User defined nodal displacement of numerical mesh for analysis of screw machines in FLUENT

International Conference on Compressors and their Systems 2019

Basha, N.^1 ; Kovacevic, A.^1 ; Rane, S.^2

University of London, United Kingdom^1

University of Oxford, Oxford, United Kingdom^2

关键词: Complex geometries;    Internal flow patterns;    Nodal displacement;    Parallel framework;    Positive displacement;    Predicted performance;    Reduce energy consumption;    Segregated algorithm;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/604/1/012012/pdf
DOI  :  10.1088/1757-899X/604/1/012012

来源: IOP

【 摘 要 】

Growing demands to reduce energy consumption are driving researchers towards in-depth analysis of positive displacement machines. Twin screw compressors are amongst the most common types of positive displacement machines. These machines have inherently complex geometry due to intricate rotor profiles used. As the details of the internal flows are difficult to obtain experimentally, Computational Fluid Dynamics (CFD) offers a good alternative for evaluation of internal flow patterns. However, implementation of CFD is challenging due complex deforming geometries. In this paper, a customised grid generator SCORGTM developed by authors is used to generate numerical meshes for commercially available solver ANSYS FLUENT. FLUENT is an unstructured solver which offers flexibility of using both segregated and coupled solution algorithms. Segregated algorithms are generally faster which results in shorter product development time. Interface with FLUENT is implemented by performing User Defined Nodal Displacements (UDND) of grids generated by SCORG in a parallel framework. For this purpose, SCORG and UDND are coupled and extended to work with FLUENT's parallel architecture. The developed code is compiled within the solver. The oil free air screw compressor with 'N' profile rotors and 3/5 lobe combination is modelled for 8000 RPM and 6000 RPM. Finally, the predicted performance values with FLUENT are compared to previously calculated CFX predictions and experimental results. FLUENT requires shorter solution time to obtain same accuracy of CFX.