1
00:00:00,060 --> 00:00:07,469
Thank you very much, John. It's delightful to be here. I was an undergraduate at Oxford and I was sort of a weakness when I come into this lecture

2
00:00:07,470 --> 00:00:15,090
theatres and some wonderful memories of my Aitchison lectures on quantum mechanics.

3
00:00:15,090 --> 00:00:20,310
If any of you had those. They were spectacular. Lots of other wonderful times.

4
00:00:21,390 --> 00:00:26,969
I mean, what? We had two feet, plasma started, plasma unleashed.

5
00:00:26,970 --> 00:00:30,590
And now I'm supposed to talk about plasma obtained.

6
00:00:31,740 --> 00:00:40,229
I'll talk a little bit about where we are, what is the current status, where we expect to be, and some of the exciting things that are happening.

7
00:00:40,230 --> 00:00:45,450
Because I'm very excited about the relationship with Oxford because as you can see,

8
00:00:45,450 --> 00:00:49,500
we have some very bright young theoreticians who are making a big dent,

9
00:00:49,830 --> 00:00:57,870
not just in explaining what's happening, but in proposing ways in which we might be able to reduce the cost and scale of fusion.

10
00:00:58,230 --> 00:01:02,010
And I think if you ask me, are we going to have fusion power?

11
00:01:03,180 --> 00:01:06,570
I'll be bold here, which is that I think fusion is possible.

12
00:01:06,660 --> 00:01:15,180
I think where we are with Jet and where we will go with ITER will demonstrate that you can actually have a self-sustained fusion burn in a device.

13
00:01:15,570 --> 00:01:20,310
What I don't know is whether we can do it at a cost that you want to pay for your electricity,

14
00:01:20,700 --> 00:01:27,540
and that requires some innovation and some steps forward that require both innovation in technology,

15
00:01:27,750 --> 00:01:31,590
innovation in engineering, and innovation in the science.

16
00:01:32,310 --> 00:01:36,330
So we talk about that and how it relates to what we've already had this morning.

17
00:01:36,510 --> 00:01:40,470
I won't give you my normal hard sell on why we should do fusion energy.

18
00:01:40,650 --> 00:01:45,300
Most of you probably got it 30 years ago. You also will know the joke.

19
00:01:47,160 --> 00:01:50,730
I don't think I have to repeat the joke because you know it so well.

20
00:01:51,540 --> 00:01:58,680
So anyway, so my basic thing is I'll talk about what the current plan for the first electricity in fusion is

21
00:01:59,670 --> 00:02:07,140
that electricity will be expensive electricity and it will not demonstrate the cost effectiveness.

22
00:02:07,440 --> 00:02:11,700
I'll talk about what Aether will do, which is the machine we're building in southern France.

23
00:02:11,700 --> 00:02:17,099
You saw a picture of it. It's very large and some of the challenges there.

24
00:02:17,100 --> 00:02:24,120
And then I'll talk about what we're trying to do to make fusion perhaps cheaper, faster, more effective.

25
00:02:25,290 --> 00:02:34,320
One of the problems that you have in developing a technology in which every step is going to cost you $10 billion,

26
00:02:34,950 --> 00:02:40,290
euros, pounds, whatever is, how many steps can you actually make?

27
00:02:40,440 --> 00:02:45,360
A lot of innovation happens in that Edison ion way in which you build something.

28
00:02:45,360 --> 00:02:50,910
It's not quite right. You tweak it, you build something else, etc. And you can't do that on the 10 billion scale.

29
00:02:51,510 --> 00:02:57,810
So a lot has to be driven by theory and very precise understanding of what the next step will actually be.

30
00:02:58,230 --> 00:03:02,790
Some. The 30 year rule about fusion.

31
00:03:04,140 --> 00:03:09,450
I can make it worse than that. We can go back to what it really all started.

32
00:03:09,990 --> 00:03:16,980
I mean, we can go back all the way to Eddington in 1920 saying that the fusion must be happening in the middle of the sun.

33
00:03:16,990 --> 00:03:23,790
But we've come a bit forward from that. When people started to think about whether you could actually use fusion as a power source.

34
00:03:24,090 --> 00:03:25,980
And there's a nice piece of history here.

35
00:03:26,940 --> 00:03:31,499
So during the the bomb project in the Second World War, of course, they were thinking a little bit about fusion.

36
00:03:31,500 --> 00:03:37,410
And famously, Edward Teller, of course, was not doing his job. He was working on the atom bomb, he was working on the H-bomb.

37
00:03:38,220 --> 00:03:40,170
And people kept getting frustrated with him.

38
00:03:40,440 --> 00:03:48,900
But part of that was thinking about they started thinking about whether you could control the fusion reaction and how you can find it.

39
00:03:49,290 --> 00:03:53,340
And when I was I was an undergraduate here, and then I went and did my Ph.D. in Princeton.

40
00:03:53,700 --> 00:04:04,800
And there was always stories around Princeton that Fermi had given a lecture on how to make a fusion reactor with magnetic confinement,

41
00:04:05,730 --> 00:04:14,460
and he gave it at a famous dream of happy conference that they had at Los Alamos in 1946.

42
00:04:14,610 --> 00:04:18,210
It's a classified conference. At this conference,

43
00:04:18,540 --> 00:04:22,350
they decide all the big names who were going back to their universities decided

44
00:04:22,950 --> 00:04:25,979
we should have a conference and everybody should give a talk about what they

45
00:04:25,980 --> 00:04:29,549
think is going to happen in the future and the issues that the US government

46
00:04:29,550 --> 00:04:34,560
should look at scientifically and what kind of new ideas need to be worked on.

47
00:04:34,830 --> 00:04:39,210
And there should be no notes from this conference because a lot of it was highly classified anyway.

48
00:04:39,630 --> 00:04:46,650
And we'll just talk now. Indeed, there are no notes from this conference because if you go to Los Alamos,

49
00:04:46,650 --> 00:04:51,299
some friends of mine who have access to all the classified files in in Los

50
00:04:51,300 --> 00:04:55,710
Alamos say that there is no record of what was talked about at this conference.

51
00:04:55,980 --> 00:05:03,240
But in fact, there is a record that record was held by the British because in 1946,

52
00:05:03,240 --> 00:05:08,400
we knew that the Americans were not going to cut us out of access to secret information.

53
00:05:09,240 --> 00:05:19,260
And we wanted to gather as much as we possibly could at that time and archive it and make sure that we had it under control.

54
00:05:19,260 --> 00:05:27,300
And this was the possibly the last time that we had access to the inner thinking of what was going on at Los Alamos.

55
00:05:27,750 --> 00:05:33,510
So there was a guy, I think his name was Philip Moon. He was a professor at Berkeley and for many years in nuclear physics.

56
00:05:33,840 --> 00:05:39,120
And he was at this conference and he was at Fermi's lecture.

57
00:05:39,780 --> 00:05:45,510
And Fermi talked about thermonuclear reactions and what you could do with thermonuclear reactions.

58
00:05:47,040 --> 00:05:51,990
Obviously, one of the things you can do with thermonuclear reactions in enhanced explosive power of a bomb.

59
00:05:52,230 --> 00:05:59,879
He talked about that. He but he also talked about how you could confine thermonuclear fuel with a

60
00:05:59,880 --> 00:06:05,730
magnetic field and a simple a simple toroidal field would not be sufficient.

