problem with fzero func

I Have problem with this part of code it returns the following error 'Function values at interval endpoints must be finite and real.' in the linen where M(m)=fzero(func,x_int).I have the latest version of matlab

%% Pressione in uscita 
if (11000>ALT) && (ALT<25000)
    T = -56.46; %C
    p_o = 1000*(22.65*exp(1.73-0.000157*ALT));
elseif ALT>=25000
    T = -131.21 + 0.00299*ALT ;
    p_o = 1000*(2.488*((T+273.1)/216.6)^-11.388);
else 
    T = 15.04 - 0.00649*ALT;
    p_o = 1000*(101.29*((T+273.1)/288.08)^5.256);
end

%%  Inizio il calcolo 
PR = p_o/p_1;
PR2 = (p_o/p_1)^((g-1)/g);
TT = (2*g*R*T_1)/(g-1);
p_t = ((2/(g+1))^(g/(g-1)))*2.068;
v_t = sqrt((2*g*R*T_1)/(g+1));
v_e = sqrt(TT*(1-PR2));

if m_dot==0
    m_dot=FT/v_e;
elseif FT==0
    FT = m_dot/v_e;
else
    fprintf('You can either set desired thrust OR mass flow rate')
end

T_e = T_1*(p_o/p_1)^((g-1)/g);
a_e = sqrt(g*R*T_e);

Me = v_e/a_e;
    
% Metodi delle caratteristiche 
TR = 35; %raggio della gola in cm 
RTOD = 180/pi;
DTOR = pi/180;
P = []; %punti sull' asse delle x 
%% Parametrizzazione funzione di Prantl meyer 
A = sqrt((g+1)/(g-1));
B = (g-1)/(g+1);
v_PM = @(x) A*atan(sqrt(B*(x^2-1))) - atan(sqrt(x^2-1));


%% Calcolo T_max, BREAK UP INTO DIVISIONS
T_max = 0.5*v_PM(Me)*RTOD;
DT = (90-T_max) - fix(90-T_max);
T(1) = DT*DTOR;
n = T_max*2;

for m = 2:T_max*2
    T(m) = (DT + (m-1))*DTOR;
    %Mach from T(i) using T(i) = v_PM (FALSE POSITION)
    x_int = [1 1.01*Me];
    func = @(x) T(m) - v_PM(x);
    M(m) = fzero(func,x_int);
    P(m) = 0 + TR*tan(T(m)); %X-AXIS POINTS
    %RRSLOPES
    RR(m) = -TR/P(m);
    %LR slopes
    LR(m) = tan(T(m)+asin(1/M(m)));
    SL(m) = -RR(m);
end

[Matt J]

For which m in the loop does this occur? Also, have you checked that func(x_int(1)) and func(x_int(2)) are both finite and real, as the error message suggests?

[Mathieu NOE]

hello Francesco

would be glad to help , but we miss a bunch of initialisation of data to run this code

Answer 1

[Shiva Kalyan Diwakaruni]

Hi,

The error 'Function values at interval endpoints must be finite and real.' could be due to because some of input data could be 'Inf' or some values could be negative which may lead to imaginary values.

Hence you could add a breakpoint at the line containing M(m)=fzero(func,x_int) and try out calling func() with different input values , where at some point would lead to the error and you could handle the case for such input values .

hope it helps,

thanks.