Kinetic impact induced fusion

Antius · 2016-09-23 09:26:11

Possibly a more practical approach to nuclear fusion than laser driven inertial confinement. Coupling between the pellet and driver (rail gun) is much more efficient than laser induced fusion, because there is minimum pellet surface ablation and less energy loss through thermal radiation at the pellet surface. A mass driver is also much more energy efficient than a laser, so overall energy losses are much lower. The catch is that the projectile must be accelerated to 1000km/s to achieve sufficient KE to produce a stable burn on impact. That suggests either horrendous acceleration (hence the diamond bullet) or a very long accelerator. Still, this may be the most practical approach for pure fusion nuclear-pulse propulsion. It would be tough to miniaturise this technology sufficiently for a ground launched Orion though. Minimal realistic accelerator length is at least a few km. So this would only work for really big ships, presumably launched from the ocean. I would be interested to know if required pellet velocity dcreases with scale. For a kinetic impact, coupling should generally increase in efficiency as the size of colliding pellets increases and surface energy losses should go down. A solid projectile is more difficult to accelerate than a plasma but has vastly superior energy density.

http://www-pub.iaea.org/MTCD/Meetings/F … _p7-30.pdf

Last edited by Antius (2016-09-23 09:34:34)

Tom Kalbfus · 2016-09-23 11:48:32

Would the Rama Starship, that I outlined in the other thread be big enough to use this kind of propulsion?

Just as a review it is 54 km by 16 km wide on the inside or 34 miles long and 10 miles wide in Imperial Units. What if we made it 36 miles long and 12 miles wide on the outside? One can accelerate pellets along the inside and collide them in reaction chambers to produce thrust. If the rim is 16 km long, it is 50 km in circumference. To get it up to 1000 km/sec would require circle the rim 20 times per second, the centripedal acceleration would be 12,882,170 times Earth's gravity. Well actually 1000 km/sec of relative velocity is required so we can cut that down to 500 km/second going in opposite directions, and colliding pellet with pellet, in which case the centripedal acceleration for both pellets would be much reduced at "only" 3,220,542 times the force of gravity at Earth's surface. Each pellet would have to be precisely aimed to collide within the reaction chamber, or maybe behind an Orion Style pusher plate. When the pellets hit, you get a nuclear explosion. Some heat exchangers can turn some of that waste heat into electricity for use in the generation of artificial sunlight and other things such as car charging stations and the like on the inside. the space between the inner and outer hulls is where we would put the apparatus and fuel supply.

kbd512 · 2016-09-23 12:25:54

If the particle to accelerate is small enough, this might be feasible. The potential problem I see is the switching technology. We've been working on that issue for some time now, but I don't know exactly where we're at. The highest plasma velocity achieved to date that I'm aware of is 100km/s using new switching technology, so we need to improve velocity by a factor of ten to achieve fusion.

Antius · 2016-09-23 14:54:21

Tom Kalbfus wrote:

Would the Rama Starship, that I outlined in the other thread be big enough to use this kind of propulsion?
http://ebruneton.free.fr/rama3/rama-northpole.jpg
Just as a review it is 54 km by 16 km wide on the inside or 34 miles long and 10 miles wide in Imperial Units. What if we made it 36 miles long and 12 miles wide on the outside? One can accelerate pellets along the inside and collide them in reaction chambers to produce thrust. If the rim is 16 km long, it is 50 km in circumference. To get it up to 1000 km/sec would require circle the rim 20 times per second, the centripedal acceleration would be 12,882,170 times Earth's gravity. Well actually 1000 km/sec of relative velocity is required so we can cut that down to 500 km/second going in opposite directions, and colliding pellet with pellet, in which case the centripedal acceleration for both pellets would be much reduced at "only" 3,220,542 times the force of gravity at Earth's surface. Each pellet would have to be precisely aimed to collide within the reaction chamber, or maybe behind an Orion Style pusher plate. When the pellets hit, you get a nuclear explosion. Some heat exchangers can turn some of that waste heat into electricity for use in the generation of artificial sunlight and other things such as car charging stations and the like on the inside. the space between the inner and outer hulls is where we would put the apparatus and fuel supply.

