1
00:00:16,090 --> 00:00:20,730
So, Michael did a really lovely job of introducing basic concepts and fusion.

2
00:00:21,030 --> 00:00:24,030
Some of the problems we're facing with fusion.

3
00:00:24,360 --> 00:00:29,440
And I really want to think about where is he going next? How can we take it further?

4
00:00:29,460 --> 00:00:37,770
What are the devices that we're going to consider? So I hope by the end of this talk, I will at least somehow convince some of you that Stellarator,

5
00:00:37,800 --> 00:00:42,150
which I will introduce in a moment, could potentially be the future of fusion.

6
00:00:42,810 --> 00:00:49,410
Okay. So before I get started, there are obviously other different types of fusion devices.

7
00:00:49,740 --> 00:00:55,050
One being an actual component fusion, which actually we'll talk about in the last topic of the day.

8
00:00:55,650 --> 00:00:58,200
But we also have other areas of research.

9
00:00:58,260 --> 00:01:05,430
I'm putting them on the slides here just to be transparent that, you know, tokamak accelerators, etc. are not the only form of fusion.

10
00:01:05,760 --> 00:01:09,540
Other things do exist, but I'm not really gonna spend any time on these.

11
00:01:09,780 --> 00:01:12,990
I just want to point out there are other areas of active research.

12
00:01:14,040 --> 00:01:18,570
So arguably the most popularised fusion device is the Tokamak.

13
00:01:19,230 --> 00:01:22,740
Most of you probably would have heard of it. It's the one that makes it into the news.

14
00:01:22,950 --> 00:01:27,480
It's one that has a lot of fancy science results, maybe overhyped, as we've heard.

15
00:01:27,490 --> 00:01:30,810
But, you know, this is arguably the most famous fusion device.

16
00:01:31,350 --> 00:01:36,100
So a really quick recap of what a tokamak is. It's basically a doughnut.

17
00:01:36,120 --> 00:01:42,210
We say it has this as a little symmetry. I'm sure you all know what it means, but it means it's symmetric with two pi.

18
00:01:42,840 --> 00:01:51,180
So how do they work? Well, we for most modern day tokamaks, we have these capacitors in which the Tokamak is sandwiched in between.

19
00:01:51,610 --> 00:01:56,309
And we discharge a current. This current induces a toroidal electric field.

20
00:01:56,310 --> 00:01:59,970
So that's one that goes around the doughnut. So it'd be like the jam of the doughnut.

21
00:02:00,600 --> 00:02:06,809
This electric field increases the current. The current very nicely produces these colloidal magnetic fields.

22
00:02:06,810 --> 00:02:14,130
So that's like the basically the shape of those red poke at the pointer, those red circles up here.

23
00:02:14,370 --> 00:02:21,550
So this would be the colloidal clay that is generated by the current that goes through or part of it is and this is themselves confining, right?

24
00:02:21,570 --> 00:02:26,640
It keeps it in this doughnut shape. Those magnetic fields contain on our tokamak.

25
00:02:27,660 --> 00:02:33,719
So that's what I took my first very quick refresh to those we can think back to Michael talk you know and that

26
00:02:33,720 --> 00:02:40,710
has been talked about tokamaks have their own problems and they are numerous and I'm not going to get over that.

27
00:02:40,980 --> 00:02:48,000
But some real main ones that we have to consider, all one we actually need to charge them up is means they are discontinuous and use.

28
00:02:48,450 --> 00:02:56,070
This is extremely problematic when we think about you know from a commercial standpoint we don't want to be connecting discontinuous problem.

29
00:02:57,420 --> 00:03:07,040
Any second now, it will hopefully come back up. We don't want a discontinuous power supply to our national grid.

30
00:03:10,780 --> 00:03:17,020
The irony is not lost on me. You know, this is why we need stuttering.

31
00:03:18,400 --> 00:03:25,140
So, like I mentioned, they have this steroid occurrence that is actually extremely problematic.

32
00:03:25,150 --> 00:03:30,600
It wasn't mentioned too much by my uncle, but this is extremely problematic because they drive their own instabilities.

33
00:03:30,610 --> 00:03:34,510
There is a whole host of instabilities associated with this toroidal current.

34
00:03:34,900 --> 00:03:42,250
They lead to these violent outbursts that you saw on that on that video that Michael played when we had those flashes of white light,

35
00:03:42,550 --> 00:03:45,640
that was the violent disruptions, again, partly due to these.

36
00:03:46,060 --> 00:03:53,800
So that that's one problem. And the assignment problem I really want to mention is that Tokamaks have this empirically observed density limit.

37
00:03:54,100 --> 00:04:01,420
That means we hit a wall of the density that we're allowed. And for some reason that we're not entirely sure about yet, we can't go beyond this.

38
00:04:01,840 --> 00:04:08,739
And as you can imagine, that is really bad for fusion. We want high densities because then fusion is more likely to happen and the likelihood of

39
00:04:08,740 --> 00:04:13,360
a reaction actually occurring is going to be greater if we have these high densities.

40
00:04:13,810 --> 00:04:20,410
So these are some problems of tokamaks. So I'm going to propose to you an alternative that is rising,

41
00:04:20,440 --> 00:04:26,580
arguably in popularity amongst some and all people in the in the field is that that's the stellarator.

42
00:04:26,590 --> 00:04:32,740
So how can we think what is the stellarator? So the way I'm going to explain this to you is the following.

43
00:04:33,070 --> 00:04:39,550
I take my tokamak, I twist it, I stretch it, and maybe compress it in some way.

44
00:04:39,790 --> 00:04:43,210
Just kind of really moulder around and I'm going to get a stellarator.

45
00:04:43,690 --> 00:04:48,200
So what am I actually doing? What? What do I mean when I say I'm twisting, stretching, etcetera?

46
00:04:48,820 --> 00:04:53,830
Well, these purple surfaces that you see here are the flux surfaces that Michael spoke about earlier.

