Pixel

The pixel object is used to define the geometry of the system being optimized. There are two major motivations for setting up the geometry input in this manner:

1. Traditional and some advanced nuclear system designs often leverage repeating lattice structures

2) The most inspiring goal for this optimization work is to optimize the shape of fuel for fuel performance. The methodology for which requires the pixelation of geometric space such that the algorithm can step towards a continuous shape in the limit of infinitesimal pixel size.

The pixel object houses information about the materials/regions within the repeating unit. A number of methods are also included that perform functions such as applying the update of materials given some set of transformed optimization parameters. This object-oriented structure also helps with bookeeping whenever the number of pixels, and therefore optimization parameters, becomes very large.

class ADAM.pixel.pixel(region_definition, parameter_definition, material_df_base, pixel_id, temperature)

Summary of pixel class.

More on pixel class.

apply_optimization_parameters_to_material_definitions:

Applies the optimization parameters to the base material values.

write_material_string:

Creates the updated material string attribute.

combine_derivatives_wrt_nuclides:

Combine derivatives for each nuclide in a material.

combine_region_derivatives:

Combines derivatives for each region with a given optimization parameter.

Methods

apply_density_factors(density_factors)	Generates updated isotopic concentration values for each region within a pixel.
combine_derivatives_wrt_nuclides(...)	Combines absolute sensitivities/derivatives with respect to each nuclide within each region of the pixel object.
combine_region_derivatives()	Combines the derivatives across regions for the same parameters given the function for bomining derivatives in the objective_function_definition module.

apply_density_factors(density_factors)

Generates updated isotopic concentration values for each region within a pixel.

The updated_materials_definition attribute for the pixel object is created. This dataframe is created by multiplying the density factors (transformed optimization parameters) to the base isotopic concentrations (material_df_base).

\[N_{updated} = N_{base}*T(\theta)\]

Where T is the user defined transformation function.

Parameters

density_factors (DataFrame) – Transformed optimization parameters, aka density factors.

Return type

None.

combine_derivatives_wrt_nuclides(material_dict_base)

Combines absolute sensitivities/derivatives with respect to each nuclide within each region of the pixel object.

This function goes into the

Parameters

material_dict_base (TYPE) – DESCRIPTION.

Return type

None.

combine_region_derivatives()

Combines the derivatives across regions for the same parameters given the function for bomining derivatives in the objective_function_definition module.

Return type

None.