Contour plot using Griddata for a curved geometry

Hi,

I am trying to plot my velocity on a 2d plane, taken from a slice of my simulation results, I've already extracted the data on the slice, (v_r(y,z)) however when I plot it via griddata the interpolations causes a strightline, where there should be a curve, I thought about making the points lying outside the curve values NaN using inpolygon although have had no success as of yet. Could anyone advise on how to overcome this problem ?

This is the code i've used to plot the slice. Thank you for any help

clc

clear all

WMLES = importdata('30DegreePSyz.xlsx');

variables_names = WMLES.colheaders;

WMLES_data = array2table(WMLES.data,'VariableNames',variables_names);

z = WMLES_data.z;

y = WMLES_data.y;

ps = WMLES_data.ps;

spacing = 1000;

zi = linspace(min(z), max(z), spacing);

yi = linspace(min(y), max(y), spacing);

[zm,ym] = meshgrid(zi, yi);

WFm = griddata(z,y,ps, zm, ym,'cubic');

figure (1)

pcolor(zm, ym, WFm)

shading interp;

colormap('jet');

% xlim([-0.01 0.14])

% ylim([0.18 0.26])

caxis([0 1])

colorbar

colorbar('Ticks',[0 1])

Answer 1

[Ayush]

Hi Shaun,

It is my understanding that you are unable to plot a curve in your contour plot using the “griddata” function. It is currently resulting to an interpolation that causes a straight line.

Here are some possible workarounds to resolve this issue:

1. Use a higher resolution for the grid, by increasing the value of “spacing” to get a finer grid which would facilitate more points and accuracy. You can refer to the below documentation to know more about the “linspace” function:

Generate linearly spaced vector - MATLAB linspace - MathWorks India

2. Use a different interpolation method. Try using “linear” or “natural” as it may provide different results. Experiment with them to choose which one works best for your data. You can refer to the below documentation to know more about the “method” parameter:

Interpolate 2-D or 3-D scattered data - MATLAB griddata - MathWorks India

3. To exclude points outside the curve from interpolation by setting their values “NaN”, you need to do it before the “griddata” function using the “inpolygon” function. You can refer to the documentation to know more about the “inpolygon” function:

Points located inside or on edge of polygonal region - MATLAB inpolygon - MathWorks India

Here is the code snippet to perform the same:

% Define the polygon that encloses the desired curve 
x_polygon = [x1, x2, x3, ...]; % x-coordinates of the polygon vertices 
    y_polygon = [y1, y2, y3, ...]; % y-coordinates of the polygon vertices 
    % Determine points lying outside the curve 
    outside_curve = ~inpolygon(z, y, x_polygon, y_polygon); 
    % Set values outside the curve to NaN 
    ps(outside_curve) = NaN; 
    % Perform interpolation 
    WFm = griddata(z, y, ps, zm, ym, 'cubic'); 

Hope it helps,

Regards,

Ayush Misra