Integrated Test and Evaluation for the 21 st Century *

Transcription

1 USAF Developmental and Evaluation Summit November 2004, Woodland Hills, California AIAA Integrated and Evaluation for the 21 st Century * Marcus L. Skelley U.S. Air Force, Arnold Engineering Development Center, Arnold AFB, TN Tommie F. Langham and William L. Peters ** Aerospace ing Alliance, Arnold Engineering Development Center, Arnold AFB, TN Aerospace ground testing has evolved over the past century. Large- and small-scale tunnels were in the mainstream for most of the 20 th century, utilized usually for different aspects of the technology and flight systems development process. High-performance computers have now enhanced flight and ground testing, thus adding the term computational to the already existing field of fluid dynamics. Although, modeling and simulation (M&S) have played important roles in the flight systems development process for a long time, the process has been further enabled by the explosive growth in computational power and speed now available to the analyst. Recent initiatives, such as Simulate,, and Evaluate Process (STEP), seek to optimize the combination of flight test, ground test, and M&S usage to streamline the test and evaluation process, reduce risk, and minimize redundancy. The target is a triple partnership, allowing for a smaller combined effort that will result in equal effectiveness. The streamlined test and evaluation process, in parallel with explosive computer processing growth, predicts the tightly woven triple ring of flight test, ground test, and M&S illustrated below. The challenge of the future is to bring about this synergistic trio of development capabilities to address the developmental needs of new aerospace systems. Capability GT Ground Modeling & Simulation Time GRD M&S GRD TEST FLT M&S Although it has been suggested that wind tunnels might become obsolete in the coming decades, this may not happen. Smaller tunnels may be more heavily utilized for validation of analytical and computational tool sets, while larger tunnels are dedicated to limited demonstration activities designed to minimize risks of early transition to flight development and operational testing. Both ground testing and M&S will generate an ever-increasing volume of data, which in time will require a substantial increase in data management capability. The * The research reported herein was performed by the Arnold Engineering Development Center (AEDC), Air Force Materiel Command. Work and analysis for this research were performed by personnel of Aerospace ing Alliance, the operations, maintenance, information management, and support contractor for AEDC. Further reproduction is authorized to satisfy needs of the U. S. Government. Senior Project Manager, Systems Division, Arnold Engineering Development Center, Arnold AFB TN Technical Fellow, Integrated and Evaluation Department, Arnold Engineering Development Center, Arnold AFB TN, AIAA Associate Fellow ** Technical Fellow, Integrated and Evaluation Department, Arnold Engineering Development Center, Arnold AFB TN, AIAA Senior Member 1 This material is declared a work of the U.S. Government and is not subject to copyright protection in the United States.

2 increased volume of data is complicated by encryption requirements for high-speed data transfer and constantly evolving software standards and formats. Greater competition, fewer Department of Defense (DoD) programs, smaller budgets, and greater computational capabilities require a proactive and flexible strategy. With the downsizing of defense programs, shrinking budgets, and the concerns of the Office of Secretary of Defense (OSD) regarding redundant testing facilities, it is essential to optimize future investments. These environmental factors require today s test and evaluation (T&E) professionals to plot a course well into the next century. The Vision for AEDC in the 21st Century is based on the following: * National defense planning guidance * Extrapolation of the state of technology * Estimates of national and international facility capabilities * And, most importantly, customer needs This paper describes: 1) the current state and planned improvements of TT&E at the Arnold Engineering Development Center (AEDC) and 2) a vision for test and evaluation across the T&E centers within the development test and evaluation (DT&E) weapon system program development cycle. Nomenclature 6-DOF = Six degree-of-freedom AAC = Air Armament Center AMDMS = Armament Munitions Digital Modeling and Simulation AEDC = Arnold Engineering Development Center AFB = Air Force Base AFFTC = Air Force Center AFSEO = Air Force Seek Eagle Office AIDACS = Advanced Instrumentation Data and Control System CFD = Computational fluid dynamics CLRANC = Clearance distance calculation DoD = Department of Defense DT&E = Developmental test andevaluation EMD = Engineering, Manufacturing and Development FLIP TGP = Flow-Field Load Influence Prediction Trajectory Generation Program HPC = High-performance computing IT&E = Integrated test and evaluation JASSM = Joint Air-to-Surface Standoff Missile JDAM = Joint Direct Attack Munition JSF = Joint Strike Fighter J-UCAS = Joint Unmanned Combat Air System M&S = Modeling and simulation MASTER = Modeling and simulation test and evaluation resources MDA = Missile distributed airloads OFP = Operational Program OSD = Office of the Secretary of Defense P-LOCAAS = Powered Low-Cost Autonomous Attack System PSP = Pressure-sensitive paint PWT = Propulsion Wind Tunnel Facility SDD = System design and development SPO = System Program Office 2

3 STEP = Simulate,, Evaluate Process T&E = and evaluation TVIS = Trajectory visualization VKF = Von Karman Facility W I. Introduction ith the present emphasis on streamlining acquisition of weapon systems, ground and flight test centers must formulate an improved integrated test and evaluation (IT&E) approach to support systems development. The Arnold Engineering Development Center (AEDC) has aggressively accepted the challenge and is continuously evolving and developing an IT&E approach to support aerospace system development efforts. IT&E is the integration of modeling tools, including computations and engineering methods, in direct support of ground and flight tests of aircraft and aircraft systems. Integrating the modeling tools directly with ground and flight tests enables the design of a better test program with validation and/or extrapolation of the results and assists in the planning and making of decisions needed for a more efficient or cost-effective test. AEDC uses the IT&E concept to couple the complementary strengths of ground testing and analysis to resolve issues associated with the development of aerospace systems. Computational fluid dynamics (CFD), engineering methods codes, and specific analysis processes are tools that formalize the analytical techniques that go into the IT&E approach. The IT&E concept is in keeping with the Department of Defense (DoD)- advocated Simulation,, and Evaluate Process (STEP). IT&E takes on a more important role when used to integrate the subsystems such as the airframe, propulsion system, weapons, and avionics that make up the flight vehicle. Early IT&E concepts had limited acceptance in the ground testing and flight testing environments because of the lack of fidelity and confidence in the computations coupled with the long lead times required to accomplish a CFD solution or set of solutions. Prior to around 1990 most store separation predictions were derived from wind tunnel testing using a quasi-steady captive trajectory system and/or grid matrix approach. The details of this process are presented in a following section. One of the earliest examples of development and application of the IT&E concept at the AEDC involved the development of a synergistic combination of experimental, CFD, and engineering methods to store-separation simulations for the F-15E by K. S. Keen. This example was presented at the 28 th Aerospace Sciences Meeting in an AIAA paper entitled A New Approach to Computational Aircraft/Store Weapons Integration. 1 This implementation of computations into the testing environment was coined as IT&E in the ground testing environment by Dr. E. M. Kraft, now the Technical Advisor to the AEDC Base Commander. In the fourteen years since that time AEDC has developed a set of reusable software for the production simulation of internal and external flow-field definitions, propulsion systems, and weapons separation from arbitrary aircraft. These tools have matured to the point that they are being used to define and optimize the required weapon separation test matrices, validate the test results, and provide near real-time assessment of the results. Early on, the assumption was made that the interference of the aircraft on the store could be separated from the free-stream aerodynamics of the store. Therefore, in the simulation of a trajectory, the aerodynamics of a store in the interference flow field of the aircraft could be determined by a linear combination of the free-stream aerodynamics of the store and the interference of the aircraft on the store (grid matrix). This means of representing the interference of the aircraft on the store has changed in relation to the type of data that are available. Throughout the development of the weapon trajectory simulation codes, the basic weapon trajectory prediction module of the codes has remained the same, while the inputs or the form of the inputs to represent the forces and moments acting on the store have continued to evolve and change. AEDC s unique mixture of personnel, computational tools, and facilities has been used to place the Center at the forefront of aerodynamic performance testing and related modeling and simulation. The need exists today and is projected to exist tomorrow to meet the nation s continuing ground test and evaluation requirements. The emphasis has been on developing and applying reusable computational tools that provide robust, timely results in the pursuit of an integrated test and analysis capability. Over the last two decades, a number of computational tools have been developed or enhanced for application to air vehicle external and inlet flow phenomena and weapons separation testing and evaluation programs. Initially, the tools were developed to complement weapons separation testing as a way of conducting parametric assessment with sensitivity to atmospheric and/or weapon manufacturing tolerances. and Evaluation plans (ground simulation activities) at AEDC have been developed and executed in support of many developing weapons systems. Specific systems for which IT&E plans have been developed and executed or currently are being executed include the F-22 weapons separation program, Joint Strike Fighter (JSF) aircraft store loads, inlet and lift fan flow performance, weapons separation programs, and the Joint Unmanned Combat Air Sys- 3

4 tem (J-UCAS) weapons separation program. AEDC has also supported the weapons separation requirements for a number of new and/or improved missile systems: AIM-120, Joint Direct Attack Munition (JDAM), AGM-130, AIM- 9X, Small Diameter Bomb, Powered Low Cost Autonomous Attack System (P-LOCAAS), and (Joint Air-to-Surface Standoff Missile (JASSM). In addition, AEDC works on a number of efforts for the integration of developing weapons on legacy aircraft. AEDC has worked jointly with the Air Force Seek Eagle Office (AFSEO) toward utilization of common processes and tools for the weapons separation and clearance process. The primary objectives of IT&E are risk mitigation, optimizing concurrent planning, testing, computational applications, and system modeling during the development and validation of the ground-based simulation schedules. Direct access and utilization of the high-performance computing (HPC) network provides additional and necessary computational resources in applying computational tools in order to understand more about the cause and effects relationship during the aerodynamic/inlet performance and weapon separation process. Implementation of risk mitigation activities improved the potential for adhering to both cost and timetable constraints. Utilization of the resident corporate knowledge at AEDC has provided: a) valuable insight into airframe development issues gained on externally similar vehicles; and b) insight and knowledge of the carrier aircraft s potential effects on the motion of the respective separation vehicle as it translates through its flow field. A. Wind Tunnel Overview The current weapons system development cycle depicted in Fig. 1 involves a sequential but transitional set of development processes that move from aerodynamic predictions, wind tunnel testing (or ground testing), and airframe design and manufacturing to final flight testing. The wind tunnel testing is part of both the concept development phase and the first part of the system design and development phase (formerly the engineering, manufacturing, and development phase). AEDC utilizes a number of small and large wind tunnels to assist in acquiring test information to support aircraft, missiles, engines, and other flight system components. The PWT and VKF facilities consist of the Propulsion II. The Current State of Ground ing Integrated and Analysis Concept Demo Aerodynamics Predictions Wind Tunnel ing Predict - Simulate - Build - Wind Tunnel 16T and 16S, the Aerodynamic Wind Tunnel 4T, and Supersonic/Hypersonic Tunnels A, B, and C. General descriptors for the tunnels are shown in Table 1. The wind tunnels are closed-circuit, continuous-flow, variabledensity units. They are operated in a production-oriented environment designed to provide high throughput and productivity, although research-oriented tests have been accomplished on occasion. A wide variety of test, evaluation, and analysis capabilities are required to support AEDC s broad mission objectives. Table 1. Wind Tunnel Descriptions - Airframe Design SDD Manufacturing Improved Aero Modeling and Simulations for EMD ing Figure 1. Current Weapons System Development Cycle Tunnel Mach Number Reynolds Number 10-06,/ft Section Size 16T ft by 16 ft 16S ft by 16 ft 4T ft by 4 ft A in. by 40 in. B 6, in. by 50 in. C 4,8, in. by 50 in. Tunnels 16T and 16S are intended primarily for aerodynamic testing of large-scale aircraft or missiles and fullscale forebody/inlet/engine integration. Tunnel 4T is used predominantly for the store separation and aerodynamic testing of smaller-scale models. Tunnels A, B, and C support a number of the aforementioned types of tests at small scale but also can be used for thermal and materials testing. All tunnels can be used for sting-mounted force and pressure tests of aerodynamic models to support performance or dynamic stability testing. Both the 16T and 16S units can 4

5 provide test environments for evaluation of propulsion integration effects of proposed vehicle designs. This can be done by drawing air into inlets and by expelling gases out the exhaust of a model (either cold flow or actual burning/ expanding gases). B. Force Accounting Methodology Current wind tunnel T&E techniques used for aircraft involve a process of integrating test and analysis information from the wind tunnel, empirical modeling techniques, and CFD. These data are integrated by the weapons system manufacturer on the basis of a vehicle force and moment accounting methodology, a sample of which is depicted in Fig. 2. The wind tunnel test objectives include determining basic aircraft aerodynamic stability and control characteristics, including the effects of vehicle attitude, control surface deflection, and propulsion system integration effects on aerodynamic performance. ing techniques employed include: empirical sizing and placement of boundary-layer transition strips on leading edges; model support system interference studies done by using incremental tests; exhaust jet simulation using high-pressure air for nozzle exhaust; and other engine bay effects. The importance of a good force-accounting system is that it permits consideration of every single force and moment that acts on the air vehicle once and only once. With the advent of computer systems and networks, the current practice is to develop an electronic ground test database of wind tunnel information from various wind tunnel tests and facilities; this database is used to analyze total aircraft performance prior to flight for vehicles such as the JSF. Wind tunnel test configurations are developed in rapid fashion based upon real-time analysis of wind tunnel data to permit optimization of the aircraft configuration. Without a good force-accounting system, the vehicle performance prediction is subject to error and speculation, and the advantage of the wind tunnel to support flight vehicle development is hampered. 2 The wind tunnel tests that feed the ground test database come from a variety of test models and corresponding objectives, as indicated in Fig. 3. Six major types of tests, with corresponding objec- Sting & Distortion Model Throttle Protuberance Roughness & RN Adjustments Dummy Sting Independent & Surface Propulsion Drags: Wind Tunnel To Operation Quality Eng. Bay Vent To Full Scale Altitude Adjustment Undistorted Distorted LEX Spoiler altitude f(m, Alt) to Full Scale Bleed f(m CDmin CDmin Trimmed Force & Moment Sting & Distortion s C L, C D = f(m, α ) Aero Model Trimmed C Reference Trimmed L,C D Aerodynamic Reference Drag Polars Database Conditions C L,C D =f(m, α ) C L,C D f(m, α,alt,cg) Critical Inlet CG Effects Spill Drag α, C D Store Drag f(m, C L ) Inlet Drag f(m) f(m, C L,CG) Predicted Aircraft Performance Subcritical Power Compressor Inlet Ramp Bleed Inlet Spill Drag Extraction Bleed & ECS Heat f(m, MFR) Exchanger Drag Engine Cycle Deck Net with Derivatives Net Propulsive for Recovery, Bleed, Thrust Force and HPX FNP Jet Effects Model f(m, Alt, PLA) Inlet/Airframe Model Nozzle Drag & Jet Effects Inlet α = 0 0 f(m, PLA) C L, C D = f (M, PLA) Figure 2. Sample Manufacturer s Aircraft Force Accounting System Figure 3. Aircraft Wind Tunnel Types 5

6 tives, are typically conducted for development of aircraft whose mission is primarily in the subsonic, transonic, or supersonic Mach number range. 3 Each of the wind tunnel models is designed and fabricated on the basis of the individual test objectives, the required output test information, and the required data uncertainty. These factors function as inputs to the design of the model concept, the support system in the wind tunnel, the size of the wind tunnel required, and instrumentation measurements required. Design and fabrication of the wind tunnel model may take several months and often include subcomponent tests to ensure that the model will provide the appropriate level of data quality. At AEDC, the practice of forming an integrated project team that includes the customer aircraft development team ensures that the wind tunnel test facilities are prepared to effectively support the planned testing. This includes ensuring that the model system within the test facilities will provide correct test conditions as well as adequate interfaces of hardware systems, instrumentation, controls, and data reduction. The team will determine the most efficient method of conducting the test program with the objectives of reducing the duration of the test to minimize cost and schedule while ensuring that the technical objectives are met. C. Current Wind Process Quality, Cost and Data Acquisition Rates 1. Data Quality The quality of results from the AEDC wind tunnels has improved through the years as better understanding of issues such as support system interference, wind tunnel flow quality and turbulence, and effects such as Reynolds number and boundary layers have been achieved. Currently, for transonic testing in Tunnels 16T and 4T, an inventory of new force measurement balances, installed electronically scanned pressure systems, and new component data acquisition systems help to provide the highest quality of test data. Instrumentation currently used for direct measurement of aerodynamic performance information for fighter aircraft configurations include modern six-component force and moment balances of either shell or single-piece design that have been developed to minimize the effects of temperature. Electronically scanned pressure transducers that compensate for temperature effects are used to measure wind tunnel surface and duct pressures. They are also used in numerical pressure times area force-and-moment integrations and in assisting engineers in characterizing the aerodynamic flow. Many flow visualization tools are being employed to help characterize details of aerodynamic flow, including oil flow, flow tufts, and, sporadically, new test techniques such as pressure-sensitive paint (PSP). An example of historical data quality improvement is depicted in Fig. 4 for afterbody tests from the early 70s until the present Year Figure 4. Data Uncertainty for Freestream Mach Number and Afterbody Drag in Tunnel 16T 2. Data Costs 300 The cost of wind tunnel information has diminished through the years as test productivity All tests except CTS and Grid Adjusted for Inflation 250 improvements have been pursued and developed as indicated in Fig. 5. One of the latest improvements then have reduced costs is the use of com- 200 puter systems on the moveable wind tunnel test 150 sections; this innovation permits investigators to check out a test article without disconnecting and 100 reconnecting leads when the wind tunnel test begins. One of the newest techniques enabling 50 faster acquisition of data is the use of continuous sweep. Conventional aerodynamic testing for 0 high-quality test information typically has required the test article to move and then pause to FISCAL YEAR acquire test data at a given attitude. New techniques in data acquisition and networking now Continuous Sweep Figure 5. Data Point Cost Trends for 16T Excluding permit test data from force balances, pressures, and model attitude to be completely synchronized to acquire test data while the test article continues to pitch and roll. $'s/atp Mach or CD x 10 C1 MONITOR PITCH SPEED ESP X-DUCERS COMPOSITE CREW DIFFUSER MOD TVA CONTRACT HAAS CTS PCC HAAS CONTROL H.P. AIR VAX-AMAPS VAX-CTS PRETEST CKOUT PITCH BOOM CONTINUOUS PITCH COMPUTER COMMO TVA C0NTRACT Mach No. CD*10 Poly. (Mach No.) Poly. (CD*10) SUSTAINMENT UPGRADES 6

7 This technique is very significant as it can easily multiply the number of data points or data acquisitions per one degree of change in pitch angle by a factor of ten. 3. Data Acquisition Rates The rate of collection of test data varies by test type in Tunnel 16T, but several improvements have reduced the total time required to prepare and complete a wind tunnel test program. Preparation times for standard force and moment tests have been somewhat reduced by use of e-business processes for collecting and implementing test customer requirements, but the most substantial improvements have been in the reduced data acquisition times. This is a result of automated tunnel and model-positioning systems, reflective memory networks, and on-cart data acquisition systems, as indicated in Fig. 6. The polars per operating hour of 16T have nearly doubled in the last 20 years. OP/OSH All tests except CTS and Grid C1 MONITOR PITCH SPEED ESP X-DUCERS COMPOSITE CREW DIFFUSER MOD TVA CONTRACT CTS PCC HAAS CONTROL H.P. AIR VAX-AMAPS VAX-CTS PRETEST CKOUT Fiscal Year Figure 6. Polars per Operating Hour Trend in Tunnel 16T III. The Current State of Evaluation A. Computational Overview The general technical areas of propulsion and aeromechanics are the two most important areas in which evaluations are undertaken at AEDC because of the sheer volume of work that takes place in those areas. Under the general umbrella of aeromechanics, store integration, aerodynamic performance, inlet performance, and jet effects have become intensive areas of growth at AEDC. Over the past two decades aircraft store integration has become the most intensive area of growth for computations and evaluation. This has come about primarily through the marrying of the wind tunnel with modeling and simulation (M&S) using the IT&E methodology. AEDC has developed an arsenal of M&S tools for application to the aeromechanics area. This IT&E concept was not initially accepted in the ground testing and flight testing environments because of lack of confidence in the computations and because of the long lead times required to accomplish a CFD solution or set of solutions. Based on the demonstrated accuracy of the Trajectory Generation Prediction code, an overall methodology for integrating CFD and analytical methods into the store clearance process was established. 