Talk:Nuclear fusion
|
|
Please attribute the criticisms of fusion power, or provide better arguments for them. At the moment I'd strongly disagree with them. --Robert Merkel
Here's one link
http://www.ornl.gov/~webworks/cpr/rpt/104912.pdf
Right at the beginning of the article there's a really bad mistake: it says that "strong force only operates over short distances, unlike the electromagnetic force". It is in fact the other way around: electric force gets weaker and weaker at large distances, while the strong one is overwhelming at large distances (because of confinement), getting weaker and weaker at short distances (because of asymptotic freedom). I lack the style and experience to edit the article, but wanted to make this small suggestion...(neko)
| Contents |
Belated Move from Talk:Fusion
- The following two 'graphs apparently created Talk:Fusion in the edit "M 15:43, 2002 Feb 25 . . Conversion script (Automated conversion)":
- could somebody add something about tokamaks laser fusion and any other current research here please. Also, please check my physics as I am not that sure of myself here. Thanks
- What is the energy for DT reactions?
Misc
Here's another
http://vlt.ucsd.edu/stacey.pdf
I'll rephrase the criticism, but it was a hot topic a few years back when the Department of Energy was deciding funding priorities for future energy research.
Has anybody considered the possible social and environmental costs of obtaining an extremely cheap source of energy? For example, the economic changes that might result, and the environmental damage that could be done if energy costs are no longer a barrier to huge civil engineering projects? -- Heron
Coal plants source of radioactive waste? check change by 66 ...
Yep. If ORNL says it, it must be so: [1] (http://www.ornl.gov/ORNLReview/rev26-34/text/colmain.html) Graft
Can we include something about the possibility of extracting Helium-3 from the lunar surface? This has been bandied about for years by adocates of space exploration and space commercilization, and recently has been advocated by some within the Chinese government. RK
I think there are a few bugs in this page. I corrected one for the D+D reaction, in the 2nd case the new nucleus is a 3He, not a T. Plus, shouldn't the atomic numbers be written top-left instead of top right of the symbol (in the equations)?
what else is needed for stellar fusion?
In the section of this article discussing gravitational confinement, one finds this assertion:
"Some simple math can demonstrate that the mass of fuel needed to make a star using the D-D reaction is about the size of the Moon."
Certainly Jupiter contains that mass of hydrogen and more. But it isn't a star. So what else is necessary to induce the fusion reaction?
- If the "fuel" assumption that yields that result is some typical non-stellar isotopic assortment, the answer may be interesting. If it refers to something else (pure deuterium would be a logical choice for such a calculation), Jupiter should not be expected to qualify. And "and more" is irrelevant since concentration is likely to be crucial.
- Also regarding composition, neutron-absorbing substances can poison fusion. (The control rods in fission reactors similarly poison the fission reaction.) This is a plot point in (the novel but IIRC not the film of) The Sum of All Fears.
- BTW, "the size of the Moon" is an odd way to put the "mass of fuel": that size under what T & P conditions? It takes no calculation to deduce that a lunar-mass of H2 or D2, compressed to a lunar-diameter sphere, and located near the Earth's or Moon's orbit would dissipate by reaching escape velocity, since we know the moon is far from holding an atmosphere even of heavier gases. (Even Mars is believed to have lost most of its H2O from dissociation and the escape of the H2 that results.) Even if H2 released on the Moon escapes slowly, we know the mass i've specified would expand immediately, producing
- Of course i don't mean to assume that the D-D statement is accurate! If it is, it isn't clearly enough stated, IMO.--Jerzy(t) 21:54, 2004 Mar 6 (UTC)
How can nTτ be about 10^14 s/cm³? There is unit inconsistency. It would have to be s cm^-3 K.
- 10^14 is nτ and actually is s cm^-3. I will fix the reference to make it consistent. Trelvis 19:16, May 17, 2004 (UTC)
College Research Paper - Peer Review
Has been moved here: /Pompura paper
Lawson Criterion
As it currently stands, there's a typo in the section on the Lawson criterion. The text states <math>nT\tau<math> must be greater than <math>10^{14} s/cm^3<math>. The units don't match the quantity, and - I think - should be <math>s-eV/cm^3<math> or <math>s-keV/cm^3<math>. Do any of the current watchers have the correct units on hand?
Also for reference, the textbook Choudhuri. Physics of Fluids and Plasmas: An Introduction for Astrophysicists. Cambridge University Press: Cambridge, 1998. describes the Lawson criterion as the restriction <math> n \tau > 10^{16} s/cm^3<math>, but I don't know what temperature he assumes.
