Domain error. To compute complex results, make at least one input complex, e.g. 'power(complex(a),b)'.

Hi, Im getting the following error:

Domain error. To compute complex results, make at least one input complex, e.g. 'power(complex(a),b)'.

function y = fcn(u, Pm)

Cd = 1; %Discharge Coefficient (determined expermentally = actual mass flow/ideal mass flow)) [-]

P0 = 1.01*10^5; %Ambient pressure [Pa]

R = abs(287.1); %Gas constant [J/(kg*K)]

T0 = abs(293.15); %Atmospheric temp [K]

gamma = 1.4; %exponent [-]

q = (Cd*u*P0)/(sqrt(R*T0));

w = (Pm/P0)^(1/gamma);

e = (2*gamma)/(gamma-1);

f = (1.4-1)/1.4;

r = 1-(Pm/P0);

r = r^f;

x6 = q*w*(e*r)^(0.5);

y = x6;

Could anyone please help me? I do not require a complex solution (as far as I know, it should be just a number) and i've tried a lot: complex() and abs() (to try to tell matlab that it is a positive double).

thank you so much.

[Torsten]

In which greater context is this function embedded ? Seems it is called for negative values of Pm which makes w, r x6 and y complex-valued.

[Mars de Bruin]

it is inbedded in a simulink file that is quite big. but even there limiting Pm>=0 doesn't work. Pm is not, cannot and will not be a negative double... furthermore, abs(Pm) does not work.....

Also Im so sorry I forgot to mention the next error code:

Error in power.m (line 83) coder.internal.error('Coder:toolbox:power_domainError'); Error in power.m (line 71) Error in mpower.m (line 44)

Error in 'ICEF/Mass air flow trough throttle valve/Formula 7.6' (line 14)

Line 14 is r = r^f;

(Which works if I replace 'f' with a 1 but doesn't work when I replace it with 0.5 for example. (??)

[Torsten]

Pm is not, cannot and will not be a negative double

Might be. Then Pm becomes greater than P0 during the solution process which makes r negative.

Line 14 is r = r^f;

(Which works if I replace 'f' with a 1 but doesn't work when I replace it with 0.5 for example. (??)

Yes, r seems to become negative. Thus r^f becomes complex for non-integers f.

[Mars de Bruin]

you are an absolute legend. thank you so much.

I just had to look further than line 14 and actually see what could give negative values (like you said).

replacing r = 1-(Pm/P0); with r = abs(1-(Pm/P0)); and that worked!