entime.f

99 lines · 1 definitions · on GitHub (7be4133)

Python version: entime.py, 56 lines, on the right (switch to Python above the code).

Variables on this page (6)

atime, hlsuro, json, tefflo, tridl, tridt

Purpose

This file limits the time step so that the surface changes from one model to the next stay within the envelope "triangle": the changes of $\log T_\mathrm{eff}$ and $\log L$ must not exceed the triangle's size, so that the surface boundary condition is not extrapolated far outside the envelope integrations.

How it fits

Inputs: the arguments DELTS (the previous step), HL (luminosity), TEFFL, M; the previous model's surface luminosity HLSURO and TEFFLO (OLDMOD_MOD, saved by HPOINT before the rezoning, and by main.f and STARIN); the settings TRIDT, TRIDL (the triangle's sizes, formerly /CENV/) and ATIME(13) (the limit, formerly /CTLIM/), from NMLPAR_MOD.
Who calls it: HTIMER, which sets the next time step.
Output: EDELT, the time step this criterion allows.

The method

$$f = \max\left(\frac{|\Delta\log T_\mathrm{eff}|}{\Delta T_\mathrm{tri}}, \frac{|\Delta\log L|}{\Delta L_\mathrm{tri}}\right), \qquad f \leftarrow \min\!\left(\max\left(f, \frac{1}{a}\right), a\right), \qquad \mathtt{EDELT} = \frac{\Delta t}{f},$$ with $a$ = ATIME(13), the largest factor by which the step may change; $f = 1$ if nothing changed. A previous step of 0 thus gives EDELT = 0, which HTIMER does not catch (htimer.f, suspected bug 1). The triangle's sizes are also copied into HMAX(1:2), which nothing reads (unused?). The header's output line, "time step for requested change in P,T,R,L," is copied from PTIME: Only $T_\mathrm{eff}$ and $L$ are tested here (settled from the code, Nov 9, 2026: line 68 ↗; lines 73–95 ↗).

Definitions

subroutineENTIME

lines 6–98 · GitHub

Time step from the envelope triangle: scales the last step so that $\Delta\log T_\mathrm{eff}$ and $\Delta\log L$ stay within the triangle, by a factor bounded by ATIME(13).

Uses modules DIMENSIONS_MOD, NMLPAR_MOD, OLDMOD_MOD

Called by HTIMER

Compiler warnings (1)

gfortran -Wall -Wextra, compiling only; nothing was run.

line 90: Equality comparison for REAL(8) [-Wcompare-reals]

Change log

The changes since rp_1

Changed since rp_1. Declarations only: COMMON blocks to modules, IMPLICIT NONE; the unused COMMON blocks are gone, which settles that half of former question 1.

Changed since a8a5950. $\Delta\log L$ is taken against the saved surface luminosity HLSURO instead of HLO(MO): HPOINT re-splines HLO onto the new mesh, so HLO(MO) was an interior point or a stale entry whenever the number of points changed (this settles HPOINT's former question 6).