Check these numbers?
According to our section "The fusion triple product", :
- the nuclei only undergo fusion once in every 1029 seconds. However, the fact that the Sun contains 1059 nuclei means that the net reaction rate is actually quite high, and since the Sun is around for billions of years, eventually the fuel is used up and the total energy released is huge.
This doesn't seem right. If the reaction rate per nucleus is 10-29 s-1, which is order of 10-21 yr-1, then it will take a lot more than mere billions of years to use up the fuel. Indeed the claim that: 1059 nuclei means that the net reaction rate is actually quite high also seems false; our Sun article gives the Sun a mass of 1.9891 × 1030 kg (agreeing with the <math>4\pi^2 r^3 = GMT^2<math> law), which (given 99% H, and NA H nuclei to the gram) would result in only about 1057 nuclei. 1057 nuclei × 10-29 s-1 = 1028 fusions per second, and about 40 kg of fuel consumed per second - way too low! It seems likely that the 59 should be 57, but I'm not sure what the 29 should be. I suppose it could be approximated from power output, but if someone already knows, please fix it! Securiger 10:01, 28 Nov 2004 (UTC)
- Another way to put it: if we allow 6.5 MeV (<math>10^{-12}<math>J) for one quarter of 4p→4He + 2e (which I think is about right, and is certainly the right order of magnitude), then <math>10^{-12} J/nucleus \times 10^{57} nuclei \times 10^{-29} s^{-1} = 10^{16} W<math>, which, spread over a sphere of area <math>4\pi R_e^2 = 2.8 \times 10^{23} m^2<math> corresponds to a "solar constant" of 35 nanowatts/m2. Our reaction rate seems to be out by 10 orders of magnitude! Securiger 10:28, 28 Nov 2004 (UTC)
- There are several different fusion reactions that happen in parallel. See [2] (http://www.tim-thompson.com/fusion.html) for a discussion of the different rates. Also, the cross-section-velocity product at 20 keV (200,000,000 K) for deuterium-tritum fusion (which I think is the main energy-producing reaction) is <math>4.2 x 10^{-22}<math> (from Goldston & Rutherford. Introduction to Plasma Physics. IOP Publishing, Ltd: Philadelphia, 1997.) For a density of <math>10^{21} m^{-3}<math>, this gives a reaction rate of 0.42 per second per nucleon... Unfortunately, I don't know the solar core's parameters, although the surface is around 5 eV. A detailed (but technical) treatment of fusion cross-sections in a solar context is given at [3] (http://www.sns.ias.edu/~jnb/Papers/Preprints/Solarfusion/paper.pdf). Also, [4] (http://www.int.washington.edu/talks/WorkShops/Neutrino02/Plenary/People/Bahcall/SSMandexpts.pdf) gives a power output for a net 4p -> He reaction of 25 MeV/nucleon (one eV = <math>1.6 x 10^{-19}J<math>) One final reference, [5] (http://www.shef.ac.uk/physics/teaching/phy303/phy303-7.html) is fairly accessible, but I should be getting back to work. Will try to come back here & help distill the numbers into something manageable later.SMesser 17:41, 30 Nov 2004 (UTC)
- From the Plasma Formulary: Solar Radius = <math>6.96 x 10^{8}<math> m, the upward mass flux is <math> 1.6 x 10^{-8} kg/m^2-s<math>, and mass = <math>1.99 x 10^{30}<math> kg, for a particle confinement time of <math> 2.04 x 10^{19} <math> seconds = <math> 6.47 x 10^{11} <math> years. An estimate of the energy confinement time can be reached by comparing the sun's luminosity (<math>3.83 x 10^{26} <math> J/s) with the energy content of the sun, which is approximately twice its mass times its temperature. (Since temperature is energy / particle, and a zeroth-order model of the sun has only protons and electrons, with the protons contributing nearly all the mass.) To do this calculation properly, you'd need estimates of the densities, temperatures,and compositions of the sun at different depths, but a quick-and-dirty estimate can be reached by using estimates of the dense solar core (below) and of the less-dense surface (above). A geometric mean of the three values gives a mean temperature of 400 eV, for a rough thermal energy of <math>2 x 10^{41}<math> J, for a confinement time of <math>4 x 10^{14}<math> sec = 10 million years. Those are estimates of the <math>\tau <math> factor in the triple product, with the energy confinement time being more relevant for the Lawson criterion. Also from the formulary, we can estimate a rough solar density of 1410 <math>kg/m^3<math>, although that's certainly lower than the core value (see below). This gives a proton density of <math> 8.42 x 10^{29} m^{-3}<math>, which provides a lower limit on n.SMesser 18:04, 3 Jan 2005 (UTC)
