Final Report Summary - MAS_LAB (Multipurpose aircraft simulation laboratory)
Executive summary
The 'Multipurpose aircraft simulation laboratory (MAS_LAB) project is intended for use as a generic aircraft model by the 'Clean Sky - System for green operation' consortium, within a mathematical optimisation tool, to develop technologies to reduce emissions of carbon dioxide (CO2), nitrogen oxides (NOX) and noise in the way the aircraft manages its trajectory. MAS_LAB is a software tool implementing the flight dynamic models for three different aircrafts: a civil commercial mainline, a regional aircraft and a business jet. It is intended to be a tool for flight trajectory optimisation and validation and trajectory guidance functions validation, when purposely interfaced to a simulator environment (Airlab), according to the Clean Sky intendments. It is supposed to be integrated and run in a MS Windows XP operating system, sharing processing resources with other applications. Quasi real-time and accelerated functioning are both contemplated.
MAS_LAB major components are:
1. Simulated air vehicles
2. Simulated automatic flight control system (AFCS).
MAS_LAB is supposed to be interfaced with:
1. the Airlab components, encompassing the flight management system (FMS), the flight control unit (FCU), the radio management panel (RMP) and the cockpit (sidestick/yoke, pedals, gear extension/retraction commands, secondary surfaces control surfaces such as flaps, slats and airbrakes)
2. a separately procured environmental impact model
3. a separately procured atmosphere model
4. a separately procured sensors model.
The three aircraft models are related to:
1. Boeing 747-100, in the large aircraft category
2. ATR 42-500, in the regional turbo-prop category
3. Cessna Citation CJ3, in the business jet category.
Moreover, the possibility of developing a simplified model to be parameterised as a generic aircraft model has been investigated.
Project context and objectives
The tendency of the past to build and test in aeronautics has clearly been overridden by a very extensive simulation activity for many reasons, one of these being the risk associated to safety-critical or safety-related systeMSc Simulation is hence strategic to develop and test all the key aircraft elements, such as flight control systems and is now being used effectively also beyond the design purposes, to conceive and evaluate mission strategies, for example, human-machine-interfaces or operational procedures. During the last decade, many commercial simulation environments and advanced software codes have been released and made available to research centres and industrial entities, which now share very powerful and versatile simulation tools. It is now possible to simulate very complex multi-variable multi-dynamic non linear aircraft systems in real time on a single personal computer. There are tools for testing, V&V of the related flight control systems and for automatic generation of executable codes, which comply with the main avionic standards (such as the DO-178B).
In this context, MAS_LAB is an ultimate state-of-the-art multi-aircraft simulator, primarily meant to be interfaced with Airlab, a pilot operated flight simulator, developed within the framework of the two major European programmes for civil air transport, Sesar and Clean Sky joint technology initiative (JTI). Scope of the MAS_LAB project was to develop a mathematical model for three different aircraft categories, wide-body, business jet and regional transport aircraft, to test new technologies such as an innovative FMS, designed to reduce emissions (CO2, NOx) and noise through the four-dimensional (4D) trajectory optimisation.
Within the integrated system, in fact, the outcomes of the flight simulator are post-processed by two external modules:
1. an environmental model to compute the noise related to the aircraft operations as well as the amount of carbon dioxide, nitrogen oxides and water vapour released in the high atmosphere
2. a cost model to evaluate the economical impact of innovative trajectory on the commercial flight operators.
The MAS_LAB project consists, more in detail, in the design of a so called master model representing four models, three of which are referred to existing aircrafts: B747-100; ATR 42-500 and Cessna Citation CJ3. The fourth model refers to a virtual aircraft, for which the parameters can be defined by the user within limited ranges imposed by physical constraints. The latter is referred to as the generic model. As the simulator was primarily meant to be used for the validation of a novel FSM, it needed accurate modelling of all aircraft subsystems that have a significant impact on flying qualities and fuel consumption.
More specifically, the project was divided into six work packages (WPs), according to the following scheme:
WP number: One
WP title: Definition of model requirements and functional architecture
WP Objective: The objective of this WP was to define the simulator specifications in terms of major features, such as hardware/software, development environment, but also in terms of what the simulator must be able to model and why
Number of Deliverables: One
Deliverable D1.1: Model technical specification (MTS): This specification establishes requirements, for a generic aircraft flight dynamics model, nomenclature MAS_LAB, for a civil commercial mainline, a regional aircraft and a business jet. The MTS document includes the model requirements and interface specification.
WP Number: Two
WP title: Model V&V
WP Objective: The objective of this WP was to verify and validate the model according to the definitions given by the verification, validation and accreditation (VV&A) recommended practice guide of the Department of Defence (DoD) model and simulation coordination office. More precisely, the objective of this WP was to:
1. verify and validate and the dynamic model of the 747-100 and of the Cessna CJ3
2. verify the dynamic model of the ATR42-500 and all the AFCSs.
Number of Deliverables: Six
D2.1: Model validation test plan: This document defines the verification and validation (V&V) plan MAS_LAB. Within the document, the objectives of V&V are defined and the methodology to achieve them is described. The V&V Plan provides the basis for further expected V&V deliverables planned (2.2 - Validation test report and 2.3 - reference test data files). In detail, in order to establish the level of accuracy of the mathematical models developed, the Federal Aviation Administration (FAA) part 60 Appendix A 'Qualification performance standards for airplane full flight simulators (FFS)' was used as reference to set up the right passing criteria for the tests accomplished on MAS_LAB. In particular the document establishes the standards for airplane FFS evaluation and qualification. The national simulator programme management (NSPM) is responsible for the development, application and implementation of the standards contained within the document. The regulation presents the most stringent existing requirements that are used to certify the level (A, B, C or D) of the FFS. Even though MAS_LAB is not a FFS, it was decided to use this regulation as reference to guarantee the best results for the project purpose. As the four models have been developed differently, according to the characteristics of the data available for each aircraft, also the V&V tests are defined and scheduled differently. D2.1 is specifically detailed for the B747-100 model and its AFCS suite.
D2.2: Validation test report: This document represents the validation test report for the B747-100 model for MAS_LAB. Tests refer to D2.1. Within the document, the achieved results are described and commented: table containing all the tests, their description, the systems involved and the limitations imposed by the regulation used as a reference are reported into two main sections: the first is related to the aircraft dynamic behaviour and performance, whereas the second concerns the AFCS mode response analysis.
D2.3: Reference test data files: xlsx file (attachment to D2.2)
D2.4: Model validation test plan update: This document represents the validation test plan for the Cessna CJ3 and ATR42-500 models and their AFCS suites. It shares the main understanding of the V&V process with D2.1. This V&V plan provides the basis for further expected V&V deliverables (2.5 - Validation test report).
D2.5: Final release of the validation test report update: This document represents the validation test report for the Cessna CJ3 and ATR42-500 models. This document assesses the level of accuracy, coherence and fidelity reached for the Cessna CJ3 and ATR42-500 models, according to the validation test plan specified in D2.4.
D2.6: Reference tests data files update: xlsx file (attachment to D2.5).
WP number: Three
WP title: Simulator implementation
WP objective: The objective of this WP was to implement the simulator, including model and interfaces, the static and dynamic performance, flying qualities, control reactions and AFCS of the four aircrafts Boeing 747-100, ATR 42-500, Cessna CJ3 and generic aircraft
Number of Deliverables: Five
D3.1: Initial aircraft simulation model software: This software represents the initial release of the MAS_LAB simulator, which includes the Boeing 747-100 model and its AFCS suite. In this first release of the software the majority of the common utilities and simulation tools have been implemented. These parts are mainly related to the communication interfaces between MAS_LAB and the users, including the vectorworks component object model (VCOM) interface; the master model architecture; the AFCS modes implementation and handling; the AFCS validation strategies; the utilities to extract performance data.
D3.2: User manual - Aircraft model description: This document describes all the aircraft model implemented for MAS_LAB. The description is articulated as to account for aircraft model components; AFCS model components; model internal and external interfaces; input and output data files; operative sequences.
D3.3: User manual - Aircraft user guide: This document describes the B747-100 model implemented for MAS_LAB. The description is articulated as to account for aircraft short description; aircraft limitations; operating information; AFCS modes; performance.
D3.4: Complete aircraft simulation model software: This software represents the initial release of the MAS_LAB simulator, which includes the CJ3 and ATR42 models and their AFCS suite.
D3.5: Aircraft performance computation tool: The performance data computation tool, a dedicated Matlab tool with a graphical user interface (GUI) implemented by Polito on Thales specifications, which calculates the specific excess power and fuel flow for different flight profiles and guidance modes.
WP number: Four
WP title: Problem fixing, updating, maintenance
WP objective: The objective of this WP was to maintain and improve the simulator software while updating the related manuals in successive releases. Feedbacks from partners and end users were to be organised and harmonised every time the simulator software and the related manuals are released. Comments were to be shared among the partners and taken under consideration for fixing problems or improve the simulator quality, reliability and portability.
