= Measurement principle
A \b[measurement principle] is the phenomenon on which a measurement is based. The VIM gives the <thermoelectric effect> used to measure <temperature> as a physical example.<International vocabulary of metrology by JCGM>[VIM3entry2.4, definition and example1]{ref} Naming that phenomenon identifies the physical connection that makes an observable signal informative about the <measurand>.
A <thermocouple> uses the <Seebeck effect> to produce an <electric potential difference> associated with temperatures along a circuit of dissimilar conductors. In a chosen approximation with constant coefficient $S$, the <measurement model> can be written
$$
V=S(T_h-T_r),\qquad T_h=T_r+V/S,
$$
where $T_r$ is the reference-junction temperature. For an original ideal example, choose $S=40\,\mathrm{\mu V\,K^{-1}}$, $T_r=20\,{}^{\circ}\mathrm C$ and $V=2.00\,\mathrm{mV}$. The inferred temperature difference is $50\,\mathrm K$, so $T_h=70\,{}^{\circ}\mathrm C$. These stipulated values illustrate the model, not the calibration of a particular thermocouple type.
The reference condition matters physically: the same signal at $T_r=25\,{}^{\circ}\mathrm C$ would imply $T_h=75\,{}^{\circ}\mathrm C$ in this model. A readout that neglects the reference temperature can give a precise-looking indication without measuring the intended temperature correctly. The <measurement principle> explains why the signal arises; it does not by itself supply all corrections or the <measurement uncertainty>.
The <measurement method> describes how operations are organized around the principle, for example comparing an output with a reference. The <measurement procedure> specifies the actual conductors, junction arrangement, temperature references, readout settings, settling interval and calculations. Different measuring systems can share a <measurement principle> while implementing different methods and procedures.
Back to article page
