SpaceX Dragon XL

[yg1968]

There is some interesting information about Dragon XL in today's GAO Report:

The project's cost and schedule depends on the timing of NASA providing SpaceX with authority to proceed with work on the task order. According to project officials, NASA did not provide SpaceX with authority to proceed in October 2020, as originally planned, due to funding constraints from operating under a continuing resolution and NASA having other funding priorities. As a result, there is a risk that the logistics mission may not be capable of supporting the Artemis III mission at the end of 2024. If NASA had provided SpaceX with the authority to proceed in October 2020, SpaceX would have planned to develop the logistics vehicle and launch it in October 2024 to support the Artemis III mission. Project officials said they are evaluating whether using a fast transit capability, which increases the cost of SpaceX's task order, could help the project support the Artemis III mission time frames. This capability increases the speed that the logistics vehicle arrives in lunar orbit by using expendable rather than reusable first stages for all three cores of the Falcon Heavy to increase launch capability.

https://www.gao.gov/assets/gao-21-306.pdf

[butters]

That's nonsense. A faster departure velocity for NRHO might save days, but it won't save months.

[cohberg]

That's nonsense. A faster departure velocity for NRHO might save days, but it won't save months.

This choice represents a mission tradeoff - the fast transit consists of a tranit duration of three days and requires 450 m/s to perform the NRHO insertion burn; the slow transit consists of a transit duration of 120 days and a 30 m/s NRHO insertion burn.

*The numbers could be different for Dragon.

[yg1968]

That's nonsense. A faster departure velocity for NRHO might save days, but it won't save months.

If you read, the GAO's comments relating to the PPE, there is a lot of schedule risks related to it, so it likely won't matter. I understand why NASA prefers not to use Gateway for Artemis III. I suspect that Gateway won't be ready until 2026. 

https://forum.nasaspaceflight.com/index.php?topic=51452.msg2241208#new

[gongora]

Another note from the GAO report:

In addition, project officials said SpaceX is conducting ground-based radiation testing on select components to inform potential design trades NASA may need to make to extend mission duration.

[DigitalMan]

Interesting, gongora.

How does one go about doing radiation testing?

- X-Ray machines?
- Visit Chernobyl or Fukushima?
- High-flying aircraft?

I suppose the DOE could be of great help for this.

[eeergo]

Interesting, gongora.

How does one go about doing radiation testing?

- X-Ray machines?
- Visit Chernobyl or Fukushima?
- High-flying aircraft?

I suppose the DOE could be of great help for this.

Far too low values in #2, also non-representative radiation 'flavors'. Likewise for #3, although it's just more expensive while more representative.

They probably stick the components in/near nuclear reactors (especially for neutrons), or better still utilize particle accelerators, where you can better tune which radiation, with which energy and under which additional backgrounds, you're applying. "X-ray" machines are at the enf of the day a low-energy, small subclass of particle accelerators. You need more penetrating/higher stopping-power radiation (think neutrons, protons, atomic nuclei, high energy electrons or gammas) to account for deep space effects.

[ugordan]

They probably stick the components in/near nuclear reactors (especially for neutrons)

AFAIK there aren't any neutrons in deep space given that free neutrons have lifetimes far too short to reach Earth from any plausible neutron source, even if you put in a ludicrous amount of time dilation (and hence energy into them).

[eeergo]

They probably stick the components in/near nuclear reactors (especially for neutrons)

AFAIK there aren't any neutrons in deep space given that free neutrons have lifetimes far too short to reach Earth from any plausible neutron source, even if you put in a ludicrous amount of time dilation (and hence energy into them).

But there are "fresh" ones once suitable "cosmic rays", such as protons or atomic nuclei, collide with condensed materials, due to spallation. Neutrons are an essential part of space radiation effects (both biologically, electronically and wrt material ageing).

[scientist]

Interesting, gongora.

How does one go about doing radiation testing?

- X-Ray machines?
- Visit Chernobyl or Fukushima?
- High-flying aircraft?

I suppose the DOE could be of great help for this.