47
00:04:54,250 --> 00:05:00,370
We hope and pray that all particles are going to be confined, these purple surfaces.

48
00:05:01,150 --> 00:05:06,010
And the idea is they're allowed to steam along these surfaces, but we don't want them drifting off of them.

49
00:05:06,490 --> 00:05:10,600
And a tokamak, generally speaking, has these like nested surfaces.

50
00:05:10,900 --> 00:05:14,660
So the way you can think about this is if you think about one of the Russian dolls.

51
00:05:14,920 --> 00:05:20,200
But we have one is the other side. The other is like the other. We hope that we have these kind of purple surfaces.

52
00:05:20,200 --> 00:05:24,489
One is like the other is like the other. And I'm going to take each of my purple surfaces.

53
00:05:24,490 --> 00:05:31,360
I'm going to twist it in some funky way, and I'm going to get some crazy geometry that has no is no longer acce symmetric.

54
00:05:32,020 --> 00:05:36,280
So one thing I really want to draw your attention to here, because it's going to come up later in the talk.

55
00:05:36,720 --> 00:05:40,930
You know, it is, you know, tokamak nice and symmetric. It has this by symmetry.

56
00:05:41,260 --> 00:05:44,530
And look how simple the shape is. It looks nice, right?

57
00:05:45,040 --> 00:05:48,850
The stellarator. Now the hand has this very nontrivial geometry.

58
00:05:49,270 --> 00:05:53,200
And as such, the coils that can go around the outside have to match.

59
00:05:53,200 --> 00:05:56,620
This complicated geometry has to recreate these magnetic surfaces.

60
00:05:56,890 --> 00:06:04,630
So they are just going to be more complicated and it's not as trivial a problem to create the shape in a stellarator as it is in a tokamak.

61
00:06:05,740 --> 00:06:11,500
So that's my vague introduction to what a stellarator is. I hope it gives you some kind of a feel for what they look like.

62
00:06:12,790 --> 00:06:14,259
But you know, as you can imagine,

63
00:06:14,260 --> 00:06:20,470
I can't just twist my device in whatever way I want and expect to get a fusion reaction that's going to work perfectly.

64
00:06:20,500 --> 00:06:23,710
You can just imagine that not everything is going to work.

65
00:06:23,890 --> 00:06:27,760
So we have to ask ourselves, what devices are we actually allowed?

66
00:06:28,240 --> 00:06:34,600
So Michael spoke about in his talk very kindly leading up to mine that we have these regions of good and bad curvature.

67
00:06:34,900 --> 00:06:41,110
So on the bad curvature region, this is where those instabilities all unstable.

68
00:06:41,380 --> 00:06:43,870
They grow in amplitude. This is really problematic.

69
00:06:44,230 --> 00:06:52,570
But on the good curvature region, we have, you know, stability, the amplitudes of turbulence, the suppressed, etc. we like that side.

70
00:06:53,170 --> 00:06:58,990
So the tokamaks, the way that they deal with the fact that we have turbulence, etc., as Michael very kindly explained earlier,

71
00:06:59,350 --> 00:07:04,780
was they take stuff from the outside where it's unstable and these are all the way to the inside where it's stable.

72
00:07:05,110 --> 00:07:11,310
And the jargon within the fusion community to describe this is something called the safety factor, which I put up here.

73
00:07:11,320 --> 00:07:17,229
It's called Q for some unknown reason, and it's just basically telling us the number of toroidal turns.

74
00:07:17,230 --> 00:07:23,799
So the number of times we go down way along the doughnut bars, over the number of colloidal terms,

75
00:07:23,800 --> 00:07:31,030
which is the number of times we go the short way around the doughnut. And for tokamaks, generally speaking, we want this to be greater than one.

76
00:07:31,210 --> 00:07:37,330
And the reason we want this is because then the particles will on average see good curvature.

77
00:07:37,750 --> 00:07:41,050
You know, overall they will those perturbations will be stable.

78
00:07:41,860 --> 00:07:44,920
So that's what we do for Tokamaks. You know, I care about Stellar.

79
00:07:45,100 --> 00:07:49,089
So what do I do to celebrate as well? You know, the not actually symmetric.

80
00:07:49,090 --> 00:07:53,050
And so it's not as trivial as it is in the Tokamak.

81
00:07:53,380 --> 00:07:59,830
We don't necessarily have an intuitive feel of what the good and bad curvatures are in the stellarator.

82
00:08:00,730 --> 00:08:04,930
So we have to be more intelligent. We have to think a bit more.

83
00:08:05,050 --> 00:08:09,250
And I was very proud of this shape. But we have to think outside the non actually symmetric.

84
00:08:11,940 --> 00:08:14,459
And I will tell you, I laughed at this when I came up with it.

85
00:08:14,460 --> 00:08:22,330
Say, well, let me say before I even decide what I'm not allowed to like, before I tell you what I am and what I want for my salary.

86
00:08:22,740 --> 00:08:25,890
We have to think what are excluded? What are we not allowed to do?

87
00:08:26,310 --> 00:08:30,230
So in order to think, well, I can't have, I need to know what I'm worried about.

88
00:08:30,240 --> 00:08:34,800
Right. So there are two main problems that I want to talk about today that we have to be very worried about.

89
00:08:35,220 --> 00:08:40,140
So they are magnetic drift and this thing could be a classical transport, which hasn't been introduced yet.

90
00:08:40,140 --> 00:08:43,770
But I will introduce you to it in a minute. So the first one.

91
00:08:45,180 --> 00:08:51,270
Is magnetic dressed again, Michael, very kindly, almost leading up to this presentation like it was planned.

92
00:08:52,020 --> 00:08:58,170
Introduce the magnetic dress earlier. So the first one just to highlight here, something Michael sticks out is the equals B dressed.

93
00:08:58,560 --> 00:09:05,460
And this is purely a dress that acts radially and it's a result of the fact we have electric and magnetic fields in all systems.

