About GUI usage - Githubissues

code-whale commented 3 years ago

Hi,I am using GUI interface for simulation, and I have the following questions: 1.What is the unit in the Hamilton option? What does the value of mu_SPIN represent?For example, in the micro-magnetic simulation, the exchange effect is 8e-11J/m, how should I set this value in the Spirit GUI interface? I noticed that there is a Number of shells, could you teach me what this is used for? 2.After I changed the value in the z direction in the geometric parameter, I found that the magnetic moment could not be spread over the grid as in Webapp. Is this because of the display problem? 3.The geometry-surface option in the visualization only works when z=1, but fails when z is other values. Could you please tell me the reason? 4.Finally, I want to ask besides https://spirit-docs.readthedocs.io/en/latest/README.html provided in GUI tutorial: are there any other can refer to the tutorial? , I don't have a good understanding of how to write plot and input files, can you provide a more detailed tutorial? Thank you for your help!

code-whale commented 3 years ago

Also, I would like to ask if Spirit can simulate multi-layer structure, and if so, how to set it

MSallermann commented 3 years ago

I would recommend to give the input file specification a look https://spirit-docs.readthedocs.io/en/latest/core/docs/Input.html . You can make sense of some stuff in the GUI from there. For plots and python scripting, I would recommend to look at the examples in the ui-python folder. To answer the rest of your questions:

The units of the exchange in the Hamiltonian is meV per unique pair. The conversion of an exchange stiffness from micromagnetics into this system depends on the lattice structure and your pair configuration etc.
The "number of shells" option is simply a shortcut to define interaction pairs based on positions of the spins. If you have e.g 2 shells, the numbers can can enter are the interaction parameters J1 and J2 of the two shells. E.g in a square lattice:
```
J2  J1  J2           
J1   X  J1
J2  J1  J2
```
mu_s is the magnetic moment of the spins in multiples of Bohrs magneton. It is used for dipole-dipole interactions and the external magnetic field
Changing the geometry at runtime can lead to a bit of wonky behaviour. In your case I would try to deactivate and then reactivate the surface renderer

code-whale commented 3 years ago

@MSallermann ,Thank you very much for your help. I will further study the first few questions! I have a question about the multi-layer structure: can Spirit simulate FM-NM-FM model like Co-NM-Co, i.e. the finite element or finite difference method is to define the parameters of different regions. Thanks again for your reply!

MSallermann commented 3 years ago

I am not exactly sure what your requirements are.

Generally speaking, you can define a multilayer by adding additional basis atoms to your unit cell and defining the interaction pairs accordingly. For this, you have to edit the inputfile. It is not possible to do it directly in the GUI.

code-whale commented 3 years ago

@MSallermann ，I'm sorry that I have to deal with some personal matters these days, so I didn't reply to your message in time.Thank you for your help and patience. My problem is mainly the difference between micromagnetism and atomic dynamics. They all use the LLG equation. However, the micromagnetic simulation uses finite element or finite difference method to divide the grid, and spirit mainly considers the interaction between atoms, but in the micromagnetic simulation, the influence of demagnetization energy is generally considered, but I can not find the calculation of demagnetization energy in spirit, so what I want to teach is that if there is no demagnetization energy in spirit, will the calculation result be close to the actual result Forgive me for the lack of knowledge in this area, looking forward to your reply!

GPMueller commented 3 years ago

In the input file specs you can find the demagnetization energy under "dipole-dipole" interaction (note that others sometimes also call it "dipolar" or "stray-field" interaction). In the source code, the implementations have names containing "dipole".

To be clear: you can definitely scale the magnetic moments etc. such that the results of an atomistic simulation match very well with a micromagnetic simulation (in certain cases exactly) and you can do this with Spirit. However, the differences between atomistic and micromagnetic equations and how to translate between them are things you should best learn by reading papers, books and theses - explaining all of this is beyond the scope of GitHub issues.

code-whale commented 3 years ago

@GPMueller ，Thank you for your help! I will go to the paper reading and learning! I will carry on the study of input file first! Because I've only done some simple simulations in the GUI so far, using its own interface! Thanks again for your help!

GPMueller commented 3 years ago

Closing due to inactivity. Please open new issues for specific topics, if necessary.

spirit-code / spirit

About GUI usage #579