Procedures

добавляение новой точнки траектории

$y(i)=y(x(i))$ the data to be approximated.
$n$ number of points in the input data.
$m$ number of coefficients of decomposition over base functions $f(k,x)$ :
$y(x)=sum_1^m [b(k)*f(k,x)]$
$b(i)$ found decomposition coefficients

сохраняет массивы расределения и скорости

процедура закапывания

fst(j)=pnxlogjc04/pi4/vt3

pdec1=-pdecv*dfdv

calculate components of dielectric tensor

Given a,b,c(1:n), as output from routine chebft(), and given n, the desired degree of approximation (length of c to be used), this routine returns the array cder(1:n), the Chebyshev coefficients of the derivative of the function whose coefficients are c(1:n).

Chebyshev evaluation: All arguments are input. c(1:m) is an array of Chebyshev coefficients, the first m elements of c output from chebft (which must have been called with the same a and b). The Chebyshev polynomial evaluated and the result is returned as the function value.

Chebyshev fit: Given a function func, lower and upper limits of the interval [a,b], and a maximum degree n, this routine computes the n coefficients c(k) such that func(x) approximately = SUMM_(k=1)^(k=n)[c(k)*T(k-1)(y)]-c(1)/2, where y and x are related by (5.8.10). This routine is to be used with moderately large n (e.g., 30 or 50), the array of c’s subsequently to be truncated at the smaller value m such that c(m+1) and subsequent elements are negligible. Parameters: Maximum expected value of n, and ð.

создание сетки тепловых скоростей

! adde=zze(dd/dens(j))weight ! e2perp(i,j)=e2perp(i,j)+adde

вычисляет znakstart !!! w*dH/dw=wdhdw:

! зачем save ?????

!!! old variant:

!!! pn=fn1(pa,fnr) pn=fn2(pa,fnr,fnrr) !!!!!!!!!!!!!!!!!!!!!!!!!!!!! dl=bs2-4d0asbs

деление спектра на две части

pintld=-pintld4*dfdv

!old variant ! call raspr(v,abs(jr),iv,df) ! if(iv.eq.0) iv=1 !!!!!!!!!!!!!!!!!!!!!!!! dfsr=vdf(vrt-vlf) vsr=v(vrt-vlf) pintld=-dland(i)(dflf+dfrt)/2d0 pdec1=-pdecv*df

outj(i) = LH driven current density, MA/m^2 dndt(i) = d^2Jr1/dt^2/E, MA/m^2/sec^2/(V/m), ~runaway d(el.density)/dt/E djdt(i) = dJr2/dt, time drivative of runaway current Jr2, MA/m^2/sec outjrun(i) = LH driven runaway current density, MA/m^2 outnerun(i) = runaway electron density/10^19 m^-3

solve eqs. starting from xbeg ystart(1) = tet ystart(2) = xm x1 = xbeg rini x2 = xend

var if(dabs(dydx(3)).ne.zero) h=dabs(hmin1/dydx(3))/hdrob1 if(dabs(y(3)-rexi).gt.rrange.or.nstp.eq.maxstep4) then ! exit !sav2008

!!!

вычисление значения полинома и его производной

вычисление значения полинома и первой и второй производной

find achieved radial points jbeg-jend

распределение Максвелла с альфа-частицами $$f(v) = \frac{1}{\sqrt{2\pi}} \exp(-\frac{1}{2} v^2 (1.0 + \frac{1}{2} \cdot alfa_2 \cdot v^2))$$ и его производная $$dfmaxw = - v \cdot (1.0 + alfa_2 \cdot v^2) \cdot f(v)$$

$$alfa = \sqrt{alfa_2}$$ $$api = 2 \cdot alfa \cdot \exp({-\frac{1}{4 alfa_2}})$$ $$b = 2 - erf(0.5/alfa) + api$$ $$f = psiq(v, alfa_2)$$ $$fmaxw = \frac{f+api}{b \sqrt{2\pi}}$$

plasma density, cm^-3

plasma density and its derivative

plasma density and its first and second derivatives

конструктор для FokkerPlanck1D

инициализация диффузии для схемы савельева

!!!!!!!!!! solve problem !!!!!!!!!!!!!!!!!!!!!!!!!!

electron temperature, erg ft=y ! kev

ion temperature, kev

распределение Максвелла $$f(v) = \frac{1}{\sqrt{2\pi}} \exp(-\frac{1}{2} v^2 ))$$ и его производная $$dfmaxw = - v \cdot f(v)$$

нет описания

что-то про гаусс

v12<v<=v2

инициализация траетории

!!!!!! ptkev=ft(zero)/0.16d-8 !Te in keV ccurnr=pqepqe0.333d-9/pme

variant amy=(btor/q)rho(drho/dr) is a function of "minor radius" r=rh(i). Poloidal magnetic field: B_pol=amy(r)*sqrt(g22/g), where g is determinant of 3D metric tensor and g22 is the (22) element of the tensor, normalized on ABC^4 and ABC^2, correspondingly.

вычисление полной мощности спектра интегрирование методом трапеций

save Iteration Result to file

implicit real8 (a-h,o-z) write(,) write(,)'ccur',ccur,' curdir=',curdir,' nr=',nr write(,*)'cu_out, MA=',cu_out,' cfull, A=',cfull close(111)

линейная аппроксимация TODO будет удобнее если переделать в функцию

что делает?

что делает

approximation of input LH spectrum

memorize trajectory point

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!sav2008 pn=fn(r) pn=fn1(r,fnr) pn=fn2(r,fnr,fnrr) !sav2008 !!!!!!!read data !!!!!!!!!!!!
!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!sign of driven current in right coordinate system {dro,dteta,dfi}: !!!curdir=+1.0 for current drive in positive direction "dfi" !!!curdir=-1.0 for current drive in negative direction "dfi" !!!spectrum Nz>0 is along dfi>0 and Nz<0 is along dfi<0 !!!it is also OK if Npar is used instead of Nz, but for Btor>0, that is along dfi>0 curdir=-dble(ispectr) !!!!!!!!!!!!! begin iterations !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! find nevyazka

$$psiq=exp(ksiV**2)*erfcc(ksiV)*exp(-0.25/alfa2)$$

polynomial extrapolation

!!!!!!!!!!! read physical parameters !!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!! read parameters for alphas calculation !!!!!!!!!!!!!!!!!!! !!!!!!!!!!! read numerical parameters !!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!! read options !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!! read grill parameters and input LH spectrum !!!!!!!!!!!! !!!!!!!!!!! checking initial parameters !!!!!!!!!!!!!!!!!!!!!!!!!!!!!

recalculate f' for a new mesh ! if(vrj(i).gt.vvj(i0)) exit

сброс счетчика

метод рунге-кутта, нужны подробности

метод рунге-кутта

rational extrapolation зачем save ????

-- fill abc matrix

разностная схема Савельева для уравнения Фоккера-Планка

создает трехдиагональнйю матрицу

!writing trajectories into a file pintld=-dland(i)(dflf+dfrt)/2d0 pdec1=-pdecvdf

сохранение в файл информации о траектории

write lcms

write spectrum to file

сохраняет массивы расределения и скорости

z_effective profile