Books
Marta, A. C.; Suleman, A. (Ed.)
IDMEC, Lisboa, Portugal, 2023, ISBN: 978-989-53599-4-3.
@book{Marta:AeroBest2023,
title = {Proceedings of the AeroBest 2023 - II ECCOMAS Thematic Conference on Multidisciplinary Design Optimization of Aerospace Systems},
editor = {A. C. Marta and A. Suleman},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Book_AeroBest2023_proceedings.pdf, PDF
https://aerobest2023.idmec.tecnico.ulisboa.pt/, WWW},
isbn = {978-989-53599-4-3},
year = {2023},
date = {2023-07-01},
urldate = {2023-07-01},
publisher = {IDMEC},
address = {Lisboa, Portugal},
series = {ECCOMAS Thematic Conference},
abstract = {This Book of Proceedings contains the full manuscripts on the works presented at AeroBest 2023, the second edition of the ECCOMAS Thematic Conference on Multidisciplinary Design Optimization of Aerospace Systems. This document is a testimony of the latest developments in the field, covering a wide range of MDO computational methods, tools and processes and applications spanning civil aviation and space systems. Following the successful debut in 2021, which was held online due to the COVID-19 pandemic, this second edition gathered in Lisbon almost 60 participants, who assisted to 38 presentations, including four keynote lectures by prominent speakers from academia and industry, who shared the state-of-the-art on the computational methods, tools and processes and experimental facilities in the field of study, and outlined the technological development roadmap for the future of sustainable aviation and space technology. The participants from academia, government laboratories and industry highlighted the growing awareness and acceptance that 21st -century aerospace design problems must simultaneously take into account multiple physical disciplines using a seamless coupling and a friendly user-interface, while mitigating the inevitable demanding computational resources. Being a thematic conference, relatively small but highly focused and specialized, resulted in a dynamic environment that fostered interesting and productive exchanges of ideas and discussions. A final word of appreciation to all participants, in particular to those whose work is documented in these proceedings, that made the conference possible.
Keywords: Aerospace Design and Integrated Systems, Systems Engineering and Integration, Discipline Analysis Models, Multi-Disciplinary Optimization, Design Optimization},
howpublished = {https://www.eccomas.org/publications/conference-proceedings/},
keywords = {},
pubstate = {published},
tppubtype = {book}
}
This Book of Proceedings contains the full manuscripts on the works presented at AeroBest 2023, the second edition of the ECCOMAS Thematic Conference on Multidisciplinary Design Optimization of Aerospace Systems. This document is a testimony of the latest developments in the field, covering a wide range of MDO computational methods, tools and processes and applications spanning civil aviation and space systems. Following the successful debut in 2021, which was held online due to the COVID-19 pandemic, this second edition gathered in Lisbon almost 60 participants, who assisted to 38 presentations, including four keynote lectures by prominent speakers from academia and industry, who shared the state-of-the-art on the computational methods, tools and processes and experimental facilities in the field of study, and outlined the technological development roadmap for the future of sustainable aviation and space technology. The participants from academia, government laboratories and industry highlighted the growing awareness and acceptance that 21st -century aerospace design problems must simultaneously take into account multiple physical disciplines using a seamless coupling and a friendly user-interface, while mitigating the inevitable demanding computational resources. Being a thematic conference, relatively small but highly focused and specialized, resulted in a dynamic environment that fostered interesting and productive exchanges of ideas and discussions. A final word of appreciation to all participants, in particular to those whose work is documented in these proceedings, that made the conference possible.
Keywords: Aerospace Design and Integrated Systems, Systems Engineering and Integration, Discipline Analysis Models, Multi-Disciplinary Optimization, Design Optimization
Keywords: Aerospace Design and Integrated Systems, Systems Engineering and Integration, Discipline Analysis Models, Multi-Disciplinary Optimization, Design Optimization
Marta, A. C.; Suleman, A. (Ed.)
IDMEC, Lisboa, Portugal, 2021, ISBN: 978-989-99424-8-6.
@book{Marta:AeroBest2021,
title = {Proceedings of the AeroBest 2021 - International Conference on Multidisciplinary Design Optimization of Aerospace Systems},
editor = {A. C. Marta and A. Suleman},
url = {https://www.eccomas.org/wp-content/uploads/sites/15/2021/11/AeroBest2021_proceedings-1.pdf, PDF ECCOMAS
https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Book_AeroBest2021_proceedings.pdf, PDF
https://aerobest2021.idmec.tecnico.ulisboa.pt/, WWW},
isbn = {978-989-99424-8-6},
year = {2021},
date = {2021-07-01},
urldate = {2021-07-01},
publisher = {IDMEC},
address = {Lisboa, Portugal},
series = {ECCOMAS Thematic Conference},
abstract = {The thematic conference scope encompasses design optimization and inverse problems applied to Aerospace systems, with focus on aircraft and spacecraft. Aerospace design spans from discipline-level conceptual and preliminary design studies up to system level studies of complete aircraft or spacecraft. Multidisciplinary analysis and optimization of conventional and novel configurations is sought. This includes the coupling of disciplines, such as acoustics, aerodynamics, heat and mass transfer, dynamics and control, performance and structural mechanics. The discussion of architectures of discipline coupling is also within the scope of the conference. Focus on numerical optimization algorithms (deterministic or heuristic), discipline analysis models (high-fidelity, low-fidelity and surrogate models) and sensitivity analysis techniques is also desired. The nature of the coupled design problems includes size, shape and topology optimization, using deterministic, reliability-based or robust design optimization approaches..
Keywords: Aerospace Design and Integrated Systems, Systems Engineering and Integration, Discipline Analysis Models, Multi-Disciplinary Optimization, Design Optimization},
howpublished = {https://www.eccomas.org/publications/conference-proceedings/},
keywords = {},
pubstate = {published},
tppubtype = {book}
}
The thematic conference scope encompasses design optimization and inverse problems applied to Aerospace systems, with focus on aircraft and spacecraft. Aerospace design spans from discipline-level conceptual and preliminary design studies up to system level studies of complete aircraft or spacecraft. Multidisciplinary analysis and optimization of conventional and novel configurations is sought. This includes the coupling of disciplines, such as acoustics, aerodynamics, heat and mass transfer, dynamics and control, performance and structural mechanics. The discussion of architectures of discipline coupling is also within the scope of the conference. Focus on numerical optimization algorithms (deterministic or heuristic), discipline analysis models (high-fidelity, low-fidelity and surrogate models) and sensitivity analysis techniques is also desired. The nature of the coupled design problems includes size, shape and topology optimization, using deterministic, reliability-based or robust design optimization approaches..
Keywords: Aerospace Design and Integrated Systems, Systems Engineering and Integration, Discipline Analysis Models, Multi-Disciplinary Optimization, Design Optimization
Keywords: Aerospace Design and Integrated Systems, Systems Engineering and Integration, Discipline Analysis Models, Multi-Disciplinary Optimization, Design Optimization
Journal Articles
Pacheco, J.; Marta, A. C.; Eça, L.
Wind tunnel testing of a Formula Student vehicle for checking CFD simulation trends Journal Article
In: Proceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering, vol. 238, no. 14, pp. 4347–4363, 2024, ISSN: 0954-4070.
@article{Pacheco:2024:JAUTO,
title = {Wind tunnel testing of a Formula Student vehicle for checking CFD simulation trends},
author = {J. Pacheco and A. C. Marta and L. Eça},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_JournalArticle_Pacheco_2024_JAUTO.pdf, Full-text PDF
https://journals.sagepub.com/doi/10.1177/09544070231203076},
doi = {10.1177/09544070231203076},
issn = {0954-4070},
year = {2024},
date = {2024-12-01},
urldate = {2024-12-01},
journal = {Proceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering},
volume = {238},
number = {14},
pages = {4347–4363},
abstract = {The aerodynamic performance analysis of Formula Student racecars has been mostly done by teams with CFD tools, for time and cost savings, that often lack proper validation. To address this, the FST Lisboa team performed a detailed wind tunnel (WT) test campaign, using a one-third scale model, under different configurations, including variable ride heights, bullhorn appendix, and rear wing flap settings, also replicated in CFD. The simulations used RANS with the SST k-omega turbulence model, with a 13.7 million polyhedral mesh for the test chamber region domain. Both experimental and numerical errors were estimated from the instrumentation and mesh convergence analysis, respectively. Comparisons were made between WT and CFD both in terms of local flow, using tufts for flow visualization, and global flow, using lift, drag, and pitching moment coefficients. Overall, the numerical streamlines agreed very well with the orientations of the tufts in experiments, but some discrepancies were found in regions of cross-flow and high-frequency unsteadiness, mainly caused by limitations of the visualization technique. The gamma transition model in CFD was abandoned as it could not replicate the WT observations. In terms of aerodynamic coefficients, a strong correlation was found between WT and CFD. The parametric studies revealed that the simulations captured the experimental sensitivity to each car setting parameter studied but the uncertainties did not enable a full quantitative evaluation of the aerodynamic performance. The drag reduction system significantly impacted the aerodynamic balance of the racecar, while the current bullhorn design proved to be ineffective. The ride height increase led to higher downforce, mostly due to the higher pitch angle of the vehicle, with negligible variation of the aerodynamic balance. This work validated the team CFD studies, building confidence in that trends estimated in numerical parametric studies are likely to be translated to the real prototype performance.
Keywords: Six-component force balance, Flow visualization, Experimental procedure, Numerical errors, Drag reduction system, Ride height, Bullhorn},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
The aerodynamic performance analysis of Formula Student racecars has been mostly done by teams with CFD tools, for time and cost savings, that often lack proper validation. To address this, the FST Lisboa team performed a detailed wind tunnel (WT) test campaign, using a one-third scale model, under different configurations, including variable ride heights, bullhorn appendix, and rear wing flap settings, also replicated in CFD. The simulations used RANS with the SST k-omega turbulence model, with a 13.7 million polyhedral mesh for the test chamber region domain. Both experimental and numerical errors were estimated from the instrumentation and mesh convergence analysis, respectively. Comparisons were made between WT and CFD both in terms of local flow, using tufts for flow visualization, and global flow, using lift, drag, and pitching moment coefficients. Overall, the numerical streamlines agreed very well with the orientations of the tufts in experiments, but some discrepancies were found in regions of cross-flow and high-frequency unsteadiness, mainly caused by limitations of the visualization technique. The gamma transition model in CFD was abandoned as it could not replicate the WT observations. In terms of aerodynamic coefficients, a strong correlation was found between WT and CFD. The parametric studies revealed that the simulations captured the experimental sensitivity to each car setting parameter studied but the uncertainties did not enable a full quantitative evaluation of the aerodynamic performance. The drag reduction system significantly impacted the aerodynamic balance of the racecar, while the current bullhorn design proved to be ineffective. The ride height increase led to higher downforce, mostly due to the higher pitch angle of the vehicle, with negligible variation of the aerodynamic balance. This work validated the team CFD studies, building confidence in that trends estimated in numerical parametric studies are likely to be translated to the real prototype performance.
