Flow-rate prediction independent of the regime in a dynamic two-phase flow system using a simple pulse height spectrum of a detector and Artificial Neural Networks

• Published in: Nuclear instruments & methods in physics research. Section A, Accelerators, spectrometers, detectors and associated equipment Accelerators, spectrometers, detectors and associated equipment, 2021-11, Vol.1017, p.165794, Article 165794
• Main Authors: Aarabi Jeshvaghani, P., Khorsandi, M., Feghhi, S.A.H.
• Format: Article
• Language: English
• Subjects: Artificial Neural Networks; Gamma-ray densitometer; NaI(Tl) Detector; Single beam-Single detector; Test loop; Two-phase flows
• ISSN: 0168-9002; 1872-9576
• DOI: 10.1016/j.nima.2021.165794

In this work, the air and water flow-rates were accurately predicted within a two-phase flow loop using features extracted from a simple detector spectrum, independently of the changes in the flow regime. In this regard, a new method based on a single beam-single detector using single-energy of gamma-rays was proposed. The gamma-ray attenuation setup combined with Artificial Neural Network (ANN) was used to predict the flow-rates in various regime of gas–liquid two-phase flows such as bubble, plug, slug, annular and dispersed regimes. Moreover, the ANN was developed based on four features extracted from the recorded pulse height spectrum in the dynamic condition of the fluids. The results showed that the air and water flow-rates can be measured with an average of Mean Relative Error (MRE) less than 4.5%. Overall results revealed that using the proposed method, prediction of the flow-rates can be successfully carried out independent of their regimes.