bond_style table: Bond length < table inner cutoff(../bond_table.cpp:618)

LAMMPS LAMMPS Mailing List Mirror
1

Dear Lammps users,

I recently downloaded the latest stable version of lammps (7Aug19) built with the MOLECULE package. I get the following error message using bond_style table. (The inner cutoff of this potential table is 11.5 A)

ERROR on proc 3: Bond length < table inner cutoff: type 1 length 11.4634 (…/bond_table.cpp:618)

I am confused. In the manual it says: “If the actual bond length is ever smaller than the LO value or larger than the HI value, then the bond energy and force is evaluated as if the bond were the LO or HI length.”, So I was expecting for the offending bond (11.4634 A) to be treated as 11.5 A bond but I get an error instead. This error does not show up if I use my previous lammps version (14May16).

I wonder if I am missing something or this is a “bug” in the code or in the manual… What do you think?

Best regards,

Pablo.

2

Pablo,

I would call this a “bug” in the manual. The behavior was changed in October 2016 ( https://github.com/lammps/lammps/commit/991034b632c7c36a283082416e2575eefbd4d27e ) .
The previous behavior is problematic, as it can result in undefined behavior and users won’t know about it. It is easy to emulate the previous behavior by creating a suitably extended table, but people that are not aware of the non-error behavior when crossing tabulation boundaries would get bogus trajectories without any warning.

We will update the documentation. Thanks for reporting.

Axel.

3

Dear Axel,

Thanks a lot for your quick response and explanation. I will create an extended table to handle the extreme cases (which is what we were doing for the rest of degrees of freedom of the system)

Best,
Pablo.