[Pw_forum] phonons

[bothina hamad]

Dear users,
                     
I am starting phonon calculations of a full Heusler alloy ( 4 atoms per unit 
cell). The alloy is  half-metallic (metallic in majority spin channel and 
semiconducting in the minority). It is also a ferromagnet with a total magnetic 
moment 2Bohr magnetons. 


Using a small grid  2x2x2, I got the phonon calculations working smoothly and I 
got the phonon density of states. However, I could not  reach convergence for 
denser grids. In these calculations, I then tried to split the points and make 
calculations for each q point in the grid separately, I found the the 
convergence occur for the 3 q-points ( the same points as 2x2x2 grid), however 
no  convergence is reached for the other point and the calculations keep 
oscillating
until 100 iterations then stop. I tried to change some parameters such as 
decreasing the tr2_ph=1.0d-14 to tr2_ph=1.0d-18, and also 

by decreasing the mixing to 0.1, still the problem is not solved.
 
Do you have  any idea that can solve the problem

Best wishes

Dr. Bothina Hamad
 Physics Department
 University of Jordan
 Amman-11942, Jordan
 Tel:  +962-6-5355000
 ext#22030 & 22023
 Fax:+962-6-5348932