The accelerator would need to be linear and bigger is better. Even at 54km long, acceleration would need to be close to 1million g. This is about the limit of what physical materials can withstand.

Antius · 2016-09-24 07:03:59

I carried out a few scoping calculations. I am initially assuming a silicon carbide impactor. It is relatively cheap and has crushing strength of 2GPa. For a right circular cylinder bullet (L = D) weighing 1 gram, D = 0.735cm. Assuming a design factor of 2, the maximum tolerable pressure on the prjectile is 1GPa, corresponding to a maximum possible acceleration of 40million m/s2. The required accelerator length is 12.5km. For a 0.1gram bullet of the same proportions, maximum possible acceleration is 91million m/s2 and accelerator length is 5.5km. Basically, reducing projectile mass by a factor of 10 cuts the required accelerator length by roughly half. So a 1mg projectile could be accelerated by a mass driver about 1km long and would deliver an explosive yield of about 7tonnes TNT. Maybe enough for a ground launch Orion if pulse frequency is high enough.

Tom Kalbfus · 2016-09-24 08:59:37

Antius wrote:

I carried out a few scoping calculations. I am initially assuming a silicon carbide impactor. It is relatively cheap and has crushing strength of 2GPa. For a right circular cylinder bullet (L = D) weighing 1 gram, D = 0.735cm. Assuming a design factor of 2, the maximum tolerable pressure on the prjectile is 1GPa, corresponding to a maximum possible acceleration of 40million m/s2. The required accelerator length is 12.5km. For a 0.1gram bullet of the same proportions, maximum possible acceleration is 91million m/s2 and accelerator length is 5.5km. Basically, reducing projectile mass by a factor of 10 cuts the required accelerator length by roughly half. So a 1mg projectile could be accelerated by a mass driver about 1km long and would deliver an explosive yield of about 7tonnes TNT. Maybe enough for a ground launch Orion if pulse frequency is high enough.

Easier to use plutonium for a ground launch. I think nuclear fusion is better for interplanetary and interstellar trips than for ground launch, where you have a whole host of better options than nuclear fusion. What you want for a ground launch I high acceleration to overcome Earth's gravity and also to increase your velocity above that. With interstellar travel using generation ships, I think nuclear fusion is best. Fusion is relatively cheap compared to the other options. Lasers and other beamed propulsion methods required sustained homefront efforts, and a generation ship would take centuries to reach its destination, that is asking a bit much for the society that sent the ship. Nuclear fusion is contained within the ship, and it is capable of reaching 5% of the speed of light and then slowing down upon arrival. Antimatter is expensive and dangerous to store for long periods of time, say a couple centuries, fusion fuel (Deuterium, Tritium, and Helium-3) can be stored in tanks made of matter without exploding! the Interstellar ramjet has problems related to drag, has a slow acceleration rate, and a maximum speed, and it would be much easier just to use stored fuel for a generation ship like Rama.

Antius · 2016-09-24 13:09:08

Plutonium fission bombs are a terrible idea for ground launch. The main point of pursuing pure fusion rather than two-phase devices is to avoid contaminating the biosphere with fission products and heavy actinides. This problem has stalled nuclear pulse propulsion for nearly 60 years.

If we can crack the pure fusion problem, then the sort of colony planting missions GW was talking about become practical and politically achievable. Fast neutrons will activate some of the debris, but overall radiotoxicity is orders of magnitude lower than even the cleanest bomb propelled design. Far out is space, away from the Earth there are more options. But what we most need at this point is cheap ways of getting large items into orbit, like small O'Neill colonies and SPS factories. These would require a lot of inorbit assembly if we launched them using Falcon heavy and costs would be unaffordable for the scale of infrastructure needed. We need something that can lift 100,000 tonne payloads for a few tens of dollars per kg and take it up to L5 with minimal additional complication. Its hard to see any other option than nuclear pulse, but it will never be politically workable if it is fission based.