61
00:06:07,800 --> 00:06:16,440
When I went to Princeton, this was known as Fermi's Theorem, and nobody knew, nobody had actually ever seen where Fermi actually wrote this down,

62
00:06:16,710 --> 00:06:22,740
that you can't just take a solenoid, bend it into a into toroidal solenoid, put a plasma in it and it'll be stable.

63
00:06:22,750 --> 00:06:26,550
It doesn't work. The drift, the particles just drift right out of it.

64
00:06:26,910 --> 00:06:33,750
And Fermi proved it in this lecture. And the only notes we have of it were this classified report that Moon took

65
00:06:33,750 --> 00:06:38,340
during the day and handed to somebody from some organisation at Night-Time,

66
00:06:38,340 --> 00:06:46,500
and they took the notes away and typed them up. And you can see it was for many years a secret report and it was declassified.

67
00:06:46,860 --> 00:06:56,790
And it was this was in a cockcroft archive at Churchill College and Dan Cleary of Science magazine.

68
00:06:57,720 --> 00:07:02,459
I told him about this and about how we couldn't ever find any record of this.

69
00:07:02,460 --> 00:07:05,520
And he was digging through Cockcroft notes.

70
00:07:05,520 --> 00:07:12,390
He found this. It's in Cockcroft notes and it's also in Jp Thomson's notes from those times.

71
00:07:12,570 --> 00:07:16,110
So this is Fermi talking about it in 1946,

72
00:07:16,440 --> 00:07:25,020
but actually by 1946 somebody at Oxford was thinking about it because Peter Tournament was at Oxford in those days,

73
00:07:25,290 --> 00:07:30,210
young Australian, he was Australian, not Kiwi.

74
00:07:31,140 --> 00:07:40,200
And he, he was thinking about whether you could do deuterium, deuterium, fusion in discharges and what you would need to do.

75
00:07:40,770 --> 00:07:46,620
Tournament was a thought with his hands a type of physicist.

76
00:07:46,800 --> 00:07:53,160
They can be very productive. I'm not being theoretically snobbish at this point, but and he hadn't done many calculations,

77
00:07:53,160 --> 00:07:59,190
and so he was probably a little ambitious about what he actually thought would happen by passing a current through.

78
00:07:59,710 --> 00:08:03,250
Through a glass tube and etc. at that time.

79
00:08:03,490 --> 00:08:07,840
But he gradually worked his way towards an understanding of what you need,

80
00:08:07,840 --> 00:08:17,530
and it was probably really the first actual experiment on magnetically confined fusion where in the Clarendon Laboratory in 46, 47.

81
00:08:18,280 --> 00:08:27,310
And then there was a similar group working at Imperial College where and cousins working under G.P. Thompson at that time.

82
00:08:28,060 --> 00:08:30,250
This was the beginning of magnetic confinement.

83
00:08:30,520 --> 00:08:37,840
And at that time, people made these simple estimates, as Alex showed, that if you didn't have any turbulence,

84
00:08:38,560 --> 00:08:43,240
the confinement time of something that wasn't very big would be sufficient.

85
00:08:43,660 --> 00:08:50,469
Confinement time, by the way, is the time basically it takes to go from the middle to the outside of the device.

86
00:08:50,470 --> 00:08:55,940
And I'll use it all the time and we always give it the symbol title ee.

87
00:08:57,760 --> 00:09:07,300
If you took your magnetically confined plasma and you turned off all external sources of heat and any fusion heating, and you let it cool down.

88
00:09:07,540 --> 00:09:16,749
This is the time it would take to lose half of its temperature, but it's also the time it takes heat to diffuse from the middle to the outside on jet.

89
00:09:16,750 --> 00:09:20,110
That's about one second on a heater.

90
00:09:20,110 --> 00:09:30,370
It has to be about three, 3 to 4 seconds. And a reactor about 4 seconds is a typical confinement time with conditions that we know at this time.

91
00:09:31,870 --> 00:09:38,799
It was understood right at the beginning, and people were very optimistic about Fusion in the late forties because they did

92
00:09:38,800 --> 00:09:43,540
simple calculations like the ones not not not that Alex's talk was simple or anything.

93
00:09:44,170 --> 00:09:51,670
It was. But the simple calculation of the diffusion across a device that was done by people.

94
00:09:51,670 --> 00:10:00,249
And here's Lyman Spitzer, who in the around about 1950 started to think about magnetic confinement at that time

95
00:10:00,250 --> 00:10:05,620
because there was an announcement that somebody in Argentina had actually done fusion.

96
00:10:05,800 --> 00:10:10,950
It was in the front page of the New York Times, and he was a bit surprised by that announcement.

97
00:10:10,960 --> 00:10:19,000
And he started to do calculations and he came up with the first design of that three dimensional object, the Stellarator.

98
00:10:19,570 --> 00:10:23,890
Here it is. It's only the dome. The tube is a figure of eight.

99
00:10:24,100 --> 00:10:34,090
The diameter of the tube is 100 centimetres, so a metre across the actual plasma had a radius of about 40 centimetres inside that tube.

100
00:10:34,930 --> 00:10:38,770
Right. 40 centimetres would be much smaller than jet. Right.

101
00:10:38,780 --> 00:10:46,240
And he'd done this wasn't just sort of without calculation, done detailed calculations of what the,

102
00:10:46,600 --> 00:10:49,960
what we call classical diffusion in that non turbulent diffusion.

103
00:10:50,230 --> 00:10:53,680
And he'd come up with this is the size of the device. It was very long,

104
00:10:55,030 --> 00:11:04,299
this is 50 metres around etc. and it was twisted into a figure of eight because he wanted to get round Fermi's theory that

105
00:11:04,300 --> 00:11:13,300
a simple toroidal device with just toroidal field going around this in one direction would not confine the particles.

106
00:11:13,540 --> 00:11:16,209
What he did here was cancel the drifts of the particles,

107
00:11:16,210 --> 00:11:20,260
so the particles followed the field line and drifted upwards on one half of the figure of eight,

108
00:11:20,260 --> 00:11:25,300
and then they drifted downwards on the other half of the figure of eight. And those two things cancel each other.

109
00:11:25,630 --> 00:11:29,680
And this was Spitzer's design in 1951.

110
00:11:31,150 --> 00:11:41,860
By this time, Atonement had built several pinches discharges here, and everything had moved off to Harwell by that time up the street.

111
00:11:43,060 --> 00:11:47,140
Oh, by the way, for people who don't know, that thing in Spitzer's hand is a slide rule.

112
00:11:50,040 --> 00:12:01,020
And these are journals. And in Harwell in the mid fifties we built this device, which was called Z2,

113
00:12:01,620 --> 00:12:06,360
which was the biggest toroidal pitch and for many years was the best fusion device out.

114
00:12:06,690 --> 00:12:13,910
And it's very it's a, as the leader of a large organisation and I have to do a lot of PR.

115
00:12:15,540 --> 00:12:26,850
It's a it's a solitary lesson to read about Z2 because at some point due to electric fields in the joins probably of the vacuum vessel,

116
00:12:27,240 --> 00:12:36,540
they were getting high energy ions generated inside this and those high energy ions were producing some fusion neutrons.

117
00:12:37,800 --> 00:12:44,430
So Cockcroft got on television and told the world that we would make energy that was too cheap to metre.

118
00:12:45,090 --> 00:12:48,719
And then they discovered these weren't thermonuclear neutrons at all,

119
00:12:48,720 --> 00:12:53,280
and they were just from straight field acceleration and there was no way that was scaling.

