Research Workflow
A Zettelkasten-inspired system for theoretical and computational research.
Capturing Research Ideas
When an idea strikes, dump it in inbox.md:
# inbox.md
- [ ] Idea: What if we apply Monte Carlo to the XY model with disorder? :research: :idea:
- [ ] Question: How does DFT handle strongly correlated systems? :research: :question:
- [ ] Follow up: Re-read Anderson localization paper :literature:
Processing into Zettelkasten
After processing, create atomic notes:
# research/concepts/anderson-localization.md
# Anderson Localization
## Definition
The absence of diffusion of waves in a disordered medium.
## Key Physics
In a sufficiently disordered system, quantum interference effects cause
wavefunctions to become exponentially localized...
## Mathematical Framework
$$\psi(r) \sim e^{-|r-r_0|/\xi}$$
where $\xi$ is the localization length.
## Computational Approaches
- Exact diagonalization for small systems
- Transfer matrix method for quasi-1D
- Kernel polynomial method for large systems
## Related Concepts
- [[metal-insulator-transition]]
- [[weak-localization]]
- [[random-matrix-theory]]
## Key Papers
- [[research/literature/anderson1958-absence-diffusion]]
- [[research/literature/abrahams1979-scaling-theory]]
---
:concept: :condensed-matter: :disorder:
Literature Management
When reading a paper:
- Create note:
research/literature/[author][year]-[short-title].md - Fill in template as you read
- Link to relevant concept notes
- Add to
research/inbox.md: follow-up papers to read
Computational Research Notes
Example structure for research/methods/molecular-dynamics.md:
# Molecular Dynamics
Overview: Classical simulation of particle motion using Newton's equations.
Algorithm:
- Initialize positions and velocities
- Calculate forces from potential
- Integrate equations of motion
- Repeat
Code Snippet (Python/ASE):
from ase import Atoms
from ase.md.velocitydistribution import MaxwellBoltzmannDistribution
from ase.md.verlet import VelocityVerlet
# Setup
atoms = ...
MaxwellBoltzmannDistribution(atoms, temperature_K=300)
# Run
dyn = VelocityVerlet(atoms, timestep=1.0 * units.fs)
dyn.run(steps=1000)
Common Pitfalls:
- Time step too large → energy drift
- Poor thermostat choice → incorrect ensemble
- Finite size effects
Packages: ASE, LAMMPS, GROMACS
Related concepts: verlet-integration, thermostats, periodic-boundary-conditions
Tags: :method: :simulation: :computational:
Thesis Progress Tracking
Create school/thesis/progress.md with this structure:
| Section | Content |
|---|---|
| Working Title | Your thesis title |
| Committee | Advisor + members |
| Timeline | Milestones table (lit review, proposal, defense, etc.) |
| Chapters | Checklist of chapters |
| Current Focus | What you're working on this week |
| Open Questions | Research questions to resolve |
| Meeting Log | Date, discussed, action items |
Example Timeline Table:
| Milestone | Target | Status |
|---|---|---|
| Literature review | ||
| Proposal | ||
| First results | ||
| Draft to advisor | ||
| Defense |
Tags: :thesis: