Rocket thrust

If polywell fusion is developed, in what ways will the world change for better or worse? Discuss.

Moderators: tonybarry, MSimon

MSimon
Posts: 14334
Joined: Mon Jul 16, 2007 7:37 pm
Location: Rockford, Illinois
Contact:

Post by MSimon »

Mindblast wrote:
clonan wrote:Mindblast,

Would you really want all that spattering on your grids?
Well i'm not sure.. i mean the usual figures for a polywell are like 20% thermal power anyway. I guess a lot if not most of that would be on the grids, so obviously they would have to be able to handle like 1.2GW thermal load in a 6GW reactor. For running some fans the first 5-10 minutes of the flight you don't even need 1.2GW power.

If that fails.. maybe it would be possible to put a REB/ARC heater into the coolant loop to superheat the coolant using the electricity generated by the reactor.
The latest thinking is that in a pBj machine with sufficiently high B fields (3T for a 1 m bore - linearly [inverse] less for bigger bores) the only loading on the coils would be neutron impingement. No significant alpha impingement. Even with 6 MeV alphas. Currently experimental MgB SCs with 1 m bores are in the 9 T range. I have done the calculations with 3T coils and 6 MeV alphas and the gyro radius is on the order of .1 m. Actually it would be less than that as the field is higher near the coil casings.
Engineering is the art of making what you want from what you can get at a profit.

MSimon
Posts: 14334
Joined: Mon Jul 16, 2007 7:37 pm
Location: Rockford, Illinois
Contact:

Post by MSimon »

93143 wrote:I estimate a 6 GW reactor would require at least 350 tonnes of shielding, if it is to be operated near civilians. This represents a 3½" shell of lead on a ~10 metre sphere, to mitigate gamma radiation (about 1 cm of lead per factor of 10 reduction) from a thin boron-10 neutron absorber with maybe 6" of water (50 tonnes) ahead of it as a moderator (60x10^12 n/s coming off the reactor). If my calculations are correct, concrete would be more than twice as heavy as lead, because of the decreased stopping power against half-MeV gammas.

Is MSimon around? Does that sound halfway plausible? Or would you want more margin in a passenger vehicle?


The thing is, reducing the output power doesn't substantially decrease the amount of shielding required. It's logarithmic, so the higher the reactor power, the more mass-efficient the shielding is.

This is not for 70-seat regional jets. I'd love to see a truly monstrous fusion-powered aircraft, hypersonic or otherwise, that can afford to pack one (or two! or three!!) of these, and actually uses full power at takeoff...
In a rocket, reaction mass is available as shielding. And for gammas the only thing that counts is mass. (There are some details for low Z species that complicate the calculation but roughly - gammas are shielded by mass).

The shielding doesn't have to be a full 360 degree shield. Only passengers need to be shielded.
Engineering is the art of making what you want from what you can get at a profit.

clonan
Posts: 91
Joined: Tue Jun 10, 2008 4:16 pm

Post by clonan »

MSimon wrote:
The shielding doesn't have to be a full 360 degree shield. Only passengers need to be shielded.
Unless you are trying to do a SSTO ship, powered by the polywell at launch...

The people on the ground might disagree with you...

D Tibbets
Posts: 2775
Joined: Thu Jun 26, 2008 6:52 am

Post by D Tibbets »

MSimon wrote:
Mindblast wrote:
clonan wrote:Mindblast,

Would you really want all that spattering on your grids?
Well i'm not sure.. i mean the usual figures for a polywell are like 20% thermal power anyway. I guess a lot if not most of that would be on the grids, so obviously they would have to be able to handle like 1.2GW thermal load in a 6GW reactor. For running some fans the first 5-10 minutes of the flight you don't even need 1.2GW power.