120
00:12:53,670 --> 00:12:56,879
And I think it was probably the worst moment of Cockcroft life,

121
00:12:56,880 --> 00:13:04,140
and it certainly set back the fusion cause because in fact it was a very good machine and we could have learned probably even more from it.

122
00:13:04,560 --> 00:13:11,970
But my lab was commissioned before we realised they were not thermonuclear earth neutrons.

123
00:13:12,720 --> 00:13:18,930
So good things come up. So once the first electricity,

124
00:13:20,640 --> 00:13:29,430
the jet results of the late nineties where we almost got to energy breakeven and the conditions

125
00:13:29,430 --> 00:13:35,880
in the middle of the device when we when we got 16 megawatts out of it but 230 million degrees,

126
00:13:36,330 --> 00:13:42,450
those conditions are almost identical to what you're going to need on ITER to actually generate a full fusion burn.

127
00:13:43,140 --> 00:13:49,110
So we actually produced all the conditions necessary for fusion in that device at that time.

128
00:13:50,370 --> 00:14:00,980
It's what we didn't do was sustain the reaction by the fusion process itself that has made us feel that we're on the right track.

129
00:14:01,680 --> 00:14:10,680
And this last two years we have produced a fusion roadmap that the European Commission has said will drive forward the programme at this point.

130
00:14:10,950 --> 00:14:17,990
And the point of that fusion roadmap is to deliver a demonstration reactor in the 2014.

131
00:14:18,960 --> 00:14:24,780
This would be actual produce some electricity at probably quite phenomenal cost to the electricity.

132
00:14:25,020 --> 00:14:35,729
And it's a rather large and I hate to say this clumsy machine in that it uses only at this point existing technology.

133
00:14:35,730 --> 00:14:42,840
And there are some serious issues with this machine, but the idea is to put a mission in front of us to actually deliver some electricity.

134
00:14:43,020 --> 00:14:48,690
And then I won't have to stand in front of people who tell me it's 30 years away and always will be.

135
00:14:49,860 --> 00:14:54,560
So this is the plan 2014 to get today.

136
00:14:54,700 --> 00:15:01,980
This machine is very large. In fact, ITER is six metres from there to the middle of the plasma.

137
00:15:02,280 --> 00:15:06,220
This device may be as large as nine metres between there and that.

138
00:15:06,450 --> 00:15:14,340
These are superconducting magnets which we hope will produce up to about six Tesla in the middle of the device.

139
00:15:15,750 --> 00:15:21,150
That's a little less than double of what a jet currently has.

140
00:15:22,410 --> 00:15:30,990
One of the key issues about this device will be all the power loadings on the walls and on the device.

141
00:15:31,230 --> 00:15:37,980
The neutrons crossing the edge of this device will give something like two megawatts per square metre of neutron flux in the wall.

142
00:15:38,340 --> 00:15:47,129
That means every atom inside the steel wall will be moved ten times to 20 times

143
00:15:47,130 --> 00:15:51,450
a year from its equilibrium position in the lattice of the steel in the wall.

144
00:15:56,250 --> 00:16:01,920
The first self-sustained fusion burn. This is a bit more certain this would be eater.

145
00:16:03,210 --> 00:16:08,430
Eater. I'd love not to talk about Eater today.

146
00:16:09,450 --> 00:16:13,559
Eater is suffering from some real management issues at the moment.

147
00:16:13,560 --> 00:16:20,160
It's a collaboration between seven large international partners.

148
00:16:20,190 --> 00:16:25,320
45% is paid by Europe. Money coming directly from the European Commission.

149
00:16:25,560 --> 00:16:31,230
Actually, 9% comes from France and the rest of the 45% from Europe.

150
00:16:31,440 --> 00:16:35,100
And you can see if you know your flags, you can see what the other partners are.

151
00:16:36,030 --> 00:16:41,189
This machine is designed to get off to almost self-sustained fusion.

152
00:16:41,190 --> 00:16:46,450
But and one of the questions is, will it get completely self-sustained fusion done or not?

153
00:16:46,770 --> 00:16:56,640
Describe what that is in a moment. It is absolutely at the edge of what we could do in terms of technology at the moment.

154
00:16:56,940 --> 00:17:04,319
These superconducting magnets, which you can see wrapped around the yellow hole, which is where the plasma goes, these are niobium,

155
00:17:04,320 --> 00:17:15,660
tin magnets at four degrees Kelvin producing all magnet, a field of 13.5 Tesla and in the middle of the plasma, 5.2 Tesla.

156
00:17:15,930 --> 00:17:23,760
Because of the overall falloff of the magnetic field, that is pretty much as strong as you can get out of niobium tin at this point.

157
00:17:24,180 --> 00:17:27,360
And so it's right at the edge of technological capability.

158
00:17:27,360 --> 00:17:33,390
The plasma car will be 15 milligrams of current and jet goes up.

159
00:17:34,290 --> 00:17:38,129
It can go above four, it can even go above five mg.

160
00:17:38,130 --> 00:17:41,700
But 15 mega EPS is a considerable extension from that.

161
00:17:41,910 --> 00:17:48,630
The energy stored in the magnetic system of E2 is about 40 gigajoule of energy.

162
00:17:48,840 --> 00:17:51,900
That's equivalent to about ten tonnes of TNT.

163
00:17:53,910 --> 00:17:59,340
It's power amplification. That's the amount of power going out compared to the amount going in.

164
00:17:59,580 --> 00:18:05,850
The baseline target is ten. But I'm going to show you simulations that show it go to full ignition.

165
00:18:05,850 --> 00:18:09,639
And a second. The cost is very useful.

166
00:18:09,640 --> 00:18:19,080
The use of the greater than sine is somewhere you learn things you see as an undergraduate physicist that that's useful for life afterwards.

167
00:18:19,350 --> 00:18:21,690
The official start date is 2020.

168
00:18:21,720 --> 00:18:28,560
You can hear it from me and I'm prepared to stick my neck out on this one, that it won't be 2020, it will be more like 2023.

169
00:18:28,860 --> 00:18:35,129
It's very late. It's very hard to move projects like this forward and to manage projects like this.

170
00:18:35,130 --> 00:18:37,800
And I think as a community we need to up our game.

171
00:18:40,830 --> 00:18:48,900
Just for those of you who don't understand roughly what happens in a discharge here, this is a terrible slide, isn't it?

172
00:18:50,820 --> 00:18:52,560
You have a tube of plasma,

173
00:18:53,910 --> 00:19:03,060
and what you do in that tube of plasma is you add it to heat that tube of plasma as you energise these coils that make it to a little solenoid.

174
00:19:03,300 --> 00:19:07,080
So you make a field that goes around the toroidal direction.

175
00:19:07,080 --> 00:19:13,050
So this, this field around like this, you do that with superconducting magnet.

176
00:19:13,060 --> 00:19:18,600
So it's on all the time. So the beginning of a shot, actually, there is no plasma there or anything.

177
00:19:18,600 --> 00:19:23,370
You puff in your fuels, deuterium and tritium into the hole in the middle.

178
00:19:23,940 --> 00:19:31,889
The field's already there. Then what you do is you swing the amount of magnetic flux going down through the middle.

179
00:19:31,890 --> 00:19:35,270
This is done on a transformer winding in in each.

180
00:19:35,610 --> 00:19:41,130
It'll just be coils going down through the middle that change the magnetic flux through that loop.

181
00:19:41,640 --> 00:19:44,580
That induces, of course, an EMF around that loop.