1C 2C 3C$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$ 4C ENTIME 5C$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$ 6 SUBROUTINE ENTIME(DELTS,HL,TEFFL,M,EDELT) 7C 10/15/26 MHP via Claude 8C code changed to take JSON from DIMENSIONS_MOD instead of declaring 9C PARAMETER(JSON=5000) here. 10 USE DIMENSIONS_MOD 11C 10/14/26 MHP via Claude 12C code changed to convert COMMON/OLDMOD/ into OLDMOD_MOD. 13 USE OLDMOD_MOD 14C 10/14/26 MHP via Claude 15C code changed to convert COMMON/NMLPAR/ into NMLPAR_MOD. 16 USE NMLPAR_MOD 17C 10/20/26 MHP via Claude 18C code changed to replace implicit typing with IMPLICIT NONE and 19C explicit declarations. 20 IMPLICIT NONE 21 REAL*8 DELTS,EDELT,FAC,HL,HMAX,TEFFL,TEMP,TESTL,TESTT 22 INTEGER M 23 24C 10/13/26 MHP via Claude: removed the unused COMMON/CONSTANTS/ declaration 25C (no member referenced here; residue of the obsolete-option removals). 26C 9/24/26 MHP via Claude 27C code changed to remove dead CMIXL from COMMON/CONST3/; it is now 28C passed as an explicit argument where actually used 29C 10/12/26 MHP via Claude 30C code changed to extend COMMON/NMLPAR/ to 234 members, adding the 31C namelist variables that the physics routines read from 27 COMMON 32C blocks which also held derived or working members. Those blocks keep 33C only their non-namelist members; five of them -- NEWMX, VMULT, 34C PMMWIND, CCOUT1 and ENVGEN -- had nothing left and are gone. 35C Derivation-only namelist variables from the same blocks went to 36C COMMON/NMLBYPASS/ instead, not here. 37C 38C Declared identically in all 121 files that hold any of these members. 39C Generated from one member list, so every copy is byte-identical; the 40C linker's 'size of symbol changed' warning is the check that they are. 41C 10/13/26 MHP via Claude 42C code changed to drop COMMON/CTLIM/, declared here but never used. 43C 10/13/26 MHP via Claude: removed the unused COMMON/CENV/ declaration 44C (no member referenced here; residue of the obsolete-option removals). 45 46 DIMENSION HL(JSON) 47 DIMENSION HMAX(4) ! NHMAX(4) 48 SAVE 49 50C THIS SR LIMITS THE TIMESTEP BASED ON THE REQUIREMENT THAT THE CHANGES FROM 51C ONE MODEL TO THE NEXT IN T,AND L BE LESS THAN THE SIZE OF THE ENVELOPE TRIANGLE 52C (TRIL AND TRIT). THIS PREVENTS EXTRAPOLATION OUTSIDE OF THE TRIANGLE 53C FOR SMALL ENVELOPE TRIANGLES. 54C 55C INPUT VARIABLES 56C DECLARED VARIABLES : 57C DELTS : PREVIOUS MODEL TIMESTEP. 58C HL : RUN OF LUMINOSITY(AS A FUNCTION OF MASS) IN THE CURRENT MODEL. 59C TEFFL: EFFECTIVE TENPERATURE OF CURRENT MODEL 60C M : NUMBER OF POINTS IN THE MODEL. 61C VARIABLES IN COMMON BLOCKS : 62C TRIL: DELT L OF ENVELOPE TRIANGLE 63C TRIT: DELT T OF ENVELOPE TRIANGLE 64C ATIME(13): maximum percentage change in the timestep from the previous one 65C HLO : RUN OF LUMINOSITY(AS A FUNCTION OF MASS) IN THE PREVIOUS MODEL. 66C TEFFLO : EFFECTIVE TEMPERATURE OF PREVIOUS MODEL 67C OUTPUT VARIABLES : 68C EDELT : TIME STEP FOR REQUESTED CHANGE IN P,T,R,L. 69C 70C FIND MAXIMUM ABSOLUTE TIME DIFFERENCES FOR EACH QUANTITY 71C 72C TEMPERATURE 73 HMAX(1)=TRIDT 74C LUMINSOITY 75 HMAX(2)=TRIDL 76 77 TESTT = ABS(TEFFLO-TEFFL) 78C 10/29/26 MHP via Claude 79C code changed to compare with the saved surface L, HLSURO, not HLO(MO): 80C HLO is re-splined onto the new mesh by HPOINT, so MO no longer indexes 81C its surface. 82 TESTL = ABS(DLOG10(HLSURO)-DLOG10(HL(M))) 83 84C NOW ACTUALLY LIMIT THE TIMESTEP BY A FACTOR THAT REDUCES THE 85C TIME CHANGES IN ALL QUANTITIES TO THE TRIANGLE VALUES OR LESS 86 FAC = TESTT/TRIDT 87 IF (TESTL/TRIDL .GT.FAC) FAC=TESTL/TRIDL 88C IF NO CHANGE FROM PREVIOUS MODEL,SET EDELT TO TIMESTEP 89C STORED IN THE PREVIOUS MODEL. 90 IF (FAC .EQ. 0.0D0) FAC=1.0D0 91C USE ATIME(13) AS THE GLOBAL FACTOR FOR LIMITING TIMESTEP CHANGES 92 TEMP = ATIME(13) 93 IF (FAC.GT.TEMP) FAC=TEMP 94 IF (FAC.LT.1.0D0/TEMP) FAC=1.0D0/TEMP 95 EDELT = DELTS/FAC 96 97 RETURN 98 END 99

1"""Python version of entime.f (YREC 7be4133): the time step allowed by the 2envelope triangle.""" 3 4import math 5 6import numpy as np 7 8from pyrec import nmlpar_mod 9from pyrec import oldmod_mod 10 11# SAVE: the work array HMAX(4), filled and never read (Fortran lines 12# 47–48). 13hmax = np.zeros(4 + 1) 14 15 16def entime(delts: float, hl: np.ndarray, teffl: float, m: int, edelt: float) -> float: 17 """Scale the last step so that the changes of log Teff and log L from 18 the previous model stay within the envelope triangle, by a factor 19 bounded by atime[13]. 20 21 Translates: entime.f#ENTIME (7be4133) 22 23 Arguments: delts, the previous step; hl, the luminosities; teffl, 24 log Teff; m, the number of shells; edelt (set only). 25 26 Returns edelt, the step this criterion allows. 27 28 """ 29 30 # The triangle's sizes, copied into HMAX, which nothing reads (Fortran 31 # lines 73–75). 32 hmax[1] = nmlpar_mod.tridt 33 hmax[2] = nmlpar_mod.tridl 34 35 # The changes of log Teff and log L since the previous model (Fortran 36 # lines 77–82). 37 testt = abs(oldmod_mod.tefflo - teffl) 38 testl = abs(math.log10(oldmod_mod.hlsuro) - math.log10(hl[m])) 39 40 # The factor, within [1/atime[13], atime[13]], and the step (Fortran 41 # lines 86–95). 42 # Suspected bug (see the page, htimer.f#b1): a previous step of 0 gives 43 # edelt = 0, which HTIMER does not catch. 44 fac = testt / nmlpar_mod.tridt 45 if testl / nmlpar_mod.tridl > fac: 46 fac = testl / nmlpar_mod.tridl 47 if fac == 0.0: 48 fac = 1.0 49 temp = nmlpar_mod.atime[13] 50 if fac > temp: 51 fac = temp 52 if fac < 1.0 / temp: 53 fac = 1.0 / temp 54 edelt = delts / fac 55 56 return edelt