Number of Deliverables: Five
D4.1: Problem report and model modification request: Scope of this document is to collect problems and model modification requests following the deliverable of the MAS_LAB Boeing 747-100 mathematical model (D3.4).
D4.2: MTS update: Scope of the document is to provide an updating for the MTS, deliverable D1.1 where the AFCS modes had not been addressed. This document, hence, outlines the proposed AFCS functioning modes to be implemented within the MAS_LAB simulation model. This includes both autopilot and autothrottle functions. Whereas specific modes may not be active for some aircraft, only modes described here are modelled. This document is the result of an activity conducted in cooperation with Thales and reported in the document series MAS_LAB AFCS modes, whose main purpose was to ensure a common understanding and agreement on the autopilot modes.
D4.3: Final release of the aircraft simulation model software, with problem fixes: The final release of the aircraft simulation model software, with problem fixes (D4.3) has been issued as the final MAS_LAB software, version 3.4. It includes the generic model for the three categories.
D4.4: Final model user manual and aircraft flight manual update: The final model user manual and aircraft flight manual update has been issued in three different documents: D4.4a for Boeing 747-100, D4.4b for Cessna CJ3 and D4.4c for ATR 42-500.
D4.5: Numerical model software and data package update for problem fixes or evolutions: The numerical model software and data package update for problem fixes or evolutions (D4.5) has been issued to include the generic model for the three categories in the performance tool. A guide on how to select parameters of the generic model both for the MAS_LAB software ver.3.4 and for the performance tool has been included.
WP number: Five
WP title: Knowledge and technological transfer
WP objective: The objective of this WP was to introduce the knowledge of the state-of-the-art simulation technology within the consortium and promote the dissemination of the technical and scientific results achieved during the project, among the partners and the international community also in non aeronautical contexts.
Number of Deliverables: None
WP number: Six
WP title: Coordination
WP objective: The objective of this WP was to ensure that the tasks are execute within the specific deadlines and economical constraints. Main actions have been to keep accurate financial records and submit these in a timely manner at reporting milestones; verify the deliverable quality and timeliness; attend regular project meetings; monitor the project progress; coordinate and supervise the activities of the WP leaders; write and submit regular progress of the project; observance of quality requirements.
Number of Deliverables: None
Project results:
The main results of the foreground obtained during the MAS_LAB are:
1. the knowledge gained by the research leaders
2. the post-docs and the PhD and undergraduate students educated and trained during the project
3. the software and documentation developed for the project
4. the methodologies implemented to obtained the expected results
5. the papers and reports published.
In what follows, a brief description will be given on how MAS_LAB project collected the state-of-the-art science and technology foregrounds to produce a powerful tool to support the design of innovative on-board systems, able to introduce substantial environmental and economic benefits in flight operations.
In May 1997 one of the major company dedicated in software development for simulation and model-based design, The Mathworks, released on a public file transfer protocol (FTP) server the flight dynamics and control (FDC) toolbox for Matlab and Simulink. It was a free toolbox which provides flexible models and tools for flight simulation, flight dynamics analysis and flight control system design.
The toolbox was built around a generic non-linear aircraft model, with a modular design that provided maximum flexibility to the user. The aircraft model was implemented as a Simulink block-diagram, which could be accessed both from the Matlab workspace and directly through the graphical user-interface of Simulink. The model was configured for the DeHavilland DHC-2 Beaver aircraft, but could be adapted for many different kinds of airplanes if required. Since then the toolbox has converted in an open source project and undergone an extensive updating process which has lead, through the years, to a very complex model which complies with the latest Matlab/Simulink releases. The toolbox also includes analytical Matlab routines for extracting steady-state flight conditions and determining linearised models for user-specified operating points
The Modelling process for MAS_LAB
In accordance with the FDC toolbox structure, MAS_LAB was developed to be modular and several modules can be identified: AFCS (AFCS, modelling the entirety of autopilot suites), Actuators (modelling the operation of aerodynamic control surfaces), Engine (modelling the propulsion system), Ground (modelling the landing gear), Gravity (modelling gravitational forces on the aircraft), Aerodynamics (modelling the aerodynamic forces on the aircraft, including the effects of the operation of control surfaces), Atmosphere (implementing an ISA atmosphere model), Fuel System (modelling the fuel consumption and its effects on inertial properties), Equations of Motion (implementing the 6DOF equations of motion) and Sensors (modelling on-board sensors that provide data to the AFCS).
Two modules have proved to be particularly critical and required extensive discussion among the partners: the propulsion system and the AFCS suite. For the 4D trajectory evaluation, in fact, an accurate modelling of the propulsion system was considered paramount, for correct estimation of the aircraft performance and fuel consumption. High-fidelity modelling of the propulsion system would require the aero-thermo-dynamic simulation of the entire engine and its components. This would require extensive knowledge of the engine characteristics, which belongs to data usually not disclosed by the engine companies. The engine model developed for MAS_LAB is based on a methodology which uses the off-line estimation of the engine main parameter trend in the entire operating range, in terms of Mach number and altitude. It is important to notice that the three aircraft selected to represent the three different categories are equipped with two different engine technologies, the turbofan and the turboprop, which implied differentiated modelling. This approach is interesting, as it is effective and computationally efficient and can be successfully adopt both for a turbofan and a turboprop engines. The methodologies used for the simulation of the two different propulsion systems present small differences, for this reason they will be described separately.
As for the two turbofan engines (JT9D-3 and FJ44-3A), the mathematical model reported in [1] was used as main reference. It consists mainly in mapping the following relations: throttle (TH) to power lever angle (PLA), PLA to engine pressure ratio (EPR) and EPR to thrust. The critical part in this modelling process is to obtain the complete database for the relations mentioned above from a relatively small set of available data. This last issue has been addressed through the use of commercial software such as the gas turbine simulation programme (GSP) developed by the Netherlands National Aerospace Laboratory (NLR). From data publicly available in [2], engine components working maps were theoretically obtained and then used off-line in the GSP to generate the database.
Once the engine and the aircraft dynamic mathematical models were fully validated, the fuel consumption models could be implemented by means of a heuristic function: 'TSFC_adim=f(Thrust_adim,Mach)' and validated with the flight manual data found in [3] and [4].
Regarding the turboprop engine, (PW 127-E) the difficulty was increased by the even greater lack of data; for example, data regarding the airfoils and pitch angles used for propeller blades, but also the gearbox reduction ratio, are not publicly disclosed. With reference to the formula reported in [5] and considering data available in [6] a simple model architecture was implemented to compute the following relation: 'Thrust=f(rpm%,TQ%,V_e,T)', where rpm% and TQ% are the pilot inputs. As for the fuel consumption, the same architecture described above for the turbofan has been implemented, substituting Thrust with TQ% and Mach with H altitude.
Also for the AFCS modelling, the main issue was the t lack of literature dedicated to the development of autopilots. More precisely, whereas theoretical aspects of autopilots are extensively analysed ([7] and [8]), the actual modelling and development of an autopilot is not a common subject for literature, mainly because autopilots are a commercial products. Moreover autopilot functionalities depend on the specific aircraft and differ significantly from one category to another. For example the Boeing 747 is provided with an autothrottle, but that is not the case for the ATR 42 and the Cessna Citation. In order to simplify the model, it was decided that all modelled aircraft would be provided with the same AFCS capabilities; this also allowed to streamline the interface between the simulator and the FMSC
The AFCS functionality can be split between three separate areas of operation: longitudinal control laws, lateral-directional control laws and autothrottle. The following control laws were implemented for the longitudinal axis:
1. pitch hold (Manual); maintains a fixed pitch angle
2. altitude hold; captures and maintains the desired altitude
3. vertical speed hold; captures and maintains the desired vertical speed
4. indicated air speed (IAS) hold; captures and maintains the desired airspeed by manoeuvring the aircraft along the longitudinal axis
5. glide-slope; captures and follows an instrument landing system (ILS) radio signal
6. take-off/go-around (TOGA); maintains a pre-determined airspeed that ensures that the maximum safe climb rate is achieved.
The following control laws were implemented for the lateral-directional axis:
1. roll-hold (Manual); maintains a fixed roll angle
2. heading hold; captures and maintains the desired heading
3. track-hold; captures and maintains the desired track
4. localiser; captures and follows an ILS or very high frequency (VHF) omnidirectional radio range (VOR) radio signal
5. TOGA; maintains level wings (zero roll angle).
The following control laws were implemented for the autothrottle:
1. thrust reference; switches between several pre-determined thrust values, such as TOGA, climb and idle
2. speed hold; captures and maintains the desired airspeed, by operating on the throttle
3. Mach hold; captures and maintains the desired Mach number, by operating on the throttle.
In modern aircraft, autopilot mode activation and switching is governed by a large set of safety rules; for example, switching between certain modes could be prohibited, or the activation of a mode might require specific conditions to be met. Within MAS_LAB, the most important of such rules were implemented as a mode selection logic that governs mode activation. This was modelled using Stateflow charts and continuously determines the current active mode for each AFCS axis by combining pilot inputs and flight data with the embedded rules.
