Question about defining the surface normal vector in viDMExchange

[Arturo]

Hi everyone,

I have a question about how the viDMExchange module works. In particular, it requires specifying the surface normal vector n. I'm a bit unsure about how n should be defined.

Suppose I have a thin film lying in the x-y plane with a thickness of 3 cells. For the top cell, should n be (0, 0, 1), for the bottom cell (0, 0, -1), and for the middle cell (0, 0, 0)? .  Should similar considerations be applied to the lateral surfaces as well?

thanks   

Mario

[Serban Lepadatu]

Hi Mario,

Technically it's possible to configure it in the way you describe, but physically I'm not sure if it's correct. This is an interfacial exchange interaction and it's meant to be used with thin films. The idea of viDMExchange module is to allow the user to configure the symmetry axis, so it's a generalization of the iDMExchange module for which the symmetry axis is z, i.e. only allows interfacial DMI for thin films stacked along the z direction. With viDMExchange you control this symmetry axis with the Ddir parameter. As an example, you could have a tube with the Ddir parameter defined radially outwards, which allows supporting skyrmions on a curved surface (e.g. https://www.boris-spintronics.uk/example-11/) - this is defined by setting a parameter spatial variation for Ddir (with a user-defined text equation).

Regards,

Serban