Keywords: Six-component force balance, Flow visualization, Experimental procedure, Numerical errors, Drag reduction system, Ride height, Bullhorn
Keywords: Six-component force balance, Flow visualization, Experimental procedure, Numerical errors, Drag reduction system, Ride height, Bullhorn
Matos, N. M. B.; Marta, A. C.
Longitudinal motion system identification of a fixed-wing unmanned aerial vehicle using limited unplanned flight data Journal Article
In: Aerospace, vol. 11, no. 12, pp. 959, 2024, ISSN: 2226-4310.
@article{Matos:2024:Aerospace,
title = {Longitudinal motion system identification of a fixed-wing unmanned aerial vehicle using limited unplanned flight data},
author = {N. M. B. Matos and A. C. Marta},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_JournalArticle_Matos_2024_Aerospace.pdf, Full-text PDF
https://www.mdpi.com/2226-4310/11/12/959},
doi = {10.3390/aerospace11120959},
issn = {2226-4310},
year = {2024},
date = {2024-11-01},
urldate = {2024-11-01},
journal = {Aerospace},
volume = {11},
number = {12},
pages = {959},
abstract = {Acquiring knowledge of aircraft flight dynamics is crucial for simulation, control, mission performance and safety assurance analysis. In the fast-paced UAV market, long flight testing campaigns are hard to achieve, leaving limited controlled flight data and a significant amount of unplanned flight data. This work delves into the application of system identification techniques on unplanned flight data when faced with a shortage of dedicated flight test data. Based on a medium-sized, fixed-wing UAV, it focuses on the system identification of longitudinal dynamics using structural routine flight test data of pitch down and pitch up manoeuvres with no specific guidelines for the control inputs given. The proposed solution uses first- and second-order parameter-based models to build a non-linear dynamic model which, using a least square error optimisation algorithm in a time domain formulation, has its parameters tuned to converge the model behaviour with the real aircraft dynamics. The optimisation uses a combination of pitch, altitude, airspeed and pitch rate responses as a measure of model accuracy. Very significant improvements regarding the UAV model response are found when trimmed flight manoeuvres are used, resulting in proper estimation of important aerodynamic and control derivatives. Pitching moment and control derivatives are shown to be the crucial parameters. However, difficulties in estimation are shown for untrimmed flight manoeuvres. Better results were obtained when using multiple manoeuvres simultaneously in the optimisation error metric, as opposed to single manoeuvres that led to system bias. The proposed system identification procedure can be applied to any fixed-wing UAV without the need for specific flight testing campaigns.
Keywords: Digital twin, Aircraft dynamics, Flight control, Flight test, System identification, Aerodynamic derivatives, Control derivatives, Optimisation},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Acquiring knowledge of aircraft flight dynamics is crucial for simulation, control, mission performance and safety assurance analysis. In the fast-paced UAV market, long flight testing campaigns are hard to achieve, leaving limited controlled flight data and a significant amount of unplanned flight data. This work delves into the application of system identification techniques on unplanned flight data when faced with a shortage of dedicated flight test data. Based on a medium-sized, fixed-wing UAV, it focuses on the system identification of longitudinal dynamics using structural routine flight test data of pitch down and pitch up manoeuvres with no specific guidelines for the control inputs given. The proposed solution uses first- and second-order parameter-based models to build a non-linear dynamic model which, using a least square error optimisation algorithm in a time domain formulation, has its parameters tuned to converge the model behaviour with the real aircraft dynamics. The optimisation uses a combination of pitch, altitude, airspeed and pitch rate responses as a measure of model accuracy. Very significant improvements regarding the UAV model response are found when trimmed flight manoeuvres are used, resulting in proper estimation of important aerodynamic and control derivatives. Pitching moment and control derivatives are shown to be the crucial parameters. However, difficulties in estimation are shown for untrimmed flight manoeuvres. Better results were obtained when using multiple manoeuvres simultaneously in the optimisation error metric, as opposed to single manoeuvres that led to system bias. The proposed system identification procedure can be applied to any fixed-wing UAV without the need for specific flight testing campaigns.
Keywords: Digital twin, Aircraft dynamics, Flight control, Flight test, System identification, Aerodynamic derivatives, Control derivatives, Optimisation
Keywords: Digital twin, Aircraft dynamics, Flight control, Flight test, System identification, Aerodynamic derivatives, Control derivatives, Optimisation
Portugal, M.; Marta, A. C.
Optimal multi-sensor obstacle detection system for small fixed-wing UAVs Journal Article
In: Modelling, vol. 5, no. 1, pp. 16–36, 2024, ISSN: 2673-3951.
@article{Portugal:2024:Modelling,
title = {Optimal multi-sensor obstacle detection system for small fixed-wing UAVs},
author = {M. Portugal and A. C. Marta},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_JournalArticle_Portugal_2024_Modelling.pdf, Full-text PDF
https://www.mdpi.com/2673-3951/5/1/2},
doi = {10.3390/modelling5010002},
issn = {2673-3951},
year = {2024},
date = {2024-07-01},
urldate = {2024-07-01},
journal = {Modelling},
volume = {5},
number = {1},
pages = {16–36},
abstract = {The safety enhancement of small fixed-wing UAVs regarding obstacle detection is addressed using optimization techniques to find the best sensor orientations of different multi-sensor configurations. Four types of sensors for obstacle detection are modeled, namely an ultrasonic sensor, laser rangefinder, LIDAR, and RADAR, using specifications from commercially available models. The simulation environment developed includes collision avoidance with the Potential Fields method. An optimization study is conducted using a genetic algorithm that identifies the best sensor sets and respective orientations relative to the UAV longitudinal axis for the highest obstacle avoidance success rate. The UAV performance is found to be critical for the solutions found, and its speed is considered in the range of 5-15 m/s with a turning rate limited to 45°/s. Forty collision scenarios with both stationary and moving obstacles are randomly generated. Among the combinations of the sensors studied, 12 sensor sets are presented. The ultrasonic sensors prove to be inadequate due to their very limited range, while the laser rangefinders benefit from extended range but have a narrow field of view. In contrast, LIDAR and RADAR emerge as promising options with significant ranges and wide field of views. The best configurations involve a front-facing LIDAR complemented with two laser rangefinders oriented at ± 10º or two RADARs oriented at ± 28º.
Keywords: Sense and avoidance, Collision avoidance, Optimization, Ultrasonic sensor, Laser rangefinder, LIDAR, RADAR},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
The safety enhancement of small fixed-wing UAVs regarding obstacle detection is addressed using optimization techniques to find the best sensor orientations of different multi-sensor configurations. Four types of sensors for obstacle detection are modeled, namely an ultrasonic sensor, laser rangefinder, LIDAR, and RADAR, using specifications from commercially available models. The simulation environment developed includes collision avoidance with the Potential Fields method. An optimization study is conducted using a genetic algorithm that identifies the best sensor sets and respective orientations relative to the UAV longitudinal axis for the highest obstacle avoidance success rate. The UAV performance is found to be critical for the solutions found, and its speed is considered in the range of 5-15 m/s with a turning rate limited to 45°/s. Forty collision scenarios with both stationary and moving obstacles are randomly generated. Among the combinations of the sensors studied, 12 sensor sets are presented. The ultrasonic sensors prove to be inadequate due to their very limited range, while the laser rangefinders benefit from extended range but have a narrow field of view. In contrast, LIDAR and RADAR emerge as promising options with significant ranges and wide field of views. The best configurations involve a front-facing LIDAR complemented with two laser rangefinders oriented at ± 10º or two RADARs oriented at ± 28º.
Keywords: Sense and avoidance, Collision avoidance, Optimization, Ultrasonic sensor, Laser rangefinder, LIDAR, RADAR
Keywords: Sense and avoidance, Collision avoidance, Optimization, Ultrasonic sensor, Laser rangefinder, LIDAR, RADAR
Matos, N. M. B.; Marta, A. C.
Concurrent trajectory optimization and aircraft design for the Air Cargo Challenge competition Journal Article
In: Aerospace, vol. 9, no. 7, pp. 378, 2022, ISSN: 2226-4310.
@article{Matos:2022:Aerospace,
title = {Concurrent trajectory optimization and aircraft design for the Air Cargo Challenge competition},
author = {N. M. B. Matos and A. C. Marta},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_JournalArticle_Matos_2022_Aerospace.pdf, Full-text PDF
https://www.mdpi.com/2226-4310/9/7/378},
doi = {10.3390/aerospace9070378},
issn = {2226-4310},
year = {2022},
date = {2022-07-01},
urldate = {2022-07-01},
journal = {Aerospace},
volume = {9},
number = {7},
pages = {378},
abstract = {A coupled aerostructural aircraft design and trajectory optimization framework is developed for the Air Cargo Challenge competition to maximize the expected score based on cargo carried, altitude achieved and distance traveled. Its modular architecture makes it easily adaptable to any problem where the performance depends not only on the design of the aircraft but also on its flight trajectory. It is based on OpenAeroStruct, an aerostructural solver that uses analytic derivatives for efficient gradient-based optimization. A trajectory optimization module using a collocation method is coupled with the option of using b-splines to increase computational efficiency together with an experimentally-based power decay model that accurately determines the aircraft propulsive response to control input depending on the battery discharge level. The optimization problem totaled 206 variables and 283 constraints and was solved in less than 7 h on a standard computer with 12% reduction when using b-splines for trajectory control variables. The results revealed the need to consider the multi-objective total score to account for the different score components and highlighted the importance of the payload level and chosen trajectory. The wing area should be increased within allowable limits to maximize payload capacity, climb to maximum target height should be the focus of the first 60 s of flight and full throttle should be avoided in cruise to reduce losses and extend flight distance. The framework proved to be a valuable tool for students to easily obtain guidelines for both the model aircraft design and control to maximize the competition score.