Best way is to use particle accelerators to irradiate the components with electron, proton and ion beams. But beam time availability is typically limited, especially for heavier ions and higher energy beams.

Fast and cheap way to do it is for example with a strong Co-60 gamma ray source, and then use computer simulations to translate the result into the actual expected particle flux environment in space.

[ugordan]

They probably stick the components in/near nuclear reactors (especially for neutrons)

AFAIK there aren't any neutrons in deep space given that free neutrons have lifetimes far too short to reach Earth from any plausible neutron source, even if you put in a ludicrous amount of time dilation (and hence energy into them).

But there are "fresh" ones once suitable "cosmic rays", such as protons or atomic nuclei, collide with condensed materials, due to spallation. Neutrons are an essential part of space radiation effects (both biologically, electronically and wrt material ageing).

Yes, but then you're not testing against the actual cosmic ray background since you're inputting neutrons, vs. them being produced via spalling in whatever mix of materials you're testing so the actual doses wouldn't be representative nor would any side effects of that neutron production in situ match the "real world".

[eeergo]

They probably stick the components in/near nuclear reactors (especially for neutrons)

AFAIK there aren't any neutrons in deep space given that free neutrons have lifetimes far too short to reach Earth from any plausible neutron source, even if you put in a ludicrous amount of time dilation (and hence energy into them).

But there are "fresh" ones once suitable "cosmic rays", such as protons or atomic nuclei, collide with condensed materials, due to spallation. Neutrons are an essential part of space radiation effects (both biologically, electronically and wrt material ageing).

Yes, but then you're not testing against the actual cosmic ray background since you're inputting neutrons, vs. them being produced via spalling in whatever mix of materials you're testing so the actual doses wouldn't be representative nor would any side effects of that neutron production in situ match the "real world".

Sorry to insist, but no: you know (pretty well) the energy spectrum of the incoming cosmic radiation (in LEO, deep space or whatever), within certain error bounds and tolerances, and which kinds of spallation products that flux will produce, particularly when knowing by design the materials projected to be present around it. You also know the neutron output of a reactor, and its spectrum, to very high precision, so you can adjust where to put the component and how to shield it to make it as representative a test as possible. MCNP / Scale-style software can inform as to the representativity of the doses received. That would mimic spallation neutrons without actually spallating on anything (it'd be cumbersome to test the component with the actual stuff around it up to several meters). If you really want an in-depth analysis of actual spallation, you can try putting representative stuff around the component and stick it in a particle beam, but that will come with many other sets of limitations and uncertainties, so it's a complementary approach, not necessarily a better substitute.

[Robotbeat]

You can use beam lines to recreate the same spectra as deep space. That’s how it’s done.

[eeergo]

You can use beam lines to recreate the same spectra as deep space. That’s how it’s done.

Believe me, I know for a fact reactors are also used, no need for such definitive statements (the gist of which I also stated, btw). Not sure about this particular instance for Dragon XL, of course.

OTOH, beam lines have huge limitations wrt simulating cosmic ray spectra. Far cry from "recreating the same spectrum".

[1]

Interesting, gongora.

How does one go about doing radiation testing?

- X-Ray machines?
- Visit Chernobyl or Fukushima?
- High-flying aircraft?

I suppose the DOE could be of great help for this.

Should probably make a dedicated thread for this, but in general, radiation concerns follow three* avenues.

1) total ionizing dose (TID) effects
2) displacement damage (DD) effects
3) single event effects (SEE)

Items 1 and 2 a degradative effects that manifest increasingly over time. Often, it is sufficient to deal with items 1 and 2 on a component by component basis. Item 3 is are rapid onset events that may or may not persist. Because of this, it is necessary to deal with SEE on a system-wise/signal chain basis.