182
00:19:45,180 --> 00:19:53,160
And that EMF is enough to break down the gas of deuterium and tritium and what you get is a current flowing around that loop.

183
00:19:54,390 --> 00:19:57,690
Now the great thing about that current is like currents attract.

184
00:19:58,230 --> 00:20:05,970
And so we think of it as a watch of current, the basic, basic force that holds all of this together, the pinched force, as it's called,

185
00:20:06,270 --> 00:20:13,410
is the fact that the like the current going around in the toroidal direction attracts itself, pulls itself in off the wall.

186
00:20:13,560 --> 00:20:18,330
And it's the self attraction of currents that pulls the plasma away from the wall

187
00:20:18,540 --> 00:20:24,960
and supports the pressure gradient because the middle of this plasma is hot,

188
00:20:25,110 --> 00:20:28,799
the edge is cold, and it's pushing out. Right.

189
00:20:28,800 --> 00:20:32,730
And you've got to have a force that's pulling in to hold it together.

190
00:20:32,910 --> 00:20:36,090
And that's the fundamental principle of how the Tokamak works.

191
00:20:36,300 --> 00:20:43,770
Once you can get it into that state, you can start to heat this plasma up to 70 clear temperatures.

192
00:20:43,770 --> 00:20:48,959
And I'll show you how that works in a moment with this is just a picture of the

193
00:20:48,960 --> 00:20:53,970
magnetic configuration that that current will produce because you've got a field going.

194
00:20:54,010 --> 00:20:59,139
Around the long way and the current in the plasma produces a twist in that field.

195
00:20:59,140 --> 00:21:02,200
And Felix showed that the field lines look sort of like that.

196
00:21:02,200 --> 00:21:05,890
These are terrible slides and out of focus. So I'm sorry about that.

197
00:21:07,210 --> 00:21:16,780
So what that field does to the particle orbits inside the plasma is this is a picture, for instance, of the particle orbits.

198
00:21:16,780 --> 00:21:19,790
And this was what Fermi actually was doing.

199
00:21:19,810 --> 00:21:26,950
He was showing you what these particle orbits would look like inside an actually symmetric machine and calculating their drift motion.

200
00:21:27,520 --> 00:21:30,980
And a drift motion here is a particle such as, you know,

201
00:21:31,540 --> 00:21:36,099
just something a maybe a first year Oxford undergraduate could do would integrate the

202
00:21:36,100 --> 00:21:42,069
equations of motion for the particle in these fields and then plot it on a nice picture.

203
00:21:42,070 --> 00:21:50,320
Here you can see this is actually the motion of of an energetic alpha particle in each like field here,

204
00:21:50,980 --> 00:21:56,559
because it's got a nice big fat spiral here and it's bouncing backwards and forwards inside here,

205
00:21:56,560 --> 00:22:01,840
but remaining confined inside the machine so that those are the orbits.

206
00:22:01,840 --> 00:22:08,050
Once you get now as you heat the plasma up and the temperature goes up, it collides less and less.

207
00:22:08,380 --> 00:22:17,410
And a typical main free path in jet might be several kilometres before it actually collides with another particle.

208
00:22:17,410 --> 00:22:21,340
So it will go round the machine many, many times before it actually collides.

209
00:22:24,800 --> 00:22:29,510
So inside his his that pink bit of the plasma again inside.

210
00:22:29,930 --> 00:22:38,240
Peter, you've got a field from the magnetic field of the coils here of 5.2 Tesla.

211
00:22:38,540 --> 00:22:41,810
That's equivalent to a magnetic pressure of about 100 atmospheres.

212
00:22:42,410 --> 00:22:45,799
So force is really quite large in the coils here.

213
00:22:45,800 --> 00:22:54,860
The central temperature of the plasma here in a indicative e to shot is in about 20 kilowatts.

214
00:22:55,280 --> 00:23:02,239
20 kilowatts is about 200 million degrees. And the plasma pressure is about seven atmospheres there.

215
00:23:02,240 --> 00:23:10,190
So the hot bit of the plasma is pushing outwards. And part of the magnetic pressure is pushing inwards and holding that in place.

216
00:23:10,490 --> 00:23:17,629
So that's fundamentally what is happening inside there and the reaction you want to do.

217
00:23:17,630 --> 00:23:24,370
I'm going to go over this again, even though, Felix, did the reaction you want to do this is a little gift from God.

218
00:23:24,380 --> 00:23:37,280
I used to give a lecture when I was teaching at UCLA about fusion being tantalisingly close, but sometimes not close enough.

219
00:23:37,550 --> 00:23:45,320
And one of the tantalising things is that this reaction has a spectacularly large cross-section deuterium and tritium.

220
00:23:45,740 --> 00:23:58,750
It's 100 times the next largest fusion cross-section, which is a D helium three tritium plus deuterium coming together.

221
00:23:58,760 --> 00:24:03,500
Of course, you've got to get them together against the Coulomb repulsion, so they have to ram together hard enough.

222
00:24:03,890 --> 00:24:09,560
And for an instance, they make helium five in the spin free half state.

223
00:24:10,160 --> 00:24:12,620
J Three half state. And it's a resonant interaction.

224
00:24:12,620 --> 00:24:20,980
And that's why you have such a large cross-section and that helium five then splits up into helium four natural helium and a neutron.

225
00:24:20,990 --> 00:24:29,420
And the point I'm making here is this is charged and this is not so in your fusion reactor you have two different particles,

226
00:24:29,420 --> 00:24:38,389
4/5 of the energy because of momentum and energy conservation comes out of the neutron and one fifth comes out as the alpha particle,

227
00:24:38,390 --> 00:24:46,010
the helium nucleus. Right now, the helium nucleus is charged, so it stays inside the magnetic bottle.

228
00:24:46,250 --> 00:24:48,350
It's confined by the magnetic field.

229
00:24:48,680 --> 00:24:57,110
The neutron, on the other hand, doesn't see the magnetic field, crosses the magnetic field and into the wall of your device.

230
00:24:59,120 --> 00:25:03,889
Now, the problem with this reaction, I mean, it's a gift to have such a large cross-section.

231
00:25:03,890 --> 00:25:08,690
But the problem is this tritium, a really remarkable nucleus.

232
00:25:08,690 --> 00:25:12,799
Really, when you think about it, it means you've got one proton and two neutrons.

233
00:25:12,800 --> 00:25:18,170
It's sort of very asymmetric. Tritium doesn't exist in nature.

234
00:25:18,170 --> 00:25:22,750
It has a half life of 12.4 years, so you just don't find it.

235
00:25:22,940 --> 00:25:27,380
You've got to make it. And the way to make it is to bombard lithium.

236
00:25:29,090 --> 00:25:34,700
If a neutron hits lithium six, you will get helium and tritium out.

237
00:25:35,000 --> 00:25:40,220
And you can take that tritium and you can put it back in and do your fusion reaction with it.

238
00:25:40,520 --> 00:25:47,540
So tritium actually isn't a fuel. Tritium is only an intermediary because your fusion reactor has to do two reactions.

239
00:25:47,540 --> 00:25:52,280
It has to do this one and that one, and tritium is produced here, consumed here.

240
00:25:52,520 --> 00:25:56,900
And so it's not actually a fuel. The fuels are lithium and deuterium.

241
00:25:57,680 --> 00:26:02,960
There is 60 billion years worth of of deuterium, fuel.

