Modeling of ternary ion exchange and stress evolution in lithium-containing glass

A computational model of ternary ion exchange (IOX) for strengthening glass is proposed to predict the cation concentration and residual stress distributions in glass after ternary IOX. The comparison between theoretical predictions and experimental results indicated the validates the model. Additionally, it provides a method to determine ion diffusivity and volume expansion through ternary IOX experiments. Simulations of K–Na–Li ternary IOX were conducted using the parameters calibrated based on experimental results from a thick silicate glass. Then the process parameters were changed to clarify their influences. Key findings reveal that for thick glass (where lateral expansion is negligible), the optimum ratio of K⁺ and Na⁺ concentrations in a molten salt is 2:1. We further consolidate the effects of process parameters by training a neural network (NN) and demonstrate that the NN can be a surrogate model to replace the time-consuming simulations, which could be more adaptable by the glass industry.