Author Topic: FRC + Imploding Plasma Liner Fusion for the Fusion Rocket (NIAC2)  (Read 23696 times)

Offline alexterrell

  • Full Member
  • ****
  • Posts: 1283
  • Liked: 0

To put 134 metric tonnes in perspective that's 2 SLS flights, 6 Delta IV Heavy flights or 9 Skylon flights, or 5 Atlas V heavy flights (if it ever gets funded).

If the vehicle can be fully reused (which it seems it can) this is not ridiculously expensive (but it's not cheap).  :(


Personally, I think speed is overrated.

I like to see this engine take 1,000 tons from LEO to Mars on a Hohmann trajectory rather than 70 tons in 50 days or so.

Maybe take the fast route back when it doesn't have all the landing modules and cargos etc.

Offline Elmar Moelzer

  • Full Member
  • ****
  • Posts: 1677
  • Liked: 91
Personally, I think speed is overrated.
I disagree. Speed is less relevant for cargo and unmanned science missions, but for manned missions, speed is the key to everything. More speed means less consumables needed on the flight. It means less exposure to radiation. It means less exposure to a low gravity environment and it means less risk overall. It means that you need a smaller habitat. More speed would also allow for more mission flexibility in regards to timing.
Just to repeat that again. I am talking about manned missions.

Offline IRobot

  • Full Member
  • ****
  • Posts: 754
  • Liked: 29
  • Portugal & Germany
One question: if we wanted to double the average thrust, what would be the path?

Scale up the engine? Shorten time between pulses?

Putting two engines in tandem does not seem to be viable.

I don't see why you couldn't put 2 engines in tandem. Four would be better, then fire them 2 at a time to prevent bending moments.

Other than that, the lithium ring is the smallest they could make it, so it could be expanded. But I think increasing the firing rate would be easiest, and it doesn't need any increase in capacitor banks.
Well, for tandem pulsing engines you have two problems. First either they are very well synchronized or they need their own shock absorver. Then if one of them is not working, you need to gymbal the other.

Offline john smith 19

  • Senior Member
  • *****
  • Posts: 2156
  • Liked: 41
  • Moderation policy Linton Barwick
  • Everyplaceelse
A lot of these problems can be worked out in sounding rocket or zero g plane flights long before a full vehicle is designed.
Hmmm, I see the problem being the opposite: you cannot do representative testing of the reload mechanism on a small scale (sounding rocket), and/or small time-scale (aircraft). This is a large complex mechanism that includes multi-meter transport of components, cold-welding them with very fine-alignment, in vacuum zero g. Just two full cycles will take ~28 seconds.
Note that cycle is the maximum time it takes. it could be run experimentally somewhat faster.

Quote
Not to say it can't be done, but it was Elmar who suggested upthread we think of this mechanism as akin to placing a robot assembly line (i.e. well understood and highly-reliable) in orbit. I think that analogy is a good one.
If the you're using cold welding that's a technique from the beverage can making industry. That uses fixed automation to mfg 1000 000 of cans a day.

Likewise the launch system can be viewed (loosely  :) ) as a big relay or door latch. Such devices have operating lives measured in decades.

Yes this system is much bigger than a door latch or a beverage can but it's also much slower to mfg and needing a much shorter lifespan (years not decades) than a door opener.
Quote
To get confidence in (man-rate) a mechanism that is required to run for thousands of cycles to keep your crew safe, I think you need to launch a full-scale test article and run it for a representative period. While the mechanism is hoped to be just a few tonnes, a long-run test will require a lot of liner components, and that's going to be heavy.
Full size maybe not. Full duration would be a good idea.
Quote
As for reliability and recovering from a jam, given the number of failure points and the likelihood of in-flight repair being minimal, I think you just include two or three FDRs on the spacecraft to get full redundancy.
I've never understood the fondness for building interplanetary spacecraft (and/or) their engines be made in single units.  :(
"Solids are a branch of fireworks, not rocketry. :-) :-) ", Henry Spencer 1/28/11
Averse to bold? You must be in marketing.

