SpaceX Starship/Super Heavy Engineering General Thread 4

[InterestedEngineer]

I know I'm assuming perfect circulation of the LOX against the downcomer so that the temperature of the LOX next to the downcomer is always 80K

This is an interesting analysis but it applies a virtually steady state condition of LOX near the boiling point of LN2 to a dynamic cooling of the LCH4. You are assuming the LOX is an infinite heat sink at 80 K. Both assumptions are highly suspect.

The exit condition of the LOX from the "hippo" chillers may be close to 80 K depending on relative flow rates of LN2/LOX and the heat exchanger area/dwell time/effectiveness etc but it is highly doubtful hasn't warmed up enough to significantly close the gap with the freezing point of LCH4 by the time it reaches the vehicle and fills the tank.

The volume of the LOX around the downcomer is something like 50 times the mass of the LCH4 in the downcomer so it's not quite infinite but it is pretty large heat sink.

My guess is assuming circulation of the LOX isn't optimal.  So the LOX next to the downcomer warms up, stretching the time to freeze out by at least a factor of 2.

I think also the saving item is that new LCH4 is supplied via the downcomer as the LCH4 evaporates above and is exhausted out the side of the rocket.

Topping off LCH4 from the stage 0 tanks (confusingly from the bottom) with slightly warmer LCH4 would provide the necessary heat to prevent the downcomer from freezing.

That's something that is controllable in the Stage0 software.

[shark0302]

I know I'm assuming perfect circulation of the LOX against the downcomer so that the temperature of the LOX next to the downcomer is always 80K

This is an interesting analysis but it applies a virtually steady state condition of LOX near the boiling point of LN2 to a dynamic cooling of the LCH4. You are assuming the LOX is an infinite heat sink at 80 K. Both assumptions are highly suspect.

The exit condition of the LOX from the "hippo" chillers may be close to 80 K depending on relative flow rates of LN2/LOX and the heat exchanger area/dwell time/effectiveness etc but it is highly doubtful hasn't warmed up enough to significantly close the gap with the freezing point of LCH4 by the time it reaches the vehicle and fills the tank.

The volume of the LOX around the downcomer is something like 50 times the mass of the LCH4 in the downcomer so it's not quite infinite but it is pretty large heat sink.

My guess is assuming circulation of the LOX isn't optimal.  So the LOX next to the downcomer warms up, stretching the time to freeze out by at least a factor of 2.

I think also the saving item is that new LCH4 is supplied via the downcomer as the LCH4 evaporates above and is exhausted out the side of the rocket.

Topping off LCH4 from the stage 0 tanks (confusingly from the bottom) with slightly warmer LCH4 would provide the necessary heat to prevent the downcomer from freezing.

That's something that is controllable in the Stage0 software.

Would the solution be fill via downcomer? Cause that makes sense to me.

[TheRadicalModerate]

Would the solution be fill via downcomer? Cause that makes sense to me.

Yeah, I think that's where I was confused.  That'll work for launch, as long as you don't let the system sit too long on the pad after filling is complete.  There's likely a max allowable hold time, which isn't a big deal.

There are still some cases where things could get dicey:

1) On-orbit time between burns.  I'd guess that LCH4 that's in the downcomer stays in the downcomer in microgravity, and blobs of LOX will accumulate around it.

2) Sitting on the lunar or martian surface.

3) In the depot, especially when it's full.

My guess is that all of these can be mitigated by a recirculation system:  run a small line from the top of the LCH4 tank to the bottom of the downcomer, then pump LCH4 into the bottom of the downcomer at a rate sufficient to keep everything from freezing. 

[warp99]

Would the solution be fill via downcomer? Cause that makes sense to me.

Yeah, I think that's where I was confused.  That'll work for launch, as long as you don't let the system sit too long on the pad after filling is complete.  There's likely a max allowable hold time, which isn't a big deal.

There are still some cases where things could get dicey:

1) On-orbit time between burns.  I'd guess that LCH4 that's in the downcomer stays in the downcomer in microgravity, and blobs of LOX will accumulate around it.

2) Sitting on the lunar or martian surface.