Keywords: Multidisciplinary design optimization, Multi-objective optimization, Aerostructural design, Battery discharge model, Collocation method, Optimal control},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
A coupled aerostructural aircraft design and trajectory optimization framework is developed for the Air Cargo Challenge competition to maximize the expected score based on cargo carried, altitude achieved and distance traveled. Its modular architecture makes it easily adaptable to any problem where the performance depends not only on the design of the aircraft but also on its flight trajectory. It is based on OpenAeroStruct, an aerostructural solver that uses analytic derivatives for efficient gradient-based optimization. A trajectory optimization module using a collocation method is coupled with the option of using b-splines to increase computational efficiency together with an experimentally-based power decay model that accurately determines the aircraft propulsive response to control input depending on the battery discharge level. The optimization problem totaled 206 variables and 283 constraints and was solved in less than 7 h on a standard computer with 12% reduction when using b-splines for trajectory control variables. The results revealed the need to consider the multi-objective total score to account for the different score components and highlighted the importance of the payload level and chosen trajectory. The wing area should be increased within allowable limits to maximize payload capacity, climb to maximum target height should be the focus of the first 60 s of flight and full throttle should be avoided in cruise to reduce losses and extend flight distance. The framework proved to be a valuable tool for students to easily obtain guidelines for both the model aircraft design and control to maximize the competition score.
Keywords: Multidisciplinary design optimization, Multi-objective optimization, Aerostructural design, Battery discharge model, Collocation method, Optimal control
Keywords: Multidisciplinary design optimization, Multi-objective optimization, Aerostructural design, Battery discharge model, Collocation method, Optimal control
Morgado, F.; Marta, A. C.; Gil, P.
Multistage rocket preliminary design and trajectory optimization using a multidisciplinary approach Journal Article
In: Structural and Multidisciplinary Optimization, pp. 1–31, 2022, ISSN: 1615-147X.
@article{Morgado:2022:SAMO,
title = {Multistage rocket preliminary design and trajectory optimization using a multidisciplinary approach},
author = {F. Morgado and A. C. Marta and P. Gil},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_JournalArticle_Morgado_2022_SAMO.pdf, Full-text PDF
https://link.springer.com/article/10.1007/s00158-022-03285-y},
doi = {10.1007/s00158-022-03285-y},
issn = {1615-147X},
year = {2022},
date = {2022-02-01},
urldate = {2022-02-01},
journal = {Structural and Multidisciplinary Optimization},
pages = {1--31},
abstract = {A procedure for rocket preliminary design was developed using a multidisciplinary coupled approach that simultaneously finds the optimal design and trajectory parameters for a given representative insertion in orbit launch mission. Given the nature of the performance metrics and design space, and the distinct design and trajectory problems, heuristic methods were used in a multilevel design optimization architecture. For the design, a continuous genetic algorithm able to perform parallel optimization was developed and benchmarked. The results were obtained with mass and sizing models, required to estimate the rocket structure, and created using historical data regression. For the trajectory, once defined its assumptions, the optimality equations are deduced and the optimal values are found using a particle swarm optimization. The multidisciplinary optimization procedure was demonstrated by designing a small launch vehicle and comparing it to a state-of-the-art existing rocket. Promising results were obtained in both design and trajectory optimization, with the imposed constraints adequately handled and the optimal rocket preliminary design with appropriate optimal trajectory found with an affordable computational cost.
Keywords: Launch vehicle, Coupled approach, Multilevel optimization, Two-point boundary-value problem, Genetic algorithm, Particle swarm optimization},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
A procedure for rocket preliminary design was developed using a multidisciplinary coupled approach that simultaneously finds the optimal design and trajectory parameters for a given representative insertion in orbit launch mission. Given the nature of the performance metrics and design space, and the distinct design and trajectory problems, heuristic methods were used in a multilevel design optimization architecture. For the design, a continuous genetic algorithm able to perform parallel optimization was developed and benchmarked. The results were obtained with mass and sizing models, required to estimate the rocket structure, and created using historical data regression. For the trajectory, once defined its assumptions, the optimality equations are deduced and the optimal values are found using a particle swarm optimization. The multidisciplinary optimization procedure was demonstrated by designing a small launch vehicle and comparing it to a state-of-the-art existing rocket. Promising results were obtained in both design and trajectory optimization, with the imposed constraints adequately handled and the optimal rocket preliminary design with appropriate optimal trajectory found with an affordable computational cost.
Keywords: Launch vehicle, Coupled approach, Multilevel optimization, Two-point boundary-value problem, Genetic algorithm, Particle swarm optimization
Keywords: Launch vehicle, Coupled approach, Multilevel optimization, Two-point boundary-value problem, Genetic algorithm, Particle swarm optimization
Campos, L. M. B. C.; Marta, A. C.
On the vibrations of pyramidal beams with rectangular cross-section and application to unswept wings Journal Article
In: The Quarterly Journal Of Mechanics And Applied Mathematics, vol. 74, no. 1, pp. 1–31, 2021, ISSN: 0033-5614.
@article{Campos:2021:QJMAM,
title = {On the vibrations of pyramidal beams with rectangular cross-section and application to unswept wings},
author = {L. M. B. C. Campos and A. C. Marta},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_JournalArticle_Campos_2021_QJMAM.pdf, Full-text PDF
https://academic.oup.com/qjmam/article-abstract/74/1/1/6126085},
doi = {10.1093/qjmam/hbaa017},
issn = {0033-5614},
year = {2021},
date = {2021-02-01},
urldate = {2021-02-01},
journal = {The Quarterly Journal Of Mechanics And Applied Mathematics},
volume = {74},
number = {1},
pages = {1--31},
abstract = {The bending frequencies of an unswept wing are calculated based on the model of a beam clamped at the root and free at the tip. For a tapered wing with straight leading- and trailing-edges, the chord is a linear function of the span; the same linear function of the span applies to thickness, in the case of constant thickness-to-chord ratio. The latter is usually small, so that the beam differs from the more frequent cases of a conical beam with a circular cross-section or a prismatic beam with a square cross-section. Thus, the bending modes of a non-uniform beam are considered, with mass and area moment of inertia which are respectively quadratic and quartic functions of the span. There is no exact solution expressible in finite terms using elementary functions, and thus power series expansions are used. The bending frequencies are calculated for a delta wing and compared with a rectangular wing, with the same span, mean chord and thickness, mass density and Young’s modulus. It is shown that the fundamental frequency is higher by a factor 4.96 for the delta wing; it is also shown that the general case of the tapered wing is intermediate between the delta and the rectangular wing. Lastly, the analytical results obtained for the bending modes are compared with numerical modal analyses of general tapered wing beams using high-fidelity finite-element model software.
Keywords: Cantilever beam, Modal analysis, Transverse vibrations, Bending frequencies, Analytical solution, Power series expansion, Frobenius–Fuchs series, Gamma function},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
The bending frequencies of an unswept wing are calculated based on the model of a beam clamped at the root and free at the tip. For a tapered wing with straight leading- and trailing-edges, the chord is a linear function of the span; the same linear function of the span applies to thickness, in the case of constant thickness-to-chord ratio. The latter is usually small, so that the beam differs from the more frequent cases of a conical beam with a circular cross-section or a prismatic beam with a square cross-section. Thus, the bending modes of a non-uniform beam are considered, with mass and area moment of inertia which are respectively quadratic and quartic functions of the span. There is no exact solution expressible in finite terms using elementary functions, and thus power series expansions are used. The bending frequencies are calculated for a delta wing and compared with a rectangular wing, with the same span, mean chord and thickness, mass density and Young’s modulus. It is shown that the fundamental frequency is higher by a factor 4.96 for the delta wing; it is also shown that the general case of the tapered wing is intermediate between the delta and the rectangular wing. Lastly, the analytical results obtained for the bending modes are compared with numerical modal analyses of general tapered wing beams using high-fidelity finite-element model software.
Keywords: Cantilever beam, Modal analysis, Transverse vibrations, Bending frequencies, Analytical solution, Power series expansion, Frobenius–Fuchs series, Gamma function
Keywords: Cantilever beam, Modal analysis, Transverse vibrations, Bending frequencies, Analytical solution, Power series expansion, Frobenius–Fuchs series, Gamma function
Alexandre, D.; Marino, L.; Marta, A. C.; Graziani, G.; Piva, R.
On the feasibility of Rayleigh cycles for dynamic soaring trajectories Journal Article
In: PLoS One, vol. 15, no. 3, pp. e0229746, 2020, ISSN: 1932-6203.
@article{Alexandre:2020:PLOS,
title = {On the feasibility of Rayleigh cycles for dynamic soaring trajectories},
author = {D. Alexandre and L. Marino and A. C. Marta and G. Graziani and R. Piva},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_JournalArticle_Alexandre_2020_PLOS.pdf, Full-text PDF
https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0229746},
doi = {10.1371/journal.pone.0229746},
issn = {1932-6203},
year = {2020},
date = {2020-03-01},
urldate = {2020-03-01},
journal = {PLoS One},
volume = {15},
number = {3},
pages = {e0229746},
abstract = {Dynamic soaring is a flight technique used by albatrosses and other birds to cover large distances without the expenditure of energy, which is extracted from the available wind condi-
tions, as brightly perceived five centuries ago by Leonardo da Vinci. Closed dynamic soaring trajectories use spatial variations of wind speed to travel, in principle, indefinitely over a prescribed area. The application of the concept of closed dynamic soaring trajectories to aerial vehicles, such as UAVs, may provide a solution to improve the endurance in certain missions. The main limitation of dynamic soaring is its dependence on the wind characteristics. More than one century ago, Lord Rayleigh proposed a very simple model, based on the repeated crossing of a step wind profile, presently known as Rayleigh cycle, that provides a clear explanation of the physical phenomenon. The present paper studies the feasibility of closed, single-loop, energy-neutral trajectories for a broad set of wind and vehicle conditions. Through the use of trajectory optimization methods, it was possible to see how the shape of the wind profile, the initial flight conditions and the vehicle constraints influence the required wind strength to perform dynamic soaring trajectories and consequently their feasibility. It was possible to conclude that there are optimal values for the initial airspeed and initial height of the vehicle, that minimize the required wind strength. In addition, it was seen how the structural and aerodynamic constraints of the vehicle affect dynamic soaring at high and low airspeeds respectively. Finally, some new trajectories that can be performed in conditions of excess wind are proposed. The purpose is to maximize the time spent aloft and the path length while maintaining the concept of single-loop, energy-neutral trajectories, making them especially useful for aerial vehicles surveillance applications.
