Replacing SLS/Orion using Starship HLS and Crew Dragon (AI data allowed)

[Roy_H]

If you are moving domes forward to carry more fuel and reduce astronaut space, ...

If you look at the spreadsheet values pictured, you will note that the fuel capacity of the LD is exactly the same as the HLS at 1,200 tons. This is really coincidence more than planning. The way Starships are built, ring by ring, I find it difficult to believe it is a major job to leave a ring or two out. At this rate the LD would actually be shorter than the HLS. Even if it was the same size, just removing the landing thrusters, side air lock, elevator components would result in significant mass (and cost) savings. SpaceX doesn't seem to have a problem extending fuel tank sizes as demonstrated by later versions of Starship.

[crandles57]

If you are moving domes forward to carry more fuel and reduce astronaut space, ...

If you look at the spreadsheet values pictured, you will note that the fuel capacity of the LD is exactly the same as the HLS at 1,200 tons. This is really coincidence more than planning. The way Starships are built, ring by ring, I find it difficult to believe it is a major job to leave a ring or two out. At this rate the LD would actually be shorter than the HLS. Even if it was the same size, just removing the landing thrusters, side air lock, elevator components would result in significant mass (and cost) savings. SpaceX doesn't seem to have a problem extending fuel tank sizes as demonstrated by later versions of Starship.

About that 1200 tons for HLS and LD:

I think we are pretty sure HLS for Artemis III will have 1600 tons fuel at launch. We are less sure about the dry mass but suggestions seem around 130+ tons for a tanker arriving with 150 tons of fuel so the mass fraction is 280/1350. With 1200 ton fuel you can only do  ~280*1200/1350 tons of payload + dry mass yet you have 180 ton dry mass ship (presumably including your 50 ton radiation shielding) and 100 ton payload. You would have to trim that payload to get to orbit with 0 fuel left.

I am also slightly dubious of 130 ton tanker minus TPS, header tanks & fins + 15 ton life support and human occupancy payload + 50 ton radiation shielding + MYI and boil off minimisation + waist thrusters - a couple? of rings is as little as 180 tons.

These are minor nits on the numbers you used which can be fixed by reducing the 100 ton payload which has always seemed nice but excessive to me.

A shorter stubbier HLS makes a lot of sense for lunar landing and ascent. The longer length needed for launch to Earth orbit is wasteful once it has been launched. One way to navigate that is to have the HLS as a 3rd stage but then the problem is when is that substantial development work going to be done? The simpler solution is to accept the standard v3 starship size with 1493 ton boiling point fuel capacity.

Maybe your numbers make more sense than I realise?

[Vultur]

I don't see any reason to have 50 tons of radiation shielding. These aren't multi year missions.

[Roy_H]

I don't see any reason to have 50 tons of radiation shielding. These aren't multi year missions.

From my first posts on page 1: Look at page 32 for table from NASA of recommended radiation shielding for various conditions on this link: https://www.nasa.gov/wp-content/uploads/2020/10/2022-01-05_nasa-std-3001_vol.1_rev._b_final_draft_with_signature_010522.pdf

It gives recommended thickness in water, but PE is marginally more effective. So 15cm of water thickness recommended for beyond earth orbit and < 6 months is equivalent to 14cm of PE. Now I did a very rough calculation for all of crew quarters on HLS, but much less would be required if only for sleeping area as a radiation shelter to be used during radiation storms or transiting Van Alan Belts.

Yes, fewer ST flights for fueling would be required if less shielding is used.

[Roy_H]

About that 1200 tons for HLS and LD:

All of my values for dry mass are just my guess, I have no official source to draw on. Part of the reason I provided the spreadsheet set up so it is easy to put your own values in. It should be noted that the 1,200 tons shown is not at sub-cooled temperature. I have targeted 95°K to 100°k for fuel temperature as both fuels can be the same temperature in liquid form if the CH4 is at 1 bar and the O2 at 6 bar. This is desirable because the tanks share the same bulkhead. This means the HLS and LD tanks would have to be larger than the original 1,200 ton rating at sub cooled temperatures due to fuel expansion.

To the argument that it may not be possible to include 100 tons payload in HLS on initial launch to orbit, I agree this is marginal. However I do not expect the first Moon landing to have a large payload, and subsequent flights will have payload added to the HLS via Starship at 200km LEO.

[Twark_Main]

I don't see any reason to have 50 tons of radiation shielding. These aren't multi year missions.

From my first posts on page 1: Look at page 32 for table from NASA of recommended radiation shielding for various conditions on this link: https://www.nasa.gov/wp-content/uploads/2020/10/2022-01-05_nasa-std-3001_vol.1_rev._b_final_draft_with_signature_010522.pdf

It gives recommended thickness in water, but PE is marginally more effective. So 15cm of water thickness recommended for beyond earth orbit and < 6 months is equivalent to 14cm of PE. Now I did a very rough calculation for all of crew quarters on HLS, but much less would be required if only for sleeping area as a radiation shelter to be used during radiation storms or transiting Van Alan Belts.

Yes, fewer ST flights for fueling would be required if less shielding is used.

This says more about how unrealistic NASA requirements are, or how unrealistic it is to shield the entire habitable volume. 

The correct engineering trade-off should be that taking 1 kg of radiation shielding mass and allocating it differently would give the same number of statistical years of life (based on the overall systems risk analysis) as deleting 1 kg of radiation shielding.

[math needed]

[Roy_H]

I don't see any reason to have 50 tons of radiation shielding. These aren't multi year missions.

From my first posts on page 1: Look at page 32 for table from NASA of recommended radiation shielding for various conditions on this link: https://www.nasa.gov/wp-content/uploads/2020/10/2022-01-05_nasa-std-3001_vol.1_rev._b_final_draft_with_signature_010522.pdf

It gives recommended thickness in water, but PE is marginally more effective. So 15cm of water thickness recommended for beyond earth orbit and < 6 months is equivalent to 14cm of PE. Now I did a very rough calculation for all of crew quarters on HLS, but much less would be required if only for sleeping area as a radiation shelter to be used during radiation storms or transiting Van Alan Belts.

Yes, fewer ST flights for fueling would be required if less shielding is used.

This says more about how unrealistic NASA requirements are, or how unrealistic it is to shield the entire habitable volume. 

The correct engineering trade-off should be that taking 1 kg of radiation shielding mass and allocating it differently would give the same number of statistical years of life (based on the overall systems risk analysis) as deleting 1 kg of radiation shielding.

[math needed]

If I interpret your statement correctly, you are recommending shielding that would be 50% effective, that is half way between levels we get on Earth to the level of no shielding in space. I suspect that is more shielding than you think, and a lot more than NASA's recommendations.

In terms of how unrealistic, it is simply how many dollars you want to spend for crew protection. Cost of a few more ST flights plus the cost of the PE shielding. What I have shown is that it does not cost so much in mass that the mission cannot be performed. I have no objection if the astronauts feel it is excessive and they want less shielding. If you were flying, how much shielding would you want?

[Twark_Main]

I don't see any reason to have 50 tons of radiation shielding. These aren't multi year missions.

From my first posts on page 1: Look at page 32 for table from NASA of recommended radiation shielding for various conditions on this link: https://www.nasa.gov/wp-content/uploads/2020/10/2022-01-05_nasa-std-3001_vol.1_rev._b_final_draft_with_signature_010522.pdf

It gives recommended thickness in water, but PE is marginally more effective. So 15cm of water thickness recommended for beyond earth orbit and < 6 months is equivalent to 14cm of PE. Now I did a very rough calculation for all of crew quarters on HLS, but much less would be required if only for sleeping area as a radiation shelter to be used during radiation storms or transiting Van Alan Belts.

Yes, fewer ST flights for fueling would be required if less shielding is used.

This says more about how unrealistic NASA requirements are, or how unrealistic it is to shield the entire habitable volume. 

The correct engineering trade-off should be that taking 1 kg of radiation shielding mass and allocating it differently would give the same number of statistical years of life (based on the overall systems risk analysis) as deleting 1 kg of radiation shielding.

[math needed]

If I interpret your statement correctly, you are recommending shielding that would be 50% effective, that is half way between levels we get on Earth to the level of no shielding in space.

