, Christopher M. Smith
The behavior of engineering structures in fire is commonly studied through large-scale experiments. Full-field, noncontact measurement techniques such as Digital Image Correlation (DIC) are potentially ideal for such experiments; however, the presence of light emitted by the flames, thermal radiation from the heated structure, and convective thermal gradients in the air make this a challenging application for DIC. A simple method has been developed to enable the use of DIC in large, low-soot, fires using narrow-spectrum blue light and spectrally- matched bandpass optical filters to increase signal-to-noise ratio and filter undesired radiant energy before it reaches the camera. The method is applied to full-scale experiments in which a 6 m long W16x26 steel beam is supported over a 700 kW fire from a natural gas diffusion burner. The resulting images are temporally and spatially averaged during post-processing to smooth out false distortions of the images caused by the thermal gradients in and around the flames before DIC techniques are applied to resolve strain.
digital image correlation, DIC, fire, narrow-spectrum illumination, blue light