Offline alexterrell

  • Full Member
  • ****
  • Posts: 1283
  • Liked: 0
Personally, I think speed is overrated.
I disagree. Speed is less relevant for cargo and unmanned science missions, but for manned missions, speed is the key to everything. More speed means less consumables needed on the flight. It means less exposure to radiation. It means less exposure to a low gravity environment and it means less risk overall. It means that you need a smaller habitat. More speed would also allow for more mission flexibility in regards to timing.
Just to repeat that again. I am talking about manned missions.
I agree on the advantages, but consider, for about 100 tons of Lithium fuel:

Fast trajectory return, delta V 40,000m/s, inert mass = 80 tons
Hohmann trajectory return, delta V 4000m/s, inert mass = 1,100 tons

Now, 1,100 tons is ambitious for early trips to Mars, but for Mars colonisation, that could be really useful. With that mass, consumables are trivial, you can use a torus to provide gravity, and you can have a shielded area in the middle for solar flares.



Offline Elmar Moelzer

  • Full Member
  • ****
  • Posts: 1677
  • Liked: 91
Personally, I think speed is overrated.
I disagree. Speed is less relevant for cargo and unmanned science missions, but for manned missions, speed is the key to everything. More speed means less consumables needed on the flight. It means less exposure to radiation. It means less exposure to a low gravity environment and it means less risk overall. It means that you need a smaller habitat. More speed would also allow for more mission flexibility in regards to timing.
Just to repeat that again. I am talking about manned missions.
I agree on the advantages, but consider, for about 100 tons of Lithium fuel:

Fast trajectory return, delta V 40,000m/s, inert mass = 80 tons
Hohmann trajectory return, delta V 4000m/s, inert mass = 1,100 tons

Now, 1,100 tons is ambitious for early trips to Mars, but for Mars colonisation, that could be really useful. With that mass, consumables are trivial, you can use a torus to provide gravity, and you can have a shielded area in the middle for solar flares.
Besides the logistical problems of long term missions with large space craft, the biggest problem with this is getting all that mass into orbit. Personally, I consider that the bigger challenge of BEO exploration program right now. I hope that SpaceX can contribute to that side of things.

Offline john smith 19

  • Senior Member
  • *****
  • Posts: 2156
  • Liked: 41
  • Moderation policy Linton Barwick
  • Everyplaceelse
To put 134 metric tonnes in perspective that's 2 SLS flights, 6 Delta IV Heavy flights or 9 Skylon flights, or 5 Atlas V heavy flights (if it ever gets funded).
If the vehicle can be fully reused (which it seems it can) this is not ridiculously expensive (but it's not cheap).  :(
Or 3 Falcon Heavy flights, which would even leave some extra margin. For a reusable mars mission architecture that allows for 90 day transfers and if all goes well, even 30 day transfers, that is cheap, at least in my book.
Yes I forgot about this one.  :) It's a fair option, and probably as likely (if not more more so) to go to development than the Atlas V heavy.

Of course if some kind of Raptor FH became viable..   :) That's (potentially) in SLS territory...

That said I do believe a lot of the sub system can be tested as secondary payloads on various LV's.

Short term I'd guess the top (near term options) are a Delta IV Heavy (24 tonnes) and the FH (53 tonnes?).

A single launch to put a whole experimental vehicle into orbit? IDK. Tough definitely
"Solids are a branch of fireworks, not rocketry. :-) :-) ", Henry Spencer 1/28/11
Averse to bold? You must be in marketing.

Offline Elmar Moelzer

  • Full Member
  • ****
  • Posts: 1677
  • Liked: 91

That said I do believe a lot of the sub system can be tested as secondary payloads on various LV's.

I think they are planning that, actually. Sending mission using an untested engine on a trip to mars does not seem to make much sense ;)

Offline john smith 19

  • Senior Member
  • *****
  • Posts: 2156
  • Liked: 41
  • Moderation policy Linton Barwick
  • Everyplaceelse

That said I do believe a lot of the sub system can be tested as secondary payloads on various LV's.

I think they are planning that, actually. Sending mission using an untested engine on a trip to mars does not seem to make much sense ;)
Indeed. There is a very wide spectrum of testing between what's currently running in the lab and a full Mars vehicle(s).  :)  I think a lot of these systems can have their TRL's raised (at relatively low cost) before anyone has to start looking fundraising for a multi $Bn payload.
"Solids are a branch of fireworks, not rocketry. :-) :-) ", Henry Spencer 1/28/11
Averse to bold? You must be in marketing.

Offline a_langwich

  • Full Member
  • ***
  • Posts: 391
  • Liked: 36
Personally, I think speed is overrated.
I disagree. Speed is less relevant for cargo and unmanned science missions, but for manned missions, speed is the key to everything. More speed means less consumables needed on the flight. It means less exposure to radiation. It means less exposure to a low gravity environment and it means less risk overall. It means that you need a smaller habitat. More speed would also allow for more mission flexibility in regards to timing.
Just to repeat that again. I am talking about manned missions.

