DRACO: NASA and DARPA nuclear propulsion collaboration

[Vultur]

I am thinking more and more that space nuclear in any form isn't a thing for this generation.

Either Starship (and/or other rockets on a similar model) end up working, or not. If so, I think solar and battery, which are being actively and heavily developed, will end up winning out over nuclear. If not, beyond LEO space activities will probably progress at a pretty slow pace.

For beyond Mars, yeah, nuclear would be needed. But that's a long long way away, and probably not worth developing now.

you mean nuclear propulsion or nuclear power?  The thread title is "propulsion", but think you mean "electrical power"

I meant both, actually. I don't think either will have much place in the next, say, 30 years.

[InterestedEngineer]

I am thinking more and more that space nuclear in any form isn't a thing for this generation.

Either Starship (and/or other rockets on a similar model) end up working, or not. If so, I think solar and battery, which are being actively and heavily developed, will end up winning out over nuclear. If not, beyond LEO space activities will probably progress at a pretty slow pace.

For beyond Mars, yeah, nuclear would be needed. But that's a long long way away, and probably not worth developing now.

you mean nuclear propulsion or nuclear power?  The thread title is "propulsion", but think you mean "electrical power"

I meant both, actually. I don't think either will have much place in the next, say, 30 years.

Nuclear electric power on planetary body surfaces will make sense, especially Mars and beyond, for large baseline power that can e.g. survive a 1 month dust storm.

Nuclear Thermal Propulsion makes zero sense in the era of refuelable aerobraking chemical rockets.

So completely two different topics.  This one is the propulsion topic.

[edzieba]

There are no technical barriers to aerobraking with NTRs. There may be safety/environmental ones for aerobraking in Earth's atmosphere (and contamination concerns for doing so elsewhere), but those are mostly the same ones as for launch in the first place.
Then there are destinations without an atmosphere where aerobraking is not an option.

As for nuclear-electric propulsion, that still has its place in the outer solar system where solar-electric is not viable, and for instances where operation in shadow is desirable (e.g. visiting L2 locations of bodies).

[tbellman]

There are no technical barriers to aerobraking with NTRs. There may be safety/environmental ones for aerobraking in Earth's atmosphere (and contamination concerns for doing so elsewhere), but those are mostly the same ones as for launch in the first place.

I suspect InterestedEngineer is actually thinking of aerocapture, or at least aerobraking where you shed lots of velocity per pass.  (Note for InterestedEngineer: aerobraking is when you are already in orbit, and just wants to lower that orbit, while aerocapture is when you arrive at a planet and wants to capture into orbit around the planet.)

Slow aerobraking, where you are shed just a little bit of velocity each pass, and thus don't subject the spacecraft to high aerodynamic loads or high heating, and can spend dozens of orbits to lower your apoapsis, is indeed not a problem for NTRs.  We have done this with at least one solar powered spacecraft at Mars (but I don't remember which at the moment), and those are at least as sensitive as NTRs.

Aerocapture, or fast aerobraking, on the other hand, are more problematic.  There are no hard barriers that make it impossible, but there are other technical constraints with NTRs that make it more challenging.

The basic problem is that nuclear thermal rockets tend to have fairly high dry mass, and can ill afford the shielding and stronger structures needed to survive the aerodynamic loads and heating.

First, NTRs tend to use hydrogen as propellant, otherwise their specific impulse drops very quickly.  And hydrogen has very low density, requiring large and thus heavy tanks.  Keeping the hydrogen liquid also requires insulation, which makes the tanks even heavier.

Another issue is that you often have to mount the engine on a long boom or truss, to get the reactor far away from the payload, in order to lower the radation the payload is exposed to.  The resulting geometry can be difficult to shield, and gives you high bending forces when subjected to the aero loads.

Most NTR designs I have seen even use drop tanks, as they would otherwise have too high dry mass.  And that is difficult to combine with heat shields.

[GuessWho]

NTRs don’t need to rely on aerocapture.  One only wants to aerocapture when you can’t afford the propellant mass to brake propulsively. And aerocapture is a risky process even when you fully understand the atmosphere you are dealing with.  Always better to propulsively capture.  Aerobraking is fine for orbital adjusts once in orbit if time is not critical.  There  are no barriers to doing that with NTRs. LH2 thermal management considerations play but the total energy deposition is low and LH2 has a high specific heat.

Second, you have no refueling infrastructure anywhere that aerocapture (or aerobraking) might serve.  All propellant needed for the return trip has been carried from day one.  The one exception might be if you can launch mostly dry and then load propellant once on orbit.  It maximizes the specific payload mass for that flight system but you still have then added cost/mass of getting that propellant into orbit in the first place.

[Robotbeat]

That is the conventional wisdom, but the performance hit from NOT doing aerocapture or direct entry is so great that it basically neuters any benefit of NTR if you don’t do it.

NTR really doesn’t help you for Mars unless you’re doing aerocapture/direct-entry as well. (And there’s no reason you can’t do orbital rendezvous for refueling, BTW, since you have to do orbital rendezvous anyway for a typical Mars orbit rendezvous NASA architecture.)

The other piece of this puzzle is that fully reusable rockets, besides making in-orbit mass so cheap that it makes Isp not matter as much, are ALSO aerocapture/direct-entry-capable upper stages. You get the aero almost “for free” as part of upper stage reuse. This is often ignored in these kinds of analyses.

What’s also sort of glossed over is that with nuclear thermal, you actually need MORE staging, not less, due to the high dry mass. A single chemical stage with pressure stabilized tanks using a high density propellant may actually have a higher delta-v for small payloads than a nuclear thermal stage does, because the nuclear thermal stage has much greater dry mass and high gravity losses. So to actually get the benefit of nuclear thermal, your vehicle needs to do a bunch of staging with drop tanks and multiple stages and ends up looking like corn on the cob. Particularly if you’re planning on avoiding all aerocapture and doing it all propulsively.

[InterestedEngineer]

There are no technical barriers to aerobraking with NTRs. There may be safety/environmental ones for aerobraking in Earth's atmosphere (and contamination concerns for doing so elsewhere), but those are mostly the same ones as for launch in the first place.
Then there are destinations without an atmosphere where aerobraking is not an option.

As for nuclear-electric propulsion, that still has its place in the outer solar system where solar-electric is not viable, and for instances where operation in shadow is desirable (e.g. visiting L2 locations of bodies).

I've outlined all the technical barriers and project management barriers in prior posts, I won't bother again.  They are all killers to the idea of aerobraking NTRs.

[InterestedEngineer]

the dry mass overhead for Starship braking is < 20t (10t heat shield, < 10t flaps and misc).

That's far less than the dry mass overhead of running Hydrogen in cryo tanks.   1000t of Methalox tanking is about 40t.  1000t of LH2 tanking is about 90t.  90-40 = 50t.  Plus the tanks are so gargantuan coming up with an EDL version is likely impossible.

LH2 simply can't win anything but an Isp score.  It's probably the stupidest propellant ever proposed, in that people keep pushing it long after Methalox has been demonstrated and mission plans worked out.

(and it doesn't matter if it's Hydrolox or NTR).

[Robotbeat]

Also, that’s not parasitic mass for Starship since it’s a lander as well, and landers need a heatshield.

[edzieba]

There are no technical barriers to aerobraking with NTRs. There may be safety/environmental ones for aerobraking in Earth's atmosphere (and contamination concerns for doing so elsewhere), but those are mostly the same ones as for launch in the first place.
Then there are destinations without an atmosphere where aerobraking is not an option.

As for nuclear-electric propulsion, that still has its place in the outer solar system where solar-electric is not viable, and for instances where operation in shadow is desirable (e.g. visiting L2 locations of bodies).

I've outlined all the technical barriers and project management barriers in prior posts, I won't bother again.  They are all killers to the idea of aerobraking NTRs.

I've posted my rebuttals before too. The claims of impossibility don't hold up.

[Robotbeat]

I agree it’s possible. In fact, if you DON’T aerobrake, it won’t be competitive.

[InterestedEngineer]

I agree it’s possible. In fact, if you DON’T aerobrake, it won’t be competitive.

Please explain how you will develop and *test* an aerobrake capable NTR, particular use case going to mars without excessive propulsive braking.  Use an Isp of 800 and compare it to Starship equivalent.  Make sure you include COM calculations issues.

I've done this. Repeatedly.  You can't make it work, from a technical and project management standpoint and financial standpoint.

