View Video Presentation: https://doi.org/10.2514/6.2023-3314.vid By virtue of their efficiency, gradient-based algorithms can play a profound role in assessing wing designers overcome the challenges of decarbonizing the aviation industry. This role becomes more valuable when non-linearities increase in the corresponding physical discipline. The coupled aero-elastic adjoint approach, is used to efficiently compute the gradients of aerodynamic cost functions with respect to shape design parameters. This approach helps aerodynamic designers engage the effects of wing structure flexibility, earlier in the design process. To compute the gradients via the coupled aero-elastic adjoint approach, a fixed-point iteration between the aerodynamic adjoint equation and the structure adjoint equation is carried out. The cost of computing the coupled aero-elastic adjoint for a specific cost function is, hence, higher than that of computing the aerodynamic adjoint. In practice it costs three to ten times more, based on the structure flexibility. The aircraft industry tends currently to build wings with higher structure flexibility. This is the result of using composite materials instead of metal, in order to save structural mass, or advancing towards higher aspect ratio wings in order to improve the aerodynamic performance. These trends, which result in increased wing flexibility, motivate investigating the benefit of the coupled aero-elastic adjoint approach, over the aerodynamic adjoint approach. In the literature, some studies show the importance of employing the coupled adjoint, while others show barely any benefit for it. Hence, the aim of this study is to investigate, for the test case at hand, the relation between the wing flexibility and the necessity to employ the relatively costly, but accurate, coupled aero-elastic adjoint approach. Five structure models were generated with different elasticities and were employed in two sets of optimizations; the first uses the aerodynamic adjoint to compute the gradients, and the other uses the coupled aero-elastic adjoint. This investigation was done for unconstrained as well constrained optimizations. The investigation showed that for the unconstrained optimization and for the design task at hand, a structure flexibility, that results in a normalized wing-tip deformation of 8% or higher, requires the use of the coupled aeroelastic adjoint. The normalized wing-tip deformation is defined, in this context, as the upward wing-tip deformation divided by half the wing span. For the constrained optimization, the investigation showed that the coupled aeroelastic adjoint always brought more improvements than the aerodynamic adjoint.

Composite structures have shown a prominent impact in the aircraft structural design. With an increasing shift towards incorporating more composite materials in the primary aircraft structure it is imperative to have corresponding design tools to simplify the design process. In the present work, a simplified implementation for composite optimization has been developed within the DLR-AE (German Aerospace Centre, Institute of Aeroelasticity) automated aeroelastic structural design framework cpacs-MONA. This paper presents the results of structural optimization of a high aspect ratio composite wing aircraft model developed in the DLR project ATLAs. The generation of almost all involved simulation models for this study is done using the in-house DLR tool ModGen. An aeroelastic trim analysis is conducted for various manoeuvre and gust conditions. A load selection process is used to determine the most relevant sizing load cases. A comparison is made between the optimization results of a composite wing and an aluminium wing to demonstrate the more favourable strength to weight ratio of the composite wing. A manoeuvre load alleviation procedure has been introduced in the load calculation process. The results show further weight savings in the design process when load alleviation is utilized due to reduction in the span wise bending moment.

Industry-relevant multidisciplinary design and optimization (MDO) processes engage numerous computationally intensive, disciplinary analysis and design, tasks. This motivates not only the use of high-performance computing resources for tackling the design and optimization problems, but also the search for efficient MDO algorithms and formulations. On the algorithmic side, the MDO chains developed in this work are gradient-based. Such algorithms are known for their efficiency in finding optima when compared to gradient-free algorithms. Concerning the formulations, several studies in the literature showed the effect of MDO formulation on the convergence efficiency. However, these studies used mostly model problems rather than realistic aircraft design problems. The target of our study is to understand the effect of MDO formulations on the convergence efficiency and the improvement of a realistic MDO task, and to compare the results to a promising data-driven scalable methodology for testing MDO formulations. Two MDO formulations are highlighted and investigated. The first one is driven by one optimizer that controls all design parameters. The second is driven by two optimizers; one responsible for the cruise flight performance analysis and the other responsible for the structure sizing process. The use of commercial solvers forces neglecting part of the cross disciplinary gradients in both realistic MDO tasks. The work is performed on the research aircraft configuration; AIRBUS’ XRF-1. The results of the realistic optimizations showed that engaging two optimizers is more efficient, for the given optimization task. The data-driven scalable methodology approach showed that the one optimizer formulation is more efficient when the exact set of gradients is considered.

