Processing math: 100%

User Tools

Site Tools


exercises:common:sgcp

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revisionPrevious revision
Next revision
Previous revision
exercises:common:sgcp [2025/06/19 02:27] jglanexercises:common:sgcp [2025/06/19 02:54] (current) – [2. Comparison with CPMD and BOMD] jglan
Line 9: Line 9:
 [[ https://doi.org/10.1021/acs.jpclett.0c01025 |J. Phys. Chem. Lett. 2020, 11, 9, 3724–3730]] [[ https://doi.org/10.1021/acs.jpclett.0c01025 |J. Phys. Chem. Lett. 2020, 11, 9, 3724–3730]]
  
-===Introduction===+====1. Introduction====
 Second Generation CPMD (2ndG CPMD) is a molecular dynamics method that combines the efficiency of Car-Parrinello MD (CPMD) with the accuracy of Born-Oppenheimer MD (BOMD). It avoids fully self-consistent field (SCF) optimizations at each time step while enabling larger integration steps and maintaining accuracy close to BOMD. Second Generation CPMD (2ndG CPMD) is a molecular dynamics method that combines the efficiency of Car-Parrinello MD (CPMD) with the accuracy of Born-Oppenheimer MD (BOMD). It avoids fully self-consistent field (SCF) optimizations at each time step while enabling larger integration steps and maintaining accuracy close to BOMD.
  
-| Feature                          | CPMD          | BOMD         2ndG CPMD                + 
-|----------------------------------|---------------|--------------|--------------------------+**Goal:** Retain the efficiency of CPMD while achieving BOMD-level accuracy. 
-| SCF at each step                 ❌ No          | ✅ Yes        | ❌ No (predictor-corrector) |+ 
 +- **Efficiency**:  Large time steps ; No full SCF loops   
 + 
 +- **Accuracy**:  Forces nearly indistinguishable from BOMD   
 + 
 +- **Stability**:  Effective for systems with vanishing band gaps   
 + 
 +- **Error Control**:  Controlled deviation from BO surface using adaptive correction 
 + 
 +====2. Comparison with CPMD and BOMD==== 
 + 
 +| Feature                          | CPMD          | BOMD         SGCP                | 
 +| SCF at each step                 | No          | Yes        |  Partially (predictor-corrector) |
 | Time step                        | Small (~0.1 fs) | Large (~1 fs) | Large (~1–2 fs)           | | Time step                        | Small (~0.1 fs) | Large (~1 fs) | Large (~1–2 fs)           |
 | Conserved quantity preservation  | Excellent     | Reasonable   | Excellent                 | | Conserved quantity preservation  | Excellent     | Reasonable   | Excellent                 |
Line 20: Line 32:
 | Works for small-gap systems      | Poor          | Good         | Good                      | | Works for small-gap systems      | Poor          | Good         | Good                      |
  
 +====3. ASPC Method====
 +ASPC Method: Always Stable Predictor Corrector
 +
 +ASPC is a **Gear-type integrator** for electronic wavefunctions:
 +
 +Predictor:
 +
 +\[
 +C_p(t_n) = \sum_{m=1}^{K} (-1)^{m+1} \cdot m \cdot B_m \cdot P_S(t_{n-m})
 +\]
 +
 +where:
 +- Bm: Kolafa predictor coefficients
 +- PS: projection onto the overlap matrix S
 +
 +Corrector:
 +
 +\[
 +C(t_n) = \omega \cdot \min[C_p(t_n)] + (1 - \omega) \cdot C_p(t_n), \quad \omega = \frac{K}{2K - 1}
 +\]
 +
 +Langevin Dynamics & Dissipation Compensation
 +
 +Because ASPC introduces small dissipation, **Langevin-type equations** are used to stabilize the dynamics:
 +
 +\[
 +M_I \ddot{R}_I = F_\text{BO} - (\gamma_D + \gamma_L)\dot{R}_I + \Xi_I
 +\]
 +
 +- γD: implicit friction from ASPC  
 +- γL: Langevin thermostat  
 +- ΞI: Langevin random noise
 +
 +
 +
 +====4. How to Set Up in CP2K====
 +
 +
 +| Parameter              | Purpose                                  | Notes                                      |
 +| EXTRAPOLATION_ORDER  | Higher gives better predictor            | 1–4 typical, 0 for metallic is more stable                               |
 +| MAX_SCF_HIST        | Controls SCF correction                  | ≥2 helps smoother convergence              |
 +| STEPSIZE             | Time step in fs                          | ~0.5–2 fs depending on system                |
 +| PRECONDITIONER       | Affects SCF convergence                  | `FULL_SINGLE_INVERSE` slightly better      |
 +| NOISY_GAMMA (γ_D)    | ASPC dissipation compensation            | Adjust to control drift in T and energy    |
 +| GAMMA (γ_L)          | Langevin thermostat strength             | Set to 0 for dissipation-only integration  |
 +
 +1. ASPC Extrapolation
 +
 +  &FORCE_EVAL
 +    &DFT
 +       &QS
 +        EXTRAPOLATION ASPC
 +        EXTRAPOLATION_ORDER 0 # Higher gives better corrector 
 +       &END QS
 +       &SCF
 +        MAX_SCF_HIST 2
 +       &END SCF
 +    &END DFT
 +  &END FORCE_EVAL
 +
 +2. Langevin Thermostat
 +
 +  &MOTION
 +    &MD
 +      ENSEMBLE LANGEVIN
 +      &LANGEVIN
 +        GAMMA 0.005         ! γ_L
 +        NOISY_GAMMA 4.0E-4  ! γ_D
 +      &END LANGEVIN
 +    &END MD
 +  &END MOTION
 +
 +3. Atom-Specific γ_D (Optional)
 +
 +  &THERMAL_REGION
 +    DO_LANGEVIN_DEFAULT TRUE
 +    &DEFINE_REGION
 +      TEMPERATURE 500
 +      NOISY_GAMMA_REGION 4.E-4
 +      LIST 577..745
 +    &END DEFINE_REGION
 +  &END THERMAL_REGION
  
exercises/common/sgcp.1750300053.txt.gz · Last modified: by jglan