One Month to Mars -- Methods for Very Fast Settler Transit

[chopsticks]

I have no idea what's going on here.

[Phil Stooke]

Wake me up when you've got it working.

[LMT]

Wake me up when you've got it working.

Hi, Dr. Phil.  Have you walked through Winterberg's "3F" design?  I'm curious to see how it might be adapted for OP purpose.  1 2

Or alternate, if better.

[LMT]

Boca Chica Fusion

Applying Winterberg 2015, a minimized "3F" system from Winterberg 2004 has some workable fusion parameters:

1.  Each shot's energy release is only ~ 20% of Winterberg 2015.

2.  A thin parabolic steel reflector / magnetic mirror with ~ 1.5 m diameter bears a blast pressure of ~ 100 MPa.  (cf. vacuum Raptor bell diameter 2.4 m and Raptor v3 chamber pressure 35 MPa.)

3.  A reflector's surrounding disperse-medium structural disk with ~ 4.6 m diameter has radial (hoop) stress of ~ 14 MPa, requiring a reinforcing hoop ~ 3 cm thick. *

It's starting to look like something SpaceX could prototype at Boca Chica.

Image:  Starship SN1 rings.  @bocachicagal for NSF.

* It seems Winterberg 2015 p. 17 has a mistake (red); hoops are thinner than indicated.

[InterestedEngineer]

One-month transit would remove the need for artificial gravity, cut consumable payload, and relieve psychological stress.  A practical consequence is increased passenger capacity.  If one-month transit raised a ship's passenger count by 50%, the number of crewed ships required for settlement would be cut by 1/3.  The corresponding fleet savings would justify considerable investment in fast transit infrastructure.

From a software engineer's perspective, you are trading an O(N²) problem for an O(N) problem.  Not a good tradeoff except for very small N where some other variable may become more important.

N in this case is velocity, and the energy you need to expend (or expel).

Making more spacecraft is O(N) (if you solve the mfg line problem, which SpaceX is doing right now).

Making spacecraft faster is O(N²)

Solving the problem with O(N) energy is going to be far easier than solving it in O(N²) energy

So, at a fundamental level, trying to solve the problem of transit to Mars like this is not a good plan.

[LMT]

One-month transit would remove the need for artificial gravity, cut consumable payload, and relieve psychological stress.  A practical consequence is increased passenger capacity.  If one-month transit raised a ship's passenger count by 50%, the number of crewed ships required for settlement would be cut by 1/3.  The corresponding fleet savings would justify considerable investment in fast transit infrastructure.

From a software engineer's perspective, you are trading an O(N²) problem for an O(N) problem.  Not a good tradeoff except for very small N where some other variable may become more important.

N in this case is velocity, and the energy you need to expend (or expel).

Making more spacecraft is O(N) (if you solve the mfg line problem, which SpaceX is doing right now).

Making spacecraft faster is O(N²)

Solving the problem with O(N) energy is going to be far easier than solving it in O(N²) energy

So, at a fundamental level, trying to solve the problem of transit to Mars like this is not a good plan.

Your notion falls apart when you try to apply it here, with numbers.

[Coastal Ron]

...
It's starting to look like something SpaceX could prototype at Boca Chica.

SpaceX consumes a lot of stainless steel for Starship production, but that doesn't mean they are the only organization that can form and weld stainless steel.

As for your proposal, and just like all the other proposals of yours that you think SpaceX should be working on, if SpaceX thought they were a good idea they would have already been working on them, meaning... 

[InterestedEngineer]

One-month transit would remove the need for artificial gravity, cut consumable payload, and relieve psychological stress.  A practical consequence is increased passenger capacity.  If one-month transit raised a ship's passenger count by 50%, the number of crewed ships required for settlement would be cut by 1/3.  The corresponding fleet savings would justify considerable investment in fast transit infrastructure.

From a software engineer's perspective, you are trading an O(N²) problem for an O(N) problem.  Not a good tradeoff except for very small N where some other variable may become more important.

N in this case is velocity, and the energy you need to expend (or expel).

Making more spacecraft is O(N) (if you solve the mfg line problem, which SpaceX is doing right now).

Making spacecraft faster is O(N²)

Solving the problem with O(N) energy is going to be far easier than solving it in O(N²) energy

So, at a fundamental level, trying to solve the problem of transit to Mars like this is not a good plan.

Your notion falls apart when you try to apply it here, with numbers.

There are plenty of numbers in the thread, but current projected deltaV from LEO is about 4km/sec and all the other deltaV to LMO is aerobrake so not counted towards rockets.

Where 1 month requires 25km/sec acceleration and then another 25km/sec braking, or 50km/sec deltaV.

That's 150 times the amount of energy expenditure, for what, maybe 2 times the payload because of a faster transit?

So there it is, applied with numbers.

[LMT]

...
It's starting to look like something SpaceX could prototype at Boca Chica.

SpaceX consumes a lot so stainless steel for Starship production, but that doesn't mean they are the only organization that can form and weld stainless steel.

As for your proposal, and just like all the other proposals of yours that you think SpaceX should be working on, if SpaceX thought they were a good idea they would have already been working on them, meaning... 

-

The Helicity Space plot of specific thrust power vs. Isp is updated with Winterberg's "3F" system (red), using MOX numbers posted above. 

Several joking posters just couldn't stick that pin, and don't understand, so I'll note the basic reason for this extraordinary outlier:  DT fusion inherently outperforms every other useful form of fusion, as seen in the plot of reactivity vs. temperature.  People tell emotional stories, with animations and pitch decks even, but the fact remains.

Also shown:  a characteristic product -- specific thrust power times Isp -- gives one measure of an engine's fit to OP purpose.  Higher is better.  MOX is convenient, not optimal.  Values are averaged where a range is projected.

[LMT]

One-month transit would remove the need for artificial gravity, cut consumable payload, and relieve psychological stress.  A practical consequence is increased passenger capacity.  If one-month transit raised a ship's passenger count by 50%, the number of crewed ships required for settlement would be cut by 1/3.  The corresponding fleet savings would justify considerable investment in fast transit infrastructure.

From a software engineer's perspective, you are trading an O(N²) problem for an O(N) problem.  Not a good tradeoff except for very small N where some other variable may become more important.

N in this case is velocity, and the energy you need to expend (or expel).

Making more spacecraft is O(N) (if you solve the mfg line problem, which SpaceX is doing right now).

Making spacecraft faster is O(N²)

Solving the problem with O(N) energy is going to be far easier than solving it in O(N²) energy

So, at a fundamental level, trying to solve the problem of transit to Mars like this is not a good plan.

Your notion falls apart when you try to apply it here, with numbers.

There are plenty of numbers in the thread, but current projected deltaV from LEO is about 4km/sec and all the other deltaV to LMO is aerobrake so not counted towards rockets.

Where 1 month requires 25km/sec acceleration and then another 25km/sec braking, or 50km/sec deltaV.

That's 150 times the amount of energy expenditure, for what, maybe 2 times the payload because of a faster transit?

So there it is, applied with numbers.

Who forgot that construction and operation of a vast interplanetary settlement Starship fleet... will cost a bit?

[Coastal Ron]

...
It's starting to look like something SpaceX could prototype at Boca Chica.

SpaceX consumes a lot so stainless steel for Starship production, but that doesn't mean they are the only organization that can form and weld stainless steel.

As for your proposal, and just like all the other proposals of yours that you think SpaceX should be working on, if SpaceX thought they were a good idea they would have already been working on them, meaning... 

The Helicity Space plot of specific thrust power vs. Isp is updated with Winterberg's "3F" system (red), using MOX numbers posted above...

Why are quoting MY post when you don't even address anything I wrote. It's almost like you created this public thread for yourself, and not to really discuss or debate the topic... 

[InterestedEngineer]

So if I understand their specific thrust power correctly, I get 4.2GW of thrust at an Isp of 59,000 for a 300t engine.  Is that correct?  This is for the best setup DHe3.

(300000*50000/3600) ~= 4.2e6 W

So solving for mass flow, 4.2GJ/sec = 1/2mdot * ( 579,000 m/sec)², I get a mass flow rate of .025kg/sec

Force = .025kg/sec * 579,000m/sec = 14.5kN.

Now suppose the craft masses 300t and the payload masses 200t, or 500t.

50km/sec of deltaV (what a 1 month transfer requires) is a mass ratio of 1.09, so 45-50t of fuel.

So our average mass is 525t at 14.5kN, so our acceleration is .028m/s² or 2.4km/sec per day.  Since it'll take weeks to get to velocity I don't think impulsive burn simplifications are going to work here.  Still, we are at the right order of magnitude.

If we fix burn time at 30 days, with the mass flow of .025kg/sec we end up with a fuel of 65t, and a max velocity of 35km/sec and an average velocity of 18km/sec.  It requires 25km/sec, so not fast enough.

I can only conclude that the best Helicity drive proposal, the one that requires 10s of tons of extremely rare He3, won't get to Mars in 30 days or less.   But it'll probably do 40 days.

the sad part is I don't see an economical way for Helicity to get to their final configuration, assuming the He3 cost problem can be solved.   All the in-between tech has a far lower ROI and ongoing cost compared to just use layered refueling of Starships.  There are no viable missions that would pay of the multi-billion dollar development effort.

Contrast this to how Elon is funding Starship.  Bootstrapping cash flow from reusable rockets to meet a multi-billion dollar/year need (high speed rural internet), plus NASA funding, to get his Mars rocket.

I don't see a pathway for Helicity to do that.  Starship will have crashed the costs for doing anything in the Solar System to where the dev costs for anything else won't have a reasonable ROI.

At the same time, I am rooting for them, it's the best fusion drive proposal I've seen in decades.   Magnetic Peristalsis, if it can be made to work, is a great idea.

[LMT]

Magnetic Peristalsis, if it can be made to work, is a great idea.

Helicity tech wouldn't give an economical settler tug, but 3F / MOX would.  "Great idea", yes?  Try ballparking 3F / MOX numbers:  e.g., tug propellant mass for one month to Mars.  Include tug turnaround.

[InterestedEngineer]

Magnetic Peristalsis, if it can be made to work, is a great idea.

Helicity tech wouldn't give an economical settler tug, but 3F / MOX would.  "Great idea", yes?  Try ballparking 3F / MOX numbers:  e.g., tug propellant mass for one month to Mars.  Include tug turnaround.

It's your proposal, you should benchmark it

[LMT]

Magnetic Peristalsis, if it can be made to work, is a great idea.

Helicity tech wouldn't give an economical settler tug, but 3F / MOX would.  "Great idea", yes?  Try ballparking 3F / MOX numbers:  e.g., tug propellant mass for one month to Mars.  Include tug turnaround.

It's your proposal, you should benchmark it

Incorporating another efficiency:

You could leave Starship main tanks dry in LEO and save $ billions more per synod.  Just burn straight from LEO on 3F tug:  i.e., deliver all 29 km/s, for Mars in a month, on fusion.

With same-day turnaround, a notional 50-ton tug sends a 400-ton Starship on its way with 200 t of methalox or a bit more, depending on Starship header-tank top-off requirement.

Is it a "great idea" yet?

[LMT]

Shock Absorber

A pulsed fusion drive concept typically applies shock absorbers.  With 3F tug microexplosions, a superconducting regenerative shock absorber system might be best.  What are some mass-efficient designs for such shock absorbers?

Image:  Project Orion, Matthew Paul Cushman.

[Vultur]

It's not about understanding of the rocket equation (and it's rather insulting to say so). It's entirely about how long it actually takes major new technologies to be developed & put into practice on a large scale.

It's also pretty well known that D-T fusion is the best, disregarding the problem that tritium has to be manufactured & is radioactive & highly regulated (I don't think D-T is aneutronic either). But how many people really believe that fusion propulsion will be well enough tested to commit lots of people to it in 15 years? I'd be very impressed if it had even propelled a test satellite by then.

It's not a rocket equation problem: there are technologies less exotic than fusion that can give the required Isp. Known electric propulsion can do it.

The problem is that getting the required thrust with that Isp means either:
- deploying utterly enormous, super-thin-film solar panels in space (solar-electric);
- developing a space reactor with specific power (W/kg) orders of magnitude better than the demonstrated ones (nuclear-electric)
- violating international treaties (Orion style nuclear pulse);
- doing crazy stuff with nuclear materials that probably no one will let you do (gas-core nuclear thermal or nuclear-salt-water rocket);
- relying on fusion being successfully developed (any non-Project Orion fusion drive)
- building massive in-space infrastructure (various beamed possibilities)

If you gave me infinite budget and a mandate to do it in 15 years, I'd go for the first one; it seems the less crazy risk (though I'd give it a very low chance of working on schedule - just less low than all the other possibilities; but at least less-than-complete success would still be useful, unlike some of the other possibilities, and I wouldn't irradiate anything with development accidents).

