The strengthening time was 1, 2 and 3 h, with all glass being strengthened under the same temperature. The glass was then cooled, which resulted in the surface of the glass becoming compressed, while the center core remained in a state of tension. The strength of the glass increased after chemical strengthening compared to the non-strengthened glass, but it was decreased by increasing the chemical strengthening times of 1, 2 and 3 h from 668.3 ± 88.43, 645.5 ± 82.37 and 623.9 ± 80.49 MPa, respectively, as shown in
Figure 7. Significantly, the bending strength was observed to form compressive stresses at the surface when the chemical strengthening mechanism of a Na
+ ion was substituted with a K
+ ion. However, the longer the chemical strengthening time conditions, the more likely it was that the bending strength decreased by about 20 MPa (see
Table 5). These results are related to the loss of surface compressive stress, which may be attributed to the difference in temperature when cooled to ambient temperature during ion exchange [
19,
24,
25,
26]. In general, high compressive stress at the surface and low tension at the center can improve the strength of chemically strengthened glass. Moreover, Kerper and Scuderi investigated the mechanical properties of chemically strengthened glasses at high temperatures [
27]. It was found that the elastic modulus decreased with increasing temperature, with a sharp inflection slightly above room temperature. The ion exchange of K
+ ions by chemical strengthening layer is about 200 µm. In the BT method, the mechanical strength value is the compressive stress of the surface layer, resulting in a difference value depending on the chemical strengthening treatment conditions. In the UPET method, there was a limit in measuring the difference in acoustic speed for both sides of the 200 µm, so there was some difference in the error range, as in the BT method. In other words, it was confirmed that both the BT and UPET measurements allowed for the derivation of measurement errors. Moreover, the IET method was able to measure minute variations in natural frequency of the glass.
Table 6 compares the dynamic elastic modulus values using an IET method in the same conditions of chemical strengthening, showing that it was about 3 to 5 GPa lower than that of the non-strengthened glass. In addition, the longer the strengthening time, the lower the elastic modulus value. This is the physical meaning of elastic modulus, and it is known that Equation (12) is formed in relation to the atomic force [
28]:
where:
F is the atomic force; Δ
r (
r −
r0) is the movement between atoms;
r0 is the equilibrium interatomic spacing.
In other words, based on the spring model according to the ion exchange of K
+ ions by chemical strengthening, the difference in elastic waves can be seen by the interaction of atomic forces as shown in
Figure 7, which indicates that the K
+ (0.231 nm, 63 °C) ions have a larger atomic radius and a lower melting point than Na
+ (0.186 nm, 98 °C). It was previously reported that the exchange of K
+ ions by strengthening has two competing processes: one is the generation of stress from “stuffing” large foreign ions into small host ion sites in the glass, and the other is the relaxation of stress using viscous flow [
29]. The exchange of large ions such as K
+ for comparatively smaller ions such as Na
+ in glass at temperatures below the material’s strain point leaves the surface of the glass in a state of high compression. Because glass products usually break due to excessively applied tension acting on a surface flaw, the introduction of a high surface compression strengthens the glass [
30]. Furthermore, the chemically strengthened glass has a compressive stress layer formed on the surface by ion exchange treatment (see
Figure 8a). Distribution of K
+ and Na
+ ions after heat treatment at 420 °C for 3 h is shown in
Figure 8b,c. It is obvious that the flow depth of K
+ and Na
+ ions was less than 10 µm. The surface of the glass is ion–exchanged to form a surface layer in which compressive stress remains. Specifically, a metal ion with a small ionic radius which exists near the glass plate’s surface is converted to an ion with a larger ionic radius by ion exchange at a temperature lower than the glass transition point, as illustrated in
Figure 9. The exchange depth, the concentration of ions that can be replaced with large ions, and any stress or structural relaxation that may occur determine the magnitudes of the residual stress. Thereby, compressive stress remains on the surface of the glass, meaning the strength of the glass can be increased. Unlike tempered glass, the surface layer is thin with a high stress at the surface. The tensile stress inside the glass is low and almost constant with thickness. The maximum compressive stress at the surface can be calculated by Equation (13) below:
where
E is the elastic modulus;
ν is the Poisson’s ratio; Δ
V/
V is the relative volume increase due to an exchange of smaller ions for ions with larger sizes. Hence, understanding the effect of compressive residual stress on glass fractures is not only of long-standing fundamental interest, but also important for controlling the mechanical properties of glass products.