  then truly bad things will almost inevitably happen!

Based on previous grid fin rocket blast damage,  and the “missing fin” location in the version 3 Starship/Superheavy,  I expected to see the stage do its flip in the plane of the missing grid fin,  so that the other grid fins are farther from the upper stage rocket blast during the flip,  and thus avoid damage to them.  This is depicted as the “EXPECTED” geometry shown in Figure 1.  

What I saw in the video was the “OBSERVED” pattern in the figure:  a flip just about 90 degrees away from the direction I expected to see!  I have to conclude that the hot staging event did not happen according to plan! 

And so I must also wonder if that deviation had anything to do with the lost engine on the upper stage Starship,  and with the multiple engine failures seen on the Superheavy booster as the flip proceeded!  Not keeping the propellants properly settled in the after ends of the tanks would very probably cause engine failures on the Superheavy,  which is exactly what we all saw in the video.  I am less sure about the engine-out in the Starship.

Figure 1 – Author’s Observations About Hot Staging on Flight 12

Now the Raptor-3 version of these engines,  whether sea level or vacuum,  operates at a higher chamber pressure,  and a significantly-increased thrust level,  than the previous version 2 Raptors used in earlier test flights.  The earlier flights used 3 engines on the booster during the flip,  Flight 12 used 5.  And the Starship second stage thrust at full throttle is also significantly higher.  So,  the hot staging event forces are all higher. 

That means the materials are most likely being pushed “right to the limits”,  in order to stay as lightweight as possible.  And these new Raptor-3 engines reportedly use more 3-D-printed metal parts than even the previous Version-2 Raptors used. 

This author is an old retired engineer (and teacher).  His knowledge may be obsolete,  but he does remember that very early on,  3-D-printed metal parts were weaker than their forged counterparts.  Later on,  printed-part strengths improved to equal forged values,  although ductile plastic elongation capability still fell short of that of forged parts.  This is depicted in Figure 2.  Whether that is still true is not known to this author. 

It is the plastic elongation capability that confers toughness against shock load forces and impact forces!  The part survives,  although it distorts,  if the elongation at failure is large.  If elongation is insufficient,  the part still fails suddenly,  in a sort-of brittle fashion.  

Figure 2 – Typical Metal Alloy Stress-Strain Behaviors

This author has to wonder if the shock loads from the violent hot staging process,  especially one not proceeding to plan,  might not have cracked something on the one vacuum Raptor that failed on the upper stage Starship. 

Telemetry said it lit and immediately shut back down.  There was some sort of “smoke” seen persisting about the chamber end of the engine,  plus a reddish glow at one point near its exit,  plus another reddish glow across the way on the engine bay skirt.  That “smoke” was most probably leaking propellant vapors.

Were those red glows the result of a fire in the engine bay,  from propellants leaking from the feed piping or turbopump shells near the head end of that engine?  It would have to be both fuel (liquid methane) and oxidizer (liquid oxygen) leaking,  for there to be an engine bay fire at all,  because staging takes place where the atmosphere is so thin,  that it is first cousin to vacuum!  No one has said so,  but it sure looked like a fire to this old engineer.

If there were leaking propellants,  that might also explain why SpaceX cancelled the in-space engine re-light test.  With leaks happening at the shut-down engine,  there might not have been enough propellants left aboard,  after engine cutoff,  to do both that test and make a landing burn at splashdown!

The task facing SpaceX is to look at all these results,  and figure out what actually happened,  because there were definitely things about this flight test that did not go according to plan!  Then they have to fix whatever went wrong.  Finding and fixing troubles is what testing is all about,  as this old retired engineer knows all too well! 

I wish SpaceX well doing these things.  And I heartily congratulate them yet again,  for the bulk of the test flight,  which went right!

-----  

Search code DDMMYYYY format:   xx052026

Search keywords:  launch,  space program

-----   

Update 6-3-2026:  I was saddened by the explosion of Blue Origin's "New Glenn" heavy lift rocket recently,  especially with all the damage that it did.  I am glad no one was hurt or killed by it.  They are months away from trying again,  having to completely rebuild that launch pad,  in addition to finding out and correcting what went wrong  with the rocket.  Such is part of experimental flight test,  though!

SpaceX and Blue Origin are the only two NASA contractors trying to create a lunar lander vehicle.   Both are grounded for failure investigations,  SpaceX for its Superheavy booster failure to recover,  and Blue Origin for the static fire explosion.  This delays NASA's plans for moon landings by at least those same several months.  

Sure would be nice if there were more than just two lunar lander contractors to choose from,  wouldn't it?  

As for heavy-lift satellite launch services,  there's SpaceX with its Falcon-9 and Falcon-Heavy,  there was soon going to be Blue Origin with its "New Glenn",  and there is ULA with its new Vulcan that replaced its Atlas-5.  Except that ULA's Vulcan is grounded pending investigation and correction of 2 serious solid booster problems,  out of 4 flight test attempts.  

It sure would be nice if there were more than 3 heavy-lift launch suppliers,  wouldn't it?  Especially with 2 of them currently grounded.

In 2020 hindsight,  the mania for corporate growth by acquisitions and mergers,  unfettered by any monopoly busting,  was a massive governmental policy failure,  dating back to during World War 2.  

You do notice,  do you not,  that nobody in the congress or the administration is talking about that?  

You can correct that,  by voting for someone else.  Anybody else would be better than what you have.  

But you'd better do it soon!  If the current crowd stays in power after the November mid-terms,  there will be no more honest elections in America,  only rigged shams!  

Dictatorship looms over you!