Generally speaking, the vast majority of radiation you'll encounter in space will be due to relatively low energy protons (mainly solar protons and trapped protons if you're near a magnetic field). The immediate concern with be with items 1 and 2. Protons striking an object will cause a combination of TID and DD effects simultaneously. TID because protons have an electric charge and secondary effects like bremsstrahlung will cause the struck object to be subjected to ionizing EM photons, and DD because the mass of the proton itself can cause physics damage to the atomic structure of an object. Because these protons are relatively low energy and not very penetrating, TID and DD effects are often mitigated by adding shielding to circuits or your spacecraft as a whole. Or, just using more toleratant technologies.

Testing on the ground is done to isolate the effects of items 1 and 2 from each other. Total ionizing dose degradation is measured by exposing your test sample to gamma radiation; often from the Co-60 source mentioned above. Displacement damage degradation is independently measured by exposing another test sample to neutrons from a reactor. Over the decades, a lot of effort has been put into normalizing the natural space environments against these two tests. Thus, based upon the response to both tests, and overall part response can be predicted for a given orbit and mission lifetime. This is all ultimately baked in to your worst-case circuit analysis where the designers must assign sufficient margin to the system to ensure operational ability for the duration of the mission.

Item 3 effects mainly bistable circuitry, where your system can be unexpectedly kicked from one configuration into another.  The purpose of SEE testing is to understand the effects of very highly energetic protons (cosmic rays) or natural heavier ions from some other origin. These are essentially so energetic that no reasonable amount of shielding can be used to block them. You don't get a lot of total dose or displacement damage effects from cosmic rays, because they're relatively few in number. However, if they hit your circuit in just the right spot, they will cause an impact to your system which can propagate far beyond the immediate device impacted. When possible, integrated circuits are designed to be SEE resistant, but often one simply has to settle for a 'good enough' circuit where the anticipated impact rate is sufficiently low.

For single event effects testing, circuits are exposed to accelerated particles, usually heavy ions,) at varying energies. What this actually translates into is measuring the amount of energy deposited within a certain depth of the active layers of your circuity. Testing begins at high energy, and then the ion energy is reduced. The upset rates slowly drop until a threshold value where upset rates suddenly plummet. This then informs the tester where the SEE onset threshold lies (you actually try to fit weibull curves to the data), and then one can calculate an upset rate based upon the anticipated environment. If the onset threshold is sufficiently low, then testing with accelerated protons is needed to probe lower energies with higher fidelity. If the threshold is still too low, then the part cannot be used as it can never be assumed to be in a nominal state of operation.

In general, resistance to one type of effect does not promise resistance to any other kind. CMOS circuitry, for example, is generally very resistance to displacement damage, but may not be resistant to total ionizing dose. Thus, the need for independent testing, because simple exposure to protons would leave the tester unsure of which effect is causing issues in the circuit. Silicon-Germanium is generally resistant to both TID and DD effects, but is highly sensitive to SEE.

Although a rocket's lifetime may only be hours (if not minutes) the existing Dragon and Starlink efforts will provide spacex with a lot of info about the longevity of their systems on orbit from a radiation standpoint. They should have plenty of internal competence to design DragonXL to last.


*there also exist prompt dose and neutron induced upset concerns, but these generally only apply to weapons environments and can be safely ignored in regards to Dragon.

[Comga]

Interesting, gongora.

How does one go about doing radiation testing?

- X-Ray machines?
- Visit Chernobyl or Fukushima?
- High-flying aircraft?

I suppose the DOE could be of great help for this.

Besides member "1"'s very long and detailed post, and several guesses and chatty replies, the short answer is that radiation testing is available as a commercial service from several providers.  Some of these are universities or federal labs.  Some are commercial outfits.  They provide calibrated doses and fluences of the various types of radiation listed, electrons, protons, gamma rays, all at various energies.  There are test zones into which samples and subsystems, either unpowered or powered, can be placed.

A few years ago I sent a bunch of blue emitting LEDs on small circuit boards to LLNL, if memory serves, for proton irradiation at a series of doses.  This was for a NASA mission.  (The result was that they were not affected by doses well above that expected on the mission.) We were piggy-backing on a day of testing of other components from another program.