1 In general, every application of CFD as part of the M&S application in both ground test and flight test must include a thorough verification or benchmarking process to ensure that the computational simulations accurately represent the conceptual model. This verification process establishes the investigators confidence in utilizing these tools for extrapolation outside of the existing reference database. Other examples exist where computational tools have been proposed and applied in support of basic aerodynamic configuration assessment during wind tunnel testing. The development of a successful aircraft or weapon system requires the assurance of airframe-propulsion system compatibility early in the development cycle to prevent costly engine instabilities or performance deficiencies in fielded systems. Historically, wind tunnel tests and turbine engine tests have been employed to evaluate inlet-engine compatibility during the design and development cycle. Wind tunnel tests predicted the flow conditions that the engine would receive from the external airframe and inlet system combination when subjected to the flight environment. Turbine engine tests then helped predict the engine response to the flight environment, including the effects of pressure distortions. Screens with porosity distributed to match specific flow distortion patterns provided the coupling between the wind tunnel test and the engine test in this lengthy process. As the separate wind tunnel and turbine engine tests did not account for all of the interactions between the airframe, inlet, and engine, the final assessment did not arrive until the vehicle development had progressed to the flight test stage. Computational tools are under development to bridge the gap between the inlet testing and engine testing. Computational support will provide understanding of the flow field, as well as of the cause and effect relationship between inlet and engine. A wide variety of aerodynamic flow characteristics have been simulated at AEDC. Fluid dynamic flow phenomena have been simulated at subsonic, transonic, supersonic, and hypersonic conditions. Calculations for viscous, inviscid, laminar, turbulent, steady, chemically reacting, and time-varying flow fields have been performed. PITCH BOOM CONTINUOUS PITCH COMPUTER COMMO TVA C0NTRACT 7

8 B. Efficiencies in the Computational Tools and Hardware Extremely complex geometries and a large variety of flow phenomena are routinely modeled with CFD tools. The chimera domain decomposition methodology (structured flow solvers) used at AEDC allows configurations to be divided into smaller, more easily modeled components. After all the flight vehicles components are created, they are computationally reassembled to obtain the flow-field solution. Flow solver algorithms are continually upgraded, and new computer programs are constructed as required to meet customer needs. 4 Visualization and animation software is available which can present the massive amounts of flow-field information (data) in a more readily understood fashion. This capability was realized with the advent of the High Performance Computing Modernization Office in The computational throughput improvements have been very significant, as shown in Fig. 7. A paper presented at the 18 th Aerospace Ground ing Conference in June examined the development of the technologies combined in the IT&E approach to weapons integration at the AEDC. In 1995, AEDC was chosen by the Department of Defense to be one of 12 HPC centers in the U.S. As an HPC center, AEDC was identified as a distributed resource center with state-of-the-art computational hardware and a high-speed data link to the Defense Figure 7. Improvements in Computational Throughput Research and Engineering Network. In 1995, AEDC's computing resources were equivalent to 30 gigaflop years with approximately 100 gigabytes of storage. At that time, AEDC had eight CPUs with a total of 8 gigabytes of memory. These resources have increased significantly over the past nine years. Now, AEDC s available computing resources are equivalent to 781 gigaflop years, 5.6 terabytes of storage, and 192 CPUs with 192 gigabytes of memory. The storage space is used to archive CFD solutions and test data from the various AEDC test units. The increase in available resources has ensured that the throughput and storage capacity meet the requirements of AEDC's customers for more effective program decision making and risk management. AEDC s computational resources have continued to reflect the industry quantum leap in speed and throughput. With the computational throughput resulting from the infusion of HPC supercomputers and remote access to large HPC computing centers, AEDC has been able to apply the IT&E concept to many weapon systems. The cooperation of AEDC customers and, in many cases, the collaborative work being conducted within their organizations has contributed to the development of this concept. During development testing for the JDAM weapon system, certain modifications were made and evaluated for the attachment of model hardware to the wind tunnel sting support system. The use of M&S tools to assess and determine the impact of wind tunnel support hardware, representative results of which are shown in Fig. 8, was demonstrated and documented in Forces Blue Boxes - Software Enhancements Yellow Boxes - Hardware Upgrades CRAY XMP 0.9 gflop AEDC HPC DC 32 gflop AEDC HPC DC 112 gflop XAIR ADI Flow Solver AEDC HPC DC 264 gflop Unfactored Algorithm *AFOSR) Wall Functions (AEDC) MPI Parrallel (PET) Moments Multigrid (AEDC) AEDC HPC DC 781 gflop DiRTlib (PET) In Memory (PET) Altered Afterbody, Large Sting, Area- Matched Fins M =1.4 Altered Afterbody, Large Sting, Area - Matched Fins M =1.4 a. Force Coefficient Computations b. Moment Coefficient Computation Figure 8. Comparison of Coefficient Compuations vs Ground Results 8

9 Ref. 6. The results of the AEDC computational study of afterbody effects for the JDAM were presented at the 16 th AIAA Applied Aerodynamics Conference in June The majority of IT&E applications have occurred within the store separation area. Specific applications of the IT&E process and the complexity of these applications are demonstrated in Fig. 9. Complexity Ratio (Compared to Euler F-15E) 1.00E E E E E+00 F-16 Viscous grid pts. Turnaround = 5 days F-18 Viscous grid pts. B-2 Turnaround = 2 weeks Euler grid pts. Turnaround = 3 weeks F-15E Euler grid pts. Turnaround = 4 weeks Year Support for flight and ground testing is a primary application of M&S at AEDC. Corrections for wind tunnel model support interference, wall interference, corrections for geometric deviations of the scaled model from the fullscale article, and corrections for imperfections in optical hypersonic flow measurements have all been obtained using M&S tools. Store separation analysis is performed using flow-field