Keywords: Closed trajectories, Energy-neutral trajectories, Wind profile, Trajectory optimization, UAV surveillance, Extended endurance},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Dynamic soaring is a flight technique used by albatrosses and other birds to cover large distances without the expenditure of energy, which is extracted from the available wind condi-
tions, as brightly perceived five centuries ago by Leonardo da Vinci. Closed dynamic soaring trajectories use spatial variations of wind speed to travel, in principle, indefinitely over a prescribed area. The application of the concept of closed dynamic soaring trajectories to aerial vehicles, such as UAVs, may provide a solution to improve the endurance in certain missions. The main limitation of dynamic soaring is its dependence on the wind characteristics. More than one century ago, Lord Rayleigh proposed a very simple model, based on the repeated crossing of a step wind profile, presently known as Rayleigh cycle, that provides a clear explanation of the physical phenomenon. The present paper studies the feasibility of closed, single-loop, energy-neutral trajectories for a broad set of wind and vehicle conditions. Through the use of trajectory optimization methods, it was possible to see how the shape of the wind profile, the initial flight conditions and the vehicle constraints influence the required wind strength to perform dynamic soaring trajectories and consequently their feasibility. It was possible to conclude that there are optimal values for the initial airspeed and initial height of the vehicle, that minimize the required wind strength. In addition, it was seen how the structural and aerodynamic constraints of the vehicle affect dynamic soaring at high and low airspeeds respectively. Finally, some new trajectories that can be performed in conditions of excess wind are proposed. The purpose is to maximize the time spent aloft and the path length while maintaining the concept of single-loop, energy-neutral trajectories, making them especially useful for aerial vehicles surveillance applications.
Keywords: Closed trajectories, Energy-neutral trajectories, Wind profile, Trajectory optimization, UAV surveillance, Extended endurance
tions, as brightly perceived five centuries ago by Leonardo da Vinci. Closed dynamic soaring trajectories use spatial variations of wind speed to travel, in principle, indefinitely over a prescribed area. The application of the concept of closed dynamic soaring trajectories to aerial vehicles, such as UAVs, may provide a solution to improve the endurance in certain missions. The main limitation of dynamic soaring is its dependence on the wind characteristics. More than one century ago, Lord Rayleigh proposed a very simple model, based on the repeated crossing of a step wind profile, presently known as Rayleigh cycle, that provides a clear explanation of the physical phenomenon. The present paper studies the feasibility of closed, single-loop, energy-neutral trajectories for a broad set of wind and vehicle conditions. Through the use of trajectory optimization methods, it was possible to see how the shape of the wind profile, the initial flight conditions and the vehicle constraints influence the required wind strength to perform dynamic soaring trajectories and consequently their feasibility. It was possible to conclude that there are optimal values for the initial airspeed and initial height of the vehicle, that minimize the required wind strength. In addition, it was seen how the structural and aerodynamic constraints of the vehicle affect dynamic soaring at high and low airspeeds respectively. Finally, some new trajectories that can be performed in conditions of excess wind are proposed. The purpose is to maximize the time spent aloft and the path length while maintaining the concept of single-loop, energy-neutral trajectories, making them especially useful for aerial vehicles surveillance applications.
Keywords: Closed trajectories, Energy-neutral trajectories, Wind profile, Trajectory optimization, UAV surveillance, Extended endurance
Rodrigues, S. S.; Marta, A. C.
Adjoint-based shape sensitivity of multi-row turbomachinery Journal Article
In: Structural and Multidisciplinary Optimization, vol. 61, no. 2, pp. 837–853, 2020, ISSN: 1615-147X.
@article{Rodrigues:2020:SAMO,
title = {Adjoint-based shape sensitivity of multi-row turbomachinery},
author = {S. S. Rodrigues and A. C. Marta},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_JournalArticle_Rodrigues_2020_SAMO.pdf, Full-text PDF
https://link.springer.com/article/10.1007%2Fs00158-019-02386-5},
doi = {10.1007/s00158-019-02386-5},
issn = {1615-147X},
year = {2020},
date = {2020-02-01},
journal = {Structural and Multidisciplinary Optimization},
volume = {61},
number = {2},
pages = {837--853},
abstract = {The performance sensitivity of a low-pressure turbine stator-rotor stage of a commercial jet engine to its blades and hub shapes is analyzed. The derivatives of various metrics, such as isentropic efficiency, total pressure ratio, total pressure loss, and loss coefficient are computed using an adjoint solver capable of handling multi-row analyses. The importance of considering the coupled stator-rotor stage is highlighted from the computed sensitivities, which reveal that performing individual stator or rotor row component optimization may lead to an unexpected performance loss of the whole stage. The obtained coupled sensitivities are then used in an endwall contouring application, consisting of two Hicks-Henne bumps applied on the hub surface of the rotor to maximize stage efficiency, at locations selected where the impact is found to be the highest from the adjoint-based sensitivity analysis. The performance gain obtained with the bumps is correlated to the changes in the flow field, in particular with the effect on secondary flows.
Keywords: Multistage coupling, Sensitivity analysis, Aerodynamic shape optimization, Endwall contouring, Gas turbine, Jet engine},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
The performance sensitivity of a low-pressure turbine stator-rotor stage of a commercial jet engine to its blades and hub shapes is analyzed. The derivatives of various metrics, such as isentropic efficiency, total pressure ratio, total pressure loss, and loss coefficient are computed using an adjoint solver capable of handling multi-row analyses. The importance of considering the coupled stator-rotor stage is highlighted from the computed sensitivities, which reveal that performing individual stator or rotor row component optimization may lead to an unexpected performance loss of the whole stage. The obtained coupled sensitivities are then used in an endwall contouring application, consisting of two Hicks-Henne bumps applied on the hub surface of the rotor to maximize stage efficiency, at locations selected where the impact is found to be the highest from the adjoint-based sensitivity analysis. The performance gain obtained with the bumps is correlated to the changes in the flow field, in particular with the effect on secondary flows.
Keywords: Multistage coupling, Sensitivity analysis, Aerodynamic shape optimization, Endwall contouring, Gas turbine, Jet engine
Keywords: Multistage coupling, Sensitivity analysis, Aerodynamic shape optimization, Endwall contouring, Gas turbine, Jet engine
Rodrigues, S. S.; Marta, A. C.
On addressing wind turbine noise with after-market shape blade add-ons Journal Article
In: Renewable Energy, vol. 140, pp. 602–614, 2019, ISSN: 0960-1481.
@article{Rodrigues:2019:RENE,
title = {On addressing wind turbine noise with after-market shape blade add-ons},
author = {S. S. Rodrigues and A. C. Marta},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_JournalArticle_Rodrigues_2019_RENE.pdf, Full-text PDF
https://www.sciencedirect.com/science/article/pii/S0960148119303611},
doi = {10.1016/j.renene.2019.03.056},
issn = {0960-1481},
year = {2019},
date = {2019-09-01},
journal = {Renewable Energy},
volume = {140},
pages = {602--614},
abstract = {When stricter noise limits are enforced to legacy wind turbines already deployed, actions need to be taken. In this paper, we present a solution of retrofitting wind turbine blades with additional outer layer skins that change their aeroacoustic footprint. An optimization design framework produces add-ons shapes that, when attached to blades, reduce their noise without compromising aerodynamic performance. The Blade Element Momentum theory is used to predict the aerodynamic performance and generated noise is predicted using semi-empirical models. Two competing metrics are analyzed, Annual Energy Production and Overall Sound Pressure Level, using a multi-objective genetic algorithm. The add-on shapes are parameterized using NURBS totaling 54 design variables. The AOC 15/50 wind turbine is used as a test case and optimal solutions selected from the Pareto front are discussed. The after-market add-on approach produces solutions that range from an increase of 8.7% in energy production to a decrease of 3.5 dB(A) in noise levels, with an estimated blade weight increase of less than 4%. While the add-on approaches typically fall short in terms of performance when compared to a new blade design, this retrofiting is expected to be a competitive alternative when compared to the cost of replacing the whole blade.
Keywords: Noise reduction, Aeroacoustic analysis, Airfoil self-noise, Design optimization, Retrofitting, Multi-objective optimization},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
When stricter noise limits are enforced to legacy wind turbines already deployed, actions need to be taken. In this paper, we present a solution of retrofitting wind turbine blades with additional outer layer skins that change their aeroacoustic footprint. An optimization design framework produces add-ons shapes that, when attached to blades, reduce their noise without compromising aerodynamic performance. The Blade Element Momentum theory is used to predict the aerodynamic performance and generated noise is predicted using semi-empirical models. Two competing metrics are analyzed, Annual Energy Production and Overall Sound Pressure Level, using a multi-objective genetic algorithm. The add-on shapes are parameterized using NURBS totaling 54 design variables. The AOC 15/50 wind turbine is used as a test case and optimal solutions selected from the Pareto front are discussed. The after-market add-on approach produces solutions that range from an increase of 8.7% in energy production to a decrease of 3.5 dB(A) in noise levels, with an estimated blade weight increase of less than 4%. While the add-on approaches typically fall short in terms of performance when compared to a new blade design, this retrofiting is expected to be a competitive alternative when compared to the cost of replacing the whole blade.
Keywords: Noise reduction, Aeroacoustic analysis, Airfoil self-noise, Design optimization, Retrofitting, Multi-objective optimization
Keywords: Noise reduction, Aeroacoustic analysis, Airfoil self-noise, Design optimization, Retrofitting, Multi-objective optimization
Campos, L. M. B. C.; Marta, A. C.
On the extension of cylindrical acoustic waves to acoustic-vortical-entropy waves in a flow with rigid body swirl Journal Article
In: Journal of Sound and Vibration, vol. 437, pp. 389–409, 2018, ISSN: 0022-460X.
@article{Campos:2018:JSV,
title = {On the extension of cylindrical acoustic waves to acoustic-vortical-entropy waves in a flow with rigid body swirl},
author = {L. M. B. C. Campos and A. C. Marta},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_JournalArticle_Campos_2018_JSV.pdf, Full-text PDF
https://www.sciencedirect.com/science/article/pii/S0022460X18305960},
doi = {10.1016/j.jsv.2018.09.017},
issn = {0022-460X},
year = {2018},
date = {2018-12-01},
journal = {Journal of Sound and Vibration},
volume = {437},
pages = {389--409},
abstract = {The noise of jet and rocket engines involves the coupling of sound to swirling flows and to heat exchanges leading in the more complex cases of triple interactions to acoustic-
vortical-entropy (AVE) waves. The present paper presents the derivation of the AVE equation for axisymmetric linear non-dissipative, compressible perturbations of a non-homentropic, swirling mean flow, with constant axial velocity and constant angular velocity for a perfect gas with constant density. The axisymmetric AVE wave equation is obtained for the radial velocity perturbation, specifying its radial dependence for any frequency and axial wavenumber. The AVE wave equation in the case of zero axial wavenumber, corresponding to cylindrical AVE waves, has no singularities for finite radius, including the sonic radius, where the isothermal Mach number for the swirl velocity is unity. The exact solution of the AVE wave equation for the fundamental axisymmetric mode with zero axial wavenumber is obtained without any restriction on frequency, as series expansions of Gaussian hypergeometric type: (i) covering the whole flow region; (ii) specifying the wave field at the sonic radius; (iii) specifying near-axis and asymptotic scaling for small and large radius. Using polarization relations among wave variables specifies exactly and allows the plotting of the perturbations of: (i,ii) the radial and azimuthal velocity; (iii,iv) pressure and mass density; (v,vi) entropy and temperature. Thus the extension of cylindrical acoustic waves, that are specified by Bessel functions, to cylindrical acoustic-vortical-entropy waves, is specified by Gaussian hypergeometric functions.