94
00:09:06,090 --> 00:09:10,590
We have that rugby dress that Michael spoke about earlier where we have these

95
00:09:10,590 --> 00:09:14,250
dress that are occurring purely as a result of a magnetic field gradient.

96
00:09:14,610 --> 00:09:16,860
And again, just to remind you back to Michael's talk,

97
00:09:17,250 --> 00:09:22,860
this is occurring because where there is a weaker magnetic field, it's a lot bigger and stronger magnetic field.

98
00:09:22,860 --> 00:09:26,990
The normal weight is a smaller. And again, this means that these dress out of our device.

99
00:09:27,510 --> 00:09:29,760
But there's one additional thing that we need to consider,

100
00:09:29,760 --> 00:09:35,760
and this is very inventively called the curvature dress, and it arises because of curvature.

101
00:09:36,360 --> 00:09:40,799
And so the way you can think about this is the particles are streaming along and they

102
00:09:40,800 --> 00:09:46,650
experience some kind of fictitious centrifugal force outwards purely as a result of curvature.

103
00:09:47,370 --> 00:09:54,150
So these are something that we need to contend with, something we need to be aware of these these little losses in our devices.

104
00:09:54,990 --> 00:09:57,990
The other thing I want to introduce here is an external transport.

105
00:09:58,230 --> 00:10:02,190
So let's take a moment to think about what this actually is.

106
00:10:02,580 --> 00:10:07,620
So on the board here, I have a schematic of the magnetic field, all the talk of that.

107
00:10:08,010 --> 00:10:12,540
So it's quite a crude schematic. So please forgive me slightly, but the general just isn't here.

108
00:10:12,930 --> 00:10:17,130
So we have the magnetic field as a function of along the field line.

109
00:10:17,430 --> 00:10:21,180
So we have regions where the magnetic field is stronger and this is going to correspond

110
00:10:21,210 --> 00:10:24,990
to that good curvature region which is on the inside of our torque network.

111
00:10:24,990 --> 00:10:26,160
Magnetic field is stronger.

112
00:10:26,760 --> 00:10:32,690
We have this region of bad curvature which is corresponding to the outside of our device, whether magnetic field is weaker.

113
00:10:33,980 --> 00:10:36,950
So I'm now going to invite you to consider some different particles.

114
00:10:37,610 --> 00:10:42,650
So let's start off with a particle that has enough energy to sample the entire magnetic field.

115
00:10:43,040 --> 00:10:47,330
I'm kind of representing this as a red line on the field line.

116
00:10:47,660 --> 00:10:51,290
However, I know energy and field are not quite like this as a square.

117
00:10:52,340 --> 00:10:57,050
But generally speaking, imagine my particle has enough energy to sample anything on the field line.

118
00:10:57,830 --> 00:11:01,850
Well, as you can imagine, it's just going to stream all the way around my device.

119
00:11:01,880 --> 00:11:06,260
It's going to go along the field line almost as if it was never impeded.

120
00:11:06,660 --> 00:11:11,250
It'll just keep going. Okay. But obviously not enough.

121
00:11:11,280 --> 00:11:15,140
Nothing's ever that simple. So we might have a different type of particle, right?

122
00:11:15,560 --> 00:11:21,020
Let's now consider a particle that only has enough energy to sample some of the magnetic field.

123
00:11:22,160 --> 00:11:28,730
Well, in this case, it doesn't have enough energy to go to the good character regions or beyond sitting in a steaming round.

124
00:11:29,090 --> 00:11:34,940
Instead, it gets trapped inside, basically a potential well, and it's just going to bounce back and forth, back and forth.

125
00:11:34,970 --> 00:11:39,440
And it's going to be confined to, mostly speaking, the Bad Cabbage region.

126
00:11:40,670 --> 00:11:44,990
So this introduces two different types of particle trajectories.

127
00:11:45,230 --> 00:11:51,800
So the first to do with the particle that can sample the entire magnetic field is called passing particles.

128
00:11:52,040 --> 00:12:01,160
So this flat surface here is a flat surface, as I spoke about earlier, and I've only just taken a cut in my in my tokamak to look at it.

129
00:12:01,520 --> 00:12:05,540
But they would just go all the way around. They have no problem. They can just do the entire thing.

130
00:12:06,230 --> 00:12:13,549
The other ones, the trap particles, which don't have enough energy to sample the whole thing, are going to bounce within this magnetic.

131
00:12:13,550 --> 00:12:16,340
Well, right. I'm going to go back and forth and back and forth.

132
00:12:17,410 --> 00:12:24,970
And if this is kind of when the classical starts coming in as we talk about these orbits and how they relate to collisions.

133
00:12:25,390 --> 00:12:30,280
So collisions tend to enhance these types of stress of these orbits.

134
00:12:30,580 --> 00:12:33,430
And that is, generally speaking, what we call neo classical transport.

135
00:12:33,850 --> 00:12:40,329
It's not entirely important for this talk to really fully understand every detail of the classical transport.

136
00:12:40,330 --> 00:12:43,090
But I really want to give you an overview of what it's talking about.

137
00:12:43,570 --> 00:12:49,479
So these drifts outwards will cause electric fields, these clothes again in the perturbations, etcetera, etcetera.

138
00:12:49,480 --> 00:12:53,650
And we end up finding that there is more transport because of these types of orbits.

139
00:12:54,430 --> 00:12:59,589
And just so you can see what these really look like on the on the left here,

140
00:12:59,590 --> 00:13:06,160
I have the passing particles and you can see the black lines are giving us particle trajectories and they just go all the way round.

141
00:13:06,280 --> 00:13:15,250
They have no problem. They're very happy. On the right here, I have my trap particles and again, the black lines that are showing us the trajectories.

142
00:13:15,700 --> 00:13:20,440
And you see, they kind of bounce back and forth and they never make it to this blue side at the top of that.

143
00:13:20,920 --> 00:13:25,480
So this is just how they would look in our device. So this is passing trap particles.

144
00:13:26,490 --> 00:13:31,320
Okay. So now let's have a look at a schematic for Stellarator magnetic field.