information provided by M&S, coupled with load-predictive tools based on engineering methods which model the store dynamics. M&S (viscous CFD) is also used to provide time-varying flow-field simulations, such as the simultaneous launch of multiple stores, the flow field about a maneuvering aircraft, and the flow in and around an aircraft weapons bay. Wind tunnel testing from the weapon carriage position in the bay may result in significant compromise of the separation effects. With the advent of increased use of aircraft with weapons bays, M&S tool applications are routinely used to support the determination of the store aerodynamic loads at the carriage position and in the flow field out to the point (referred to as the park position ) to which the support system can achieve with minimum interference. The F/A-22 store separation IT&E story is a demonstration of the combination of ground testing, M& S, and flight testing. AEDC worked with the F/ A-22 System Program Office (SPO) and Lockheed Martin to develop an IT&E plan for F/A-22 store separation of externally carried and released pylon and fuel tank combinations. The early validation of the simulation process during development came from the direct comparison of the simulations with wind tunnel data, as shown in Fig. 10. For the F/ A-22 validation process, experimental data from the wind tunnel were used as input to account for the freestream aerodynamics of the store as well as the flow-field effects from the aircraft, as shown in Fig. 11. The comparisons of the computations with wind tunnel data are provided in Refs. 7 and 8. AEDC Figure 9. HPC Impact on Development Systems Application CFD AIRCRAFT FLOWFIELD ARCHIVED FUEL TANK DATA REDUCED RISK UCAVATD UCAVATD F-33 Euler grid pts. Turnaround = 3 days CFD TUMBLING TANK VALIDATE CFD MODEL DISTRIBUTED LOADS UCAV Viscous grid pts. Turnaround = 3 days JSF Viscous grid pts. Turnaround = 3 days WASP Viscous Unsteady grid pts. Turnaround = 14 days Complexity = (Number of CPUs) *(Number of Grid Points) *(Number of Solutions/Week) Technical Risk Mitigation CFD A/C-TANK INTERFERENCE MODEL HEAVY SCALING L3 PRESSURE TEST FLIP TGP JETTISONS 4T CTS & FLOW-FIELD PROBE PRE-SCREEN JETTISONS 16T CTS JETTISON & FREESTREAM FINAL VALIDATION Figure 10. F/A-22 Ground Validation Process 4T CTS VALIDATE DISTRIBUTED LOADS INCREASED UNDERSTANDING 4T DYNAMIC DROP 9

10 Freestream Aerodynamics CFD and Traj Prog + Loads Prog Separation Predictions Flow Field + + CFD and Probe/Grid Correlated with W/T Results Figure 11. IT&E Store Separation Process Captive/Grid Trajectory and CFD Carriage Loads Interference W/T CTS/Grid added a virtual camera capability to the Trajectory Visualization (TVIS) software package used for ground simulation of flight-test store separation film data. 9 This capability allowed the weapons system manufacturer to assess all proposed store separation flight test matrices. Prior to the first store release, camera locations and aim angles, were optimized for best viewing. An AIAA paper, Validating Weapon Separation Prediction Software Using F/A-22 Missile Launch Results 10 has been developed for presentation at the USAF Development & Evaluation Summit, November 16-18, The validation of the simulation process against flight test data was a part of an Air Force Materiel Command-funded joint effort between the Air Force Center (AFFTC) and AEDC. The joint effort is the Modeling and Simulation & Evaluation Resources (MASTER) program, described in more detail in this paper. The Air Force Seek Eagle Office (AFSEO) is responsible for the store certification recommendation process for Air Force systems when the DT&E store separation work is completed. AEDC works in close partnership with AFSEO, which has the responsibility for the certification of aircraft/stores compatibility on particular aircraft as they become operational during production and deployment. AEDC s role in store integration spans concept definition studies through the development phase of aircraft and stores. Under the auspices of the AFSEO, AEDC provides ground test and analytical support to the AFSEO store separation and ballistic accuracy certification processes. C. Data Management Data management is a critical asset for maintaining and managing massive amounts of data, whether they be ground test or M&S-generated results. Currently the ground test and M&S, information are archived on the AEDC mass storage system for future use. A true databasing system is not presently a part of AEDC s requirement structure. Concurrent access to real-time test, M&S, and historically archived results are essential for AEDC customers to make informed assessments and decisions during ground testing processes. With this strong advocacy, AEDC s Systems Division drafted a technology statement of need for the development and/or improvement in the data mining, graphics display, and visualization software employed. 1. Storage The current mass storage devices can handle 5.6 terabytes of storage. The system monitors and maintains the integrity of the stored information. This storage space is used to archive CFD solutions and test data from the various test units at AEDC. As noted above, the increase in the available AEDC resources have ensured that the throughput and storage capacity meet the requirements to provide AEDC's customers more effective program decision making and risk management. 2. DATAMINE The requirement for expedient real-time mining and display of test data has evolved with the ever-increasing wind tunnel data acquisition rates. The concept of the DATAMINE software was identified as a requirement for AEDC Technology program in FY02 to improve customer satisfaction, to minimize errors introduced in the wind tunnel data acquisition process, and to provide mining tools to search and access historical test data in common data files. The DATAMINE software design was planned and executed along a three-year spiral development path. This concept provided the developers a path by which each spiral completion was applied in the testing environment to establish the usefulness of the tool and to provide guidance for the next spiral. This process was followed as the tool was introduced into each test type within the wind tunnels area. Concurrent with the enhancements for mining and display application, a data validation manager was put into place to ensure that accurate data were being acquired. This was accomplished with the incorporation of data mining interface/protocols to common and relevant expectation models for dynamic online comparison with wind tunnel data. Examples of expectation models include the engineering methods aerodynamic prediction codes, Missile Distributed Air-Loads and Aerodynamic Prediction 02 (MDA, AP02), NASA data characterization software, historical data, and aerodynamic models based on historical or predicted data. 10

