19981207: ``npt'' run with reduced step size

GOAL

``npt'' run with reduced step size to see if r.m.s. on temperature, cell size and pressure come out better.

Results

This is to be compared with 19981130, in which the time step was $0.002\;ps$. Here the time step was $0.0005\;ps$ (both time constants for temperature and pressure control were $0.5\;ps$).

These are the results from the OUTPUT file of DL_POLY:

 run terminated after  420000 steps. final averages calculated over  400000 steps.


 ----------------  .   -------------
                                   .   
     step     eng_tot    temp_tot  .        eng_tet
 time(ps)      eng_pv    temp_rot  .        vir_tet
 cpu  (s)      volume    temp_shl  .          press
                                   .   
 ----------------  .   -------------
   420000 -1.0757E+04  1.7590E+03  .     0.0000E+00
  210.000 -1.0759E+04  0.0000E+00  .     0.0000E+00
   287.72  4.2516E+03  4.5589E+01  .     5.7442E-02
                                   .   
  r.m.s.   1.6141E-02  6.9824E+01  .     0.0000E+00
 fluctn.   2.9336E+01  0.0000E+00  .     0.0000E+00
           2.5381E+01  3.1975E+00  .     1.0831E+01
 ----------------      -------------
                                   
....


            Average cell vectors           |        r.m.s. fluctuations
                                           |               
16.1999910959  0.0000000000   0.0000000000 | 3.2231E-02  0.0000E+00  0.0000E+00
 0.0000000000 16.1999910959   0.0000000000 | 0.0000E+00  3.2231E-02  0.0000E+00
 0.0000000000  0.0000000000  16.1999910959 | 0.0000E+00  0.0000E+00  3.2231E-02
                                           

Does not seem to have had any benefit. Note the big r.m.s. fluctuations on temperature and pressure...

Comparing with 19981130 is not straigthforward, since I've lost the OUTPUT file for that experiment! I could do similar plots here, but the time expense does not seem to be worth of.