The thermal behaviour of glass bulb sprinklers was investigated in a heated wind tunnel under various test conditions. The response parameters considered were the response time index (RTI), the conduction parameter (C) and the change of phase parameter (CHP). Combinations of one, two and three response parameters were used to predict the response times of various wind-tunnel test conditions, and for slow, medium and fast growth rate fires. These calculations indicate that the wind-tunnel results cannot be fully explained in the framework of the thermal response models applied. Tests with a high precondition temperature of the sprinkler resulted in higher RTI values compared with cases with normal precondition temperatures. A plausible explanation for the higher RTI values could be a time delay caused by temperature gradients within the glass bulb. This indicates that the CHP parameter previously explained in terms of work needed to shatter the glass bulb, more likely reflects an affect of thermal time delay. Calculated heat release rates at sprinkler response under a growing fire condition were used to evaluate the thermal response models. For the fast, medium and slow fire growth rates, the heat release rates at sprinkler response were found to be similar, whether based on calculations with two parameters (RTI and C) or with all three parameters (RTI, C, CHP). When using the RTI value only, however, the predicted rate of heat release at sprinkler response decreased significantly.
Read full abstract