User: rhaas Date: 2013/03/27 08:49 PM

Modified: /trunk/ param.ccl, schedule.ccl /trunk/src/ GRHydro_BondiM.c

Log: GRHydro_Init_Data: add routine to force boundary data to Bondi

From: Roland Haas rhaas@tapir.caltech.edu

File Changes:

Directory: /trunk/src/ ======================

File [modified]: GRHydro_BondiM.c Delta lines: +46 -8 =================================================================== --- trunk/src/GRHydro_BondiM.c 2013-03-28 01:49:03 UTC (rev 204) +++ trunk/src/GRHydro_BondiM.c 2013-03-28 01:49:05 UTC (rev 205) @@ -890,22 +890,41 @@ -- driver routine for the Bondi solution;

***********************************************************************************/ -static void GRHydro_BondiM_Internal(CCTK_ARGUMENTS, CCTK_REAL range_min, CCTK_REAL range_max); +static void GRHydro_BondiM_Internal(CCTK_ARGUMENTS, + CCTK_REAL range_min, CCTK_REAL range_max, + const int range_imin[], const int range_imax[]);

void GRHydro_BondiM(CCTK_ARGUMENTS) { - GRHydro_BondiM_Internal(CCTK_PASS_CTOC, 1e100, -1e100); + DECLARE_CCTK_ARGUMENTS; + const int null[3] = {0,0,0}; + GRHydro_BondiM_Internal(CCTK_PASS_CTOC, 1e100, -1e100, cctk_lsh, null); }

void GRHydro_BondiM_Range(CCTK_ARGUMENTS) { DECLARE_CCTK_PARAMETERS; + DECLARE_CCTK_ARGUMENTS; + const int null[3] = {0,0,0}; GRHydro_BondiM_Internal(CCTK_PASS_CTOC, bondi_freeze_inner_radius, - bondi_freeze_outer_radius); + bondi_freeze_outer_radius, + cctk_lsh, null); }

+void GRHydro_BondiM_Boundary(CCTK_ARGUMENTS) +{ + DECLARE_CCTK_PARAMETERS; + DECLARE_CCTK_ARGUMENTS; + const int imax[3] = {cctk_lsh[0]-cctk_nghostzones[0], + cctk_lsh[1]-cctk_nghostzones[1], + cctk_lsh[2]-cctk_nghostzones[2]}; + GRHydro_BondiM_Internal(CCTK_PASS_CTOC, 1e100, -1e100, + cctk_nghostzones, imax); +}

-static void GRHydro_BondiM_Internal(CCTK_ARGUMENTS, CCTK_REAL range_min, CCTK_REAL range_max) +static void GRHydro_BondiM_Internal(CCTK_ARGUMENTS, + CCTK_REAL range_min, CCTK_REAL range_max, + const int range_imin[], const int range_imax[]) { DECLARE_CCTK_ARGUMENTS; DECLARE_CCTK_PARAMETERS; @@ -920,7 +939,7 @@

int nbondi;

- const int size = cctk_lsh[0] * cctk_lsh[1] * cctk_lsh[2]; + const int size = cctk_ash[0] * cctk_ash[1] * cctk_ash[2]; int i, imin, j;

# define velx(i_) vel[i_ + 0*size] @@ -952,7 +971,7 @@ } else if( CCTK_EQUALS(bondi_Bvec_method, "transform")) { bvec_method = BVEC_TRANSFORM; } else { - CCTK_VWarn(CCTK_WARN_ABORT, __LINE__, __FILE__, CCTK_THORNSTRING, + CCTK_VError(__LINE__, __FILE__, CCTK_THORNSTRING, "Unknown bvec setup method '%s'", bondi_Bvec_method); }

@@ -1109,8 +1128,12 @@ *********************************************************************************/ xpos[TT] = 0.;

- for(i=0; i < size; i++) { + for(int kk = 0 ; kk < cctk_lsh[2] ; kk++) { + for(int jj = 0 ; jj < cctk_lsh[1] ; jj++) { + for(int ii = 0 ; ii < cctk_lsh[0] ; ii++) {

+ i = CCTK_GFINDEX3D(cctkGH, ii,jj,kk); + xpos[XX] = x[i] ; xpos[YY] = y[i] ; xpos[ZZ] = z[i] ; @@ -1152,8 +1175,21 @@

rspher = x_spher[RR];

+ // the conditions here are a bit strange. The logic goes like this: + // range_min,range_max (and range_imin, range_imax) are radii (and indices) + // bounding the volume where we want to evolve the fields. So during + // evolution we only reset to ID data if we are NOT inside this volume. For + // ID we pass special values for min/max such that these conditions never + // trigger. + // TODO: do this more nicely. The fact that a comment is needed should tell + // me that this is getting too complex. if(rspher > range_min && rspher < range_max) continue; + if(ii >= range_imin[0] && ii < range_imax[0] && + jj >= range_imin[1] && jj < range_imax[1] && + kk >= range_imin[2] && kk < range_imax[2]) { + continue; + }

/* Find nearest point in the Bondi solution : */ j = (int) ( 0.5 + (log10(rspher) - logrmin) / dlogr ) ; @@ -1266,7 +1302,9 @@ if(psidc) psidc[i] = 0.;

- } + } // ii + } // jj + } // kk

free(r_bondi); free(logr_bondi); free(rho_bondi); free(u_bondi); free(v_bondi);

Directory: /trunk/ ==================

File [modified]: param.ccl Delta lines: +4 -0 =================================================================== --- trunk/param.ccl 2013-03-28 01:49:03 UTC (rev 204) +++ trunk/param.ccl 2013-03-28 01:49:05 UTC (rev 205) @@ -420,6 +420,10 @@ *:* :: "any value" } 1e300

+CCTK_BOOLEAN bondi_overwrite_boundary "reset data to initial data in outer boundary in boundary condition" +{ +} "no" + # For Poloidal Magnetic field test:

CCTK_REAL poloidal_A_b "Vector potential strength"

File [modified]: schedule.ccl Delta lines: +10 -0 =================================================================== --- trunk/schedule.ccl 2013-03-28 01:49:03 UTC (rev 204) +++ trunk/schedule.ccl 2013-03-28 01:49:05 UTC (rev 205) @@ -257,6 +257,16 @@ } "force analytic solution outside anulus" }

+if(bondi_overwrite_boundary ) { +# this is a terrible HACK. We really should at least schedule ourselves in +# Boundaries and look at cctk_bbox and the symmetry flags +# by now Cactus is convenient CCTK_LOOP macros for this... +SCHEDULE GRHydro_BondiM_Boundary IN HydroBase_Boundaries BEFORE HydroBase_Select_Boundaries +{ + LANG: C +} "force analytic solution in boundaries" +} + if (CCTK_EQUALS(initial_Bvec, "poloidalmagfield") || CCTK_EQUALS(initial_Avec, "poloidalmagfield")) { # SCHEDULE GRHydro_PoloidalMagFieldM AT CCTK_INITIAL AFTER IN HydroBase_Initial AFTER rnsid_init AFTER TOV_Initial_Data after CCCC_StarMapper_InitialData