By virtue of using efficient methods to compute the design sensitivities, such as the coupled adjoint methods, gradient-based optimization techniques allow aircraft designers to efficiently obtain an optimum design, that satisfies all considered constraints. The more constraints and disciplines engaged, the higher the reliability of the optimization output. Decision makers in aircraft industry, however, prefer to look at Pareto fronts rather than one optimum design before making their critical decisions, since such diagrams provide better understanding of the main trade-offs and compromises between the different targets that the aircraft is supposed to satisfy. Gradient-free algorithms have better reputation for generating Pareto fronts because these fronts result naturally. These algorithms are, however, relatively inefficient, even for single disciplinary high-fidelity optimizations. They are restricted by the number of design parameters and the size of the design space. Engaging more disciplines, more design parameters and higher fidelity computational models, to increase the reliability of the outputs, can only increase the computational cost of gradient-free algorithms, and quickly announce them to be not usable for the generation of Pareto fronts. This study investigates the generation of Pareto fronts efficiently, using gradient-based algorithms on the industry-relevant aircraft; AIRBUS XRF-1. The approach starts several optimizations with different weighting of objectives, in parallel. The multidisciplinary design chain, which is used in this study, was connected by experts in aerodynamics, structure and loads, propulsion and overall aircraft design. The generation of Pareto front was completed successfully for a multi-point multidisciplinary optimization task, where the use of gradient-free algorithms would have been computationally infeasible. As expected and mentioned in the literature, some regions of the Pareto front were not satisfactorily covered by employing this approach, which shows the necessity to look into different gradient-based approaches that search for optimal designs along the Pareto front.

The purpose of this paper is to investigate the influence of the engine position and mass as well as the pylon stiffness on the aeroelastic stability of a long-range wide-body transport aircraft. As reference configuration, DLR’s (German Aerospace Center/Deutsches Zentrum für Luft und Raumfahrt) generic aircraft configuration DLR-D250 is taken. The structural, mass, loads, and optimization models for the reference and a modified configuration with different engine and pylon parameters are set up using DLR’s automatized aeroelastic design process cpacs-MONA. At first, the cpacs-MONA process with its capabilities for parametric modeling of the complete aircraft and in particular the set-up of a generic elastic pylon model is unfolded. Then, the influence of the modified engine-wing parameters on the flight loads of the main wing is examined. The resulting loads are afterward used to structurally optimize the two configurations component wise. Finally, the results of post-cpacs-MONA flutter analyses performed for the two optimized aircraft configurations with the different engine and pylon characteristics are discussed. It is shown that the higher mass and the changed position of the engine slightly increased the flutter speed. Although the lowest flutter speeds for both configurations occur at a flutter phenomenon of the horizontal tail-plane outside of the aeroelastic stability envelope.

Within the DLR project VicToria various high fidelity-based MDO processes were set-up as applicalble methods for aircraft design. Apart from aerodynamic optimization using high fidelity-based CFD analysis, the sub-processes overall aircraft design synthesis, loads analysis, and structural optimization were part of the MDO processes. The presented paper expounds such MDO sub-processes in order to exhibit their contributions and capabilities for the respected MDO process.

Cybermatrix is a novel approach to aircraft design through multidisciplinary optimization, developed within the DLR project VicToria. It combines three aspects: representing a design problem by an approximate Karush-Kuhn-Tucker system, distributing the rows of the system among disciplinary groups, and employing large computational resources and many humans experts in a parallel fashion. For demonstration an optimization of a long-range, twin-engine transport aircraft has been performed.

It is assumed, that the presence of nonlinear effects in transonic flow influences the aerodynamic load distribution and hence must be considered in the aeroelastic loads analysis of maneuvers. In this study a steady pull-up and two quasi-static rolling maneuvers are examined using the Reynolds-Averaged-Navier-Stokes (RANS) equations and compared to the vortex lattice method (VLM). In contrast to the RANS equations, the VLM is widely used for fast analysis of many load cases but does not include transonic effects. The pitch and roll rates are accounted for by the flow solver. A fully elastic transport aircraft configuration is analyzed. Elastic deformations are calculated by the modal approach and are loosely coupled with the aerodynamic model.

The DLR project VicToria brings together disciplinary methods and tools of different fidelity for collaborative multidisciplinary design optimization (MDO) of long-range passenger aircraft configurations, necessitating the use of high-performance computing. Three different approaches are being followed to master complex interactions of disciplines and software aspects: an integrated aero-structural wing optimization based on high-fidelity methods, a multi-fidelity gradient-based approach capable of efficiently dealing with many design parameters and many load cases, and a many-discipline highly-parallel approach, which is a novel approach towards computationally demanding and collaboration intensive MDO. The XRF-1, an Airbus provided research aircraft configuration representing a typical long-range wide-body aircraft, is used as a common test case to demonstrate the different MDO strategies. Additional results are presented for the NASA Common Research Model (CRM) to show their flexibility. Parametric disciplinary models are used in terms of overall aircraft design synthesis, loads analysis, flutter, structural analysis and optimization, engine design, and aircraft performance. The different MDO strategies are shown to be effective in dealing with complex, real-world MDO problems in a highly collaborative, cross-institutional design environment, involving many disciplinary groups and experts and a mix of commercial and in-house design and analysis software.

Over the past decade, profound attention was given to exploring the benefits of engaging numerical multidisciplinary design optimization in aircraft design. Due to its importance, aerostructural wing design optimization is the most visited multidisciplinary problem in research institutes. To deal with this problem efficiently, gradient-based algorithms are popularly used. The complexity of the gradient-based aerostructural optimization, however, forced researchers to apply several simplifications to the problem formulation, such as neglecting engine effects or oversimplifying the loads process into few predefined load cases. The authors of this paper aim at running a gradients-based multidisciplinary design optimization of a commercial aircraft while including a powered engine and engaging a comprehensive, multi-fidelity loads process, subject to flutter as well as overall aircraft design constraints. The work, done by experts in the mentioned fields, is performed on a commercial aircraft provided by Airbus with many industry-relevant constraints. The results, showed the necessity to include a comprehensive loads process during the optimization. Additionally, it was concluded that engaging powered engines during the optimization is inevitable to come up with realistic designs; significantly different design geometries resulted when engaging a powered engine than when running the optimization with a flow-through nacelle.

The highly parameterized process cpacs-MONA for structural and aeroelastic design of aircraft configurations is presented. The parameterized design process is based on the Common Parametric Aircraft Configuration Schema (CPACS), developed by DLR, where almost all aircraft parameters like mission, geometry, structure, and material are defined. The process stands out for its independent use for structural and aeroelastic design on the one hand and its integration into high fidelity based multidisciplinary optimization tasks on the other hand. The process consists of preliminary mass and loads estimation based on conceptual design methods followed by a parameterized set-up of simulation models and an optimization model. They are used for a comprehensive loads analysis followed by a component wise structural optimization. The latter takes stress, strain, buckling and control surface efficiency as constraints into account. The structural simulation model consist of shell and beam elements for the wing-like component and the fuselage in order to represent the construction of the load carrying structure of the complete aircraft appropriately. Such depth of modelling allows also for the use of well-established structural optimization methods. MONA stands for ModGen, the in-house parametric model generation computer program, and MSC Nastran, used for loads and aeroelastic analysis as well as for structural optimization. cpacs-MONA is also integrated in MDO tasks, where other disciplines like aerodynamics and overall aircraft design are involved. Therein aerodynamic data from high fidelity CFD calculations are used to improve the vortex lattice based aerodynamic method that is applied of the aeroelastic loads analysis. The MDO tasks where cpcas-MONA is integrated can be gradient-free and gradient-based. For the gradient-based MDO task cpacs-MONA delivers to the system optimization level the structural variables, the structural responses the corresponding sensitivities. As structural optimization task alone can deal with a high number of variables and constraints, for the MDO task with other disciplines, like aerodynamics, methods were developed where the number of constraints is reduced that are taken into account on the system level. Two applications are presented for cpacs-MONA. In the first application cpacs-MONA is applied as independent and stand-alone structural and aeroelastic design process. For a long range wide body transport aircraft configuration the design loads case are estimated after a comprehensive loads analysis campaign. For the structural design especially the structural requirements and alternative strategies to achieve sufficient control surface effectiveness within the flight envelope are investigated. In the second application cpacs-MONA is part of a gradient-free optimization task with aerodynamics and the structure to be optimized simultaneously. The presented results show that the aerodynamic optimum lead on the one hand to a heavier wing mass, but investigating the structural optimization results more closely, also local benefits of the new aerodynamic design regarding the wing structure can be found.

Aviation is undergoing a transformation, fueled by the Corona pandemic, and methods to evaluate new technologies for more economical and environment-friendly flight in a timelier manner and to enable new aircraft to be designed (almost) exclusively using computers are sought after. The DLR project Victoria brings together disciplinary methods and tools of different fidelity for collaborative multidisciplinary design optimization (MDO) of long-range passenger aircraft configurations, necessitating the use of high-performance computing. Three different approaches are being followed to master complex interactions of disciplines and software aspects: an integrated aero-structural wing optimization based on high-fidelity methods, a multi-fidelity gradient-based approach capable of efficiently dealing with many design parameters and many load cases, and a many-discipline highly-parallel approach, which is a novel approach towards computationally demanding and collaboration intensive MDO. The XRF-1, an Airbus provided research aircraft configuration representing a typical long-range wide-body aircraft, is used as a common test case to demonstrate the different MDO strategies. Parametric disciplinary models are used in terms of overall aircraft design synthesis, loads analysis, flutter, structural analysis and optimization, engine design, and aircraft performance. The different MDO strategies are shown to be effective in dealing with complex, real-world MDO problems in a highly collaborative, cross-institutional design environment, involving many disciplinary groups and experts and a mix of commercial and in-house design and analysis software. © 2021 32nd Congress of the International Council of the Aeronautical Sciences, ICAS 2021. All rights reserved.