- From [6] (http://www.sns.ias.edu/~jnb/Papers/Preprints/Solarfusion/paper.pdf), solar fusion reactions typically take place at temperatures of 5 - 30 keV, which should be a decent estimate of T. If we take a proton-proton collision energy of 12 keV, this paper's equations 7 and 8 give a cross section of <math> 2.7 x 10^{-23} <math> barns or <math> 3.3 x 10^{-51} m^2<math>- pretty tiny, which is at the very least consistent with my understanding that the p-p reaction is the rate-limiting part of the process.SMesser 18:04, 3 Jan 2005 (UTC)
- [7] (http://www.shef.ac.uk/physics/teaching/phy303/phy303-7.html) quotes a solar hydrogen density of <math>10^5 kg/m^3<math>, which is presumably taken at the core. This translates to a proton density of <math> 6 x 10^{31} m^{-3}<math>. Combining this with the estimates of cross section and core density and temperature (above), I get a typical proton-proton reaction time of <math>6.0 x 10^7<math> seconds = 1.9 years. (i.e. about 37% of the solar core's protons will fuse in 1.9 years. This may or may not be consistent, depending on the size of the core relative to the sun.)SMesser 18:04, 3 Jan 2005 (UTC)
- Looking back to the sun's luminosity, if each fusion reaction provides 25 MeV, <math>3.83 x 10^{26}<math>J/s implies <math>9.6 x 10^{37}<math> He nuclei produced per second, or <math> 6.4 x 10^{11}<math> kg/s of H fusing to form He. This implies a solar lifetime on the order of <math>3.1 x 10^{18}<math> sec = <math> 9.9 x 10^{10}<math> years, which is about 20 times too long. To be consistent with the 1.9-year estimate for the fusion core, it also predicts a mass of <math>3.8 x 10^{19}<math> kg of hydrogen in the core.SMesser 18:04, 3 Jan 2005 (UTC)
- Can anyone out there check my math, please? Thanks. SMesser 18:04, 3 Jan 2005 (UTC)
- Taking another stab at this with numbers from the last paragraph above... I'll estimate the mean lifetime of a proton using the sun's luminosity and density: <math>9.6 x 10^{37}<math> He nuclei produced per second implies <math>3.8 x 10^{38}<math> H nuclei used per second. Since hydrogen makes about 75% of the sun and the mass is <math>1.99 x 10^{30}<math> kg with <math>1.7 x 10^{-27}<math> kg per proton, there should be <math>8.9 x 10^{56}<math> protons present. So the typical time-till-fusion of a typical proton is more-or-less the ration of the number of protons to the rate at which they're fusing: <math>2.4 x 10^{18}<math> seconds.
- Since it's been a few weeks and I'm the only person posting with hard numbers, I'm going to toss the fallout from this last bit on the page. (Hopefully the lack of posts isn't because I've accidentally marked some of the edits as minor. If it was, I'm sorry.) Anyhow, I'm also going to cross-post on the talk:sun page to ask for more data. SMesser 21:08, 21 Jan 2005 (UTC)
- Saw the note on Talk:Sun and have been pondering this. For the estimate of number of particles in the sun, it looks like you're about on track with 1057 - this link has some calculations:[8] (http://www.uclan.ac.uk/facs/science/physastr/x99/PAM98/UCert/Ch09/9_3the~1.htm). As for the proton-proton reaction rate/lifetime of a proton in the sun, my way of working it is pretty much the same as yours - solar luminosity is 2.9×1026W, and by my calculation each fusion reaction liberates 0.7% of the rest mass of four protons, or 4.64e-29kg. This equals 4.176e-12J. So, to support the solar luminosity, there must be 6.944e37 fusion reactions per second. The number of protons divided by four times the reaction rate (because each reaction requires four protons) gives us a proton lifetime of 4.32e18s, which is near enough similar to your estimate. This is 136 billion years, which means that during the solar lifetime of 10 billion years, rather less than 10% of the protons present would fuse. I believe that is a plausible result. Worldtraveller 18:16, 27 Jan 2005 (UTC)
Stop! Most of this belongs under Fusion power
The Fusion disambiguation page says
- For the combining of two atomic nuclei into a single nucleus (with possible emission of radioactivity), see nuclear fusion
- cold fusion refers to a once hypothetical but now largely discredited form of nuclear fusion
- See fusion power for a discussion of using nuclear fusion as an energy source
