! FPBASIC.DFT Basic default namelist parameters for FPP ! Greg Hammett 29-Oct-1987 ! $fppin comments=60*' ' ! up to 60 lines * 80 characters of comments nshell=10 !# of radial shells to be used in the FPP calculations ! nshell=20 is the current maximum. 10 shells is a ! reasonable value. Note that the beam deposition ! calculation time scales like nshell**3! ! Multiple ion species information: The first ion in this list is ! assumed to be the "main" thermal ion species, for which FRANTIC will ! calculate a neutral density. nions=3 ! actual number of ions species used (maximum is MXIONS) nAions=2,1,12 ! list of Atomic Mass A_i of each ion species nZions=1,1,6 ! list of Atomic charge Z_i of each ion species namions='D','H','C' ! label for each ion species (CHARACTER*6) nfasti=2 ! index i of the fast ion to be Fokker-Planck'd ! nfasti is the ion species for which we will solve the bounce-averaged ! Fokker-Planck/Quasilinear equation. This could be the fast ions ! produced by neutral beam injection or ICRF minority heating, or could ! be the Maxwellian plus energetic tail for second harmonic ICRF ! heating of a majority ion species. ! Can specify the complete density and temperature profiles for ! each species (This option is used if Te0=0.0): Trelec= 20*0.0 ! T_e(r) electron temperature profile (eV) ! trelec(maxshe) drelec= 20*0.0 ! n_e(r) electron density profile (/cm**3) ! drelec(maxshe) Trions=20*0.0, 20*0.0, ! T_i(r,i) ion temperature profiles (eV) 20*0.0 ! trions(maxshe,mxions) Trionspa=60*-1.0 ! T_i_parallel(r,mxions) (eV) ! if < 0, will be set to Trions drions=20*0.0,20*0.0, ! n_i(r,i) ion density profiles (/cm**3) 20*0.0 ! drions(maxshe,mxions) ! ! Or, if Te0>0 then FPP will use parabolic-to-a-power formulas for the ! temperature and density profiles. The parabolic formulas are of ! the form: ! Te(r) = (Te0-Tea)*(1-r**2/a**2)**Teexp + Tea ! Te0=0.0 !central Te(0) for parabolic formula (eV) Tea=0.0 !edge Te(a) for parabolic formula (eV) Teexp=0.0 !exponent for parabolic formula for Te Delec0=0.0 !central ne(0) for parabolic formula (/cm**3) deleca=0.0 !edge ne(a) for parabolic formula (/cm**3) delecexp=0.0 !exponent for parabolic formula for ne Ti0=10*0.0 !list of ions' Ti(0) for parabolic formula (eV) Tia=10*0.0 !list of ions' Ti(a) for parabolic formula (eV) Tiexp=10*0.0 !list of ions' parabolic exponent for Ti(r) ! ! Ti0,Tia, and Tiexp are to be specified for each ion species. ! rations=10*0.0 ! fraction n_i(r)/n_e(r) for each ion species rzeff=0.0 ! >0 will adjust the impurity and main ion fractions ! impurity is the largest Z ion, main is the first ion. ! LAntic=1 ! 1=FRANTIC 2=GNUSUB 3=flat 0=no neutrals AN00=1.e8 !edge neutral density for FRANTIC B.C. (/cm^3) Tedg=0.25 !edge neutral temperature for FRANTIC B.C. (eV) ! !********************************************************************** ! ! Description of plasma geometry: ! ! There are several possible models for the plasma geometry. ! To use a simpler model, leave all of the parameters of the higher ! model set to zero. ! Rmaj=0.0 ! Major radius of the plasma outer surface (cm) ! (Btor is evaluated at Rmaj, so Rmaj must be set ! for all geometry models.) ! ***************** The simplest plasma geometry model**************** ! is for an elliptically shaped cross-section with half-width a ! and half-height b: rmin=0.0 ! area-equivalent plasma minor radius (sqrt(a*b)) (cm) fpshift0=0.0 ! Magnetic axis Shafranov shift relative to Rmaj (cm) fpellip=1.0 ! ellipticity of plasma (b/a) ! ************* A more complicated plasma geometry model ************** ! allows for a general fourier moments representions of ! the outer plasma boundary, with parabolic extrapolations ! to the center. The plasma boundary fourier moments ! have the same defitions as the VV fourier moments, so see the ! definitions of VVRmoms and VVZmoms below: ! plasma boundary Fourier moments (in cm) for i=0 to mxgmom: ! (mxgmom=5 at the moment). pRmoma(0)=6*0.0 ! R(th) = Sum pRmoma(i) cos(i th) pZmoma(0)=6*0.0 ! Z(th) = Sum pZmoma(i) sin(i th) ! pZmoma(0) must be 0.0 !fpshift0=0.0 ! Magnetic axis Shafranov shift (cm) ! (fpshift0 is also in the simplest model above.) fpellip0=1.0 ! ellipticity near plasma center expmom=1.0 ! parabolic exponent used in extrapolating edge shape ! to the center. For example, the shafranov shift as a ! function of minor radius will scale like ! fpshift0*(1-r**2/a**2)**expmom. !Rmaj=0.0 ! Major radius of the plasma outer surface (cm) ! (Btor is evaluated at Rmaj, so Rmaj must be set ! for all geometry models.) ! ********** A completely general plasma geometry model ************** ! Allows the fourier moments to be general functions of minor ! radius (defined here at the outer boundary of each radial zone). ! (FPP's minor radius coordinate is equally spaced in ! Sqrt(Toroidal Flux). Since we are ignoring diamagnetic effects, ! this is equivalent to saying that the minor radius is equally ! spaced in Sqrt(Cross-sectional area).) ! ! plasma Fourier moments (in cm) for ir=0 to maxshe and i=0 to mxgmom: ! (mxgmom=5 at the moment, must also set moments for the magnetic axis ! at ir=0). (126=21 radial * 6 moments:) pRmom(0,0)=126*0.0 ! R(ir,th) = Sum pRmoma(ir,i) cos(i th) pZmom(0,0)=126*0.0 ! Z(ir,th) = Sum pZmoma(ir,i) sin(i th) ! pZmoma(0) must be 0.0 !Rmaj=0.0 ! Major radius of the plasma outer surface (cm) ! (Btor is evaluated at Rmaj, so Rmaj must be set ! for all geometry models.) ! ! All of FPP has been upgraded to general plasma geometry except for ! the radial transport routines RDJACO, RDOPER, RDSTEP, and RADNEO, ! and the adhoc rf power profile model RFPROF, which still work ! in volume equivalent cylindrical coordinates. ! (The full RF power profile model SPRUCE has been upgraded for ! general geometry.) ! ! Also note that FPP is presently assuming B = Btor*Rmaj/R, ignoring ! diamagetic and poloidal field corretions to B. ! ! Also, major parts of the code still assume that the maximum R ! occurs at theta=0, and the minimum R occurs at theta=pi, i.e., ! the code assumes that there is only a single magnetic well, ! i.e., the code ignores any particles trapped in an indentation. ! !********************************************************************** Btor=0.0 ! Toroidal magnetic field strength at Rmaj (Gauss) ! a simple model for the inverse rotation transform q, ! based on the central a and edge q: qax=1.0 ! q(r=0) q on axis. Placur=0.0 ! Plasma current Ip (kAmps) used only to find the edge q qpro=20*0.0 ! or a full q(r) profile can optionally be specified here ! (leave qpro(i)=0.0 to use the simpler q model.) ! (qpro(i) is q at the outer boundary of radial zone i) RinWal=0.0 ! Major Radius of the inner wall (cm) ! Vloopfp=0.0 ! Loop voltage around the torus (Volts) cratio=1.0 ! compression ratio ! !