If that fails.. maybe it would be possible to put a REB/ARC heater into the coolant loop to superheat the coolant using the electricity generated by the reactor.
The latest thinking is that in a pBj machine with sufficiently high B fields (3T for a 1 m bore - linearly [inverse] less for bigger bores) the only loading on the coils would be neutron impingement. No significant alpha impingement. Even with 6 MeV alphas. Currently experimental MgB SCs with 1 m bores are in the 9 T range. I have done the calculations with 3T coils and 6 MeV alphas and the gyro radius is on the order of .1 m. Actually it would be less than that as the field is higher near the coil casings.
The magrid with P-B11 may not have much charged particle impacts. But the collection grids would have to have some compromises in the range of alpha particles it could decellerate to almost zero speeds befor impact. Also, alot of that 20% thermal load would probably come from x-rays hitting everything.

Dan Tibbets
To error is human... and I'm very human.

MSimon
Posts: 14334
Joined: Mon Jul 16, 2007 7:37 pm
Location: Rockford, Illinois
Contact:

Post by MSimon »

D Tibbets wrote:
MSimon wrote:
Mindblast wrote: Well i'm not sure.. i mean the usual figures for a polywell are like 20% thermal power anyway. I guess a lot if not most of that would be on the grids, so obviously they would have to be able to handle like 1.2GW thermal load in a 6GW reactor. For running some fans the first 5-10 minutes of the flight you don't even need 1.2GW power.

If that fails.. maybe it would be possible to put a REB/ARC heater into the coolant loop to superheat the coolant using the electricity generated by the reactor.
The latest thinking is that in a pBj machine with sufficiently high B fields (3T for a 1 m bore - linearly [inverse] less for bigger bores) the only loading on the coils would be neutron impingement. No significant alpha impingement. Even with 6 MeV alphas. Currently experimental MgB SCs with 1 m bores are in the 9 T range. I have done the calculations with 3T coils and 6 MeV alphas and the gyro radius is on the order of .1 m. Actually it would be less than that as the field is higher near the coil casings.
The magrid with P-B11 may not have much charged particle impacts. But the collection grids would have to have some compromises in the range of alpha particles it could decellerate to almost zero speeds befor impact. Also, alot of that 20% thermal load would probably come from x-rays hitting everything.

Dan Tibbets
If X-Rays are 20% of total power and the intercepted area is 20% a maximum of 4% of the reactor power will impinge on the coils. I don't see much of a problem unless there is significant absorption in the He4 cooling circuit. I have a paper BOE design for handling 20% of the reactor power going into the grids. Something on the order of 5% to 10% would be a cake walk.

Edited for clarity. I left a few words out.
Last edited by MSimon on Sat Nov 14, 2009 6:05 am, edited 1 time in total.
Engineering is the art of making what you want from what you can get at a profit.

DeltaV
Posts: 2245
Joined: Mon Oct 12, 2009 5:05 am

Post by DeltaV »

If no heat exchanger material could withstand the REB, would heating the air/propellant with a REB-based free-electron laser generate excessive ozone at low altitudes?

A FEL beam could be focused to form a plasma blob on the centerline of air/propellant flow in a duct, which would reduce thermal loads on the "combustion chamber" walls, somewhat like the magnetic field does for QED-ARC.

http://www.answers.com/topic/free-electron-laser

"A free-electron laser, or FEL, is a laser that shares the same optical properties as conventional lasers such as emitting a beam consisting of coherent electromagnetic radiation which can reach high power, but which uses some very different operating principles to form the beam. Unlike gas, liquid, or solid-state lasers such as diode lasers, in which electrons are excited in bound atomic or molecular states, FELs use a relativistic electron beam as the lasing medium which moves freely through a magnetic structure, hence the term free electron.[1] The free-electron laser has the widest frequency range of any laser type, and can be widely tunable,[2] currently ranging in wavelength from microwaves, through terahertz radiation and infrared, to the visible spectrum, to ultraviolet, to X-rays[3]"

"FEL technology is considered by the US Navy as a good candidate for an antimissile directed-energy weapon. Significant progress is being made in increasing FEL power levels (the Thomas Jefferson National Accelerator Facility's FEL has demonstrated greater than 14 kW [12]) and it should be possible to build compact multi-megawatt class FEL weapons.[13] On June 9 2009 the Office of Naval Research announced it had awarded Raytheon a contract to develop a 100 kW experimental FEL.[14]"

93143
Posts: 1142
Joined: Fri Oct 19, 2007 7:51 pm

Post by 93143 »

MSimon wrote:In a rocket, reaction mass is available as shielding.
No, it isn't. The shielding has to still be there at burnout.