With solar-electric, unlike anything nuclear, you could do a rapid development cycle since you can accept failures.

So given that mandate, here's what I would do.

I'd go talk to the VASIMR people, the people workint on Gateway PPE, and dig into the old HiPEP research and see who could give me the best confidence about scaling up to much higher thrust/power levels (hundreds of megawatts).

I'd go to the thin film solar cell companies and researchers and see just how thin I could get cells that could still be reasonably long-lasting and manufactured in very large quantities. This article ( https://spectrum.ieee.org/thin-film-solar-panels ) claims solar cells can be made 2 to 3 micrometers thick. Can I get tons of reliable cells at that thickness, and what's their efficiency?

And (the hardest part likely) work on expanding the size of soft structures we can unfold in space. Get everyone available who's worked on solar sails involved: NASA, Planetary Society, JAXA, whoever else. Start building test vehicles and flying them; pay the extra to bump back whatever payloads SpaceX will let me bump back, so we can get results as fast as possible and use them to design the next test vehicle.

I think very-thin-film solar on a solar sail type deployment (similar to what JAXA's IKAROS did, but vastly larger) would have far better W/kg in the inner solar system than any space nuclear reactor that could be built in the next few decades.

[Density of silicon is about 2.5 grams per cubic centimeter: that's 5 grams per square meter if it's 2 micrometers thick.
Solar constant is about 1360 W/m^2 at Earth, more like 590 at 1.52 AU. Say 950 average over the whole trip ...
Assume 10% efficiency, very poor for solar cells but we're pushing for the thinnest possible.
So 95 W/m^2.
That's 19 W/g or 19 kW/kg - but that's just the cells. If we have to triple that mass because of power cables and deployment booms ... that's stillover 6 kW/kg. No way you're getting that out of nuclear reactors in space before 2050.

At this rate, a 50-ton power module (which could be launched by one Starship, leaving mass budget for the actual engines) would give 300 MW. Of course, it would also unfold to over 3 square kilometers...

[whitelancer64]

This article ( https://spectrum.ieee.org/thin-film-solar-panels ) claims solar cells can be made 2 to 3 micrometers thick. Can I get tons of reliable cells at that thickness, and what's their efficiency?

For the record, this paper I just looked up says their 3 micrometer thick solar cell tested at 12.3% efficient.

https://www.sciencedirect.com/science/article/pii/S2211285520300227

[LMT]

It's not about understanding of the rocket equation (and it's rather insulting to say so)....

...getting the required thrust with that Isp means either:
- deploying utterly enormous, super-thin-film solar panels in space...

Matching fusion Isp at 13,000 s with electric engine, you'd get ~ 10 mN/kW.  That fails OP requirement, obviously. 

You're making a story without numbers there.  You might read the Electric Thruster thread for relevant basics, e.g., the fundamental performance curve.

[Vultur]

It's not about understanding of the rocket equation (and it's rather insulting to say so)....

...getting the required thrust with that Isp means either:
- deploying utterly enormous, super-thin-film solar panels in space...

Matching fusion Isp at 13,000 s with electric engine you'd get ~ 10 mN/kW.  That fails OP requirement, obviously. 

You're making a story without numbers there.  You might read the Electric Thruster thread for relevant basics, e.g., the fundamental performance curve.

a) there are lots of numbers in that post. Please don't be so condescending. But I admit I didn't go through the full math.

b) You don't need 13k Isp for Mars; in fact, it's rather too high to be efficient. According to your own numbers:

fwiw, an optimal 29 km/s impulsive departure burn gives Mars arrival in 30 days with Vinf magnitude (entry speed) of ~ 32 km/s.

29 km/s + 32 km/s = Total mission delta-v 63 km/s. If you can use aerobraking/entry for say 8 km/s of the deceleration, that becomes more like 55 km/s.

If we accept a mass ratio of 6, we need an effective exhaust velocity of 30.7 km/s, or Isp 3130.

If we need a mass ratio of 3 (far less than an interplanetary Starship would have) we instead need ~50.6 km/s or Isp ~5160.

That's way more achievable.

And some of the VASIMR articles were talking about 39 days to Mars, so using electric propulsion for fast transits isn't just me being crazy. The power source is challenging, but I think very thin film solar is less development risk than fusion drive. Also, if you don't quite make it, 50-60 days to Mars is still useful, but if the fusion drive doesn't work you have nothing.