I can't get my color bar to be the range that I want it at.

I have the code show below and I am trying to get my colorbar so it shows ranges between 10^4 to 10^7 because that will make it easier to read. However, when It try to set the range to that using the clim operation it just shows me values from 1*10^6 to 10*10^6 making the graph look bad. can anyone help me?

A = 7.516*10^3; % These are placholders for the experiment done in the paper
B = -0.6738;
C = 5.339;
% A,B, and C are regression constants
% N is the life of the part.
BigA = [];
BigSigma=[];
BigN  = [];
% When R = 0.1
W = logspace(4,7,4)
for N= [10^4,10^5,10^6,10^7]
    K = A.*(N).^B + C;
    sigma_max = (5.392*10^3).*N.^-0.1606 + 190.847;
    R = 0.1;
    A_0 = ((K/((1-R)*sigma_max))^2)*(1/pi); % This is really sqrt(A_0).
    A_1 = logspace(0,4); % This is really sqrt(a) 
    sigma_w = (1-R).*sigma_max.*sqrt(A_0./((A_1/1000000) + A_0));
    % sigma_w is the cyclic stress where defect size sqrt(A) wont propogate.
    % sqrt(A_0) is the critical defect size below which a surface notch will
    % not propogate at the endurance stress.
    % sqrt(A) is the observed surface notch size.
    BigA = [BigA;A_1];
    BigSigma = [BigSigma;sigma_w];
    BigN = [BigN;ones(1,length(A_1))*N];
end
contour(BigA,BigSigma,BigN,1000)
xlabel("Sqrt(A_0) (um)"), ylabel("Stress Range (MPA)")
xlim([10^0,10^4])
ylim([10^1,10^4])
grid on
clim([10^4 10^7])
colorbar
set(gca, 'XScale','log')
set(gca, 'YScale','log')

The graph ends up looking like this

Answer 1

[Walter Roberson]

On the scale of 10^4 to 10^7, 10^4 occupies 1/1000 of the range. Starting at 10^4 is a 1 to 2 pixel difference.

Notice the colorbar labels do not start at 0. That is because 0 is outside of clim.

Why doesn't the labeling start at 10^4 then? Because the labeling is automatic, and the automatic chooser chooses multiples of 10^6 and 10^4 is not a multiple of 10^6

If it really bugs you then

cb = colorbar;
cb.Ticks = [10^4, 10^6:10^6:10^7];

[Sean Marcus]

Sorry for bothing you but I tried using your code and i'm still getting the same graph. My problem with it right now is that you can't really tell the difference between the top like 75% of the graph making it not very readable. This is the bottom of my code that I tried to change with what you gave me.

contour(BigA,BigSigma,BigN,1000)

xlabel("Sqrt(A_0) (um)"), ylabel("Stress Range (MPA)")

xlim([10^0,10^4])

ylim([10^1,10^4])

grid on

cb = colorbar;

cb.Ticks = [10^4, 10^6:10^6:10^7];

set(gca, 'XScale','log')

set(gca, 'YScale','log')

hold on

Answer 2

[Voss]

If you are interested in using a logarithmic color scale, it might be something like this:

A = 7.516*10^3; % These are placholders for the experiment done in the paper
B = -0.6738;
C = 5.339;
% A,B, and C are regression constants
% N is the life of the part.
BigA = [];
BigSigma=[];
BigN  = [];
% When R = 0.1
W = logspace(4,7,4);
for N= [10^4,10^5,10^6,10^7]
    K = A.*(N).^B + C;
    sigma_max = (5.392*10^3).*N.^-0.1606 + 190.847;
    R = 0.1;
    A_0 = ((K/((1-R)*sigma_max))^2)*(1/pi); % This is really sqrt(A_0).
    A_1 = logspace(0,4); % This is really sqrt(a) 
    sigma_w = (1-R).*sigma_max.*sqrt(A_0./((A_1/1000000) + A_0));
    % sigma_w is the cyclic stress where defect size sqrt(A) wont propogate.
    % sqrt(A_0) is the critical defect size below which a surface notch will
    % not propogate at the endurance stress.
    % sqrt(A) is the observed surface notch size.
    BigA = [BigA;A_1];
    BigSigma = [BigSigma;sigma_w];
    BigN = [BigN;ones(1,length(A_1))*N];
end
% contour(BigA,BigSigma,BigN,1000)
contour(BigA,BigSigma,log10(BigN),50)
xlabel("Sqrt(A_0) (um)"), ylabel("Stress Range (MPA)")
xlim([10^0,10^4])
ylim([10^1,10^4])
grid on
cl = [4 7];
clim(cl)
ticks = cl(1):cl(2);
colorbar('Ticks',ticks,'TickLabels',"10^"+ticks);
set(gca, 'XScale','log', 'YScale','log')