The V&V process for MAS_LAB
The V&V process implemented for MAS_LAB was based on documentation published by the Defence Modelling and Simulation Office (DMSO) of the American Department of Defense (DoD). These recommended practice guides (RPG) on Verification, Validation and Accreditation (VV&A) [9] allow to establish guidelines and methodologies that can be expected to result in meaningful V&V campaigns. The documentation is publicly available online at the following internet address: http://vva.msco.mil/Default.htm.
In order to define the objectives of the entire V&V process for MAS_LAB, a number of definitions and theoretical concepts from the RPGs needs to be outlined. The first definition is that of modelling and simulation (M&S), which is defined as follows: 'the use of models, including emulators, prototypes, simulators and stimulators, either statically or over time, to develop data as a basis for making managerial or technical decisions'. While the terms M&S are often used interchangeably, a distinction exists: a model is defined as 'a physical, mathematical, or otherwise logical representation of a system, entity, phenomenon, or process', while modelling is the 'application of a standard, rigorous, structured methodology to create and validate a model'. Note that the implementation of a model is normally considered to be static; producing output from the model requires a simulation. A simulation is defined as 'a method for implementing a model over time.' Separating the definition of the model from the simulation is an extremely useful method for developing analytical tools. This modular approach, combined with well-defined interfaces, allows updating models as necessary without the need for updating the simulation software. It also supports a more thorough approach to V&V by allowing a V&V practitioner to separate the search for errors associated with the model from errors associated with the implementation of time.
Within MAS_LAB, the simulation environment is provided by the Matlab/Simulink software package. As this is a widely used simulation platform, which is well known and trusted by both the MAS_LAB user and developer, it was assumed that V&V of the platform was outside the scope of the V&V process. This means that the V&V process was focussed on determining whether the model itself was correct, rather than on checking whether the simulation environment provided sufficient fidelity. However, due to its nature, the MAS_LAB model had to be tested under both static and dynamic conditions, since many aircraft characteristics are dynamic in nature.
It is important at this stage to distinguish between V&V, since the two terms are often used together and incorrectly treated as if they were interchangeable.
Verification is defined as 'the process of determining that a model or simulation implementation accurately represents the developer's conceptual description and specification'. Verification also evaluates the extent to which the model or simulation has been developed using sound and established software engineering techniques.
Validation is defined as 'the process of determining the degree to which a model or simulation is an accurate representation of the real world from the perspective of the intended uses of the model or simulation'.
Within MAS_LAB, verification consisted in ensuring that the developed models:
1. were bug-free, both statically and dynamically
2. provided consistent output for all input conditions
3. were correctly interfaced with external components.
Validation instead consisted in ensuring that the models:
1. reproduced the modelled aircraft with an acceptable degree of fidelity
2. provided output data where the error compared to the real-world was sufficiently small so that the simulation was useful for the purposes of the user.
The RPGs define a large number of useful V&V techniques, such as debugging, assertion checking, special input testing, graphical comparison, functional testing and face validation and acceptance testing. This last technique is usually the final stage of each model delivery but is usually the most important, as it consists in testing the model operationally with the actual hardware and data to determine whether all requirements specified in the legal contract are satisfied. Requirements are usually related to standards which imply the availability of pre-defined input/output data-sets collected during specifically designed flight test. This means that if flight data are available: validation can be performing according to any of these standards:
1. Australia: FSD-1, Operational standards and requirements, approved flight simulators
2. Canada: TP9685, Aeroplane and rotorcraft simulator manual
3. France: Projet d'arrêté relatif à l'agrément des simulateurs de vol 1988
4. United Kingdom: CAP 453, Aeroplane flight simulators: Approval requirements
5. United States: Advisory circular 120-40B airplane simulator qualification.
All these standards agree that the only mean of performing validation is through actual data, provided or verified and authorised only by the aircraft manufacturer. Validation on separate subsystems, moreover, is not accepted as a way of validation. This process is obviously extremely expensive and time consuming and it is performed only if it actually gives economical/practical advantages, for example because it allows the simulator to be used extensively instead of the real A/C, such as for the Training Simulators. The aircraft constructors have a direct advantage in promoting Training Simulators, as they can be seen as a powerful tool to qualify pilots on their aircrafts, increasing acceptance and demand among the airliners. The advantage is so tangible that aircraft constructors accept to grant a sale of flight test data to the simulator developers. For each aircraft budgets are in the million EUR range.
MAS_LAB falls in a different category, the laboratory simulators, where simulation data are used for prediction. Flight data were not available for the three aircraft but validation at a subsystem level was not considered acceptable or practicable. The problem of gathering reliable data and define an acceptable process for effective validation of the three aircraft models was a central issue for MAS_LAB and required extensive discussion among the partners.
Five main data sources were identified:
1. Scientific literature, in the form of book chapters and papers, which is particularly useful to implement the special input testing technique by providing example responses to specific input configurations. In particular, an extensive Boeing 747-100 documentation was found including responses to commands as well as modal parameters in different configuration and flight conditions.
2. Data included in the Eurocontrol Base of Aircraft Data (BADA), from the Eurocontrol Experimental Centre, for which a license was requested and obtained for the length of the research project. BADA contains summary performance tables of true air speed climb/descent rates and fuel consumption at various flight levels for many different specific aircrafts. This data can be used as reference data both for special input testing and functional testing techniques.
3. Data released from the National Transportation Safety Board (NTSB) and related to flight data (including but not limited to flight data recorders) recorded as a consequence of minor to major accidents. In this case, flight data from minor accidents, especially when related to the final phases of flight, can be used to test the simulator in realistic scenarios.
4. Data released in the aircraft flight manuals in forms of tables for different flight conditions. This data is usually related to the fuel consumption and performance prediction, which is a central issue for MAS_LAB.
5. An American University, which was funded as sub-contractor, acquired, for a previous project, a wide set of geometric, inertial and flight data form Cessna. These data have undergone a process identifier (PID) process for the aerodynamic coefficients estimation. Following these procedure a flight simulator was implemented and certified according to the FAA regulation (level six) by the American University. Due to the proprietary nature of the Cessna CJ3 flight data, however, the numerical estimates of the aerodynamic coefficients for the CJ3 were not directly available and were not released in any form. A process has been purposely defined and implemented to perform validation on the business jet model, without violating confidentiality agreements and property rights.
References:
[1] C. Rodney Hanke and Donald R. Nordwall, The simulation of a jumbo jet Transport aircraft Volume II: modelling data, D6 -30643, The Boeing company Wichita division Wichita, Kansas, September 1970.
[1] Jane's Aero Engine, http://www.janes.com.
[2] Boeing Document D6-13703, Boeing 747 Flight Manual, December 30, 1979.
[3] Citation CJ3 Flight Planning Guide, May 2006.
[4] Raymer D. P., Aircraft Design: A Conceptual Approach, 3rd ed., AIAA Education Series, 1999.
[5] ATR 42-500 Flight Manuals, F.C.O.M. 3.05.00 March 1993.
[6] Stevens B. and Lewis F., Aircraft Control and Simulation, J. Wiley and Sons, 2004.
[7] Etkin B. and Reid L., Dynamics of flight: Stability and Control, J. Wiley and Sons, 1996.
[8] 'Flight control systems: practical issues in design and implementation', Edited by Roger Pratt, Institution of Electrical Engineers, IEE Control Engineering Series, 2000.
[9] US Department of Defense, 'Validation, Verification and Accreditation Recommended Practice Guide', last accessed on 30 January 2013
Potential impact:
A team has been established, working in the total respect of the main research principles. Young researchers have been valorised, to support the growth of highly specialised profiles. In this context, no racial, gender, religious of political discrimination has been admitted in the selection of the new personnel. All the principles and dictates of the working discipline have been respected, for the employees and the environment work, both for the Politecnico and the companies (subcontractors) involved. No one has been somehow forced to act against his/her willing or political/ethical/religious precepts. On the contrary, differences have been valorised and harmonised within the group, as a form of cultural enrichment.
As in the tradition of a free and independent University, no rigid daily scheduling has been enforced. The group members had at their disposal all the most advanced communication tools to help them staying constantly connected with the rest of the group without having to commute daily or move permanently from their residence, causing discomfort for themselves or for their families. This is considered a strong measure to reduce the work cost, help reconcile work and private life and promote gender equality.
As far as the dissemination of project results is concerned, the group has worked respecting the principles of rights and duties of a research centre such as the Politecnico di Torino, which is to divulge and spread the most advanced knowledge. The group firmly believes in the importance to promulgate advance courses and lectures for highly specialised profiles (PhD students) which are expected to be absorbed by big industries, as well as small and medium sized enterprises (SMEs). As in the tradition of the group, moreover, dissemination has been pursued through technical and scientific reports as well as the attendance of international conferences.
