diff --git a/+af_tools/+utils/parsefileinputs.m b/+af_tools/+utils/parsefileinputs.m
index 72856f14eb3be26b9dcdfe93727a844b364d67f2..625308c041af6ab659cffde56eb648fbaf31669d 100644
--- a/+af_tools/+utils/parsefileinputs.m
+++ b/+af_tools/+utils/parsefileinputs.m
@@ -36,6 +36,7 @@ function [filenames, filepaths, idxOpts] = parsefileinputs(optList, filetype, va
ext = ['*', filetype];
[filenames, filepaths] = uigetfile(ext, 'Select all dat-files to aggregate', ...
'MultiSelect', 'on');
+ filenames = string(filenames);
filepaths = repmat(string(filepaths), 1, length(filenames));
idxOpts = 1;
diff --git a/+af_tools/+utils/spacedvector.m b/+af_tools/+utils/spacedvector.m
index ecf48e348549431c67a49ebcbd736d0704dbd3de..eed470dbbb476285f56210468c92a3ef3cdc7709 100644
--- a/+af_tools/+utils/spacedvector.m
+++ b/+af_tools/+utils/spacedvector.m
@@ -1,7 +1,14 @@
function vect = spacedvector(spacing, nPoints)
% SPACEDVECTOR Returns a vector based on chosen spacing and number of points
+ %
+ % Spacing:
+ % - linear : points spaced linearily
+ % - halfcosine : more dense at beginning and end of vector
+ % - cosine : more dense at the beginning and more spaced out at the end
+ % - sine : more dense at the end and more spaced out at the beginning
+ %
% -----
- % (c) Copyright 2022 University of Liege
+ % (c) Copyright 2022-2023 University of Liege
% Author: Thomas Lambert <t.lambert@uliege.be>
% ULiege - Aeroelasticity and Experimental Aerodynamics
% Apache 2.0 License
@@ -18,6 +25,9 @@ function vect = spacedvector(spacing, nPoints)
case 'cosine'
beta = linspace(0, pi / 2, nPoints);
vect = 1 - cos(beta);
+ case 'sine'
+ beta = linspace(0, pi / 2, nPoints);
+ vect = sin(beta);
end
end
diff --git a/+af_tools/@Polar/Polar.m b/+af_tools/@Polar/Polar.m
index ecedbd173811c1f3c0eb25ad8271c40baa501e51..196137adcfac9ea8e661ebeb470116429c92dbd8 100644
--- a/+af_tools/@Polar/Polar.m
+++ b/+af_tools/@Polar/Polar.m
@@ -81,6 +81,7 @@ classdef Polar < handle
Stall = struct('aoa', [], 'cl', [], 'cd', []) % Stall point data
Zero = struct('aoa', [], 'cd', []) % Zero-lift point data
Lin = struct('slope', [], 'aoaRange', [], 'clRange', []) % Lift linear range data
+ extrapMethod (1, :) {mustBeMember(extrapMethod, {'none', 'spline', 'viterna'})} = 'spline'
end
% ----------------------------------------------------------------------------------------------
@@ -144,7 +145,7 @@ classdef Polar < handle
%
% See also: af_tools.findcllinearrange.
- [self.Lin.clRange, self.Lin.aoaRange, self.Lin.slope] = ...
+ [self.Lin.aoaRange, self.Lin.clRange, self.Lin.slope] = ...
af_tools.findcllinearrange(self);
end
diff --git a/+af_tools/@Polar/getcoeffs.m b/+af_tools/@Polar/getcoeffs.m
index e4da05fb7894d1174e8787ef74ec1115cea19435..88096a2ade63ffedacae6d12bf667558ebb633c4 100644
--- a/+af_tools/@Polar/getcoeffs.m
+++ b/+af_tools/@Polar/getcoeffs.m
@@ -1,7 +1,12 @@
-function [cl, cd, cm] = getcoeffs(self, aoa, re, polarType)
+function [cl, cd, cm] = getcoeffs(self, aoa, re, extrapMethod)
% getcoeffs Interpolate polars and returns coefficients for given AoA and Reynolds.
% This function returns does mesh interpolation of the polars and returns the three
% coefficients (lift, drag, moment) for the input angles of attacks and reynolds.
