Study on transient flow analysis of viscosity fluid in spinning metering pump

This paper presents a hybrid fluid solver which couples a grid-free viscous vortex particle method with gridbased unstructured Computational Fluid Dynamics (CFD) solvers. The solver was developed to address the challenging rotor wake and aerodynamic interaction problem. The viscous vortex particle model is a state-of-the-art rotor wake modeling technology that can accurately solve for the complicated rotor wake variation while the unstructured Computational Fluid Dynamics (CFD) solvers can accurately predict the near-body flow. In this coupled solution, the vortex particle method is used to transport vorticity over the flow field while the Computational Fluid Dynamics (CFD) solvers are used to solve near-body flow and provide vorticity sources to the vortex particle method. The wake effect is fed back to the CFD solution to form a coupled solution. This paper discusses the fundamental aspects of the methodology and the associated algorithms that were used to couple the viscous vortex particle method and the unstructured CFD solvers. The developed hybrid tool was used to investigate the unsteady aerodynamic interactions between the helicopter rotor and body and excellent correlations between the simulation results and the measurements were achieved.

Reinforcement learning (RL) is highly suitable for devising control strategies in the context of dynamical systems. A prominent instance of such a dynamical system is the system of equations governing fluid dynamics. Recent research results indicate that RL-augmented computational fluid dynamics (CFD) solvers can exceed the current state of the art, for example in the field of turbulence modeling. However, while in supervised learning, the training data can be generated a priori in an offline manner, RL requires constant run-time interaction and data exchange with the CFD solver during training. In order to leverage the potential of RL-enhanced CFD, the interaction between the CFD solver and the RL algorithm thus has to be implemented efficiently on high-performance computing (HPC) hardware. To this end, we present Relexi as a scalable RL framework that bridges the gap between machine learning workflows and modern CFD solvers on HPC systems, providing both components with its specialized hardware. Relexi is built with modularity in mind and allows easy integration of various HPC solvers by means of the in-memory data transfer provided by the SmartSim library. Here, we demonstrate that the Relexi framework can scale up to hundreds of parallel environments on thousands of cores. This allows to leverage modern HPC resources to either enable larger problems or faster turnaround times. Finally, we demonstrate the potential of an RL-augmented CFD solver by finding a control strategy for optimal eddy viscosity selection in large eddy simulations.

An efficient methodology for the simulation of nonlinear irregular waves in computational fluid dynamics solvers based on the high order spectral method with an application with OpenFOAM

<div class="section abstract"><div class="htmlview paragraph">In this work, the multi-disciplinary problem arising from fluid sloshing within a partially filled tanker truck undergoing lateral acceleration is investigated through the use of multiphysics coupling between a computational fluid dynamics (CFD) solver and a multi-body dynamics (MBD) solver. This application represents a challenging test case for simulation technology within the design of commercial vehicles and is intended to demonstrate a novel approach in the field of computer aided engineering.</div><div class="htmlview paragraph">Computer aided engineering is playing a more predominant role in the design process for commercial and passenger vehicles. Better understanding of the real time loading and responses on a vehicle during intended or unintended use can result in improved design and reduced cost over traditional designs that relied heavily on assumed loads. Liquid sloshing within the cargo tank of a commercial tanker truck results in increased loading on the vehicle's suspension when undergoing acceleration maneuvers. The change in loading can have a significant effect on the design of the vehicles suspension components and braking components. The ability to investigate the fully coupled behavior of the mechanical and fluid systems is a key technology to enable improved designs for these types of applications.</div><div class="htmlview paragraph">The following paper presents a multiphysics analysis of a simplified tanker truck undergoing a lane change maneuver. Bi-directionally coupled CFD and MBD solvers are used to compute the response of the vehicle during a lane change maneuver. The distribution of the liquid within the cargo tank is computed by the CFD solver, <i>AcuSolve</i> using an Arbitrary Lagrange-Eulerian (ALE) mesh motion approach. The forces resulting from the sloshing are then passed to the MBD solver, <i>MotionSolve</i> and the response of the tanker truck is computed. This exchange of forces and displacements occurs at run time and is enabled through a socket connection between the two solvers.</div></div>