The foreground results are primarily the algorithms, the software, the models and the methods implemented to obtain the expected results. Dissemination during the project has been organised with two main objectives:
1. to recruit and train researchers who had actively supported the MAS_LAB project or parts of it: in this case dissemination has been carried out through dedicated meeting during which the project documentation and the relevant bibliography has been made available, discussed and commented;
2. to communicate effectively and ensure common understanding among the parties involved in the MAS_LAB project: in this case a glossary has been prepared, discussed, approved and included in the first deliverable of the project (D1.1). One of the criticality of the project, in fact, was that the parties had different backgrounds, besides belonging to different countries and kind of structures (academy and industry). The objective difficulties, moreover, was that different producers (for example Boeing and Airbus) have different names for systems with similar functionality. A common base was established among the parties, also, through periodic meetings in which the main theoretical issued were discussed from the different perspectives.
During the project communication with users outside the project has been mainly pursued at an academic level, within undergraduate courses in the flight mechanic area. In particular, a four-hour seminar has been prepared and proposed to the BSc students within the course aircraft guidance and control (eight credits). The seminar AFCS for modern civil aircraft is focussed on the most practical issued related to the AFCS use, especially in the category of the large aircraft. The outline is the following:
1. AFCS functions in civil aircraft - SAS, CAS and autopilots: the command and control systems of two large aircraft (Boeing 474-100 and Airbus 330-300) are described in details and the main differences are outlined. Autopilots modes, their interaction and their limitations are presented. The simplified control schemes of the main autopilots are introduced and commented.
2. Conceptual differences for the main constructors (Boeing and Airbus)
3. Pilot interface with the AFCS: the interfaces for the main constructors (Boeing and Airbus) are presented. Flight director (FD), mode select panel (MSP) or mode control panel (MCP), FCU, (multipurpose) control display unit, FMS are discussed in details
4. Typical mission profiles and interaction with the AFCS: practical examples of Boeing 747-100, Boeing 747-400, Airbus 330-300.
As outlined in the list of dissemination activities many of the topics addressed during the project had been the subject of undergraduate theses. In particular four MSc and three BSc theses have been completed. Their content will be briefly described hereinafter:
1. Title: Modelling and validation of the B747-100 AFCS suite for the implementation of an innovative FMS
Author: Mario Cassaro
Faculty Advisor: Manuela Battipede
Abstract: This report is the result of a graduation research, which has been carried out within the Clean Sky project for the development of MAS_LAB. The main objectives of this work were to verify and validate a B747-100 nonlinear aircraft dynamic model and also to implement and validate the AFCS. To achieve the proposed objectives it has been decided that classical control techniques would be used to develop the autopilot controllers. The steps to achieved the proposed objectives can be listed as follows:
1. non linear model implementation, validation and trimming;
2. model linearisation and dynamic response investigation;
3. stability and control augmentation system design;
4. control logic development;
5. non-linear testing of the complete model.
Title: Mathematical model turboprop aircraft for the implementation of an innovative FMS 5L
Author: John Paronitti
Faculty Advisors: Manuela Battipede, Paolo Maggiore
Abstract: This dissertation is the outcome of a eight-month long research work carried out at the Politecnico di Torino's Aerospace Engineering Department, whose aim was the implementation of a turboprop aircraft mathematical model that will be used for the development of an innovative FMSC The first objective of this dissertation was the calculation the aerodynamic coefficients of a simplified geometric model of a modern commercial turboprop aircraft such as the ATR 42-500 through the United States Air Force-developed commercial software digital DATCOM. Due to the inherent limitations of the DATCOM, it was very important to determine the plausibility of the results of its output file, they have therefore been compared with similar plots found in literature. The second objective was the development of the propulsion, aerodynamic, flight controls and dynamic models of the ATR 42 that had to be implemented as modules into the MAS_LAB master model, in order to calculate the forces and the moments the aircraft is subject to.
Title: Implementation of a mathematical model for the AFCS analysis for a large aircraft
Author: Davide Bietto
Faculty Advisors: Manuela Battipede, Paolo Maggiore
Abstract: This thesis describes the work carried out within the Clean Sky MAS_LAB project In particular, the task carried out and reported here concerns the implementation of the flight control systems (longitudinal, lateral and directional controls, flaps and stabiliser) and the propulsion system of the Boeing 747-100 and the validation of these models at a subsystem level. It should be noted that all the work presented here, concerns one of the most complex model in the state-of-the-art and the first model implemented and tested in the MAS_LAB project. This thesis, hence, represent a reference for the other models (ATR42-500, CJ3 and the Generic) to be implemented verified and validated. The reader should be familiar with classical control theory as well as with the flight mechanics and dynamics main concepts. Matlab and Simulink are the basic modelling tools.
Title: Non linear dynamic modelling of the Cessna Citation CJ3 and implementation of the relevant autopilots
Author: Francesco Trifilò
Faculty Advisor: Manuela Battipede
Abstract
This thesis has been developed within the European research project called Clean Sky and was carried out at the Department of Mechanical and Aerospace Engineering (DIMEAS) of the Politecnico di Torino. The main objectives consisted in implementing and validating the Cessna Citation CJ3 nonlinear dynamic model and the relevant AFCS. The V&V process has been performed using specifications reported in MIL-DTL-9490E. The implemented model of the aircraft is part of the flight simulator platform called MAS_LAB. The steps involved in the development of the project are as follows:
1. implementation of the geometric model and its aerodynamic database obtained using DATCOM;
2. model linearisation and dynamic response investigation;
3. stability and control augmentation system design;
4. V&V of the complete non-linear model.
Title: Validation of a turbojet engine dynamic model for a for fuel consumption analysis
Author: Daniele Mazzotta
Faculty Advisors: Manuela Battipede, Michele Ferlauto
Abstract: The main objective of this BSc Thesis was to validate the lumped element mathematical model of the Pratt&Whitney JT9D-3A turboprop engine, which equips the B747-100. The engine mathematical model is intended primarily for the evaluation of the trust and fuel consumption in the whole aircraft flight envelope and is part of a Clean Sky project, for which the main aim is to implement and validate a flight simulator, the MAS_LAB. The validation process consists in the identification of the coherence of data calculated through the use of commercial software such as the GSP developed by the NLR. These data in fact are extrapolated from the data available in the current literature, which is relevant to particular flight conditions, such a full throttle or zero speed. The behaviour of data outside this flight condition has been analysed and corrected at a subsystem level and with the model integrated in the full aircraft simulator.
Title: Implementation of a virtual FCU panel for a flight simulator
Author: Marco Sasanelli
Faculty Advisors: Manuela Battipede, Matteo Vazzola
Abstract: The objective of this BSc thesis is to implement a software interface for MAS_LAB, with the same layout and the functionalities of the FCU panel. In particular the panel is used as a human-machine interface, to set autopilot parameters such as cruise speed, vertical speed or flight path angle; heading angle; altitude; autopilot mode transition; autothrottle mode transition. The thesis is divided into three parts: the first part is dedicated to a description of the commercial software LabVIEW and in particular to the add-on application Simulink interface tool (SIT) through which the FCU panel has been implemented, interfaced and made run simultaneously with the MAS_LAB master model. The second part consists in the description of the features and functionalities of the FCU Panel developed for the MAS_LAB project. The third part is dedicated to the description of the programming details. Each virtual element is analysed: criticalities in the implementation are discussed and validation tests are presented. A final test consists in the evaluation of the performance of the FCU panel interface when functioning in real time as an embedded software.
Title: Feasibility analysis of a parametric mathematical model of a conventional aircraft
Author: Laura Novaro Mascarello
Faculty Advisors: Manuela Battipede
Abstract: The main objective of this BSc thesis is to provide a parameterisation of some geometric features of a generic aircraft, identifying mathematical relationships between the geometric features and several aircraft performance parameters. The choice of the parameters to be analysed is not random: on the contrary it is based to an appropriate strategy, partially echoes the preliminary design philosophy of the aviation companies when faced with a new project. The following steps were accomplished:
1. statistical analysis over the geometrical and performance data of aircraft belonging to the categories of large aircraft, business jet and regional aircraft
2. detailed comparison between geometric and performance data of similar large aircraft made by the two most important aviation companies Boeing and Airbus
3. partially parameterisation of the flight control block implemented in MAS_LAB using Matlab/Simulink programming language
4. identification of the influence of bounded gross weight variation over the geometrical and performance parameters of a generic aircraft.
As far as the exploitation of results is concerned, results have been made available for all the JU partners: the simulator modularity makes the platform suitable for different types of aircraft with different operational goals within and beyond the scope of the Clean Sky JTI.