-----   

Launch to Low Orbit Study

This study explores launch to low circular Earth orbit at low inclination.  It encompasses two different propellant combinations,  and 1-shot throwaway designs versus recoverable and reusable designs.  It considers both single-stage-to-orbit (SSTO) and two-stage-to-orbit (TSTO) configurations.  It includes a mixed-propellant TSTO as 1-shot and reusable.  This study used the new worksheet “both” added to the “stage studies.xlsx” spreadsheet.  See “Launch Vehicle Rough-Out”,  5-18-2026 on this site, for spreadsheet descriptions.

The two propellant combinations are oxygen-hydrogen (LOX-LH2) and oxygen-methane (LOX-LCH4).  These bound the performance problem,  be it SSTO or TSTO,  and regardless of whether 1-shot or reusable.  The mixed-propellant option looks at a LOX-LCH4 first stage,  and a LOX-LH2 second stage,  in the TSTO,  for both 1-shot and reusable designs.

This study is based on the performance levels of modest-technology engine designs,  be they sea level/ascent-capable,  or vacuum-capable.  They have lower max chamber pressures Pc,  lower pressure turndown ratios P-TDR,  and a cycle resulting in some turbo-pump bleed drive gas dumped overboard.  One can always do a little better with higher-technology versions of these same engines,  especially with full-flow cycles. 

The difference between 1-shot and reusable designs is two-fold:  (1) slightly-different design delta-vee (dV) requirements for reusable versus 1-shot,  and (2) significantly different stage inert fractions for reusable versus 1-shot. 

See Figure 1 for the basic mission concepts,  and Figure 2 for the orbital mechanics data.  All figures are at the end of this article. 

The engine performance analysis results for the four baseline engine types are included as Appendix A below.  These are easily re-scaleable to other sizes and thrust levels.  The thrust level used was 220,500 lb = 100 metric ton-force.  Areas and flow rates are proportional to thrust.  Dimensions are proportional to the square root of thrust.

The numbers used are summarized in Figure 3 below.  It covers both the SSTO and TSTO configurations.  The specific impulse (Isp) values came from the engine sizing analyses in the appendix.  They were reported to the nearest-second of Isp,  but without any rounding-up.  The inert fractions were simply presumed,  with modern 1-shot stages typically near 0.05 (5%).  The reusable stages must have higher inert fractions,  reflecting the inert additions for entry and descent aerodynamic control,  minimal landing legs or fittings,  and in the case of SSTO or TSTO 2^nd stages,  heat shielding to survive full-speed entries.  These were simply presumed as educated guesses for this study.

There is one fundamental governing equation here,  regarding mass numbers:  the sum of the payload fraction,  inert fraction,  and propellant fraction,  must be 1.000!   The dV values for the reusables are larger by the presumed landing budgets.  All the other components of dV are shared by both 1-shot and reusable designs.  For this study,  all the gravity and drag losses are borne by the SSTO,  and by the TSTO 1^st stage.  These were each presumed as 5% of perigee speed for the transfer ellipse as a measure of the orbital mechanical energy.

The basic bounding study results are given in Figure 4 below.  They comprise a data table for the SSTO configurations,  and plots for the TSTO configurations.  The SSTO table includes both 1-shot and reusable configurations,  and LOX-LH2 and LOX-LCH4 propellant combinations.  Mass ratio MR = exp(dV/Vex),  where Vex = 9.80667*Isp/1000.

The SSTO using LH2 as a 1-shot,  looks rather competitive in terms of payload fraction.  The SSTO using LCH4 as a 1-shot is technically feasible,  but has a very unattractive,  tiny payload fraction.  Neither of the SSTO reusables is technically feasible at all,  whether H2 or CH4,  with their negative payload fractions.

The all-H2 TSTO looks very attractive in terms of payload fraction as a 1-shot,  and is still quite competitive as a reusable.  The all-CH4 TSTO as a 1-shot,  is also feasible,  and more-or-less comparable in terms of payload fraction with the all-H2 reusable.  The all-CH4 reusable TSTO is mostly technically feasible except at the lowest staging velocities,   but it is not very competitive in terms of payload fraction.

Those TSTO results as bounds bring up the question of using H2 in one stage and CH4 in the other.  Only H2 in the 2^nd stage and CH4 in the 1st stage makes any sense,  since the 1^st stage Isp is lower,  but that 1^st stage also shoulders a lower dV requirement.  That combination was run as both a 1-shot and a reusable,  with results given in Figure 5 below. 

6% mixed vs 8% all-H2 payload fraction is pretty comparable for the 1-shot TSTO,  and 3.5% mixed vs 5% all-H2 is pretty comparable for the reusable TSTO.  The mixed results are closer to the upper-bound all-H2 results than the lower-bound all-CH4 results,  whether you look at 1-shot or reusable.  What you “buy” with that slight drop with mixed,  is a smaller 1^st stage-as-LCH4 volume,  sitting on the launch pad.  The 2^nd stage is the same H2 configuration,  either way.  The mixed-propellant reusable TSTO payload fraction is not as sensitive to staging V,  compared to all the other configurations here.  2 km/s is “typical”.

Conclusions and Caveats 

The highest payload fractions are 1-shot (not surprisingly).  The “best” TSTO 1-shot payload fractions are all-H2 at 8+%,  with mixed 1-shot not very far behind at ~6.5%.  The all-CH4 TSTO 1-shot falls significantly short of these at about 4%.  The 1-shot H2 SSTO is also comparable at ~5.7%,  but the reusable H2 SSTO has essentially zero payload fraction.  Neither of the CH4 SSTO’s was even technically feasible at all.