The flow ripple, which is the source of noise in an axial piston pump, is widely studied today with the computational fluid dynamic(CFD) technology development. In the traditional CFD modeling, the fluid compressibility, which strongly influences the accuracy of the flow ripple simulation results, is often neglected. So a compressible sub-model was added with user defined function(UDF) in the CFD model to predict the flow ripple. At the same time, a test rig of flow ripple was built to study the validity of simulation. The flow ripple of pump was tested with different working parameters, including the rotation speed and the working pressure. The comparisons with experimental results show that the validity of the CFD model with compressible hydraulic oil is acceptable in analyzing the flow ripple characteristics. In this paper, the improved CFD model increases the accuracy of flow ripple rate to about one-magnitude order. Therefore, the compressible model of hydraulic oil is necessary in the flow ripple investigation of CFD simulation. The compressibility of hydraulic oil has significant effect on flow ripple, and the compression ripple takes about 88% of the total flow ripple of pump. Leakage ripple has the lowest proportion of about 4%, and geometrical ripple leakage ripple takes the remnant 8%. Besides, the influence of working parameters was investigated through the CFD simulations and experimental measurements. Comparison results show that the amplitude of flow ripple grows with the increasing of rotation speed and working pressure, and the flow ripple rate is independent of the rotation speed. However, flow ripple rate of piston pump grows with the increasing of working pressure, because the leakage ripple will increase with the pressure growing. The investigation on flow ripple of an axial piston pump using compressible hydraulic oil provides a more validity simulation model for the CFD analyzing and is beneficial to further understanding of the flow ripple characteristics in an axial piston pump.

Since the last two decades, aerospace agencies around the world have started planning Space-based Solar Power Systems (SSPS) as an alternative power source [1]–[9]. The use of Space Shuttle Orbiter type re-usable launch vehicles will enable the completion of this project in a limited time span with economic feasibility. This would leave less time between successive launches, making it imperative that the aero-thermodynamic analysis of these vehicles be fast and accurate. Currently, aero-thermodynamic analysis is done by high fidelity Computational Fluid Dynamics (CFD) solvers which are accurate but take significantly more time to give necessary results. Therefore, the present CFD solvers might not be useful tools for this mission to be completed in limited time. In this work a low cost, quick and reasonably accurate model is developed when compared with the CFD results [10]. Pressure coefficients and surface temperatures from this code are compared with results from the CFD solver of ‘Air Force Research Lab, Wright-Patterson’ [10]. The model developed in this work is based on hypersonic theories which meet the requirements of this mission.

For industrial design and the improvement of fluid flow simulations, computational fluid dynamics (CFD) solvers offer practical functions and conveniences. However, because iterative simulations demand lengthy computation times and a considerable amount of memory for sophisticated calculations, CFD solvers are not economically viable. Such limitations are overcome by CFD data-driven learning models based on neural networks, which lower the trade-off between accurate simulation performance and model complexity. Deep neural networks (DNNs) or convolutional neural networks (CNNs) are good illustrations of deep learning-based CFD models for fluid flow modeling. However, improving the accuracy of fluid flow reconstruction or estimation in these earlier methods is crucial. Based on interpolated feature data generation and a deep U-Net learning model, this work suggests a rapid laminar flow prediction model for inference of Naiver–Stokes solutions. The simulated dataset consists of 2D obstacles in various positions and orientations, including cylinders, triangles, rectangles, and pentagons. The accuracy of estimating velocities and pressure fields with minimal relative errors can be improved using this cutting-edge technique in training and testing procedures. Tasks involving CFD design and optimization should benefit from the experimental findings.

Spectral Wave Explicit Navier-Stokes Equations for wave-structure interactions using two-phase Computational Fluid Dynamics solvers

In this study, we present results from a numerical model of a full-scale fracture propagation test where the pipe sections are filled with impure, dense liquid-phase carbon dioxide. All the pipe sections had a 24″ outer diameter and a diameter/thickness ratio of ∼32. A near symmetric telescopic set-up with increasing toughness in the West and East directions was applied. Due to the near symmetric conditions in both set-up and results, only the East direction is modelled in the numerical study. The numerical model is built in the framework of the commercial finite element (FE) software LS-DYNA. The fluid dynamics is solved using an in-house computational fluid dynamics (CFD) solver which is coupled with the FE solver through a user-defined loading subroutine. As part of the coupling scheme, the FE model sends the crack opening profile to the CFD solver which returns the pressure from the fluid. The pipeline is discretized by shell elements, while the backfill is represented by the smoothed-particle hydrodynamics (SPH) method. The steel pipe is described by the J2 constitutive model and an energy-based fracture criterion, while the Mohr-Coulomb material model is applied for the backfill material. The CFD solver applies a one-dimensional homogeneous equilibrium model where the thermodynamic properties of the CO2 are represented by the Peng-Robinson equation-of-state (EOS). The results from the simulations in terms of crack velocity and pressure agree well with the experimental data for the low and medium toughness pipe sections, while a conservative prediction is given for the high-toughness section. Further work for strengthening the reliability of the model to predict the arrest vs. no-arrest boundary of a running ductile fracture is addressed.