242
00:26:03,260 --> 00:26:13,520
If we power the whole planet with fusion in seawater and there is a thousand years worth of lithium in the Indian mountains,

243
00:26:13,520 --> 00:26:21,740
in lithium carbonate, it's in salty brines in the Andes Mountains, and there's 30 million years worth of lithium in seawater.

244
00:26:22,730 --> 00:26:26,880
So it's very abundant fuel sources. Okay.

245
00:26:27,260 --> 00:26:36,620
Important point here is you react to has to have this thing going on at about 250 million degrees or between 150 million and 250 million degrees.

246
00:26:36,860 --> 00:26:41,210
And this going on in your walls to produce your tritium to go around.

247
00:26:41,450 --> 00:26:46,070
This happens in something we call blankets in the wall where you deliberately put

248
00:26:46,700 --> 00:26:52,490
lithium into the blanket and breed your tritium in the wall out of the blanket.

249
00:26:53,600 --> 00:27:04,249
So in E to the middle is hot enough that that reaction will be going on at around about five megawatts of power,

250
00:27:04,250 --> 00:27:13,549
permit the cubed of the plasma in the middle of each other and it's producing the helium and the neutron.

251
00:27:13,550 --> 00:27:20,180
From that reaction, the helium gets trapped and stays in the plasma and it has 3.5 megawatts of energy.

252
00:27:20,330 --> 00:27:23,750
The plasma has 20 kilowatts. So that helium goes through the plasma.

253
00:27:24,490 --> 00:27:29,350
Getting that up just by collisional processes and depositing its energy in the plasma.

254
00:27:29,500 --> 00:27:37,900
So that's about 100 megawatts of heating at full power in each hour, and the power from the neutrons comes into the world.

255
00:27:39,100 --> 00:27:48,740
We will not breed most of the tritium for ITER, but we will have six panels on the wall with what we should call the test blanket modules.

256
00:27:49,150 --> 00:27:55,270
Six panels on the wall, testing out six concepts for the breeding blanket on the wall of Eater.

257
00:27:55,480 --> 00:27:58,960
And so of Europe has two of those panels.

258
00:27:59,410 --> 00:28:05,290
The US has decided not to have a panel on the wall and one of the other partners has decided not to have a panel on the wall.

259
00:28:05,290 --> 00:28:13,300
But basically the partners have experiments on the wall to test their concepts that breeding tritium from lithium in the wall.

260
00:28:13,720 --> 00:28:22,030
Key point here, self heating. If you get enough fusion, a fifth of the power will come out of self heating of the plasma.

261
00:28:23,440 --> 00:28:32,049
A nice, handy, dandy formula for the fusion power per metre cubed of plasma can be expressed in these units

262
00:28:32,050 --> 00:28:36,460
where that is the pressure of the plasma in atmospheres and this is megawatts per metric acute.

263
00:28:37,180 --> 00:28:42,630
In order to make a power station, you've got to get at least megawatts per metre cubed, if not more.

264
00:28:42,640 --> 00:28:46,870
Otherwise are going to have to build something the size of Wembley Stadium in order to get the gigawatt out.

265
00:28:47,920 --> 00:28:56,710
So this is a nice, handy dandy formula that works. It's an approximation to all the cross-sections, but a fifth of that fusion power, right,

266
00:28:56,740 --> 00:29:07,270
which is approximately p squared over 50, has to be balanced by the losses which are proportional to the energy in the plasma.

267
00:29:07,270 --> 00:29:16,330
So the pressure divided by this confinement time, the time that it takes for the turbulence to work the heat out.

268
00:29:16,540 --> 00:29:23,230
So this would say if you wanted to get self-heating, you have to get enough fusion to beat the losses due to turbulence.

269
00:29:24,700 --> 00:29:27,520
And if you look at that and this is a time in seconds,

270
00:29:27,850 --> 00:29:36,250
it says you've got to have a confinement time of several seconds in order to be in self-sustained kind of regime.

271
00:29:36,340 --> 00:29:41,380
That's just estimation. This is sort of state of the art simulation of Ito.

272
00:29:42,070 --> 00:29:45,700
This is somebody who's taken all the models of the turbulent transport,

273
00:29:46,090 --> 00:29:52,060
the heating, etc., and put it in a great big CO to model the whole system of ITO.

274
00:29:52,930 --> 00:30:02,020
Okay, so this is time along the bottom here in seconds and let's see what's happening here and fusion power up this side.

275
00:30:02,650 --> 00:30:09,459
Now what you do in ITR is the magnets are on the coils around their superconducting, they're on all the time.

276
00:30:09,460 --> 00:30:15,940
Basically you put in your deuterium and tritium in here and you induce the

277
00:30:15,940 --> 00:30:20,140
current in the plasma by changing the flux through the middle of the Taurus.

278
00:30:20,350 --> 00:30:24,430
So by this time you've actually got a plasma going right here.

279
00:30:25,060 --> 00:30:29,740
But this plasma is it about 10 million degrees, thereabouts.

280
00:30:30,220 --> 00:30:32,500
It's what we call ultimately heated at this point.

281
00:30:32,890 --> 00:30:43,060
At this point here, what this is, is the heater beams going on, the heater beams on E to a beams of neutral particles that come in from the outside.

282
00:30:43,060 --> 00:30:48,010
They slam into the plasma, the heat, the plasma up to the temperatures.

283
00:30:48,010 --> 00:30:52,780
For fusion, they have to be neutral because if they were charged, they wouldn't get through the magnetic field.

284
00:30:53,290 --> 00:30:57,509
So what you have is you have an accelerator that accelerates the charge particle

285
00:30:57,510 --> 00:31:03,190
over neutralise the charged particle to send it in as a neutral atom on jet.

286
00:31:03,190 --> 00:31:08,090
We do that by accelerating negative ions. Sorry, positive chance.

287
00:31:08,110 --> 00:31:15,009
Get the right way around. I'm looking at Barry right here of positive ions and then attaching an electron

288
00:31:15,010 --> 00:31:18,580
to it by passing it through a gas and then injecting it into the plasma.

289
00:31:19,030 --> 00:31:28,329
On Aether they will attach an extra electron to the atom and accelerate it as a negative ion and knock that extra electron off.

290
00:31:28,330 --> 00:31:32,830
And it will come in as a negative, as a as a neutral particle.

291
00:31:33,220 --> 00:31:40,629
So what you've got here is 70 megawatts of power coming into the plasma as those neutral particles,

292
00:31:40,630 --> 00:31:46,000
they slam into the middle of the plasma, get ionised and deposit their energy into the middle of the plasma.

293
00:31:46,300 --> 00:31:53,200
Then the temperature on each of shoots up the temperature I need to shoot up.

294
00:31:53,200 --> 00:31:58,030
And I think in this case, we're getting to about 200 to 230 million degrees.

295
00:31:58,270 --> 00:32:06,249
So 20 to 23 kilowatts at this point. By this time, the deuterium, tritium, fusion power has gone up at the beginning.

296
00:32:06,250 --> 00:32:09,070
It's exponential, and then it goes up as a power law.

297
00:32:09,280 --> 00:32:16,810
So you can see it really is dramatic how much fusion power comes up as the as the temperature goes up, we'll take the red line for obvious reasons.

298
00:32:18,670 --> 00:32:23,650
In the red line you'll see the modelling here has relaxation oscillations in the middle of the class.

299
00:32:24,220 --> 00:32:29,140
We do not model the turbulence directly here. We take the results of those turbulence codes.