3) In the depot, especially when it's full.

My guess is that all of these can be mitigated by a recirculation system:  run a small line from the top of the LCH4 tank to the bottom of the downcomer, then pump LCH4 into the bottom of the downcomer at a rate sufficient to keep everything from freezing.

The easy solution for the ship is to do what they do for the F9 downcomer which is to have a double walled tube.  The space between the tubes can be pumped down to vacuum or filled with carbon dioxide which then freezes out on the walls to provide a vacuum as soon as propellant is loaded. 

They could do the same for the booster but it is likely not to be necessary.  There is always liquid methane flowing through the downcomer - either filling the methane tank before launch or flowing down to the engines after launch.  The boostback burn starts less than ten seconds after MECO so the longest delay without propellant flow will be between the end of the boostback burn and the landing burn with a nearly empty main LOX tank.  I would expect globules of LOX to move away from the downcomer as the contact surface warms up and eventually evaporates a thin film of LOX.   

Incidentally there is a lot of discussion of 80K LOX here but F9 LOX is subcooled to 67K and I would expect it to be the same for Starship.  The ullage space above the LN2 in the kettle boilers is pumped down to about 0.1 bar to drop the boiling point of the LN2 to 62K in order to achieve this. 

Similarly the liquid methane will be cooled to about 97K so the temperature difference is about 30K.

[TheRadicalModerate]

The easy solution for the ship is to do what they do for the F9 downcomer which is to have a double walled tube.  The space between the tubes can be pumped down to vacuum or filled with carbon dioxide which then freezes out on the walls to provide a vacuum as soon as propellant is loaded. 

That sounds like a pretty good solution.

Let's figure out how long it'll last.  If we use Tlox=80K, Tlch4=67K, and stainless steel emissivity of 0.57, then

J = (0.57)*(5.670E-8)*(80-67)⁴ = 9.2E-4W/m²

If we use a 0.75m x 13m Starship downcomer, that has 5.7m³ or 12.7t of LCH4 in it.  Up thread, I calculated that you'd need cooling of roughly 83kJ/kg to take the methane down to its freezing point.  That'd take more than 1000 days to freeze with radiative cooling alone.  Even if we're off by a factor of 2, we're still good.

I'm sure there are sneak conductive pathways that will overwhelm this number, but I'm convinced that this is a non-problem for the time being--as long as the downcomer truly is a double-walled dewar.

[InterestedEngineer]

Let's figure out how long it'll last.  If we use Tlox=80K, Tlch4=67K, and stainless steel emissivity of 0.57, then

In the last two years I've seen 4 different numbers for stainless steel emissivity, ranging from 0.15 to 0.57

 

[TheRadicalModerate]

Let's figure out how long it'll last.  If we use Tlox=80K, Tlch4=67K, and stainless steel emissivity of 0.57, then

In the last two years I've seen 4 different numbers for stainless steel emissivity, ranging from 0.15 to 0.57

 

Yeah, it's all over the place.  A lot of it has to do with how highly polished it is.  HeatLoad = fractionReflected + fractionConducted + fractionEmitted.  The conductivity is what it is, and has a much narrower range of values.  Reflectivity off a specular surface is obviously much higher than off of a Lambertian surface.  So the emissivity is heavily dependent on the reflectivity.

Fairly good assumption for all Starship pieces-parts:  They're not polished.

[Guillerz]

Is there a list of changes from the S7 to the S9?

[sebk]

Let's figure out how long it'll last.  If we use Tlox=80K, Tlch4=67K, and stainless steel emissivity of 0.57, then

In the last two years I've seen 4 different numbers for stainless steel emissivity, ranging from 0.15 to 0.57

 

Yeah, it's all over the place.  A lot of it has to do with how highly polished it is.  HeatLoad = fractionReflected + fractionConducted + fractionEmitted.  The conductivity is what it is, and has a much narrower range of values.  Reflectivity off a specular surface is obviously much higher than off of a Lambertian surface.  So the emissivity is heavily dependent on the reflectivity.

Fairly good assumption for all Starship pieces-parts:  They're not polished.

