dl_mg_utils Module Reference

auxiliary subroutines More...

Functions/Subroutines
subroutine	dl_mg_init_nonlin (eq_type, lambda, temp, n, c, q, rho, steric_weight, der_pot, expcap, linearised, use_fas, neutralisation_method, neutralisation_ion_ratios, ierror)
	Initialises the non-linear parameters. It can be called multiple times, e.g. if beta needs to be changed in continuation calculation. More...
subroutine	compute_rho_sum (rho)
subroutine	compute_steric_weight_sum (has_stw, steric_weight)
	computes the accesible volume More...
subroutine	test_charge_neutrality (rho, pot, steric_weight)
	called by the PBE solver More...
subroutine	remove_zero_mode_source (mg, eq_type, remove_mode, rho, steric_weight, der_pot)
	removes zero modes from source term. Needed when using full PBC More...
subroutine	remove_zero_mode_potential (mg, eq_type)
	removes zero mode from the potential. Needed when using full PBC at the moment this is called only inside vcycle for mgp it migh need to be extend for arbitrary arrays with optional args N.B. the potential shift is applied also to the halo points to save a halo exchange. This is so because this sub is called immediately after mg_relax which does a full halo exchange. If used in other location the halo shift might need to be adjusted More...
real(wp) function	compute_norm (mg, component, scale)
subroutine	return_betamu_gamma_ion_concentrations (n, betamu, ion_concentration, neutralisation_ratios, steric_weight_average)
	needed in dl_mg_solver_pbe if the arguments betamu or gammas are present set them to the values used in code added the ion concentrartions used internally (it might differ from the input value for counter-ion neutralisation, when using PBC) More...
subroutine	check_error_code (ierror)
	checks if the first error code bubbled up to the top level or if it was replaced More...
subroutine	mg_convergence_test (mg, level, test_tag, iter, converged, f_norm, res_norm, sol)
	convergence tests, for details see dl_mg_convergence_params More...
subroutine	set_boundary_values (mg, v, v_type, val, aval, av_type)
subroutine	set_mass_term (mg, der_pot, ierr)
	computes mass term at non-zero potential used in linearized PBEs (Newton and deferred correction) More...

Detailed Description

auxiliary subroutines

Function/Subroutine Documentation

check_error_code()

subroutine dl_mg_utils::check_error_code

(

integer, intent(inout), optional

ierror

)

checks if the first error code bubbled up to the top level or if it was replaced

compute_norm()

real(wp) function dl_mg_utils::compute_norm	(	type(mg_t), intent(in), target	mg,
		character(len=1), intent(in)	component,
		logical, intent(in), optional	scale
	)

Todo:

a_ptr can have contiguous attribute

compute_rho_sum()

subroutine dl_mg_utils::compute_rho_sum

(

real(wp), dimension(isx:,isy:,isz:), intent(in)

rho

)

compute_steric_weight_sum()

subroutine dl_mg_utils::compute_steric_weight_sum	(	logical, intent(out)	has_stw,
		real(wp), dimension(isx:, isy:, isz:), intent(in), optional	steric_weight
	)

computes the accesible volume

\[ ~ \int_V steric\_weight(r) dv / (V/(nx*ny*nz)) \]

Attention: this sum is not multipliplied with dv = dx * dy * dz because it's used as a normalisation factor for quatites which are also not multiplied with dv

dl_mg_init_nonlin()

subroutine dl_mg_utils::dl_mg_init_nonlin	(	integer, intent(out)	eq_type,
		real(wp), intent(in)	lambda,
		real(wp), intent(in)	temp,
		integer, intent(in)	n,
		real(wp), dimension(n), intent(in)	c,
		real(wp), dimension(n), intent(in)	q,
		real(wp), dimension(:,:,:), intent(in)	rho,
		real(wp), dimension(:,:,:), intent(in), optional	steric_weight,
		real(wp), dimension(:,:,:), intent(in), optional	der_pot,
		real(wp), intent(in), optional	expcap,
		logical, intent(in), optional	linearised,
		logical, intent(in), optional	use_fas,
		integer, intent(in), optional	neutralisation_method,
		real(wp), dimension(n), intent(in), optional	neutralisation_ion_ratios,
		integer, intent(out), optional	ierror
	)

Initialises the non-linear parameters. It can be called multiple times, e.g. if beta needs to be changed in continuation calculation.

