Summary for: SteelLibrary < handle
Class summary
SteelLibrary Static class for creating electrical steel Material objects.
Data referenced from Tata Steel website at some point, see https://www.tatasteeleurope.com/ts/engineering/products/electrical-steel
The underlying data is stored in an Excel file, under the ‘Materials’ folder. Modifying this data file is not recommended, as changes will be lost with any EMDtool update, unless special care is taken not to overwrite the modified fiel.
Usage:
mat = SteelLibrary.create( grade_name )
where grade_name = string or char. See SteelLibrary.grades for available grades.
mat = SteelLibrary.create(grade_name, ‘plot_fit’, true)
to also plot the resulting loss fit
Properties
.SteelLibrary/fname is a property.
.SteelLibrary.grades is a property.
Methods
Class methods are listed below. Inherited methods are not included.
.create Create new material from the lib.
material_object = create(grade_name) create a Material object of the corresponding grade. See SteelLibrary.grades for the available grades.
material_object = create(grade_name, ‘optimize_exponents’, true) to also optimize the time-domain generalized Steinmetz model exponents, used for the hysteresis losses.
See also SteelLibrary.parse_data_from_Excel and fit_losses.fit_losses for extra optional arguments.
Optional arguments include:
- ‘fit_analytical_BH’ : false (DEFAULT) / true. Replace material BH data by fitting an analytical BH curve to the data with
this.generate_BH_langevin. - ‘file_to_use’ : use a custom datasheet file (.xlsx) name/path.
.fit_losses Fit loss data.
[coeffs, c_opt] = fit_losses(this, B, fs, Ws, varargin) fits the frequency-domain Bertotti model to the given material loss data.
Input arguments:
- B : a 1 x Nb array of flux density values, assumed to be peak-of-sine.
- fs : a 1 x Nf array of frequencies.
- Ws : a Nb x Nf array of loss densities, as W/kg.
Returns:
- coeffs : coefficients for the Bertotti loss model, with the loss density expressed as
coeffs(1)*f^a*B^b + coeffs(2)*f^2*B^2 + coeffs(3)*f^1.5*B^1.5 - copt : optimized hysteresis loss exponents
[a b]. See below.
Optional key-value arguments:
- ‘plot_fit’ : visualize fit. Default false.
- ‘include_excess’ : include the excess loss term. Default false.
- ‘normalize_losses’ : Perform normalized least-squares when fitting the losses. When set to true, better agreement is typically obtained at low flux densities (corresponding to better prediction of the hysteresis loss coefficient), at the cost of decreased accuracy at high loss densities. Default false.
- ‘optimize_exponents’ : optimize the hysteresis loss exponents [a,b]. Defaults to false.
.generate_BH_langevin Fit analytical Langevin curve.
[Bnew, Hnew] = generate_BH_langevin(this, B, H) returns a new BH curve, consisting of a single Langevin function fitted to the measured data. The fitted curve will go exactly through the (B(end), H(end).
Note: The resulting curve may fail to well approximate the original curve at lower field strengths. User discretion is advised.
.get_grades List available steel grades.
.parse_data_from_Excel Load Material data from Excel.
[B, H, Ws, fs, Bint, Hint] = parse_data_from_Excel(this, gradename) where B : a 1xNb array of flux density values for tabulated loss data. Assumed to be peak-of-sine values. H : a 1xNb array of field strength values corresponding to B. Can be all-empty, see below. Ws : a Nb x Nf array of loss values, expressed as W/kg. fs : 1 x Nf array of frequencies for the tabuled loss data. Bint : Data for material BH curve interpolation, see next bullet. Hint : Data for material BH curve interpolation. If not found in the Excel, then Bint = B, Hint = H.