That is, the Nuclear fusion page is only for fusion as a nuclear process. All aspects of fusion as an energy source should be moved to Fusion power. I should say "integrated into Fusion power" since the pages are 80% redundant. I think such a division makes good sense. Art Carlson 17:56, 2005 Jan 30 (UTC)
Tunneling
I think the sentance "additional effects, such as quantum tunneling, lower the energy barrier slightly" is incorrect. I don't know about nuclear chemistry, but in "regular" physical chemistry tunneling is a process by which species "tunnel" through an energy barrier without lowering the barrier. I think it is more correct to say that the "additional effects, such as quantum tunneling, lower the effective activation energy slightly". The difference being that effective activation energy (perhaps there is a more formal term for it) reflects the required ammount of energy put into the system and is independant of the actual height of the thermodynamic barrier. Fearofcarpet 17:34, 18 Mar 2005 (UTC)
latest condensation
Regarding the latest edit which got rid of most of the explanations of types of fusion confinement: I rather think we should keep this. Yes the individual articles have detailed information on each method but there's no reason we can't have a short synopsis of all the methods here. I welcome Art's edits very much and am glad we have a real plasma physicist here to make these edts, I just get slightly nervous when he gets on with big swaths of deletions. --Deglr6328 07:12, 24 Mar 2005 (UTC)
- Addendum- I also have to say I don't really like the whole removal [9] (http://en.wikipedia.org/w/index.php?title=Nuclear_fusion&diff=11303183&oldid=11302763) of "fusion triple product" and "breakeven and ignition". Why couldn't these just be improved a bit instead of being deleted? Quell my fears Art! :o) --Deglr6328 07:19, 24 Mar 2005 (UTC)
Thanks for the flowers. I realize I came out redecorating with napalm. I'm surprized I got away with most of it. The issue here is where things belong. I'm not sure having separate articles for nuclear fusion and for fusion power is a good idea, but as long as we do, there should be some system to it. I am trying to draw the line between fusion as a nuclear process and fusion as (potentially) used for electricity production. I discovered that even this line is not so easy to draw. I put the <σv> business in nuclear fusion because it can be discussed (nearly) idependent of any particular confinement scheme, even though it is only of interest if you have power production in mind. That's why I reduced discussions of confinement methods, triple product, breakeven, and ignition to a bear minimum. What do you think is best? One big article for everything? Substantial duplication? Or some line of demarcation with cross references and minimal overlap? Art Carlson 09:29, 2005 Mar 24 (UTC)
- I say separate articles are good with some slight duplication necessary for thoroughness of discussion in each article. Wouldn't "breakeven (well... maybe not breakeven) and ignition" be ok to keep here because, for example the term "ignition" can be applied to fusion process in stars? Is the fusion triple product section really THAT inappropriate for this article insted of being in fusion power? Really, I'm asking non-rhetorically 'cause I'm not certain. If it (and breakeven/ignition stuff) does have to go can you incorporate it into fusion power? Also, I would love it if you could provide an explanation/description in the magnetic mirror article of why the "baseball" experiments were considered (largely anyway...?) faiures.--Deglr6328 06:38, 25 Mar 2005 (UTC)
Fixing the table of reactions
Could someone please fix the table of reactions? The energy released by 3He+3He is missing, and it's not clear what's meant by energy values in parentheses next to the products (eg. D + T -> 4He(3.5MeV) + n (11.4 MeV)). If someone's feeling really ambitious, it would be nice to know what the Lawson criterion is at some "typical" temperature (eg. DT has 10^20 sec/m^3 at 10keV). Thanks much. --NDL
Seconded, 3He should be listed on all the tables ASAP. its all over the @#$@# moon (http://www.yfiles.com/helium3.htm) and is one of very few Aneutronic fusion reactions (aka: safe).
--Myren
Some of this is on aneutronic fusion (which talks about why aneutronic fusion maybe isn't as safe as it seems). BTW, do you know about the nonequilibrium fusion study (cited there)? It's worth reading as it essentially rules out a lot of the nice-sounding reactions. --Andrew 02:21, Apr 22, 2005 (UTC)