Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Planewave cutoff energy convergence!

  1. Nov 10, 2011 #1
    Hello everyone,

    I have some problems with the convergence for the planewave cutoff energy, I tried to find the correct PWcE for a system with 2 atoms, I've determined the minimum k-point mesh with success, but when I compute the Total energy varying the PWcE, it seems that the convergence will be achieved till too high values of PWcE, should I change the pseudopotentials for the atoms?

    I use DACAPO,

    I'm Mexican by the way, so sorry if I made a mistake while I wrote this.

  2. jcsd
  3. Nov 11, 2011 #2
    If the pseudopotential is hard then you may not achieve convergence until reaching to very high cutoff (~700+ eV). On the other extreme ultrasoft pseudopotentials tend to converge with much lower values. Another factor is the type of element you are studying.
    Check the manual of your code and the recommended values of the pseudopotential.
  4. Nov 11, 2011 #3
    Thanks for your response, but in fact I'm using ultrasoft pseudopotentials (Vanderbilt), It seems that the Total energy converge at 400 eV but when I compute the total Energy with higher PWcE the tot E fall again to lower values. I didnt allow the atoms to relax, so the unique converge criteria is the total Energy, should I change to a relaxation calculation?

  5. Nov 11, 2011 #4
    There is no need to conduct the energy convergence test in a relaxation simulation. A single point calculation is good enough to test energy convergence.

    Can you tell us what the total energy difference per atom between the 400eV cutoff calculation and the more conservative one?
  6. Nov 11, 2011 #5

    Of course! If I compute the total energy difference between the calculations done with a planewave cutoff energy of 320 and 400 eV , I get like 36 KJ/mol; something similar but in the interval of 400 - 480 eV (PWcE) I get about 4 KJ/mol; and the same with a PWcE range 480-700 eV, I get 17 KJ/mol (I've already done some calculations with higher pwce, up to 880 eV, and at this point (starting in 700) the plot seems to be almost asymptotic). But it seems to be a kind of weird to me, because I was hoping for a lower pwce value, it makes any sense?
    could I take the 400 eV value as the planewave cutoff energy?

    by the way I use a 6 atoms periodic model!
    Last edited: Nov 11, 2011
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook