返回词条

阿尔芬波 · 历史版本

版本编辑者时间修改说明
第 6 版站方导入2026/11/9 06:58:15百科字段核查:fields.content
第 5 版站方导入2026/10/18 13:22:40按统一模板整理并保留字段来源
第 4 版站方导入2026/10/18 12:33:38按统一模板整理并保留字段来源
第 3 版站方导入2026/10/15 22:42:12按统一模板整理并保留字段来源
第 2 版站方导入2026/10/10 19:10:45根据原始来源更新词条
第 1 版站方导入2026/10/10 19:00:12根据原始来源更新词条

第 1 版

阿尔芬波

IV. NONLINEAR ION LANDAU DAMPING OF KINETIC ALFVEN WAVES

As noted in Sec. I, the results obtained in Sec. II are rather general and, therefore, are also useful for describing nonlinear processes associated with low-frequency wave heating. To illustrate this potential application, we consider here nonlinear ion Landau damping of the mode-converted kinetic (shear) Alfven waves. [1]^{[1]} The major difference between our work and Ref. 16 is that we include the diamagnetic drift effects. Another minor difference is that no small ion-larmor-radius expansion is taken here.

For the sake of simplicity, we adopt the WKB description, Eqs. (49) and (50), and ignore the geometrical effects. Furthermore, FoF_{o} is taken to be Maxwellian and for the present purpose we can assume a weak-turbulence subsidiary ordering as well as ignore the compressional Alfven perturbations δA1=0\delta A_{1} = 0 [i.e., β≡(plasma pressure/magnetic pressure)<<1\beta \equiv (\text{plasma pressure/magnetic pressure}) << 1 ]. Thus, the relevant equations are

δF~(k∼)=−(q/T)δΦ~(k∼)F0+δH~0(k∼)exp⁡(1k⋅y×e∥/Ω),(B1)\delta \widetilde {F} (\underset {\sim} {k}) = - (q / T) \delta \widetilde {\Phi} (\underset {\sim} {k}) F _ {0} + \delta \widetilde {H} _ {0} (\underset {\sim} {k}) \exp (1 k \cdot y \times e _ {\parallel} / \Omega),\tag{B1}

and

(ωk∼−k∥v∥)δHo(k∼)=(q/T)(ω+ω∗)k∼FoδLˉ(k∼)Jo(γ)−t(c/B)∑k′+k′′∼[δLˉ(k′∼)Jo(γ′)δHˉo(k′′∼)(k∼>k′∼)∗e∥],(82)\begin{array}{l} \left(\omega_ {\underset {\sim} {k}} - k _ {\parallel} v _ {\parallel}\right) \delta H _ {o} (\underset {\sim} {k}) = (q / T) \left(\omega + \omega_ {*}\right) _ {\underset {\sim} {k}} F _ {o} \delta \bar {L} (\underset {\sim} {k}) J _ {o} (\gamma) \\ - t (c / B) \sum_ {\underset {\sim} {k ^ {\prime} + k ^ {\prime \prime}}} \left[ \delta \bar {L} (\underset {\sim} {k ^ {\prime}}) J _ {o} (\gamma^ {\prime}) \delta \bar {H} _ {o} (\underset {\sim} {k ^ {\prime \prime}}) (\underset {\sim} {k} > \underset {\sim} {k ^ {\prime}}) * e _ {\parallel} \right], \end{array}\tag{82}

where δL~(k∼)=δϕ~(k∼)−v∥δA~∥(k∼)/c,ω∗k=kycT/eBLn,Ln−1={dln⁡N0/dx}\delta\tilde{L}(\underset{\sim}{k}) = \delta\tilde{\phi}(\underset{\sim}{k}) - v_{\parallel}\delta\tilde{A}_{\parallel}(\underset{\sim}{k})/c, \omega_{*k} = k_{y}cT/eBL_{n}, L_{n}^{-1} = \{d\ln N_{0}/dx\} , xx is the nonuniformity direction, and temperature gradients are ignored. Expanding δH~0=δH0(1)+δH0(2)+δH0(3)+…\delta\tilde{H}_{0} = \delta H_{0}^{(1)} + \delta H_{0}^{(2)} + \delta H_{0}^{(3)} + \ldots , we have, in the linear order,