The mixed reusable TSTO is almost as good in terms of payload fraction at ~3.5% as the reusable all-H2 TSTO at ~5%,  with reusables having much lower operating costs! 

The mixed reusable TSTO is less sensitive to staging V than the rest,  and it would have the lowest stage 1 volume sitting on the launch pad. 

Caveat:  inert fractions were presumed as educated-guess values,  and no inert buildup analyses were done to verify those numbers!

Caveat:  no thrust requirements were determined for any of these stages,  no numbers of engines and their thrust levels were determined,  so there was no determination of dimensions,  and no determination of whether the engines would fit behind the stages! 

Additional Related Information

The author cross-plotted his results for the TSTO,  separated into two plots,  one for 1-shot designs and the other for all-reusable designs.  This was to determine the sensitivities to propellant selection,  but bear in mind that we are examining effects that are “down in the weeds” compared to engine Isp and stage inert fraction effects. 

The author ran one more design that was mixed-propellant with a methane stage 1 and a hydrogen stage 2.  The 1^st stage was calculated as reusable,  with those dV,  Isp,  and inert fraction values,  while the 2^nd stage was calculated as 1-shot,  with those dV,  Isp,  and inert fraction values. 

Like the both-stages-reusable case with mixed propellants,  overall payload fraction is well below the both-stages-1-shot values,  and very near the both-stages-reusable values.  However,  it exceeds the both-reusable only at rather high staging speeds,  and in any event would be more expensive to operate,  with its 1-shot 2^nd stage.  Again,  this is “down in the weeds” compared to engine Isp and inert fraction effects.  See Appendix B below.

About the Author

The author had two 20-year careers,  the first in aerospace/defense new product development work,  and the second in mostly teaching,  with some civil engineering and aviation work thrown in.  The change was necessitated by the huge aerospace/defense workforce drawdown that took place shortly after the fall of the Soviet Union.  For most of these careers,  he had BS and MS degrees in aerospace engineering.  He obtained a PhD in general engineering late in life.  He is now long-retired,  only doing some consulting now and then.  Contact him by email at gwj5886@gmail.com. 

Figure 1 – Basic Mission Concepts

Figure 2 – Orbital Mechanics Defines Most of the dV Requirements

Figure 3 --  The Numbers As-Used In This Study

Figure 4 – The Bounding Results for Both SSTO and TSTO/Both Stages Same Propellants

Figure 5 – Results for TSTO with LOX-H2 Stage 2,  and LOX-LCH4 Stage 1

-----   

search code DDMMYYYY format:   19052026

search keywords:  launch,  space program

-----   

Appendix A --  Rough-Sized Baseline Engines

These were two LOX-LH2 engines sized for ascent and for vacuum,  and two LOX-LCH4 engines sized for ascent and for vacuum.  These were sized from propellant combination c*,  using a specific heat ratio of 1.20,  and the appropriate nozzle throat to deliver a required thrust.  We are only interested in the specific impulse (Isp),  for this study.  Figures A-1 through A-4 below are the reported results for these sizing calculations. 

The ascent nozzles were sized to be on the verge of flow separation in the exit bell,  when operating at 80% of max Pc,  while testing at sea level.  That corresponds to an exit area ratio of 35.88:1,  when max Pc is the rather modest 2000 psia.

The vacuum nozzles were sized to an arbitrary exit area ratio of 150:1.

An 18-8-degree curved bell was presumed,  with throat discharge coefficient of 0.995. 

The engine cycle need not figure into any of this,  except as its dumped bleed fraction.  5% (0.05) was presumed for that.  The value of “Pc” here is that taken just before the contraction from chamber to throat.

Nozzle separation:  Psep/Pc = (1.5 * Pe/Pc)^0.8333,  where Pa > Psep is separated.

Figure A-1 – Hydrogen Ascent Engine

Figure A-2 – Hydrogen Vacuum Engine

Figure A-3 – Methane Ascent Engine

Figure A-4 – Methane Vacuum Engine

-----   

Appendix B – Related Information

Figure B-1 – Sensitivities to Propellants Selected in Stages,  For 1-Shot and For Reusable

Figure B-2 – Effects of 1-Shot Stage 2 with Reusable Stage 1 with Mixed Propellants

-----   

Launch Vehicle Rough-Out

This author has tried several times to automate pencil-and-paper design analyses with spreadsheet software,  for the purpose of rough-sizing Earth-to-orbit launch vehicles,  be they 1 or 2 stage.  None of those were as successful and suitable as he would like,  until now.  Bear in mind that these spreadsheet results are not real performance estimates,  only bounding calculations!  The choice of stage inert mass fractions is entirely arbitrary at this level of analysis,  totally unrefined by any sort of inerts-buildup design activity.

In this particular case,  the spreadsheet file is an Excel spreadsheet file named “stage studies.xlsx”.  It has 3 worksheets within,  1 being for single-stage-to-orbit (SSTO) scenarios,  the other 2 worksheets for two-stage-to-orbit (TSTO) scenarios. 

The worksheet named “SSTO” creates plotted trends of payload mass fraction vs ascent-averaged specific impulse (Isp),  parametric upon values of vehicle inert mass fraction W_inert/W_ignition.  Metric units are presumed,  being metric tons (m.ton) for masses,  and speeds in kilometers per second (km/s).  Isp,  measured in seconds (s),  is not defined in terms of consistent units!  The corresponding effective exhaust velocity Vex is,  and is measured in km/s to match the other speeds.  See Figure 1. 

