The answer is in the "dynamics.f" routine, where all MD algorithms are written.
The temperature you enter in your input is the "target" temperature. The code moves atoms so that their kinetic energy is consistent (on average) with your target temperature, but you are not to expect it to match it at every MD step... the Nose thermostat takes/gives kinetic energy as the calculation goes, precisely to achieve the target temperature. Your Nose Mass should not be too small, to actually let the system explore a range of temperatures. Once you have sufficient MD steps, you can then discard many at the start of the run (when thermalization is just being set up), and then get an average temperature. In all my runs, I get the average temperature to be within 1% of the target one, after tens of thousands of MD steps for good averaging. Your supercell should also be large enough to get meaningful results. You should also read Nose papers on the subject. Going at it expecting certain numbers and without a grasp of theory could be dangerous and not recommendable. ----------------------------------------- Salvador Barraza-Lopez Associate Professor of Physics University of Arkansas https://wordpressua.uark.edu/sbarraza/ ________________________________ From: siesta-l-requ...@uam.es <siesta-l-requ...@uam.es> on behalf of pooja pu <pupooj...@gmail.com> Sent: Saturday, May 9, 2020 4:31 AM To: siesta-l@uam.es <siesta-l@uam.es> Subject: [SIESTA-L] Abinitio MD simulations Respected all I'm new to MD simulations using siesta There are different types of MD runs like nose, Verlet. Using nose however temperature is changing in every step and not becoming constant up-to 2ps while it is said calculation using nose thermostat are at constant temperature ? How is thermal stability of a system checked using MD simulations in siesta? Please let me know. Any help and guidance would be highly appreciated. Please if anyone knows the answer please help me. Thanks With regards Pooja kapoor India
-- SIESTA is supported by the Spanish Research Agency (AEI) and by the European H2020 MaX Centre of Excellence (http://www.max-centre.eu/)