145
00:13:32,040 --> 00:13:36,110
So you can see it looks very similar to a tokamak shape.

146
00:13:36,930 --> 00:13:41,169
It's got this same kind of like oval arching structure, but on top of it,

147
00:13:41,170 --> 00:13:46,530
it has these like little undulations on it, and that's arising because of the geometry of our device.

148
00:13:47,070 --> 00:13:50,370
So now when we think about all types of particles.

149
00:13:50,790 --> 00:13:56,070
Well, the first is all passing particles again. They have no problem making sample the entire magnetic field.

150
00:13:56,430 --> 00:14:05,310
This is fine. But the particles, because of these more complicated geometries, they have the potential to get trapped in these small wells as well.

151
00:14:05,850 --> 00:14:10,799
And so I just kind of want to show you a video quickly of what this kind of looks like.

152
00:14:10,800 --> 00:14:16,680
So this is a device that is not optimal. So the red lines are showing the particle factories.

153
00:14:16,680 --> 00:14:20,219
And currently we're looking at the passing particles.

154
00:14:20,220 --> 00:14:24,000
As you see, they don't really have much of a problem. They stick to our device quite nicely.

155
00:14:24,090 --> 00:14:29,670
They're not really the big concern here, but these trap particles as we're about to see.

156
00:14:31,480 --> 00:14:35,470
They start off on our socks off so that they start off being quite happy.

157
00:14:35,800 --> 00:14:40,660
You'll see they very quickly drift radially out of out of the flux office.

158
00:14:41,050 --> 00:14:44,170
This is problematic. This is leading to losses in our device.

159
00:14:44,170 --> 00:14:47,700
It's poor confinement in general. This is not good. This isn't what we want.

160
00:14:48,460 --> 00:14:54,910
So this is a non optimised thing. You can see this type of problem is now arising and it's obviously going to be more

161
00:14:54,910 --> 00:15:00,490
complicated in a device like a stellarator because the geometry is more complicated.

162
00:15:01,000 --> 00:15:09,940
Okay. So you just quickly summarise neo classical transport is this interaction between collisions and geometry.

163
00:15:10,540 --> 00:15:15,190
And honestly, we have very little control over this part of our equation.

164
00:15:15,310 --> 00:15:21,130
I mean, it's a very crude equation, but we have very little control over I can't tell my particles how to collide.

165
00:15:22,610 --> 00:15:26,540
Alex, you might not disagree with me, but I can't tell my pocket, Clyde.

166
00:15:26,570 --> 00:15:34,220
I can't tell them how to interact. Well, what I do have a lot of control over, especially for stellar eaters, is the geometry.

167
00:15:34,550 --> 00:15:39,470
So that's what I'm really going to say to something. So, you know, I've told you, these are my problems.

168
00:15:39,710 --> 00:15:44,220
What do I demand in order for these things not to actually be a problem within my salary?

169
00:15:44,900 --> 00:15:49,100
So you can maybe just think hard for 5 minutes and you might come up with the idea.

170
00:15:49,610 --> 00:15:53,840
As Michael kind of alluded to earlier, we want these average drifts to go to zero.

171
00:15:54,320 --> 00:15:58,160
That means that as my particle goes around the device, in theory,

172
00:15:58,160 --> 00:16:02,059
it would be following its field line and at some point it might draw out away

173
00:16:02,060 --> 00:16:06,469
from my device and other points on sort of mind in the good character region.

174
00:16:06,470 --> 00:16:11,480
It might drift back onto the field line overall during its entire order.

175
00:16:11,510 --> 00:16:19,550
I want that radio address to go to zero. So basically, I want those things to not leave the magnetic surface on average.

176
00:16:19,580 --> 00:16:25,490
That is the general gist of what I want. You might think that's nice and obvious, which if you think hard, it might happen.

177
00:16:26,210 --> 00:16:33,770
Sadly, I didn't come up with it. And I'm sure you all remember from your physics undergrad that we absolutely love symmetries in physics.

178
00:16:34,130 --> 00:16:35,480
Whenever there's a symmetry,

179
00:16:35,780 --> 00:16:42,590
we normally have some kind of concept of quantity associated with it and the symmetry associated with these types of things.

180
00:16:43,610 --> 00:16:47,360
I'm going to focus on in this talk anyway something called quasi symmetry.

181
00:16:47,870 --> 00:16:54,560
It is a subset of a wider part, a wider group of symmetries, but it's one that I'm really just going to focus on today.

182
00:16:55,130 --> 00:17:01,100
And so for these symmetries, I'm saying that, B, have a continuous symmetry in the sun quarter system.

183
00:17:01,610 --> 00:17:09,260
And if my device is quasi symmetric, it means that these radio dressing are going to average to zero, which is exactly what I want from my device.

184
00:17:10,100 --> 00:17:15,799
So this is going to be something I'm going to focus on. So what do I mean when I say the causes metric?

185
00:17:15,800 --> 00:17:22,010
I told you this complicated thing that, you know, in the sun coordinate system, etc., etc., but what am I actually telling me?

186
00:17:22,460 --> 00:17:29,960
It means that if I take my I stellarator here, which arguably has very complicated geometry compared to our nice simple tokamak,

187
00:17:30,350 --> 00:17:33,920
I take a little cut down one of the sentence. I'm going to cut it about here.

188
00:17:34,370 --> 00:17:39,470
I'm going to unfold it. Unwrap it. I'm going to apply my coordinate transformation.

189
00:17:40,040 --> 00:17:45,230
I'm going to get something that looks like this. So would you believe that these two are the same regardless?

190
00:17:45,240 --> 00:17:54,470
They are. But the idea is, if I have a particle streaming along here, if my device is quasi symmetric, I can make one of these transformations.

191
00:17:54,770 --> 00:18:02,390
And my particle wouldn't know. It's not in an axis symmetric system, so it doesn't know that it's not a tokamak under this coordinate transformation.

192
00:18:03,410 --> 00:18:08,059
And so this is how we get these types of symmetries. You know, there's been a lot of work into it.

