From the large scale atmospheric turbulent motion to blood flow in microvessels, our daily lives are dependent on various modes of transport phenomena. Different techniques exist that help us to study and understand these complex phenomena experimentally. Flow visualization is one of the most used methods that provide a good qualitative understanding of complex fluid behavior. By definition, flow visualization is the process of making the physics of fluid flows visible. With the development of digital and computational imagery, flow visualization has transitioned from being a qualitative measure to a quantitative measure of transport phenomena. In this research, we have incorporated computational imagery techniques to delineate different scales of transport phenomena such as the flow of convective hot air around a flame and the blood flow inside a mouse’s brain. The visualization was done by generating granular patterns of dark and bright spots, noise, and imposing it on the flow field. For the fire experiments, the granular pattern was created by illuminating a printed noise pattern using white non-coherent light. The hot air around the flame caused fluctuations in air density affecting the air’s refractive index, leading to distortion of the background image. By comparing distorted and undistorted images using optical flow algorithms, the flow field was visualized. A similar algorithm can be deployed to visualize the flow inside a mouse’s brain. Instead of using noncoherent light, the granular pattern was generated by using a laser beam. This granular pattern, produced by random interference effects in laser light, is commonly referred to as laser speckle pattern. When the speckle pattern is generated on blood vessels, the flow of blood causes fluctuations in this pattern leading to a blurriness associated with the blood vessel. This blurriness is later processed using computer algorithms which can be correlated to the motion of the red blood cells. This presentation will describe the pipeline for creating flow visualization of these multiscale transport phenomena and relevant results.

Flow visualization is one of the main methods of understanding flow behavior in experimental fluid mechanics. Most of the visualization methods provides a good qualitative information about the flow behavior. With The developments in Computer sciences especially in fields of artificial Intelligence and computer vision, Image processing has been used as an important tool in experimental fluid mechanics to quantify the qualitative data obtained by flow visualization. One of the common use of computer vision is in “Particle Image Velocimetry”. PIV uses block matching algorithm to detect motion of the fluid between two frames by calculation the displacement of small particles. PIV has many downsides and is not applicable in many fluid mechanic studies such as two phase flow. Because of the downsides of PIV it is better to use other visualization technics and more advanced image processing algorithms. Therefore, Optical flow measurement is used for the advance image processing algorithm which provides flow properties in more variety of applications. Having these algorithms developed, it was used to study bubble behavior in a turbulent flow. Using the optical flow and edge detection algorithms, perimeter velocity of the bubble, and geometric properties of the bubble was obtained in various frames. It was found that the turbulent intensity of the bubble’s perimeter velocity will reach to a common value when the bubble breaks and reaches its smallest possible size.