The purpose of this article is to provide the derivation of the heat transfer matrix of ISO 13786 from the solution of the one-dimensional heat equation (partial differential equation, PDE). The matrix refers to building components as defined in ISO 13786. Inside the same regulation the heat transfer matrix is given without a proof and no derivation is available online.
In the one-dimensional heat equation considered in this case, T(x,t) is the temperature which is a function of the x-coordinate and time.
(1) ![]()
This partial differential equation is solved by the so-called method of separation of variables. The solution is supposed to be made of the product of two other functions of one of the variables x and t separately:
(2) ![]()
The constant inside the T(x, t) function is neglected:
(3) ![]()
The PDE becomes then:
(4) ![]()
Now, the T(x,t) function is replaced by the two separate functions inside the PDE and the product rule is employed:
(5) ![]()
The above equation is rearranged as follows:
(6) ![]()
We assume now that μ is a complex number with the real part equal to 0.
(7) ![]()
Now the above PDE can be split into two separate ordinary differential equations (ODE) which can be solved separately:
(8) 
(9) 
(10) ![]()
(11) 
We introduce the new variable ω (angular velocity), which is equal to:
(12) ![]()
The power of the exponentials inside the X(x) function can be rewritten as:
(13) ![]()
(14) ![]()
δ is defined as:
(15) 
The two differential equations above can be written as:
(16) 
Now, the solution of the one-dimensional heat equation can be written:
(17) ![]()
where
(18) ![]()
The following parameters are introduced:
θ0: maximum temperature at side n of the building component
q0: maximum heat flux density at side n of the building component
θL: maximum temperature at side m of the building component
qL: maximum heat flux density at side m of the building component
The origin of the x-axis is located on side n while the x-value on the opposite side m is equal to L. Therefore, the solution can be written on the two boundaries as:
(19) ![]()
(20) ![]()
(21) ![]()
(22) ![]()
(23) ![]()
(24) ![]()
(25) ![]()
(26) ![]()
The four equations obtained are as follows:
(27) ![Rendered by QuickLaTeX.com \begin{equation*} \begin{cases} \theta_0 = (A' + B') \\ q_0 = \frac{\lambda}{\delta} (1+j) (A'-B') \\ \theta_L = \left [ A' e^{\frac{1}{\delta} (1+j) L} + B' e^{-\frac{1}{\delta} (1+j) L} \right ] \\ q_{L} = -\lambda \left [ A' \frac{1}{\delta} (1+j) e^{\frac{1}{\delta} (1+j) L} - B' \frac{1}{\delta} (1+j) e^{- \frac{1}{\delta} (1+j) L} \right ] \end{cases}\end{equation*}](https://www.fabrizioberloco.com/wp-content/ql-cache/quicklatex.com-ffe0175116eb7adb04e77d16624b078a_l3.png)
Now we rearrange the first two equations so that parameters A’ and B’ are isolated on left-hand side of each of the two equations:
(28) 
Those two parameters are then replaced in the third and fourth equations:
(29) 
The following properties of hyperbolic functions with complex argument are used in order to obtain the final form of the matrix provided inside regulation ISO 13786:
(30) 
After a few calculations and after relacing parameter ξ, defined as:
(31) ![]()
the final form is obtained:
(32) ![]()
where:
(33) ![Rendered by QuickLaTeX.com \begin{equation*} M_{1,1} = M_{2,2} = cosh(\xi)cos(\xi) + j sinh(\xi) sin(\xi) \end{equation*} \begin{equation*} M_{1,2} = \frac{\delta}{2\lambda} \left [ sinh(\xi)cos(\xi) + cosh(\xi)sin(\xi) + j \left ( cosh(\xi) sin (\xi) - sinh(\xi) cos(\xi) \right ) \right ] \end{equation*} \begin{equation*} M_{2,1} = - \frac{\lambda}{\delta} \left [ sinh(\xi)cos(\xi) - cosh(\xi)sin(\xi) + j \left ( cosh(\xi) sin (\xi) + sinh(\xi) cos(\xi) \right ) \right ] \end{equation*}](https://www.fabrizioberloco.com/wp-content/ql-cache/quicklatex.com-b3c19f793a45f7be17a6ad7e4bc40748_l3.png)
Relationship between complex solution and spatial solution
References
Parametri Dinamici UNI EN ISO 13786 – Prof. Marco Manzan
Luca Bonifazi – Comportamento termico dinamico di strutture opache

