Measurement dynamics and data treatment in fire testing

Abstract

Fire testing is and will remain of paramount importance in helping industry, regulators and all other fire safety stakeholders to achieve a safer world. For this to happen, test data must be stable and reliable, regardless of the test facility that produced the data, its various operators, the measurement equipment used, and so on. The instruments and sensors used to measure heat release, smoke production and mass loss rate are susceptible to mechanical and/or electrical disturbances and all have their own dynamic behaviour. This leads to a variation in the test results. A scatter that is only due to the measurement.

This paper examines the time and frequency domain behaviour of various measurement components and hardware & software filters in use today. It proposes an instrument-independent method of eliminating noise and other interference, describes a method of adjusting the response time of the various input signals to a common value applicable to all test institutes, and proposes a method of accurately synchronising these signals in time.

The result is unambiguous qualitative measurement data that can be confidently compared with data from other sample sets or between different laboratories and researchers. It improves the stability and reproducibility of test results leading to a more stable classification of materials.