MODELING OF EXERGY EFFICIENCY PERFORMANCES OF COUNTER FLOW RANQUE-HILSCH VORTEX TUBES WITH DIFFERENT GEOMETRIC CONSTRUCTIONS USING ARTIFICIAL NEURAL NETWORKS

In this experimental study, the exergy efficiency performances of counter flow type Ranque-Hilsch vortex tubes (RHVT), with a length to diameter ratio of 10-18, were investigated for RHVT made of aluminum, brass, and steel having the internal diameter (D) of 9 mm. Cross section of nozzle was 0.002x0.002 m(2) and the number of nozzles (Nn) was 3. Flow was controlled with a valve on the hot outlet side, and the valve at the hot outlet side was changed from a nearly closed position from its nearly open position. The exergy efficiency of RHVT for various L/D ratios (10-18) and various ksi values (0.1-0.9) were determined under 440 kPa pressurized air. As a new approach, this study proposes determining the exergy efficiency by using artificial neural networks (ANN). As ANN input parameters, L/D, ksi, total outlet exergy and total lost exergy were used, while the exergy efficiency was the output parameter. The actual values and ANN results show that ANN can be successfully used for the exergy efficiency performances of counterflow Ranque-Hilsch vortex tube.