Kinetics of Moisture Expansion in Fired Clay Ceramics: A $(\mathrm{\text{Time}}{)}^{1/4}$ Law

We report the results of experiments to estimate the moisture expansion of several modern and ancient clay brick ceramics from the time of manufacture. From these data we propose a new expansion law, in which the expansive strain increases as $(\mathrm{\text{age}}{)}^{1/4}$ approximately. Such time dependence is consistent with a mechanism in which expansion arises from a diffusion-controlled rehydration reaction on a linear or low-dimension structure. Our results provide new guidance for the engineering design of masonry and suggest a possible new method for archaeological dating of ceramics.

Figure 1

The expansion of a freshly fired brick aging under ambient conditions (approximately $20°\mathrm{C}$, $50\%$ relative humidity) in the laboratory. The solid line is the power-law fit $ϵ=9.9\times {10}^{-5}(t/\mathrm{d}{)}^{0.24}$. (a) Linear axes. (b) Log axes; logarithmic (dashed line) and polynomial (dotted line) regression fits are shown for comparison.

Figure 2

The expansion of freshly fired bricks subject to different treatments. $A$: steaming at $100°\mathrm{C}$ for 6 h; $B$: aging in ambient laboratory conditions for 104 d; $C$: aging under water at $20°\mathrm{C}$ for 104 d; $D$: aging under water at $50°\mathrm{C}$ for 104 d; $E$: aging in $100\%$ relative humidity at $20°\mathrm{C}$ for 104 d; $F$: aging at $105°\mathrm{C}$ in a ventilated oven for 104 d; $G$: alternate vacuum saturation with water and oven drying, total of 14 separate saturations over 104 d; $H$: steam autoclaving at 1.15 MPa, $186°\mathrm{C}$ for 4 h.

Figure 3

Strain $ϵ$ produced by steam autoclaving and heating four clay bricks of different ages. Zero strain corresponds to the initial condition of each brick. The solid bars represent strains resulting from steam autoclaving as in Fig. 2. The open bars represent strains resulting from heating at $450°\mathrm{C}$ for 2 h. The cycle of treatments in each case was autoclaving, heating, autoclaving, reading from the left for each sample. 1: Kiln fresh brick; 2: 20 yr old brick; 3: 19th century brick (120 yr old); 4: Roman brick (1900 yr old).

Figure 4

The total expansive strain $ϵ$ (solid squares) and fractional mass gain $\Delta m/m_0$ (open squares) from the time of original manufacture for bricks tested. The power-law regression equations are $ϵ=9.87\times {10}^{-5}(t/d{)}^{0.24}$ and $\Delta m/m_0=5.7\times {10}^{-4}(t/d{)}^{0.22}$ where $t$ is the time. The dashed curve is the best fit logarithmic function. The dotted horizontal line shows the strain produced by autoclaving a freshly fired brick as in Fig. 2.