------------------------------------------------------------------------ ! Neutral Beam injection parameters: ! numnbi=0 ! The number of neutral beam injector sightlines in use. ! numnbi must be .le. maxnbi !p ra rbeams(maxnbi) list of tangency radii of the beams (cm) !p ra vbeams(maxnbi) list of injection voltages of the beams (Volts) !p ra pbeams(maxnbi) list of injection powers of each beam (MegaWatts) !p ra Bmixfull(maxnbi) fraction of the beam particles at the full energy !p ra Bmixhalf(maxnbi) fraction of the beam particles at the full energy trise=0.0 ! The beam current rise time (s) !p r toff Time at which the beams go off (s) !p r FallR Beam Gaussian half width in R (cm) !p r FallZ Beam Gaussian half height in Z (cm) !p r BmEdgHW Beam cutoff half width (cm) !p r BmEdgHH Beam cuttoff half height (cm) ! !------------------------------------------------------------------------ ! CX Detector parameters (set for TFTR's vertically viewing detectors): ! Ldir=0 ! direction of CX detectors (1=horizontal,0=vert) nline=0 ! # of CX detector sightlines (0 for none) ! nline must be .le. mxslin ! must also satisfy (nline*nencx*nshell*maxxsi/2 < mxsens =20000) ! tans = 194 208 222 244 270 297 310 323 ! Rtan's of CX detectors (cm) ! rcrs and phicrs are used only if 3-D neutrals are present: RCrsfp=0.0 ! MAJOR RADIUS AT WHICH ALL SIGHT LINES CROSS. (CM) PhiCrsfp=0.0 ! PHI ANGLE AT WHICH ALL SIGHT LINES CROSS. (DEGREES) Lbeam=0 !0=ignore 1=read 3-D neutral profiles made by NBEAM ! This option is so old I am not sure it works right anymore. Leave ! Lbeam=0 to use simple, toroidally symmetric neutral density profiles. ! Elim2=150.e3 ! Maximum energy for the CX spectra calculation (eV) bkgmax=0.0 ! turn 0.0=off/RF 1.0=on/NBI the background Maxwellian ! ! subroutine spect adds in a background maxwellian when calculating the ! cx spectrum. bkgmax gives the density of the background maxwellian as a ! fraction of main plasma ion density. bkgmax=1 includes the regular ! background for NBI cases, while bkgmax=0 is appropriate for ICRF cases ! where the "maxwellian component" is already included in the f ! calculated by FPPRF. ! ! nencx=0 ! # of energies in CX spectra (=0 to default to full spectrum) nshcx=0 ! find CX spectra due to only radius nshcx (=0 for all) ! ! run CX detection routines with arbitrary neutral density profile: Ldnb=.f. !.t.= use arbitrary n0(R), .f. = use frantic n0(r) dnbneu=20*0.0 ! arbitrary n0(R) for CX detection (/cm^3) ! dnbneu(2*maxshe) dnbRmj=40*0.0 !as a function of Major Radius (cm) ! dnbRmj(2*maxshe) ndnb=0 ! # of points in the dnbneu profile ! actually, this is an arbitrary n0(R) of major radius only for ! horizontal CX detectors (Ldir=1). For vertically viewing (Ldir=0) ! CX detectors, this is n0(r) of minor radius. Latten=.t. !.t.=include .f.=ignore standard signal attenuation ! !----------------------------------------------------------------------- ! controls on the CX spectrum plots: ! LlnFpl=1 ! 