+ % It is possible to specify an extrapolation method as well if results outside the original
+ % polars need to be retrieved. In that case, it is strongly suggested to use the 'Viterna'
+ % method as it is the only one that tries to emulate the proper aerodynamic behavior of the
+ % airfoil in the extreme ranges. If no extrapMethod is specified, self.extrapMethod will be
+ % used (default is 'spline').
%
% Note
% The inputs can be supplied as vectors of same size. In that case, the output will be the
@@ -16,9 +21,9 @@ function [cl, cd, cm] = getcoeffs(self, aoa, re, polarType)
% [cl, cd, cm] = Airfoil.getcoeffs(aoa, re)
%
% Input
- % aoa : Angles of attacks, [rad]
- % re : Reynolds numbers, [-]
- % polarType : Type of Polar to use ('original', 'extended')
+ % aoa : Angles of attacks, [rad]
+ % re : Reynolds numbers, [-]
+ % extrapMethod : Extrapolation method ('none', 'spline', 'viterna')
%
% Output
% cl : Lift coefficients, [-]
@@ -45,28 +50,36 @@ function [cl, cd, cm] = getcoeffs(self, aoa, re, polarType)
narginchk(3, 4);
% Defaults and constants
- ALLOWED_TYPES = {'original', 'extended'};
- DEF.POLAR_TYPE = 'extended';
+ ALLOWED_EXTRAP = {'none', 'spline', 'viterna'};
+ EXTENDED_POLARS = {'viterna'};
% Input validation
if nargin == 3
- polarType = DEF.POLAR_TYPE;
+ extrapMethod = self.extrapMethod;
end
validateattributes(aoa, {'double'}, {'row', 'nonempty'}, mfilename(), 'aoa');
validateattributes(re, {'double'}, {'row', 'nonempty', 'size', size(aoa)}, mfilename(), 're');
- polarType = validatestring(polarType, ALLOWED_TYPES, mfilename(), 'polarType');
+ extrapMethod = validatestring(extrapMethod, ALLOWED_EXTRAP, mfilename(), 'extrapMethod');
+
+ if any(strcmpi(extrapMethod, EXTENDED_POLARS))
+ polarType = 'extended';
+ if isempty(self.Ext.aoa)
+ self.extendpolar;
+ end
+ else
+ polarType = 'original';
+ end
% Interpolate coefficients
% -------------------------------------------
% Select correct polar
Pol = selectpolar(self, polarType);
-
if numel(self.reynolds) == 1
- cl = checkandinterp1(Pol, 'cl', aoa);
- cd = checkandinterp1(Pol, 'cd', aoa);
- cm = checkandinterp1(Pol, 'cm', aoa);
+ cl = checkandinterp1(Pol, 'cl', aoa, extrapMethod);
+ cd = checkandinterp1(Pol, 'cd', aoa, extrapMethod);
+ cm = checkandinterp1(Pol, 'cm', aoa, extrapMethod);
else
% Create meshgrid with the polar data points
[X, Y] = meshgrid(Pol.aoa(:, 1), Pol.reynolds);
@@ -75,9 +88,9 @@ function [cl, cd, cm] = getcoeffs(self, aoa, re, polarType)
[Xq, Yq] = meshgrid(aoa', re);
% 2D interpolation
- cl = checkandinterp2(X, Y, Pol.cl, Xq, Yq);
- cd = checkandinterp2(X, Y, Pol.cd, Xq, Yq);
- cm = checkandinterp2(X, Y, Pol.cm, Xq, Yq);
+ cl = checkandinterp2(X, Y, Pol.cl, Xq, Yq, extrapMethod);
+ cd = checkandinterp2(X, Y, Pol.cd, Xq, Yq, extrapMethod);
+ cm = checkandinterp2(X, Y, Pol.cm, Xq, Yq, extrapMethod);
end
end
@@ -99,28 +112,39 @@ function Polar = selectpolar(self, polarType)
end
-function out = checkandinterp1(Pol, field, aoa)
+function out = checkandinterp1(Pol, field, aoa, extrapMethod)
% checkandinterp1 Checks if field is not empty and interpolate (1D)
% If field is empty, return NaN vector of the size of AoA
val = Pol.(field);
if ~isempty(val)
- out = interp1(Pol.aoa, val, aoa, 'spline');
+ if strcmpi(extrapMethod, 'none')
+ out = interp1(Pol.aoa, val, aoa, 'spline', NaN);
+ else
+ out = interp1(Pol.aoa, val, aoa, 'spline');
+ end
else
out = nan(size(aoa));
end
end
-function out = checkandinterp2(X, Y, val, Xq, Yq)