Yes, surface finish is the dominant factor. But also many figures are temperature dependent. There's no big difference between cryo and room temp. But many considerations of SS emissivity and Starship were around reentry and emissivity at 1000K is quite different from the one at the room temp.

[TheRadicalModerate]

Yes, surface finish is the dominant factor. But also many figures are temperature dependent. There's no big difference between cryo and room temp. But many considerations of SS emissivity and Starship were around reentry and emissivity at 1000K is quite different from the one at the room temp.

Emissivity is relatively temperature-invariant, isn’t it?  It’s a measure of how close to black a grey body is. Radiant flux, however, scales as the fourth power of temperature.

[tbellman]

Emissivity is relatively temperature-invariant, isn’t it?  It’s a measure of how close to black a grey body is. Radiant flux, however, scales as the fourth power of temperature.

Surfaces can have quite different emissivity at different wavelengths.  And the wavelengths a black-body emits at (or more exactly, the distribution of radiation over different wavelengths), depends on its temperature.  The combination of those two facts means that the effective emissivity constant for a specific surface can, and often does, change with temperature.

[Slarty1080]

Can anyone explain what the COPV's in the chines on Superheavy are used for and any reference for this? I thought that they were for pressurization of the header tanks for landings as autogenous pressurization won't be available initially. I'm not sure if they have other uses as well such as for attitude control thrust or pressurizing the main tanks or whether this has changed during development. 

[BT52]

Can anyone explain what the COPV's in the chines on Superheavy are used for and any reference for this? I thought that they were for pressurization of the header tanks for landings as autogenous pressurization won't be available initially. I'm not sure if they have other uses as well such as for attitude control thrust or pressurizing the main tanks or whether this has changed during development.

If i m correct mainly for purging and pump prespining.  And maybe even for some gas sealing in rotary seals. Although in full flow pump assembly that is not so important. So in short. Nitrogen and helium gas COPV should be there. Pressure is provided autogenously. Electric TVC eliminates some complexities further in chine area.

[InterestedEngineer]

Can anyone explain what the COPV's in the chines on Superheavy are used for and any reference for this? I thought that they were for pressurization of the header tanks for landings as autogenous pressurization won't be available initially. I'm not sure if they have other uses as well such as for attitude control thrust or pressurizing the main tanks or whether this has changed during development.

If i m correct mainly for purging and pump prespining.  And maybe even for some gas sealing in rotary seals. Although in full flow pump assembly that is not so important. So in short. Nitrogen and helium gas COPV should be there. Pressure is provided autogenously. Electric TVC eliminates some complexities further in chine area.

High pressure gas for restarting the center Booster engines for return and and landing burns

[Slarty1080]

Can anyone explain what the COPV's in the chines on Superheavy are used for and any reference for this? I thought that they were for pressurization of the header tanks for landings as autogenous pressurization won't be available initially. I'm not sure if they have other uses as well such as for attitude control thrust or pressurizing the main tanks or whether this has changed during development.

If i m correct mainly for purging and pump prespining.  And maybe even for some gas sealing in rotary seals. Although in full flow pump assembly that is not so important. So in short. Nitrogen and helium gas COPV should be there. Pressure is provided autogenously. Electric TVC eliminates some complexities further in chine area.

High pressure gas for restarting the center Booster engines for return and and landing burns

So the gas is used for spin up but not pressurisation. I wonder how they manage the first few seconds of start up? After a few seconds I imagine the autogenous gas is available for pressurization, but it must take time for autogenous gas pressure to build? Starting with a small head space in the header tank its volume would very rapidly expand and its pressure rapidly decrease unless very promptly repressurizion was applied.

Perhaps there was an issue with one of the Starships that I am confusing this with and maybe autogenous pressurization is just super quick from the get go.

[warp99]

Can anyone explain what the COPV's in the chines on Superheavy are used for and any reference for this? I thought that they were for pressurization of the header tanks for landings as autogenous pressurization won't be available initially. I'm not sure if they have other uses as well such as for attitude control thrust or pressurizing the main tanks or whether this has changed during development.