1 = new size ln(f) plot 0=old size log10(f) plot fmin=20. ! min ln(f) on plot (f is in Davis-Mueller-Keane units) fmax=40. ! max ln(f) on plot Emax=0.0 ! maximum E on plot (0= default to Elim2) (eV) NxminCX=100 ! xmin of plot in screen units (to overlay data exactly) NxmaxCX=900 ! xmax of plot in screen units (to overlay data exactly) NyminCX=120 ! ymin of plot in screen units (to overlay data exactly) NymaxCX=720 ! ymax of plot in screen units (to overlay data exactly) LCXRTP=0 ! plot vs. Rtan: 0=nothing 1=ln(f_cx), 2=linear f fcxlmax=0.0 ! max linear f_cx (for LCXRTP=2), 0=self-scale Rtmin=0.0 ! Min Rtan in f_cx vs. Rtan Plot (cm) Rtmax=0.0 ! Max Rtan in f_cx vs. Rtan plot (cm) ! !------------------------------------------------------------------------ ! FISHBONE MODEL ! LFishMdl=0 !Fishbone model 0=off. See FISH.FOR for other options tfish=0.0 !TIME BTWN FISHBONES (SEC'S) wfish2=0.0 !FULL WIDTH OF LOSS REGIONS IN VPAR/V FOR MODEL 2 wfish3=0.0 !FULL WIDTH OF LOSS REGION IN VPAR/V FOR MODEL 3 fshfrc=0.0 !FRACTION OF PARTICLES IN LOSS REGION THAT SURVIVE A FISH-BONE efmin=0.0 !MINIMUM ENERGY AFFECTED BY A FISH-BONE efmax=0.0 !MAXIMUM ENERGY AFFECTED BY A FISH-BONE rfmin=0.0 !MINIMUM MINOR RADIUS AFFECTED BY A FISH-BONE rfmax=0.0 !MAXIMUM MINOR RADIUS AFFECTED BY A FISH-BONE ! !------------------------------------------------------------------------ ! ICRF parameters: ! prftot=0.0 ! total input rf power (Watts) freqrf=47.0e6 ! frequency of the RF generator (Hz) rkpar=0.07 ! abs(kpar) AT THE ANTENNA (+/- kpar used in code) (1/cm) nharm=1 ! 1 = fundamental, 2 = 2cd harmonic ICRF heating concmi=0.0 ! (0.0=default) nmin/ne for the fundamental minority species ! defaults to the same value as cbeam unless the "beam" is a 2cd harmonic ! majority species. concmi is used to normalize the P/n profile in ! RFPROF, and is used only if lshoot=0.0. LSHOOT=1 ! ICRF model 1 or -1 =Smithe's SPRUCE 2=SHOOT 0=Guess ! LSHOOT=1 or -1 Smithe's general geometry fast-wave propagation code ! LSHOOT=2 Smithe and Colestock's circular fast-wave code ! LSHOOT=0 Hand-entered guesses at the ICRF power profile, etc. ! LSHOOT = -1, 1 or 2 requires the following parameters ! (which are currently set to TFTR parameters, corrected 9/22/89): ! antRcen=364. ! major radius R of the antenna center (cm) antHH=40. ! antenna half height (cm) ! The antenna is assumed to be curved to lie parallel to the ! vacuum vessel. Its length (along its curvature) may be longer ! than its height (along a straight line from tip to tip). ! For a low-field launch, antRcen should be close to the ! outer VV wall. For a high-field launch, antRcen should be ! close to the inner VV wall. ! antwidth=10. ! toroidal width of each antenna strap (cm) antsep=37. ! toroidal separation between the strap centers (cm) ! antsep is measured from the center of one strap to the center ! of the next. SPRUCE assumes that the 2 straps are out of phse, ! so rkpar should be approximately set to pi/antsep. ! ! SPRUCE presently assumes that the antenna current has a poloidal ! dependence of cos(1.3*k*r*theta) on the antenna (and zero beyond the ! antenna), where k is the free space wavenumber. The factor of 1.3 is ! an approximate fudge for Faraday shield effects, etc. The current ! profile can be adjusted by modifying a few lines in ZDILE.FOR. vacden=1.e11 ! "vacuum" plasma density (/cm**3) ! by raising vacden a bit, one can sometimes reduce ! the edge-heating which sometimes occurs during central He3 ! minority heating due to the shear-Alfven/fundamental-D ! resonance at the high-field edge. This shear-Alfven ! resonance, and whether or not it is real, is not well ! understood. vactemp=100. ! "vacuum" plasma temperature (eV) ! finite plasma parameters are needed in the "vacuum" between the ! plasma edge and the vacuum vessel to avoid complications in ! the ICRF wave propagation calculation. ! rkpcorr=1.0 ! irrelevant as an input variable. ! ! LSHOOT = 0 requires the following guesses at ICRF fields: ! Erfpro(1)=20*1. ! ERFpro(maxshe) is E_plus(r) or Prf(r) LRFmod=1 ! ERFPRO is actually (1 or 3)=E-plus 2=Power profile shape ! If LRFmod=1, then ERFpro is just the Eplus(r) profile shape, its ! magnitude will be renormalized to give the specified PRFTOT, using ! the Stix formula for Prf. ! If LRFmod=3, then ERFpro is the actual Eplus(r), including magnitude. ! LKmodel=0 !k_perp model 0= cold plasma k_perp +-2=input rkperp(r) ! (-2 will turn off Eminus for benchmarking purposes.) rkperp=20*0. ! ICRF k-perp(r) profile (cm^-1) llstix=.f. ! .f. = Best log(Lambda) formulas. ! .t. = Stix's log(Lambda) formulas (for benchmarking). !************************************************************* !VV shape description in Fourier moments (in cm) for i=0 to mxvvmoms: ! (mxvvmoms=5 at the moment). ! ! Suggested parameters for TFTR's near circular VV: VVRmoms(0)=265.,103.,0.,0.,0.,0. ! R(th) = Sum VVRmoms(i) cos(i th) VVZmoms(0)=0.0, 103.,0.,0.,0.,0. ! Z(th) = Sum VVZmoms(i) sin(i th) ! VVZmoms(0) must be 0.0 ! A 2-moment fit to the JET VV: !VVRmoms(0)=292.8, 134.5, 10.4 ! R(th) = Sum VVRmoms(i) cos(i th) !VVZmoms(0)= 0.0, 209.6, -16.2 ! Z(th) = Sum VVZmoms(i) sin(i th) ! ! For comparison, the Lao-Hirschman 2-moment representation is of ! the form: ! ! R(th) = R0 + a cos(th) + R2 cos(2 th) ! Z(th) = E ( a sin(th) - R2 sin(2 th) ) ! ! Describing the VV shape with the 4 parameters Rmin, Rmax, ! Rtop, and Ztop (these last two give the position of the top of ! the VV) we can calculate the Lao-Hirschman coefficients via: ! ! a = (Rmax-Rmin)/2 ! Rx = (Rmax+Rmin)/2 ! ! The next two equations can be solved by combining them together ! to make a cubic equation. Or for small d they can be solved ! iteratively by using R2=0.0 as an initial guess: ! ! d = (Rx-R2-Rtop)/a ! a measure of "D-ness" or triangularity ! R2 = 3 a d / (9 - 8 d**2) ! ! Once the above two equations are solved, we can then find: ! ! R0 = Rx-R2 ! E = 3 (Ztop/a) (9 - 8 d**2) / (9 - 4 d**2)**(3/2) ! ! These TFTR VV parameters are of neccessity approximate: ! ! Greg Hammett, 23-Sep-1989: ! (based on conversations with Mike Bell and Cynthia Phillips and ! and Kingston Owens.) ! ! Here are a few relevant positions in TFTR (note these distances ! are in m, while FPP wants distances in cm): ! ! R=1.507 m Inner part of Vacuum Vessel ! R=1.62 m Inner bellows cover plates (but they've been removed for ! the carbon inner bumper limiter) ! R=1.656 m Inner bumper limiter ! R=3.598 m Outer RF limiter ! R=3.64 m RF antenna (approximate and variable) ! R=3.68 m Outer part of bellows covers plate (but they've been ! removed from the bays which have antennas). ! R=3.793 m Outer part of Vacuum Vessel ! ! But to put the conducting wall boundary conditions in the ! approximately right place for the ICRF code (using the bellows ! covers plate as the relevant conductor), one might set: ! ! VVRmaj=265. ! VVamin=103. ! ! If you use the VV instead as the relevant conductor, you get: ! ! VVRmaj=265. ! VVamin=114.3 ! !Suggested defaults for TFTR were: !VVamin=103. ! Vacuum Vessel minor radius (cm) !VVRmaj=265. ! Vaccuum vessel major radius (cm) ! VVamin and VVRmaj are obsolete, use VVRmoms and VVZmoms. ! antamin and anttheta are obsolete, use antRcen and AntHH. !End of VV shape description. !