Engineering topics
Pandya, Shishir
Publications and source records attributed to Pandya, Shishir.
Skylon Aerodynamics and SABRE Plumes
An independent partial assessment is provided of the technical viability of the Skylon aerospace plane concept, developed by Reaction Engines Limited (REL). The objectives are to verify REL's engineering estimates of airframe aerodynamics during powered flight and to assess the impact of Synergetic Air-Breathing Rocket Engine (SABRE) plumes on the aft fuselage. Pressure lift and drag coefficients derived from simulations conducted with Euler equations for unpowered flight compare very well with those REL computed with engineering methods. The REL coefficients for powered flight are increasingly less acceptable as the freestream Mach number is increased beyond 8.5, because the engineering estimates did not account for the increasing favorable (in terms of drag and lift coefficients) effect of underexpanded rocket engine plumes on the aft fuselage. At Mach numbers greater than 8.5, the thermal environment around the aft fuselage is a known unknown−a potential design and/or performance risk issue. The adverse effects of shock waves on the aft fuselage and plumeinduced flow separation are other potential risks. The development of an operational reusable launcher from the Skylon concept necessitates the judicious use of a combination of engineering methods, advanced methods based on required physics or analytical fidelity, test data, and independent assessments.
High Fidelity Simulation of a Computer Room
This viewgraph presentation reviews NASA's Columbia supercomputer and the mesh technology used to test the adequacy of the fluid and cooling of a computer room. A technical description of the Columbia supercomputer is also presented along with its performance capability.
Automated CFD Parameter Studies on Distributed Parallel Computers
This paper presents viewgraphs of an automated process of running CFD parameter studies on distributed parallel computers.
High Energy Boundary Conditions for a Cartesian Mesh Euler Solver
Inlets and exhaust nozzles are common place in the world of flight. Yet, many aerodynamic simulation packages do not provide a method of modelling such high energy boundaries in the flow field. For the purposes of aerodynamic simulation, inlets and exhausts are often fared over and it is assumed that the flow differences resulting from this assumption are minimal. While this is an adequate assumption for the prediction of lift, the lack of a plume behind the aircraft creates an evacuated base region thus effecting both drag and pitching moment values. In addition, the flow in the base region is often mis-predicted resulting in incorrect base drag. In order to accurately predict these quantities, a method for specifying inlet and exhaust conditions needs to be available in aerodynamic simulation packages. A method for a first approximation of a plume without accounting for chemical reactions is added to the Cartesian mesh based aerodynamic simulation package CART3D. The method consists of 3 steps. In the first step, a components approach where each triangle is assigned a component number is used. Here, a method for marking the inlet or exhaust plane triangles as separate components is discussed. In step two, the flow solver is modified to accept a reference state for the components marked inlet or exhaust. In the third step, the flow solver uses these separated components and the reference state to compute the correct flow condition at that triangle. The present method is implemented in the CART3D package which consists of a set of tools for generating a Cartesian volume mesh from a set of component triangulations. The Euler equations are solved on the resulting unstructured Cartesian mesh. The present methods is implemented in this package and its usefulness is demonstrated with two validation cases. A generic missile body is also presented to show the usefulness of the method on a real world geometry.
Parametric Powered-Lift Navier-Stokes Computations
The goal of this work is to enable the computation of large numbers of unsteady high-fidelity flow simulations for a YAV-8B Harrier aircraft in ground effect by improving the solution process and taking advantage of NASA parallel supercomputers. The YAV-8B Harrier aircraft can take off and land vertically, or utilize short runways by directing its four exhaust nozzles toward the ground. Transition to forward flight is achieved by rotating these nozzles into a horizontal position.
Parametric Time-Dependent Navier-Stokes Computations for a YAV-8B Harrier in Ground Effect
The Harrier Jump Jet has the distinction of being the only powered-lift aircraft in the free world to achieve operational status and to have flown in combat. This V/STOL aircraft can take-off and land vertically or utilize very short runways by directing its four exhaust nozzles towards the ground. Transition to forward flight is achieved by rotating these nozzles into a horizontal position. Powered-lift vehicles have certain advantages over conventional strike fighters. Their V/STOL capabilities allow for safer carrier operations, smaller carrier size, and quick reaction time for troop support. Moreover, they are not dependent on vulnerable land-based runways. The AV-8A Harrier first entered service in the British Royal Air Force (RAF) during 1969, and the U.S. Marine Corps (USMC) in 1971. The AV-8B was a redesign to achieve improved payload capacity, range, and accuracy. This modified design first entered service with the USMC and RAF in 1985. The success and unique capabilities of the Harrier has prompted the design of a powered-lift version of the Joint Strike Fighter (JSF). The flowfield for the Harrier near the ground during low-speed or hover flight operations is very complex and time-dependent. A sketch of this flowfield is shown. Warm air from the fan is exhausted from the front nozzles, while a hot air/fuel mixture from the engine is exhausted from the rear nozzles. These jets strike the ground and move out radially forming a ground jet-flow. The ambient freestream, due to low-speed forward flight or - headwind during hover, opposes the jet-flow. This interaction causes the flow to separate and form a ground vortex. The multiple jets also interact with each other near the ground and form an upwash or jet fountain, which strikes the underside of the fuselage. If the aircraft is sufficiently close to the ground, the inlet can ingest ground debris and hot gases from the fountain and ground vortex. This Hot Gas Ingestion (HGI) can cause a sudden loss of thrust (powered lift), and the vehicle may crash. The high-speed jet flow along the ground can also entrain the ambient flow, resulting in a low pressure region underneath the vehicle. The accompanied loss of lift is referred to as the suckdown effect. The ground vortex may also be highly unsteady, dramatically changing its size and position with time at low frequencies, e.g., 1 Hz.
Parametric Study of a YAV-8B Harrier in Ground Effect using Time-Dependent Navier-Stokes Computations
A process is described which enables the generation of 35 time-dependent viscous solutions for a YAV-8B Harrier in ground effect in one week. Overset grids are used to model the complex geometry of the Harrier aircraft and the interaction of its jets with the ground plane and low-speed ambient flow. The time required to complete this parametric study is drastically reduced through the use of process automation, modern computational platforms, and parallel computing. Moreover, a dual-time-stepping algorithm is described which improves solution robustness. Unsteady flow visualization and a frequency domain analysis are also used to identify and correlated key flow structures with the time variation of lift.
Automated CFD Parameter Studies on Distributed Parallel Computers
The objective of the current work is to build a prototype software system which will automated the process of running CFD jobs on Information Power Grid (IPG) resources. This system should remove the need for user monitoring and intervention of every single CFD job. It should enable the use of many different computers to populate a massive run matrix in the shortest time possible. Such a software system has been developed, and is known as the AeroDB script system. The approach taken for the development of AeroDB was to build several discrete modules. These include a database, a job-launcher module, a run-manager module to monitor each individual job, and a web-based user portal for monitoring of the progress of the parameter study. The details of the design of AeroDB are presented in the following section. The following section provides the results of a parameter study which was performed using AeroDB for the analysis of a reusable launch vehicle (RLV). The paper concludes with a section on the lessons learned in this effort, and ideas for future work in this area.
Unsteady Computations of a Jet in a Crossflow with Ground Effect
A numerical study of a jet in crossflow with ground effect is conducted using OVERFLOW with dual time-stepping and low Mach number preconditioning. The results of the numerical study are compared to an experiment to show that the numerical methods are capable of capturing the dominant features of the flow field as well as the unsteadiness associated with the ground vortex.
Automation of a N-S S and C Database Generation for the Harrier in Ground Effect
A method of automating the generation of a time-dependent, Navier-Stokes static stability and control database for the Harrier aircraft in ground effect is outlined. Reusable, lightweight components arc described which allow different facets of the computational fluid dynamic simulation process to utilize a consistent interface to a remote database. These components also allow changes and customizations to easily be facilitated into the solution process to enhance performance, without relying upon third-party support. An analysis of the multi-level parallel solver OVERFLOW-MLP is presented, and the results indicate that it is feasible to utilize large numbers of processors (= 100) even with a grid system with relatively small number of cells (= 10(exp 6)). A more detailed discussion of the simulation process, as well as refined data for the scaling of the OVERFLOW-MLP flow solver will be included in the full paper.