300
00:32:29,140 --> 00:32:33,490
We fit them to empirical formulas, and we put them into these systems codes.

301
00:32:34,000 --> 00:32:38,980
And so at this point, we go through now what's happening to the heating as we step down to 50.

302
00:32:39,520 --> 00:32:47,350
Because by the time you've got up to here, the fusion reactions are producing enough heat to be part of the self-sustaining into the plasma.

303
00:32:47,620 --> 00:32:54,550
And actually, what you can do in this red, blue and green case is you can turn off the power of 400.

304
00:32:54,880 --> 00:32:59,020
And the plasma will self-sustained. That's the burning plasma.

305
00:32:59,260 --> 00:33:05,260
That's the plasma in which the fusion itself is providing the heat to keep the plasma hot, to keep it going.

306
00:33:05,530 --> 00:33:11,410
And that is what it really has to do. They'll tell you they wanted to get two Q equals ten.

307
00:33:11,650 --> 00:33:17,770
I'll tell you what I want them to do is to get to a really properly what we call ignited state of self-sustaining,

308
00:33:17,770 --> 00:33:28,750
that like this recording this is simulations from Princeton Group and a scan two was the one that they considered to have the best model, best models.

309
00:33:28,780 --> 00:33:36,790
What they did is they degraded that model. And they said, okay, if that model and our uncertainty in the theory is not very good,

310
00:33:37,090 --> 00:33:44,360
what they did is they lowered one of the boundary values on the temperature and we got skin free here.

311
00:33:44,380 --> 00:33:50,530
This can free would be exciting. Not as exciting as that because it's still self-sustained fusion.

312
00:33:50,980 --> 00:33:55,420
No energy and basically energy out. So that's lovely.

313
00:33:55,930 --> 00:34:04,809
But Skin four would be disastrous. So what happened in scattered forms can form as interesting as you produce fusion reactions.

314
00:34:04,810 --> 00:34:09,850
You make helium that helium heats the plasma, but it is a pollutant.

315
00:34:10,120 --> 00:34:14,440
It's diluting your fuel. It's ash. It's the result of fusion reactions.

316
00:34:14,650 --> 00:34:19,780
So if the heat in builds up, you'll poison the reaction and the tokamak will go out.

317
00:34:20,920 --> 00:34:29,980
And if you so, if. If the turbulence does not flush the helium out after it's slowed down and deposit its energy in there,

318
00:34:30,220 --> 00:34:38,440
you will get a build up of ash and you're talking about will go out in the case of scan for he turned off the turbulent diffusion

319
00:34:38,440 --> 00:34:45,160
of the helium and what happened is it built up and then you can see it went out even before you turned off the heating power.

320
00:34:46,090 --> 00:34:51,790
Right. We don't know exactly how to turn this will interact with the alpha particles.

321
00:34:52,390 --> 00:34:58,240
And that's something we wish to learn from jet more about in the in the next few years.

322
00:34:59,710 --> 00:35:06,560
So that's that's the dramatic thing that could happen on E to each is an experiment.

323
00:35:06,580 --> 00:35:09,700
It's not a demonstration. It's an experiment.

324
00:35:10,120 --> 00:35:20,650
We think that it's this kind of behaviour we will get to, but we have to be aware that we don't know everything that we do over the timescale.

325
00:35:20,890 --> 00:35:26,799
Is that really 400, 400 C at full power?

326
00:35:26,800 --> 00:35:34,000
You'll be able to do something in the thousand to 2000 seconds before heating of various components will be too much.

327
00:35:34,000 --> 00:35:37,240
And you have to turn the machine off. So.

328
00:35:37,470 --> 00:35:41,200
So what was the longest sustained time?

329
00:35:41,380 --> 00:35:45,260
So, James, I'm going to show you that in a second.

330
00:35:45,270 --> 00:35:49,010
But this is about this notion of self-sustaining diffusion here.

331
00:35:49,480 --> 00:35:55,050
Now, looking from this point of view to the decomposition, thinking about 280.

332
00:35:55,420 --> 00:35:59,390
Yeah. So what are we. Well, the plasma, right.

333
00:35:59,400 --> 00:36:05,520
Is it about 23 K and all the reactions are happening in the tail of the Maxwell.

334
00:36:05,910 --> 00:36:12,210
As often happens in these cases, you don't want to hit the temperature to be at the peak of the cross-section of dty,

335
00:36:12,870 --> 00:36:16,530
which is which is about 60 K is the peak.

336
00:36:16,890 --> 00:36:24,360
Because when you get up there, the radiation from the plasma is the best place to sit is a bit lower temperature in the peak of

337
00:36:24,370 --> 00:36:29,190
the diffusion cross-section because the radiation cross-sections are really coming up steeply.

338
00:36:29,280 --> 00:36:30,750
By the time you get to 60 K,

339
00:36:31,290 --> 00:36:38,490
so the optimum temperature to run it is less than the peak of the actual fusion cross-section because you want to keep the radiation down,

340
00:36:39,390 --> 00:36:44,640
particularly the synchrotron radiation. Sorry. The other thing I would mention the poisoning output.

341
00:36:44,970 --> 00:36:50,040
Mm hmm. What about the need for the neutral atoms of deuterium?

342
00:36:50,880 --> 00:36:54,150
Well, you can actually inject tritium, so you're just injecting fuel.

343
00:36:54,910 --> 00:37:00,900
The other way we put fuel into it is we make frozen pellets, and we fly them in as fast as you can.

344
00:37:01,290 --> 00:37:08,010
And they cross before they oblate. They cross a certain distance and that's how you get deuterium and tritium into the middle of the plasma.

345
00:37:08,760 --> 00:37:17,310
One of the questions we don't know the answer to this is will the turbulence differentially transport tritium and deuterium differently?

346
00:37:17,700 --> 00:37:22,920
So we would like to keep 5050 deuterium and tritium everywhere in the plasma.

347
00:37:23,730 --> 00:37:27,480
But if the deuterium is better transported by turbulence than the tritium,

348
00:37:27,720 --> 00:37:32,250
that you'll end up with not 5050, and then you won't have an optimal situation.

349
00:37:32,580 --> 00:37:35,610
So lots of questions that we really don't know the answer to.

350
00:37:35,870 --> 00:37:42,809
And one of the reasons that we're going to do another blast, a huge blast of deuterium and tritium in jet towards the end of this decade.

351
00:37:42,810 --> 00:37:49,080
And I'm about to describe describe that. How do we know this will happen?

352
00:37:51,300 --> 00:37:54,330
Barry smiling at me. Do we know it will happen?

353
00:37:54,510 --> 00:38:02,880
We know it will happen. Partly from empirical extrapolation, from where we are, partly from the theoretical models that have been being described,

354
00:38:03,060 --> 00:38:09,690
and partly from dimensionless dimensionally similar experiments.

355
00:38:09,870 --> 00:38:18,570
So if you if you write the answers in dimensionless quantities and you try and produce a jet discharge that looks like at each discharge,

356
00:38:18,960 --> 00:38:24,300
but just the change is in in the scaling, right.

357
00:38:24,300 --> 00:38:27,720
You should be able to understand what the scale up is.

358
00:38:28,680 --> 00:38:37,950
ETA is twice the size of JET. So the most important machine in the world by far is the next biggest one to each of which is Jet.

359
00:38:38,370 --> 00:38:42,600
And Jet is the only machine in the world that can currently run in tritium.