This subroutine needes to be called also in the case of linearised PBE.

Note: from physical point of view it is sensible to assume that the solvent is electrically neutral, i.e. \( \sum_i c_i q_i = 0 \), however the neutrality condition is not checked.

Todo:

Test for charge neutrality?

for arguments description see dl_mg::dl_mg_solver_pbe

mg_convergence_test()

subroutine dl_mg_utils::mg_convergence_test	(	type (mg_t), intent(inout), target	mg,
		integer, intent(in)	level,
		integer, intent(in)	test_tag,
		integer, intent(in)	iter,
		integer, intent(out)	converged,
		real(wp), intent(in), optional	f_norm,
		real(wp), intent(in), optional	res_norm,
		real(wp), dimension(isx:,isy:,isz:), intent(in), optional, target	sol
	)

convergence tests, for details see dl_mg_convergence_params

remove_zero_mode_potential()

subroutine dl_mg_utils::remove_zero_mode_potential	(	type(mg_t), intent(inout), target	mg,
		integer, intent(in)	eq_type
	)

removes zero mode from the potential. Needed when using full PBC at the moment this is called only inside vcycle for mgp it migh need to be extend for arbitrary arrays with optional args N.B. the potential shift is applied also to the halo points to save a halo exchange. This is so because this sub is called immediately after mg_relax which does a full halo exchange. If used in other location the halo shift might need to be adjusted

remove_zero_mode_source()

subroutine dl_mg_utils::remove_zero_mode_source	(	type(mg_t), intent(inout), target	mg,
		integer, intent(in)	eq_type,
		integer, intent(in)	remove_mode,
		real(wp), dimension(isx:, isy:, isz:), intent(inout), optional, target	rho,
		real(wp), dimension(isx:, isy:, isz:), intent(in), optional, target	steric_weight,
		real(wp), dimension(isx:, isy:, isz:), intent(in), optional	der_pot
	)

removes zero modes from source term. Needed when using full PBC

Parameters

[in]

remove_mode

1 (first call) or 2 for call in inner loops (defco or newton)

Todo:

analyse furter eq_linear_pbe_pot in as primary input

return_betamu_gamma_ion_concentrations()

subroutine dl_mg_utils::return_betamu_gamma_ion_concentrations	(	integer, intent(in)	n,
		real(wp), dimension(n), intent(out), optional	betamu,
		real(wp), dimension(n), intent(out), optional	ion_concentration,
		real(wp), dimension(:), intent(out), optional	neutralisation_ratios,
		real(wp), intent(out), optional	steric_weight_average
	)

needed in dl_mg_solver_pbe if the arguments betamu or gammas are present set them to the values used in code added the ion concentrartions used internally (it might differ from the input value for counter-ion neutralisation, when using PBC)

set_boundary_values()

subroutine dl_mg_utils::set_boundary_values	(	type(mg_t), intent(inout)	mg,
		real(wp), dimension(:,:,:), intent(inout)	v,
		integer, intent(in)	v_type,
		real(wp), intent(in), optional	val,
		real(wp), dimension(:,:,:), intent(in), optional	aval,
		integer, intent(in), optional	av_type
	)

set_mass_term()

subroutine dl_mg_utils::set_mass_term	(	type(mg_t), dimension(:), intent(inout)	mg,
		real(wp), dimension(isx:, isy:, isz:), intent(in)	der_pot,
		integer, intent(out), optional	ierr
	)

computes mass term at non-zero potential used in linearized PBEs (Newton and deferred correction)

test_charge_neutrality()

subroutine dl_mg_utils::test_charge_neutrality	(	real(wp), dimension(:,:,:), intent(in)	rho,
		real(wp), dimension(:,:,:), intent(in)	pot,
		real(wp), dimension(:,:,:), intent(in), optional	steric_weight
	)

called by the PBE solver