δHo(1)(k)=qT(ω+ω∗)kωk−kvllFoδLˉ(k)Jo(γ).(83)\delta H _ {o} ^ {(1)} (k) = \frac {q}{T} \frac {\left(\omega + \omega_ {*}\right) k}{\omega_ {k} - k v _ {l l}} F _ {o} \delta \bar {L} (k) J _ {o} (\gamma).\tag{83}

To calculate δHo(2),(3)\delta H_{o}^{(2),(3)} , we let (ωk∼,k)(\omega_{\underset{\sim}{k}}, k) and (ωk∼′,k′)(\omega_{\underset{\sim}{k}^{\prime}}, k^{\prime}) be the normal modes, while (ωq=ωk∼−ωk∼′,q=k−k′)(\omega_{q} = \omega_{\underset{\sim}{k}} - \omega_{\underset{\sim}{k}^{\prime}}, q = k - k^{\prime}) is the virtual or quasi mode. Also, we have in mind

∣ωq∣∼∣q∥v1∣∼∣k∥,k∥′∣v1<<∣ω∗q∣∼∣ωk,k′∣<<∣k∥,k∥′∣ve.(84)\left| \omega_ {q} \right| \sim \left| q _ {\parallel} v _ {1} \right| \sim \left| k _ {\parallel}, k _ {\parallel} ^ {\prime} \right| v _ {1} < < \left| \omega_ {* q} \right| \sim \left| \omega_ {k, k ^ {\prime}} \right| < < \left| k _ {\parallel}, k _ {\parallel} ^ {\prime} \right| v _ {e}.\tag{84}

Let us now calculate the q-mode response. For the ions we find, noting Eq. (84), that

δHo1(2)(q∼)=[ω∗qe/(q∼−q∥v∥)]δP~q∼δΦ~(k∼)δΦ~(k′∼)Jo(γ)Jo(γ′)Fo1,(85)\delta \mathrm{H} _ {\mathrm{o} 1} ^ {(2)} (\underset {\sim} {q}) = [ \omega_ {* q e} / (\underset {\sim} {q} - q _ {\parallel} v _ {\parallel}) ] \delta \widetilde {P} _ {\underset {\sim} {q}} \delta \widetilde {\Phi} (\underset {\sim} {k}) \delta \widetilde {\Phi} (\underset {\sim} {k ^ {\prime}}) J _ {\mathrm{o}} (\gamma) J _ {\mathrm{o}} (\gamma^ {\prime}) F _ {\mathrm{o} 1},\tag{85}

where

δP~q‾=1(Ω1/ωk‾∗)ρs2[(k‾×k‾′)⋅e‾∥],(86)\delta \widetilde {P} _ {\underline {{q}}} = 1 (\Omega_ {1} / \omega_ {\underline {{k}} ^ {*}}) \rho_ {s} ^ {2} [ (\underline {{k}} \times \underline {{k}} ^ {\prime}) \cdot \underline {{e}} _ {\parallel} ] \quad ,\tag{86}

and δδ~≡eδϕ~/Te\delta\tilde{\delta} \equiv e\delta\tilde{\phi}/T_e . In deriving Eq. (A5), we have observed that ∣δϕ~∣∼∣ωδAˉ∣∣/ck∣∣|\delta\tilde{\phi}| \sim |\omega\delta\bar{A}_||/ck_|| for the k and k' modes. As to electron nonlinearity, it is 0(∣ω/ω∗e∣q)0(|\omega/\omega_{*e}|_q) smaller and, hence, negligible. Furthermore, for β≪1\beta \ll 1 , we have ∣ωq∣≪∣q∥vA∣|\omega_q| \ll |q_\parallel v_A| , and the q-mode response is predominantly electrostatic. Here VAV_A is the Aifven speed. The quasi-neutrality condition then yields