193
00:18:08,060 --> 00:18:13,430
And I feel like it's slightly more understood these days that this is what we require for confinement.

194
00:18:14,140 --> 00:18:20,510
And just to show you that this isn't kind of something that people have, scratch their heads over and actually had applications.

195
00:18:20,840 --> 00:18:26,059
These are some quasi symmetric devices that exist, like some of them have actually been built.

196
00:18:26,060 --> 00:18:29,810
So w seven max have been built in Germany is one that I would focus on today.

197
00:18:30,110 --> 00:18:33,260
You know, some of these devices exist in the real world and have gone on.

198
00:18:33,560 --> 00:18:35,850
So this is a real advancement for theory, right?

199
00:18:35,870 --> 00:18:42,830
We made these predictions, made these gluons on our symmetries, and we came up with shapes that are allowed for our stellarator.

200
00:18:44,000 --> 00:18:49,040
Okay. So I've told you the demands. I told you the symmetries that I require for my stellarator.

201
00:18:49,370 --> 00:18:52,550
But I guess why do we want stellarator anyway?

202
00:18:53,060 --> 00:18:57,240
Tokamaks are more favourable, you might ask. Surely them will save for a reason.

203
00:18:57,770 --> 00:19:00,530
And to. To explain to you why I think we won't celebrate.

204
00:19:00,770 --> 00:19:08,660
I think it's important to understand this kind of rat race between tokamaks and celebrate is so a little bit of a history of stellarator.

205
00:19:09,710 --> 00:19:15,470
They actually were conceptualised by Spitzer in 1951, before the Tokamak was.

206
00:19:15,710 --> 00:19:18,760
So people thought accelerators were going to work us.

207
00:19:19,220 --> 00:19:23,209
And there was kind of a funny anecdote that Spitzer was on.

208
00:19:23,210 --> 00:19:30,500
It was skiing. He was on a ski lift. And between the time of the bottom and the top of the ski lift, he had completely outgrew Tokamaks.

209
00:19:30,500 --> 00:19:34,600
He had decided they weren't going to work and it was due to this toroidal current that they have.

210
00:19:34,610 --> 00:19:40,980
He said that too unstable. They're not going to work. And so the initial influx of interest was actually accelerated.

211
00:19:41,160 --> 00:19:47,069
So this is Spitzer himself with the first stellarator, and I really want to highlight the size of this thing.

212
00:19:47,070 --> 00:19:55,740
It would probably fit happily on this table, but following this, there was actually an influx of interest in Stellarator.

213
00:19:56,550 --> 00:20:00,990
And, you know, they didn't want to don that. That's just another story.

214
00:20:02,070 --> 00:20:10,799
But following that introduction, the Soviet Union in 1968 unveiled to the world the Tokamak, and it was just superior.

215
00:20:10,800 --> 00:20:14,370
It had better confinement, had better fusion properties in general.

216
00:20:14,370 --> 00:20:18,540
It was just far better than one of these tabletop devices.

217
00:20:18,540 --> 00:20:22,110
And that's because this was extremely lossy. It didn't have good confinement.

218
00:20:22,410 --> 00:20:27,059
So the Tokamak took over. And you might ask, okay, great.

219
00:20:27,060 --> 00:20:31,080
So they both existed. Why? Why did Tokamaks fly off the shelf?

220
00:20:31,440 --> 00:20:34,860
Why did everyone pick up Tokamak Research and not continue with both?

221
00:20:35,520 --> 00:20:38,550
And I think it's nicely encapsulated by the statement.

222
00:20:39,060 --> 00:20:45,000
I try to avoid hard work when things that complicated, that is often a sign that there is a better way to do it.

223
00:20:46,290 --> 00:20:50,730
And I think if you ask anyone who was with me, I live my life by this philosophy.

224
00:20:51,390 --> 00:20:58,350
So I mean, if we're being completely honest, it stellarator is initially they were neoclassical dominated.

225
00:20:58,350 --> 00:21:05,070
So those programs that I told you about mean classical transport that was a real problem to solve is so that was problematic.

226
00:21:05,340 --> 00:21:09,629
The Soviet Union tokamaks that were unveiled to the world which are superior.

227
00:21:09,630 --> 00:21:12,840
Like I said, they had better confinement, better fusion properties in general.

228
00:21:13,080 --> 00:21:18,570
They looked like they were going to work better. But I think one thing is just that they were objectively simpler.

229
00:21:18,660 --> 00:21:22,379
You know, they looked more attractive to both engineers, to physicists.

230
00:21:22,380 --> 00:21:27,720
We feel like we can approach the more they are just objectively simpler, and I think that makes them more attractive.

231
00:21:29,050 --> 00:21:39,400
But you know, with numerical developments with other people at Oxford, we have been able to really make advancements in stellarator physics.

232
00:21:39,790 --> 00:21:44,800
So we're now actually able to optimise for that new classical transport that I spoke about earlier.

233
00:21:45,190 --> 00:21:49,750
So this isn't just something I'm saying we can do, it's something that has been done.

234
00:21:50,260 --> 00:21:58,510
So there exist accelerator, like I mentioned, in Rice out in Germany, could find signs of an X or w7x for sure.

235
00:21:59,050 --> 00:22:02,680
And this was the conceptual shape that you can see. It's very complicated.

236
00:22:03,100 --> 00:22:08,560
The yellow is the magnetic field or the magnetic flux axis, and these blue,

237
00:22:08,860 --> 00:22:13,510
horrible squiggly things are the coils that are required to produce that surface.

238
00:22:13,750 --> 00:22:22,390
Please note just how horrible they are. It is non-trivial and it is much more complicated than the lovely tokamak which has this beautiful symmetry.

239
00:22:23,320 --> 00:22:26,590
But, you know, these things do exist. This was the concept of it.

240
00:22:27,530 --> 00:22:29,060
This was the construction of it.

