When I find out spin-spin correlation i.e. <S_i.S_{i+1}>(nearest-neighbour), even for large system size it shows an oscillatory behaviour. Though it is oscillating around the correct value which is “-0.44314”, but shouldn’t it show a constant value in the bulk?? Because this behaviour even for large system size leads to a false dimerization in the bulk.
What model is it (e.g. is it spin 1/2 or spin 1?) and how does the behavior change as you consider larger system sizes? If it’s the S=1/2 case, that is a gapless system so the edge effects may decay very slowly. Do your results change if you perform more DMRG sweeps and make the calculation more accurate?
It is spin 1/2 case… Even if I increase the system size and sweeps the decay of these oscillations is very small.
I think this may just be the correct behavior of this system, due to the combination of open boundaries and the fact that it is a critical / gapless system. Does that answer your question about it?
yes. I understand your point.
Thank you.