360
00:38:43,740 --> 00:38:49,950
We had one other which was TFT in the States and they weren't as good as us, so they closed the machine down.

361
00:38:51,150 --> 00:39:02,560
But it's very important in the lead up to each that we use JET as effectively as possible to reduce the risk in the program.

362
00:39:02,970 --> 00:39:10,050
Each program has risks in it, and we can make those less by making sure that on Jet we discover everything necessary.

363
00:39:10,290 --> 00:39:22,500
Going into that, the extrapolation from Jet to E2 is only a factor of two in size, but it is a regime that we've not really been into.

364
00:39:22,740 --> 00:39:29,460
So we are using physics as much as possible to predict that performance, and we hope we will get there.

365
00:39:30,120 --> 00:39:39,179
I will say in the laser world, the extrapolation to NIF from the previous machine Nobel was a huge extrapolation,

366
00:39:39,180 --> 00:39:42,809
and one of the problems that the NIF programme is having at the moment is that

367
00:39:42,810 --> 00:39:47,160
they had nothing in between and it's very important at the moment to run.

368
00:39:47,730 --> 00:39:53,880
This is a picture inside Jet at the moment and three years ago we replaced the inside wall of Jet.

369
00:39:54,570 --> 00:40:01,500
So it was understood in 2009 a report to the commission chaired by Albrecht Bogner, who used to be the head of DSC.

370
00:40:02,310 --> 00:40:08,940
Albrecht, one of his committee, said it is absolutely important to keep jet running right up to the beginning of of each year

371
00:40:08,940 --> 00:40:15,780
because what we will learn from JET will allow us to have expertise going into each operation.

372
00:40:16,680 --> 00:40:22,020
But a key thing is to change jet so that it has the same raw materials as Ito.

373
00:40:22,260 --> 00:40:28,680
And what we've done is this all used to be carbon on the inside of jet, these tiles, etc. all used to be carbon fibre composites.

374
00:40:28,950 --> 00:40:33,179
And this is now beryllium and this is tungsten down here.

375
00:40:33,180 --> 00:40:37,320
This is the what we call the diverter or the system at the bottom.

376
00:40:37,500 --> 00:40:45,120
And this is the robotic arms which we did all the work with. So we now can maintain a jet entirely from externally.

377
00:40:45,360 --> 00:40:50,280
And this is important as we go into tritium operation again, because once you've operating with tritium,

378
00:40:50,280 --> 00:40:55,050
the inside of a jet becomes a hazardous environment and you can't send people inside.

379
00:40:55,230 --> 00:41:04,650
So this capability that we've developed over the last decade and a half in main maintenance in the machine robotically allows us to go forward.

380
00:41:05,100 --> 00:41:16,470
And the key experiments go back to those 1997 experiments, which we call DP one, when we actually did some fusion power out of jet of considerable.

381
00:41:16,650 --> 00:41:20,370
We've run tritium since then in 2003, but they weren't full power.

382
00:41:20,370 --> 00:41:27,060
Tritium shocks. This was the famous shop 16 megawatts here.

383
00:41:27,870 --> 00:41:31,379
But you can see 2 seconds and actually just be honest about it.

384
00:41:31,380 --> 00:41:33,000
It wasn't the best two shots.

385
00:41:33,000 --> 00:41:39,240
I mean, it's a fantastic amount of computing power, but something happened to it right there and it plunged right back down again.

386
00:41:40,980 --> 00:41:44,309
And so it is very transitory in that.

387
00:41:44,310 --> 00:41:49,380
And there were about 24 megawatts of power going in at that time to get that 16 megawatts out.

388
00:41:49,650 --> 00:41:56,580
A much better shot, actually. And something we've used in the prediction is this one where you turn the machine on,

389
00:41:56,580 --> 00:42:02,820
you ramp up the power, you get fusion power, 4.5 megawatts and you turn it off at the end of the shot.

390
00:42:03,090 --> 00:42:10,290
You cannot run longer than this widget because the systems one will run at full power for longer than about five or 6 seconds.

391
00:42:10,590 --> 00:42:16,320
And so that's really all you can do. This was the best the Americans could do, say no more.

392
00:42:18,750 --> 00:42:28,620
But we've been given permission to do tritium again. So about a month ago we launched the project to re engage all of our tritium capability.

393
00:42:28,620 --> 00:42:33,030
We still have tritium on site and we've kept our capability through the years.

394
00:42:33,360 --> 00:42:40,470
But now the plan is to start tritium operation in 2017 and to break our records.

395
00:42:40,980 --> 00:42:45,780
And what the predictions of the code say is that we can probably get up to about.

396
00:42:46,160 --> 00:42:50,450
20 megawatts for about five or 6 seconds.

397
00:42:50,960 --> 00:42:56,090
It says 2015 because I was promised we could start in 2015 and now it's 2017.

398
00:42:56,840 --> 00:43:04,820
So sorry about that. But I think it's very important to do this because there's a lot of questions that we don't know about for the tritium operation.

399
00:43:06,710 --> 00:43:10,730
I'm running out of time, so I will move on very quickly.

400
00:43:11,240 --> 00:43:18,319
Question Can we make it smaller and cheaper? The problems is to the size of jet.

401
00:43:18,320 --> 00:43:24,260
The size of it are set by being big enough that the time it takes turbulence to move

402
00:43:24,260 --> 00:43:29,480
the hot plasma to the edge of the cold plasma to the middle is 3 to 4 seconds.

403
00:43:30,980 --> 00:43:40,220
And if we could go back to almost ten months free operation, the size of the machine would shrink to tens of centimetres instead of many metres,

404
00:43:41,630 --> 00:43:45,410
and it would make a lot of difference in the development of fusion.

405
00:43:45,560 --> 00:43:51,920
I think it's crucial to get to electricity as soon as possible. So I'm fully in support of demo and of beta.

406
00:43:52,370 --> 00:43:59,540
But what we need to be doing right now with the more academic side of things is looking to ways to make it smaller, cheaper, faster, etc.

407
00:44:00,960 --> 00:44:05,480
I can't guarantee anything here, but could we go towards more?

408
00:44:05,630 --> 00:44:13,280
Well, at least reduced tokamaks. And what we've done with Oxford is to assemble the best.

409
00:44:14,090 --> 00:44:24,230
I think we now have the best theoretical plasma physics fusion plasma physics group in the world here at this collection check in power.

410
00:44:24,920 --> 00:44:29,569
Michael Bonds, who's joining from Texas, Edmund Haycock, who's it to join our effort?

411
00:44:29,570 --> 00:44:35,960
Maudlin really done some absolutely terrific work. And I want you to just highlight what that was and what we were thinking.

412
00:44:36,560 --> 00:44:41,800
Okay. So the the turbulence is really quite small scale compared to the size of the device.

413
00:44:41,810 --> 00:44:45,710
In fact, the typical eddy size is the size of these of a radio.

414
00:44:46,190 --> 00:44:52,340
It's a few times that and it means there are about a thousand eddies in the size of each.

415
00:44:53,360 --> 00:44:58,790
But terms is really small scale. It's not these big turbulent eddies, etc., it's this small scale stuff.

416
00:44:59,270 --> 00:45:05,570
And how can you change that? Because you've got this free energy source, which is it's hot there, it's cold there.

417
00:45:05,780 --> 00:45:07,610
Can you really reduce that?

418
00:45:07,850 --> 00:45:17,479
Well, one of the things and I think I'm going to skip through this because it's been covered before, but this term, this is fully kinetic.