This paper presents an approach to multi-disciplinary optimization (MDO) of transport aircraft that attempts to strike a balance between two broad classes of MDO approaches: those arising from the formal optimization background, and those coming from the aircraft design background. It starts from the observation that any kind of numerical design process can be viewed as an approximation of a formal optimization process, where Jacobians of cost functions may be inexact and are often not explicitly computed. Based on that, a specific MDO problem representation and a highly parallel process assembly and execution protocol (the “cybermatrix” protocol) is defined, as well as one possible realization on high-performance computing (HPC) resources. The approach is applied to an optimization of a long-range transport aircraft, employing disciplinary subprocesses for high-fidelity aerodynamic design of wing airfoil shapes, structural sizing of lifting surfaces, and determination and evaluation of design loads.

The DLR project VicToria deals also with high fidelity-based MDO with various approaches. Aside from aerodynamic optimization, further sub-processes are part of the developed MDO processes. They belong to overall aircraft design synthesis, loads analysis, and structural sizing and optimization. The Paper lays out the mentioned MDO sub-processes in order show their contribution and capabilities for the selected MDO process, but also their complexity when dealing with a high fidelity based MDO approach.

The DLR project VicToria brings together design methods that necessite use of high-performance computing, for the multi-disciplinary optimization of a long-range transport aircraft. Three different approaches are followed: the integrated aerostructural wing optimization, the multi-fidelity gradient-based approach, and the many-discipline highly-parallel approach. A common test case is used: XRF1, an Airbus provided research configuration representing a typical long-range wide-body aircraft.

Background: The consideration of composite design in an early design phase has become more and more necessary in aircraft design. In the scope of multidisciplinary optimisation, a composite aeroelastic reference model is often desired to investigate multidisciplinary effects on a near-industrial application. Objective: The aim is to generate a benchmark aeroelastic model by developing a robust design process for composite configurations. Method: This is done by using a continuous gradient based optimisation with lamination parameters as design variables followed by a discrete stacking sequence retrieval combined with a comprehensive load analysis routine comprising manoeuvre, gust, landing loads and a manoeuvre loads alleviation system. Simulation Models: The process utilises a GFEM/Dynamic, a DLM and a condensed FEM model based on the reference XRF1 configuration. Results: Results presented show the optimised wing-box design, comparing the effect of a manoeuvre load alleviation system, the effect of including blending constraints in the optimisation, and the change of the structrual design when moving to a stacking sequence design.

DLR's work on developing a distributed collaborative MDO environment is presented. A multi-level Approach combining high-fidelity MDA for aerodynamics and structures with conceptual aircraft design methods is employed. Configuration-specific sizing loads are evaluated and used for sizing the structure. A gradient-free optimization algorithm is used to optimize the fuel burn of a generic long-range wide-body transport aircraft configuration with 9 shape parameters. The results show a truly multidisciplinary improvement of the modified design. The result of a gradient-free high-fidelity MDO with preselected load cases and five shape parameters is also presented, comparing a full mission analysis with results for the Breguet range equation.

A combined aerodynamic and structural, gradient-based optimization has been performed on the NASA/Boeing Common Research Model civil transport aircraft configuration. The computation of aerodynamic performance parameters includes a Reynolds-averaged Navier-Stokes CFD solver, coupling to a linear static structural analysis using the finite element method to take into account aero-elastic effects. Aerodynamic performance gradients are computed using the adjoint approach. Within each optimization iteration, the wing's structure is sized via a gradient-based algorithm and an updated structure model is forwarded for the performance analysis. In this pilot study wing profile shape is optimized in order to study engine installation effects. This setting was able to improve the aerodynamic performance by 4%.

Critical loads play an essential role in early aircraft design studies. In order to determine a consistent envelope of critical design loads, a framework has to be introduced in which the dynamic models for loads calculation and the static models for aircraft structural design have to be evaluated and analyzed in an iterative manner. In the course of the DLR project Digital-X a process was established that generically builds an initial aircraft structural model and sizes it in a global optimization loop using updated dynamic models for calculating the trim, gust and maneuver loads necessary to ensure structural integrity as prescribed by the Federal Aviation Regulations (FAR). In this paper the fully automated aeroelastic analysis, flight loads prediction and structural sizing process is presented that determines the design loads used in the Digital-X multidisciplinary design optimization (MDO) framework.