Error :Nonscalar arrays of function handles are not allowed; use cell arrays instead.

Hi everyone,

trying to plot the graph like in the picture but got Error using this code:

Es=200000;  %Mpa
Esh=8500; %Mpa
fy=500;  %Mpa
fsu=750; %Mpa
epssh=0.009;
epssu=0.075;
P=Esh*((epssu-epssh)/(fsu-fy));
epsy=fy/Es;

epscmv = linspace(0.1, 10, 500)*1E-3;
for i=1:numel(epscmv);
    epscm = epscmv(i);
    sigmaSteel(i)=@(epscm) Es*epscm .* (epscm<=epsy) + fy .* (epscm>epsy & epscm<=epssh) + fsu+(fy-fsu)*abs((epssu-epscm)./(epssu-epssh))^(1/P) .* (epscm>epssh & epscm<=epssu) + 0 .* (epscm>epsu);
end

plot(epscmv, sigmaSteel)
grid on

Thank you very much

[Stephen23]

"trying to plot the graph like in the picture..."

Then why are you trying to create an array of function handles?

[Shimon Katzman]

Dear Sthephen.

1.Im new to Matlab.

2.Later I will need to use this function, this is just the beggining of the code that i am writing.

[Shimon Katzman]

Es=200000;  %Mpa
Esh=8500; %Mpa
fy=500;  %Mpa
fsu=750; %Mpa
epssh=0.009;
epssu=0.075;
P=Esh*((epssu-epssh)/(fsu-fy));
epsy=fy/Es;

epscmv = linspace(0, 10, 5000)*1E-3;
for i=1:numel(epscmv);
    epscm = epscmv(i);
    sigmaSteel(i)= Es*epscm .* (epscm<=epsy)  +fy .* (epscm>epsy & epscm<=epssh) + fsu+(fy-fsu)*abs((epssu-epscm)./(epssu-epssh))^(1/P) .* (epscm>epssh & epscm<=epssu) + 0 .* (epscm>epssu);
end

plot(epscmv, sigmaSteel)
grid on

Answer 1

[Star Strider]

The loop is not necessary.

This required a few corrections (a typographical error and to vectorise the exponentiation), and now appears to work:

Es=200000;  %Mpa
Esh=8500; %Mpa
fy=500;  %Mpa
fsu=750; %Mpa
epssh=0.009;
epssu=0.075;
P=Esh*((epssu-epssh)/(fsu-fy));
epsy=fy/Es;
epscmv = linspace(0.1, 10, 500)*1E-3;
sigmaSteel=@(epscm) Es*epscm .* (epscm<=epsy) + fy .* (epscm>epsy & epscm<=epssh) + fsu+(fy-fsu)*abs((epssu-epscm)./(epssu-epssh)).^(1/P) .* (epscm>epssh & epscm<=epssu) + 0 .* (epscm>epssu);

figure
plot(epscmv, sigmaSteel(epscmv))
grid on

It does not exactly reproduce the plot in the 1.jpg attachment, although it appears to be reasonably close.