RKKY coupling between FM with different structure

36 views
Skip to first unread message

Hying

unread,
Nov 23, 2025, 8:39:34 PMNov 23
to mumax2
Dear all,
From mumax3 workshop, the prefactor and shift functions are defined as following for FM(1nm)/SL(1nm)/FM(1nm): 
prefactorZ := Const( (2 * AFMAex) / (cellsize*cellsize*Ms))
up := Mul(prefactorZ,Mul(Add(Mul(Const(-1),m),Shifted(m,0,0, 2)),Shifted(Const(1),0,0,2)))
down :=  Mul(prefactorZ,Mul(Add(Mul(Const(-1),m),Shifted(m,0,0,-2)),Shifted(Const(1),0,0,-2)))
freelayerthickness=2e-9
setgridsize(256,64,3)
setcellsize(1e-9,1e-9,1e-9)
// define 2 layers
defregion(1,layer(2))
defregion(2,layer(0))
setgeom(layer(0).add(layer(2)))

I’m not sure that How should I modify the up and down functions for my structure: FM(2nm)/SL(1nm)/FM(2nm):
freelayerthickness=4e-9
setgridsize(8,8,5)
setcellsize(3.75e-9,3.75e-9,1e-9)
// define 2 layers
top:=cylinder(30e-9,2e-9).transl(0,0,1.5e-9)
bottom:=cylinder(30e-9,2e-9).transl(0,0,-1.5e-9)
defRegion(1, bottom)
defRegion(2, top)
setgeom(top.add(bottom))
 
Thank you very much in advance for any help!
Best, Hying

Josh Lauzier

unread,
Nov 29, 2025, 5:32:42 PM (13 days ago) Nov 29
to mumax2
Hi,

You would want to mask the custom parts to only effect the two layers on either side of the spacer, as discussed in these previous discussions. So for your structure, something like:


up := Mul(prefactorZ,Mul(Add(Mul(Const(-1),m),Shifted(m,0,0, 2)),Shifted(Const(1),0,0,2)))
down :=  Mul(prefactorZ,Mul(Add(Mul(Const(-1),m),Shifted(m,0,0,-2)),Shifted(Const(1),0,0,-2))) 

Maskeddown:= Masked( layer(1),down)
Maskedup:= Masked (layer(3),up)
AddFieldTerm(Maskeddown)
AddFieldTerm(MaskedUp)

(Actually, in your special case, you technically don't actually even need to bother to do this. There is nothing for layer 0 and layer 4 to couple to, so you can get away with being lazy. But if you had anymore than 5 layers, this is how you would do it)

Best,
Josh L.

Hying

unread,
Dec 5, 2025, 11:15:20 PM (6 days ago) Dec 5
to mumax2
Hi Josh,
Thank you very much for your reply!
I did the test, and it's the same as you said. I got the same switching time for the simulations. There is nothing for layer 0 and layer4 to couple!
 
Thank you again and best wishes!
Hying
Reply all
Reply to author
Forward
0 new messages