Summary Reconstructive fire testing is an important tool used by fire investigators to determine the cause, origin, and progression of a particular fire. Accurate reconstruction of the fire requires the laboratory structure to be outfitted with materials that, in terms of contribution to fire growth, perform similarly to the original materials found at the fire scene. Therefore, a procedure was developed to enable fire investigators to select these replacement materials on the basis of a quantitative assessment of their relative fire performance. This procedure consists of gram‐scale and/or milligram‐scale standard testing accompanied by inverse numerical modeling of these tests, which is used to obtain relevant material properties. A numerical model composed of a detailed pyrolysis submodel and empirical flame heat feedback submodels, which were developed in this study, is subsequently employed to simulate the early stages of the Room Corner Test, which was selected to represent full‐scale material performance. The results of these simulations demonstrate that this procedure can successfully differentiate between fire growth propensities of several commercially available medium density fiberboards.

Statements from witnesses to a fire are often crucial to the investigation of the fire. Given a large enough sample of witness statements, and by using a systematic approach to the analysis of witness statements, it is possible to provide a clearer picture of the events of a fire and assist in the determination of the fire origin, even when no witness observed the fire in the incipient stage. This paper presents a method for analyzing witness statements for fire investigations in the framework of the scientific method, as recommended by NFPA 921. This type of analysis is appropriate for testing fire origin hypotheses. A case study is presented involving a fire at a 129 unit apartment complex, for which interviews of over 90% of the apartment units are available. These statements have been condensed, mapped against a floor plan of the building, and analyzed such that fire origin hypotheses can be tested. The case study presented organized witness statements from a large fire incident in a manner that allowed the investigators to observe trends in the statements without discarding data that, at first, did not appear relevant or consistent with other statements.