11 The DATAMINE software, as incorporated in the flight systems test and valuation processes, is provided at the request of AEDC customers with their final data package. The customers implement an executable version of the AEDC mining and plotting software on their computer system to continue detailed analysis of the test results. 3. Data Visualization (TVIS) The Trajectory Visualization (TVIS) software has become a very popular tool for AEDC s store separation customers. Once set up and integrated with DATAMINE, the TVIS software provides an efficient and effective method for visualizing and comparing any trajectory, measured or predicted, from multiple data sets. The computer resources required to implement TVIS within DATAMINE are based upon the summation of every possible configuration combination rather than displaying specific configurations. Also, the manual setup of the different customer airplane/ weapon/loading configurations is too time intensive. With trajectory visualization now easily accomplished, the customers are increasingly interested in having access to a time history of the minimum distance between the weapon and the parent aircraft. A software tool, CLRANC, exists for accomplishing minimum distance calculations with TVIS-compatible geometry, but the process is not integrated with TVIS/DATAMINE. Minimum distance time history is not currently part of the data product. IV. The Vision of the Future As noted in the preceding sections, both ground testing and computational evaluation have grown and evolved over the past two decades, becoming not only more efficient and quicker, but also more effective. Along with flight testing, the ultimate goal of all developmental testing is to produce a complete picture of an optimized flight vehicle. In the past, each type computations, ground test, and flight test was one piece of the puzzle. Computations filled a certain niche, but they were too slow, and models were too unsophisticated to make accurate predictions. The wind tunnel and other ground test results played a role in the design efforts, but these were largely ignored once the prototype vehicle was built and flight testing began. Each type of testing in succession becomes an order of magnitude higher in cost. With computational power growing by leaps and bounds and improvements continuing in the instrumentation and processes of ground and flight testing, all three test methodologies have begun to leverage off the strengths of the others. Computational analysis reveals benign areas of the flight envelope that can be dropped from wind tunnel and flight tests. Ground tests point the way to areas of concern that need more detailed computer solutions or areas of the envelope that flight tests must investigate. testing provides a real world environment that can show where analytical and ground-based models may be based on inaccurate predictions and assumptions, thus leading to future improvements. This synergy, shown in Fig. 12, produces an iterative process that ultimately decreases the time, and thereby the cost, of the acquisition process. Figure 12. IT&E Diagram The following conclusions and summary were extracted from a paper by Mr. Fredrick Webster, USAF, 412 th Wing, AFFTC, at Edwards AFB. Many predictive models require updating with data derived from testing the actual article. The AFFTC has learned that the maximum benefit from using M&S in the T&E process requires that updating be accomplished as testing progresses. Model updating from test data requires a robust data acquisition and analysis capability. The ability to update models as testing progresses requires careful pre-test planning and proper provisioning. The provisioning should include consideration of proper data collection, instrumentation, test progression rate, and especially, the experience of the update team. 11 AEDC, in concert with AFFTC, is starting to reap the benefits of this IT&E approach. The MASTER program, mentionedpreviously, was begun in FY00 between the two test centers, and has furthered the goals and processes of integrating computer simulations, wind tunnel and engine ground testing, and flight testing. The program is centered on the F/A-22 development process. A multitude of data was produced at AEDC during the early development stages of the aircraft in the 1990s. Computational codes were used for modeling different aspects of the development, most 11

12 Free-Stream Aerodynamics CFD and + Traj Prog + Loads Prog Separation Predictions Flowfield CFD and Probe/Grid + Correlated with W/T Results Captive/Grid Trajectory a nd CFD Carriage Loads Interference W/T CTS/Grid notably the store separation characteristics of the aircraft. These codes were continually updated and validated against results from the wind tunnel. This approach saved the F/A-22 SPO over $8M in wind tunnel testing on points that were predicted by computational analysis. It also laid the groundwork for the coming flight test program. Now, through the MASTER program, the simulation process will be further validated by the flight test program at AFFTC. That validation will be directly applicable to the use of the codes by AFSEO after the aircraft becomes operational. In addition, a corollary benefit of the MASTER program is the closer working relationship between AEDC and AFFTC. By learning processes used by both test centers, investigators will achieve better understanding, which will result in better application of tools and, in the end, a better final product to the warfighter. The lessons of F/A-22 and the MASTER program are already being applied to the next generation stealth fighter, the F-35 JSF. A schematic of the MASTER process is shown in Fig.13. Integrated & Evaluation M&S ing AEDC Wind Tunnel/Analysis Model Accurately predict