241
00:22:29,630 --> 00:22:38,330
Notice not only how big these things are, as I'm sure Michael kind of hammered home earlier, but notice how complicated the shape is.

242
00:22:38,330 --> 00:22:44,149
It's not nice. It's not. And this at each point in our stellarator as we go around, this will change.

243
00:22:44,150 --> 00:22:48,170
Shape is not going to be the same here as it is here or here now.

244
00:22:48,170 --> 00:22:52,550
It changes shape as we go around the device. It looks messy.

245
00:22:52,940 --> 00:22:59,509
But, you know, physicists like continued, they had, you know, was the iron.

246
00:22:59,510 --> 00:23:03,770
Well, they continued forward and we actually have built their designs overnight.

247
00:23:03,770 --> 00:23:11,989
So I say we as if I had anything to do with it, but we built out Windows seven X, it exists, it's now in operation and it is producing results.

248
00:23:11,990 --> 00:23:15,290
We experiment on it. You know, it is a device that now exists.

249
00:23:15,290 --> 00:23:20,450
And this is a real triumph of theory because we have been able to optimise in the classical transport.

250
00:23:21,860 --> 00:23:27,590
Okay. So this is just the mall side by side. Although from inception to completion and now it exists.

251
00:23:28,100 --> 00:23:34,310
Okay, so what I would want to evaluate, you might say, right, how are they any better than Tokamaks?

252
00:23:34,550 --> 00:23:38,060
Well, one thing is that they are driven by these external coils.

253
00:23:38,480 --> 00:23:41,719
That's great, because we can now have continuous operation. You remember that?

254
00:23:41,720 --> 00:23:45,320
I told you that tokamaks are discontinuous and these we don't have that problem

255
00:23:45,320 --> 00:23:48,690
to celebrate because we can completely drive them by these external current.

256
00:23:48,800 --> 00:23:55,940
The external coils which are driven by currents with more improvements in materials, we can use superconducting coils.

257
00:23:56,270 --> 00:24:02,540
This is great for power output. It reduces the power we need to drive these devices.

258
00:24:03,320 --> 00:24:09,050
They don't have this toroidal current or at least we can design them to not have these two little currents.

259
00:24:09,670 --> 00:24:16,910
You know, I told you that this was a real problem for Tokamaks because it draws a whole load of instabilities and some major disruptions.

260
00:24:17,270 --> 00:24:25,239
But Stellarator is we can avoid that completely. That's very. They have an empirically observed high density limit than Tokamaks I told you before,

261
00:24:25,240 --> 00:24:30,670
the Tokamaks have this wall of density that we don't really understand, but we can't go beyond it.

262
00:24:31,360 --> 00:24:35,679
We don't seem to have such a rule in Saturday celebrations that there will be problems that we

263
00:24:35,680 --> 00:24:40,450
don't seem to have such work and currently tokamaks are not producing good enough confinement.

264
00:24:40,480 --> 00:24:47,320
We had arguments earlier that they're not doing great and there is the potential the STELLARATOR could fill that gap.

265
00:24:48,170 --> 00:24:51,700
Well, you know, I'm telling you, kind of a beautiful picture of salary.

266
00:24:51,780 --> 00:24:58,380
Is that going to fix all of our problems? But as you expect to get anything for every single advantage that we have, we are at a disadvantage.

267
00:25:00,520 --> 00:25:04,490
So you know, stellarator, they do have more complicated geometries.

268
00:25:04,600 --> 00:25:09,640
I cannot hide that this is not a super attractive feature in terms of engineering, etc.

269
00:25:10,120 --> 00:25:14,140
They are they are just complicated. That is something we just kind of have to deal with.

270
00:25:14,800 --> 00:25:17,890
They, you know, they don't have these self-generated currents.

271
00:25:17,890 --> 00:25:20,560
And although I've been telling you how problematic these currents are,

272
00:25:20,800 --> 00:25:25,660
there are people within the community who actually think these these currents could be helpful.

273
00:25:26,080 --> 00:25:32,950
They they generate this colloidal carrot of this field that goes round the short way around tokamak, which is self confining.

274
00:25:33,430 --> 00:25:36,850
At the moment, it's only 10 to 20%. That actually helps.

275
00:25:37,210 --> 00:25:40,710
But some people really want to take advantage of this. They want to see if they can push it further.

276
00:25:41,060 --> 00:25:45,820
Now, that could potentially be a problem or at least a disadvantage compared to Tokamaks.

277
00:25:46,850 --> 00:25:49,680
They have these big radios. We don't know if that's going to be a problem.

278
00:25:49,710 --> 00:25:54,620
The accelerators are comparatively feeling kind of new in terms of how much we understand them.

279
00:25:54,770 --> 00:25:57,800
That's still to be worked on that.

280
00:25:58,400 --> 00:26:03,379
And then the final disadvantage accelerators, which up until now are really shoved under the rug.

281
00:26:03,380 --> 00:26:08,970
And I didn't want to mention it because it is a very embarrassing point for those of us who like elevators,

282
00:26:09,210 --> 00:26:12,440
is that we're not actually guaranteed to have these massive accesses.

283
00:26:12,830 --> 00:26:16,940
So, Topamax, I told you we had these kind of like Russian dolls of magnetic surfaces.

284
00:26:17,420 --> 00:26:23,030
We can't guarantee this the stuttering, because we can say that we can we can demand that we have at least one,

285
00:26:23,360 --> 00:26:28,160
maybe a couple, but there's no 100% guarantee that we can have these nice nested fluxes.

286
00:26:28,680 --> 00:26:32,180
You know, we do need to be cautious of that. Okay.

287
00:26:32,870 --> 00:26:38,109
So there seems to be as many disadvantages as I have for advantages over salary.

288
00:26:38,110 --> 00:26:43,610
It isn't. All right. I'm telling you that every thing I say that's good about a salary in the same breath.

289
00:26:43,610 --> 00:26:49,580
I'm telling you the bad things about them. And you may be very well inclined to think that this is the slowest race in humanity.