List of websites:
Piero Gili
Mechanical and Aerospace Department (DIMEAS)
Politecnico di Torino
Corso Duca degli Abruzzi, 24
10129 Torino (TO)
Tel. +39-011-5646854
Fax. +39-011-5646899
piero.gili@polito.it
http://www.polito.it
Manuela Battipede
Mechanical and Aerospace Department (DIMEAS)
Politecnico di Torino
Corso Duca degli Abruzzi, 24
10129 Torino (TO)
Tel. +39-011-5646874
Fax. +39-011-5646899
manuela.battipede@polito.it
http://www.polito.it
Michele Visone
BLUE Group Engineering and Design
Via Coroglio, 57
c/o BIC città della scienza
80124 Napoli (NA)
Tel. +39-081-6171534
Fax. +39-081-7352433
m.visone@blue-group.it
bluenapoli@blue-group.it
http://www.blue-group.it
The 'Multipurpose aircraft simulation laboratory (MAS_LAB) project is intended for use as a generic aircraft model by the 'Clean Sky - System for green operation' consortium, within a mathematical optimisation tool, to develop technologies to reduce emissions of carbon dioxide (CO2), nitrogen oxides (NOX) and noise in the way the aircraft manages its trajectory. MAS_LAB is a software tool implementing the flight dynamic models for three different aircrafts: a civil commercial mainline, a regional aircraft and a business jet. It is intended to be a tool for flight trajectory optimisation and validation and trajectory guidance functions validation, when purposely interfaced to a simulator environment (Airlab), according to the Clean Sky intendments. It is supposed to be integrated and run in a MS Windows XP operating system, sharing processing resources with other applications. Quasi real-time and accelerated functioning are both contemplated.
MAS_LAB major components are:
1. Simulated air vehicles
2. Simulated automatic flight control system (AFCS).
MAS_LAB is supposed to be interfaced with:
1. the Airlab components, encompassing the flight management system (FMS), the flight control unit (FCU), the radio management panel (RMP) and the cockpit (sidestick/yoke, pedals, gear extension/retraction commands, secondary surfaces control surfaces such as flaps, slats and airbrakes)
2. a separately procured environmental impact model
3. a separately procured atmosphere model
4. a separately procured sensors model.
The three aircraft models are related to:
1. Boeing 747-100, in the large aircraft category
2. ATR 42-500, in the regional turbo-prop category
3. Cessna Citation CJ3, in the business jet category.
Moreover, the possibility of developing a simplified model to be parameterised as a generic aircraft model has been investigated.
Project context and objectives
The tendency of the past to build and test in aeronautics has clearly been overridden by a very extensive simulation activity for many reasons, one of these being the risk associated to safety-critical or safety-related systeMSc Simulation is hence strategic to develop and test all the key aircraft elements, such as flight control systems and is now being used effectively also beyond the design purposes, to conceive and evaluate mission strategies, for example, human-machine-interfaces or operational procedures. During the last decade, many commercial simulation environments and advanced software codes have been released and made available to research centres and industrial entities, which now share very powerful and versatile simulation tools. It is now possible to simulate very complex multi-variable multi-dynamic non linear aircraft systems in real time on a single personal computer. There are tools for testing, V&V of the related flight control systems and for automatic generation of executable codes, which comply with the main avionic standards (such as the DO-178B).
In this context, MAS_LAB is an ultimate state-of-the-art multi-aircraft simulator, primarily meant to be interfaced with Airlab, a pilot operated flight simulator, developed within the framework of the two major European programmes for civil air transport, Sesar and Clean Sky joint technology initiative (JTI). Scope of the MAS_LAB project was to develop a mathematical model for three different aircraft categories, wide-body, business jet and regional transport aircraft, to test new technologies such as an innovative FMS, designed to reduce emissions (CO2, NOx) and noise through the four-dimensional (4D) trajectory optimisation.
Within the integrated system, in fact, the outcomes of the flight simulator are post-processed by two external modules:
1. an environmental model to compute the noise related to the aircraft operations as well as the amount of carbon dioxide, nitrogen oxides and water vapour released in the high atmosphere
2. a cost model to evaluate the economical impact of innovative trajectory on the commercial flight operators.
The MAS_LAB project consists, more in detail, in the design of a so called master model representing four models, three of which are referred to existing aircrafts: B747-100; ATR 42-500 and Cessna Citation CJ3. The fourth model refers to a virtual aircraft, for which the parameters can be defined by the user within limited ranges imposed by physical constraints. The latter is referred to as the generic model. As the simulator was primarily meant to be used for the validation of a novel FSM, it needed accurate modelling of all aircraft subsystems that have a significant impact on flying qualities and fuel consumption.
More specifically, the project was divided into six work packages (WPs), according to the following scheme:
WP number: One
WP title: Definition of model requirements and functional architecture
WP Objective: The objective of this WP was to define the simulator specifications in terms of major features, such as hardware/software, development environment, but also in terms of what the simulator must be able to model and why
Number of Deliverables: One
Deliverable D1.1: Model technical specification (MTS): This specification establishes requirements, for a generic aircraft flight dynamics model, nomenclature MAS_LAB, for a civil commercial mainline, a regional aircraft and a business jet. The MTS document includes the model requirements and interface specification.
WP Number: Two
WP title: Model V&V
WP Objective: The objective of this WP was to verify and validate the model according to the definitions given by the verification, validation and accreditation (VV&A) recommended practice guide of the Department of Defence (DoD) model and simulation coordination office. More precisely, the objective of this WP was to:
1. verify and validate and the dynamic model of the 747-100 and of the Cessna CJ3
2. verify the dynamic model of the ATR42-500 and all the AFCSs.
Number of Deliverables: Six
D2.1: Model validation test plan: This document defines the verification and validation (V&V) plan MAS_LAB. Within the document, the objectives of V&V are defined and the methodology to achieve them is described. The V&V Plan provides the basis for further expected V&V deliverables planned (2.2 - Validation test report and 2.3 - reference test data files). In detail, in order to establish the level of accuracy of the mathematical models developed, the Federal Aviation Administration (FAA) part 60 Appendix A 'Qualification performance standards for airplane full flight simulators (FFS)' was used as reference to set up the right passing criteria for the tests accomplished on MAS_LAB. In particular the document establishes the standards for airplane FFS evaluation and qualification. The national simulator programme management (NSPM) is responsible for the development, application and implementation of the standards contained within the document. The regulation presents the most stringent existing requirements that are used to certify the level (A, B, C or D) of the FFS. Even though MAS_LAB is not a FFS, it was decided to use this regulation as reference to guarantee the best results for the project purpose. As the four models have been developed differently, according to the characteristics of the data available for each aircraft, also the V&V tests are defined and scheduled differently. D2.1 is specifically detailed for the B747-100 model and its AFCS suite.
D2.2: Validation test report: This document represents the validation test report for the B747-100 model for MAS_LAB. Tests refer to D2.1. Within the document, the achieved results are described and commented: table containing all the tests, their description, the systems involved and the limitations imposed by the regulation used as a reference are reported into two main sections: the first is related to the aircraft dynamic behaviour and performance, whereas the second concerns the AFCS mode response analysis.
D2.3: Reference test data files: xlsx file (attachment to D2.2)
D2.4: Model validation test plan update: This document represents the validation test plan for the Cessna CJ3 and ATR42-500 models and their AFCS suites. It shares the main understanding of the V&V process with D2.1. This V&V plan provides the basis for further expected V&V deliverables (2.5 - Validation test report).
D2.5: Final release of the validation test report update: This document represents the validation test report for the Cessna CJ3 and ATR42-500 models. This document assesses the level of accuracy, coherence and fidelity reached for the Cessna CJ3 and ATR42-500 models, according to the validation test plan specified in D2.4.
D2.6: Reference tests data files update: xlsx file (attachment to D2.5).
WP number: Three
WP title: Simulator implementation
WP objective: The objective of this WP was to implement the simulator, including model and interfaces, the static and dynamic performance, flying qualities, control reactions and AFCS of the four aircrafts Boeing 747-100, ATR 42-500, Cessna CJ3 and generic aircraft
Number of Deliverables: Five
D3.1: Initial aircraft simulation model software: This software represents the initial release of the MAS_LAB simulator, which includes the Boeing 747-100 model and its AFCS suite. In this first release of the software the majority of the common utilities and simulation tools have been implemented. These parts are mainly related to the communication interfaces between MAS_LAB and the users, including the vectorworks component object model (VCOM) interface; the master model architecture; the AFCS modes implementation and handling; the AFCS validation strategies; the utilities to extract performance data.
D3.2: User manual - Aircraft model description: This document describes all the aircraft model implemented for MAS_LAB. The description is articulated as to account for aircraft model components; AFCS model components; model internal and external interfaces; input and output data files; operative sequences.
D3.3: User manual - Aircraft user guide: This document describes the B747-100 model implemented for MAS_LAB. The description is articulated as to account for aircraft short description; aircraft limitations; operating information; AFCS modes; performance.
D3.4: Complete aircraft simulation model software: This software represents the initial release of the MAS_LAB simulator, which includes the CJ3 and ATR42 models and their AFCS suite.
D3.5: Aircraft performance computation tool: The performance data computation tool, a dedicated Matlab tool with a graphical user interface (GUI) implemented by Polito on Thales specifications, which calculates the specific excess power and fuel flow for different flight profiles and guidance modes.
WP number: Four
WP title: Problem fixing, updating, maintenance
WP objective: The objective of this WP was to maintain and improve the simulator software while updating the related manuals in successive releases. Feedbacks from partners and end users were to be organised and harmonised every time the simulator software and the related manuals are released. Comments were to be shared among the partners and taken under consideration for fixing problems or improve the simulator quality, reliability and portability.
Number of Deliverables: Five
D4.1: Problem report and model modification request: Scope of this document is to collect problems and model modification requests following the deliverable of the MAS_LAB Boeing 747-100 mathematical model (D3.4).
