Grid generation for complex high-lift configurations

An efe cient and user-friendly approach is presented for constructing unstructured surface grids directly on three-dimensional CAD surface data. This system has the following features: 1 ) use of stereolithography data as surface dee nitions, that is, background grids for surface meshing; 2 ) automatic reconstruction of geometric features for initial front setup; 3 ) graphical user interface for easily controlling surface mesh density by inserting line sources and point sources on the surface; 4 ) adoption of an advancing-front surface triangulation method, which minimizes the necessity to divide the surface into a number of patches; and 5 ) outer boundary generation from templates in the system. The resulting system has been applied to several airplane cone gurations, and a notable reduction in turnaround time has been achieved. OMPUTATIONAL e uid dynamics (CFD) has become an indispensable design and analysis tool for many different types of geometric cone gurations and e eld regimes. The effectiveness of CFD usage in an actual design environment, however, is often limited. Most of these limitations are due to dife culties in generating the computational grid in a reasonable amount of time. The grid generation method should be automated as much aspossible. This is true for volume grid generation; it is now possible to create volume grids automatically. For the surface grid generation, however, user interventions are often required because the surface meshdirectlyaffectsthesolutionaccuraciesinCFDproblems,especiallyfortheaerodynamicevaluationofairplanes.Theaerodynamic performance is usually evaluated by surface values, such as pressure,skin friction,and so on.Thesevalues arehighlyaffected by the dee nition of sound geometric features, as well as the surface grid density. This suggests that surface grid generation methods must easily facilitate changes in grid density to obtain the best result with limited computational resources. Although controllability is often incompatible with the automation of grid generation, it is essential for surface meshing.

The National Grid Project

Numerical simulations have been performed for a simplified high-lift (SHL) version of the Common Research Model (CRM) configuration, where the Fowler flaps of the conventional high-lift (CRM-HL) configuration are replaced by a set of simple hinged flaps. These hinged flaps are equipped with integrated modular active flow control (AFC) cartridges on the suction surface, and the resulting geometry is known as the CRM-SHL-AFC configuration. The main objective is to make use of AFC devices on the CRM-SHL-AFC configuration to recover the aerodynamic performance (lift) of the CRM-HL configuration. In the current paper, a Lattice Boltzmann method-based computational fluid dynamics (CFD) code, known as PowerFLOWQ is used to simulate the entire flow field associated with the CRM-SHL-AFC configuration equipped with several different types of AFC devices. The transonic version of the PowerFLOWQ code that has been validated for high speed flows is used to accurately simulate the flow field generated by the high-momentum actuators required to mitigate reversed flow regions on the suction surfaces of the main wing and the flap. The numerical solutions predict the expected trends in aerodynamic forces as the actuation levels are increased. More efficient AFC systems and actuator arrangements emerged based on the parametric studies performed prior to a Fall 2018 wind tunnel test. Preliminary comparisons of the numerical solutions for lift and surface pressures are presented here with the experimental data, demonstrating the usefulness of CFD for predicting the flow field and lift characteristics of AFC-enabled high-lift configurations. \n\n\n\n\n

An Elliptic Multiblock grid generation technique is presented. Boundary orthogonality is imposed by oating the grid points along the boundary. A formulation is presented to maintain the grid spacing at the boundaries. For the case of three dimensions an improved search algorithm maintains the grid orthogonality on the surface along with grid spacing. The speci cation of orthogonal boundary condition as a natural outcome of grid generation process preserves the grid quality obtained by the standard elliptic solver. Grid clustering is done algebraically, thereby de-coupling it from the basic equation in the grid generation process. Elliptic smoothing is used to spread out the clustered grid points adaptively resulting in a very smooth, clustered grid. The advantage of oating grid points at the boundary is exploited to yield a customized clustered grid with smoother grid distribution. User interaction is aided with the help of a TCL (Tool Command Language) interpreter. Since the surface grid is generated along with the volume grid, the grid generation script can be readily reused to generate grids of similar topological structure. The features of grid generator has been demonstrated with typical examples.

