Development of a simplified thermal analysis procedure for insulating glass units
Klam, Jeremy Wayne
MetadataShow full item record
A percentage of insulating glass (IG) units break each year due to thermally induced perimeter stresses. The glass industry has known about this problem for many years and an ASTM standard has recently been developed for the design of monolithic glass plates for thermal stresses induced by solar irradiance. It is believed that a similar standard can be developed for IG units if a proper understanding of IG thermal stresses can be developed. The objective of this research is to improve understandings of IG thermal stresses and compare the IG thermal stresses with those that develop in monolithic glass plates given similar environmental conditions. The major difference between the analysis of a monolithic glass plate and an IG unit is energy exchange due to conduction, natural convection, and long wave radiation through the gas space cavity. In IG units, conduction, natural convection, and long wave radiation combine in a nonlinear fashion that frequently requires iterative numerical analyses for determining thermal stresses in certain situations. To simplify the gas space energy exchange, a numerical propagation procedure was developed. The numerical propagation procedure combines the nonlinear effects of conduction, natural convection, and long wave radiation into a single value. Use of this single value closely approximates the nonlinear nature of the gas space energy exchange and simplifies the numerical analysis. The numerical propagation procedure was then coupled with finite element analysis to estimate thermal stresses for both monolithic glass plates and IG units. It is shown that the maximum thermal stresses that develop in IG units increase linearly with input solar irradiance during the transient phase. It is shown that an initial preload stress develops under equilibrium conditions due to the thermal bridge effects of the spacer. It is shown that IG units develop larger thermal stresses than monolithic glass plates under similar environmental conditions. Finally, it is shown that the use of low-e coatings increase IG thermal stresses and that the location of low-e coating as well as environmental conditions affect which glass plate develops larger thermal stresses.