D4.2: MTS update: Scope of the document is to provide an updating for the MTS, deliverable D1.1 where the AFCS modes had not been addressed. This document, hence, outlines the proposed AFCS functioning modes to be implemented within the MAS_LAB simulation model. This includes both autopilot and autothrottle functions. Whereas specific modes may not be active for some aircraft, only modes described here are modelled. This document is the result of an activity conducted in cooperation with Thales and reported in the document series MAS_LAB AFCS modes, whose main purpose was to ensure a common understanding and agreement on the autopilot modes.
D4.3: Final release of the aircraft simulation model software, with problem fixes: The final release of the aircraft simulation model software, with problem fixes (D4.3) has been issued as the final MAS_LAB software, version 3.4. It includes the generic model for the three categories.
D4.4: Final model user manual and aircraft flight manual update: The final model user manual and aircraft flight manual update has been issued in three different documents: D4.4a for Boeing 747-100, D4.4b for Cessna CJ3 and D4.4c for ATR 42-500.
D4.5: Numerical model software and data package update for problem fixes or evolutions: The numerical model software and data package update for problem fixes or evolutions (D4.5) has been issued to include the generic model for the three categories in the performance tool. A guide on how to select parameters of the generic model both for the MAS_LAB software ver.3.4 and for the performance tool has been included.
WP number: Five
WP title: Knowledge and technological transfer
WP objective: The objective of this WP was to introduce the knowledge of the state-of-the-art simulation technology within the consortium and promote the dissemination of the technical and scientific results achieved during the project, among the partners and the international community also in non aeronautical contexts.
Number of Deliverables: None
WP number: Six
WP title: Coordination
WP objective: The objective of this WP was to ensure that the tasks are execute within the specific deadlines and economical constraints. Main actions have been to keep accurate financial records and submit these in a timely manner at reporting milestones; verify the deliverable quality and timeliness; attend regular project meetings; monitor the project progress; coordinate and supervise the activities of the WP leaders; write and submit regular progress of the project; observance of quality requirements.
Number of Deliverables: None
Project results:
The main results of the foreground obtained during the MAS_LAB are:
1. the knowledge gained by the research leaders
2. the post-docs and the PhD and undergraduate students educated and trained during the project
3. the software and documentation developed for the project
4. the methodologies implemented to obtained the expected results
5. the papers and reports published.
In what follows, a brief description will be given on how MAS_LAB project collected the state-of-the-art science and technology foregrounds to produce a powerful tool to support the design of innovative on-board systems, able to introduce substantial environmental and economic benefits in flight operations.
In May 1997 one of the major company dedicated in software development for simulation and model-based design, The Mathworks, released on a public file transfer protocol (FTP) server the flight dynamics and control (FDC) toolbox for Matlab and Simulink. It was a free toolbox which provides flexible models and tools for flight simulation, flight dynamics analysis and flight control system design.
The toolbox was built around a generic non-linear aircraft model, with a modular design that provided maximum flexibility to the user. The aircraft model was implemented as a Simulink block-diagram, which could be accessed both from the Matlab workspace and directly through the graphical user-interface of Simulink. The model was configured for the DeHavilland DHC-2 Beaver aircraft, but could be adapted for many different kinds of airplanes if required. Since then the toolbox has converted in an open source project and undergone an extensive updating process which has lead, through the years, to a very complex model which complies with the latest Matlab/Simulink releases. The toolbox also includes analytical Matlab routines for extracting steady-state flight conditions and determining linearised models for user-specified operating points
The Modelling process for MAS_LAB
In accordance with the FDC toolbox structure, MAS_LAB was developed to be modular and several modules can be identified: AFCS (AFCS, modelling the entirety of autopilot suites), Actuators (modelling the operation of aerodynamic control surfaces), Engine (modelling the propulsion system), Ground (modelling the landing gear), Gravity (modelling gravitational forces on the aircraft), Aerodynamics (modelling the aerodynamic forces on the aircraft, including the effects of the operation of control surfaces), Atmosphere (implementing an ISA atmosphere model), Fuel System (modelling the fuel consumption and its effects on inertial properties), Equations of Motion (implementing the 6DOF equations of motion) and Sensors (modelling on-board sensors that provide data to the AFCS).
Two modules have proved to be particularly critical and required extensive discussion among the partners: the propulsion system and the AFCS suite. For the 4D trajectory evaluation, in fact, an accurate modelling of the propulsion system was considered paramount, for correct estimation of the aircraft performance and fuel consumption. High-fidelity modelling of the propulsion system would require the aero-thermo-dynamic simulation of the entire engine and its components. This would require extensive knowledge of the engine characteristics, which belongs to data usually not disclosed by the engine companies. The engine model developed for MAS_LAB is based on a methodology which uses the off-line estimation of the engine main parameter trend in the entire operating range, in terms of Mach number and altitude. It is important to notice that the three aircraft selected to represent the three different categories are equipped with two different engine technologies, the turbofan and the turboprop, which implied differentiated modelling. This approach is interesting, as it is effective and computationally efficient and can be successfully adopt both for a turbofan and a turboprop engines. The methodologies used for the simulation of the two different propulsion systems present small differences, for this reason they will be described separately.
As for the two turbofan engines (JT9D-3 and FJ44-3A), the mathematical model reported in [1] was used as main reference. It consists mainly in mapping the following relations: throttle (TH) to power lever angle (PLA), PLA to engine pressure ratio (EPR) and EPR to thrust. The critical part in this modelling process is to obtain the complete database for the relations mentioned above from a relatively small set of available data. This last issue has been addressed through the use of commercial software such as the gas turbine simulation programme (GSP) developed by the Netherlands National Aerospace Laboratory (NLR). From data publicly available in [2], engine components working maps were theoretically obtained and then used off-line in the GSP to generate the database.
Once the engine and the aircraft dynamic mathematical models were fully validated, the fuel consumption models could be implemented by means of a heuristic function: 'TSFC_adim=f(Thrust_adim,Mach)' and validated with the flight manual data found in [3] and [4].
Regarding the turboprop engine, (PW 127-E) the difficulty was increased by the even greater lack of data; for example, data regarding the airfoils and pitch angles used for propeller blades, but also the gearbox reduction ratio, are not publicly disclosed. With reference to the formula reported in [5] and considering data available in [6] a simple model architecture was implemented to compute the following relation: 'Thrust=f(rpm%,TQ%,V_e,T)', where rpm% and TQ% are the pilot inputs. As for the fuel consumption, the same architecture described above for the turbofan has been implemented, substituting Thrust with TQ% and Mach with H altitude.
Also for the AFCS modelling, the main issue was the t lack of literature dedicated to the development of autopilots. More precisely, whereas theoretical aspects of autopilots are extensively analysed ([7] and [8]), the actual modelling and development of an autopilot is not a common subject for literature, mainly because autopilots are a commercial products. Moreover autopilot functionalities depend on the specific aircraft and differ significantly from one category to another. For example the Boeing 747 is provided with an autothrottle, but that is not the case for the ATR 42 and the Cessna Citation. In order to simplify the model, it was decided that all modelled aircraft would be provided with the same AFCS capabilities; this also allowed to streamline the interface between the simulator and the FMSC
The AFCS functionality can be split between three separate areas of operation: longitudinal control laws, lateral-directional control laws and autothrottle. The following control laws were implemented for the longitudinal axis:
1. pitch hold (Manual); maintains a fixed pitch angle
2. altitude hold; captures and maintains the desired altitude
3. vertical speed hold; captures and maintains the desired vertical speed
4. indicated air speed (IAS) hold; captures and maintains the desired airspeed by manoeuvring the aircraft along the longitudinal axis
5. glide-slope; captures and follows an instrument landing system (ILS) radio signal
6. take-off/go-around (TOGA); maintains a pre-determined airspeed that ensures that the maximum safe climb rate is achieved.
The following control laws were implemented for the lateral-directional axis:
1. roll-hold (Manual); maintains a fixed roll angle
2. heading hold; captures and maintains the desired heading
3. track-hold; captures and maintains the desired track
4. localiser; captures and follows an ILS or very high frequency (VHF) omnidirectional radio range (VOR) radio signal
5. TOGA; maintains level wings (zero roll angle).
The following control laws were implemented for the autothrottle:
1. thrust reference; switches between several pre-determined thrust values, such as TOGA, climb and idle
2. speed hold; captures and maintains the desired airspeed, by operating on the throttle
3. Mach hold; captures and maintains the desired Mach number, by operating on the throttle.
In modern aircraft, autopilot mode activation and switching is governed by a large set of safety rules; for example, switching between certain modes could be prohibited, or the activation of a mode might require specific conditions to be met. Within MAS_LAB, the most important of such rules were implemented as a mode selection logic that governs mode activation. This was modelled using Stateflow charts and continuously determines the current active mode for each AFCS axis by combining pilot inputs and flight data with the embedded rules.