δΦ~q(1)=−(F1/Γoq)δP~qδΦ~(k)δΦ~(k′),(87)\delta \widetilde {\Phi} _ {q} ^ {(1)} = - (F _ {1} / \Gamma_ {o q}) \delta \widetilde {P} _ {q} \delta \widetilde {\Phi} (k) \delta \widetilde {\Phi} (k ^ {\prime}) \quad ,\tag{87}

where

F1(k,k′)=2π∫0∞v1dv1Jo(γ)Jo(γ′)Jo(γ)FM′(v1)≡⟨Jo(γ)Jo(γ′)Jo(γ)⟩⊥,(88)F _ {1} (k, k ^ {\prime}) = 2 \pi \int_ {0} ^ {\infty} v _ {1} d v _ {1} J _ {o} (\gamma) J _ {o} (\gamma^ {\prime}) J _ {o} (\gamma) F _ {M ^ {\prime}} (v _ {1}) \equiv \langle J _ {o} (\gamma) J _ {o} (\gamma^ {\prime}) J _ {o} (\gamma) \rangle_ {\perp},\tag{88} roq=Io(b)exp⁡(−b),(89)r _ {o q} = I _ {o} (b) \exp (- b),\tag{89}

rg=q⊥v⊥/Ωir_{g} = q_{\perp}v_{\perp}/\Omega_{i} and bg=q⊥2p12/2b_{g} = q_{\perp}^{2}p_{1}^{2}/2 . Equations (83) and (87) then give δHoi(1)(g)\delta H_{oi}^{(1)}(g) .

For the k-mode, the ion nonlinear response can be shown to be

δHo1(3)(k∼)=∣δP~q∼∣2∣δΦ~(k′∼)∣2δΦ~(k∼)Fo1⋅(ω∗eω−q,vj}q∼⋅[Jo2(γ′)Jo(γ)−F1Jo(γq∼)Jo(γ′)/Toq].(90)\begin{array}{r l} \delta H _ {o 1} ^ {(3)} (\underset {\sim} {k}) & = | \delta \widetilde {P} _ {\underset {\sim} {q}} | ^ {2} | \delta \widetilde {\Phi} (\underset {\sim} {k ^ {\prime}}) | ^ {2} \delta \widetilde {\Phi} (\underset {\sim} {k}) F _ {o 1} \cdot \left(\frac {\omega_ {* e}}{\omega - q _ {, v _ {j}}} \right\rbrace_ {\underset {\sim} {q}} \\ & \quad \cdot [ J _ {o} ^ {2} (\gamma^ {\prime}) J _ {o} (\gamma) - F _ {1} J _ {o} (\gamma_ {\underset {\sim} {q}}) J _ {o} (\gamma^ {\prime}) / T _ {o q} ] \quad . \end{array}\tag{90}

From Eq. (90) we can readily calculate δn1(3)(k∼)\delta n_{1}^{(3)}(\underset{\sim}{k}) . Meanwhile, the electron nonlinearity, again is found to be O(∣ω/ω∗e∣g∼)O(\left|\omega/\omega_{*e}\right|_{\underset{\sim}{g}}) smaller. Substituting δn1(3)(k∼)\delta n_{1}^{(3)}(\underset{\sim}{k}) and the linear response into the quasi-neutrality condition and parallel Ampere's law, it is straightforward to derive the following nonlinear dispersion relation

