Skip to main content

Multi-Objective Robust Assessment of heLicopter Improvements

Final Report Summary - MORALI (Multi-Objective Robust Assessment of heLicopter Improvements)

Executive Summary:
The MORALI project took place from 2011-2014 and was embedded in the Cleansky Joint Technology Initiativen. It answered a call for proposals resulting from requirements defined by Eurocopter Germany, now Airbus Helicopters Germany. The proposal consortium consisted of two parties: University of Stuttgart from Germany, represented by the helicopter and aeroacoustics group of the Institute for Aerodynamics and Gasdynamics, as the coordinator and taking responsibility for aerodynamic simulations on different fidelity levels, and as the second partner MACROS Solutions from Bulgaria, taking care of advanced optimisation technology in a general framework.
The goal of the project was to establish helicopter rotor design capabilities at significantly advanced levels. Although other aspects as structural stability, mass and construction cost are finally certainly of great importance as well, the rotor as the main means of generating lift, propulsion and control is mostly aerodynamics driven, so the project concentrated on this aspect only, considering other factors only by appropriate boundary conditions. Of course, later on those other factors can be included as well, as soon as models become available and are accessible to numerical treatment.

Even on the aerodynamic side along the complexity of the problem is tremendous. About 20-30 design variables at least allow a fine-tuning of a given rotor to get optimal performance. However, “optimal performance” is difficult to define, as several flight states (hover, forward flight at different speeds, start and approach) have to be taken in to account, on terms as power requirements, loads and acoustics. Furthermore, a reliable evaluation at all degrees of freedom necessitates extremely demanding computational fluid dynamics, allowing only for a literally handful of variants to be considered in detail. Other approaches, as blade element theory or free wake simulations, are several orders of magnitude faster, but may mispredict the performance dependence on some parameters due to missing physical phenomena.
As stated, the University of Stuttgart took responsibility for the aerodynamic simulation on all modelling levels, enhancing the tools in order to improve the physical modelling (dynamic stall for blade element theory, transition prediction for CFD), to enable another method (free-wake model) or to boost performance (trim acceleration). Another task was the appropriate definition of performance – at least in view of power input – for a reliable differentiation between variants, where a new optimisation goal could be defined to drive the process. However, the exact balance between power at different speeds and acoustics is a strategic decision of industry, specific for a certain product with defined missions, and thus delegated to them, in this case Airbus Helicopters.
The optimisation procedure itself was taken care of and targeted by MACROS Solutions, who adapted their optimiser to the specific needs of MORALI. They regard the problem as high-dimensional, with various function evaluations at different reliability levels, from low (depending on the parameter, but very cheap) to high (and extremely costly). The tool can generate a new parameter set, which is then to be evaluated aerodynamically, and the result fed back to the optimiser to drive the process further. The goal here is to blend and integrate the different reliability and cost levels to create the best design with minimal computational effort.
In summary, the project was a huge success. In spite of some intermediate technical problems and political difficulties, all deliverables were deployed and all mile stones reached. Indeed, the technical program was more than fulfilled.. The improved tool chain is now in active use at Airbus Helicopters and supports the development of upcoming new products.

Project Context and Objectives:
The MORALI project started in January 2011 and lasted for a duration of 48 months until December 2014. The strategic goal of MORALI was to improve the helicopter rotor design capability, including comprehensive analysis and evaluation skills of different designs, simulation competence at various modelling levels, and automated optimisation support.

