In the previous chapters, we discussed slow crack growth in the presence of a water environment (a relative humidity of 100%). What happens at lower relative humidities? In this case, reaction rate is slowed but, nonetheless, still present. Studies to this effect were carried out by Wiederhorn1 for soda-lime glass. These studies showed that for relative humidity greater than 1%, growth rate was directly and linearly proportional to relative humidity; for example, at 50% relative humidity, growth rate is reduced by 2. This is evidenced in the plot of Fig. 14.1.
Note that, in principle, 100% humidity and liquid water should exhibit similar degrees of corrosion; however, the plot in Fig. 14.1 shows that corrosion levels are not quite the same, as is often observed in other materials as well. This is perhaps due to mobility of the reaction products during the corrosion process. Thus, dynamic fatigue testing in water is the most conservative approach.
Similar studies have been conducted for fused-silica glass. In this case, the rate is drastically reduced at very low relative humidities, as shown in Fig. 14.2. At 50% relative humidity, a typical environment at room-temperature rate is approximatelyone-quarter that of the fully moist condition. Although some studies choose to modify the flaw growth exponent to account for humidity, rate reduction can be used by modification of the crack velocity material constant, a method that is preferred by the author. However, for other materials, it is not clear how humidity affects crack growth rate, as such studies are not common. Indeed, while most ceramics exhibit slow crack growth, the chemical reaction causing stress corrosion [i.e., proton transfer and the like (Chapter 4)] may be different. Thus, it has been common practice to assume a 100% moist environment, unless the environmental parameters are well known.
Online access to SPIE eBooks is limited to subscribing institutions.