And reserves are still useless for this, because (a) the shadow shield will probably be at or near the top of the fuel tank and will thus get drained first, and (b) you might need the reserves.

Far better to just use a dedicated shield.
And for gammas the only thing that counts is mass. (There are some details for low Z species that complicate the calculation but roughly - gammas are shielded by mass).
I think I checked it once, for that specific energy of gamma ray. The post is here: viewtopic.php?p=10250#10250

IIRC I read the numbers off a graph somewhere, so they may be a bit shaky... Lots of low-Z nuclei in concrete, though...
The shielding doesn't have to be a full 360 degree shield. Only passengers need to be shielded.
Maybe for a big launcher that no one is going to get anywhere near during the launch. A single-core spaceplane would be far easier to use if it had a full 360° shield, though unfortunately the shield is probably too heavy. Ground crews are going to need bunkers...

If you could get the core size down to 6 m diameter instead of 10, the shielding mass would be less than 150 mT... still pretty heavy; more than half the mass of Bussard's notional SSTO...

A large fusion-powered airplane (which is actually what I was talking about) needs to be able to operate from an airport. So it would probably absolutely need 360° shielding on the reactor core(s).

Remember, you're trying to get more than 6 orders of magnitude reduction in neutron radiation intensity. Taking a 3-metre core radius as a baseline, this means an unshielded core has to be over 3 km away, neglecting the effect of the atmosphere.

For larger clusters the shielding mass efficiency starts going up even if the reactor size doesn't increase. Multiple cores in a single linear sleeve doesn't get you anything, but I calculate that a 5x5x6 cluster (150 cores, 900 GW) would require about 35% as much shielding mass as 150 individually-shielded reactors. So a Hypernova-class launcher might not have as much trouble getting a plausible power-to-weight ratio... not that anyone should be standing close enough to worry about neutrons during the launch anyway...

...but in space the shielding would be very useful, as you could get near the thing while it was under power. Maybe to land a starfighter in one of the docking bays... Another nice thing about such a large cluster is that it gives you fine-grained throttle control even if individual cores can't be throttled at all...

Yes, I know I sound like I'm smoking something. No, you can't have any...
DeltaV wrote:If no heat exchanger material could withstand the REB
That's probably a safe bet.

...

Free electron lasers would probably make great Polywell-powered weapons. For propulsion applications, the trouble is that a laser beam doesn't get absorbed that easily by air; a fair chunk of it will keep on going and hit something solid, with tragic results.

I was thinking of REB-heating an intermediate fluid, which would then be run through the heat exchanger in the engine at extreme temperatures.

It doesn't have to be a thermodynamic cycle, because you don't have to get any power out. If you want power out, you take it from the Brayton cycle represented by the heat-exchanger-driven airbreathing engine.

You could even use a refrigeration loop to dump waste heat into the intermediate fluid before it gets REB-heated. If the engine heat exchanger is a counterflow design, it might get the final temperature of the intermediate fluid tolerably low (yes, counterflow is heavier, but the reduced refrigeration requirements should help make up for that; this is quite a refrigerator we're talking about here)...

Considering that the energy in question is used rather than dumped, perhaps that would make it a heat pump rather than a refrigerator? This is essentially a zero-mass-overboard All Regeneratively Cooled design, at least until things start getting too fast and hot for it to work properly...

MSimon
Posts: 14334
Joined: Mon Jul 16, 2007 7:37 pm
Location: Rockford, Illinois
Contact:

Post by MSimon »

No, it isn't. The shielding has to still be there at burnout.