Figure 1 --  Image of the “SSTO” Worksheet

The plots respond automatically to changes in the yellow-highlighted user inputs for a spread of Isp values,  and a required velocity increment capability “rq dV”,  km/s.  That last  is the end of all burns speed upon reaching orbit,  with drag and gravity losses added to it,  plus a small budget to cover anything else.  The input Isp range covers whatever ascent-averaged Isp might obtain,  for any given propellant combination and engine technology.  Generally speaking,  the user need not change the Isp spread inputs as shown. 

For a given ascent-averaged Isp,  and a given vehicle inert fraction,  the corresponding payload fraction can be read right off the plot to about 2 significant figures! 

There is an aid for figuring vehicle weight statements,  located top right.  It has yellow-highlighted inputs for vehicle inert and payload fractions.  This takes two forms:  you know a known (yellow-highlighted) liftoff mass,  or you know a (yellow-highlighted) delivered payload mass.  It generates the correspond weight statement,  either way,  whichever is chosen.

To bottom right is an aid for determining the “rq dV” needed as an input for the main worksheet calculation.  There is a yellow-highlighted input for the actual speed at entry into orbit,  plus two yellow-highlighted inputs for the percentages of that speed,  that are the gravity and drag losses.  There is one other yellow-highlighted input for a small budget to cover anything else,  such as landing burns and rendezvous burns.   It sums the 4 components to create the “rq dV” value,  which is essentially the speed needed,  factored up to cover all the loss and budget items.  That is what is used for the rocket equation-based calculation for design mass ratio MR.

Figure 2 below shows a part of the “SSTO” worksheet and the plots it generates.  This image has been annotated to indicate where the appropriate bands of ascent-averaged Isp are,  corresponding to LOX-LCH4 and LOX-LH2 propulsion.  LOX is liquid oxygen.  LCH4 is liquid methane.  LH2 is liquid hydrogen. 

Note that there is no indicated feasibility (positive payload fraction) for inert fractions much above 5%.  That low fraction corresponds to an expendable stage design.  And note that the payload fraction with LOX-LCH4 is substantially less than that with LOX-LH2.   What that really says is that you actually can build an SSTO to reach low Earth orbit,  but only as an expendable,  and it cannot be truly competitive unless you use LOX-LH2 propulsion.

Figure 3 below is an image of the “TSTO” worksheet,  which creates plots of payload fraction vs stage Isp,  parametric on stage inert fraction,  for both stages.  Top left are the yellow-highlighted user inputs,  which include orbit entry speed and staging speed,  plus the losses to be added to each stage’s effective dV requirement. The plots and the data are generated automatically from these inputs. 

The first stage must ascend through the atmosphere,  needing an ascent-averaged Isp for a sea level-capable engine design,  while the second stage makes its burns essentially exoatmospherically,  using a vacuum-capable engine.  

Figure 2 – Example Results of the “SSTO” Worksheet

Figure 3 – Image of the “TSTO” Worksheet

There are two plots,  one for each stage.  They are both plots of stage payload fraction vs a range of stage average Isp,  parametric on stage inert fraction. 

User instructions are given on the page.  There was NOT room to include stage and overall weight statements.  Those were included as a separate worksheet “TSTO wts”.

The example TSTO here is a crude approximation of the earlier block 1 or block 2 versions of SpaceX’s “Starship/Superheavy” vehicle,  that is still in experimental development flight test,  as of this writing.  Figure 4 is an image of a png file containing the two plots from the “TSTO” worksheet,  annotated to reflect the vehicle characteristics also sketched.  The old Windows “Paintbrush 2-D” software was used to generate this png file.

Figure 4 – Example Results of the “TSTO” Worksheet

For each stage,  the payload fraction is read from the plot at the appropriate Isp and inert fraction.  The best way to do this is to sketch-in the appropriate curve for the inert fraction.  Then read up from the appropriate Isp until you hit that inert curve.  Then read across left to the payload fraction scale,  to estimate the payload fraction.  Once you have the appropriate payload fraction for each stage (along with its inert fraction),  you can go to worksheet “TSTO wts” and run weight statements for each stage.

Figure 5 shows an image of the “TSTO wts” worksheet.  The user inputs are the yellow-highlighted items.  They are the payload and inert fractions for each of the two stages,  plus an estimate of the lift-off mass.  This worksheet creates no plots.  This worksheet also creates a sort of overall weight statement for the entire two-stage vehicle,  and computes its overall payload fraction,  that being delivered payload divided by liftoff mass.  User instructions are listed on the page.  If you know payload instead,  just iterate lift-off mass until you get the payload you desire.

Figure 5 – Image of the “TSTO wts” Worksheet

For the “Starship/Superheavy” example already shown in Figure 4 above,  the results are indicated just below in Figure 6.  That is an image of just the weight statement calculation blocks,  annotated.  It is amazing how close these results are,  to the actual test vehicles flown so far,  given just how crude these input data values are,  that were used here.

Figure 6 --  Example Results of the “TSTO wts” Worksheet

One could actually run a sort of sensitivity study with this worksheet and the “TSTO” worksheet together.  One would do this by varying the inert fraction somewhat,  and seeing how the rest of the results are affected.  You must do this on both worksheets together in sequence,  because inert fraction affects payload fraction,  with the propellant fraction set by Isp and required dV.  Note that 1 = payload fraction + inert fraction + propellant fraction.

Critical Issues NOT Addressed By This Level of Analysis

(#1) As indicated in the first paragraph,  stage inert fractions are merely assumed!  Verifying these requires more detailed design analysis of the inerts build-up for an actual vehicle design concept.  The more realistic the assumed values are,  the more realistic the results.

(#2) Ascent-averaged and vacuum Isp values are simply assumed in this analysis.  To verify those values,  one must rough out the actual nozzle designs,  and engine cycle characteristics,  for his propellant combination.  Better assumed values are better results.

(#3) In this analysis there is no attempt to size thrust requirements,  and determine the numbers of engines,  their thrust ratings,  and their turndown ratios,  for each stage.  If the engines will not fit behind the stage,  Isp and dV do not matter,  the design is infeasible.

How to Obtain this Spreadsheet File (“stage studies.xlsx”)

To obtain a copy of this spreadsheet,  either contact the author directly by email,  or go to the Mars Society’s New Mars forums,  for the links to a free download of this,  and many other related things.  That website is https://newmars.com/forums/.  Once there,  scroll down to the Acheron Labs section,  and select the “Interplanetary Transportation” topic.  Within that topic,  select the thread titled “orbital mechanics class traditional”.  The links to the online drop box are in those postings.  Those downloads are free!  This particular spreadsheet is one of the items stored under “lesson 8B”,  post 45.

About the Author

The author is well-qualified to create spreadsheets like this,  or those lessons on the forums,  and much more.  He obtained BS and MS degrees in aerospace engineering long ago at UT Austin,  and then went to work in the defense industry for 20 years,  doing mainly rocket and ramjet work,  mostly as pencil-and-paper engineering,  starting in the slide rule days (see Figure 7 below).  He even did some industrial launch vehicle work while still a graduate student.  His majors were aerodynamics,  thermodynamics,  and propulsion.  He obtained a terminal doctorate late in life,  in general engineering.

That aerospace/defense career was cut short by the enormous defense industry drawdown after the fall of the Soviet Union.  The author had to do something else for a living,  being thrown onto the job market along with nearly 2 million other defense engineers,  from an industry that could no longer employ any of them.

His second 20-year career was mostly teaching,  at all levels from public school to university.  This had some civil engineering and aviation work mixed in with the teaching.  He taught mathematics,  physics,  and engineering at a variety of institutions.  Those included Bosqueville and McGregor high schools,  Minnesota State University,  Baylor University,  Texas State Technical College,  and McLennan Community College.

The author is now long-retired.  He still builds and sells custom farm implements that he invented,  which kill prickly pear cactus out of farm and ranch pastures,  without pickup and disposal,  and without chemicals.  He also still occasionally consults in topics like ramjet propulsion.

Figure 7 --  Old-Time Engineering Design Tools

-----  

Search code DDMMYYYY format      18052026

Search keywords                                       launch, space program

-----   

Update (same day) 5-18-2026:

The author has added another worksheet to the “stage studies.xlsx” spreadsheet file,  that does both SSTO point sizing and TSTO sizing versus staging velocity.  These are still only bounding calculations,  and are no better than the quality of the Isp and inert fraction values assumed. 

These TSTO trends with staging velocity are of interest toward design,  but bear in mind that these values are “down in the weeds” relative to the effects of both engine Isp values and stage inert fraction values!  There is still no assessment of stage thrust requirements,  or whether the engines would actually fit behind the stages. 

The user instructions for this added worksheet are on the worksheet.  As set up,  it analyzes multiple cases.  An overall view of this worksheet and the plots it automatically generates from these cases is given in Figure 8 below.   A more close-up view of the user-input portion of the worksheet is given in Figure 9 below,  which is more useful for dealing with the inputs.

Looking at Figure 9,  the top left yellow inputs are for SSTO,  as 1-shot and as reusable.   Those calculations are simpler,  and the results are highlighted blue in that same little block. 

Just below it are the inputs block for the TSTO cases versus stage velocity Vstg.  Vstg is not an input,  it is a wide range of values “built-in” to the worksheet.   Although,  they could be changed (those are highlighted green).  The cases themselves are the calculation blocks that stretch far to the right on the worksheet,   something more apparent in Figure 8. 

The bottom block of yellow inputs are the names and identifying values for the various cases this worksheet is set up to compute.  These are for the different possible stage propellant selections,  and whether the stage is reusable or not. 

The main inputs the user should be concerned with,  are the engine Isp values,  the stage inert fractions,  and the mission dV components that add up to stage dV requirements.  Thos dV values are currently set up for low circular orbit at 300 km altitude,  and low inclination. 

The first of the several calculation blocks that stretch to the right is where the user inputs for speeds,  losses,  and budgets are recombined into the 1-shot and the reusable stage dV values versus staging speed.  A plot of stage dV requirements versus staging speed is one of the plots generated automatically.     The other plots,  also generated automatically,  compare various cases in the format of overall payload fraction versus staging speed. 

Figure 8  --  Overall View of the Added Worksheet “Both”

Figure 9 --  More Closeup View of the Input Portion of the Added Worksheet “Both”

Watch this site for a future posting that deals with the results generated by this added worksheet. 

-----  

About the Artemis-2 Mission

See the 4-13-2026 update at the end,  for some early photos of Artemis-2 post landing,  plus my preliminary assessment of heat shield damage from them.

------------   

The following picture is a NASA illustration of the mission.  It does not give reliable distances,  nor does it show when the vehicle crosses,  or is outside,  the Van Allen radiation belts.  Outside the Van Allen belts,  the vehicle is also outside the Earth’s magnetic field,  and therefore at risk for radiation exposure,  should a solar flare’s coronal mass ejection (CME) event occur,  and happen to hit them.  

The next two figures together show the sequence of actual orbits being used,  after I looked around to various sources on line,  for real distance information.  I could not fit all the information into one figure.  That is why there are two figures that I hand-drew.

The NASA space launch system (SLS) rocket is a two-stage launch vehicle whose first stage core is assisted by two large solid boosters (SRB’s),  as was the space shuttle.  This first stage booster and its SRB’s essentially put the interim cryogenic propellant second stage (ICPS),  plus Orion capsule and its service module (SM),  onto a transfer ellipse with an apogee at the desired apogee altitude,  and a perigee that is pretty close to being a surface-grazing orbit.  There is no second stage ICPS burn to get onto that transfer ellipse. 

The ICPS makes a small burn at that desired apogee,  which raises the perigee up to the desired perigee altitude just outside the atmosphere,   but still quite low.  That apogee is actually within the inner of the two Van Allen radiation belts!    Without that perigee-raising burn at the initial apogee,  the Orion would inevitably re-enter the atmosphere on its first-pass return to perigee! 

The ICPS second stage then makes another,  larger burn at the newly-raised perigee point,  which raises the apogee to a very high elliptical orbit apogee,  actually outside both of the Van Allen belts.  They extend from about 1000 km,  to over 50,000 km,  altitudes.

The ICPS stage then makes one last “small” burn to put it into a “graveyard orbit”.  That graveyard orbit was not defined anywhere that I could access,  so I could not determine the magnitude of that burn.  All of this that I could find,  is in my first hand-drawn figure,  here: 

My second hand-drawn figure just below shows the transition from this high elliptic orbit to the actual transfer trajectory to the moon.  I used an ellipse from the low 185 km perigee at Earth to an apogee at the average distance of the moon to approximate the actual figure-8 trajectory,  for the purpose of estimating the departure delta-vee (dV) supplied by the Orion Service Module (SM) propulsion.  That dV is shown in that second hand-drawn figure. 

The actual transfer trajectory starts with this ellipse,  but gets distorted into the figure-8 shape by the gravity of the moon passing by,  making those details a 3-body problem one can only solve by finite-difference methods on the computer.  It loops around the moon in almost a polar orientation,  somewhere near 6500 km altitude,  behind the moon.  Then it free-returns to Earth.  A minor course correction from the SM is needed,  just before it gets jettisoned,  to ensure hitting Earth’s atmosphere at just the right angle,  for the free-return direct entry.

That direct re-entry returning from the moon is more demanding than one from low Earth orbit like the space shuttle endured,  because the speed at entry interface (about 140 km altitude) is higher,  at essentially the perigee speed of the transfer ellipse model shown above,  or right at 10.94 km/s,  maybe even 10.98 km/s.  That speed is very nearly Earth escape speed,  which is 11.18 km/s at the surface,  and 11.07 km/s at the entry interface altitude of 140 km.  Only about 90 to 130 m/s different!

The test history of the Orion capsule and its heat shield points to a disturbing possibility of heat shield damage possibly happening on this Artemis-2 mission!  The first Orion flew atop a Delta-IV launch vehicle uncrewed,  for a re-entry test,  among other things.  This was before Artemis,  and was named experimental flight test 1,  or EFT-1.  It had a heat shield manufactured of the same materials,  and built exactly the same way,  as the Apollo heat shields.  It did fine,  but that is an expensive,  labor-intensive manufacturing process. 

NASA used the same basic ablative material,  but manufactured in a completely different way,  for building two Orion capsules at once:  designated for the Artemis-1 and -2 missions.  They built the Artemis-2 heat shield before flight-testing the revised manufacturing process for it on the uncrewed Artemis-1,  launched by an SLS rocket.    They expected it to do fine,  but it did not,  unexpectedly shedding chunks of char,  leaving craters in the heat shield,  as the photo just below shows.  This is an official NASA photo of the Artemis-1 heat shield,  as recovered after the uncrewed Artemis-1 flight. 

The streaks on the heat shield point to a “source”,  to photo right on the heat shield,  which would be the stagnation point deliberately located off-center,  for generating a slight lift force during entry.  That force is small,  but it can “fine tune” the re-entry trajectory shape,  by rolling the vehicle to point that force where you want it.  NASA has done this for decades,  dating back to the Gemini flights of the mid 1960’s.  That is normal.

But,  if you look close,  you can see “craters”,  some large,  some small,  all over that heat shield,  where it shed chunks of the charred material from its surface!  That outcome was entirely unexpected,  and led to serious investigations at NASA,  for what to do about it. 

It must also be said that these damage craters,  as they were experienced on Artemis-1,  were not a risk for a fatal burn through!  The interior temperatures in the cabin did not vary from normal and expected,  despite the alarming damage!

NASA decided from its investigations that the two-heating-pulse “skip” entry,  that they used experimentally for Artemis-1,  was the culprit behind the chunk-shedding,  thinking that charring-material gas-evolution during the second heating pulse is what “blew” these chunks out,  leaving the craters behind.  They eliminated the skip during re-entry for Artemis-2,  getting it down to 1 heating pulse,  and decided to fly the same heat shield design,  already installed on the Artemis-2 capsule,  with a crew.

Others are not so sure about that damage mechanism.  We shall soon see. 

If Artemis-2 shows similar damage to Artemis-1,  then we (and NASA) will know that they were wrong about this damage mechanism!  Only a flight test can tell!

The problem here is shedding a second chunk from the bottom of one of the larger craters.  Should that happen,  the probability of a fatal burn through becomes very significant indeed!   Such would be a very low-probability event,  but that probability is not zero!

Last year,  I sent my concerns about this problem,  plus a low-cost means to stop the chunk-shedding,  even for a two-pulse skip re-entry,  to the entry heat protection group at NASA-Houston,  and again this year directly to the new Administrator,  at his DC office.  I do know that the Houston heat protection group thought I was right about my concern,  and about my proposed “fix”. 

So far,  NASA has not officially chosen to explore my alternative,  and,  as near as I can tell,  has already started construction of the Artemis-3 capsule with the same Artemis-1/Artemis-2 heat shield design.

Personally,  I would not ask a crew to fly with an unresolved risk that I already thought that I knew how to mitigate!  I think that is unethical!  Apparently,  there are high-level managers at NASA who disagree with that assessment. 

All I can say to them is this:  “there is nothing as expensive as a dead crew,  especially one dead from a bad management decision”.

What I sent the new Administrator is what was posted here on “exrocketman” 1 March 2026,  under the title “Ramjet Data Re:  Heat Shields”.  I knew a lot about ablative heat protection in ramjets and in solid rockets.  Some of that overlaps re-entry heat shields!

I did the 2-body ellipse orbital calculations illustrated here,  with a simple Excel spreadsheet “orbit basics spreadsheet.xlsx”,  which is part of the course materials posted online for free download,  via the Mars Society’s New Mars forums.  It simply automates the classic 2-body textbook equations.  While that New Mars forums site is down as of this writing,  you can still get that spreadsheet from me.  Just email me for it.  Anybody can do what I did here.

-----   

search code DDMMYYYY format:   31032026

search keywords:  aerothermo, launch, radiation, space program

-----   

Update 4-11-2026:  They made it back safe.  Of that I am glad.  I am eagerly awaiting photos of the condition of their heat shield.  Once I see such,  I will update here again.

-----   

Update 4-13-2026:  Enough photos have surfaced of the Artemis-2 heat shield to make a preliminary assessment.  Starting with Figure A,  which is too blurry to make out very much,  other than the side facing the camera is not the side with the hatch and windows.  There is one tie-down pad zone that seems to have suffered damage (the whitish thing near the rim.  The other one near it did not,  being the slightly-dark spot below it and to its right,  also near the rim.  The other two tie-down pads are not in this view.   It is unclear whether the capsule is being hoisted,  or is still descending on its main chutes.

Figure A – View of Artemis-2 Orion Capsule

Figure B is a clearer version but of limited view dimension,  probably enhanced for clarity,  and apparently made from the blurry photo in Figure A,  intended to examine the whitish hold down pad damage.  It looks like a cavity in the heat shield,  burned out around where that hold-down pad got destroyed. 

To its left and lower left,  I see 4 dark,  small cavities that appear to be craters left behind from char chunks lost,  similar to those seen on the Artemis-1 heat shield.  They are just smaller and fewer than the hundred or so seen on Artemis-1.  Bear in mind that these few are on a small portion of the entire heat shield!  There could well be several,  to many, more!

Up on the lateral side of the capsule,  where the Avcoat protection was thinner,  I do see exposed metal in at least two places,  where the charred heat shielding was lost,  and that exposed metal looks distorted,  as if it did indeed see overheating during entry.

Figure B – Detail Near Hold-Down Pad Damage,  Enhanced for Clarity

Figure C is a photo taken during the post-splashdown extraction of the crew.   This is on the other side of the capsule,  where the windows and the hatch are located,  opposite side from that examined in Figures A and B.   The heat shield itself is hidden in this view.  I wish to point out that the lateral side of the capsule suffered very much less apparent damage than that seen in Figure B.  It would appear that the Figure B side saw more attached flow due to maneuvering angle-of-attack,  than the Figure C side,  thus charring the thinner lateral tiles through,  and so losing some of them.  

Figure C – Photo Taken During Crew Extraction,  Other Side of Capsule

The last figure,  Figure D below,  is NOT of an Orion capsule,  but an Apollo capsule,  specifically the one from Apollo-11.  In it,  one can very clearly see the reinforcing hex in the heat shield charred surface,  plus a variety of plugs that close openings in that heat shield. 

Similar to Orion,  there is more erosion near the “compression shear pad”  (same function as the “hold down pad” on Orion.  That kind of thing is apparently not unexpected. 

What I want to point out is the oxidizer dump plug,  which does not seem to have had any hex in it.  That location shows more erosion than the heat shield around it,  which emphasizes the positive reinforcement function served by the hex.  

Figure D – Minor Damages Seen on the Apollo-11 Heat Shield

My conclusions are three-fold: 

(1) Beneficial erosion reduction was indeed obtained by deleting the two-heating-pulse skip entry,  just as thought by NASA. 

(2) While erosion was reduced by deleting the skip,  there was still some char chunk-shedding going on,  which NASA hoped would not happen.  Plus,  there was some damage to the lateral side of the capsule. 

(3) The need for the reinforcing hex has been apparent since Apollo,  which did NOT do skip entries.

My recommendation still stands:  NASA,  put the reinforcing hex into your Avcoat tiles.  Use the extrusion press the way I suggested,  to do that without manual hand-gunning.  

-----   

Update 4-20-2026:  I have been able to confirm that Figure A just above is indeed a CBS-taken photo of Artemis-2 still hanging from its chutes,  just immediately prior to splashdown.  That one is blurry,  and shows a white splotch near the rim that sparked a lot of questions and discussions.  I am inferring that CBS did digital enhancement to clear up the blurriness,  and zoom-in on the white splotch.  That would be Figure B just above.  It is quite clearly derived from the original in Figure A.  