flight test performance Apply Predictions Matrix Actual On-site AEDC Analysis: Anomaly resolution Model correction M&S F-22 Store Separation ing Reduce Matrix $$/Risk Savings ing & Ground Data for Model Validation Ground ing Free-Stream Aerodynamics CFD and Traj Prog + Loads Prog Separation Predictions Figure 13. The MASTER Iterative Process Flowfield + + CFD and Probe/Grid Correlated with W/T Results Ground ing Captive/Grid Trajectory and CFD Carriage Loads Interference W/T CTS/Grid In partnership with the Air Armament Center (AAC) at Eglin AFB, AEDC has also been supporting work on the Armament Munitions Digital Modeling and Simulation (AMDMS) project since FY03. The purpose of this project is to produce and validate a set of digital models and simulations for munitions that are interoperable, portable, and easily correlated. With much of the modeling and simulation capabilities today parceled out among different organizations with different missions, it is paramount that the testing centers have standardization and commonality between software tools and hardware to provide the foundation of cooperation and partnership. This cross-coupled use of models should lead to reduced time throughout the testing process and, along with that, lower costs. Another exciting program under way at AEDC to further the goals of IT&E is the Advanced Instrumentation Data and Control System program, or AIDACS. It began in FY02 with multiple tasks all aimed at a process that came to be known as fly the mission, that is, conduct a full, continuous mission profile for a vehicle there in the wind tunnel. That necessitates automation of plant controls so that the wind tunnel system can respond to the changing conditions of the vehicle in a real-time updating environment. It also requires integration of the operations functions required to operate the tunnel and respond to the test article. This includes data acquisition, utilities and all other systems. The main thrust of the program is to develop and apply new technologies in instrumentation, controls, and processes to make integration a reality, resulting in a faster loop, as illustrated in Fig. 14. This AIDACS philosophy is embodied by example in the continuous sweep mode test technique in tunnel 16T described earlier. Also partially funded under the AIDACS program is a new test technique involving pressure sensitive paint or PSP. This process involves painting a wind tunnel model with a paint that changes luminescence depending upon the airloads present at the local area at any given time. Digital cameras measure the distribution of luminescence on the surface of the model. This distribution is then converted by computational image processing to pressure or pressure 12

13 coefficient data. The image data from all of the different cameras are mapped onto a 3-D grid of the model. This process could ultimately do away with the need for pressure loads testing, saving millions of dollars for a large aircraft development program. Pressure distribution data could easily be taken on a standard force and moment model at the same time as lift and drag are being determined, eliminating wind tunnel testing and improving overall acquisition cycle time. These data could also be used to validate CFD results for pressure distribution. Figure 15 shows an early comparison between PSP data and CFD data taken on an F-16. The next steps for IT&E point to an even more exciting and greater cost-saving environment. Computing power will continue to grow and is not expected to plateau in the near future. With such Customer power behind it, computational codes will continue to make more accurate predictions in less time. Soon the time for a fully viscous CFD aircraft solution may be reduced from less than a week or days or possibly hours. This speed, along with the accuracy that comes from the iterative process, is the key to making computational analysis and CFD results a true partner with ground and flight testing. An important aspect of this partnership will be having the analysis available during actual testing and being able to use it in a real-time environment. This can take place not only onsite but also at remote locations. Having data available at one s fingertips both from analysis products and from previous testing, will provide the test decision makers with the ability to make more informed decisions and quickly change the test s direction when the situation warrants it. This scenario, especially the ability to have analysts off site, remote from the testing facility, must continue to be refined, taking into consideration bandwidth and security concerns. However, the dollar savings both in travel funds and time that could be realized, will continue to drive test management in that direction. V. Conclusions Ground testing, flight testing, and modeling and simulation must continue to be fused into a routine by addressing the evolution of analytical models, flexibility of the processes, and judicious consideration of the data from previous tests and models. The systematic utilization of these information sources will allow customers to benefit from a procedure that is seamless and easily used, as shown in Fig. 16. In order for this to occur, the evolution of analytical models into flexible processes that enable the application of any of these models at any point in the system development is needed. Also, the immediate availability during testing, of data from past tests and models during testing of results will create a more time-efficient system. For these reasons, customer confidence in the testing process will increase the customer interaction during testing. By building upon past experiences and using flexible pro- Plant Systems Dynamic Plan Article Real-Time Planning Affect next hours or days test (vs next month) critical regions... "envelope" management Distribution More SPEED Validation Data Data Acquisition & Evaluation Process Integrated End-to-End Figure 14. The AIDACS Loop CFD Customer Requirements Increased Experience Base System Evaluation Validated Model PSP Mach.9 4 Angle of Attack Figure 16. Requirements into Knowledge Pressure Compare Results Point Figure 15. PSP vs CFD Comparison Planning & Preparation Analytical Model Calibrated Model Wind Tunnel Plan Wind Tunnel Plan Information to support program decisions 13