You compute the heat path between a heat sink and a component; the material conductivities are known, the geometry is known — and the result deviates markedly from measurement. What is usually missing is the contact resistance.
Why it exists
No surface is truly flat. When two surfaces are pressed together the real contact area is a small fraction of the apparent area; the gaps between are filled with air (or with an interface material). Heat has to pass through those narrow contact points, and that creates a resistance.
The resistance depends on surface roughness and flatness, contact pressure, material hardness, and the conductivity and thickness of the interface material.
A sense of magnitude
The thermal resistance of a thin metal wall is usually negligible; the resistance of a contact interface in the same assembly can be several times that wall's. Solving the solid in detail while neglecting contact resistance therefore spends precision in the wrong place.
The interface material should be thin
Thermal grease or a pad conducts far better than air but far worse than metal. The ideal thickness is therefore the thinnest that still fills the gaps. Too much grease increases the resistance — a common and counter-intuitive error.
Modelling
Contact resistance is defined directly as a thermal resistance at the interface between two regions; this definition belongs in every conjugate analysis that solves a solid. See the conjugate heat transfer note.
The value comes from the literature or from manufacturer data, and its uncertainty is large. The right approach is to run two plausible extremes and report how much the result moves. If the hot spot temperature depends strongly on that value, this is a design risk and should be stated.
References
- JEDEC Solid State Technology Association. JESD51 — Methodology for the Thermal Measurement of Component Packages (Single Semiconductor Device).
- Shah, R. K. & Sekulić, D. P. (2003). Fundamentals of Heat Exchanger Design. John Wiley & Sons. DOI: 10.1002/9780470172605