+function out = checkandinterp2(X, Y, val, Xq, Yq, extrapMethod)
% checkandinterp1 Checks if field is not empty and interpolate (1D)
% If field is empty, return NaN vector of the size of AoA
+ warning('off', 'MATLAB:interp2:NaNstrip'); % Warning if original polars contain NaN already
+
if ~isempty(val)
- out = diag(interp2(X, Y, val', Xq, Yq, 'spline'));
+ if strcmpi(extrapMethod, 'none')
+ out = diag(interp2(X, Y, val', Xq, Yq, 'spline', NaN));
+ else
+ out = diag(interp2(X, Y, val', Xq, Yq, 'spline'));
+ end
else
out = nan(length(Xq));
end
+ warning('on', 'MATLAB:interp2:NaNstrip');
end
diff --git a/+af_tools/formatairfoilcoord.m b/+af_tools/formatairfoilcoord.m
index 77d3b44b447d4b6809edcf3247398756ad3c1f97..3b5e00b54092edb76c6df7cc5555d9aa6cd248b6 100644
--- a/+af_tools/formatairfoilcoord.m
+++ b/+af_tools/formatairfoilcoord.m
@@ -139,7 +139,7 @@ function Dat = formatairfoilcoord(varargin)
end
% Output structure
- Dat(i).path = fullpaths;
+ Dat(i).path = fullpaths{i};
Dat(i).file = allFileNames{i};
Dat(i).airfoil = char(cellstr(tmpAirfoil{:}));
Dat(i).format = outputFormat;
diff --git a/+af_tools/nacacamber.m b/+af_tools/nacacamber.m
index 94ec3b81616fa676aa041331bc10bc46326abd0f..ab13a45b06a6654c98d1ae2b06f7bd9311c266a2 100644
--- a/+af_tools/nacacamber.m
+++ b/+af_tools/nacacamber.m
@@ -6,13 +6,13 @@ function [coord, gradY] = nacacamber(varargin)
% -----
%
% Usage:
- % [X, Y] = NACACAMBER prompts the user for the airfoil digits, then calculates the coordinates
- % of the camberline using 100 points spaced with the half-cosine rule.
+ % [coord, gradY] = NACACAMBER prompts the user for the airfoil digits, then calculates the
+ % coordinates of the camberline using 100 points spaced with the half-cosine rule.
%
- % [X, Y] = NACACAMBER(DIGITS) calculates the coordinates of the camberline for the NACA
- % airfoil specified by DIGITS, without further input.
+ % [coord, gradY] = NACACAMBER(DIGITS) calculates the coordinates of the camberline for the
+ % NACA airfoil specified by DIGITS, without further input.
%
- % [X, Y] = NACACAMBER(DIGITS, NPOINTS) calculates the coordinates of the NACA airfoil
+ % [coord, gradY] = NACACAMBER(DIGITS, NPOINTS) calculates the coordinates of the NACA airfoil
% specified by DIGITS, using NPOINTS points spaced with the half-cosine rule.
%
% [...] = NACACAMBER(..., 'spacing', SP) uses the defined spacing rule to determine to
@@ -20,14 +20,14 @@ function [coord, gradY] = nacacamber(varargin)
% 'linear' - linear
% 'halfcosine' - (default) finer at leading and trailing edges
% 'cosine' - finer at leading edge, coarser at the trailing edge
+ % 'sine' - finer at trailing edge, coarser at the leading edge
%
% Inputs:
% digits : Character vector representing the NACA airfoil (ex: '0012', '24120', '24012')
% nPoints : Number of output coordinates
%
% Output:
- % xc : Camberline X coordinates (0 <= X <= 1)
- % yc : Camberline Y coordinates
+ % coord : Camberline coordinates [nPts x 2]
% gradY : Gradient of Y (required to get the whole airfoil coordinates)
%
% Example:
@@ -41,7 +41,7 @@ function [coord, gradY] = nacacamber(varargin)
% <a href="https://gitlab.uliege.be/am-dept/matlab_airfoil_toolbox">Documentation (online)</a>
% -------------------------------------------
- % (c) Copyright 2022 University of Liege
+ % (c) Copyright 2022-2023 University of Liege
% Author: Thomas Lambert <t.lambert@uliege.be>
% ULiege - Aeroelasticity and Experimental Aerodynamics
% MIT License