lincs-order 4 is too conservative for some force fields
For example, ClayFF has bonds only in the hydroxyl of its silicate. If I understand correctly, it would only need a LINCS order of 1.
Is there a way we can usefully observe that the bonded connectivity is low, and to have grompp suggest that the user can use a lower lincs-order? In turn, that affects the minimum cell size that the domain decomposition requires.
That would also impact a simulation of diatomic gases, etc.