elastic constants using an NVT - why no poisson effect?


I am seeing some studies attempting to obtain the C11 elastic constant using an fix deform in the x direction in an NVT. An NVT thermostat controls the lateral strains to zero, thus not allowing for poisson’s effect. Then the slope of stress/strain can be used to get C11. It seems very odd to me to not allow for poisson’s contraction when applying a uniaxial deformation.

When I replicate this with an NPT obviously the results are very different. Why is it ok to fix strains in lateral axes to zero when finding C11 from an NVT. This appears to also be what is done in the in.elastic script so perhaps Dr. Thompson can comment.

Thanks very much.