What I see is essentially a lot of unexpected lateral-side damage in Figure B,  plus some cracking and about 4 small chunk-loss craters,  on a very small portion of the main bottom of the heat shield.  

What I have noticed is that NASA has not jumped to release photos taken all over that heat shield,  to quiet the fears sparked by the Artemis-1 damage.  The only ones we have seen show the lateral side with the windows and hatch,  where things look quite good.  None show the other lateral side where there was attached flow with higher heating while flying at angle-of-attack,  and none show the main heat shield on the bottom.  

My suspicion remains:  I think there was still some chunk-loss cratering,  and cracking,  on the main heat shield bottom,  just fewer craters,  and of smaller size.  So it really did do better with the revised trajectory,  just not good enough!

I have a new suspicion:  the damage on the back lateral side threatens a burn-through into the pressure cabin shell inside,  and was apparently quite unexpected.  Deleting the skip increased the peak heating load,  and that may have been a bit too much for that backside lateral wall,  where flow was attached at angle of attack,  causing a lot more "scrubbing action".  

If I am right,  neither suspicion threatens Artemis-3,  which is an Earth-orbital mission that sees far less heating upon re-entry.  But is does threaten Artemis-4,  the intended first landing mission to the moon.  It must return from the moon just like Artemis-2 just did.  

They really,  really need to fix this heat shield problem "right" before Artemis-4 flies!  Otherwise,  they are playing the odds with crews' lives.  Sooner or later,  that kills crews.  We have already seen it!

-----   

Update 4-23-2026:  Here are some more photos that have surfaced since the Artemis-2 splashdown.  We have not yet seen anything “official” about Artemis-2’s heat shield condition and performance from NASA,  not as of this writing! 

Figure I below is supposed to be a photo taken of lateral-wall damage seen,  as pointed out by the mission commander Reid Wiseman.  I do not know exactly where this photo came from!  The tan color surface is stuff that is almost unheated.  The black stuff is material that saw enough heating to actually char.   There is some sort of localized crater in the lateral wall heat shield that he is pointing to. 

Note the black thing near his head in the image.  That is either the hatch opening or one of the recessed windows,  which would have been in the separated flow region on the opposite side of the capsule,  from where the stagnation point was biased toward the rim. 

Damages might well have been more severe on the other side of the capsule,  adjacent to the stagnation point that was offset toward the rim,  where heating would have been much higher.  We do not know:  no images from there have been released by NASA.

We have seen (so far) no images of the other side of the capsule at all,  where flow was attached and heating higher,   other than in the unofficial enhanced CBS image taken just before splashdown. 

Figure II below is a NASA photo taken of the heat shield by the Navy divers,  while the capsule was still in the water,  probably before the astronauts were recovered,  and certainly before the capsule itself was recovered!  It was posted on LinkedIn by NASA’s Steve Yoon.  You can tell by the surface streaks on the heatshield,  that the stagnation point was toward the top left in this image,  but actually outside the view of the camera,  from when the image was taken. 

The basic heat shield surface in this photo looks quite intact,  so far as this image discloses!  Which means NASA was correct in saying that the change in entry trajectory from skip to non-skip reduced base heat shield damage!  There is no doubt about that!

You can see an inverted-U-shape mark on the heat shield,  which is located around where one of the hold-down pads was located.  You must look very close to see the other one visible in this view,  to image top right.  Yet,  it is there.  Neither of these is the whitish “stain” and erosion cavity that was seen in the CBS image and its zoom-in enhancement.  Those other two pad locations are not visible in this image’s view.  They are on the other side.

Figure III below is something I found posted on LinkedIn,  purported to be damage on the lateral capsule wall of Artemis-2,  but I do not believe that claim!  If this is even a photo of an Orion at all,  it has to be from the initial flight test EFT-1,  before the Artemis program began!  I say that because it is quite evident that there was a reinforcing hex in the pictured lateral heat shield!  There was such on EFT-1,  since it was built the same way as Apollo. There was no such reinforcing hex installed on either the Artemis-1 or Artemis-2 Orion capsules!  Those were insulated with cast and machined Avcoat tiles that had no reinforcing hex.

We are not going to see the real evidence,  until NASA actually releases photos of the entire base heat shield,  and photos of the lateral wall heat shield that include the attached flow side,  opposite where the windows and the hatch were located.  Such may not be forthcoming for some time.  But the laws (and ethics) require that such evidence actually be released to the public!

What I suspect,  but have zero evidence for,  is that the base heat shield still showed some chunk-shedding craters over on the stagnation-near-rim side,  just less than Artemis-1.  The enhanced CBS image does indicate that to be the case.  If that is true,  it says that NASA still needs to restore the reinforcing hex to the bonded Avcoat tiles that they want to use.

The other thing I suspect,  but still have zero evidence for,  is that increasing the heating by deleting the skip,  pushed one lateral side of the capsule into a situation of inadequate insulation thickness.  That would be the side away from the windows,  where attached flow might really happen.  This would be based upon the burn-throughs and total char losses seen in the enhanced CBS pre-splashdown image.  And perhaps that Figure III image,  or maybe not. 

We cannot know any better than my “informed speculations”,  until the full suite of “official” images might be released. 

Figure I – Photo of Reid Wiseman Pointing to Damage on Lateral Wall of Artemis-2 Capsule

Figure II – Navy Diver Photo of Artemis-2 Heat Shield in the Water,  via Steven Yoon, NASA

Figure III – LinkedIn Posting Photo,  of Lateral-Wall Heating Damage to a Space Capsule

-----