Andrew J. Lofthouse

Prediction of Aerodynamic Characteristics of Ram-Air Parachutes

The focus of this work is on the computational methodology for aerodynamic modeling of ram-air parachutes and increasing confidence and understanding in their concept designs including new parachute control methods. The complex geometries of ram-air parachutes are modeled by two-dimensional rigid airfoil geometries with or without trailing-edge deflections and bleed air spoilers. The aerodynamic forces are then calculated from steady or unsteady Reynolds-averaged Navier–Stokes simulations using Cobalt and Kestrel flow solvers. The effects of the grid size and type, the time step, and the choice of solver parameters are investigated. The flow solvers are then used to study the flow around three-dimensional wings with open/closed ram-air inlets by comparing lift and drag coefficients with available experimental data. The results show that computational fluid dynamics simulations are a valuable aid in understanding the flow structure of ram-air parachutes, which resemble a rectangular wing with open inlets. ...

Vortical Flow Prediction of the AVT-183 Diamond Wing

The objective of this paper is to assess the potential and limitations of current practice in computational fluid dynamic modeling for predicting vortical flowfields over a generic 53degree swept diamond wing with rounded leading edges. This wing was designed under STO AVT Task Group 183 and is based on the SACCON UCAV geometry, which is a lambda wing with complex span-wise distributions of thickness and leading edge radius and with a linear twist outboard of the first trailing edge break. The SACCON wing trailing-edge was swept forward by 26.5-degrees to form a diamond-shaped planform that is used in this study. This new wing also has a constant NACA 64A-006 airfoil section with a leading edge radius of 0.264 percent chord and has no twist. CFD simulations were performed for various angles of attack at a Mach number of 0.15 and a Reynolds number of 2.7× 10 based on the mean aerodynamic chord to match experiments. CFD simulations were run with different turbulence models and with a limited assessment of Delayed Detached Eddy Simulation. The wind tunnel experiments of the diamond wing were carried out in the Institute of Aerodynamics and Fluid Mechanics of the Technische Universitat Munchen, Germany and include aerodynamic lift, drag, and pitch moment measurements as well as span-wise pressure distributions at different chord-wise locations. This data set is used to validate the CFD results. The results presented demonstrate that the CFD compare well with the experiments at small angles of attack; the pitch moment predicted by the SARC turbulence model provide a better match to experimental results than the SA model at moderate angles of attack; and at high angles of attack, CFD predictions are not as good. The flow visualization results show that a leading-edge vortex is formed above the upper surface of the wing at an angle of attack of about eight degrees. This vortex becomes larger and stronger when the angle of attack is increased. With increasing angle of attack, the vortex formation point moves upstream and the vortex core moves inboard towards the wing center. Finally, the results show that the flow over the diamond wing is steady throughout the range of angles of attack.

Static and Dynamic Simulations of a Generic UCAV Geometry Using the Kestrel Flow Solver

The Kestrel CFD solver is used to characterize the static and dynamic aerodynamic behavior of a generic UCAV configuration with control surfaces, named the DLR-F19. Computational results are compared to wind tunnel data from experiments performed in DNW-NWB wind tunnel at several different angles-of-attack, ranging from 0 to 30 degrees, with no flap deflections and flap deflections for maximum roll. Results for several angles-ofsideslip are also presented. The aerodynamics around the UCAV configuration are highly nonlinear, even at moderate angles-of-attack, due to vortex interactions. In general, good agreement between computational results and the wind tunnel experimental data for total force coefficients is achieved using relatively inexpensive steady-state simulations with the SARC turbulence model for the static cases, and unsteady simulations with the SARC turbulence model for the dynamic pitching motions. However, unsteady simulations using a DDES turbulence model are required for angels-of-attack greater than 20-degrees. The computational models have difficulty predicting the finer details of the nonlinear moments, most likely due to the challenge of accurately predicting separation and vortex breakdown. This work is a contribution to the NATO RTO Task Group AVT-201 on Stability and Control estimation methods.

Computational Fluid Dynamics for the Aerodynamic Design and Modeling of a Ram-Air Parachute with Bleed-Air Actuators

Abstract : The activation of spoilers on the upper surface of a wing is a relatively new method of achieving longitudinal and lateral control of ram-air canopies, often referred to as parafoils. No numerical studies, however, have fully investigated the 3-D aerodynamic performance of these bleed-air actuators. Simulation results are presented for a finite span, ram-air canopy geometry and several configurations amenable for comparison with wind tunnel experiments and flight test data. Assessments are also made between the resulting flowfields and previously reported 2-D computational fluid dynamics (CFD) results. Evaluations include cases with an asymmetrical trailing edge deflection to simulate a turn and the inclusion of two different bleed-air vents. Lift, drag and roll coefficients correlate well with wind tunnel and flight test results.

Numerical Study of Ram Air Airfoils and Upper Surface Bleed-Air Control

Abstract : Natick Soldier Research, Development, and Engineering Center has been leading a Modeling and Simulation effort to develop high fidelity simulations of ram-air parachute systems to complement the design and analysis of new and existing airdrop systems. In this paper an unsteady numerical study of two-dimensional, rigid, ram-air sections with an array of upper surface bleed-air actuators is presented. Aerodynamic forces and lift-to-drag ratios of a modified Clark-Y ram-air airfoil are calculated from unsteady Reynolds-Averaged Navier-Stokes (RANS) simulations, using the Kestrel and Cobalt flow solvers. The flow fields exhibit a complicated cavity flow coupling with the airfoil profile. Variations in the locations and number of bleed air actuators and trailing edge deflection yield time averaged L/D values between 1.24 and 59.14, and strongly support the utility of the bleed air actuators for use as an enhanced lateral/longitudinal control mechanism. Additionally, these initial results emphasize the requirements for prudent mesh generation and the performance of unsteady calculations for ram-air canopy simulations.

