Page 29 - Handbook of Thermal Analysis of Construction Materials

Section 2.0 - Classical Techniques 13

Therefore, a plot of the logarithm of the heating rate versus
reciprocal temperature gives a straight line with a slope equal to -E /R and
a
an intercept equal to β, assuming a first order reaction.
Thermoanalytical methods, such as TG, where degradation of a
material can be measured under conditions that accelerate its rate and the
resulting parameters extrapolate to predict a service lifetime could have
great commercial importance [41] in the construction industry. They could be
used not only for planning economic replacement before catastrophic
failure occurs or avoiding premature replacement, but also for developing
specifications for quality assurance and control tests and formulations.
If the E /R value and the rate, at a given temperature, are known,
a
rates at any other temperature may be obtained and failure predictions can
be made. A typical computer routine calculates the activation energies
using Eq. (8). Once a failure criterion is selected (e.g., 5% weight loss), the
logarithms of the times to reach failure are calculated at various tempera-
tures. These plots are used to predict times to failure at service temperatures
that are outside the range of experimental temperature measurements. Such
predictions depend on the reaction mechanism remaining unchanged over
the entire range of extrapolation. However, these routines are frequently
questionable. [40] Weight loss usually reaches a measurable rate only when
temperatures are high enough for considerable molecular movement to
occur. Therefore, extrapolation of kinetic equations parameters obtained at
these temperatures, through temperature ranges where phase and large
viscosity changes take place down to service temperature where material
diffusion limits the kinetics, results in false predictions. [41]
According to Flynn, [40] differential scanning calorimetry and ther-
momechanical analysis techniques may give more reliable correlation
between natural and accelerated aging than TG. Therefore, the accelerated
aging experiments should take into account factors such as determining the
property whose deterioration is responsible for failure; chemical groups or
morphological characteristics susceptible to attack; attacking agents; and
factors accelerating the deterioration through intensification, sensitivity of
the technique, and reliability of the measurements as well as relevance of
the extrapolation. Also, it is important to validate the procedure used by
comparing the predictions from the proposed method with those from
methods that measure another physical property, data from actual service,
or from long-term aging experiments.