Abstract:
Self-engineering (SE) systems have valuable abilities to register and respond to lost function and return it. A self-cleaning (SC) system was designed for effective automated cleaning of a heat exchanger (HX) fouled by brewing wort. The system uses temperature outputs in a Digital Twin (DT) simulation and a controller to identify when fouling occurs and trigger a cleaning response. This paper utilises the SE complexity framework and investigates the effectiveness of different complexity designs. Three levels are created for each factor of the framework (repeatability, redundancy and self-control). For repeatability, the number of cleaning cycles was changed, while for redundancy, the flow rate was changed. For self-control, the cleaning mechanism was changed; pulses and foam balls were both used as the cleaning mechanisms. Balls were used to block pipes and redirect flow. An orthogonal matrix is used to reduce the number of experiments. SC effectiveness was measured for each cleaning cycle, and the results were evaluated. Cleaning with the max flow rate (0.21 kg s−1) and using balls and pulses together provided the most effective cleaning, while the worst was with a low flow rate (0.09 kg s−1) and just pulses. Further experiments verified these results and showed that better cleaning settings could lower water use in cleaning. A longer simulation demonstrated when the SC system would be stopped. © 2022