Antius · 2016-09-24 13:26:08

kbd512 wrote:

If the particle to accelerate is small enough, this might be feasible. The potential problem I see is the switching technology. We've been working on that issue for some time now, but I don't know exactly where we're at. The highest plasma velocity achieved to date that I'm aware of is 100km/s using new switching technology, so we need to improve velocity by a factor of ten to achieve fusion.

Some mass driver designs used light beams to trigger the accelerator coils as the projectile passed. Reactive devices like this will always have relatively long time constants, even if switches are solid state. One option would be precise calculation of the projectile position as a function of time and simply activating the coils in series based on calculation rather than measurement. This requires some amount of quality control in the manufacture of projectiles to very precise mass, but it gets around the switching problem.

Tom Kalbfus · 2016-09-24 13:55:08

Antius wrote:

Plutonium fission bombs are a terrible idea for ground launch. The main point of pursuing pure fusion rather than two-phase devices is to avoid contaminating the biosphere with fission products and heavy actinides. This problem has stalled nuclear pulse propulsion for nearly 60 years.
If we can crack the pure fusion problem, then the sort of colony planting missions GW was talking about become practical and politically achievable. Fast neutrons will activate some of the debris, but overall radiotoxicity is orders of magnitude lower than even the cleanest bomb propelled design. Far out is space, away from the Earth there are more options. But what we most need at this point is cheap ways of getting large items into orbit, like small O'Neill colonies and SPS factories. These would require a lot of inorbit assembly if we launched them using Falcon heavy and costs would be unaffordable for the scale of infrastructure needed. We need something that can lift 100,000 tonne payloads for a few tens of dollars per kg and take it up to L5 with minimal additional complication. Its hard to see any other option than nuclear pulse, but it will never be politically workable if it is fission based.

Space%20Elevator%20Text.jpg

Antius · 2016-09-27 04:40:28

Here is a link to a paper written by Winterberg, some 30 years old now:

http://www.degruyter.com/downloadpdf/j/ … 6-0307.xml

The concept employs a hollow projectile filled with propellant and accelerated using a coil gun. During acceleration, the propellant vaporises, cooling the projectile and forming an expanding plasma bubble behind the projectile further boosting acceleration. Speeds of a few hundred km/s are possible with accelerator length of 260m. Projectile mass is ~1gram. If a 1gram projectile is capable of fusing 1gram of fuel, then the resultant explosive yield is 70KT. If the combined target and projectile have mass of 1kg and 70% of neutron energy is absorbed in the target, then exhaust velocity is 20,000km/s and total impulse is 20million N-s per pulse. If pulse rate is 10 per second, then the pulse engine will accelerate 10,000 tonnes of material at 2g acceleration.

JoshNH4H · 2016-10-27 09:50:45

One potential issue with kinetic impact fusion is that most materials are not stiff enough for it to work really well. If you take a "normal" material and smash it into another "normal" material at high speed most of the energy will be absorbed by mechanical deformation of the material as opposed to being transferred to the target. Furthermore, because the characteristic time (speed of sound in the material over size of the shell, in this case) is much longer than the characteristic time for inertial confinement fusion, you're not likely to be able to efficiently transfer energy to the target.

New Mars Forums

Announcement

#1 2016-09-23 09:26:11

Kinetic impact induced fusion

#2 2016-09-23 11:48:32

Re: Kinetic impact induced fusion

#3 2016-09-23 12:25:54

Re: Kinetic impact induced fusion

#4 2016-09-23 14:54:21

Re: Kinetic impact induced fusion

#5 2016-09-24 07:03:59

Re: Kinetic impact induced fusion

#6 2016-09-24 08:59:37

Re: Kinetic impact induced fusion

#7 2016-09-24 13:09:08

Re: Kinetic impact induced fusion

#8 2016-09-24 13:26:08

Re: Kinetic impact induced fusion

#9 2016-09-24 13:55:08

Re: Kinetic impact induced fusion

#10 2016-09-27 04:40:28

Re: Kinetic impact induced fusion

#11 2016-10-27 09:50:45

Re: Kinetic impact induced fusion

Board footer