Search

In transient computational fluid dynamics (CFD) simulations, the continuity of the flow field is an essential prerequisite. However, continuous flows can be separated under certain conditions, such as the process from valve opening to reclosure. The current method often leaves a narrow gap to estimate the full closing status, which will introduce a deviation. To address this issue, a full closing numerical simulation method (FCNSM) is developed to solve the problem of simulation between discontinuous flow field (DFF) and continuous flow field (CFF). The matrix laboratory (matlab) program has been used to communicate fluent as a server session to call the files fluent and automatically execute text-based user interface (TUI) commands. The radial basis function (RBF) is used to construct the relationship between the variables of the flow field and the coordinates of mesh nodes, which can achieve the data transmission from a DFF to a CFF. Automatic stopping of transient calculations is achieved by passing variables among matlab program, scheme language, and user-defined functions (UDF) when a physical quantity reaches a set value. Based on this method, a transient simulation with a dynamic mesh of a 2-D model regarding a pressure relief valve (PRV) is performed to simulate the process of the valve from full closing to reclosure, the flow characteristics through the PRV are obtained using this method. This study makes it possible to use FCNSM for understanding dynamic characteristics from DFF to CFF.

A pressurized vessel-pipe-safety valve (PVPSV) system is a common configuration for many energy management systems and a better understanding of their dynamics is helpful for system design and optimization. In this paper, a method for high-fidelity CFD modelling is presented which can be used to predict dynamic responses of PVPSV systems. For modelling, regions from the vessel outlet to the safety valve exit flange are modeled using a CFD approach; the pressure vessel is set as the boundary and the movement of the valve disc is represented by a 1-D rigid body motion model. Simulations are performed and both stable and unstable operation are investigated. To establish accuracy, an experimental test rig is designed and constructed to measure the motion of the valve disc and the pressures at different system locations. Comparisons are performed for different dynamic modes and good agreement is obtained, supporting the accuracy of the high-fidelity model in reproducing the dynamic response of PVPSV systems. With the developed model, influences of other variables, such as piping length and safety valve configurations can also be evaluated. The method presented in this paper can also be used to develop CFD models for other similar systems and should facilitate system design and optimization.

The flow force acting on a valve disk plays an important role in the overall performance of pressure safety valves (PSVs). To quantify the disk force, computational fluid dynamics (CFD) methods have been widely implemented. In this paper, the capability of CFD models, and the identification of the most suitable turbulence models' geometry modeling and mesh requirements have been assessed to establish the accuracy of CFD models for disk force prediction. For validation purposes, a PSV disk force measuring rig was designed and constructed to obtain the steady-state flow forces exerted on the valve disk at different valve openings. The CFD model assessment is achieved by comparing the simulation results to experimental measurements; this is achieved in two stages. Stage 1 investigates the use of Reynolds averaged Navier–Stokes (RANS)-based turbulence models where two-dimensional (2D) simulations are performed with five turbulence models. The results indicate that a variety of force results are produced by different turbulence models, among which the shear stress transport (SST) k–ω was found to have the best performance. Stage 2 investigates meshing and the use of symmetry and geometry simplifications; 2D, 1/8 three-dimensional (3D) and 1/2 3D mesh models are examined. The results indicate that the 1/8 3D mesh model is the optimal choice, owing to its higher accuracy and reasonable grid scale. The studies performed in this paper extend the knowledge of compressible flow force prediction, and should facilitate the design or optimization of PSVs.