I have only stripped out D3D and put in ITER character as a replacement - coils are correct but limiting surface is not yet accurate
Thanks to Dave Greeenwood for figuring out the LF95 & PGF90 switches and lib ordering so pgplot (g77-gcc) build works properly with either the Lahey or Portland compiler.
Status & Purpose
I believe the code is working correctly.
My purpose is to provide a tool D3D workers are familiar, which will aid us in assessing ITER.
Tools are provided so you can take p’ & ff’ from your favorite equilibria and produce the equivalents using the ITER coil set.
I hope this will provide part of the basis for our judging the adequacy of the ITERdesign.
Linux 129x129 only
location /d3/EFITI
subdirectories are bin cvs efund idl ITER pgplot shape & src
Call is /d3/EFITI/bin/efiti129d device k-file
Restrictions single k-file input (mode 2 in usual EFITD
Graphics
Pgplot graphics devices /xs or /cps
if /xs menu items 1-page 4-page are inactive and page 4 is not yet filled
if /cps, the pgplot.ps is moved to pshotnum.time at end of execution
It may be necessary to set environment:
PGPLOT_DIR=/d3/EFITI/pgplot
PGPLOT_FOREGROUND=black
PGPLOT_BACKGROUND=white
PGPLOT_FONT=/d3/EFITI/pgplot/grfont.dat
PGPLOT_RGB=/d3/EFITI/pgplot/rgb.txt
Old lessons being relearned
Fixed boundary solutions to rshotnum.time produce
xshotnum.time+1 a free boundary input file
mshotnum.time+11 a reconstruction input file.
The difference is primarily iconvr= 3 or 2
Fixed boundary is working well
Free boundary is working well
Reconstruction may need practice
Utilities
fixbndy.pro (& fixrbndy.pro)
FIXBNDY, gfile KFFCUR=? KPPCUR=?
Produces nicely formatted Cα,Xα, & Cγ,Xγ arrays to be appended to an input IN1 namelist.
shapel create a fixed boundary input file
/d3/EFITI/ITER is intended as a work area and has sample input/output files
Try it
I would like to collect interesting cases that our based on D3D results in – /d3/EFITI/ITER/cases
Please put input files (along with A and G-files) their for well-converged cases.
An example of output follows.
Recommendations (or work) for improvement are welcome.
Minutia
Response tables in efund with extension .itr are native linun (little-endian)
dprobe.dat => rprobe.dat
Symmetrize and fitdelz are in &in1 and default to symmetrize=F, fitdelz=T
Because of the asymmetry of the outer coils, zelip=0.5 is about right, NOT zelip=0. Perhaps (zelip=.5 relip=6.2 aelip=2. eelip=.16). Recall zelip is measured in D3D.
In my limited experience you need to work harder to get a good result than for D3D.
.