Analysis of non-conservative interpolation techniques in overset grid finite-volume methods

Overset grid technology development at NASA Ames Research Center

In this paper, a computational fluid dynamics trimming method is proposed and compared with wind tunnel experiment and the blade element method. The NASA’s generic ROBIN helicopter model is adopted for transient simulations to obtain the final main rotor trimming conditions. Totally three steps were applied to the computational fluid dynamics method. The first step is associated with no cyclic pitch motion, the second is regarding pure longitudinal cyclic pitch motion and the last is concerning with pure lateral cyclic pitch motion. At the same time, a simple linear equation system between the roll and pitching moment was established to get the final longitudinal and lateral cyclic pitch angles for the main rotor through the above three steps. An overset grid approach was used where the volume around each blade was modeled in an individual overset grid region. The rotor rotation was resolved with three degrees per time-step. Turbulence was modeled with the well-known SST K-omega model with second-order convection. The helicopter was in straight forward flight with an advance ratio of [Formula: see text]. Three sources of stick angles, which are also called rotor trimming angles, were shown and compared with each other. And the corresponding results were also plotted with a type of history plot in the computational fluid dynamics condition. In the simulations, the results became quasi periodic after about 1.5 rotations and four rotor rotations were simulated for each case. The pitch moment coefficient and roll moment coefficient were all trimmed to about zero by the computational fluid dynamics trimming method while moments were not removed thoroughly in the other two source conditions.

A novel (geometrical) parametrization procedure using solutions to a suitably chosen fourth order partial differential equation is used to define a class of airplane configurations. Inclusive in this definition are surface grids, volume grids, and grid sensitivity. The general airplane configuration has wing, fuselage, vertical tail and horizontal tail. The design variables are incorporated into the boundary conditions, and the solution is expressed as a Fourier series. The fuselage has circular cross section, and the radius is an algebraic function of four design parameters and an independent computational variable. Volume grids are obtained through an application of the Control Point Form method. A graphic interface software is developed which dynamically changes the surface of the airplane configuration with the change in input design variable. The software is made user friendly and is targeted towards the initial conceptual development of any aerodynamic configurations. Grid sensitivity with respect to surface design parameters and aerodynamic sensitivity coefficients based on potential flow is obtained using an Automatic Differentiation precompiler software tool ADIFOR. Aerodynamic shape optimization of the complete aircraft with twenty four design variables is performed. Unstructured and structured volume grids and Euler solutions are obtained with standard software to demonstrate the feasibility of the new surface definition.

Numerical analysis of a propeller during heave motion in cavitating flow

Computational Fluid Dynamics (CFD) methods have been developed into effective fluid simulation means to be used on the hydraulic design in the field of nuclear reactor. However, it is difficult to generate suitable mesh and select turbulence model for simulation because of the complex geometry structure and flow behavior in the pressure vessel. Based on CFX software, a hydraulic computation model of typical pressurized water reactor is established and the flow distribution at core inlet is analyzed. The simplified geometry model consists cold legs, downcomer, lower plenum, secondary support component, core support plate, and lower core plate. The computation model is divided into three parts for mesh generation, including the part of inlet and downcomer, the part of lower plenum and core inlet section, and the part of core. In order to reach the independence of grid several methods of mesh generation which contains different mesh density at local key parties are investigated to screen out the suitable mesh scheme. The k-ε, k-ω, and SST k-ω turbulence model are respectively used for simulation and the sensitivity of turbulence model at different locations of flow field is analyzed. The results show that the mass flow rate of the near wall flow field, computed by using k-ω turbulence model, is consistent with SST k-ω model, while the mass flow rate of central flow field computed by using k-ε turbulence agrees with SST k-ω model. The result computed by using k-ε turbulence model shows relatively uniform flow distribution at core inlet, which is more consistent with the measured data with the average difference of 3.1%. By using the k-ε model, the probability distribution of the difference between the calculated results and the experimental values follows the law of Normal distribution. The final coolant flow distribution at each orifice is evaluated, and the maximum normalized flow flux is found at center orifice while the flow rate at the edge of core is relatively lower.

