Theory and Method
Steady-state conductivity measurement rests on a simple relationship. Drive a known quantity of heat through a specimen of known thickness and area, wait until temperatures stop changing, measure the temperature difference across the specimen, and thermal conductivity follows directly from those four quantities. Nothing about the calculation is difficult. The difficulty lies entirely in knowing how much of the heat you supplied actually crossed the specimen rather than escaping somewhere else.
That question is what the guard exists to answer. A heater in the middle of a plate sends heat in every direction, and the portion leaking sideways never passes through the measured area at all. Counting it anyway inflates the calculated conductivity, and the error grows worse as the material gets more insulating, precisely where accuracy matters most. The guarded hot plate surrounds the central metered heater with a separately powered guard ring, held at the same temperature as the metered section. With no temperature difference between the two, there is no driving force for lateral flow, and the heat measured at the central heater has nowhere to go except straight through the specimen. Establishing that condition physically, rather than correcting for it afterward, is why the method serves as the absolute reference and why heat flow meter instruments are calibrated against it.
Specimen thickness sits in a trade-off that explains the preferred value. A thin specimen develops only a small temperature difference, so contact resistance at the plate faces becomes significant relative to the resistance being measured. A thick one develops a large difference but loses proportionally more heat at its exposed edges, and takes considerably longer to reach steady state. Twenty-five millimeters balances those effects for most insulation materials, which is why the apparatus lists it as preferred within its 10 to 45 mm range. Ambient stability matters for a related reason: guard balance is maintained by a control loop working against the surrounding air, so a laboratory whose temperature drifts through the day makes that loop work harder and lengthens the time to a stable reading.