Context

The development of a new helicopter rotor is – as any aerospace design task – clearly a multi-disciplinary and also multi-objective problem. While aerodynamic efficiency is obviously very important, other factors like structural stability, dynamic compatibility to the rest of the system, manufacturing issues and, last but not least, cost, do also influence the outcome. All these issues have to be taken into account before successful industrialisation of a specific configuration, consolidating the combined expertise of engineers experienced in different disciplines. An automated process chain delivering reliable results helps those engineers to explore substantially more effective the huge design space of a rotor blade geometry. Such tools significantly support the development of new products, in order to generate better results in a shorter time frame.
As in probably any engineering process, no analytical solution to the design problem is available, and the task is to balance the differing and sometimes contradicting driving forces. Even taking into account only aerodynamic issues, the requirements on different parts of the rotor in various flight conditions are sufficiently diverse to necessitate such delicate balancing.
In order to predict the performance of new rotor designs, different simulation technologies at various confidence levels are available. Momentum theory, taking only mass and radius into account, gives only first impression for prelimiinary rotor sizing and is not used further in this project, while blade element theory at least uses geometric information of the rotor blades to give a more detailed view into the aerodynamic behaviour in a time frame of seconds. Obviously, some major phenomena are not represented in such a simplified model, and thus ask for more elaborate technology to successfully generate a detailed design.
Especially the wake and thus effective rotor inflow is coarsely modelled as more or less constant, which is the reason to use a free-wake approach. The variable circulation along the blade radius and during rotation generates vorticity, shed from the trailing edge to build up the wake. This vorticity is now convected freely along the inflow as well as its own induction, and thus gives a detailed local representation of the wake structure. For example, blade-vortex interaction phenomena and their accompanying noise emission can be obtained quite successfully with such a free-wake analysis, which consumes computing time on the order of hours.
However, the fully non-linear, non-stationary and three-dimensional flow structure can be only captured by solving the Navier-Stokes equations (in the Reynolds-averaged sense only due to the high Reynolds numbers, of course) in computational fluid mechanics. Of course, the aerodynamics have to be coupled to a structure dynamics simulation in order to take blade elasticity and deformations into account (which also improves blade element theory and free-wake significantly), and the collective and cyclic control angles have to be adjusted to produce the forces and moments targeted. In the end, such simulations are able to reproduce experimental values as power requirements, deformations and noise in sufficient quality to guarantee working designs before any hardware is built. The price for this data quality are substantial computational resources, which reach several ten thousand core hours for a single rotor in a single state of flight. More elaborate configurations, including the full helicopter with fuselage, empennage and tail rotor can even run into several millions of core hours for a trimmed free flight.
All in all, we have at our disposal three simulation models of increasing accuracy and fidelity, from blade element theory to the free wake approach until fully coupled CFD, each adding about four orders of magnitude of computational effort to the previous one.
Furthermore, the geometric design of a rotor blade offers many options, from different airfoil sections along the radius to planforms with sweep and taper, not to mention anhedral or dihedral and finally twist. Even though older blades seldomly were more complicated than an extruded airfoil with some linear twist (and adding some nice tip), better efficiency and acoustic behaviour can only be obtained by more elaborate geometries. Taking just a forward/backward-swept blade as an example, we find easily some 20 parameters to describe such a shape. Obviously, a parameter space consisting of 20 degrees of freedom or dimensions is hardly to be searched exhaustively, especially if a single evaluation takes thousands of core hours, as in the CFD case. Consequently, instead of a blind search in this huge space, an optimisation tool chain helps the qualified engineer to take a well-defined way of successive improvements.
This includes blade element theory, which represents the twist distribution quite well, for example, while leaving other parameters to a – quite limited – number of CFD simulations. Even then, successive searches in 2-3 dimensional subspaces proved more efficient than tackling the full space at once, as general trends and conflicting goals can such be much more easily visualised and understood. Acoustics, as another optimisation goal, may be included by means of free-wake, to look for BVI conditions, for example, that are especially significant in this regard, but not a-priori known for a specific rotor.
To sum it up, even only the aerodynamic design of a rotor blade geometry is a very delicate and demanding process. The high dimensionality of the parameter space in conjunction with the very costly evaluation necessitates a very clever combination of different optimisation approaches. Several fidelity levels, with their reliability depending on the parameter under investigation, ask for their integration into a common tool set, in order to help a competent engineer accomplishing its task. For the time being, this task remains difficult, but the tools developed in MORALI give a decisive impetus to bring about better performing rotor blades in less time and with smaller effort.

Objectives

The overall strategic goal of the MORALI project was – as already stated – to improve the rotor design capability in industry. Itemised into sub-objectives progress on several work areas was expected:
1. Improvement of fast design methods to include more physics and thus to enhance prediction quality.
2. Acceleration of high-fidelity methods to enable earlier adoption in the design process for a better understanding of detailed phenomena.
3. Enhancement of the optimisation tool chain to make a more automated work flow feasible.
4. Assessment of the potential of innovative rotor and control concepts.
5. Application of all the tools to an industrially relevant case.
6. Implementation in an industrial context.

Structure

There were four work packages in the project. Nearly all of them (besides WP1: Management) were split into several tasks and partly subtasks.
The first work package, WP1, dealt with the management of the project and included all related tasks, as negotiation and communication with the JU, industry and the partner, knowledge dissemination and results exploitation, documentation and financial issues.
The following work package, WP2, brought forward Method Development in three different fields, namely BEM and CFD, respectively, on the simulation side and additionally the optimisation procedure. Each of these tasks consisted of different subtasks, tackling a specific area identified as in need of improvement, respectively. On the boundary element side, dynamic stall was the most important effect unconsidered yet, in addition to the local wake development by vortex tracking. For CFD it was transition modelling and further refined control of the trim procedure. The multi-objective property of the multi-dimensional optimisation process asked for new concepts regarding effective and efficient handling of the information generated.
In contrast, WP3 was concerned with the assessment of simulation results, as produced by the tools improved in WP2. It has proven insufficient to look only at singular point values as, for example, a figure of merit, especially in an automated optimisation framework. Instead, the entire information provided by a detailed simulation needed to be taken into account, giving the very vague concept of “optimal performance” a quantifiable form. Acoustics come into play here, adding another direction besides fuel consumption. Either way this includes an appropriate consideration of errors and strategies to bound them.
WP4 finally applied the full tool set to a rotor optimisation problem specified by the ITDL. As for the optimisation process several simplifications are necessary at different modelling levels, the outcome was to be assessed in terms of validity and accuracy by proven and trusted high-fidelity simulations.

