Dear all
while calculating the thermal conductivity using green kubo method of liquid electrolyte having bond angle, dihedral angle, improper angle, I got error KSpace style does not support compute centroid/stress/atom.
I used kspace_style pppm 1.0e-4
when i tried to calculate thermal conductivity with compute stress/atom simulation is running with same kspace style.
In the LAMMPS manual it is clearly written that compute centroid/stress/atom must be used if in your system having bond angle, dihedral angle, improper angle.
please help me to short out to this problem. thanks in advance
There is nothing to short out (sp!). 
The path forward is quite obvious. You have multiple choices:
thanks for your kind reply
there is any reference or article to drive the expression for computing per-atom centroid stress from lattice sum and how to implement in LAMMPS
as per my knowledge
step 1. compute the forces acting on each atom(involves summing up all pairwise forces between atoms) using lattice sum approach
step2. compute the stress tensor at each atom using lattice sum approach
σi=(1/Vi)∑(Rij)XFij
step3. compute the per-atom centroid stress by averaging the stress tensors over the neighboring atoms of each atom
σ (centoroid)i= (1/ni)∑σj
input code:
compute myforce all force
compute stress all stress/atom NULL virial
fix mystress all ave/atom 1 100 100 c_stress[1] c_stress[2] c_stress[3] c_stress[4] c_stress[5] c_stress[6] c_stress[7] c_stress[8] c_stress[9] file stress_atom.dat
compute per_atom_stress all reduce ave c_stress[1] c_stress[2] c_stress[3] c_stress[4] c_stress[5] c_stress[6] c_stress[7] c_stress[8] c_stress[9]
fix centoridstress all ave/time 1 1 1 c_per_atom_stress[*] file centroid_stress.dat
this is correct way?
I don’t understand what you are asking here.
My statements are quite clear. If there was more information available and detailed instructions on how to implement the centroid stress for kspace styles, it would probably already exist. This is first and foremost a physics problem.