Keywords: Aeroacoustics, Acoustic-vortical-entropy waves, Swirling flow instabilities, Non-isentropic perturbations},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
The noise of jet and rocket engines involves the coupling of sound to swirling flows and to heat exchanges leading in the more complex cases of triple interactions to acoustic-
vortical-entropy (AVE) waves. The present paper presents the derivation of the AVE equation for axisymmetric linear non-dissipative, compressible perturbations of a non-homentropic, swirling mean flow, with constant axial velocity and constant angular velocity for a perfect gas with constant density. The axisymmetric AVE wave equation is obtained for the radial velocity perturbation, specifying its radial dependence for any frequency and axial wavenumber. The AVE wave equation in the case of zero axial wavenumber, corresponding to cylindrical AVE waves, has no singularities for finite radius, including the sonic radius, where the isothermal Mach number for the swirl velocity is unity. The exact solution of the AVE wave equation for the fundamental axisymmetric mode with zero axial wavenumber is obtained without any restriction on frequency, as series expansions of Gaussian hypergeometric type: (i) covering the whole flow region; (ii) specifying the wave field at the sonic radius; (iii) specifying near-axis and asymptotic scaling for small and large radius. Using polarization relations among wave variables specifies exactly and allows the plotting of the perturbations of: (i,ii) the radial and azimuthal velocity; (iii,iv) pressure and mass density; (v,vi) entropy and temperature. Thus the extension of cylindrical acoustic waves, that are specified by Bessel functions, to cylindrical acoustic-vortical-entropy waves, is specified by Gaussian hypergeometric functions.
Keywords: Aeroacoustics, Acoustic-vortical-entropy waves, Swirling flow instabilities, Non-isentropic perturbations
vortical-entropy (AVE) waves. The present paper presents the derivation of the AVE equation for axisymmetric linear non-dissipative, compressible perturbations of a non-homentropic, swirling mean flow, with constant axial velocity and constant angular velocity for a perfect gas with constant density. The axisymmetric AVE wave equation is obtained for the radial velocity perturbation, specifying its radial dependence for any frequency and axial wavenumber. The AVE wave equation in the case of zero axial wavenumber, corresponding to cylindrical AVE waves, has no singularities for finite radius, including the sonic radius, where the isothermal Mach number for the swirl velocity is unity. The exact solution of the AVE wave equation for the fundamental axisymmetric mode with zero axial wavenumber is obtained without any restriction on frequency, as series expansions of Gaussian hypergeometric type: (i) covering the whole flow region; (ii) specifying the wave field at the sonic radius; (iii) specifying near-axis and asymptotic scaling for small and large radius. Using polarization relations among wave variables specifies exactly and allows the plotting of the perturbations of: (i,ii) the radial and azimuthal velocity; (iii,iv) pressure and mass density; (v,vi) entropy and temperature. Thus the extension of cylindrical acoustic waves, that are specified by Bessel functions, to cylindrical acoustic-vortical-entropy waves, is specified by Gaussian hypergeometric functions.
Keywords: Aeroacoustics, Acoustic-vortical-entropy waves, Swirling flow instabilities, Non-isentropic perturbations
Rodrigues, S. S.; Marta, A. C.
Adjoint formulation of a steady multistage turbomachinery interface using automatic differentiation Journal Article
In: Computer and Fluids, vol. 176, pp. 182–192, 2018, ISSN: 0045-7930.
@article{Rodrigues:2018:CAF,
title = {Adjoint formulation of a steady multistage turbomachinery interface using automatic differentiation},
author = {S. S. Rodrigues and A. C. Marta},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_JournalArticle_Rodrigues_2018_CAF.pdf, Full-text PDF
https://www.sciencedirect.com/science/article/pii/S0045793018306625},
doi = {10.1016/j.compfluid.2018.09.015},
issn = {0045-7930},
year = {2018},
date = {2018-11-01},
journal = {Computer and Fluids},
volume = {176},
pages = {182--192},
abstract = {The use of high-fidelity computational fluid dynamics (CFD) tools in turbomachinery design has seen a continuous increase as a result of computational power growth and numerical methods improvement. These tools are often used in optimization environments, where gradient-based optimization algorithms are the most common due to their efficiency. In cases where the optimization contains a large number of design variables, the adjoint approach for calculating the gradients is beneficial, as it provides a way of obtaining function sensitivities with a computational cost that is nearly independent of the number of design variables. The interaction between adjacent blade rows is of utmost importance for the performance of multistage turbomachines. The most commonly used method to address these effects (i.e. coupling in the simulation of multiple rows) is the mixing-plane treatment, that has become a standard industrial tool in the design environment. In this paper, the formulation and implementation of an adjoint solver for multistage turbomachinery applications are presented, namely the adjoint counterpart of the mixing-plane formulation used in the direct solver. The solver is developed using the discrete ADjoint approach, where the partial derivatives required for the assembly of the adjoint system of equations are obtained using automatic differentiation tools. The sensitivity of several performance metrics relative to neighbor blade/hub row geometry and boundary conditions are shown to highlight the physical coupling in multi-row turbomachines. The verification of the adjoint multistage solver against the finite-difference approach is performed successfully with relative differences below 1 %.
Keywords: Discrete adjoint, Mixing-plane, Algorithmic differentiation, Sensitivity analysis, Shape optimization, Endwall contouring, Operating conditions},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
The use of high-fidelity computational fluid dynamics (CFD) tools in turbomachinery design has seen a continuous increase as a result of computational power growth and numerical methods improvement. These tools are often used in optimization environments, where gradient-based optimization algorithms are the most common due to their efficiency. In cases where the optimization contains a large number of design variables, the adjoint approach for calculating the gradients is beneficial, as it provides a way of obtaining function sensitivities with a computational cost that is nearly independent of the number of design variables. The interaction between adjacent blade rows is of utmost importance for the performance of multistage turbomachines. The most commonly used method to address these effects (i.e. coupling in the simulation of multiple rows) is the mixing-plane treatment, that has become a standard industrial tool in the design environment. In this paper, the formulation and implementation of an adjoint solver for multistage turbomachinery applications are presented, namely the adjoint counterpart of the mixing-plane formulation used in the direct solver. The solver is developed using the discrete ADjoint approach, where the partial derivatives required for the assembly of the adjoint system of equations are obtained using automatic differentiation tools. The sensitivity of several performance metrics relative to neighbor blade/hub row geometry and boundary conditions are shown to highlight the physical coupling in multi-row turbomachines. The verification of the adjoint multistage solver against the finite-difference approach is performed successfully with relative differences below 1 %.
Keywords: Discrete adjoint, Mixing-plane, Algorithmic differentiation, Sensitivity analysis, Shape optimization, Endwall contouring, Operating conditions
Keywords: Discrete adjoint, Mixing-plane, Algorithmic differentiation, Sensitivity analysis, Shape optimization, Endwall contouring, Operating conditions
Campos, L. M. B. C.; Marta, A. C.
On the combined effect of atmospheric stratification and non-uniform magnetic field on magneto-sonic-gravity waves Journal Article
In: Geophysical & Astrophysical Fluid Dynamics, vol. 109, no. 2, pp. 168–198, 2015, ISSN: 1029-0419.
@article{Campos:2015:GAFD,
title = {On the combined effect of atmospheric stratification and non-uniform magnetic field on magneto-sonic-gravity waves},
author = {L. M. B. C. Campos and A. C. Marta},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_JournalArticle_Campos_2015_GAFD.pdf, Full-text PDF
http://www.tandfonline.com/doi/full/10.1080/03091929.2015.1032959},
doi = {10.1080/03091929.2015.1032959},
issn = {1029-0419},
year = {2015},
date = {2015-04-01},
journal = {Geophysical & Astrophysical Fluid Dynamics},
volume = {109},
number = {2},
pages = {168--198},
abstract = {The linear magneto-acoustic-gravity (MAG) wave equation is considered for a non-isothermal atmosphere under a non-uniform external magnetic field. The starting point is a magnetohydrostatic equilibrium with an arbitrary profiles of temperature and horizontal magnetic field as a function of altitude; this specifies the profiles of gas pressure, mass density and sound and Alfvén speeds. The wave equation is solved exactly in the case of an isothermal atmosphere with horizontal magnetic field decaying exponentially with altitude on twice the scale height. The solution for the vertical velocity perturbation is represented by confluent hypergeometric functions specifying the effect of the magnetic field in modifying the amplitude and phase of acoustic-gravity waves. It is shown that (i) in the physical conditions corresponding to the solar corona, the decrease in Alfvén speed with height leads to a decreasing spacing of nodes; this agrees with observations of ratios of periods p 2 / p 1 less than two in solar arches or loops; also (ii) the dissipation of these magnetosonic-gravity modes in the solar transition region is sufficient to heat the corona by compensating for energy losses in solar radiation. (i) and (ii) are set in the context of a tentative global picture of the possible role of MAG waves in establishing the mass and energy balances in the solar atmosphere.
Keywords: Magnetohydrodynamics, Waves, Sun, Stellar atmospheres},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
The linear magneto-acoustic-gravity (MAG) wave equation is considered for a non-isothermal atmosphere under a non-uniform external magnetic field. The starting point is a magnetohydrostatic equilibrium with an arbitrary profiles of temperature and horizontal magnetic field as a function of altitude; this specifies the profiles of gas pressure, mass density and sound and Alfvén speeds. The wave equation is solved exactly in the case of an isothermal atmosphere with horizontal magnetic field decaying exponentially with altitude on twice the scale height. The solution for the vertical velocity perturbation is represented by confluent hypergeometric functions specifying the effect of the magnetic field in modifying the amplitude and phase of acoustic-gravity waves. It is shown that (i) in the physical conditions corresponding to the solar corona, the decrease in Alfvén speed with height leads to a decreasing spacing of nodes; this agrees with observations of ratios of periods p 2 / p 1 less than two in solar arches or loops; also (ii) the dissipation of these magnetosonic-gravity modes in the solar transition region is sufficient to heat the corona by compensating for energy losses in solar radiation. (i) and (ii) are set in the context of a tentative global picture of the possible role of MAG waves in establishing the mass and energy balances in the solar atmosphere.