You are not interpreting me correctly. In fact I struggle to see how my words were misinterpreted like that.

What I actually said was that if you take 1 kg away from the radiation shield, and that 1 kg is allocated to other systems in the ideal (risk-minimizing) way, then it should give you the same number of years of statistical life as you gained by that 1 kg of shielding. In other words, the two numbers should be identical.

Otherwise the risk in the system is unbalanced, and you could reduce the total risk by reducing the radiation shield mass and re-allocating that mass to other systems.

If you were flying, how much shielding would you want?

I would want the thing I actually wrote, not the misunderstanding. 

[meekGee]

I don't see any reason to have 50 tons of radiation shielding. These aren't multi year missions.

From my first posts on page 1: Look at page 32 for table from NASA of recommended radiation shielding for various conditions on this link: https://www.nasa.gov/wp-content/uploads/2020/10/2022-01-05_nasa-std-3001_vol.1_rev._b_final_draft_with_signature_010522.pdf

It gives recommended thickness in water, but PE is marginally more effective. So 15cm of water thickness recommended for beyond earth orbit and &lt; 6 months is equivalent to 14cm of PE. Now I did a very rough calculation for all of crew quarters on HLS, but much less would be required if only for sleeping area as a radiation shelter to be used during radiation storms or transiting Van Alan Belts.

Yes, fewer ST flights for fueling would be required if less shielding is used.

This says more about how unrealistic NASA requirements are, or how unrealistic it is to shield the entire habitable volume. 

The correct engineering trade-off should be that taking 1 kg of radiation shielding mass and allocating it differently would give the same number of statistical years of life (based on the overall systems risk analysis) as deleting 1 kg of radiation shielding.

[math needed]

If I interpret your statement correctly, you are recommending shielding that would be 50% effective, that is half way between levels we get on Earth to the level of no shielding in space.

You are not interpreting me correctly. In fact I struggle to see how my words were misinterpreted like that.

What I actually said was that if you take 1 kg away from the radiation shield, and that 1 kg is allocated to other systems in the ideal (risk-minimizing) way, then it should give you the same number of years of statistical life as you gained by that 1 kg of shielding. In other words, the two numbers should be identical.

Otherwise the risk in the system is unbalanced, and you could reduce the total risk by reducing the radiation shield mass and re-allocating that mass to other systems.

If you were flying, how much shielding would you want?

I would want the thing I actually wrote, not the misunderstanding. 

How do you compare long term chronic risk (e.g. radiation) to short term acute risk (life support or landing legs)?

How do you compare risknof death and risk of severe injury?   Risk now vs. risk later?

One number does not capture all aspects of harm, and the probabilities themselves are uncertain, so trade-offs are not simple one-variable arithmetic.

[Twark_Main]

I don't see any reason to have 50 tons of radiation shielding. These aren't multi year missions.

From my first posts on page 1: Look at page 32 for table from NASA of recommended radiation shielding for various conditions on this link: https://www.nasa.gov/wp-content/uploads/2020/10/2022-01-05_nasa-std-3001_vol.1_rev._b_final_draft_with_signature_010522.pdf

It gives recommended thickness in water, but PE is marginally more effective. So 15cm of water thickness recommended for beyond earth orbit and &lt; 6 months is equivalent to 14cm of PE. Now I did a very rough calculation for all of crew quarters on HLS, but much less would be required if only for sleeping area as a radiation shelter to be used during radiation storms or transiting Van Alan Belts.

Yes, fewer ST flights for fueling would be required if less shielding is used.

This says more about how unrealistic NASA requirements are, or how unrealistic it is to shield the entire habitable volume. 

The correct engineering trade-off should be that taking 1 kg of radiation shielding mass and allocating it differently would give the same number of statistical years of life (based on the overall systems risk analysis) as deleting 1 kg of radiation shielding.

[math needed]

If I interpret your statement correctly, you are recommending shielding that would be 50% effective, that is half way between levels we get on Earth to the level of no shielding in space.

You are not interpreting me correctly. In fact I struggle to see how my words were misinterpreted like that.

What I actually said was that if you take 1 kg away from the radiation shield, and that 1 kg is allocated to other systems in the ideal (risk-minimizing) way, then it should give you the same number of years of statistical life as you gained by that 1 kg of shielding. In other words, the two numbers should be identical.

Otherwise the risk in the system is unbalanced, and you could reduce the total risk by reducing the radiation shield mass and re-allocating that mass to other systems.

If you were flying, how much shielding would you want?

I would want the thing I actually wrote, not the misunderstanding. 

How do you compare long term chronic risk (e.g. radiation) to short term acute risk (life support or landing legs)?

How do you compare risknof death and risk of severe injury?   Risk now vs. risk later?

If you prefer, use discounted (now vs later) or disability adjusted (death vs injury) numbers. We already know how to do that.

But do some math, instead of squinting and saying that X or Y amount of radiation shielding sounds about right, which certainly fails to "capture all aspects of harm."

One number does not capture all aspects of harm, and the probabilities themselves are uncertain, so trade-offs are not simple one-variable arithmetic.

What, exactly, is your proposed alternative?   

Is "winging it" (zero variable arithmetic) somehow any better?  If not, then you should be directing all this ire at Roy instead of me.... 

[Roy_H]

If I interpret your statement correctly, you are recommending shielding that would be 50% effective, that is half way between levels we get on Earth to the level of no shielding in space.

You are not interpreting me correctly. In fact I struggle to see how my words were misinterpreted like that.

What I actually said was that if you take 1 kg away from the radiation shield, and that 1 kg is allocated to other systems in the ideal (risk-minimizing) way, then it should give you the same number of years of statistical life as you gained by that 1 kg of shielding. In other words, the two numbers should be identical.

Otherwise the risk in the system is unbalanced, and you could reduce the total risk by reducing the radiation shield mass and re-allocating that mass to other systems.

Thank you for pointing out that I completely mis-understand you. However, maybe I am a little dense and still do not have a clear idea of your statement. I am aware that there are other elements that provide shielding, the ss skin, electrical equipment, supplies including clothing, space suits, drinking water, food etc. and if these items are placed on outer walls, they provide some protection to astronauts inside. However I doubt there is a 1:1 shielding value.

I do not want to get into an argument about exactly how much shielding is required/desired and the composition of the shielding. That is engineering details far beyond my very rough estimates. I only wanted to point out that 50t of shielding is possible and the mission would still be successful.

[Twark_Main]

If I interpret your statement correctly, you are recommending shielding that would be 50% effective, that is half way between levels we get on Earth to the level of no shielding in space.

You are not interpreting me correctly. In fact I struggle to see how my words were misinterpreted like that.

What I actually said was that if you take 1 kg away from the radiation shield, and that 1 kg is allocated to other systems in the ideal (risk-minimizing) way, then it should give you the same number of years of statistical life as you gained by that 1 kg of shielding. In other words, the two numbers should be identical.

Otherwise the risk in the system is unbalanced, and you could reduce the total risk by reducing the radiation shield mass and re-allocating that mass to other systems.
[/​quote]

Thank you for pointing out that I completely mis-understand you. However, maybe I am a little dense and still do not have a clear idea of your statement. I am aware that there are other elements that provide shielding, the ss skin, electrical equipment, supplies including clothing, space suits, drinking water, food etc. and if these items are placed on outer walls, they provide some protection to astronauts inside. However I doubt there is a 1:1 shielding value.

I mean that there are lots of things that can "eat mass" to improve survival: better MMOD shielding, more emergency rations, oxygen, CO2 scrubbers, medical supplies, etc.

All of these things improve survivability if you just throw more mass at them, but you have to have a "cutoff" where you decide the extra mass isn't worth it to increase safety any more.

So you have a priority list where you say "if I had an extra 500 kg I would put X into medical kit, Y into better seat shock absorption" etc.

What I'm saying is, these things should all be balanced.  Let the system "find its own level," and the radiation shield mass should compete on a level playing field with all that other risk reduction hardware.

Or, put another way, "if you can replace a ton of radiation shield with other stuff and I'm less likely to die overall, then I want that instead."

