Summary for: ThermalModel < handle
Class summary
ThermalModel Basic model-problem class for thermal problems.
Supports coupled FEA-(lumped) network models.
IMPORTANT As the FEA-part of the model is 2-dimensional, it works with loss densities and heat flux densities per unit length, rather than losses and fluxes. Importantly, any lumped-network components in the model must be scaled for 1m axial length, and the number of symmetry sectors used.
Properties
.ambient_node Network node(s) corresponding to the ambient temp.
.effective_length Effective length in the out-of-plane dimension.
.element_heat_generation Array of elementwise heat generation
values.
W / m^3
.element_thermal_conductivity Array of elementwise thermal
conductivities.
.element_volumetric_heat_capacity Array of elementwise heat
capacities.
Units in W / deg m^3
.ThermalModel/free_nodes is a property.
.matrices Struct of matrices and arrays.
For holding all kinds of useful stuff.
.mesh FEA mesh
.nodes Array of lumped nodes in this model.
.ThermalModel/number_of_mesh_nodes is a property.
.ThermalModel/number_of_network_nodes is a property.
.symmetry_sectors Number of symmetry sectors
Methods
Class methods are listed below. Inherited methods are not included.
.add_node Add node to model.
add_node(this, node)
.assemble_FEA_matrices Assemble FEA matrices.
Results stored to this.matrices.
.assemble_final_matrices Assemble and return final problem matrices.
[C, M, F] = assemble_final_matrices(this), where
- C : thermal resistivity matrix
- M : thermal capacitance matrix
- F : heat source/sink vector
The matrices contain the contributions from the pure-FEA part, and optionally summing the contributions from this.nodes, obtained by calling node.get_matrices() separately for each node.
.assemble_source_vector Assemble heat source vector, pure-FEA part.
.compute_sensitivity Simple sensitivity analysis.
compute_sensitivity(this, solution) computes simple sensitivity analysis for the given solution. The analysis is computed by going through all nodes (ThermalNode) and associated connections (ThermalNetworkConnection) in the model one by one, increasing the correction_coefficient property by 10 percent, and re-solving the model for each case. The correction coefficient is returned to its original value after solving.
The results are then printed in the command line by picking the 20 largest changes in the node temperatures, and printing associated node and connection names plus the increase in temperature. The print format is: (Node/Connection)
[dT, names, elements] = compute_sensitivity(model, solution, verbose) returns the array dT, of size number_of_network_nodes x number of nodes + connections, a string array names of the node and connection names, and an array of the nodes and connections corresponding to the names.
.ThermalModel/parse_boundary_matrix is a function.
parse_boundary_matrix(this)
.plot_thermal_conductivity Plot conductivity.
Plot elementwise thermal conductivities.
plot_thermal_conductivity(this, varargin)
.set_ambient Set ambient temperature node.
set_ambient(this, node)
.solve_steadystate Solve steady-state temperatures.
solution = solve_steadystate(this) solves the current thermal model in steady-state and returns the solution object.