As long as the shielding is continuous it doesn't matter where in the tank it is. The effect will be the same.

I was contemplating needing only about 10% of the tankage for sufficient shielding. That would leave about 2% to 3% of the total for normal burn variance. And 7% or 8% for dire emergencies.

It is kinda hard to speculate with precision when all you have to work with is guesses. But any way. That is the approach I would take.
Engineering is the art of making what you want from what you can get at a profit.

DeltaV
Posts: 2245
Joined: Mon Oct 12, 2009 5:05 am

Post by DeltaV »

93143 wrote:For propulsion applications, the trouble is that a laser beam doesn't get absorbed that easily by air; a fair chunk of it will keep on going and hit something solid, with tragic results.
That's why I said the FEL beam should be focused. High (or even medium) power lasers are known to form a plasma in air when focused. Hopefully (1) the plasma blob absorbs FEL energy better than plain air, and maybe enhanced absortion occurs with LH2, H2O, ... added to the flow (2) what's left of the diverging FEL beam leaving the focus would not cause much damage, (3) the focal region plasma would not produce the amount of ozone that a REB would, (4) the upstream flow transparency is high enough not to impede focusing, and (5) a multi-MW FEL could be made lightweight/small/efficient.
93143 wrote:If you could get the core size down to 6 m diameter instead of 10, the shielding mass would be less than 150 mT... still pretty heavy; more than half the mass of Bussard's notional SSTO...
Maybe metamaterials can help. Layers designed for particular particles and energy bands:
viewtopic.php?t=1601

93143
Posts: 1142
Joined: Fri Oct 19, 2007 7:51 pm

Post by 93143 »

MSimon wrote: As long as the shielding is continuous it doesn't matter where in the tank it is. The effect will be the same.
...what?


Oh, I get it. The entire tank is in between the reactor and the crew. That makes more sense. I'm not sure what I was thinking...
I was contemplating needing only about 10% of the tankage for sufficient shielding. That would leave about 2% to 3% of the total for normal burn variance. And 7% or 8% for dire emergencies.

It is kinda hard to speculate with precision when all you have to work with is guesses. But any way. That is the approach I would take.
Okay, so the major advantage here is that you can use up the shield if absolutely necessary. (Plus you don't need a separate structure to hold the shield.) But wouldn't that give the crew severe radiation poisoning? Unshielded, you'd get on the order of 2e5 n/cm²·s even 50 m from the core. Hard to imagine an emergency dire enough to warrant absorbing that in preference to simply aborting the mission and re-entering...

Or have I got a really skewed idea of how much fast neutron radiation constitutes a hazard?

Will 10% be enough? On Bussard's spaceplane that's only about 12 mT of hydrogen, which on a 10-metre circular cross section is only about 150 kg/m² - that's a depth of a little over two metres, or the mass equivalent of 6" of water, or the hydrogen equivalent of 4'5" of water.

I assume we're talking about an SSTO here, since dry mass isn't nearly as much of a problem in deep space...
Last edited by 93143 on Sat Nov 14, 2009 8:48 pm, edited 2 times in total.

DeltaV
Posts: 2245
Joined: Mon Oct 12, 2009 5:05 am

Post by DeltaV »

DeltaV wrote:Hopefully ... (5) a multi-MW FEL could be made lightweight/small/efficient.
http://www1.jlab.org/Ul/Publications/do ... -05-02.pdf

93143
Posts: 1142
Joined: Fri Oct 19, 2007 7:51 pm

Post by 93143 »

DeltaV wrote:
93143 wrote:For propulsion applications, the trouble is that a laser beam doesn't get absorbed that easily by air; a fair chunk of it will keep on going and hit something solid, with tragic results.
That's why I said the FEL beam should be focused. High (or even medium) power lasers are known to form a plasma in air when focused. Hopefully (1) the plasma blob absorbs FEL energy better than plain air, and maybe enhanced absortion occurs with LH2, H2O, ... added to the flow (2) what's left of the diverging FEL beam leaving the focus would not cause much damage, (3) the focal region plasma would not produce the amount of ozone that a REB would, (4) the upstream flow transparency is high enough not to impede focusing, and (5) a multi-MW FEL could be made lightweight/small/efficient.
From what I can tell from the literature, laser absorption in plasmas isn't in the 99.999% range - I'm seeing 20% - 70%, although since I'm at home I can't really do a thorough search; that's only from a couple of papers on short-pulse lasers hitting objects...

Isotropic scattering of light in the focal region could produce a large wall heating load, impossible to magnetically shield against.

Also remember that a 100 MW BFR is way down in the bad part of the power-to-shielding-weight curve. This is not a multi-MW FEL we're talking about; it's a multi-GW FEL. Residual shine-through is going to be big.

I don't like it. And I think it's really unlikely that that plasma blob is going to absorb more than half of the laser power. If you have sources that say different, please point me to them.

...never mind the fact that even a free electron laser is going to have significant inefficiencies, and thus will need to be cooled. Where is the cooling power from the reactor and the laser going in this scheme?
93143 wrote:If you could get the core size down to 6 m diameter instead of 10, the shielding mass would be less than 150 mT... still pretty heavy; more than half the mass of Bussard's notional SSTO...
Maybe metamaterials can help. Layers designed for particular particles and energy bands:
viewtopic.php?t=1601
Okay, that paper is talking about cloaking against ultracold rubidium atoms at 4.5 nanokelvins, with a de Broglie wavelength large enough (5 µm, or about 10 times larger (!) than the optical lattice) that ordinary matter can be used. Neutrons at MeV energies have wavelengths in the tens of femtometres; what would you use to cloak against that? Neutronium? Even 0.5 MeV gammas have a wavelength of only 2.5 picometres...

D Tibbets
Posts: 2775
Joined: Thu Jun 26, 2008 6:52 am

Post by D Tibbets »

93143 wrote:
MSimon wrote:In a rocket, reaction mass is available as shielding.
No, it isn't. The shielding has to still be there at burnout.

And reserves are still useless for this, because (a) the shadow shield will probably be at or near the top of the fuel tank and will thus get drained first, and (b) you might need the reserves.

Far better to just use a dedicated shield.
And for gammas the only thing that counts is mass. (There are some details for low Z species that complicate the calculation but roughly - gammas are shielded by mass).
I think I checked it once, for that specific energy of gamma ray. The post is here: viewtopic.php?p=10250#10250

IIRC I read the numbers off a graph somewhere, so they may be a bit shaky... Lots of low-Z nuclei in concrete, though...
The shielding doesn't have to be a full 360 degree shield. Only passengers need to be shielded.
Maybe for a big launcher that no one is going to get anywhere near during the launch. A single-core spaceplane would be far easier to use if it had a full 360° shield, though unfortunately the shield is probably too heavy. Ground crews are going to need bunkers...

If you could get the core size down to 6 m diameter instead of 10, the shielding mass would be less than 150 mT... still pretty heavy; more than half the mass of Bussard's notional SSTO...

A large fusion-powered airplane (which is actually what I was talking about) needs to be able to operate from an airport. So it would probably absolutely need 360° shielding on the reactor core(s).

Remember, you're trying to get more than 6 orders of magnitude reduction in neutron radiation intensity. Taking a 3-metre core radius as a baseline, this means an unshielded core has to be over 3 km away, neglecting the effect of the atmosphere.

For larger clusters the shielding mass efficiency starts going up even if the reactor size doesn't increase. Multiple cores in a single linear sleeve doesn't get you anything, but I calculate that a 5x5x6 cluster (150 cores, 900 GW) would require about 35% as much shielding mass as 150 individually-shielded reactors. So a Hypernova-class launcher might not have as much trouble getting a plausible power-to-weight ratio... not that anyone should be standing close enough to worry about neutrons during the launch anyway...