This publication presents experimental investigations on the aerodynamics of a highlift airfoil in disturbed inflow. Transient, three-dimensional disturbances are generated in a closed wind tunnel test-section. These disturbances are created by means of two vortex generating airfoils. The first creates quasi two-dimensional, transient, transversal vortices whereas the second airfoil generates a longitudinal wingtip vortex. Convecting along the test-section, these vortices interact with each other and with a two-element airfoil in high-lift configuration. Various measurement techniques, like Particle Image Velocimetry (PIV), time resolved static pressure measurements, measurements of the static pressure distribution of the high-lift airfoil, and time resolved Five-Hole-Probe (FHP) measurements of the induced disturbances are used to identify and characterize these interactions. These measurements are complemented by qualitative surface oilflow visualizations of the highlift airfoil. It is found that the interaction of the vortices with the high-lift airfoil exhibits a varying complexity with changing angle of attack. At zero incidence, the effective flow angles induced by the vortices prevail. For higher angles of attack, complex interaction effects can be observed in the time-resolving static pressure signals. By phase-locked PIV measurements, it can be shown that the flow field at the high-lift flap for small incidences is governed by the circulation effect of the main element rather than by the velocity effect of the passing vortices.

The advent of General Purpose Graphics Processing Units (GPGPUs) has spawned a lot of interest in computing high resolution flows in a shorter span of time. In combination with existing parallel programming techniques such as MPI or openMP, one is able to obtain at least one order increase in speed-up for CFD applications on stationary grids. Due to a high throughput/cost ratio, GPUs are increasingly becoming popular among the CFD community. Flow fields on helicopter systems rank among one of the most challenging, and hence are computationally demanding to simulate. Further, presence of multiple bodies moving relative to each other require use of overset grid systems, which in turn, require efficient overset grid assembly methods to support the flow solution. In this context, a computational framework for flow simulation across multiple CPU cores and multiple GPUs has been developed and tested. It is shown that the overall wall-clock time for the simulation can be considerably reduced by using multiple GPU cards. The parallel GPU framework is initially tested using both explicit and implicit schemes for the flow past a sphere. Further, a two-blade hovering rotor test case adopted from literature is used to demonstrate the capability of the code towards simulating rotorcraft flows in a short span of time.

The implementation of two different types of turbulence models for a flow solver using the Chimera overset grid method is examined. Various turbulence model characteristics, such as length scale determination and transition modeling, are found to have a significant impact on the computed pressure distribution for a multielement airfoil case. No inherent problem is found with using either algebraic or one-equation turbulence models with an overset grid scheme, but simulation of turbulence for multiple-body or complex geometry flows is very difficult regardless of the gridding method. For complex geometry flowfields, modification of the Baldwin-Lomax turbulence model is necessary to select the appropriate length scale in wall-bounded regions. The overset grid approach presents no obstacle to use of a one- or two-equation turbulence model. Both Baldwin-Lomax and Baldwin-Barth models have problems providing accurate eddy viscosity levels for complex multiple-body flowfields such as those involving the Space Shuttle.

The paper describes numerical studies by DLR devoted to the computation of the DLR F-11 high lift configuration as part of the 2nd phase of the AIAA High Lift Prediction Workshop. The model geometry is representative for a wide-body commercial aircraft with a classical three element high lift system at the wing leading and trailing edge in a landing setting. Being a follow-up of the NASA Trapezoidal wing of the 1st phase of the workshop, the F-11 wing/body wind tunnel model features only a modest overall complexity increase in the basic set-up. Yet, the most complex geometry representation includes model details, such as slat tracks, flap track fairings, and pressure tube bundles. According to the available experimental evidence, low and high Reynolds numbers have been investigated. In addition to force, moment, and pressure distributions, surface streamlines have been analyzed. The studies have been carried out using the DLR TAU code in conjunction with the Spalart-Allmaras turbulence model in the original formulation. DLR contributed to the grid generation activities by providing a set of hexahedral-based hybrid unstructured grids for all complexity stages and both Reynolds-numbers. For the most simplified configuration, a family of three grids has been generated to study grid resolution aspects. In general, a good agreement to the experimental data is obtained. It turns out that for the present wind tunnel configuration, the inclusion geometry details of the wind tunnel model, such as slat tracks and even pressure tube bundles, is essential for the agreement to the experimental evidence and the prediction of maximum lift.

