Tempered fire resistance glass is designed to be able to tolerate thermal stress on fire by giving the surface compression stress that is higher than general tempered glass. However the surface compression stress is able to be measured by optical method along Japanese Industrial Standard, but edge strength is not able to be checked without breaking glass panes. On setting glass strength, many glass panes are broken and the glass breaking stress obtained by statistically processing becomes allowable glass stress. In the previous report, it was obtained as basic knowledge that there was a strong correlation between temperature difference ΔT and surface compression stress. However, even if glass had high surface compression stress, if the edge strength of the glass is much lower, it is quite possible that the tempered fire resistance glass is broken in an early stage of fire. Therefore, tempered fire resistance glass has to be evaluated as the total strength about both the surface compression stress and glass edge strength. On this study, the glass edge strength was checked by glass breakage four-point bending tests, tempered fire resistance glass in the same production lot was used in fire tests. The surface compression stress was measured the glass that is evaluated by fire tests. To grasp the correlation between temperature difference and the total glass strength of tempered fire resistance glass, the temperature data of glass was obtained by fire tests.<br> The four-point bending test condition is shown below.<br> ·Glass size is W1,000mm×H100mm, glass thickness is 3mm～12mm , and the number of glass panes is 50 sheets.<br> ·The glass was put on the supported point of the 900mm span, and was loaded from the upper end of glass on loading point of the 300mm span as four-point bending test method.<br> The fire test condition is shown below.<br> ·Fire tests were performed for Single glass panes and Low-E Insulated Glass Units.<br> ·Concerning Low-E -IGU, tempered fire resistance glass was located on fire side(Figure7, 8).<br> ·Fire tests were performed along the ISO834 Heat Curve for each specimen.<br> ·The temperature of the glass was measured by thermo couples with a data logger.(Figure9).<br><br> Findings are as follows.<br> 1) Glass edge strength was checked by glass breakage four-point bending tests. Figure 2 shows that the total strength of tempered fire resistance glass could be expressed as surface compression stress and edge strength. On this study, the edge strength was treated as lower-limit value of the lognormal distribution 3σ_n-1.<br> 2) Figure10, 13, Table10, 11 show that the upper peak of temperature difference arrives at 400K degree for Low-E IGU, at 300K degree for Single glass without depending on the glass thickness. That is to say, the strength of tempered fire resistance glass for Low-E-IGU needs to be much stronger than Single glass.<br> 3) Figure15, 16 show that on this fire tests, fire tempered glass or Low-E glass was broken by touching each other on 10mm and 12mm thickness of tempered fire resistance glass. That is the reason why the glass that has the thickness is easy to increase temperature difference of front and back sides, and the touch is caused before glass temperature is reached at softened temperature.<br> 4) The lowest total strength for tempered fire resistance glass not to be broken could be evaluated by ΔT without depending on the glass thickness from a number of fire tests. (Formula 4: σ₀>0.6178·ΔT_max+5.161) The thermal stress of the glass was expressed in consideration of a parameter to affect the difference of temperature (Formula 3: σ₌0.608·K·ΔT)