Grid Quality and Resolution Effects on the Aerodynamic Modeling of Parachute Canopies

This work provides an overview of the grid quality and resolution effects on the aerodynamic modeling of ram-air parachute canopies. The CFD simulations were performed using the Cobalt flow solver on two dimensional canopy sections with open and closed inlets. Previous simulation results of these geometries showed that grid independence is achieved for the closed and open airfoils with grids containing around half a million and two million cells, respectively. Previous grids were either hybrid with prismatic layers near the walls or multi-block structured using algebraic grid generators. The results presented in this work show that grid independence of both geometries can be achieved with much coarser grids. These grids, however, were generated with good smoothness, wall orthogonality, and skewness qualities. The results show that the grid quality value is mainly related to the grid smoothness and does not depend on the mesh skewness or the wall orthogonality. Although a smooth grid improves the quality value and therefore the solution convergence, it does not always lead to an accurate solution. For example, the unstructured meshes with anisotropic cells near the wall have very good grid quality, however they have the worst accuracy among all meshes considered because of the poor skewness at the walls. The results also showed that in comparison to the closed inlets, the open geometry solutions are less sensitive to the initial grid spacing and number of constant spacing layers at the outside airfoil walls. Finally, the open inlet solutions do not change with the inside airfoil mesh resolution and type.

Numerical Simulation of the F-16XL at Full-Scale Flight Test Conditions Using Delayed Detached-Eddy Simulation

The F-16XL flight test data has been used for many years in the CAWAPI program series as a challenging test case for numerical prediction of military aircraft at full-scale flight conditions. The complexity of the configuration and the variety of flight test conditions available challenge most computational fluid dynamics codes and turbulence models. This study documents the use of the Kestrel flow solver with its solution-based mesh refinement capability and Delayed Detached-Eddy Simulation turbulence models for the F-16XL Flight Condition 25, a subsonic, high angle-of-attack flight condition that has proved particularly challenging for computational researchers. Spectral content of pressure data and the integrated lift, drag and pitching moment coefficients is also presented. The more advanced turbulence models, coupled with the more refined meshes, show very good agreement with flight test data and reasonable resolution of the unsteady frequency content.

Indicial Methods for the Numerical Calculation of Dynamic Derivatives

This article considers the problem of estimating dynamic derivatives of a basic finned projectile using computational fluid dynamic simulations. The indicial responses with respect to a unit step c...

Collaborative Evaluation of CFD-to-ROM Dynamic Modeling

This work presents and discusses findings from a NATO STO collaborative research on the reduced order aerodynamic modeling of a NACA 0012 airfoil and a generic swept wing UCAV. The linear and nonlinear reduced order models are created based on the superposition integrals of the step response with the derivative of its corresponding input signal. Step responses are calculated using CFD and a grid motion approach that allows separating the effects of angle of attack and sideslip angle from angular rates. This approach was previously tested using Cobalt flow solver, however to demonstrate its generalization capability, four different flow solvers are used in this study: Cobalt and Kestrel at United States Air Force Academy (USAFA), USM3D at NASA Langley Research Center (LaRC), and ENSOLV at the Netherlands National Aerospace Laboratories (NLR). Step changes in the angle of attack and pitch rate are obtained using these CFD codes. For the UCAV configuration, the lateral step responses to sideslip angle, roll and yaw rates are also calculated. The step predictions of the codes are compared with each other. Aerodynamic models are then created from these step responses and are used to predict responses to arbitrary motions (inputs). The model predictions are compared with CFD (full-order) and available experiments. The results demonstrate that step functions can be easily calculated by CFD codes. Overall, the angle-of-attack and pitch rate responses are very similar for each solver particulary at small angles of attack. Discrepancies at higher angles are probably due to differences in grids and solver numerical algorithms. The step responses show an initial jump as the grid begins to move. The initial jumps become smaller with increasing Mach number. All responses will then asymptotically reach a steady-state value. The results show that significantly fewer time steps are required to reach the steady-state solutions for the UCAV geometry than two-dimensional airfoil. Finally, the model predictions match the CFD data of different motions, all generated within the range of data used for model generation, very well.

Dynamic Store Release of Ice Models from a Cavity into Mach 2.9 Flow

An investigation was conducted into store separation from a cavity in a Mach 2.94 freestream using both experimental and computational methods. Both approaches used an open cavity with a length-to-depth ratio of 4.5, and for the sake of simplicity, release of a spherical model was analyzed. The experimental process used a piezoresistive pressure transducer to collect the time-varying content of the pressure signal, while schlieren visualization and high-speed photography capture the dynamic response of a store released from the cavity. Computationally, the OVERFLOW solver was applied with higher-order numerical methods, Chimera grids, and the delayed detached-eddy simulation/shear-stress transport hybrid turbulence model. Tests were performed in a blowdown tunnel exhausting to a vacuum, which enabled robust control of the total pressure, and computational conditions were selected to match the experiment. The studies demonstrated that the shock wave formed on the bottom surface of the sphere led to the los...