290
00:26:50,000 --> 00:26:55,250
You know, we first conceptualised harnessing the power of the sun 100 days ago.

291
00:26:55,520 --> 00:27:01,580
That was when it was first thought that maybe we could produce seasonal math to try and produce energy in the same way the Sun does.

292
00:27:01,940 --> 00:27:07,669
The first Stellarator was thought of and built about 70 years ago and 70 years on,

293
00:27:07,670 --> 00:27:13,490
we are probably no farther than the Tokamak is in terms of our advancements when facing similar problems.

294
00:27:13,970 --> 00:27:19,700
And you'd be very right to think this is a very silly race that who is winning is tortoise versus tortoise.

295
00:27:21,050 --> 00:27:25,250
But, you know, whenever we have problems, physicists have said save the day.

296
00:27:27,440 --> 00:27:30,740
And so we like to think of ourselves as superheroes from time to time.

297
00:27:31,370 --> 00:27:33,889
So, you know, I've told you that there are many problems,

298
00:27:33,890 --> 00:27:40,880
that these are ongoing active areas of research in which people are making some serious leaps in different areas.

299
00:27:41,150 --> 00:27:48,740
So I repeat, if you mess with me, not to mention that we are making new strides and understand the physics of these problems and also that

300
00:27:48,770 --> 00:27:53,390
a lot of people dedicate a lot of time to understanding fundamentals to understand the instabilities.

301
00:27:53,740 --> 00:27:58,130
Now we are slowly trying to get to grips with the physics going on here.

302
00:27:58,910 --> 00:28:04,490
We are making major strides in optimising some field configurations.

303
00:28:04,790 --> 00:28:10,700
So what I mean by this, I showed you those magnetic flux surfaces, you know, the purple ones and how they're all twisted.

304
00:28:11,120 --> 00:28:15,439
Well, I can change them in a certain way and move them about and use machine learning

305
00:28:15,440 --> 00:28:19,610
and other clever techniques to try and optimise these for my confinement.

306
00:28:20,480 --> 00:28:26,480
The other thing which I kind of want to maybe emphasise a little bit is those horrible looking coils that we had.

307
00:28:26,480 --> 00:28:31,880
Everyone in this room would be very entitled to say, they look like an engineering nightmare.

308
00:28:32,210 --> 00:28:36,350
How are we ever going to have that being a commercially viable saying?

309
00:28:36,350 --> 00:28:38,540
How are we ever going to have this working?

310
00:28:38,750 --> 00:28:45,380
Well, I'm here to tell you, physicists are saving the day because some people are working very hard to optimise these clothes for error.

311
00:28:45,890 --> 00:28:52,580
What do I mean by that? If you give me your pragmatic successes that I want and you're saying this is the surface,

312
00:28:52,580 --> 00:28:57,020
I want these give me the coils in which I have as much tolerance as possible,

313
00:28:57,380 --> 00:29:01,400
and I'll still get the same magnetic flux of this, all the same confinement properties.

314
00:29:01,850 --> 00:29:07,010
And the reason we really want this is despite us all wanting to think we can do our best job ever,

315
00:29:07,310 --> 00:29:10,730
we can never arrange these coils down to atomic precision.

316
00:29:10,730 --> 00:29:14,900
So we need to allow some kind of error because we can never get them perfect.

317
00:29:15,800 --> 00:29:22,490
And that has been some real strides in this area, meaning that these horrible looking coils are becoming more and more accessible to us.

318
00:29:23,480 --> 00:29:28,250
And again, it'd be a mistake not to mention the fact that there is research into turbulence.

319
00:29:28,610 --> 00:29:32,740
And yes, I, too, if I ever meet, God, will say why turbulence?

320
00:29:32,840 --> 00:29:34,370
As Michael put in that lovely quote,

321
00:29:34,970 --> 00:29:41,810
you might be noticing that the overarching theme here of why we are making so much progress with salary does the maybe we

322
00:29:41,810 --> 00:29:49,430
haven't made in the past is due to numerics really and it's actually something I personally spend a lot of time focusing on.

323
00:29:49,760 --> 00:29:53,480
So a large portion of my Ph.D. has been to develop code.

324
00:29:54,650 --> 00:30:02,120
So you may, if you recall, cast your mind back to Michael's talk where we had those that picture of the Tokamak.

325
00:30:02,120 --> 00:30:05,330
It was beautiful. It was green and blue and speckled.

326
00:30:05,750 --> 00:30:11,320
That was actually done by simulating a single field line and stitching them all together.

327
00:30:11,340 --> 00:30:19,760
We can recreate the whole surface. And the reason we're allowed to do this is because the that the Tokamak has this nice symmetry about it.

328
00:30:20,300 --> 00:30:22,700
Stellarator doesn't have these symmetries.

329
00:30:23,030 --> 00:30:29,570
So the codes that we currently have that have been working for Tokamaks don't necessarily apply to Stellarator as well.

330
00:30:30,020 --> 00:30:37,400
If you think just kind of vaguely each field line on my own, my surface is going to expose a different geometry, right?

331
00:30:37,700 --> 00:30:42,320
So I can't stitch them together. I can't just reproduce the same field line and hope for the best.

332
00:30:43,160 --> 00:30:50,960
So I need to be a bit more clever. And part of my research especially has been producing a code that simulates the entire flux at phase.

333
00:30:51,200 --> 00:30:56,510
So now we can look at fusion simulations of the entire thing and we can see how geometry is going to influence this.

334
00:30:57,050 --> 00:31:01,250
So like I said, we are making big strides in these in these areas.

335
00:31:01,490 --> 00:31:07,790
But largely this is due to massive advancements, which is why it's taken so long for accelerators to catch up.

336
00:31:08,540 --> 00:31:15,460
Okay. So why do I think that Celera is all going to be the future of fusion?

337
00:31:15,570 --> 00:31:23,640
You know, why do I have such strong faith in them? I'm going to draw your attention back to the schematics on the board, which I had up before.

