To calculate thermal conductivity, we have two methods. One is input heat flux and measure temperrature gradient and the other is by setting a tempearture gradient by inducing a temperature diff and measure the heat current( flux). Of these two, the fix/ heat or fix/thermal conducitvity follow the first method. Is there any way we can implement the second method?? Can we measure heat current by setting up a temperature difference initially??
Thanks
Ajit
no - LAMMPS only currently does the first method.
Steve
I think that one can perhaps create regions (bins) and fix different temperatures in the bins in order to create a temperature gradient using NVT. One would have to be careful about the number of MD time steps over which temperature control is exercised.
Then carry out a NVE simulation. Measuring the energy flux in order to study the thermal conductivity would be tricky is my guess.
I have not tried the above and it could be a stupid suggestion. Would like to hear from anybody who has tried it.
Manoj
2009/5/12 Steve Plimpton <[email protected]>