Tensile behaviour of soda-lime-silica glass and the significance of load duration – A literature review

Soda-lime-silica glass is a widely used material in society today and its strength over various loading times is of major engineering concern. This paper reviews studies from the published literature, which report on the time-dependent tensile behaviour of soda-lime-silica glass. Furthermore, current normative definitions are reviewed and compared to the literature concerned with time-dependent glass strength. In general, there exists two common applied test methods to characterise glass: (1) the static fatigue test, a constant applied stress over time, and (2) the dynamic fatigue test, a constant applied stress rate. After a thorough search of the available scientific and engineering publications, 92 articles were found to have studied glass using these two test methods. In the tests the typical setups employed to apply load on a glass specimen were the three-point and four-point-bending, and axisymmetric bending configurations. From these tests the load duration and loading rate effects on the strength of glass were investigated. For comparison purposes, all data found were normalised with respect to a reference strength, which for static fatigue is a ‘60-second’ strength, and for dynamic fatigue is a strength corresponding to a stress rate of 2.0 MPa s−1. This means that the time-dependent effect on strength are highlighted and the governing crack properties and size effects are not included. The review supports the general observations that the tensile strength of soda-lime-silica glass is strongly affected by the load duration. The static fatigue tests generally show that strength decreases with increasing load duration. Furthermore, the dynamic fatigue tests show that glass strength increases with loading rate, equivalent to a decreasing load duration. However, a significant lack of data is present at the very short and long loading times, making it difficult to draw a final conclusion at the extreme ends of the load duration and loading rate tests. Additionally, the experiments demonstrate that glass tested in air is less susceptible to static and dynamic fatigue as compared to water immersion, or in air at high relative humidity. However, for the Young's modulus, there are a limited number of studies in the literature and these studies do not highlight a conclusive outcome regarding the sensitivity on loading rates. The fatigue data also support well the load duration dependence given in the various Standards for the design of glass structures, wit