If i m correct mainly for purging and pump prespining.  And maybe even for some gas sealing in rotary seals. Although in full flow pump assembly that is not so important. So in short. Nitrogen and helium gas COPV should be there. Pressure is provided autogenously. Electric TVC eliminates some complexities further in chine area.

High pressure gas for restarting the center Booster engines for return and and landing burns

So the gas is used for spin up but not pressurisation. I wonder how they manage the first few seconds of start up? After a few seconds I imagine the autogenous gas is available for pressurization, but it must take time for autogenous gas pressure to build? Starting with a small head space in the header tank its volume would very rapidly expand and its pressure rapidly decrease unless very promptly repressurizion was applied.

Perhaps there was an issue with one of the Starships that I am confusing this with and maybe autogenous pressurization is just super quick from the get go.

The booster does not have an issue with the boostback burn since there are just a few seconds between MECO and the start of the burn and the tanks are fully pressurised.  By the time the engines are at full thrust the autogenous pressurisation will be working to maintain ullage tank pressure as propellant is used. 

For the landing burn there will have been several minutes with the engines off so the tank pressure will have dropped but possibly not too low.  The engines can start at lower thrust to prevent cavitation so that by the time they need to be at full thrust the tank pressure will be high enough to support that.  However with only two engines running it may be difficult to repressurise the main tanks.  There may be provision to pressurise just the header tanks either from the autogenous pressurisation system or by using helium as a short term expedient. 

The ship will not have an issue for the first few test flight as the engines will not be relighting.  In the longer term they will need to have reliable pressurisation before the engines light.  One option is to store ullage gas in COPVs using electric pumps during powered flight and then release to the tanks just before engine relight.  Another alternative is to have dedicated ullage gas generators which burn gaseous methane and oxygen to heat liquid methane and liquid oxygen to hot gas to pressurise the main tanks before engine start.  The exhaust from these generators could be used as ullage thrusters to settle the propellant tanks.

[darthguili]

That sounds like a pretty good solution.

Let's figure out how long it'll last.  If we use Tlox=80K, Tlch4=67K, and stainless steel emissivity of 0.57, then

J = (0.57)*(5.670E-8)*(80-67)⁴ = 9.2E-4W/m²

If we use a 0.75m x 13m Starship downcomer, that has 5.7m³ or 12.7t of LCH4 in it.  Up thread, I calculated that you'd need cooling of roughly 83kJ/kg to take the methane down to its freezing point.  That'd take more than 1000 days to freeze with radiative cooling alone.  Even if we're off by a factor of 2, we're still good.

I'm sure there are sneak conductive pathways that will overwhelm this number, but I'm convinced that this is a non-problem for the time being--as long as the downcomer truly is a double-walled dewar.

You made a mistake in your radiative exchange calculations. It should be:
J = (0.57)*(5.670E-8)*(80⁴-67⁴) which is three orders of magnitude higher. So the 1000 days become 1 day.

[darthguili]

Yes, surface finish is the dominant factor. But also many figures are temperature dependent. There's no big difference between cryo and room temp. But many considerations of SS emissivity and Starship were around reentry and emissivity at 1000K is quite different from the one at the room temp.

Emissivity is relatively temperature-invariant, isn’t it?  It’s a measure of how close to black a grey body is. Radiant flux, however, scales as the fourth power of temperature.

Emissivity of metals in general increase at high temperatures (>200°C). If I remember correctly SpaceX wants (wanted?) to use that property to radiate heat efficiently with the skin on the side of the starship which is not covered in tiles.
At cryo or anything below ambiant, I don't believe emissivity changes significantly.

[livingjw]

....  One option is to store ullage gas in COPVs using electric pumps during powered flight and then release to the tanks just before engine relight. ...

- Autogenous gases tapped off of the Raptor engine have a pressure of over 6000 psi. They are regulated down to lower pressure to feed the main tanks. A high pressure line could feed COPVs. No electric pump needed.

John

[Chris Bergin]

Massive thread. Let's go to Thread 5. So much has changed over the course of this thread

Thread 5:
https://forum.nasaspaceflight.com/index.php?topic=58830.0