The V&V process for MAS_LAB
The V&V process implemented for MAS_LAB was based on documentation published by the Defence Modelling and Simulation Office (DMSO) of the American Department of Defense (DoD). These recommended practice guides (RPG) on Verification, Validation and Accreditation (VV&A) [9] allow to establish guidelines and methodologies that can be expected to result in meaningful V&V campaigns. The documentation is publicly available online at the following internet address: http://vva.msco.mil/Default.htm.
In order to define the objectives of the entire V&V process for MAS_LAB, a number of definitions and theoretical concepts from the RPGs needs to be outlined. The first definition is that of modelling and simulation (M&S), which is defined as follows: 'the use of models, including emulators, prototypes, simulators and stimulators, either statically or over time, to develop data as a basis for making managerial or technical decisions'. While the terms M&S are often used interchangeably, a distinction exists: a model is defined as 'a physical, mathematical, or otherwise logical representation of a system, entity, phenomenon, or process', while modelling is the 'application of a standard, rigorous, structured methodology to create and validate a model'. Note that the implementation of a model is normally considered to be static; producing output from the model requires a simulation. A simulation is defined as 'a method for implementing a model over time.' Separating the definition of the model from the simulation is an extremely useful method for developing analytical tools. This modular approach, combined with well-defined interfaces, allows updating models as necessary without the need for updating the simulation software. It also supports a more thorough approach to V&V by allowing a V&V practitioner to separate the search for errors associated with the model from errors associated with the implementation of time.
Within MAS_LAB, the simulation environment is provided by the Matlab/Simulink software package. As this is a widely used simulation platform, which is well known and trusted by both the MAS_LAB user and developer, it was assumed that V&V of the platform was outside the scope of the V&V process. This means that the V&V process was focussed on determining whether the model itself was correct, rather than on checking whether the simulation environment provided sufficient fidelity. However, due to its nature, the MAS_LAB model had to be tested under both static and dynamic conditions, since many aircraft characteristics are dynamic in nature.
It is important at this stage to distinguish between V&V, since the two terms are often used together and incorrectly treated as if they were interchangeable.
Verification is defined as 'the process of determining that a model or simulation implementation accurately represents the developer's conceptual description and specification'. Verification also evaluates the extent to which the model or simulation has been developed using sound and established software engineering techniques.
Validation is defined as 'the process of determining the degree to which a model or simulation is an accurate representation of the real world from the perspective of the intended uses of the model or simulation'.
Within MAS_LAB, verification consisted in ensuring that the developed models:
1. were bug-free, both statically and dynamically
2. provided consistent output for all input conditions
3. were correctly interfaced with external components.
Validation instead consisted in ensuring that the models:
1. reproduced the modelled aircraft with an acceptable degree of fidelity
2. provided output data where the error compared to the real-world was sufficiently small so that the simulation was useful for the purposes of the user.
The RPGs define a large number of useful V&V techniques, such as debugging, assertion checking, special input testing, graphical comparison, functional testing and face validation and acceptance testing. This last technique is usually the final stage of each model delivery but is usually the most important, as it consists in testing the model operationally with the actual hardware and data to determine whether all requirements specified in the legal contract are satisfied. Requirements are usually related to standards which imply the availability of pre-defined input/output data-sets collected during specifically designed flight test. This means that if flight data are available: validation can be performing according to any of these standards:
1. Australia: FSD-1, Operational standards and requirements, approved flight simulators
2. Canada: TP9685, Aeroplane and rotorcraft simulator manual
3. France: Projet d'arrêté relatif à l'agrément des simulateurs de vol 1988
4. United Kingdom: CAP 453, Aeroplane flight simulators: Approval requirements
5. United States: Advisory circular 120-40B airplane simulator qualification.
All these standards agree that the only mean of performing validation is through actual data, provided or verified and authorised only by the aircraft manufacturer. Validation on separate subsystems, moreover, is not accepted as a way of validation. This process is obviously extremely expensive and time consuming and it is performed only if it actually gives economical/practical advantages, for example because it allows the simulator to be used extensively instead of the real A/C, such as for the Training Simulators. The aircraft constructors have a direct advantage in promoting Training Simulators, as they can be seen as a powerful tool to qualify pilots on their aircrafts, increasing acceptance and demand among the airliners. The advantage is so tangible that aircraft constructors accept to grant a sale of flight test data to the simulator developers. For each aircraft budgets are in the million EUR range.
MAS_LAB falls in a different category, the laboratory simulators, where simulation data are used for prediction. Flight data were not available for the three aircraft but validation at a subsystem level was not considered acceptable or practicable. The problem of gathering reliable data and define an acceptable process for effective validation of the three aircraft models was a central issue for MAS_LAB and required extensive discussion among the partners.
Five main data sources were identified:
1. Scientific literature, in the form of book chapters and papers, which is particularly useful to implement the special input testing technique by providing example responses to specific input configurations. In particular, an extensive Boeing 747-100 documentation was found including responses to commands as well as modal parameters in different configuration and flight conditions.
2. Data included in the Eurocontrol Base of Aircraft Data (BADA), from the Eurocontrol Experimental Centre, for which a license was requested and obtained for the length of the research project. BADA contains summary performance tables of true air speed climb/descent rates and fuel consumption at various flight levels for many different specific aircrafts. This data can be used as reference data both for special input testing and functional testing techniques.
3. Data released from the National Transportation Safety Board (NTSB) and related to flight data (including but not limited to flight data recorders) recorded as a consequence of minor to major accidents. In this case, flight data from minor accidents, especially when related to the final phases of flight, can be used to test the simulator in realistic scenarios.
4. Data released in the aircraft flight manuals in forms of tables for different flight conditions. This data is usually related to the fuel consumption and performance prediction, which is a central issue for MAS_LAB.
5. An American University, which was funded as sub-contractor, acquired, for a previous project, a wide set of geometric, inertial and flight data form Cessna. These data have undergone a process identifier (PID) process for the aerodynamic coefficients estimation. Following these procedure a flight simulator was implemented and certified according to the FAA regulation (level six) by the American University. Due to the proprietary nature of the Cessna CJ3 flight data, however, the numerical estimates of the aerodynamic coefficients for the CJ3 were not directly available and were not released in any form. A process has been purposely defined and implemented to perform validation on the business jet model, without violating confidentiality agreements and property rights.
References:
[1] C. Rodney Hanke and Donald R. Nordwall, The simulation of a jumbo jet Transport aircraft Volume II: modelling data, D6 -30643, The Boeing company Wichita division Wichita, Kansas, September 1970.
[1] Jane's Aero Engine, http://www.janes.com.
[2] Boeing Document D6-13703, Boeing 747 Flight Manual, December 30, 1979.
[3] Citation CJ3 Flight Planning Guide, May 2006.
[4] Raymer D. P., Aircraft Design: A Conceptual Approach, 3rd ed., AIAA Education Series, 1999.
[5] ATR 42-500 Flight Manuals, F.C.O.M. 3.05.00 March 1993.
[6] Stevens B. and Lewis F., Aircraft Control and Simulation, J. Wiley and Sons, 2004.
[7] Etkin B. and Reid L., Dynamics of flight: Stability and Control, J. Wiley and Sons, 1996.
[8] 'Flight control systems: practical issues in design and implementation', Edited by Roger Pratt, Institution of Electrical Engineers, IEE Control Engineering Series, 2000.
[9] US Department of Defense, 'Validation, Verification and Accreditation Recommended Practice Guide', last accessed on 30 January 2013
Potential impact:
A team has been established, working in the total respect of the main research principles. Young researchers have been valorised, to support the growth of highly specialised profiles. In this context, no racial, gender, religious of political discrimination has been admitted in the selection of the new personnel. All the principles and dictates of the working discipline have been respected, for the employees and the environment work, both for the Politecnico and the companies (subcontractors) involved. No one has been somehow forced to act against his/her willing or political/ethical/religious precepts. On the contrary, differences have been valorised and harmonised within the group, as a form of cultural enrichment.
As in the tradition of a free and independent University, no rigid daily scheduling has been enforced. The group members had at their disposal all the most advanced communication tools to help them staying constantly connected with the rest of the group without having to commute daily or move permanently from their residence, causing discomfort for themselves or for their families. This is considered a strong measure to reduce the work cost, help reconcile work and private life and promote gender equality.
As far as the dissemination of project results is concerned, the group has worked respecting the principles of rights and duties of a research centre such as the Politecnico di Torino, which is to divulge and spread the most advanced knowledge. The group firmly believes in the importance to promulgate advance courses and lectures for highly specialised profiles (PhD students) which are expected to be absorbed by big industries, as well as small and medium sized enterprises (SMEs). As in the tradition of the group, moreover, dissemination has been pursued through technical and scientific reports as well as the attendance of international conferences.
The foreground results are primarily the algorithms, the software, the models and the methods implemented to obtain the expected results. Dissemination during the project has been organised with two main objectives:
1. to recruit and train researchers who had actively supported the MAS_LAB project or parts of it: in this case dissemination has been carried out through dedicated meeting during which the project documentation and the relevant bibliography has been made available, discussed and commented;
2. to communicate effectively and ensure common understanding among the parties involved in the MAS_LAB project: in this case a glossary has been prepared, discussed, approved and included in the first deliverable of the project (D1.1). One of the criticality of the project, in fact, was that the parties had different backgrounds, besides belonging to different countries and kind of structures (academy and industry). The objective difficulties, moreover, was that different producers (for example Boeing and Airbus) have different names for systems with similar functionality. A common base was established among the parties, also, through periodic meetings in which the main theoretical issued were discussed from the different perspectives.