419
00:45:17,480 --> 00:45:25,220
There are almost no collisions. It's not a fluid. It's it's a collection of charged particles in phase space that makes up a plasma.

420
00:45:25,400 --> 00:45:33,260
And so it's truly a5d problem. And to simulate on an even on the best computers simulating in five D,

421
00:45:33,740 --> 00:45:38,390
the best way to make progress is through smart theory and not through big computers,

422
00:45:39,260 --> 00:45:45,620
because big computers get you a factor of two every now and again. And smart theory can get you a factor of a thousand every now and again.

423
00:45:46,550 --> 00:45:54,950
And the key thing here was to average out all the all the timescales to make the problem as compact as possible.

424
00:45:55,100 --> 00:46:03,830
And then it could go on a large computer and a key one was this averaging around the motion, around the free lines,

425
00:46:04,130 --> 00:46:09,020
and to produce the distribution function of charged rings as opposed to a distribution

426
00:46:09,020 --> 00:46:14,540
function of particles that reduced it from six dimensions to five dimensions,

427
00:46:14,900 --> 00:46:18,800
but also reduced the time scales that we have to solve on.

428
00:46:19,130 --> 00:46:22,310
And that was critical, those simulations.

429
00:46:22,330 --> 00:46:29,719
The second thing that reduced the number and that reduction took down the computational

430
00:46:29,720 --> 00:46:33,500
problem by about a factor of ten to the three by a factor of a thousand.

431
00:46:33,800 --> 00:46:42,050
The second one was this idea of simulating not the whole plasma at once, but only one correlated volume at a time.

432
00:46:42,800 --> 00:46:46,880
And this reduced the size of the volume that you had to solve on your supercomputer.

433
00:46:47,660 --> 00:46:58,790
This was another factor of a thousand in the calculation. You've seen this before, but the key thing here was an observation of what happened.

434
00:46:59,540 --> 00:47:05,600
This this plasma is rotating. You think it's rotating like a smoke group.

435
00:47:05,930 --> 00:47:11,270
You think it's a toroidal vortex doing this, but actually it's an inverse bipolar effect.

436
00:47:11,570 --> 00:47:16,610
It's really rotating in the long way round. And because of the striations,

437
00:47:17,000 --> 00:47:23,780
it looks like it's rotating the short way round in the middle is rotating one way and the outside is rotating the other way.

438
00:47:24,020 --> 00:47:29,210
The main rotation is being taken out. That rotation turns out to be incredibly important.

439
00:47:29,390 --> 00:47:36,410
What it does is it combs out the large eddies that want to move lots of hot stuff all the way out and back in again.

440
00:47:36,410 --> 00:47:39,560
You can see how they're being sheared apart by the sheer and the rotation.

441
00:47:39,830 --> 00:47:43,580
And in fact, this region here, there's so much shear, there's almost no turbulence.

442
00:47:44,090 --> 00:47:52,680
We observe that in JET. We observe that in mast is that the turbulence can be suppressed by the share in the rotation quite considerably.

443
00:47:53,700 --> 00:47:57,030
In fact, this is what you see in the Big Bang Theory.

444
00:47:58,560 --> 00:48:00,900
The Big Bang Theory is produced in Los Angeles,

445
00:48:00,900 --> 00:48:10,680
and I used to be a professor at UCLA before I came back to the To Run column and we were having a conference.

446
00:48:10,680 --> 00:48:19,770
I think actually Alex was there and the technical adviser to the Big Bang Theory is Dave Saltzberg, who's a particle physicist at UCLA.

447
00:48:20,040 --> 00:48:30,590
And every Monday afternoon he goes to the the set of the Big Bang Theory, and he writes things on the whiteboard for Sheldon to to to display.

448
00:48:31,500 --> 00:48:39,450
And we had been thinking about shared rotation in Tokamaks and talking about sheering apart

449
00:48:39,450 --> 00:48:44,219
the eddy structures by stretching them out and how much we could suppress the turbulence.

450
00:48:44,220 --> 00:48:52,320
By doing that, it's not obvious that by rotating and producing a sheared flow that you make the turbulence better, you might make it worse.

451
00:48:52,650 --> 00:48:55,110
And so there's all kinds of calculations going on here.

452
00:48:55,380 --> 00:49:00,810
But anyway, what Dave Saltzberg did was to take the white board that was sitting in the physics department there,

453
00:49:00,930 --> 00:49:05,729
copy down what was on it, and then put it on to the white board, the Big Bang Theory.

454
00:49:05,730 --> 00:49:08,910
And so that's an actual still. That's fine.

455
00:49:10,990 --> 00:49:14,280
Anyway, can we share away the turbulence?

456
00:49:14,640 --> 00:49:17,010
Well, his very interesting calculation.

457
00:49:17,010 --> 00:49:25,980
So about three or four years ago when when Felix Power came here and Michael Barnes was here and Edmund Haycock was just a PhD student,

458
00:49:26,250 --> 00:49:31,350
we had a project to say what would be the optimum shared rotation we put into the project.

459
00:49:31,920 --> 00:49:37,620
So along here is the rotation share in units you don't have to care about an up here is the turbulent

460
00:49:37,620 --> 00:49:45,660
heat flux and we're normally working in a plasma that has a temperature gradient of about the red curve.

461
00:49:46,260 --> 00:49:57,870
And so what you can see is as you increase the rotation shift, you come to a region here with no turbulence, no turbulent heat flux and no turbines.

462
00:49:57,870 --> 00:50:01,620
We're about here right now on the mast device column.

463
00:50:01,620 --> 00:50:08,969
If we can get down into this spot here, we should see a dramatic increase in the quality of the confinement.

464
00:50:08,970 --> 00:50:12,960
So there's a sort of optimum shear rate here that might be there.

465
00:50:13,260 --> 00:50:16,889
So there's little Haynes and the plasma sometimes does this.

466
00:50:16,890 --> 00:50:21,810
Sometimes it produces sheared parts of the plasma where there's almost no turbines.

467
00:50:22,140 --> 00:50:27,990
There's little hints that having a having a magnetic confinement device without turbines is possible.

468
00:50:28,590 --> 00:50:32,250
And what we need we need from the theoreticians now is to lead.

469
00:50:32,520 --> 00:50:36,420
Somebody earlier said, you know, other theoreticians, just tell me I'm done.

470
00:50:37,050 --> 00:50:42,360
Are the theoreticians just explaining the experiments or are they leading the experiments here?

471
00:50:42,360 --> 00:50:48,299
We want them to lead the experiments. We're building a new machine at Culham.

472
00:50:48,300 --> 00:50:51,900
It's an assembly right now. It'll be on line this time next year.

473
00:50:52,320 --> 00:50:58,740
It will rotate at Mach one, and we will hope to hit that sweet spot.

474
00:50:59,580 --> 00:51:04,409
One more thing. We have to rotate that by momentum input.

475
00:51:04,410 --> 00:51:13,920
And what Felix has shown, something quite remarkable is that turbulence sometimes can produce velocity shear without there being any momentum input.

476
00:51:15,000 --> 00:51:19,920
It transports some angular momentum inwards, the opposite of momentum outwards,

477
00:51:20,850 --> 00:51:26,520
and produces a sheared rotation inside your device with no net momentum input.

478
00:51:27,180 --> 00:51:33,989
That would be fantastic because then maybe we can encourage future fusion reactions, reactors to rotate on their own.

479
00:51:33,990 --> 00:51:34,920
So thank you very much.