The ABINIT Users Mailing List ( CLOSED )

[jzwanzig@dal.ca]

Dear Colleagues, 

I am trying to run essentially the calculation of Response Function Tutorial 
2 on a more complex material, to compute the phonon dispersion curves. For a 
variety of reasons I would prefer not to chain all the calculations one after 
another into a single run that would take days or weeks, but rather to do 
each phonon wavevector separately. In my case, obtaining the DDB for each 
phonon wavevector requires 8 calculations (2 rfdir's for each of 4 different 
atoms). Running a single one of these works, but chaining them together 
fails. For example, if I try to do just two of them in one run, with 

ndtset 2
rfdir1 1 0 0
rfdir2 0 1 0
rfphon 1
rfatpol 1 1

the calculation crashes when it compares the expected hdr_check result with 
what it reads from disk (WFK from the ground state run); the log file reports 
"hdr_check: BUG: input fform= 2 differs from disk file fform= 0". 

This behaviour is annoying, is it really a bug? Or am I trying to do 
something in violation of the code?

Thanks,
Joe Zwanziger