A comprehensive grid generation system has been developed which enables grid generation for complex aircraft configurations. This system is especially suited for generation of zonal grids without point continuity at zoneto-zone interfaces. This approach reduces the complexity of the grid generation process by allowing grids to be generated with fewer zones, and fewer unnecessary grid points, thereby reducing solution time and cost. The capabilities of the system include geometry acquisition and manipulation, grid generation on the boundaries of each zone, three-dimensional field grid generation, grid quality evaluation for the field grid, boundary condition specification, coupling parameter identification for zone-to-zone interfaces, and output directly compatible with a zonal Euler or Navier-Stokes flowfield analysis code. The system operates in an interactive environment through a userfriendly graphical interface running on several computer platforms with numerous graphics devices. In order to further improve the efficiency of the grid generation process, the system utilizes several macro operations to perform sequences of steps ranging from generating various farfield boundary shapes to generating an entire zone for an isolated forebody or diffuser type geometry. A prototype expert system has also been implemented to assist the user in identifying and correcting flaws in the field grid. Applications of the system to inlet-forebody analyses, a duct with turning vanes, and a thrust-vectoring nozzle configuration will be presented. I n t r o d u c t i o n Numerical solution of the Navier-Stokes equations is becoming increasingly important for the analysis of advanced propulsion systems and their integration with an airframe. The ability to generate a suitable grid is often the pacing i tem in obtaining solutions for these complex configurations. High quality grids are necessary to accurately predict flowfields. High quality grids a180 increase the convergence rate of the solution, thereby re* Lend Engineer Aemdynsmics ** Technical Specisiixt Aerodyrmnics t Senior Engineer Aerodynamics t Technical Specialist Propulsion 5 Senior Engineer Propulsion Copyright @ 1991 by McDonnell Douglas Cormration Published by the American Institute 01 Aeronautics and Astronautics, Inc. with permission. ducing computer resource requirements. The grid must accurately represent the geometry. It must also be tnilored to the flowfield, Le., i t must provide high density near solid boundaries, shock locations, and other regions where high Bowfield gradients may be experienced. Any flaws in the grid must be identified and eliminated before the grid is used in the flowfield solution algorithm or else costly regridding and repeated analysis will be required. The McDonnell Aircraft Computational Grid System (MACGS) was developed to address these issues. It uses an interactive zonal approach which does not require point continuity a t zone-to-zone interfaces. The zonal approach maximizes flexibility and simplifies the grid gcneration problem. In a zonal approach, the physical domain around the configuration is subdivided into simpler regions 01 zones. These zones are non-overlapping and ., boundary conforming. A zonal approach dlows various strategies for subdividing the field into zones or the use of a single zone, where appropriate. Current zonal grid generation systems, such as EAGLE (Reference 1) and GRIDGEN (Reference Z), are highly focused on generation of grids with point continuity at the zonal interfaces. Although these grid generators are very versatile and allow zones without point continuity to be generated independently, these systems lose some of their power for such cases. They do not address how adjacent zones will be treated. MACGS generates grids without point continuity at the zonal interface and computes the interpolation factors t o allow flowfield da t a to be transferred between zones. Lack of point continuity restrictions makes grid generation much more flexible because the user can distribute points where they are needed, based on flow properties within the zone rather than on where points were distributed i n an adjacent zone. This allows the user t o have as many grid points as required in a critical flowfield region, such as a boundary layer, with far fewer points in a less critical adjacent zone, such as the region upstream of an inlet. The zonal approach without point continuity was previously implemented for specific configurations in the McDonnell Aircraft Company (MCAIR) three-dimensional batch grid generator, ZGRID (Reference 3), and has been used successfully with the MCAIRzonal three-dimensional Euler or Navier-Stokes flow solver, NASTD (Reference 4). Interactive methods provide improved grids by giving the user more direct control of and feedback from the i /