Keywords: Magnetohydrodynamics, Waves, Sun, Stellar atmospheres
Keywords: Magnetohydrodynamics, Waves, Sun, Stellar atmospheres
Marta, A. C.; Shankaran, S.
Assessing turbomachinery performance sensitivity to boundary conditions using control theory Journal Article
In: AIAA Journal of Propulsion and Power, vol. 30, no. 5, pp. 1281–1294, 2014, ISSN: 0748-4658.
@article{Marta:2014:JPP,
title = {Assessing turbomachinery performance sensitivity to boundary conditions using control theory},
author = {A. C. Marta and S. Shankaran},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_JournalArticle_Marta_2014_JPP.pdf, Full-text PDF
https://arc.aiaa.org/doi/10.2514/1.B35087},
doi = {10.2514/1.B35087},
issn = {0748-4658},
year = {2014},
date = {2014-09-01},
urldate = {2014-09-01},
journal = {AIAA Journal of Propulsion and Power},
volume = {30},
number = {5},
pages = {1281--1294},
abstract = {The adjoint method is extended to assess the sensitivity of turbomachinery performance with respect to inlet and exit boundary conditions. The derivation of the adjoint and sensitivity equations are briefly derived in general form, such that they can be applied to any set of flow-governing equations. In this paper, the Reynolds-averaged Navier--Stokes equations have been used, where the k-ω turbulence model has been selected. A compressor rotor blade test case is studied for verification and demonstration of the methodology. The adjoint-based sensitivities are verified using finite differences. The sensitivities of efficiency and pressure ratio with respect to the boundary condition parameters prescribed at each computational grid node at the inlet and exit faces of the blade passage are illustrated in the form of contour plots. Circumferentially averaged values and sensitivities demonstrate that fuller, more uniform profiles lead to improved performance, in line with basic thermodynamic principles. Two examples of modifying either inlet or exit boundary profiles, according to the sensitivity data obtained, show that performance tuning can be achieved. The sensitivity assessment approach presented is shown to be accurate and extremely efficient, while providing the designer with valuable information and insights to achieve a robust design.
Keywords: Sensitivity analysis, Adjoint method, Compressor blade, Inlet conditions, Exit conditions, Efficiency, Pressure ratio},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
The adjoint method is extended to assess the sensitivity of turbomachinery performance with respect to inlet and exit boundary conditions. The derivation of the adjoint and sensitivity equations are briefly derived in general form, such that they can be applied to any set of flow-governing equations. In this paper, the Reynolds-averaged Navier--Stokes equations have been used, where the k-ω turbulence model has been selected. A compressor rotor blade test case is studied for verification and demonstration of the methodology. The adjoint-based sensitivities are verified using finite differences. The sensitivities of efficiency and pressure ratio with respect to the boundary condition parameters prescribed at each computational grid node at the inlet and exit faces of the blade passage are illustrated in the form of contour plots. Circumferentially averaged values and sensitivities demonstrate that fuller, more uniform profiles lead to improved performance, in line with basic thermodynamic principles. Two examples of modifying either inlet or exit boundary profiles, according to the sensitivity data obtained, show that performance tuning can be achieved. The sensitivity assessment approach presented is shown to be accurate and extremely efficient, while providing the designer with valuable information and insights to achieve a robust design.
Keywords: Sensitivity analysis, Adjoint method, Compressor blade, Inlet conditions, Exit conditions, Efficiency, Pressure ratio
Keywords: Sensitivity analysis, Adjoint method, Compressor blade, Inlet conditions, Exit conditions, Efficiency, Pressure ratio
Rodrigues, S. S.; Marta, A. C.
On addressing noise constraints in the design of wind turbine blades Journal Article
In: Structural and Multidisciplinary Optimization, vol. 50, no. 3, pp. 489–503, 2014, ISSN: 1615-1488.
@article{Rodrigues:2014:SAMO,
title = {On addressing noise constraints in the design of wind turbine blades},
author = {S. S. Rodrigues and A. C. Marta},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_JournalArticle_Rodrigues_2014_SAMO.pdf, Full-text PDF
http://link.springer.com/article/10.1007%2Fs00158-014-1072-4},
doi = {10.1007/s00158-014-1072-4},
issn = {1615-1488},
year = {2014},
date = {2014-09-01},
journal = {Structural and Multidisciplinary Optimization},
volume = {50},
number = {3},
pages = {489--503},
abstract = {Power production from wind energy has been increasing over the past decades, with more areas being used as wind farms and larger wind turbines (WTs) being built. With this development, awareness of the impact of wind energy on the environment and on human health has also raised. There has been a large interest in developing fast turnaround WT blade design frameworks, capable of predicting both aerodynamic and aeroacoustic performance to handle ever stricter noise criteria constraints dictated by site or local authorities. In this work, a blade element momentum theory model is used to predict the aerodynamic performance of a wind turbine, coupled to an empirical aeroacoustic noise model and boundary layer corrections. The aeroacoustic prediction code developed was validated against measurement data of the AOC 15/50 WT and included in an optimization framework using a genetic algorithm. The blade shape was parametrized using NURBS curves for the cross sectional airfoil shapes and Bézier curves for the twist and chord distributions, totaling up to 62 design variables. Two multi-objective optimization cases, both single- and multi-operating point, were performed. Optimal solutions selected from the Pareto fronts are discussed in detail. These solutions ranged from an increase in annual energy production of 15% to a reduction in noise levels of 9.8%. It was demonstrated that substantial noise reduction could be obtained at an expense of a minor aerodynamic penalty.
Keywords: Aeroacoustic, Optimization, Wind turbines, NURBS, Genetic algorithms, Pareto front},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Power production from wind energy has been increasing over the past decades, with more areas being used as wind farms and larger wind turbines (WTs) being built. With this development, awareness of the impact of wind energy on the environment and on human health has also raised. There has been a large interest in developing fast turnaround WT blade design frameworks, capable of predicting both aerodynamic and aeroacoustic performance to handle ever stricter noise criteria constraints dictated by site or local authorities. In this work, a blade element momentum theory model is used to predict the aerodynamic performance of a wind turbine, coupled to an empirical aeroacoustic noise model and boundary layer corrections. The aeroacoustic prediction code developed was validated against measurement data of the AOC 15/50 WT and included in an optimization framework using a genetic algorithm. The blade shape was parametrized using NURBS curves for the cross sectional airfoil shapes and Bézier curves for the twist and chord distributions, totaling up to 62 design variables. Two multi-objective optimization cases, both single- and multi-operating point, were performed. Optimal solutions selected from the Pareto fronts are discussed in detail. These solutions ranged from an increase in annual energy production of 15% to a reduction in noise levels of 9.8%. It was demonstrated that substantial noise reduction could be obtained at an expense of a minor aerodynamic penalty.
Keywords: Aeroacoustic, Optimization, Wind turbines, NURBS, Genetic algorithms, Pareto front
Keywords: Aeroacoustic, Optimization, Wind turbines, NURBS, Genetic algorithms, Pareto front
Campos, L. M. B. C.; Marta, A. C.
On the prevention or facilitation of buckling of beams Journal Article
In: International Journal of Mechanical Sciences, vol. 79, pp. 95–104, 2014, ISSN: 0020-7403.
@article{Campos:2014:IJMS,
title = {On the prevention or facilitation of buckling of beams},
author = {L. M. B. C. Campos and A. C. Marta},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_JournalArticle_Campos_2014_IJMS.pdf, Full-text PDF
http://www.sciencedirect.com/science/article/pii/S002074031300338X},
doi = {10.1016/j.ijmecsci.2013.12.003},
issn = {0020-7403},
year = {2014},
date = {2014-02-01},
journal = {International Journal of Mechanical Sciences},
volume = {79},
pages = {95--104},
abstract = {The present paper considers the use of linear or rotary transverse springs to: (i) prevent the buckling of beams by raising the critical buckling load and thus allow a larger axial tension; (ii) provoke the buckling of beams, by lowering the critical buckling load, facilitating the demolition of a structure. The prevention or facilitation of buckling depends on the positioning of the linear and/or rotary springs to oppose or favour bending, for example: at the tip of a (i) cantilever or clamped–free beam; at the middle of a (ii) clamped–clamped, (iii) pinned–pinned or (iv) clamped–pinned beam. In all eight of the four beam supports (i)–(iv) with either linear or rotary springs, the relation between the critical buckling load and the resilience of the spring is obtained.
Keywords: Euler buckling, Critical load, Slender columns, Spring resilience, Linear spring, Rotary spring},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
The present paper considers the use of linear or rotary transverse springs to: (i) prevent the buckling of beams by raising the critical buckling load and thus allow a larger axial tension; (ii) provoke the buckling of beams, by lowering the critical buckling load, facilitating the demolition of a structure. The prevention or facilitation of buckling depends on the positioning of the linear and/or rotary springs to oppose or favour bending, for example: at the tip of a (i) cantilever or clamped–free beam; at the middle of a (ii) clamped–clamped, (iii) pinned–pinned or (iv) clamped–pinned beam. In all eight of the four beam supports (i)–(iv) with either linear or rotary springs, the relation between the critical buckling load and the resilience of the spring is obtained.
Keywords: Euler buckling, Critical load, Slender columns, Spring resilience, Linear spring, Rotary spring
Keywords: Euler buckling, Critical load, Slender columns, Spring resilience, Linear spring, Rotary spring
Marta, A. C.; Shankaran, S.; Venugopal, P.; Barr, B.; Wang, Q.
Interpretation of adjoint solutions for turbomachinery flows Journal Article
In: AIAA Journal, vol. 51, no. 7, pp. 1733–1744, 2013, ISSN: 0001-1452.
@article{Marta:2013:AIAAJ,
title = {Interpretation of adjoint solutions for turbomachinery flows},
author = {A. C. Marta and S. Shankaran and P. Venugopal and B. Barr and Q. Wang},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_JournalArticle_Marta_2013_AIAAJ.pdf, Full-text PDF
http://arc.aiaa.org/doi/pdf/10.2514/1.J052177},
doi = {10.2514/1.J052177},
issn = {0001-1452},
year = {2013},
date = {2013-07-01},
urldate = {2013-07-01},
journal = {AIAA Journal},
volume = {51},
number = {7},
pages = {1733--1744},
abstract = {Although the mathematical derivation of the adjoint equations and their numerical implementation is well established, there is scant discussion on the understanding of the adjoint solution by itself. As this is a field solution of similar resolution of the flowfield, there is a wealth of data that can be used for design guidance. The aim is to tie the adjoint solution to the flowfield, which has physical properties. The adjoint solution of four representative cases taken from turbomachinery aerodynamic problems are used to identify the physical insight it provides. The focus is on changes related to geometry, but the changes can also be realized using other inputs to the flow solver (e.g., boundary conditions). It is shown how the adjoint counterpart of the density and velocity field can be used to provide insights into the nature of changes the designer can induce to cause improvement in the performance metric of interest. Discussion on how to use adjoint solutions for problems with constraints to further refine the changes is also included. Finally, a turbine strut problem is discussed where it is not immediately apparent what geometry changes need to be used for further evaluation with optimization algorithms. The adjoint and flow solutions are used to determine the kind of end-wall treatments that reduce the loss. These changes are then implemented to show that the loss is actually reduced. The results in this paper show there is a twofold use of the adjoint method: one for guiding the automatic optimization as such and the second for guiding the designer in the choice of the design space.
