Inf or NaN whilst trying to run a function

Hi, I'm very new to MATLAB and I am having some trouble. I'm trying to integrate an equation to solve for specific heat capacity. Could somebody please explain what this error is and how to fix it? Thanks.

T_datum_K = 25+273.15;
T_in_K = 199.3 + 273.15;
T_out_K = 359 + 273.15;
%Ar
C1_Ar = 20786;
C2_Ar = 0;
C3_Ar = 0;
C4_Ar = 0;
C5_Ar = 0;
func = @(T) C1_Ar + C2_Ar.*((C3_Ar./T)./(sinh(C3_Ar./T))).^2 + C4_Ar.*((C5_Ar./T)./(cosh(C5_Ar./T))).^2;
Cp_Ar_in = integral(func, T_datum_K, T_in_K);

Warning: Inf or NaN value encountered.

Cp_Ar_out = integral(func, T_datum_K, T_out_K);

Warning: Inf or NaN value encountered.

Answer 1

[VBBV]

T_datum_K = 25+273.15;
T_in_K = 199.3 + 273.15;
T_out_K = 359 + 273.15;
%Ar
C1_Ar = 20786;
C2_Ar = 1;
C3_Ar = 0.1;
C4_Ar = 0.1;

C5_Ar = 0.1;
func = @(T) C1_Ar + C2_Ar.*((C3_Ar./T)./(sinh(C3_Ar./T))).^2 + C4_Ar.*((C5_Ar./T)./(cosh(C5_Ar./T))).^2;
Cp_Ar_in = integral(func, T_datum_K, T_in_K)

Cp_Ar_in = 3.6232e+06

Cp_Ar_out = integral(func, T_datum_K, T_out_K)

Cp_Ar_out = 6.9429e+06

[VBBV]

Give non zero intial values to C2_Ar, C3_Ar, C4_Ar, C5_Ar, variables, since the expression contains fractions whose denominators are all zeros which will yeild NaN values

[Walter Roberson]

T_datum_K = 25+273.15;
T_in_K = 199.3 + 273.15;
T_out_K = 359 + 273.15;
%Ar
C1_Ar = 20786;
syms C2_Ar C3_Ar C4_Ar C5_Ar T real
func = @(T) C1_Ar + C2_Ar.*((C3_Ar./T)./(sinh(C3_Ar./T))).^2 + C4_Ar.*((C5_Ar./T)./(cosh(C5_Ar./T))).^2;
Cp_Ar_in = int(func, T, T_datum_K, T_in_K)

Cp_Ar_in = 

Notice the exp(40*C3_Ar/5963) - 1 in the denominator. As C3_Ar approaches 0, the exp() approaches 1, and subtracting 1 from that approaches 0, leading to a division by 0 if you are operating numerically. But there is also a C3_Ar^2 in the numerator and that can "cancel" the division, if you operate symbolically on the limit

limit(Cp_Ar_in, C3_Ar, 0)

ans = 

As C5_Ar approaches 0, the C5_Ar multiple of the terms approaches 0 so you get divisions by 0 there, with two such divisions with slightly different fractions being subtracted. But the whole thing is being multiplied by C5_Ar^2 so again you get some canceling in the limit:

limit(ans, C5_Ar, 0)

ans = 

Therefore, if you are careful to operate in the limit case, C3_Ar == 0 and C5_Ar == 0 can be accounted for. But not if you operate neively with integral()

[SL]

Got it, thank you so much for your help!