[Roy_H]

What I'm saying is, these things should all be balanced.  Let the system "find its own level," and the radiation shield mass should compete on a level playing field with all that other risk reduction hardware.

Of course, I didn't mean to suggest that anything else should be sacrificed in order to have extra shielding. I added it to the dry mass of the ship, which I expected to already include these things. I expect my "excess shielding" adds 2% to 3% of the cost of a Moon mission. Please consider it as an optional extra.

I absolutely agree with you, it should be a balanced approach.

[Twark_Main]

What I'm saying is, these things should all be balanced.  Let the system "find its own level," and the radiation shield mass should compete on a level playing field with all that other risk reduction hardware.

Of course, I didn't mean to suggest that anything else should be sacrificed in order to have extra shielding. I added it to the dry mass of the ship, which I expected to already include these things. I expect my "excess shielding" adds 2% to 3% of the cost of a Moon mission. Please consider it as an optional extra.

I absolutely agree with you, it should be a balanced approach.

Yep, I took the "50 tons" number to be mostly a placeholder (and some mass growth margin), and it sounds like that's the right way to interpret it.


To put it technically, each hardware category gets truncated once the diminishing marginal risk mitigation per mass reaches the same value across the entire vehicle. Say it ends up being 4 minutes of statistical life per kilogram. So you truncate the MMOD shield mass when the last kilogram gained 4.001 minutes but the next kilogram "only" gains 3.998 minutes. Same for medical supplies, life support consumables, and of course the radiation shield.

Basically sort the "wish list" by minutes / kg, decide how many kilograms you have, and then just keep going down the sorted list adding items until you run out of kilograms. This procedure should result in the safest vehicle possible that still fits into the mass constraints.

Obviously we're not going to do all that in this thread, but in the "real world" that's how you'd solve this problem in (something approaching) an optimal way.

[Vultur]

If you were flying, how much shielding would you want?

I would much rather spend that mass on preventing things that could lead to short term death, like spare parts, extra life support consumables, etc. I wouldn't worry too much about radiation shielding for the whole habitable volume (vs a solar storm shelter).

Part of the issue is that risk vs risk tradeoffs depend on your assumptions about chronic radiation risk. NASA probably believes in the linear-no-threshold model; I don't, so I think a lot of the risk NASA assumes is being mitigated from chronic radiation doses is risk that doesn't actually exist in the first place.

And even that is more for really long missions. For an Artemis 3 style mission, I wouldn't really worry much about radiation at all. Apollo didn't very much.

[Twark_Main]

Part of the issue is that risk vs risk tradeoffs depend on your assumptions about chronic radiation risk. NASA probably believes in the linear-no-threshold model; I don't, so I think a lot of the risk NASA assumes is being mitigated from chronic radiation doses is risk that doesn't actually exist in the first place.

I don't want to get too far down the LNT rabbit hole, but it should be said that Vultur is hardly taking a pseudoscientific position here.





The pro-LNT arguments are... rather lacking. These are not the sorts of papers you write when you stand on firm scientific ground. 

https://pubmed.ncbi.nlm.nih.gov/39222266/

https://www.nrdc.org/bio/bemnet-alemayehu/hold-fast-linear-no-threshold-radiation-protection

Anyway let's not go into a huge tangent, but hopefully these links are helpful for people who want to dig deeper.

[Vultur]

The pro-LNT arguments are... rather lacking. These are not the sorts of papers you write when you stand on firm scientific ground. 

I think that's because it's more a conservative default assumption than something that's positively supported.

IE, it's almost certainly wrong as a factual model of what happens in the real world. OTOH, "all models are wrong, but some are useful".

If one is an agency required to set *some* limit, in the absence of sufficient data about low dose/long exposure time, it is arguable that linear no threshold is a useful/safe assumption. I think that may be a reasonable argument for general-population standards (but probably not for special cases like astronauts). But in that case, the model is being used for a specific purpose - setting a regulatory standard. It may be fine for that use, but shouldn't then be considered to necessarily represent the "real" risk for other purposes.

(The opposite extreme would be the assumption that increased risk is zero, or even that risk decreases [hormesis], below the lowest level where harm is well demonstrated.

Reality is probably somewhere in between.)