, and you can have a shielded area in the middle for solar flares.


You are ignoring GCR radiation risk then.

Offline Shevek23

  • Member
  • Posts: 32
  • Liked: 0
One question: if we wanted to double the average thrust, what would be the path?

Scale up the engine? Shorten time between pulses?

Putting two engines in tandem does not seem to be viable.

Speeding up the pulse rate is not "easy," apparently not at all--if we are talking about fabricating a new half-kg lithium ring in the engine for every pulse, quite obviously not. But it's the approach that has the biggest payoff by far. Earlier press releases and published papers mention an ambition to get the rate up to once every 10 seconds (same ballpark as every 14) and I think they may have mentioned the possibility of once every second! If they did they may be regretting it now; OTOH if they can get a working system operational in space that pulses say once every 20 seconds, interest and funding to strive to improve the rate should be quite forthcoming, I'd think. Who knows what the limit would be then?

The faster the pulse rate, the greater the thrust to weight ratio. Elmar Moelzer mentions the need for the exhaust products to completely clear the system as a limit, but with the main pulse exiting at 50,000 meters/sec it's hard for me to believe this can be a factor after say 1/100 of a second, even granting there will be a bell curve or something like that of particle speeds, so some small fraction will indeed be lingering about the chamber long after most of it has departed.

More seriously, while I've found presentations where the developers predict that the overall mass of a rocket capable of the performance they are talking about (50 km sec exhaust, about half a kg per pulse, pulses in the range of 1 every 10 seconds) would be about 15 metric tons, I have no idea just how much of that is the ring-forming mechanism, or how the mass of that would scale with increasing pulse frequency (assuming the result can be accomplished at all). That it might wind up massing as the square of the frequency does not seem unreasonable to me; then the question is, how much would the baseline component capable of once every 10 seconds be? 100 kg? A tonne? That governs the point at which scaling up the pulse rate dominates the rising mass of the engine.

However I see no reason why any other component of the system (except the FRC injector, obviously) should rise in mass with it. Each pulse is the same; a magnetic nozzle system that can handle a pulse every minute should be able to handle one every second just as well...

Well, let me qualify--two things; one, the system proposed uses solar power input to drive the compression stage of the pulse--while I understand why they are going for that I've always felt the need to stress there's no fundamental reason the compression power has to come from outside the cycle! It could easily be drawn from the exhaust phase of the cycle, as a drain on the magnetic nozzle. With gains in power due to fusion in the range of 100 or more, it ought to be pretty doable. With power for the compression phase coming from outside the system, obviously that power source has to be scaled up in proportion to the pulse rate, and that's bad. (Unless we have onboard also the power generation system the same inventors are also working on, perhaps? :)) But if the compression power is drawn from the previous exhaust pulse, the shorter the time scale the better, because that's power that has to be stored. (The same is true with solar power or other external power input--it has to be accumulated in capacitors; these same capacitors could simply store the full energy needed for the next pulse from the previous one, so either way the power storage system is required--and included in the 15 tonne estimate already).

Two, obviously real world systems are never perfectly efficient; the magnetic systems will produce some waste heat, and the more cycles we get per minute, the greater the waste power is; that forms another practical limit and system that scales up in mass with the increased pulse rate.

All of this is obviously voting to bell the cat; to consider the tradeoffs involved in pushing the pulse rate up we have to first be able achieve the upper end of the range considered at all! It all depends on the system used to form and position the rings, how reliable it is, and so forth.

But until the scaling up of the formation mechanism, the FRC injector rate, and the heat rejection system starts to become a very significant fraction of the total engine mass, clearly the faster the pulse rate the better.
----
All that said, what if you can't increase the pulse rate beyond a certain point? Then I'd say that it would be better to keep the minimum scale of the engine system, and build many of them to increase the thrust, rather than to make the engine bigger in proportion to the mass of each pulse. It seems reasonable to me that it could be scaled up--that 5 kg of lithium can fuse 10 times the mass of fusible isotopes--but while there is little reason to think any component of the engine would scale up at a higher rate than linear, there is no reason to think any significant part of the mass would scale up at less than linear rate either. So we'd need a magnetic nozzle, including the compression component, 10 times the mass of the one proposed, and 10 times the mass of power storage capacitors, and so on. If compression is powered by solar power we need 10 times more panel area. Etc.