...but in space the shielding would be very useful, as you could get near the thing while it was under power. Maybe to land a starfighter in one of the docking bays... Another nice thing about such a large cluster is that it gives you fine-grained throttle control even if individual cores can't be throttled at all...

Yes, I know I sound like I'm smoking something. No, you can't have any...
DeltaV wrote:If no heat exchanger material could withstand the REB
That's probably a safe bet.

...

Free electron lasers would probably make great Polywell-powered weapons. For propulsion applications, the trouble is that a laser beam doesn't get absorbed that easily by air; a fair chunk of it will keep on going and hit something solid, with tragic results.

I was thinking of REB-heating an intermediate fluid, which would then be run through the heat exchanger in the engine at extreme temperatures.

It doesn't have to be a thermodynamic cycle, because you don't have to get any power out. If you want power out, you take it from the Brayton cycle represented by the heat-exchanger-driven airbreathing engine.

You could even use a refrigeration loop to dump waste heat into the intermediate fluid before it gets REB-heated. If the engine heat exchanger is a counterflow design, it might get the final temperature of the intermediate fluid tolerably low (yes, counterflow is heavier, but the reduced refrigeration requirements should help make up for that; this is quite a refrigerator we're talking about here)...

Considering that the energy in question is used rather than dumped, perhaps that would make it a heat pump rather than a refrigerator? This is essentially a zero-mass-overboard All Regeneratively Cooled design, at least until things start getting too fast and hot for it to work properly...
I'm splitting hairs, but shielding from gamma's has some subtilities. Ground personel exposure is easily managed by distance. The distance / bunkering used for chemical rockets may be adiquate for shielding from the reactor. Again, the takeoff runway, climb out and desent paths would be chosen to minimize danger to people on the ground-again the same as with chemical rockets.
Shielding on the ship needs to be addressed. Shielding by fuel would help. The increased weight from pure structural shielding would require greater power/ time from the reactor which would increase the radiation. Some compromize between the mission radiation shielding between the fuel and perminate structure would meet mission goals.

As far as manovering to dock with a space station, etc. The fusion thrust could be turned off well before close approach. Coasting the final distance, with terminal chemical thrusters providing the final few MPH adjustments. This assumes the reactor could be turned off and on repeatedly, or could be idled to a very low levels.

Dan Tibbets
To error is human... and I'm very human.

MSimon
Posts: 14334
Joined: Mon Jul 16, 2007 7:37 pm
Location: Rockford, Illinois
Contact:

Post by MSimon »

But wouldn't that give the crew severe radiation poisoning? Unshielded, you'd get on the order of 2e5 n/cm²·s even 50 m from the core.


Depends on how dire the emergency and how long the burn.

LD50 is around 100 to 300 rads IIRC. Keep the exposure below that for EMERGENCIES.
Engineering is the art of making what you want from what you can get at a profit.

93143
Posts: 1142
Joined: Fri Oct 19, 2007 7:51 pm

Post by 93143 »

...huh. According to my calculations, at 50 metres from an unshielded 6 GW BFR at full power, it should take about half an hour to absorb one rad.

Maybe this isn't so bad after all...

That's assuming 60e12 neutrons per second, with a neutron energy of 2.9 MeV, range 50 m yielding 31416 m² over which to distribute the neutrons. Result is 190986 n/cm²·s, and for a sitting person estimated at 4500 cm² and 70 kg, that's 12277667 n/kg·s, or 5.7e-6 J/kg·s, or 0.00057 rad/s. Assuming all the neutrons are absorbed, because I don't know how to calculate the fraction.

Is that about right?

Of course, if the reactor is closer the dose gets larger... but structure should help mitigate...

Just using water and boron-10, without the extra (heavy) gamma absorber, increases the time to a given dose by a factor of about 6 due to the lower energy of the gammas as compared with the expected neutron energy. Also, neutrons in the expected energy range are apparently 10-20 times as damaging as gammas for a given dose...

Post Reply