Dℓ(k∼)=−ω⋆eq∣q∥∣viziq[1−(1−ω⋆eω)k∼(ωk∥VAk⊥pe)k∼2]iδP~q.Φ~(k′)]2⋅G(k∼,k′∼),(91)\begin{array}{r l} {D _ {\ell} (\underset {\sim} {k})} & {= - \frac {\omega_ {\star e q}}{| q _ {\parallel} | v _ {i}} z _ {i q} [ 1 - (1 - \frac {\omega_ {\star e}}{\omega}) _ {\underset {\sim} {k}} \left(\frac {\omega}{k _ {\parallel} V _ {A} k _ {\perp} p _ {e}}\right) _ {\underset {\sim} {k}} ^ {2} ] i \delta \widetilde {P} _ {q}. \widetilde {\Phi} (k ^ {\prime}) ] ^ {2}} \\ & {\quad \cdot G (\underset {\sim} {k}, \underset {\sim} {k ^ {\prime}}) \quad ,} \end{array}\tag{91}

where Dℓ(k)D_{\ell}(k) is the linear dispersion relation given by

Dj(k)=(τ+ω∗e/ω)k(1−rok)+(1−ω∗e/ω)kD _ {j} (k) = (\tau + \omega_ {* e} / \omega) _ {k} (1 - r _ {o k}) + (1 - \omega_ {* e} / \omega) _ {k} ⋅[1−ωk∼2(τ+ω∗e/ω)k∼(1−rok)/(k∥vAk1p8)2],(92)\cdot [ 1 - \omega_ {\underset {\sim} {k}} ^ {2} (\tau + \omega_ {* e} / \omega) _ {\underset {\sim} {k}} (1 - r _ {o k}) / (k _ {\|} v _ {A} k _ {1} p _ {8}) ^ {2} ] \quad ,\tag{92} G(k∼,k′∼)=⟨J02(γ)J02(γ′)⟩⊥−F12/Γoq,(93)G (\underset {\sim} {k}, \underset {\sim} {k ^ {\prime}}) = \langle J _ {0} ^ {2} (\gamma) J _ {0} ^ {2} (\gamma^ {\prime}) \rangle_ {\perp} - F _ {1} ^ {2} / \Gamma_ {o q},\tag{93}

zig≡z(ω/g/∣qij∣vi)z_{ig} \equiv z(\omega/g / |q_{ij}| v_{i}) with Z being the standard plasma dispersion function.

In the low- β\beta limit of interest here, ∣ω∗e/kIVA∣k~<1\left|\omega_{*e}/k_{I}V_{A}\right|_{\widetilde{k}}<1 , and assuming ωk=ωrk+iωik\omega_{k}=\omega_{rk}+i\omega_{ik} with ∣ωl/ωr∣<<1\left|\omega_{l}/\omega_{r}\right|<<1 , Eq. (92) gives the known result that there exist two branches of waves; the drift waves with

ωrk,d=ω∗ekΓok/[1+τ(1−Γok)],(94)\omega_ {\mathbf {r k}, d} = \omega_ {* e k} \Gamma_ {o k} / [ 1 + \tau (1 - \Gamma_ {o k}) ] \quad ,\tag{94}

and the kinetic shear Alfven waves with

ωrk∼,A2=(k∥∇A)2[1+τ(1−Γok)]k12ρ82/τ(1−Γok).(95)\omega_ {r \underset {\sim} {k}, A} ^ {2} = (k _ {\parallel} \nabla_ {A}) ^ {2} [ 1 + \tau (1 - \Gamma_ {o k}) ] k _ {1} ^ {2} \rho_ {8} ^ {2} / \tau (1 - \Gamma_ {o k}) .\tag{95}

We note that nonlinear ion Landau damping of the drift waves, as described by Eq. (91), has previously been investigated. 17^{17} Concentrating on the kinetic shear-Alfven waves, we find