So instead, we should just install 10 engines, and pulse them in sequence to average out the thrust as much as possible. We might get some good synergy out of it--say, we only need one capacitor bank system and we operate it at 10 times the frequency.

On the scale of thrust/spacecraft ratios we are talking about, I don't think the asymmetry of thrust from ten clustered nozzles is that big a problem; it averages out obviously. And if one engine fails, we obviously can just modify the firing cycle of the rest to get degraded thrust (but not ISP--a given amount of propellent will last 10 percent longer to deliver the same impulse).

So as long as we are wishing for the next generation beyond the 2020s engine proposed, I'd say first of all wish for increased pulse rate, then think in terms of clustering standard sized units as needed.

Offline Shevek23

  • Member
  • Posts: 32
  • Liked: 0

To put 134 metric tonnes in perspective that's 2 SLS flights, 6 Delta IV Heavy flights or 9 Skylon flights, or 5 Atlas V heavy flights (if it ever gets funded).

If the vehicle can be fully reused (which it seems it can) this is not ridiculously expensive (but it's not cheap).  :(


Personally, I think speed is overrated.

I like to see this engine take 1,000 tons from LEO to Mars on a Hohmann trajectory rather than 70 tons in 50 days or so.

Maybe take the fast route back when it doesn't have all the landing modules and cargos etc.

Well, there are times and places for both!

I'd think a good test run of the engines, once designed for space and orbited, would be as a cargo hauler from LEO to Lunar operations. Shake it down on unmanned maximum payload, low speed missions to haul masses of stuff up to L2 or low Lunar orbit.

And then, consider whether we want a minimum-mass, maximum speed manned version to take explorers and developers out to those heavy caches, in halo orbit or lunar orbit or landed on the Moon, or fall back on the chemical fast trajectories already developed for Apollo.

Human beings need fast transit times, for reasons others have pointed out already.

For the Mars mission, it might make sense for the manned ship, on a 90 or even 30 day trajectory, to be the second craft launched on these rockets--the first being a large unmanned ship the human crew rendezvous with and use--hauling out the lander and perhaps even the reaction mass the fast manned ship needs to return with (obviously a bold and risky step, that last one!)

For the Earth-Moon cargo hauler application, one reason I dislike the concept of using solar power for the input to the compression is, at the beginning of the slow spiral out from Earth the craft is in shadow half the time; its engines can only work in sunlight. If they can close the cycle by drawing the input energy for the next pulse from the output and storing it, the early part of the outward spiral (or terminal part of an inward one) can chug on steadily eclipse or no eclipse, doubling the rate of progress. Obviously as it spirals out, this is less and less of a problem, but it bugs me and the solution seems so obvious. ::)

This is not a criticism of our good scientists in Washington state; their proposal is for a Mars ship, which will mainly be operating well away from Earth's shadow.

Offline aero

  • Full Member
  • ****
  • Posts: 1247
  • Liked: 33
Is it possible to start in a slightly inclined orbit configured so that the vehicle is in constant sunlight for the month or so of launch? Then use the high efficiency engine to remove the effects of the initial inclination?

I can can't describe the orbit very well, but imagine a polar orbit with the radius vector perpendicular to the Earth/Sun line. Such an orbit would keep the vehicle constantly in the sun for a few months, but of course we don't want to waste the energy that getting into or starting from a polar orbit would need. So lay the orbit down so that the vehicle is always in sunshine, but some of the sun's rays nearly graze the Earth. Such an orbit would not be an LEO because if it were then it would also need to be nearly polar.

Just a thought.

Offline Elmar Moelzer

  • Full Member
  • ****
  • Posts: 1677
  • Liked: 91
One question: if we wanted to double the average thrust, what would be the path?

Scale up the engine? Shorten time between pulses?

Putting two engines in tandem does not seem to be viable.

Speeding up the pulse rate is not "easy," apparently not at all--if we are talking about fabricating a new half-kg lithium ring in the engine for every pulse, quite obviously not. But it's the approach that has the biggest payoff by far. Earlier press releases and published papers mention an ambition to get the rate up to once every 10 seconds (same ballpark as every 14) and I think they may have mentioned the possibility of once every second! If they did they may be regretting it now; OTOH if they can get a working system operational in space that pulses say once every 20 seconds, interest and funding to strive to improve the rate should be quite forthcoming, I'd think. Who knows what the limit would be then?