Keywords: Aerodynamics, Adjoint method, Blade design, Jet engine, Shape optimization, Endwall contouring},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Although the mathematical derivation of the adjoint equations and their numerical implementation is well established, there is scant discussion on the understanding of the adjoint solution by itself. As this is a field solution of similar resolution of the flowfield, there is a wealth of data that can be used for design guidance. The aim is to tie the adjoint solution to the flowfield, which has physical properties. The adjoint solution of four representative cases taken from turbomachinery aerodynamic problems are used to identify the physical insight it provides. The focus is on changes related to geometry, but the changes can also be realized using other inputs to the flow solver (e.g., boundary conditions). It is shown how the adjoint counterpart of the density and velocity field can be used to provide insights into the nature of changes the designer can induce to cause improvement in the performance metric of interest. Discussion on how to use adjoint solutions for problems with constraints to further refine the changes is also included. Finally, a turbine strut problem is discussed where it is not immediately apparent what geometry changes need to be used for further evaluation with optimization algorithms. The adjoint and flow solutions are used to determine the kind of end-wall treatments that reduce the loss. These changes are then implemented to show that the loss is actually reduced. The results in this paper show there is a twofold use of the adjoint method: one for guiding the automatic optimization as such and the second for guiding the designer in the choice of the design space.
Keywords: Aerodynamics, Adjoint method, Blade design, Jet engine, Shape optimization, Endwall contouring
Keywords: Aerodynamics, Adjoint method, Blade design, Jet engine, Shape optimization, Endwall contouring
Marta, A. C.; Shankaran, S.
On the handling of turbulence equations in RANS adjoint solvers Journal Article
In: Computers & Fluids, vol. 74, pp. 102–113, 2013, ISSN: 0045-7930.
@article{Marta:2013:CAF,
title = {On the handling of turbulence equations in RANS adjoint solvers},
author = {A. C. Marta and S. Shankaran},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_JournalArticle_Marta_2013_CAF.pdf, Full-text PDF
http://www.sciencedirect.com/science/article/pii/S0045793013000303},
doi = {10.1016/j.compfluid.2013.01.012},
issn = {0045-7930},
year = {2013},
date = {2013-03-01},
journal = {Computers & Fluids},
volume = {74},
pages = {102--113},
abstract = {Recent developments in numerical design tools have made practical the use of gradient-based optimization using high-fidelity computational fluid dynamic simulations. Such has been made possible with the use of adjoint solvers, that can efficiently provide gradients of functions of interest with respect to design variables. However, in the presence of flows modeled by the Reynolds-Averaged Navier–Stokes (RANS) equations, the corresponding adjoint might become too complex to be fully derived or run. This has led to the use of many simplifications in the implementation of such adjoint solvers. In this paper, the constant eddy viscosity (CEV) approximation is explained and its validity tested. Two cases are used, a two-dimensional turbine vane blade and a three-dimensional transonic compressor rotor blade. The gradients computed using both the full RANS and the CEV approximation adjoints are verified against finite-differences. It is shown that the gradients differ slightly but when used in an optimization problem, the optimal solution found is nearly identical. Therefore, the CEV approximation in RANS adjoint solvers proved to be valid for engineering design problems, bringing significant advantages, such as faster implementation and less computational resources needed in terms of CPU and memory size, when compared to the full RANS adjoint solver.
Keywords: Adjoint method, Discrete approach, Turbulence models, Approximation models, Constant eddy viscosity, Shape optimization},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Recent developments in numerical design tools have made practical the use of gradient-based optimization using high-fidelity computational fluid dynamic simulations. Such has been made possible with the use of adjoint solvers, that can efficiently provide gradients of functions of interest with respect to design variables. However, in the presence of flows modeled by the Reynolds-Averaged Navier–Stokes (RANS) equations, the corresponding adjoint might become too complex to be fully derived or run. This has led to the use of many simplifications in the implementation of such adjoint solvers. In this paper, the constant eddy viscosity (CEV) approximation is explained and its validity tested. Two cases are used, a two-dimensional turbine vane blade and a three-dimensional transonic compressor rotor blade. The gradients computed using both the full RANS and the CEV approximation adjoints are verified against finite-differences. It is shown that the gradients differ slightly but when used in an optimization problem, the optimal solution found is nearly identical. Therefore, the CEV approximation in RANS adjoint solvers proved to be valid for engineering design problems, bringing significant advantages, such as faster implementation and less computational resources needed in terms of CPU and memory size, when compared to the full RANS adjoint solver.
Keywords: Adjoint method, Discrete approach, Turbulence models, Approximation models, Constant eddy viscosity, Shape optimization
Keywords: Adjoint method, Discrete approach, Turbulence models, Approximation models, Constant eddy viscosity, Shape optimization
Marta, A. C.; Alonso, J. J.
Toward optimally seeded airflow on hypersonic vehicles using control theory Journal Article
In: Computers & Fluids, vol. 39, no. 9, pp. 1562–1574, 2010, ISSN: 0045-7930.
@article{Marta:2010:CAF,
title = {Toward optimally seeded airflow on hypersonic vehicles using control theory},
author = {A. C. Marta and J. J. Alonso},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_JournalArticle_Marta_2010_CAF.pdf, Full-text PDF
http://www.sciencedirect.com/science/article/pii/S0045793010001155},
doi = {10.1016/j.compfluid.2010.05.009},
issn = {0045-7930},
year = {2010},
date = {2010-10-01},
journal = {Computers & Fluids},
volume = {39},
number = {9},
pages = {1562--1574},
abstract = {Following the renewed interest in hypersonic flight and the significant advances made recently, it is now the time to start looking at ways to optimize hypersonic vehicle designs in an efficient manner. Since the medium, in a hypersonic flow, can be locally ionized, it is possible to use electromagnetic actuators that induce an acting force to optimally control the flow. The local injection of substances that have a considerably lower ionization temperature than air into the airflow – flow seeding – leads to stronger local ionization levels at relatively low hypersonic speeds, amplifying the magnetic effects for the same imposed magnetic field intensity. Because much has been devoted to the analysis of such problems but no formal design approach as been persued to date, the main motivation for this work is to provide an efficient design framework built around high-speed magnetohydrodynamics (MHD) prediction capabilities that can be used in hypersonic control applications using magnetic effects. In particular, the design framework should provide information that leads to an optimal airflow seeding strategy in conjunction with an imposed magnetic field. The proposed framework is based on control theory, which implies developing an adjoint solver aimed to efficiently provide sensitivity analysis capability in arbitrary complex hypersonics MHD flows. Automatic differentiation tools are selectively used to develop the discrete adjoint, which make for a much shorter implementation time and greatly reduce the probability of programming errors. A generic hypersonic vehicle is used to demonstrate the sensitivity analysis capability of the implemented MHD adjoint solver. The precision of the computed adjoint-based sensitivities is established and the performance of the adjoint solver is analyzed. A sample design problem is included using a gradient-based optimizer.
Keywords: Magnetohydrodynamics, Hypersonics, Seeding, Control, Adjoint, Optimization},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Following the renewed interest in hypersonic flight and the significant advances made recently, it is now the time to start looking at ways to optimize hypersonic vehicle designs in an efficient manner. Since the medium, in a hypersonic flow, can be locally ionized, it is possible to use electromagnetic actuators that induce an acting force to optimally control the flow. The local injection of substances that have a considerably lower ionization temperature than air into the airflow – flow seeding – leads to stronger local ionization levels at relatively low hypersonic speeds, amplifying the magnetic effects for the same imposed magnetic field intensity. Because much has been devoted to the analysis of such problems but no formal design approach as been persued to date, the main motivation for this work is to provide an efficient design framework built around high-speed magnetohydrodynamics (MHD) prediction capabilities that can be used in hypersonic control applications using magnetic effects. In particular, the design framework should provide information that leads to an optimal airflow seeding strategy in conjunction with an imposed magnetic field. The proposed framework is based on control theory, which implies developing an adjoint solver aimed to efficiently provide sensitivity analysis capability in arbitrary complex hypersonics MHD flows. Automatic differentiation tools are selectively used to develop the discrete adjoint, which make for a much shorter implementation time and greatly reduce the probability of programming errors. A generic hypersonic vehicle is used to demonstrate the sensitivity analysis capability of the implemented MHD adjoint solver. The precision of the computed adjoint-based sensitivities is established and the performance of the adjoint solver is analyzed. A sample design problem is included using a gradient-based optimizer.
Keywords: Magnetohydrodynamics, Hypersonics, Seeding, Control, Adjoint, Optimization
Keywords: Magnetohydrodynamics, Hypersonics, Seeding, Control, Adjoint, Optimization
Marta, A. C.; Mader, C. A.; Martins, J. R. R. A.; Weide, E.; Alonso, J. J.
A methodology for the development of discrete adjoint solvers using automatic differentiation tools Journal Article
In: International Journal of Computational Fluid Dynamics, vol. 21, no. 9--10, pp. 307–327, 2007, ISSN: 1061-8562.