************************************************************* ! !------------------------------------------------------------------------ ! Basic FPPRF control parameters: ! tstart=0.0 ! starting time of the calculation (s) Tcalc=100.e-3 ! duration of the FPP calculation (=0 to skip). (s) dtFPP=5.e-3 ! time step of all plotting outputs. (s) nmulti=1 ! (array) number of numerical time steps per dtFPP ! for accuracy, the numerical time step dtFPP/nmulti should be about ! 1/10 (or smaller) of the slowing down time. ! ! in reality, IMP gets called 2*nmulti times between each MOMS call. ! so the true internal time step is dtFPP/nmulti/2. nmulti can ! be an array so each radial shell can have its own internal time step. ! This allows you to have more time steps on certain radii which need ! a shorter time step to maintain stability. ! LGrid=0 ! Energy grid model: 1=uniform, 0=variable EgrMax=2.e5 ! Maximum grid energy if the grid is uniform (eV) eminev=0.0 ! Emin for variable grid 0=standard (eV) deove=.1 ! dE/E for variable grid 0.1=standard demin=400 ! dE_min for var. grid (eV) (E(82)=1.e6*(dE_min/100) if ! deove=.1 and demax=infinity demax=1.e6 ! de_max for variable grid 1.e6=standard (eV) ! ! Useful formulas for choosing the grid spacing is: ! I=Int(1./deove) = index where dE_min and dE/E grids meet ! dEmin=Emax/I/(1+deove)**(maxen-I) ! where demax=infinity is assumed, and Emax=E(maxen), and the ! energy grid E(i) goes from i=0 to i=maxen. ! ! For meaningfulness, E(maxen) needs to be at least 5 times the final ! minority temperature. For accuracy, E(maxen)=10*Tmin is probably better. ! Batchm=.t. ! .t. = batch mode, do not prompt user for plot limits LLWbnd=1 ! use 1 = conservative (RF) 0=sink (NBI) lower B.C.'s nskipCX=1 ! number of skips between SPECT nplte=1 ! number of CX flux vs. E plots npltt=1 ! number of CX flux vs. time plots npltf=0 !number of f(v-precession) plots LPlotE=0 !0 = turn off plots of f(E) xmax=0.0 !upper x limit for f(E) plot ymin=1.e-4 !lower y limit for f(E) plot LPlotA=0 !0 = turn off plots of f(angle) LPlotF=0 !0 = turn off contour plots of f LPlotH=1 !1=do 0=don't plot Neutral Beam H(R)'s LPloD=0 ! 0= don't 1= do plot detector sensitivities NiAngls=50,40,26,11,1 !angle indices for f(e) plot ! NiAngls(5) (reset by FPP for rf heating) Effoa=55.e3,45.e3,35.e3,25.e3,15.e3 ! Energies for f(a) plots (eV) ! effoa(5) nshsta=1 !starting radial shell for FPP calculation (1=normal) nshsto=0 !stopping radial shell for FPP calculation (0=normal) ! if nshsto=0 initially, FPP will set it to NSHELL ! Lfget=.f. ! .true. to read f(E,xsi,r) from disk Lfsave=.f. ! .true. to save f(E,xsi,r) on disk in binary. ! (saves f in a binary format so that FPP can read it ! back in with Lfget=.t. in a later run.) Lfsavea=.f. ! .true. to save f(E,xsi,r) in ASCII for Rewoldt's ! gyrokinetic microinstability code. Lcxtouf=.f. ! .true. to save CX spectra in UFILEs (mostly for MIT.) cxonly=.f. !.f. = normal FPP, .t. = calculate CX spectrum from an old f ! lradls=.f. !.f. = turns off a Krook model for radial diffusion lcxlos=.t. !.t. = turn on the CX loss operator lcxsrc=.t. !.t. = turn on the source part of the CX operator lnsrc=.t. !.t. = turn on a source to maintain density lbador=.t. !.t. = turn on bad orbit losses fufiti=1.0 !1. = Include finite Ti effect for beam-target fusion ! for D-D, use Rob's formula for Beam + finite Ti target ! for others, use very simple estimate. Lbbfuf=0 !