(ω1ωr)k,A=(ω∗k′−ω∗k)e2τ(1−Γok)[1+τ(1−Γok)]G(k∼,k′∼)Im⁡zij∣qivi∣∣δα~(k′∼)δF~q∣2(96)\left(\frac {\omega_ {1}}{\omega_ {r}}\right) _ {k, A} = \frac {\left(\omega_ {* k ^ {\prime}} - \omega_ {* k}\right) _ {e}}{2 \tau (1 - \Gamma_ {o k}) [ 1 + \tau (1 - \Gamma_ {o k}) ]} G (\underset {\sim} {k}, \underset {\sim} {k ^ {\prime}}) \frac {\operatorname{Im} z _ {i j}}{| q _ {i} v _ {i} |} | \delta \widetilde {\alpha} (\underset {\sim} {k ^ {\prime}}) \delta \widetilde {F} _ {q} | ^ {2}\tag{96}

Comparing the ωi\omega_{i} given by Eq. (96) to that of Ref. 16 (which, we recall, does not include the ω∗\omega_{*} effects), we find the present parametric-decay growth rate is larger by a factor of 0(∣ω∗iq∣/∣q∥vi∣)>10(|\omega_{*\mathbf{iq}}| / |q_{\parallel}v_{i}|) > 1 . Furthermore, noting that Im⁡Ziq=∣qi∣viπδ(ωk′−ωk)\operatorname{Im} Z_{\mathbf{iq}} = |q_{i}| v_{i} \pi \delta (\omega_{k'} - \omega_{k}) , G(k,k′)≥0G(k, k') \geq 0 , and ωi∝(ω∗k′−ω∗k)∝(ky′−ky)\omega_{i} \propto (\omega_{*k'} - \omega_{*k}) \propto (k'_{y} - k_{y}) ; hence, the daughter wave (ωk,k)(\omega_{k}, k) has a smaller kyk_{y} but the same ω\omega as the pump wave (ωk′,k′)(\omega_{k'}, k') . This is qualitatively different from the results of Ref. 16, where ωi∝(ωk′−ωk)\omega_{i} \propto (\omega_{k'} - \omega_{k}) . In fact, as might be expected, one can show that the results of Ref. 16 are valid for ∣ω∗iq∣<∣q∥vi∣|\omega_{*\mathbf{iq}}| < |q_{\parallel}v_{i}| , i.e., in the opposite limit.

Finally, we briefly discuss the implication of the present results on the shear Alfven wave heating scheme. Since the pump wave (i.e., the mode-converted kinetic shear Alfven wave) was kx′ρ1∼0(1)k_x' \rho_1 \sim 0(1) and ky′ρ1∼0(ρ1/a)≪1k_y' \rho_1 \sim 0(\rho_1/a) \ll 1 with a being the tokamak minor radius, we have ωi=∣k×k′⋅en∣2=(kx′ky)2\omega_i = |k \times k' \cdot e_n|^2 = (k'_x k_y)^2 , and hence nonlinear coupling becomes appreciable for kyρ1∼0(1)>>ky′ρ1k_y \rho_1 \sim 0(1) >> k_y' \rho_1 ; that is, ωi<0\omega_i < 0 . Thus, our results suggest that parametric decay through nonlinear ion Landau damping will be, in general, unlikely to occur.

