Error when using if-else statement in MATLAB function block. How can I fix this error described below?

3 次查看（过去 30 天）
Delika Weragoda
Delika Weragoda 2024-5-5
回答： Sam Chak 2024-5-14
function m = fcn(Q, T_sat, T_w, sigma, rho_v, h_fg, M, P_v, R, g, beta, x, nu_novec, alpha_novec, k_m, d_pore, psi, k_f, rporous_outer, height_porous)
q_w = Q/(2*pi*rporous_outer*height_porous); %heat input converted to heat flux
if T_w>=T_sat
m = (((2*sigma)/(2-sigma))*((rho_v*(h_fg)^2)/T_sat)*((M/(2*pi*R*T_sat))^(0.5))*(1- ((P_v)/(2*rho_v*h_fg))));
else
m = (k_f*(0.5*(((g*beta*q_w*x^4)/(nu_novec*alpha_novec*k_m))*(((d_pore^2)*psi^3/(180*(1-psi^2)))/x^2))^(0.4)))/(x);
%the power of 0.4 is what triggers the complex number
end
When I run this code, I get the following error, which is triggered by the line ' m = (k_f*(0.5*(((g*beta*q_w*x^4)/(nu_novec*alpha_novec*k_m))*(((d_pore^2)*psi^3/(180*(1-psi^2)))/x^2))^(0.4)))/(x);
I've checked if any of the values result in a negative number, but none of the terms are negative, nor their result. I've run this separately, and this the same line works fine and I get a positive results without an complex number or any error? Why would this occur, does this error have to do anything with the if-else statement used? Any thoughts would be much appreciated.
I think this issue is triggered by the use of ^0.4, which cannot be changed. Why would an error/complex result occur in simulink and not in MATLAB?
Many thanks
  1 个评论
Sam Chak
Sam Chak 2024-5-6
Ran in:
Hi @Delika Weragoda
Could you please provide us with the values of the parameters? This will allow us to test the function 'fcn()' and compute 'm'.
m = fcn(1, 10, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1)
m = Inf
function m = fcn(Q, T_sat, T_w, sigma, rho_v, h_fg, M, P_v, R, g, beta, x, nu_novec, alpha_novec, k_m, d_pore, psi, k_f, rporous_outer, height_porous)
q_w = Q/(2*pi*rporous_outer*height_porous); %heat input converted to heat flux
if T_w>=T_sat
m = (((2*sigma)/(2-sigma))*((rho_v*(h_fg)^2)/T_sat)*((M/(2*pi*R*T_sat))^(0.5))*(1- ((P_v)/(2*rho_v*h_fg))));
else
m = (k_f*(0.5*(((g*beta*q_w*x^4)/(nu_novec*alpha_novec*k_m))*(((d_pore^2)*psi^3/(180*(1-psi^2)))/x^2))^(0.4)))/(x);
end
end

请先登录，再进行评论。

请先登录，再回答此问题。

回答（4 个）

Sulaymon Eshkabilov
Have you tried to change ^0.4 with: ((...)^2)^(1/5) or ((...)^2)^(0.2)
Walter Roberson
Walter Roberson 2024-5-6
m = (k_f*(0.5*(((g*beta*q_w*x^4)/(nu_novec*alpha_novec*k_m))*(((d_pore^2)*psi^3/(180*(1-psi^2)))/x^2))^(0.4)))/(x);
1 2 345 4 5 43 456 5 6 7 654 32 3 210 1 0
Simulink is calculating the expression
(0.5*(((g*beta*q_w*x^4)/(nu_novec*alpha_novec*k_m))*(((d_pore^2)*psi^3/(180*(1-psi^2)))/x^2))
as being negative. When negative is raised to the fractional power 0.4 it results in a complex value.
My guess is that psi is outside the range -1 to +1
If you want a complex result, then code
m = (k_f*complex(0.5*(((g*beta*q_w*x^4)/(nu_novec*alpha_novec*k_m))*(((d_pore^2)*psi^3/(180*(1-psi^2)))/x^2))^(0.4)))/(x);
If you do not want a complex result, you will need to protect against it, such as
if abs(psi) >= 1
m = nan;
elseif T_w>=T_sat
m = (((2*sigma)/(2-sigma))*((rho_v*(h_fg)^2)/T_sat)*((M/(2*pi*R*T_sat))^(0.5))*(1- ((P_v)/(2*rho_v*h_fg))));
else
m = (k_f*(0.5*(((g*beta*q_w*x^4)/(nu_novec*alpha_novec*k_m))*(((d_pore^2)*psi^3/(180*(1-psi^2)))/x^2))^(0.4)))/(x);
end
  9 个评论
显示 7更早的评论
Delika Weragoda
Delika Weragoda 2024-5-9
编辑：Delika Weragoda 2024-5-9
@Sam Chak So, apparently, I've made a mistake in the parenthesis, that is why the model supposedly worked, but I managed to find the issue. I believe this error in complex results in triggered from the heat sensors' input that I take to calculate the value for m, which is dynamically changing.
The variable m, is a function of Q, which is the input taken from the heat flow rate sensor. This is what is causing the error, as this signal will be a power of 0.4. Is there a way to overcome this?
function m = fcn(k_f, g, beta, x, nu_novec, alpha_novec, k_m, d_pore, psi, Q, rporous_outer, height_porous)
q_w = Q/(2*pi*rporous_outer*height_porous); %(heat flux)
m = (k_f*(0.5*(((g*beta*q_w*x^4)/(nu_novec*alpha_novec*k_m))*(((d_pore^2)*psi^3/(180*(1-psi^2)))/x^2))^(0.4)))/(x);
Please see the model below for your reference
Sam Chak
Sam Chak 2024-5-9
Hi @Delika Weragoda
m = (k_f*(0.5*(((g*beta*q_w*x^4)/(nu_novec*alpha_novec*k_m))*(((d_pore^2)*psi^3/(180*(1-psi^2)))/x^2))^(0.4)))/(x);
I tend to divide long equations and conquer them later. This approach helps 100% prevent human mistakes such as mismatched parentheses. I suggest you rewrite the equation 'm' in multiple terms and then return here for a review.
%% parameters
a = ...;
b = ...;
c = ...;
d = ...;
%% terms
term1 = ...;
term2 = ...;
term3 = ...;
%% equation
m = term1 + (term2/term3)^0.4 ...;

请先登录，再进行评论。

Delika Weragoda
Delika Weragoda 2024-5-14
@Sam Chak
I found the issue. There is a negative heat flow generated due to the thermal mass that I am using in the model. Why would something like this occur? I've tried debugging the model by monitoring the heat flow at different sections of the model. Please see the results below.
Model:
heat transfer subsystem
Heat flow results
The four heat flow sensors shown in the figure above are used to monitor the heat flow at the four locations. The negative values at the first two time steps is what triggers the error.
I would really appreciate if you can give any feedback on how to rectify this please.
Many thanks
Sam Chak
Sam Chak 2024-5-14
Hi @Delika Weragoda
Try using the Abs block to return the absolute value of the input.

类别

在 Help Center 和 File Exchange 中查找有关 Thermal Liquid Library 的更多信息

标签

产品


版本

R2022b

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by