The faster the pulse rate, the greater the thrust to weight ratio. Elmar Moelzer mentions the need for the exhaust products to completely clear the system as a limit, but with the main pulse exiting at 50,000 meters/sec it's hard for me to believe this can be a factor after say 1/100 of a second, even granting there will be a bell curve or something like that of particle speeds, so some small fraction will indeed be lingering about the chamber long after most of it has departed.

More seriously, while I've found presentations where the developers predict that the overall mass of a rocket capable of the performance they are talking about (50 km sec exhaust, about half a kg per pulse, pulses in the range of 1 every 10 seconds) would be about 15 metric tons, I have no idea just how much of that is the ring-forming mechanism, or how the mass of that would scale with increasing pulse frequency (assuming the result can be accomplished at all). That it might wind up massing as the square of the frequency does not seem unreasonable to me; then the question is, how much would the baseline component capable of once every 10 seconds be? 100 kg? A tonne? That governs the point at which scaling up the pulse rate dominates the rising mass of the engine.

However I see no reason why any other component of the system (except the FRC injector, obviously) should rise in mass with it. Each pulse is the same; a magnetic nozzle system that can handle a pulse every minute should be able to handle one every second just as well...

Well, let me qualify--two things; one, the system proposed uses solar power input to drive the compression stage of the pulse--while I understand why they are going for that I've always felt the need to stress there's no fundamental reason the compression power has to come from outside the cycle! It could easily be drawn from the exhaust phase of the cycle, as a drain on the magnetic nozzle. With gains in power due to fusion in the range of 100 or more, it ought to be pretty doable. With power for the compression phase coming from outside the system, obviously that power source has to be scaled up in proportion to the pulse rate, and that's bad. (Unless we have onboard also the power generation system the same inventors are also working on, perhaps? :)) But if the compression power is drawn from the previous exhaust pulse, the shorter the time scale the better, because that's power that has to be stored. (The same is true with solar power or other external power input--it has to be accumulated in capacitors; these same capacitors could simply store the full energy needed for the next pulse from the previous one, so either way the power storage system is required--and included in the 15 tonne estimate already).

Two, obviously real world systems are never perfectly efficient; the magnetic systems will produce some waste heat, and the more cycles we get per minute, the greater the waste power is; that forms another practical limit and system that scales up in mass with the increased pulse rate.

All of this is obviously voting to bell the cat; to consider the tradeoffs involved in pushing the pulse rate up we have to first be able achieve the upper end of the range considered at all! It all depends on the system used to form and position the rings, how reliable it is, and so forth.

But until the scaling up of the formation mechanism, the FRC injector rate, and the heat rejection system starts to become a very significant fraction of the total engine mass, clearly the faster the pulse rate the better.
----
All that said, what if you can't increase the pulse rate beyond a certain point? Then I'd say that it would be better to keep the minimum scale of the engine system, and build many of them to increase the thrust, rather than to make the engine bigger in proportion to the mass of each pulse. It seems reasonable to me that it could be scaled up--that 5 kg of lithium can fuse 10 times the mass of fusible isotopes--but while there is little reason to think any component of the engine would scale up at a higher rate than linear, there is no reason to think any significant part of the mass would scale up at less than linear rate either. So we'd need a magnetic nozzle, including the compression component, 10 times the mass of the one proposed, and 10 times the mass of power storage capacitors, and so on. If compression is powered by solar power we need 10 times more panel area. Etc.

Well, I said I am not sure how much the process can be accelerated and named one potential limitation (that might be completely off) that my layman brain could make up. I am sure that once they can show that their idea works, they will be joined by many more engineers that will work to solve the problem with the reloading. Personally, I would like to see the engine fire at 10 Hz not 1/10 of Hz one day.
In regards to the solar panels. As I mentioned before, MSNW has considered converting some of the fusion energy into energy to drive the engines and they could do so. But for a mars mission solar panels are compact and simple enough to provide that energy. Plus, dont forget that you need energy for the other systems of the spacecraft as well. In any case, there is no reason to complicate the system by introducing a complex energy conversion system, if the potential gains are too low.

Offline Stormbringer

  • Full Member
  • *
  • Posts: 183
  • Liked: 3
off the wall idea here:  How about getting the metal storm guys to design the injector system? Metal storm fires lots and lots of rounds per second. i know it has lots of barrels in parallel so the trajectory would be changing all the time but perhaps the design could be modified so that all the trajectories intersect at the laser focus spot.

« Last Edit: 10/30/2013 05:58 AM by Stormbringer »

Tags: