New Mars Forums

PhotonBytes

Member

From: USA

Registered: 2019-12-28

Posts: 29

Dual Stage 4 Grid Ion Drive possibilities

I would like to invite you to peer review my idea.

Has anyone heard of the Dual Stage 4 Grid Ion Drive? Please check out the numbers:

https://en.wikipedia.org/wiki/Dual-Stage_4-Grid
https://www.esa.int/gsp/ACT/doc/PRO/ACT … C4.4.7.pdf

Based on these numbers:

https://1drv.ms/x/s!Aotveh8OjBiliuBwbLH … Q?e=dvGeRF

I created a design of an interplanetary spacecraft that  will maximize surface area for the drive to maximize the thrust as the thrust is a function of surface area. I also noticed that solar panels are optimized for this too so on the sides of the craft that is cubic shape I also placed solar panels which would be useful when the ship is either moving slowly or in an orbit around a planet and not in transit.

The ship has to be nuclear powered and I decided for this thought experiment to make it a Fast Neutron Reactor and got the figures of wikipedia. If there are any issues with this decision please let me know. I just assumed that it would be lighter than 10 tons which I imagined to be the mass of the ship's empty weight. The nuclear fuel U238 will have to be very highly enriched and able to breed. I feel I made a mistake in my numbers because I didn't factor in the return trip so I really should have doubled the numbers of nuclear fuel and propellant, so there really should be another 10 tons on top of the 20 tons fully fueled. 10 ton empty + 20 ton of nuclear fuel and propellant should give us a very comfortable ride to Mars or Titan at 30 tons. Perhaps throw in a lander too because this ship is not designed to land on any moons or planets. Oh and maybe factor in food, water and crew another 5 tons for 35 tons. But if you fly in a convoy of at least 2 or more cubes then things like landers can just be on one or two ships and the others can store food and water. Then you're back down to an average mass of 30 tons but now you also can link up and have artificial gravity. See below for more details.

Although you still could have done it with the initial configuration but reduce the to and return trip and still be more comfortable than ISS astronauts sometimes spend at least a year there.

You could also have a cube sent ahead with a lander but it took it's sweet time controlled by AI with no crew but plenty of return fuel and propellant. Then you could fly the initial configuration ship there at warp speed and have a return cube waiting for you to take you home after your stay.

The ship is a scaffold but it can be optionally panelled and pressurized inside to maximize breathable volume allowing 4,900 cubic meters of habitable volume. The ship will need the ability to radiate away nuclear reactor heat by running coolant through the scaffolding and transfer the heat into any panels added so that it can radiate into space.

From the numbers I generated for specific impulse and my assumptions for needed fuel and mass for a ship, I came to the following conclusions about what is possible for a team to visit Mars and Titan:

Cut travel time to Mars from 7 months to 3 weeks.

Cut travel time to Titan from 7 years to 7 months.

First and last 12 hours of interplanetary travel leg will give artificial gravity to the ship especially if it has multiple deck levels. Although you will only feel 1/10th Earth's gravity at 1m/s/s it's still better than nothing. You can also rotate the single cube for some artificial gravity but there would be some  Coriolis effect because the diameter is only 17m.

Flying a 2nd, 3rd or 4th cubic ship in formation will give you the option to connect them all together for more artificial gravity by rotating a rectangular interconnected mass. At a planetary destination such as Mars or Titan they can link up in greater numbers or with longer scaffold for slower rotation in exchange for longer structure for less Coriolis effect. This would be great for Geosync orbits above Earth/Mars too so that station crew will have a 24 hour day/night.

Last edited by PhotonBytes (2023-04-22 07:15:36)

---

I play the piano

tahanson43206

Moderator

Registered: 2018-04-27

Posts: 17,322

Re: Dual Stage 4 Grid Ion Drive possibilities

For PhotonBytes Re New Topic!

Best wishes for success with this interesting thought experiment!

(th)

PhotonBytes

Member

From: USA

Registered: 2019-12-28

Posts: 29

Re: Dual Stage 4 Grid Ion Drive possibilities

Update with AI chatbot:

https://chat.openai.com/share/41ea06bb- … a33d395b7d

AI has given some additional input on my idea:
1. To maximize power generation with solar within 1AU one side should face the sun directly
2. RPM of 2.34 is required to maintain 1g with an internal centrifuge taking up a slice of the cube.

---

I play the piano

Steve Stewart

Member

From: Kansas (USA)

Registered: 2019-09-21

Posts: 161

Website

Re: Dual Stage 4 Grid Ion Drive possibilities

Hi PhotonBytes, good to see you back. I'm no rocket scientist, but your ion drive post looks interesting. I'll look it over. I just wanted to say hello and welcome you back to the forum.

kbd512

Administrator

Registered: 2015-01-02

Posts: 7,441

Re: Dual Stage 4 Grid Ion Drive possibilities

If all 20,000kg of the spacecraft wet mass is devoted to the reactor, then the reactor is producing 125kW/kg of weight.  That would be pretty remarkable, and has never been demonstrated for a reactor operating for weeks to months at a time.  The nuclear thermal rocket reactors can generate that much power for a period of minutes to hours, but the materials are subjected to extreme heat and radiation damage.  They're pretty well trashed (fuel rods are cracked) after a literal handful of hours of operation.

At 1AU, we now have thin film solar panels, that have actually flown in space (demonstrated flight heritage aboard more than 1 robotic mission, either Earth orbital or into deep space, headed to Jupiter, in the case of the Japanese mission), that can generate 1.25kW/kg of weight.

There is a NASA nuclear electric propulsion space reactor concept with considerable ground testing that demonstrated 2.5kW/kg, which Dr Zubrin previously stated was impossible.  NASA's 1.125GW NERVA XE Prime nuclear thermal rocket engines weighed in at 18,144kg, meaning 62kW/kg.  XE Prime was a flight-ready prototype with extensive ground testing and advanced ceramic materials, which have been replicated and improved for BWXT's new / upcoming nuclear thermal rocket engine being built for NASA.  Will BWXT achieve 125kW/kg?  I don't know.  That's very ambitious, but perhaps doable.

The wet mass of the entire ship will greatly exceed 20t, even if it's possible to achieve 125kW/kg from the power source.  NASA's direct-electric reactor concept was only 2.5kW/kg, as previously stated, and I think it was 70%+ efficient (don't quote me on that, it's been a long time since I read about it).  Thermal-to-electric efficiency will fall somewhere between 50% and 65% efficiency at the temperature of NERVA, which means if you need 2.5GW of input power for the ion thrusters, then you need a lot more thermal power from your reactor.

3 weeks of thrusting is 504hrs.  No reactor has ever operated at the power density you're suggesting, for that length of time.  I would be impressed beyond words if the fuel elements lasted for 24hrs.  XE Prime's nominal operating temperature was 2,230°C.  Titanium melts at 1,866°C.  I would be very careful about assuming that doubling the power output level per unit of reactor mass is doable over that length of time.  Only refractory heavy metals like Tungsten or ultra-high temperature ceramics won't melt at XE Prime's operating temperatures, never mind something with double the power density.  The fuel rod bundles are literally glowing white hot to achieve the suggested power density.  On top of that blow-torch heat, the reactor materials are subjected to a maelstrom of radiation damage.

You will have to use HEU, as you suggested, to prevent the fuel rods from cracking even faster.

I think your weight estimates for the reactor alone are highly optimistic.  The reactor core itself might only weigh 20t, ignoring the weight of the radiators, piping, shielding, coolant, and support structure.  You need to add propellant, structure, and payload on top of that.

113g/s of propellant flow equates to 205,027.2kg of Xenon over 3 weeks of thrusting, as well as the GDP of a small country, because Xenon is very expensive.  21,400s of Isp / specific impulse is achievable using Hydrogen as the propellant for these ion thrusters, but not Xenon unless the thrust is achieving plasma temperatures hotter than the surface of the Sun, with the added bonus of being almost free compared to the cost of 1kg of Xenon.  I would find 2,500s to 3,000s plausible using Xenon and 20,000s+ plausible using Hydrogen.

That means your reactor core alone, plus propellant mass, weighs 225t.  Double the reactor's weight for the radiator mass, so another 40t.  It won't need to be very large at such high temperatures, but must be made from a heavy refractory metal like Tungsten, and it's going to glow.  Modern electric aircraft motor-generators can produce 10kW to as much as 20kW/kg (this is far beyond the gravimetric power density achievable by any modern after-burning turbofan engine, but still well below high performance chemical rocket engine territory).  That means your electric generator will weigh 125,000kg.

