Higher-Order Boundary Conditions - Part 1

[cortner]

CC @thudso @jjbraun

I want to use this issue to discuss anything coming up in the implementation for the next paper.

TODO

• Linear elasticity solutions: I believe Tom has finished this, all tests pass ok
• CLE Dislocation predictor: I believe this is not yet running correctly, tentatively assigning this to Tom?
• Cauchy-Born: done
• PDE solver to solve at least for the next-order corrector
• Atomistic Green's Function
• Computation and convergence test for the dipole tensor
• Implementation of the predictor-corrector iteration for point defects
• Implementation of the predictor-corrector iteration for screw dislocations

Am I missing anything?

[cortner]

@thudso @jjbraun can either of you remind me whether we actually need the Lattice Green's function in a first instance? I can't quite remember what the conclusion was.

[thudso]

CLE dislocation predictors are running fine (tests pass); I don't think you're missing anything. If I remember right, I think we can do without the atomistic Green's function, but then we do definitely need to solve the PDE (but @jjbraun can confirm!).

[cortner]

so are you saying you are actually getting the correct decay rates for the forces?

[thudso]

Ah, sorry; I had forgotten about this issue. I will look into it. On 9 Mar 2018, at 20:58, Christoph Ortner <[email protected]<mailto:[email protected]>> wrote: so are you saying you are actually getting the correct decay rates for the forces? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub<#11 (comment)>, or mute the thread<https://github.com/notifications/unsubscribe-auth/ASQ2lwvT9aTfg6c3G43Mr5SPyyqSKbxoks5tcu0OgaJpZM4SjLV9>.

[cortner]

This shows the convergence of the dipole tensor for a vacancy in a Lennard-Jones FCC material. @jjbraun does this match your result? If not, what are possible reasons? (e.g. I didn't pre-relax, i.e. the reference crystal is not stress-free, could this be a problem?)