This is pretty normal for setting regulatory safety limits. Since the limits are being set well below "obvious acute nasty effects", there's a lot of arguable extrapolation involved. The establishment of a limit at X doesn't usually mean that some obvious bad effects happens at X+a tiny bit.

[Roy_H]

Anyway let's not go into a huge tangent, but hopefully these links are helpful for people who want to dig deeper.

So some low level of radiation is actually beneficial. The Apollo astronauts had very little radiation protection, and no health problems attributable to radiation exposure.

I think we need to do some studies, like exposing mice to various levels/types of radiation to establish some realistic levels of danger. Such studies have been done. Trying to put this into some rough perspective:

20mGy/day         lethal (long term exposure, rats)
0.137 mGy/day    actually beneficial (extended life in beagle dogs 15%)
1.37 mGy/day     NASA acceptable guide line.

From: https://pmc.ncbi.nlm.nih.gov/articles/PMC5347275
"The optimum lifespan increase appeared at 50 mGy/y. The threshold for harm (decreased lifespan) was 700 mGy/y for 50% mortality dogs and 1100 mGy/y for short-lived dogs."

[Twark_Main]

The pro-LNT arguments are... rather lacking. These are not the sorts of papers you write when you stand on firm scientific ground. 

I think that's because it's more a conservative default assumption than something that's positively supported.

IE, it's almost certainly wrong as a factual model of what happens in the real world. OTOH, "all models are wrong, but some are useful".

If one is an agency required to set *some* limit, in the absence of sufficient data about low dose/long exposure time, it is arguable that linear no threshold is a useful/safe assumption. I think that may be a reasonable argument for general-population standards (but probably not for special cases like astronauts). But in that case, the model is being used for a specific purpose - setting a regulatory standard. It may be fine for that use, but shouldn't then be considered to necessarily represent the "real" risk for other purposes.

(The opposite extreme would be the assumption that increased risk is zero, or even that risk decreases [hormesis], below the lowest level where harm is well demonstrated.

Reality is probably somewhere in between.)

This is pretty normal for setting regulatory safety limits. Since the limits are being set well below "obvious acute nasty effects", there's a lot of arguable extrapolation involved. The establishment of a limit at X doesn't usually mean that some obvious bad effects happens at X+a tiny bit.

Yep, and honestly I can sympathize. LNT is "bad science but good policy," because if there's an acceptable dose when who gets to profit off it?

The company that makes my drywall wants to "use it" so they can source more radioactive gypsum. The construction company wants to use more radioactive gravel on roads and housing development. Ditto for granite countertops, brick facades, etc. And of course the local coal power plant wants to not get penalized for radiation because it's "actually healthy."

By assuming LNT, every potential radiation exposure source gets "policed" on a level playing field. Nice and simple.

Anyway this is already too far down the rabbit hole as it is...

[Twark_Main]

Bringing this back around, while NASA currently uses LNT for cancer risk modeling, they are notably open to using non-LNT models once they have more data on low-dose radiation risks.


https://www.ncbi.nlm.nih.gov/books/NBK189539/

The estimation of human cancer and non-cancer effects at low doses (less than 100 millisievert [mSv]) is based on the epidemiological data from atomic bomb survivors together with selected data for occupational and medical exposures. There is a continued reliance on the assumption that, at these low doses, a given increment in dose produces a directly proportionate increase in the probability of the development of cancer or heritable effects that are attributable to the radiation. This relationship is described as the linear no-threshold (LNT) model. The ICRP, for example “considers that the adoption of the LNT model combined with a judged value of a dose and dose rate effectiveness factor (DDREF) provides a prudent basis for the practical purposes of radiological protection, i.e., the management of risks from low-dose radiation exposure” (ICRP, 2007, p. 51). This is an important position because the LNT hypothesis and some of the other assumptions behind the estimation of risks are based on models and projections and not on direct scientific observation.

https://ntrs.nasa.gov/api/citations/20250001731/downloads/nscr_usafsymposium_March2025.pdf

Uncertainties not yet quantified in risk projections
...
  - Alternatives to linear no-threshold at low dose (i.e. non-targeted effects)

So NASA, at least, is keeping their powder dry.