The ability of fish to maneuver in tight places, perform stable high acceleration maneuvers, and hover efficiently has inspired the development of underwater robots propelled by flexible fins mimicking those of fish. In general, fin propulsion is a challenging fluid-structure interaction (FSI) problem characterized by large structural deformation and strong added-mass effect. It was recently reported that a simplified computational model using the vortex panel method for the fluid flow is not able to accurately predict thrust generation. In this work, a high-fidelity, fluid-structure coupled computational framework is applied to predict the propulsive performance of a series of biomimetic fins of various dimensions, shapes, and stiffness. This computational framework couples a three-dimensional finite-volume Navier-Stokes computational fluid dynamics (CFD) solver and a nonlinear, finite-element computational structural dynamics (CSD) solver in a partitioned procedure. The large motion and deformation of the fluid-structure interface is handled using a validated, state-of-the-art embedded boundary method. The notorious numerical added-mass effect, that is, a numerical instability issue commonly encountered in FSI simulations involving incompressible fluid flows and light (compared to fluid) structures, is suppressed by accounting for water compressibility in the CFD model and applying a low-Mach preconditioner in the CFD solver. Both one-way and two-way coupled simulations are performed for a series of flexible fins with different thickness. Satisfactory agreement between the simulation prediction and the corresponding experimental data is achieved.

Computational fluid dynamics (CFD) solvers provide helpful components and amenities for industrial development and the advancement of fluid flow simulations. Nevertheless, CFD solvers are not advantageous since iterative simulations require high computational resources and enormous memory for complex calculations. The deep neural network-based CFD data-driven learning method eliminates these constraints by lowering the compensation between model complicatedness and precision. In this paper, we present the feasibility of predicting fluid flow velocity fields based on CFD using U-Net architecture, a subdomain of deep learning. The experimental results show that the U-Net architecture can predict fluid flow with a total loss of 0.2223, a validation loss of 0.2728, and an accuracy of 86% from our private dataset. Our U-Net model can be used to predict fluid flows, which has been proven.

<div>0D, quasi-3D, and 3D chemistry solvers with varying degrees of complexity are developed to predict the thermal runaway propagation in battery cells. The 0D solver assumes the system as homogeneous and closed. The quasi-3D solver assumes the system as homogeneous on the selection level and the 3D solver accounts all spatial inhomogeneities in the temperature and composition. Both the quasi-3D and 3D solvers are fully integrated into a computational fluid dynamic (CFD) solver and capable of predicting thermal runaway in multiple battery cells with cell-specific kinetic reaction model. As the modeling complexity increases with each solver, respectively, the accuracy and the simulation time increases. With the large amount of heat and rapid transitions from the onset of thermal runaway, the CFD solvers usually encounter difficulties in predicting the solution accurately and in extreme heat release cases the solver may diverge. A chemical time scale based adaptive time stepping is developed in this work to address the accuracy, convergence, and stability issues of the CFD solver. The proposed timescale contains in the definition the reaction rate, reaction enthalpy, and total enthalpy content of the system. As the thermal runaway progresses, the CFD solver time step is obtained dynamically from the defined timescale. The developed solvers and the adaptive time-stepping method were quite intensively tested and analyzed by using different reaction mechanisms representing different battery cells and test conditions. The analysis of the timescales and the adaptive time stepping proved quite efficient for solution accuracy, simulation time, and solver stability.</div>

Multi-Disciplinary Optimization (MDO) is widely used to handle the advanced design in several engineering applications. Such applications are commonly simulation-based, in order to capture the physics of the phenomena under study. This framework demands fast optimization algorithms as well as trustworthy numerical analyses, and a synergic integration between the two is required to obtain an efficient design process. In order to meet these needs, an adaptive Computational Fluid Dynamics (CFD) solver and a fast optimization algorithm have been developed and combined by the authors. The CFD solver is based on a high-order discontinuous Galerkin discretization while the optimization algorithm is a high-performance version of the Artificial Bee Colony method. In this work, they are used to address a typical aero-mechanical problem encountered in turbomachinery design. Interesting achievements in the considered test case are illustrated, highlighting the potential applicability of the proposed approach to other engineering problems.

An open source computational fluid dynamics (CFD) solver has been incorporated into the WindNinja modeling framework. WindNinja is widely used by wildland fire managers, as well as researchers and practitioners in other fields, such as wind energy, wind erosion, and search and rescue. Here, we describe the CFD solver and evaluate its performance against the WindNinja conservation of mass (COM) solver, and previously published large-eddy simulations (LES), for three field campaigns with varying terrain complexity: Askervein Hill, Bolund Hill, and Big Southern Butte. We also compare the effects of two model settings in the CFD solver, namely the discretization scheme used for the advection term of the momentum equation and the turbulence model, and provide guidance on model sensitivity to these settings. Additionally, we investigate the computational mesh and difficulties regarding terrain representation. Two important findings from this work are: (1) CFD solver predictions are significantly better than COM solver predictions at windward and lee side observation locations, but no difference was found in predicted speed-up at ridgetop locations between the two solvers, and (2) the choice of discretization scheme for advection has a significantly larger effect on the simulated winds than the choice of turbulence model.

Fluid analysis of cylindrical and screw type Roots vacuum pumps