During the project communication with users outside the project has been mainly pursued at an academic level, within undergraduate courses in the flight mechanic area. In particular, a four-hour seminar has been prepared and proposed to the BSc students within the course aircraft guidance and control (eight credits). The seminar AFCS for modern civil aircraft is focussed on the most practical issued related to the AFCS use, especially in the category of the large aircraft. The outline is the following:
1. AFCS functions in civil aircraft - SAS, CAS and autopilots: the command and control systems of two large aircraft (Boeing 474-100 and Airbus 330-300) are described in details and the main differences are outlined. Autopilots modes, their interaction and their limitations are presented. The simplified control schemes of the main autopilots are introduced and commented.
2. Conceptual differences for the main constructors (Boeing and Airbus)
3. Pilot interface with the AFCS: the interfaces for the main constructors (Boeing and Airbus) are presented. Flight director (FD), mode select panel (MSP) or mode control panel (MCP), FCU, (multipurpose) control display unit, FMS are discussed in details
4. Typical mission profiles and interaction with the AFCS: practical examples of Boeing 747-100, Boeing 747-400, Airbus 330-300.
As outlined in the list of dissemination activities many of the topics addressed during the project had been the subject of undergraduate theses. In particular four MSc and three BSc theses have been completed. Their content will be briefly described hereinafter:
1. Title: Modelling and validation of the B747-100 AFCS suite for the implementation of an innovative FMS
Author: Mario Cassaro
Faculty Advisor: Manuela Battipede
Abstract: This report is the result of a graduation research, which has been carried out within the Clean Sky project for the development of MAS_LAB. The main objectives of this work were to verify and validate a B747-100 nonlinear aircraft dynamic model and also to implement and validate the AFCS. To achieve the proposed objectives it has been decided that classical control techniques would be used to develop the autopilot controllers. The steps to achieved the proposed objectives can be listed as follows:
1. non linear model implementation, validation and trimming;
2. model linearisation and dynamic response investigation;
3. stability and control augmentation system design;
4. control logic development;
5. non-linear testing of the complete model.
Title: Mathematical model turboprop aircraft for the implementation of an innovative FMS 5L
Author: John Paronitti
Faculty Advisors: Manuela Battipede, Paolo Maggiore
Abstract: This dissertation is the outcome of a eight-month long research work carried out at the Politecnico di Torino's Aerospace Engineering Department, whose aim was the implementation of a turboprop aircraft mathematical model that will be used for the development of an innovative FMSC The first objective of this dissertation was the calculation the aerodynamic coefficients of a simplified geometric model of a modern commercial turboprop aircraft such as the ATR 42-500 through the United States Air Force-developed commercial software digital DATCOM. Due to the inherent limitations of the DATCOM, it was very important to determine the plausibility of the results of its output file, they have therefore been compared with similar plots found in literature. The second objective was the development of the propulsion, aerodynamic, flight controls and dynamic models of the ATR 42 that had to be implemented as modules into the MAS_LAB master model, in order to calculate the forces and the moments the aircraft is subject to.
Title: Implementation of a mathematical model for the AFCS analysis for a large aircraft
Author: Davide Bietto
Faculty Advisors: Manuela Battipede, Paolo Maggiore
Abstract: This thesis describes the work carried out within the Clean Sky MAS_LAB project In particular, the task carried out and reported here concerns the implementation of the flight control systems (longitudinal, lateral and directional controls, flaps and stabiliser) and the propulsion system of the Boeing 747-100 and the validation of these models at a subsystem level. It should be noted that all the work presented here, concerns one of the most complex model in the state-of-the-art and the first model implemented and tested in the MAS_LAB project. This thesis, hence, represent a reference for the other models (ATR42-500, CJ3 and the Generic) to be implemented verified and validated. The reader should be familiar with classical control theory as well as with the flight mechanics and dynamics main concepts. Matlab and Simulink are the basic modelling tools.
Title: Non linear dynamic modelling of the Cessna Citation CJ3 and implementation of the relevant autopilots
Author: Francesco Trifilò
Faculty Advisor: Manuela Battipede
Abstract
This thesis has been developed within the European research project called Clean Sky and was carried out at the Department of Mechanical and Aerospace Engineering (DIMEAS) of the Politecnico di Torino. The main objectives consisted in implementing and validating the Cessna Citation CJ3 nonlinear dynamic model and the relevant AFCS. The V&V process has been performed using specifications reported in MIL-DTL-9490E. The implemented model of the aircraft is part of the flight simulator platform called MAS_LAB. The steps involved in the development of the project are as follows:
1. implementation of the geometric model and its aerodynamic database obtained using DATCOM;
2. model linearisation and dynamic response investigation;
3. stability and control augmentation system design;
4. V&V of the complete non-linear model.
Title: Validation of a turbojet engine dynamic model for a for fuel consumption analysis
Author: Daniele Mazzotta
Faculty Advisors: Manuela Battipede, Michele Ferlauto
Abstract: The main objective of this BSc Thesis was to validate the lumped element mathematical model of the Pratt&Whitney JT9D-3A turboprop engine, which equips the B747-100. The engine mathematical model is intended primarily for the evaluation of the trust and fuel consumption in the whole aircraft flight envelope and is part of a Clean Sky project, for which the main aim is to implement and validate a flight simulator, the MAS_LAB. The validation process consists in the identification of the coherence of data calculated through the use of commercial software such as the GSP developed by the NLR. These data in fact are extrapolated from the data available in the current literature, which is relevant to particular flight conditions, such a full throttle or zero speed. The behaviour of data outside this flight condition has been analysed and corrected at a subsystem level and with the model integrated in the full aircraft simulator.
Title: Implementation of a virtual FCU panel for a flight simulator
Author: Marco Sasanelli
Faculty Advisors: Manuela Battipede, Matteo Vazzola
Abstract: The objective of this BSc thesis is to implement a software interface for MAS_LAB, with the same layout and the functionalities of the FCU panel. In particular the panel is used as a human-machine interface, to set autopilot parameters such as cruise speed, vertical speed or flight path angle; heading angle; altitude; autopilot mode transition; autothrottle mode transition. The thesis is divided into three parts: the first part is dedicated to a description of the commercial software LabVIEW and in particular to the add-on application Simulink interface tool (SIT) through which the FCU panel has been implemented, interfaced and made run simultaneously with the MAS_LAB master model. The second part consists in the description of the features and functionalities of the FCU Panel developed for the MAS_LAB project. The third part is dedicated to the description of the programming details. Each virtual element is analysed: criticalities in the implementation are discussed and validation tests are presented. A final test consists in the evaluation of the performance of the FCU panel interface when functioning in real time as an embedded software.
Title: Feasibility analysis of a parametric mathematical model of a conventional aircraft
Author: Laura Novaro Mascarello
Faculty Advisors: Manuela Battipede
Abstract: The main objective of this BSc thesis is to provide a parameterisation of some geometric features of a generic aircraft, identifying mathematical relationships between the geometric features and several aircraft performance parameters. The choice of the parameters to be analysed is not random: on the contrary it is based to an appropriate strategy, partially echoes the preliminary design philosophy of the aviation companies when faced with a new project. The following steps were accomplished:
1. statistical analysis over the geometrical and performance data of aircraft belonging to the categories of large aircraft, business jet and regional aircraft
2. detailed comparison between geometric and performance data of similar large aircraft made by the two most important aviation companies Boeing and Airbus
3. partially parameterisation of the flight control block implemented in MAS_LAB using Matlab/Simulink programming language
4. identification of the influence of bounded gross weight variation over the geometrical and performance parameters of a generic aircraft.
As far as the exploitation of results is concerned, results have been made available for all the JU partners: the simulator modularity makes the platform suitable for different types of aircraft with different operational goals within and beyond the scope of the Clean Sky JTI.
List of websites:
Piero Gili
Mechanical and Aerospace Department (DIMEAS)
Politecnico di Torino
Corso Duca degli Abruzzi, 24
10129 Torino (TO)
Tel. +39-011-5646854
Fax. +39-011-5646899
piero.gili@polito.it
http://www.polito.it
Manuela Battipede
Mechanical and Aerospace Department (DIMEAS)
Politecnico di Torino
Corso Duca degli Abruzzi, 24
10129 Torino (TO)
Tel. +39-011-5646874
Fax. +39-011-5646899
manuela.battipede@polito.it
http://www.polito.it
Michele Visone
BLUE Group Engineering and Design
Via Coroglio, 57
c/o BIC città della scienza
80124 Napoli (NA)
Tel. +39-081-6171534
Fax. +39-081-7352433
m.visone@blue-group.it
bluenapoli@blue-group.it
http://www.blue-group.it
