In the field of safety engineering, the propagation speed and characteristics of flames directly affect the spread and degree of fire damage. Among them, the propagation dynamics mechanism of blue flames is a core issue in fire safety research because of their full combustion and high temperature. Traditional image acquisition equipment is not sensitive enough in weak light (quantum efficiency <30%), and it is difficult to capture the transient behavior of blue flames. With its high quantum efficiency, low readout noise, and high dynamic range, sCMOS scientific cameras can effectively capture the subtle changes and transient behaviors of flames, helping to study the propagation mechanism of blue flames.
Recently, Revealer technical engineers and researchers from a safety engineering laboratory jointly carried out an observation experiment based on sCMOS scientific cameras.
1) sCMOS scientific camera: Revealer Revealer Gloria 4.2, derived from GSENSE2020BSI chip, with a resolution of 2048×2048 and a quantum efficiency of >70% in the 410nm band.
2) Combustion experimental device, with a size of 1.2m×0.8m×0.6m, can ensure the stability and repeatability of combustion by adjusting the oxygen volume fraction in the combustion environment and simulating oxygen-rich conditions. 3) Bandpass filter, with a central wavelength of 410nm and a half-width of 10nm, is used to suppress background spectral interference and improve the signal-to-noise ratio. 4) Scientific imaging analysis software RPC, used to collect and store image data obtained by sCMOS scientific cameras in real time.
1) Premixed methane-oxygen, equivalence ratio Φ=0.9, flow rate 5L/min, oxygen concentration gradient 25%~35%.
2) Set up the sCMOS scientific camera at the lateral ignition system of the combustion device, with a 50mm C-mount lens, and install the filter in front of the camera lens.
3) Adjust the focal length and aperture of the sCMOS scientific camera, increase the exposure time and gain, and ensure clear capture of the combustion area.
4) Use LabView to control ignition, gas supply and scientific camera synchronization, and the synchronization error is less than 50μs.
5) Change the oxygen concentration experimental conditions to obtain image data of blue flames under different conditions.
6) Based on the self-programmed binarization analysis method, the researchers extracted the edge contour of the blue flame, calculated the displacement of the flame at different time points based on the extracted flame edge position information, and further calculated the propagation speed of the blue flame.
The original image captured by the experiment clearly shows the wrinkle area of the flame. By adjusting the experimental conditions, it was found that under oxygen-rich conditions, the displacement speed of the blue flame was significantly higher than that of the flame under normal air conditions, and flashover was prone to occur during the propagation process.
In this experiment, the sCMOS scientific camera had a good response in the ultraviolet and blue light bands, revealing the microscopic dynamic mechanism of oxygen-rich flames and providing intuitive and detailed image data for studying the propagation mechanism of flames. In the future, researchers will combine AI algorithms to develop a flame semantic segmentation model based on the U-Net model to improve the diagnostic efficiency under complex backgrounds and provide stronger theoretical support and technical guarantees for precise prevention and control in the field of fire safety engineering.
