Hi Eric, It seems that your vc-relax calculation is not finished, it just terminate. because it not get convergence for the tress tensor or pressure when it get the max step (n=50).
the way solve this problem is get a more good start geometry or largen the step maximum. please specify nstep = 100 or larger in the input file. good luck! Best regards, =============================================== Jian SUN Physics Dept. of Nanjing University National Lab. of Solid State Microstructures 22 Hankou Road, Gulou District Nanjing, Jiangsu Province 210093 China =============================================== Message: 1 From: "Eric Abel" <[email protected]> To: pw_forum at pwscf.org Subject: Re:[Pw_forum] vc-relax Date: Mon, 11 Sep 2006 02:41:02 -0400 Reply-To: pw_forum at pwscf.org Cesar, Thank you for your response. Unfortunately if it were that simple, I wouldn't have had the need to post to the forum. my output file contains a series of scf calculations, which is to be expected, after each scf convergence it lists the atomic forces followed by the stress: total energy = -357.37450532 ryd estimated scf accuracy < 2.2E-11 ryd convergence has been achieved Forces acting on atoms (Ry/au): atom 1 type 1 force = 0.00000000 0.00000000 0.00682942 atom 2 type 1 force = 0.00000000 0.00000000 -0.00682942 atom 3 type 2 force = 0.00000000 0.00000000 0.20546159 atom 4 type 2 force = 0.00000000 0.00000000 -0.20546159 atom 5 type 3 force = 0.00000000 0.00000000 -0.08327844 atom 6 type 3 force = 0.00000000 0.00000000 0.08327844 entering subroutine stress ... total stress (ryd/bohr**3) (kbar) P= -82.53 -0.00135974 0.00000000 0.00000000 -200.02 0.00 0.00 0.00000000 -0.00071001 0.00000000 0.00 -104.45 0.00 0.00000000 0.00000000 0.00038671 0.00 0.00 56.89 after the last iteration, it lists a set of k-points followed by: Writing file tiocl.save for program phonon PWSCF : 13h57m CPU time init_run : 66.55s CPU electrons : 44803.10s CPU ( 50 calls, 896.062 s avg) forces : 1319.61s CPU ( 50 calls, 26.392 s avg) stress : 2795.36s CPU ( 50 calls, 55.907 s avg) electrons : 44803.10s CPU ( 50 calls, 896.062 s avg) c_bands : 37432.39s CPU ( 343 calls, 109.132 s avg) sum_band : 6668.94s CPU ( 343 calls, 19.443 s avg) v_of_rho : 70.93s CPU ( 688 calls, 0.103 s avg) newd : 669.21s CPU ( 345 calls, 1.940 s avg) mix_rho : 48.63s CPU ( 343 calls, 0.142 s avg) c_bands : 37432.39s CPU ( 343 calls, 109.132 s avg) init_us_2 : 514.48s CPU ( 59025 calls, 0.009 s avg) cegterg : 36838.34s CPU ( 25725 calls, 1.432 s avg) sum_band : 6668.94s CPU ( 343 calls, 19.443 s avg) becsum : 5.02s CPU ( 25725 calls, 0.000 s avg) addusdens : 625.36s CPU ( 343 calls, 1.823 s avg) wfcrot : 44.03s CPU ( 75 calls, 0.587 s avg) cegterg : 36838.34s CPU ( 25725 calls, 1.432 s avg) h_psi : 24259.17s CPU ( 85747 calls, 0.283 s avg) g_psi : 203.06s CPU ( 59947 calls, 0.003 s avg) overlap : 3791.19s CPU ( 59947 calls, 0.063 s avg) cdiaghg : 1084.00s CPU ( 60022 calls, 0.018 s avg) update : 1402.85s CPU ( 59947 calls, 0.023 s avg) last : 1466.98s CPU ( 26082 calls, 0.056 s avg) h_psi : 24259.17s CPU ( 85747 calls, 0.283 s avg) init : 96.39s CPU ( 85747 calls, 0.001 s avg) firstfft : 8568.69s CPU ( 1932983 calls, 0.004 s avg) secondfft : 8659.88s CPU ( 1932983 calls, 0.004 s avg) add_vuspsi : 2826.23s CPU ( 85747 calls, 0.033 s avg) s_psi : 2548.22s CPU ( 85747 calls, 0.030 s avg) General routines ccalbec : 4178.90s CPU ( 118972 calls, 0.035 s avg) cft3 : 128.72s CPU ( 4378 calls, 0.029 s avg) cft3s : 18657.11s CPU ( 4638404 calls, 0.004 s avg) interpolate : 33.71s CPU ( 688 calls, 0.049 s avg) davcio : 147.96s CPU ( 84750 calls, 0.002 s avg) There are a few additional lines, but I don't find the words CELL_PARAMETERS or ATOMIC_POSITIONS anywhere. My instinct tells me that perhaps something went wrong, but the program completed without complaint. I never set the verbosity level, so I assume it's set at the default. Like I said in my original post, in the 'relax' calculation, the relevant information was present. Any ideas from anyone? Eric
