Hi, Axel:
Thanks a lot for your reply. I searched online and found that somebody calculated the tip3p water viscosity as 0.3 cp using NEMD method using GROMACS. Mine using lammps is even worse. I used 1000 water molecules, and ran the simulation for at least 2 million steps. When the shear rate is too large, I saw spikes on the pxy curve. Otherwise, the fluctuation of pxy is smooth and consistently. So it looks like I do not have simulation problem. So where is wrong? Did anyone report their calculations for tip3p before? Thanks a lot.
Lqz