pyHarm.Analysis.FRF_NonLinear¶

Classes¶

FRF_NonLinear

Nonlinear forced response analysis.

Module Contents¶

class pyHarm.Analysis.FRF_NonLinear.FRF_NonLinear(inputData: dict, System: pyHarm.Systems.ABCSystem.ABCSystem, logger: logging.Logger | None = None, key: str = '', **kwargs)¶

Bases: pyHarm.Analysis.ABCAnalysis.ABCAnalysis

Nonlinear forced response analysis.

Predicts a starting point using the predictor and then solves the residual equations from the starting point.

flag_print¶

if True, prints a message after each Solve method.

Type:: bool

flag_purge¶

if True, the jacobian values are purged from the solutions stored in the solution list.

Type:: bool

analysis_options¶

input dictionary completed with the default values if keywords are missing.

Type:: dict

ds¶

initial prediction step size.

Type:: float

SolList¶

list of SystemSolution stored after each Solve method.

Type:: list[SystemSolution]

system¶

ABCSystem associated with the analysis.

Type:: ABCSystem

adaptstep¶

ABCStepSizeRule associated with the analysis.

Type:: ABCStepSizeRule

predictor¶

ABCPredictor associated with the analysis.

Type:: ABCPredictor

corrector¶

ABCCorrector associated with the analysis.

Type:: ABCCorrector

stopper¶

ABCStopCriterion associated with the analysis.

Type:: ABCStopCriterion

solver¶

ABCNLSolver associated with the analysis.

Type:: ABCNLSolver

corrector_init¶

instance of no-continuation corrector for initial solution.

Type:: ABCCorrector

reductor¶

ABCChainedReductor associated with the analysis.

Type:: ABCChainedReductor

factory_keyword: str = 'frf'¶

keyword that is used to call the creation of this class in the system factory.

Type:: str

name = 'Nonlinear FRF'¶

default¶

default dictionary for the analysis.

Type:: dict

key = ''¶

analysis_options¶

flag_print¶

flag_purge¶

logger: logging.Logger¶

ds¶

SolList = []¶

system¶

adaptstep¶

predictor¶

corrector¶

stopper¶

solver¶

corrector_init¶

reductor¶

initialise(x0=None, **kwargs)¶

First Solve on the initial guess.

Parameters:: x0 (None|str|np.ndarray) – if None x0 is the linear solution at initial angular frequency, if “null” x0 is null vector, otherwise x0 is initialised using the provided array.

_get_x0(x0)¶

CompleteSolution(sol: pyHarm.Solver.SystemSolution)¶

Completes a SystemSolution informations that just went out of the Solve method.

Parameters:: sol (SystemSolution) – solution to the system.

globalResidual(solx: numpy.ndarray, sol: pyHarm.Solver.SystemSolution)¶

Computes the residual of the whole system without the reducers.

Parameters:

solx (np.ndarray) – displacement vector for which residual is computed.
sol (SystemSolution) – actual SystemSolution being solved.

globalJacobian(solx: numpy.ndarray, sol: pyHarm.Solver.SystemSolution)¶

Computes the jacobians of the whole system without the reducers.

Parameters:

solx (np.ndarray) – displacement vector for which residual is computed.
sol (SystemSolution) – actual SystemSolution being solved.

globalJacobianwRed(solq: numpy.ndarray, sol: pyHarm.Solver.SystemSolution)¶

Computes the reduced jacobians of the whole system.

Parameters:

solq (np.ndarray) – reduced displacement vector for which residual is computed.
sol (SystemSolution) – actual SystemSolution being solved.

globalResidualwRed(solq: numpy.ndarray, sol: pyHarm.Solver.SystemSolution)¶

Computes the reduced residuals of the whole system.

Parameters:

solq (np.ndarray) – reduced displacement vector for which residual is computed.
sol (SystemSolution) – actual SystemSolution being solved.

_update_reductor(xpred_full, last_solution_pointer=None)¶

Updates the reducers

Parameters:

xpred_full (np.ndarray) – full starting displacement vector.
last_solution_pointer (SystemSolution) – SystemSolution from which the prediction has been generated.

Returns:

reduced starting displacement vector. J_red (np.ndarray): reduced jacobian matrix. output_expl_dofs (pd.DataFrame): DataFrame of the dofs after applying the reduction layers.

Return type:

xpred_red (np.ndarray)

makeStep(**kwargs)¶: Makes a step of solving : get a step size, generate a predicted point, solve the nonlinear system, save the solution.

Solve(x0=None, **kwargs)¶

Makes the whole analysis using continuation techniques until the stopping criterion is validated

Parameters:: x0 (None|str|np.ndarray) – if None x0 is the linear solution at initial angular frequency, if “null” x0 is null vector, otherwise x0 is initialised using the provided array.

purge_jacobians()¶: Purge the Jacobians of Solutions that are no longer used.

export(export_path: str, prefix: str, **kwargs) → None¶: export resutlts in csv files

_init_log()¶

_end_log()¶

_export_log()¶