Functions
	latin_hypercube (n_samples, n_dims, rng)

	unit_to_physical (u, spec)

	physical_to_unit (val, spec)

	sample_to_params (unit_vec)

	atoms_to_loop_diameter_nm (mean_n_i)

	atoms_to_void_diameter_nm (mean_n_v)

	evaluate_cost (results, sim)

	run_single (params, timeout_s=180, run_id=0)

	append_log (log_path, row_dict)

	build_row (run_id, phase, params, cost, metrics, wall)

	main ()

Variables
	SCRIPT_DIR = Path(__file__).resolve().parent

	BASE_DIR = SCRIPT_DIR.parent.parent

float	OMEGA = 1.18e-29

float	B_111 = 2.482e-10

list	PARAM_SPEC

list	PARAM_NAMES = [p[0] for p in PARAM_SPEC]

float	TARGET_DOSE = 10.0

float	TARGET_N_LOOPS = 1.0e22

float	TARGET_D_I_NM = 10.0

float	TARGET_N_VOIDS = 4.0e21

float	TARGET_D_V_NM = 3.0

int	N_MIN_TEM = 10

tuple	LOG_FIELDS

Detailed Description

calibrate.py — Autonomous parameter calibration for RadCluster_2_1.

Explores the parameter space using Latin Hypercube Sampling, runs bin-moment
cluster dynamics simulations via the C++ solver, and identifies parameter
combinations that match experimental TEM data at 10 dpa for EUROFER97:

    SIA loop density  ~ 10^22 m^-3       diameter ~ 10 nm
    Void density      ~ 3-5 x 10^21 m^-3 diameter ~ 2-4 nm

Usage
-----
    cd radcluster_code
    python codes/calibrate.py              # default 20 + 10 runs
    python codes/calibrate.py --phase1 30  # more exploration
    python codes/calibrate.py --timeout 120  # 2 min per run

Function Documentation

◆ append_log()

calibrate.append_log	(	log_path,
		row_dict
	)

Append one row to the CSV log, creating headers if needed.

Definition at line 314 of file calibrate.py.

◆ atoms_to_loop_diameter_nm()

calibrate.atoms_to_loop_diameter_nm ( mean_n_i )

Convert mean SIA cluster size (atoms) to loop diameter (nm).

Definition at line 115 of file calibrate.py.

◆ atoms_to_void_diameter_nm()

calibrate.atoms_to_void_diameter_nm ( mean_n_v )

Convert mean vacancy cluster size (atoms) to void diameter (nm).

Definition at line 123 of file calibrate.py.

◆ build_row()

calibrate.build_row	(	run_id,
		phase,
		params,
		cost,
		metrics,
		wall
	)

Assemble a log row dict.

Definition at line 324 of file calibrate.py.

◆ evaluate_cost()

calibrate.evaluate_cost	(	results,
		sim
	)

Compute scalar cost measuring deviation from experimental targets at
TARGET_DOSE dpa.  Uses TEM-visible clusters only (n,m >= N_MIN_TEM).
Returns (cost, metrics_dict) or (inf, None) on failure.

Definition at line 131 of file calibrate.py.

◆ latin_hypercube()

calibrate.latin_hypercube	(	n_samples,
		n_dims,
		rng
	)

Simple LHS: stratified random in [0,1]^d.

Definition at line 62 of file calibrate.py.

◆ main()

calibrate.main ( )

Definition at line 340 of file calibrate.py.

◆ physical_to_unit()

calibrate.physical_to_unit	(	val,
		spec
	)

Map physical value back to [0,1].

Definition at line 83 of file calibrate.py.

◆ run_single()

calibrate.run_single	(	params,
		timeout_s = `180`,
		run_id = `0`
	)

Run one simulation with the given parameter dict.
Returns (cost, metrics, wall_time) or (inf, None, wall_time).

Definition at line 239 of file calibrate.py.

◆ sample_to_params()

calibrate.sample_to_params ( unit_vec )

Convert a [0,1]^d vector to a dict of physical parameters.

Definition at line 94 of file calibrate.py.

◆ unit_to_physical()

calibrate.unit_to_physical	(	u,
		spec
	)

Map u in [0,1] to physical value according to spec.

Definition at line 72 of file calibrate.py.

Variable Documentation

◆ B_111

float calibrate.B_111 = 2.482e-10

Definition at line 33 of file calibrate.py.

◆ BASE_DIR

calibrate.BASE_DIR = SCRIPT_DIR.parent.parent

Definition at line 26 of file calibrate.py.

◆ LOG_FIELDS

tuple calibrate.LOG_FIELDS

Initial value:

=  (
    ['run_id', 'phase', 'cost', 'wall_s']
    + PARAM_NAMES
    + ['N_loops', 'N_voids', 'mean_n_i', 'mean_n_v',
       'd_i_nm', 'd_v_nm', 'delta_FP', 'delta_He']
)

Definition at line 306 of file calibrate.py.

◆ N_MIN_TEM

int calibrate.N_MIN_TEM = 10

Definition at line 112 of file calibrate.py.

◆ OMEGA

float calibrate.OMEGA = 1.18e-29

Definition at line 32 of file calibrate.py.

◆ PARAM_NAMES

list calibrate.PARAM_NAMES = [p[0] for p in PARAM_SPEC]

Definition at line 59 of file calibrate.py.

◆ PARAM_SPEC

list calibrate.PARAM_SPEC

Initial value:

=  [
    # (name,         sheet,                 lo,    hi,   scale)
    # -- Production --
    ('eta',          'production_fission',  0.15,  0.35, 'linear'),
    ('f_cl_i',       'production_fission',  0.10,  0.55, 'linear'),
    ('f_cl_v',       'production_fission',  0.02,  0.15, 'linear'),
    # -- Diffusion: 1D glide & mobility cutoff --
    ('E_m_1D',       'diffusion',           0.01,  0.20, 'linear'),
    ('i_mobile',     'diffusion',           3,     100,  'int'),
    ('L_hat',        'diffusion',           20,    100,  'linear'),
    # -- Diffusion: solute trapping (dominant effect on SIA mobility) --
    ('c_C',          'diffusion',           1e-5,  5e-4, 'log'),
    ('E_b_C_SIA',    'diffusion',           0.20,  0.50, 'linear'),
    # -- Reactions: sinks and coalescence bias --
    ('rho_d',        'reactions',           1e12,  1e14, 'log'),
    ('Z_i',          'reactions',           1.02,  1.20, 'linear'),
    ('Z_ii',         'reactions',           1.0,   10.0, 'log'),
]

Definition at line 40 of file calibrate.py.

◆ SCRIPT_DIR

calibrate.SCRIPT_DIR = Path(__file__).resolve().parent

Definition at line 25 of file calibrate.py.

◆ TARGET_D_I_NM

float calibrate.TARGET_D_I_NM = 10.0

Definition at line 107 of file calibrate.py.

◆ TARGET_D_V_NM

float calibrate.TARGET_D_V_NM = 3.0

Definition at line 109 of file calibrate.py.

◆ TARGET_DOSE

float calibrate.TARGET_DOSE = 10.0

Definition at line 105 of file calibrate.py.

◆ TARGET_N_LOOPS

float calibrate.TARGET_N_LOOPS = 1.0e22

Definition at line 106 of file calibrate.py.

◆ TARGET_N_VOIDS

float calibrate.TARGET_N_VOIDS = 4.0e21

Definition at line 108 of file calibrate.py.

Functions

Variables