Before that I had a bunch of glass and crystal samples sent to an outfit in Colorado Springs for similar dosing.

They are all done in labs with well controlled sources, not exotic locations within incidental emissions.

[mshear]

Interesting, gongora.

How does one go about doing radiation testing?

- X-Ray machines?
- Visit Chernobyl or Fukushima?
- High-flying aircraft?

I suppose the DOE could be of great help for this.

Besides member "1"'s very long and detailed post, and several guesses and chatty replies, the short answer is that radiation testing is available as a commercial service from several providers.  Some of these are universities or federal labs.  Some are commercial outfits.  They provide calibrated doses and fluences of the various types of radiation listed, electrons, protons, gamma rays, all at various energies.  There are test zones into which samples and subsystems, either unpowered or powered, can be placed.

A few years ago I sent a bunch of blue emitting LEDs on small circuit boards to LLNL, if memory serves, for proton irradiation at a series of doses.  This was for a NASA mission.  (The result was that they were not affected by doses well above that expected on the mission.) We were piggy-backing on a day of testing of other components from another program.

Before that I had a bunch of glass and crystal samples sent to an outfit in Colorado Springs for similar dosing.

They are all done in labs with well controlled sources, not exotic locations within incidental emissions.

My experience is 20+ years out of date, but similar - I interned at GSFC in the late 90s (fell into that from an argument on a sci.space Usenet group in which I - incorrectly but rationally - was arguing that RTGs shouldn't give off any significant neutron flux, the guy with whom I was arguing turned out to be a GSFC PI who was willing to take a summer intern and radiation effects on electronics was one focus of his lab) and we used an in-house (and fairly heavily scheduled) Co-60 source at Goddard for gamma / total dose testing as well as going to Brookhaven national labs to use their accelerator for heavy ion testing to simulate GCR.  In a different internship a few years later, this time at one of the DOE labs, we similarly used an in-house Co-60 source and also had an agreement with a local research reactor that had a fixture that could take devices into the core to test neutron tolerance (had to be actually in the core to get a high enough neutron flux, so fair gamma flux also) - being a DOE lab, they were concerned about terrestrial applications so worried much more about neutrons than heavy ions.  We had pretty good predictions about dose rate / total dose, but always used dosimeters as well to confirm.

(Not in any remotely related industry anymore, which I occasionally regret...)

[RoadWithoutEnd]

I understand why NASA prefers not to use Gateway for Artemis III. I suspect that Gateway won't be ready until 2026. 

I would hope they prefer not to use Gateway because it's irrelevant to landing missions, and thus a pointless source of cost and delay if included in the plan.

[NaN]

I understand why NASA prefers not to use Gateway for Artemis III. I suspect that Gateway won't be ready until 2026. 

I would hope they prefer not to use Gateway because it's irrelevant to landing missions, and thus a pointless source of cost and delay if included in the plan.

Gateway long predates Artemis and exists for its own reasons - to be a deep-space replacement for ISS, in addition to commercial station(s) in LEO.

So if you think of Artemis solely as boots on the moon, then you are correct that Gateway is a pointless source of cost and mission complexity. But on the other hand, once Gateway is available then doing it this way saves them mounting separate deep-space missions for Gateway expeditions and for lunar surface expeditions. And given the low flight rate (and high cost) of SLS/Orion they have little choice but to stitch them together until there are alternatives for getting crew up there. I think it's a pretty good compromise as long as they don't let one program's delays hold back the other, and seems like that decision has already been made.

[yg1968]

There is some nice images of Dragon XL in the tweet below:

Behold

Brand new renderings of the Gateway, an orbital outpost around the Moon that provides vital support for #Artemis astronauts on their way to the lunar surface, have been added to the library!

View the new 4k images:

https://twitter.com/NASAArtemis/status/1397553110743240708

Link to obligated amounts for the Dragon XL award:
https://www.usaspending.gov/award/CONT_AWD_80KSC020C0012_8000_-NONE-_-NONE-
https://govtribe.com/award/federal-contract-award/definitive-contract-80ksc020c0012