5.2. The kinetic Alfvén wave ion-ion streaming instability

For our second test, we consider the excitation of kinetic Alfvén waves (KAW) by ion-ion streaming in 2D-3V [54]. This is a moderately magnetized case, with an imposed external magnetic field. The instability is caused by interactions between the wave and the streaming ions. The simulation parameters are similar to those used in Ref. [54]. The mass ratio is mi/me=25m _ { i } / m _ { e } = 2 5 , which is nominal (we will vary this parameter later). We use ions as the reference species. The simulated domain is 10(di)×10(di)1 0 ( d _ { i } ) \times 1 0 ( d _ { i } ) (the unit length being the ion skin depth), with 64 64 uniformly distributed cells (with each cell size in each direction about 10 times larger than the Debye length) and periodic boundary conditions, and the average number of particles per cell of one species is 500. The external magnetic field is set to be B0=0.0667B _ { 0 } = 0 . 0 6 6 7 along the x-axis. The plasma consists of Maxwellian electrons with veT=0.0745 (βe=0.1)v _ { e T } = 0 . 0 7 4 5 ~ ( \beta _ { e } = 0 . 1 ) , and two singly charged ion components, i.e., an ambient ion component a and an ion beam component b,b , with number densities na=0.6nen _ { a } = 0 . 6 n _ { e } and nb=0.4nen _ { b } = 0 . 4 n _ { e } (where nen _ { e } is the electron density). The two ion components have vaT=1.925×10−2v _ { a T } = 1 . 9 2 5 \times 1 0 ^ { - 2 } and vbT=7.45×10−3v _ { b T } = 7 . 4 5 \times 1 0 ^ { - 3 } , and a relative streaming speed with respect to each other of vad=1.5vAv _ { a d } = 1 . 5 v _ { A } and vbd=vA{ v _ { b d } } = { v _ { A } } , with vA=me/mi/3v _ { A } = \sqrt { m _ { e } / m _ { i } } / 3 the Alfvén speed along the external magnetic field direction. The timestep is again set to Δt=0.1ωpi−1\Delta t = 0 . 1 \omega _ { p i } ^ { - 1 } (about 10 times larger than the explicit CFL for the mass ratio considered). The simulation is started without perturbing a specific wavelength. Consequently, waves with all wavelengths and with all angles to B0B _ { 0 } supported by the simulation domain are excited. Figure 6 shows the simulation result of the magnetic energy density, which is again in excellent agreement with linear theory (the growth rate for this configuration is found to be γ=0.225\gamma = 0 . 2 2 5 , using the same linear Vlasov code as in Ref. [55]).

Table 3: Solver performance for the KAW case with and without preconditioning for Lx×Ly=22(di)×22(di)L _ { x } \times L _ { y } = 2 2 ( d _ { i } ) \times 2 2 ( d _ { i } ) Nx×Ny=32×32N _ { x } \times N _ { y } = 3 2 \times 3 2 , and Npc=200N _ { p c } = 2 0 0 . For all the test cases, Δt=0.1ωpi−1\Delta t = 0 . 1 \omega _ { p i } ^ { - 1 } . Convergence data is averaged over 10 time steps.

<table><tr><td rowspan="2"> $m_i/m_e$ </td><td colspan="2">no preconditioner</td><td colspan="2">with preconditioner</td></tr><tr><td>Newton</td><td>GMRES</td><td>Newton</td><td>GMRES</td></tr><tr><td>25</td><td>4</td><td>171.9</td><td>3.2</td><td>1</td></tr><tr><td>150</td><td>4.5</td><td>764</td><td>4</td><td>2.9</td></tr><tr><td>600</td><td>7.4</td><td>4054.8</td><td>4</td><td>11.9</td></tr></table>

Table 4: Solver performance for the KAW case with and without preconditioning, with fixed ωpe/ωce=3.\omega _ { p e } / \omega _ { c e } = 3 . Other parameters are the same as in Table 3. (NC denotes “no convergence”.)

<table><tr><td rowspan="2"> $m_i/m_e$ </td><td colspan="2">no preconditioner</td><td colspan="2">with preconditioner</td></tr><tr><td>Newton</td><td>GMRES</td><td>Newton</td><td>GMRES</td></tr><tr><td>150</td><td>4.5</td><td>738</td><td>4</td><td>3</td></tr><tr><td>600</td><td>5.8</td><td>1887</td><td>4</td><td>3.9</td></tr><tr><td>1836</td><td>NC</td><td>NC</td><td>4</td><td>5.9</td></tr><tr><td>7344</td><td>NC</td><td>NC</td><td>4</td><td>12.9</td></tr></table>

参考资料