Use of regular patterns-based background oriented schlieren imaging and digital image correlation for visualization of convective airflow

Hot air rising (or convective airflow) is transparent gases (fluids) and invisible. The visualization of fluid dynamics is of considerable interest due to its widespread applications. Background-oriented schlieren (BOS) as an optical method offers numerous advantages in the visualization since this method can provide simple, non-contact, non-invasive, and non-intrusive measurements. The performance of the BOS method strongly depends on the background patterns. Implementing the BOS method usually use random dot patterns as a background and the digital image correlation (DIC) algorithm as a software tool for image processing. However, this algorithm can also analyze the regular pattern background. So far, the experimental BOS using regular patterns background and the DIC algorithm are rarely developed. In this paper, three types of regular patterns, including a checkerboard (crossed fringes), vertical and horizontal fringes patterns, have been proposed as the background in the experimental BOS. During the experiment, a camera is continuously focusing on the background patterns. At the same time, a convective airflow is in the position between the background and the camera. The digital camera records the change of background patterns with and without a convective airflow, giving them distorted and undistorted (reference) images. By comparing both images, the DIC algorithm estimates the pixel displacements. Finally, the interpolation method reconstructs the pixel displacements field into an image resembling the convective airflow. For calibration purposes, the developed BOS uses the known deflection angle of a wedge prism as a standard to assess the accuracy. As a testing, the calibrated BOS visualize the convective airflow induced by a hot object as a heat source, such as candle flame and hot plate. The results show that all the types of backgrounds in the experimental BOS has created image, visualized and distinguished the fluid dynamics of convective airflow. Also, the developed BOS can calculate quantitatively the magnitude of pixel displacements that are proportional to density (or refractive index) gradients of the convective airflow. Thus, the developed BOS can detect an invisible phenomenon and potentially measure density (or refractive index) fluids and other transparent materials.