(disabled) non zero to calculate beam-beam reactions LYTypFUS=2 !1 = linear, 2 = log neutron flux scale ! !------------------------------------------------------------------------ ! Fast Ion Radial transport: ! ! FPP will now calculate radial transport of fast ions. Two important ! components of this transport are a convective velocity v_r due ! to the asymmetric edge drag mechanism discussed in my thesis and ! my 1986 APS invited talk, and D_r due to the Goldston and Towner ! banana drift ripple diffusion process. ! mrloop=0 !# of radial transport and SHOOT evaluations (0=ignore) XNeoMult=0.0 !neoclassical D multiplier (1.0=standard, 0.0=off) convmult=0.0 !convection multiplier (1=standard, 0=off) fpdifcon=0.0 !arbitrary radial diffusion coefficient (cm**2/sec) fpdiftra=0.0 !arb. radial D for trapped ions only (cm**2/sec) fpdifdw=0.0 !arb. radial D due to drift-waves (cm**2/sec) ! (fpdifdw conserves E and xsi, while other mechanisms conserve E and mu) XRipMult=0.0 !ripple diffusion multiplier (1=standard,0=off) ! Warning: I'm not sure that the ripple transport works correctly ! right now. --Greg Hammett 17-Jan-1990. ! The above comment is probably out-of-date? --Greg Hammett 31-May-1992 ! Warning: should probably leave xneomult=0.0 until I take another ! look at the form of my radial diffusion operator. It is ! currently in the form of a Fokker-Planck operator: ! (1/r) (d^2/dr^2) ( D_r r f) ! but there should be a convective term as well (see the introductory ! comments in RDOPER.FOR to see how this happens in the map from ! Psi to r). ! Warning II: The radial transport models assume that the plasma ! flux surfaces are shifted circles. ! ! parameters for exponential ripple model: ! delta(R,y) = Ripdmin * exp( sqrt((R-RipRmin)**2+y**2)/Ripexpl ) !p r RipRmin major radius of minimum ripple (cm) !p r Ripdmin ripple (peak to average) at that point !p r RipexpL exponential scale length increase (cm) !p i NumTF # of TF coils ! ! TFTR parameters (my improvement of Steve Scott's fit): ! This fit provides an upper bound on the ripple, and fits the ripple ! well for ripples above .1%. ! RipRmin=228.0 !major radius of minimum ripple (cm) Ripdmin=1.895e-5 !ripple (peak to average) at that point RipexpL=19.17 !exponential scale length increase (cm) NumTF=20 ! # of TF coils ! ! This is a crude fit which provides a lower bound on the TFTR ripple: ! data RipRmin /228.0/ !major radius of minimum ripple (cm) ! data Ripdmin /1.5e-6/ !ripple (peak to average) at that point ! data RipexpL /13.956/ !exponential scale length increase (cm) ! data NumTF /20/ ! # of TF coils ! ! Rob's fit to PLT ripple: ! data RipRmin /112/ !major radius of minimum ripple (cm) ! data Ripdmin /1.1e-5/ !ripple (peak to average) at that point ! data RipexpL /9.97/ !exponential scale length increase (cm) ! data NumTF /18/ ! # of TF coils $end