Temperature Measurement in Fire Test Furnaces VYTENIS BABRAUSKAS* and ROBERT BRADY WILLIAMSON University of California, Berkeley A study of errors associated with temperature measurement in fire endurance test furnaces has shown that conventionally used thermocouples are subject to large time constant errors in the first 20 minutes of a standard test. T HE USUAL protocol for reporting results of standard fire endurance tests prescribes that endurance be specified to the nearest minute. Since a fire test may commonly last 1 or 2 hrs, approximately a 1 percent precision is implied. The accuracy of the needed furnace temperature measurements is widely known to be much less. In recent years several studies ~..~.~. 4 have been made in attempts to find systematic measurement errors and to compare results for different furnaces. The findings generally proved inconclusive. To shed some new light on the issue, a series of tests was run using the University of California wall furnace. With the use of only simple instrumentation useful and revealing measurements could be made, which when combined with simplified theoretical analysis provide a basis for improved furnace operations. TYPES OF THERMOCOUPLES The furnace thermocouples to be used in standard fire endurance tests (ASTM Standard El19) ~ are prescribed as follows: "thermocouples . . . enclosed in sealed porcelain tubes ~ inch (19 ram) in outside diameter and ~ inch (3 mm) in wall thickness, or, as an alternative in the case of base metal thermocouples, enclosed in sealed, standard-weight ½-inch (13-mm), black wrought steel or black wrought iron pipe .... Other types of pro- tecting tubes or pyrometers may be used that . . . give the same indi- cations .... " This prescription for furnace thermocouples has been es- sentially unchanged since the 1926 edition of the standard. A recent *Now at the Center for Fire Research, National Bureau of Standards, Gaithers- burg, Maryland. 226