Sénarmont's plate
Does heat care which way it travels? In 1847 Henri de Sénarmont coated thin crystal plates in wax and touched a hot wire to the centre. Where the plate passed the wax's melting point the wax ran clear, so the edge of the melt traced an isotherm. In glass the melt was a circle. In quartz cut along its c axis it was an ellipse.
1. Predict
A quartz plate cut along its c axis: what shape will the melt take?
2. Try
Use the bench controls: choose a plate, set the stylus temperature, turn the stage, lift the stylus or lay fresh wax. The live readings are below.
Why
Quartz carries heat faster along its c axis, 10.4 W/m·K, than across it, 6.21 W/m·K. In the same time the 62 °C isotherm reaches further along c, so the melt is an ellipse with its long axis on c, in the ratio √(10.4 ÷ 6.21) ≈ 1.29. Sénarmont measured 1.31. Fourier's law becomes q = −K∇T with K a tensor: diagonal in the crystal's own axes, with off-diagonal terms in any frame turned against them.
3. Challenge
On the Y-cut plate, turn the stage until the melt's long axis lines up with the arrow on the base, the bench's x axis. Hint: does the ellipse belong to the room or to the crystal?
The model
A 0.3 mm plate, 40 mm across with a reference flat, on three standoffs in still air. The stylus holds a 0.6 mm spot at its set temperature. Finite volumes on a 121 × 121 grid in the crystal's own frame; each face loses heat to air at h = 12 W/m²·K; ambient 22 °C; beeswax melts at 62 °C. The wax film's own heat capacity and latent heat are left out. The same model runs the lab's two-dimensional Sénarmont figure.
Made by Chandan Mettu with Claude. The book-page style follows “Every Joule” by Techartist.
Sources: quartz conductivity, 10.4 along c and 6.21 along a at 300 K, TPRC recommended values (Touloukian et al.), as cited in arXiv:1612.07049. Sénarmont's 1847 plates and his 1.31 ratio: "Thermal conductivity through the nineteenth century", arXiv:1005.2119.