No abstract

Based on data from large pool fire experiments and computational fire field model simulations, the size, shape, and character of the oxygen-starved interior in large pool fires is estimated. In the interior of the fire and near the pool surface, low average and low mean deviation temperatures were noted in experimental data for low wind conditions. These trends tend to indicate the presence of a non-combusting region. Using average and mean deviation temperature distributions (supplemented by heat flux measurements) from several data sets, the spatial extent of the vapor dome is estimated for a range of wind conditions. These estimates are compared with fire field model results of temperature and fuel/air concentration distributions. Predicted and measured temperature trends, supported by heat flux data, illustrate the importance of object placement within the fire during system fire survivability testing. The presence of this region also supplements conventional pool fire representations which are based on a continuous flame zone which extends to the pool surface.

A joint experimental and computational study was performed to evaluate the capability of the Sandia Fire Code VULCAN to predict thermocouple response temperature. Thermocouple temperatures recorded by an Inconel-sheathed thermocouple inserted into a near-adiabatic flat flame were predicted by companion VULCAN simulations. The predicted thermocouple temperatures were within 6% of the measured values, with the error primarily attributable to uncertainty in Inconel 600 emissivity and axial conduction losses along the length of the thermocouple assembly. Hence, it is recommended that future thermocouple models (for Inconel-sheathed designs) include a correction for axial conduction. Given the remarkable agreement between experiment and simulation, it is recommended that the analysis be repeated for thermocouples in flames with pollutants such as soot.4

No abstract

This PIRT exercise identifies a number of factors which can influence thermocouple readings made in fires. Identified factors are: (a) the fuel/oxidizer equivalence ratio and its effect on readings, (b) the influence of the state of oxidation and variation with time for the thermocouple sheath, (c) the convection coefficient models and how experimental readings are influenced by thermocouple diameter and yaw angle, (d) response time of a MIMS thermocouple, and (e) thermocouple end effects.