Notes on CASTEP fort.x files ---------------------------- fort.[234] (txt) ---------- fort.3 is the only one of these that CASTEP uses, the others are for the symmetrisation front-end. fort.10 (txt) ------- Ewald integral data. Length of arrays given by NEWPTS (param.inc). Read by cgion.f, used by ewaltr.f. fort.11 (txt) ------- Pseudopotential data, atomic files simply concatenated in order of occurance in fort.1[45] fort.13 (bin) ------- Contains cell size, atomic positions, k-points, RMOVE, etc. If restarting, these k-points and positions are used, not fort.15's. (This is where the relaxed ionic positions are stored.) However, fort.15's RMOVE array is used, not this one... See MJR's fort.13 to fort.15 converter to extract info from this file. fort.14 (txt) ------- Self-documenting(?) fort.15 (txt) ------- Cell dimensions, ionic positions, k-points and weights. Order of positions is POSION followed by POSIOL. POSIOL seems to be mainly ignored. fort.16 (bin) ------- Charge density in real & reciprocal space as below two complex arrays as below: WRITE (16) (CHDENR(I),I=1,NPLWV),(CHDENG(I),I=1,NPLWV) CHDENR((NZ-1)*NGX*NGY+(NY-1)*NGX+NX) Read in, then discarded, on restart, UNLESS a band structure calculation was requested. fort.17 (txt) ------- Band occupancies fort.18 (txt) ------- Non-local parts of real space pseudopotentials cat'ed together. A fort.11 must still exist for the local part which is still used. If a projector does not exist, this can be indicated by using an l value of 1000. Otherwise each potential here must have the number of projectors specified by MXRLNL. fort.2x (bin) ------- One per k-point, so max. no. of k-points is 20 if relaxing, else (highest output file automatically opened on your architecture)-19 fort.4x (bin) ------- Created only if ionic relaxation is on. One per k-point. Not needed for restart. runinf (txt) ------ Run control. If present: NITER IION POTIM NDELAY May be truncated at any point. Read at the start of each iteration.