Project Results:
The detailed results of MORALI are described with the PDF version of the final report, including figures.

Potential Impact:
The general purpose of the CleanSky JTI is to support European aerospace industry in competitiveness and sustainable growth by enabling cleaner technologies for air transport. More specifically, the MORALI project was the response to a call for proposals created to support Airbus Helicopters (then Eurocopter) as the ITD leader relevant for MORALI.
So the main driver for MORALI was the supporting action for the optimisation tool chain for future helicopter rotor blades on demand of Airbus Helicopters. Consequently, the expected impact was to considerably improve the design capabilities of the ITDL for rotorcraft with respect to improved performance, efficiency and noise generation.
All goals described in the call for proposals have been reached, with respect to technology improvements (dynamic stall treatment, transition prediction), tool developments (free wake creation, design parameter driven automatic mesh generation, optimiser integration), result production (trim acceleration, periodicity detection, goal oriented evaluation) and finally application. Of course, no specific blade design for a real future helicopter product has been made within MORALI, as this process involves many more relevant technologies and dependencies on other than aerodynamic properties, and furthermore contains considerable valuable intellectual property of the ITDL just by specifying goals and boundary conditions, but the process chain has been implemented and established at industry, in order to be applied for the next rotor generations. So MORALI´s impact is the support in improved performance, reduced operation cost and noise emissions.
This industrialisation of technology and tools is the primary impact of MORALI, according to the definition in the call for proposals and the work plan negotiated at the beginning. However, the knowledge generated is also available now at the project partners for future research, furthering the capabilities in any forthcoming projects. For example, at IAG a completely new project, funded by the DFG (Deutsche Forschungsgemeinschaft), is dedicated to research on dynamic stall within the rotating system, which would not have been possible without the fundamental work carried out within MORALI. So, besides the industrial exploitation, even at our university a new job (another PhD candidate) was created in reaction to the MORALI effort. More projects are planned, hopefully leading to further improvement of the simulation capabilities at IAG and increased understanding of aerodynamic phenomena in the context of rotorcraft, for example in the context of the just started JTI CleanSky2.
Besides, the general ideas regarding the underlying technologies have been introduced at several conferences and published in international journals. Although some very specific details, especially at the application level, had to be spared to protect sensitive intellectual property rights of Airbus Helicopters, the concepts and their interaction and interfacing are publicly availably for other interested parties as well. Thus the MORALI project has contributed to a significant increase in the general applicability of simulation technology for rotorcraft.

Specifically, the following conference contributions and journal papers directly resulted from work done within MORALI:
1. Martin Hollands and Manuel Keßler and Andree Altmikus and Ewald Krämer. Trade Study: Influence of Different Blade Shape Designs on Forward Flight and Hovering Performance of an Isolated Rotor (European Rotorcraft Forum, Gallarate, 2011)
2. Martin Hollands and Manuel Keßler and Ewald Krämer. Influence of An-/Dihedral and of Different Blade Shapes on Performance and Aeroacoustics of an Isolated Rotor (European Rotorcraft Forum, Amsterdam, 2012)
3. Patrick Kranzinger and Martin Hollands and Manuel Keßler and Sigfried Wagner and Ewald Krämer. Generation and Verification of Meshes Used in Automated Process Chains to Optimize Rotor Blades, (50th AIAA Aerospace Sciences Meeting, Nashville, 2012)
4. Hollands, Martin and Keßler, Manuel and Krämer, Ewald. Planform Design for a Five Bladed Isolated Helicopter Rotor Using Fluid-Structure Coupled CFD Simulations (31st AIAA Applied Aerodynamics Conference, San Diego, 2013)
5. Martin Hollands and Manuel Keßler and Ewald Krämer. Blade Shape Design: Interpolation Based on Fluid-Structure Coupled Simulations of an Isolated Rotor in Forward Flight (in: Dillmann, A., Heller, G., Krämer, E., Kreplin, H. P., Nitsche, W., Rist, U. (Hrsg.): Notes on Numerical Fluid Mechanics and Multidisciplinary Design, Vol. 124, New Results in Numerical and Experimental Fluid Mechanics IX. Springer, 2014)
6. A. Klein and Th. Lutz and E. Krämer and K. Richter and A. D. Gardner and A. R. M. Altmikus. Numerical Comparison of Dynamic Stall for Two-Dimensional Airfoils and an Airfoil Model in the DNW–TWG (Journal of the American Helicopter Society, Vol 57, No. 4, 2012)
More papers regarding MORALI technology or results are currently either in work or in the review process. Tools improved within MORALI (ACCO, the acoustic postprocessing tool), have been industrialised and several licenses sold to third parties. Further commercial exploitation of MORALI is within the responsibility of Airbus Helicopters.

List of Websites:

http://www.iag.uni-stuttgart.de/abteilungen/helikopter/Projekte_Heli/Preojekte_Helikopter.html

Related documents