I think 500,000kg / 500t is a more plausibly realistic reactor + shadow radiation shielding + radiator array + electric generator + engine + propellant + structural mass value than 20t.  To that, you must then add the mass of your payload, which is presumably crew habitation modules.  You still have one hell of an engine for travel to other planets, assuming the reactor's core life is about 2 months, which gives you enough energy to push a 500t to 750t payload to Mars in about the same 6 months as a "normal" chemical rocket engine powered mission, using a minor fraction of the propellant mass that would otherwise be required.  I would still categorize that as a super rapid transit system capable of powering a real ship to Mars and beyond- only 16.4% propellant mass, which is truly spectacular performance, even if you don't get there in 3 weeks.

kbd512

Administrator

Registered: 2015-01-02

Posts: 7,441

Re: Dual Stage 4 Grid Ion Drive possibilities

I forgot to include this in my previous post, but to run a reactor coolant at the temperatures that XE Prime operated at, you'll either need a metal like Cerium, which melts at "only" 795C, but boils at 3,443C, or else you need to run a gas coolant, probably something like Helium or Argon.  I'm thinking that the ability to improve power density and thus reactor core mass is pretty well at its limit with XE Prime.  Project Timberwind came up with much lighter, albeit smaller core designs with reduced engine weights, but those were all paper designs with zero actual testing behind them, unlike XE Prime (from Project Rover / NERVA) and BWXT's reactor design.

Back in the day, the "nuclear lightbulb" concept saw non-nuclear ground testing by United Aircraft Corp, I think.  That reactor was to operate at 4,600MWt and weighed in at 37,250kg, or 123,490W/kg.  The reactor's radiator provided 118MWt of cooling for structural parts of the reactor, and was projected to weigh 5,500kg.  That means 1kg of radiator mass rejected 21,455W of waste heat.  65% thermal-to-electric efficiency provides 2,990MWe, which meets your electric power requirements.  That means 1,610MW of waste heat must be rejected to space, so the radiator will weigh 75,041kg.  Could you reduce that 75t using a recompression cycle to achieve greater thermal-to-electric efficiency?  I'm sure you could, but then you add more piping and heat exchangers that are also very heavy.  That may work quite well for a land-based reactor with no mass limitations, but not so much for a spacecraft.  65% is completely doable, though.

32,750kg + 75,041kg + 149,500kg = 257,291kg
2,990,000,000We / 257,291kg (excludes propellant tank mass) = 11,621We/kg

Read through this slide deck from Sandia National Laboratory to see the materials required to operate Brayton cycle sCO2 gas turbines and heat exchangers at 700C to 800C (this is all closed-loop stuff, just like your proposed 2.5GW reactor):
Supercritical CO2 Brayton Cycles

If you read the document, lead times for the metals and machining involved are measured in years, and this was back in 2014 when the global economy was more or less fully functional.  Inconel is serious money.  Tungsten, Rhenium, and/or Osmium are on another level.  I'm wondering what the seal materials will be made from and how they'll be designed to prevent loss of coolant gas or metal, yet neither melt nor expand and contract so much over the massive temperature ranges involved, that the coolant loop doesn't leak like a sieve.  You basically need Tungsten because any lesser metal will melt or creep and then rupture.

Multi-MW Closed Cycle MHD Nuclear Space Power Via Nonequilibrium He/Xe Working Plasma

Note how they were talking about getting the vehicle's Alpha down to 1kg/1kWe, for a complete multi-MW nuclear electric power and propulsion system.  A nuclear lightbulb is about 11X better than MHD using whatever tech and at whatever scale the NASA project from the linked NTRS document above was directed towards.

Note how a nuclear lightbulb that provides the input power would weigh 257t.  No mass has been allocated for the actual ion thrusters, propellant tanks, or the engines themselves.  After you add the mass of all that stuff to the propellant mass (your total vehicle wet mass, minus payload), we're going to be at or a bit above 500t of vehicle weight.  All I know about the ion engines and electrical power processing subsystems is that they're not light.

How about a grid of 300 of the 10MW MPD thrusters that NASA and universities have played with?

The NASA prototypes were like 50kg each, so 15,000kg for all of them.  Even if they're more like 100kg each, that's only 30t.  Propellant tank mass is typically 5% of the propellant load, so 216t of the propellant and tanks.

257t + 216t + 30t = 503t

Our vehicle Alpha has now dropped to 5.5kW/kg, but we're still much better than MHD.

We still have no mass allocated for electrical wiring, power conditioning, distribution, nor the "turbo" part of the turboelectric generator.

We'll stipulate the use of CNT wiring to save weight over Copper.  Maybe we could use superconducting wiring, but that probably won't save any weight over CNT, and it will require electrical power to cool, but we have 440MWe to play with, so maybe...

Anyway, I hope this gives you some ideas to play with.