Numerical simulations of jacket side thermal-hydraulic performance for large stirred vessels. (6th January 2021)

• Record Type: Journal Article
• Authors: Kumar, Ankur; Joshi, Jyeshtharaj B.; Vitankar, Vivek
• Is Part Of: Numerical heat transfer. Volume 79:Number 7(2021)
• Page Start: 513
• Page End: 536
• Publication Date: 2021-01-06
• Subjects: Heat -- Transmission -- Measurement -- Periodicals; Mass transfer -- Measurement -- Periodicals; Numerical analysis -- Periodicals; 621.4022
• DOI: 10.1080/10407782.2021.1872254
• Languages: English
• ISSNs: 1040-7782

Abstract

Abstract: In the present work, 3-D numerical simulations have been carried out to understand the flow patterns and heat transfer on the jacket side of a stirred vessel under mixed convection conditions. Various parameters have been varied (1 ≤ Uin ≤ 20 m/s, 30 ≤ ΔT ≤ 60 K, 30 ≤ Jg ≤ 150 mm, Di = 1 and 3 m) to understand their effect on the heat transfer and pressure drop on the jacket side. SST k-ω model has been used for the numerical simulations because of its capability to predict natural convection heat transfer and low Re forced convection heat transfer. It was found that, the heat transfer coefficient increases by 40-80% by increasing the inlet fluid velocity from 1 to 5 m/s. The pressure drop increases by 120-200% and heat transfer per unit pumping power decreases by 60-75% as we increase the velocity from 1 to 5 m/s. It has also been observed that the flow inside the jacket is 3-D, nonuniform and has dead zones and hot pockets at various locations. The effects of inlet flow impinging have also been studied by varying inlet diameter and removing inlet pipe. Impinging by inlet flow on jacket wall contributes about 10-20% to heat transfer and 10-25% to pressure drop. Due to the presence of recirculating flow zones, dead zones, and impinging effect, existing correlations are not suitable for the estimations of heat transfer in large jackets. Large differences have been observed between the 3-D CFD predictions and the widely used empirical correlations. Further, the 3-D CFD predictions have been shown to be comparable to the experimental data available in the published literature. Based on simulations performed in the present work and also validated by experimental data, a new correlation has been proposed which is expected to be useful to design engineers, particularly for those, having limited computational facility.