Citation

Patton JS. J. Fire Sci. 1992; 10(4): 294-322.

Copyright

(Copyright © 1992, SAGE Publishing)

DOI

unavailable

PMID

unavailable

Abstract

The deleterious effects that combustion products generated during fires can have on normal construction materials have been well-documented. The combined effects of fire, corrosive smoke and particulate have been defined as fire corrosivity. While the effects of fire corrosivity are well-known, little quantitative information is available concerning the mechanisms involved and the degree to which materials, particularly metals, are susceptible. Consequently, a study which was conducted to begin examining the effects of fire and smoke corrosivity on metals is described and the results are presented. Various metal targets were exposed to corrosive smoke and fire particulate produced from polyvinylchloride (PVC) samples burned in a cone calorimeter. The target materials consisted of 304 stainless steel, 1010 carbon steel and 70-30 CuNi alloy. In addition to metal targets, electrical resistance probes were also utilized in the testing to monitor in-situ corrosion rates. The probe materials corresponded to the metal targets so that a comparison could be conducted. After testing, both the metal targets and corrosion probes were sectioned and prepared for analysis using standard metallographic techniques. The targets and probes were analyzed for corrosion products and depth of attack. Results from this testing show that all the metal targets proved highly susceptible to the effects of fire and smoke corrosivity attributed to the burning of PVC samples. These results are presented and compared by corrosion rates. In addition, the performance of the corrosion probes in terms of their ability to produce accurate corrosion measurements was evaluated by comparing their corrosion depth measurements to those of the metal targets. It can be concluded from these observations that the testing of structural metals for their resistance to fire corrosivity must be done over a wide range of combustion environments using a large number of targets in order to generate a statistical basis before any predictions can be made concerning a particular alloy's resistance.