[Robotbeat]

Oh, it’s incredibly challenging. I made a proposal for it, but it didn’t get funded. Nuclear thermal would only be used for the last stage (used use a chemical stage starting at high Earth orbit to get started), and the proposal also used hydrogen for crew shielding and transpiration cooling of the heatshield (ala Stoke Nova but a high lift configuration). Probably for the best.

[InterestedEngineer]

Oh, it’s incredibly challenging. I made a proposal for it, but it didn’t get funded. Nuclear thermal would only be used for the last stage (used use a chemical stage starting at high Earth orbit to get started), and the proposal also used hydrogen for crew shielding and transpiration cooling of the heatshield (ala Stoke Nova but a high lift configuration). Probably for the best.

How'd you do the iterative testing on aerobraking at Mars?  COM being much farther back thant Starship?  (Which already had problems with COM being too far back)

[edzieba]

Oh, it’s incredibly challenging. I made a proposal for it, but it didn’t get funded. Nuclear thermal would only be used for the last stage (used use a chemical stage starting at high Earth orbit to get started), and the proposal also used hydrogen for crew shielding and transpiration cooling of the heatshield (ala Stoke Nova but a high lift configuration). Probably for the best.

How'd you do the iterative testing on aerobraking at Mars?  COM being much farther back thant Starship?  (Which already had problems with COM being too far back)

Step one would be to not just blindly assume Starship is the only viable option for Mars EDL and try and glom an NTR onto it. There are plenty of other EDL architectures available, from inflatable decelerators to multipart vehicles and many others.

As for testing: test with upper atmosphere Earth EDL just as we do today. You don't need active reactor mass to test structures at subscale or even full scale, boilerplate works just fine for that.

---

Assuming SpaceX's way is the only way, or even the best way, is catastrophically poor engineering practice. They as are subject to unnecessary requirements as anyone else (e.g. the legacy 9m vehicle diameter from back when they still intended to build rocket bodies at Hawthorne).

[Robotbeat]

I think the multipart designs are terrible.

It needs to be a cleansheet design, but the Starship is FAR better than the insane multi piece designs from the 90 day study. Don’t “not invented here” the sensible idea of an integrated vehicle.

And inflatable designs struggle with a high lift to drag ratio, which is important for limiting g forces.

Integrated heatshield is the way to go.

[Robotbeat]

Oh, it’s incredibly challenging. I made a proposal for it, but it didn’t get funded. Nuclear thermal would only be used for the last stage (used use a chemical stage starting at high Earth orbit to get started), and the proposal also used hydrogen for crew shielding and transpiration cooling of the heatshield (ala Stoke Nova but a high lift configuration). Probably for the best.

How'd you do the iterative testing on aerobraking at Mars?  COM being much farther back thant Starship?  (Which already had problems with COM being too far back)

Didnt get that far (just a proposal), but if you look in the literature you can find vehicles with optimized high L/D for aerocapture.

[GuessWho]

That is the conventional wisdom, but the performance hit from NOT doing aerocapture or direct entry is so great that it basically neuters any benefit of NTR if you don’t do it.

Sorry.  That is simply not true.  I would not take the referenced NASA architectures you showed as the optimum solution space.  They are bloated designs.  Having an intimate understanding of the DRACO design, I can say that much better performance can be obtained for a lot less mass than what NASA advertises.  Just saying ...

[InterestedEngineer]

That is the conventional wisdom, but the performance hit from NOT doing aerocapture or direct entry is so great that it basically neuters any benefit of NTR if you don’t do it.

Sorry.  That is simply not true.  I would not take the referenced NASA architectures you showed as the optimum solution space.  They are bloated designs.  Having an intimate understanding of the DRACO design, I can say that much better performance can be obtained for a lot less mass than what NASA advertises.  Just saying ...

Well don't keep us in suspense, spit out some numbers and analyze them.

I am very curious where you are going to pull an extra 7-9km/sec of deltaV out of an NTR system.

[Jim]

Sorry.  That is simply not true.  I would not take the referenced NASA architectures you showed as the optimum solution space.  They are bloated designs.  Having an intimate understanding of the DRACO design, I can say that much better performance can be obtained for a lot less mass than what NASA advertises.  Just saying ...

Sorry.  But that is simply just your opinion and based on several of your last posts, they are typically wrong.

Just saying.....