338
00:31:24,030 --> 00:31:28,890
And these, again, just remind you of the magnetic field strength of a tokamak and the stellarator.

339
00:31:29,460 --> 00:31:34,040
And I spoke before about these problematic trapped particles.

340
00:31:34,110 --> 00:31:39,210
I said they are they could be worse in a stellarator because we have these more complicated geometries.

341
00:31:39,740 --> 00:31:46,860
Now, it's just hard to understand that really this actually these complicated geometries could be accelerated to make a string.

342
00:31:47,900 --> 00:31:52,200
And to demonstrate why this could be the case, I'm going to use one example of these trap particles.

343
00:31:52,730 --> 00:31:56,340
So I told you that we could get trapped here, but we could also get trapped here.

344
00:31:56,360 --> 00:32:00,680
Problematic that some people in the audience may have already asked themselves,

345
00:32:01,100 --> 00:32:10,310
why couldn't we trap up here instead where we have like in the schematic good kind of where turbulence and instabilities all stabilise.

346
00:32:10,820 --> 00:32:14,990
And this is exactly why I think Stellarator could be the future or at least could

347
00:32:14,990 --> 00:32:20,330
give us a beacon of light forward in fusion research is because some people who

348
00:32:20,330 --> 00:32:24,860
are extremely clever and probably cleverer than I am could design a shape in such

349
00:32:24,860 --> 00:32:28,759
a way to take advantage of the fact that we could cause all of our trapping.

350
00:32:28,760 --> 00:32:33,520
For example, in a good capital region, we could design autocrats as right.

351
00:32:33,560 --> 00:32:38,320
We could design accelerators in order to be favourable for fusion and favourable for life.

352
00:32:38,350 --> 00:32:44,450
So what I want out of my device. So with all these opportunities, it gives scientists, physicists,

353
00:32:44,490 --> 00:32:52,100
people who are researching this space to be more intelligent with that design and to really try and determine what they want.

354
00:32:52,450 --> 00:32:58,540
And so this is exactly why I think celebrate is all they all could at least be the teacher of magnetic confinement region.

355
00:32:58,880 --> 00:33:01,370
It's because this big a parameter space that they offer,

356
00:33:01,640 --> 00:33:07,640
despite it being horribly complicated and somewhat scary, it offers more opportunities to control.

357
00:33:08,000 --> 00:33:13,670
And it means that with more numerical advancements, we could potentially adjust it in a way that we want.

358
00:33:14,450 --> 00:33:18,430
And this isn't just me thinking, Oh, this could be great. This actually has been great.

359
00:33:18,440 --> 00:33:23,920
So to demonstrate that this is not just entirely me having wishful thinking, I want to show you another video.

360
00:33:24,970 --> 00:33:32,070
So before I showed you that video of a non optimised device, this is not one that has been optimised for that main classical transport.

361
00:33:32,320 --> 00:33:35,050
So achievement of size of next the device in Germany.

362
00:33:35,500 --> 00:33:40,050
So again, the red lines are showing the the particle directories and you can see the parts and particles.

363
00:33:40,060 --> 00:33:45,250
Again, no problems. We don't have any issues with these guys that they're quite happy on a flat surface.

364
00:33:46,380 --> 00:33:51,330
And now we're going to look at the trap particles, which before, if you remember, caused us problems.

365
00:33:51,330 --> 00:33:55,650
They radially drifted away from our device and that's extremely bad behaviour.

366
00:33:56,040 --> 00:34:02,610
So again, these are the the Trump articles, they start on the successes and you'll see as they come back around they have drifted away.

367
00:34:03,020 --> 00:34:08,190
But by the time they closed back into the same location, the radial drift has averaged to zero.

368
00:34:08,670 --> 00:34:12,840
And so this is an example of optimisation for these types of problems.

369
00:34:13,110 --> 00:34:16,560
And this actually, like I said, has been built and it has gone to operation.

370
00:34:16,930 --> 00:34:22,380
And so this shows that theory can sometimes prevail and we can make these types of strides forward.

371
00:34:23,480 --> 00:34:29,900
Okay. So, you know, just to then whet your appetite for the future accelerators and what this could mean.

372
00:34:30,350 --> 00:34:36,230
These are some wacky designs that people have come up with that potentially have very good confinement properties.

373
00:34:36,590 --> 00:34:41,240
So I wouldn't claim to to tell you anything about these confinement properties,

374
00:34:41,540 --> 00:34:44,900
but you can see they're getting a bit more crazy like this one here is wild.

375
00:34:46,400 --> 00:34:52,790
So this just shows that with more intelligence, with more numerical advances, we are actually making some strides.

376
00:34:53,090 --> 00:34:56,230
These could be the future devices that power our national grid.

377
00:34:56,250 --> 00:35:03,230
So we know. But the idea here is, I think the fact that we have more geometry, we have more control, we have more opportunities.

378
00:35:03,650 --> 00:35:08,140
You might think initially this is terrifying, but this actually could be one of the best things about Cellarity.

379
00:35:09,110 --> 00:35:16,610
And so to almost as I told you before, that we had this really slow race of tortoise races, quarters of stellarator and tokamak.

380
00:35:17,560 --> 00:35:22,810
What I'd actually like you to think it's more of a quote from Paris, where, you know, the Tokamak flew forward.

381
00:35:22,810 --> 00:35:27,130
It started off great, had a lot of advances at the beginning, but it hit this wall.

382
00:35:27,340 --> 00:35:30,610
It now is now limited, partly because of its geometry.

383
00:35:31,120 --> 00:35:34,540
You know, the Stellarator has been chugging along in the background.

384
00:35:34,750 --> 00:35:39,250
We've made some advancements. It's coming further forward. It's managed to reach up to the have.

385
00:35:39,250 --> 00:35:42,340
You never know. In the future it may completely take over.

386
00:35:42,790 --> 00:35:45,129
So with that, I want to thank you for your time,

387
00:35:45,130 --> 00:35:53,770
and I hope I've at least tried to convince you that Stellarator could be the twisted tokamak of the future.