I am having a problem determining the thermal conductivity of a layered sample. The sample is made of two materials, one whose thermal conductivity is known (λ-1) and one whose thermal conductivity is unknown (λ-2).

The thermal conductivity of the entire sample is known by means of testing. I am trying to calculate λ-2, and my first thought was to use the one-dimensional steady state diffusion equation without heat generation:

$$ \frac{d}{dx}(\lambda \frac{dT}{dx}) = 0$$

Next, I will enter the known and unknown values:

$$ \frac{d}{dx_{total}}(\lambda_{total} \frac{dT}{dx_{total}}) = \frac{d}{dx_{1}}(\lambda_{1} \frac{dT}{dx_{1}}) + \frac{d}{dx_{2}}(\lambda_{2} \frac{dT}{dx_{2}}) $$

Is this approach correct? I do not know dT (perhaps can find in machine's documentation), I'm not sure if the addition on the right side of the equation is legit; am I headed in the wrong direction?

The tested layered structure

A simplified model for analysis


1 Answer 1


The thermal resistance for a slab in steady-state heat transfer is $R_t = w/(kA)$ (K/W), where $w$ is thickness (m), $k$ thermal conductivity (W/m K), and $A$ cross-sectional area (m$^2$). For serial thermal resistors as your case

$$R_t = \sum R_{t,j} $$

For $N_j$ layers at each $k_j$ thermal conductivity and $w_j$ width, this gives

$$R_t = \frac{N_1\ w_1 + N_2\ w_2}{k_t} = \frac{N_1\ w_1}{k_1}+ \frac{N_2\ w_2}{k_2}$$

and with $w_1 = w_2$ and $N_1 = N_2 = N$ we find

$$R_t = \frac{2}{k_t} = \frac{1}{k_1} + \frac{1}{k_2}$$

  • $\begingroup$ Thank you @Jeffrey. This is related to a side project we are working on in the office, your post is a huge help. $\endgroup$
    – codenoob
    Apr 13, 2022 at 14:00
  • $\begingroup$ @codenoob FWIW, as a new contributor, the next acceptable step is to (up/down)vote or to mark as complete. $\endgroup$ Apr 14, 2022 at 0:29
  • $\begingroup$ Got it - thanks $\endgroup$
    – codenoob
    Jun 10, 2022 at 15:22

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.