THE DEVELOPMENT OF PARALEL PROCESSING OF MICROBUBBLE GENERATOR CHARACTERIZATION USING PIV METHOD, TEMPLATE MATCHING AND DISSOLVED OXYGEN DECAY RATE

• Main Author: Ibnu Salim, Taufik
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/52000
• Institution: Institut Teknologi Bandung

The development of the aquaculture industry in Indonesia has started to utilize microbubble technology for the aeration process, namely increasing dissolved oxygen levels in the water. Many aquaculture farmers use microbubble generator, both locally produced and impored. However, there are still few who carry out research and development of microbubble generator, especially until the characterization of bubble size and their hydrodynamic properties. This thesis proposes research of the continuous characterization of the microbubbles generator and their hydrodynamic properties such as the decay of dissolved oxygen levels. Microbubble characterization specializes in estimating the size distribution using the particle image velocimetry (PIV) method with image and template matching techniques. Dissolved oxygen decay measurements using dissolved oxygen (DO) sensor are carried out continuously. Then compare the DO decay to the change in the bubble size distribution over time to find the correlation coefficient. The research was carried out using reverse osmosis water into an aquarium which has a volume of 54 liters measuring 0.6 x 0.3 x 0.3 m3. Using a digital microscope camera with a resolution of 640x480, and a frame rate of 25fps. The DO level sensor uses a Lumin S Optical DO sensor from Sensorex with a measuring range of 0-20 ppm and a resolution of 0.1 ppm. The light source used is a LED light board with an intensity of 800 lumens which has three levels of brightness. A digital microscope camera records video for 30 seconds to be saved into a file server in the form of a USB flash disk using the recorder service. The storage address is stored in the Mongodb database. The pre-processing unit extracts the video into frames into the file server and adds the data to the database. The processing unit takes the data extracted from the video to calculate the vector displacement value using the template matching method. Bubble size is calculated using the Hadamard-Rybczynski equation which is then stored in a database. Both the pre-processing unit and processing unit can be added in number and run parallel in the form of a service that runs behind the scenes. The pre-processing duration takes 3.5 times the recording time. Processing duration with downsample 1/60 takes less than pre-processing time. For this reason, a minimum of 4 parallel services is required so that the analysis process runs continuously. The result of micro bubble size distribution mesurement have an average of 14.60 µm and deviation 0.11 µm. Oxygen level decay analysis resulted in a decay of -0.04 ppm / minute in the first 5 minutes and a decay of <0.01 ppm / minute after 30 minutes. There is a correlation between DO decay and the reduction in bubble size in the first five minutes with a value of 0.71.