@article{Marta:2007:IJCFD,
title = {A methodology for the development of discrete adjoint solvers using automatic differentiation tools},
author = {A. C. Marta and C. A. Mader and J. R. R. A. Martins and E. Weide and J. J. Alonso},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_JournalArticle_Marta_2007_IJCFD.pdf, Full-text PDF
https://www.tandfonline.com/doi/10.1080/10618560701678647},
doi = {10.1080/10618560701678647},
issn = {1061-8562},
year = {2007},
date = {2007-10-01},
journal = {International Journal of Computational Fluid Dynamics},
volume = {21},
number = {9--10},
pages = {307--327},
abstract = {A methodology for the rapid development of adjoint solvers for computational fluid dynamics (CFD) models is presented. The approach relies on the use of automatic differentiation (AD) tools to almost completely automate the process of development of discrete adjoint solvers. This methodology is used to produce the adjoint code for two distinct 3D CFD solvers: a cell-centred Euler solver running in single-block, single-processor mode and a multi-block, multi-processor, vertex-centred, magneto-hydrodynamics (MHD) solver. Instead of differentiating the entire source code of the CFD solvers using AD, we have applied it selectively to produce code that computes the transpose of the flux Jacobian matrix and the other partial derivatives that are necessary to compute sensitivities using an adjoint method. The discrete adjoint equations are then solved using the Portable, Extensible Toolkit for Scientific Computation (PETSc) library. The selective application of AD is the principal idea of this new methodology, which we call the AD adjoint (ADjoint). The ADjoint approach has the advantages that it is applicable to any set of governing equations and objective functions and that it is completely consistent with the gradients that would be computed by exact numerical differentiation of the original discrete solver. Furthermore, the approach does not require hand differentiation, thus avoiding the long development times typically required to develop discrete adjoint solvers for partial differential equations, as well as the errors that result from the necessary approximations used during the differentiation of complex systems of conservation laws. These advantages come at the cost of increased memory requirements for the discrete adjoint solver. However, given the amount of memory that is typically available in parallel computers and the trends toward larger numbers of multi-core processors, this disadvantage is rather small when compared with the very significant advantages that are demonstrated. The sensitivities of drag and lift coefficients with respect to different parameters obtained using the discrete adjoint solvers show excellent agreement with the benchmark results produced by the complex-step and finite-difference methods. Furthermore, the overall performance of the method is shown to be better than most conventional adjoint approaches for both CFD solvers used.
Keywords: Partial differential equations, Adjoint solver, Automatic differentiation, Discrete adjoint, Sensitivities, Gradient-based optimization},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
A methodology for the rapid development of adjoint solvers for computational fluid dynamics (CFD) models is presented. The approach relies on the use of automatic differentiation (AD) tools to almost completely automate the process of development of discrete adjoint solvers. This methodology is used to produce the adjoint code for two distinct 3D CFD solvers: a cell-centred Euler solver running in single-block, single-processor mode and a multi-block, multi-processor, vertex-centred, magneto-hydrodynamics (MHD) solver. Instead of differentiating the entire source code of the CFD solvers using AD, we have applied it selectively to produce code that computes the transpose of the flux Jacobian matrix and the other partial derivatives that are necessary to compute sensitivities using an adjoint method. The discrete adjoint equations are then solved using the Portable, Extensible Toolkit for Scientific Computation (PETSc) library. The selective application of AD is the principal idea of this new methodology, which we call the AD adjoint (ADjoint). The ADjoint approach has the advantages that it is applicable to any set of governing equations and objective functions and that it is completely consistent with the gradients that would be computed by exact numerical differentiation of the original discrete solver. Furthermore, the approach does not require hand differentiation, thus avoiding the long development times typically required to develop discrete adjoint solvers for partial differential equations, as well as the errors that result from the necessary approximations used during the differentiation of complex systems of conservation laws. These advantages come at the cost of increased memory requirements for the discrete adjoint solver. However, given the amount of memory that is typically available in parallel computers and the trends toward larger numbers of multi-core processors, this disadvantage is rather small when compared with the very significant advantages that are demonstrated. The sensitivities of drag and lift coefficients with respect to different parameters obtained using the discrete adjoint solvers show excellent agreement with the benchmark results produced by the complex-step and finite-difference methods. Furthermore, the overall performance of the method is shown to be better than most conventional adjoint approaches for both CFD solvers used.
Keywords: Partial differential equations, Adjoint solver, Automatic differentiation, Discrete adjoint, Sensitivities, Gradient-based optimization
Keywords: Partial differential equations, Adjoint solver, Automatic differentiation, Discrete adjoint, Sensitivities, Gradient-based optimization
Proceedings Articles
Cardoso, P. M.; Marta, A. C.; Matos, N. B.
Maximizing UAV range through wing aerostructural optimization: a study on chord, shape, span and composite material properties Proceedings Article
In: Proceedings of the ICEUBI 2024, International Congress on Engineering, pp. 15, Universidade de Beira Interior Covilhã, Portugal, 2024.
@inproceedings{Cardoso:2024:ICEUBI,
title = {Maximizing UAV range through wing aerostructural optimization: a study on chord, shape, span and composite material properties},
author = {P. M. Cardoso and A. C. Marta and N. B. Matos},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Cardoso_2024_ICEUBI.pdf, Full-text PDF
https://iceubi2024.pt/339-2/
https://iceubi2024.pt/},
year = {2024},
date = {2024-11-01},
urldate = {2024-11-01},
booktitle = {Proceedings of the ICEUBI 2024, International Congress on Engineering},
pages = {15},
address = {Covilhã, Portugal},
organization = {Universidade de Beira Interior},
abstract = {In the competitive UAV market, manufacturers strive to enhance performance through advanced design technologies. This study focuses on maximizing the range of a UAV through the use of a gradient-based optimization framework that couples high-fidelity Computational Fluid Dynamics and Computational Structural Dynamics models. The optimization process considers aerodynamic and structural wing design variables (DV), namely chord, airfoil shape and span, panel thickness and fiber orientation of the constitutive composite material. The discrete adjoint method is used to compute derivatives efficiently for a large number of DV and gradient-based optimization. The results demonstrate up to 9.9% increase in range, a 32% improvement in aerodynamic efficiency, despite a 114% increase in wing weight. Addressing both aerodynamic and structural disciplines concurrently offers valuable insights into the trade-offs among different design variables and leads to more efficient UAV designs.
Keywords: Multidisciplinary optimization, Aircraft design, Adjoint method, Free-form deformation, Composite materials},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
In the competitive UAV market, manufacturers strive to enhance performance through advanced design technologies. This study focuses on maximizing the range of a UAV through the use of a gradient-based optimization framework that couples high-fidelity Computational Fluid Dynamics and Computational Structural Dynamics models. The optimization process considers aerodynamic and structural wing design variables (DV), namely chord, airfoil shape and span, panel thickness and fiber orientation of the constitutive composite material. The discrete adjoint method is used to compute derivatives efficiently for a large number of DV and gradient-based optimization. The results demonstrate up to 9.9% increase in range, a 32% improvement in aerodynamic efficiency, despite a 114% increase in wing weight. Addressing both aerodynamic and structural disciplines concurrently offers valuable insights into the trade-offs among different design variables and leads to more efficient UAV designs.
Keywords: Multidisciplinary optimization, Aircraft design, Adjoint method, Free-form deformation, Composite materials
Keywords: Multidisciplinary optimization, Aircraft design, Adjoint method, Free-form deformation, Composite materials
Pacheco, L. M.; Petersson, Ö.; Marta, A. C.; Volle, F.
A high-fidelity MDO framework applied to the design of a high aspect-ratio transport wing Proceedings Article
In: Proceedings of the DLRK 2024 - Deutschen Luft- und Raumfahrtkongress, pp. 10, Deutsche Gesellschaft für Luft- und Raumfahrt Hamburg, Germany, 2024.
@inproceedings{Pacheco:2024:DLRK,
title = {A high-fidelity MDO framework applied to the design of a high aspect-ratio transport wing},
author = {L. M. Pacheco and Ö. Petersson and A. C. Marta and F. Volle},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Pacheco_2024_DLRK.pdf, Full-text PDF
https://www.dglr.de/nc/publikationen/deutscher-luft-und-raumfahrtkongress-2024-netzpublikationen/nach-autor/
https://dlrk2024.dglr.de/},
year = {2024},
date = {2024-09-01},
booktitle = {Proceedings of the DLRK 2024 - Deutschen Luft- und Raumfahrtkongress},
pages = {10},
address = {Hamburg, Germany},
organization = {Deutsche Gesellschaft für Luft- und Raumfahrt},
abstract = {A high-fidelity multidisciplinary design optimisation (MDO) framework was applied to the structural sizing and performance evaluation of high-aspect-ratio transport wings. Using Airbus Defence and Space’s MDO suite, Lagrange, two aircraft models based on the DLR F25, one with an aspect ratio of 15.5 and the other 17.4, were structurally sized through gradient-based methods. These optimisations considered the composite wing covers and spars as sizing parameters, and were subject to strength, buckling and manufacturing constraints. Following the structural sizing, performance prediction analysis for both aircraft were conducted using a high-fidelity MDO framework, which couples Lagrange with DLR’s TAU within FlowSimulator. The aim of this study was to explore the aero-structural trade-offs resulting from an increased aspect ratio. The results show that both aspect ratios exhibited similar trends in skin thickness distribution, ply share allocation, and spar and stringer sizing. However, the higher aspect ratio wing, despite experiencing a 5.51% increase in wing structural mass, demonstrated a 1.47% improvement in Breguet range, primarily due to its reduced induced drag and increased lift generation. The findings highlight the potential of high-aspect-ratio wings to enhance overall performance, albeit at the expense of increased structural weight.
Keywords: Multidisciplinary design optimisation, MDO, Structural sizing, High-fidelity performance analysis},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
A high-fidelity multidisciplinary design optimisation (MDO) framework was applied to the structural sizing and performance evaluation of high-aspect-ratio transport wings. Using Airbus Defence and Space’s MDO suite, Lagrange, two aircraft models based on the DLR F25, one with an aspect ratio of 15.5 and the other 17.4, were structurally sized through gradient-based methods. These optimisations considered the composite wing covers and spars as sizing parameters, and were subject to strength, buckling and manufacturing constraints. Following the structural sizing, performance prediction analysis for both aircraft were conducted using a high-fidelity MDO framework, which couples Lagrange with DLR’s TAU within FlowSimulator. The aim of this study was to explore the aero-structural trade-offs resulting from an increased aspect ratio. The results show that both aspect ratios exhibited similar trends in skin thickness distribution, ply share allocation, and spar and stringer sizing. However, the higher aspect ratio wing, despite experiencing a 5.51% increase in wing structural mass, demonstrated a 1.47% improvement in Breguet range, primarily due to its reduced induced drag and increased lift generation. The findings highlight the potential of high-aspect-ratio wings to enhance overall performance, albeit at the expense of increased structural weight.
Keywords: Multidisciplinary design optimisation, MDO, Structural sizing, High-fidelity performance analysis
Keywords: Multidisciplinary design optimisation, MDO, Structural sizing, High-fidelity performance analysis
Cardoso, P. M.; Marta, A. C.; Matos, N. B.
Aerostructural design of a medium-altitude medium-endurance fixed wing UAV Proceedings Article
In: Proceedings of the ECCOMAS Congress 2024 - 9th European Congress on Computational Methods in Applied Sciences and Engineering, pp. 12, APMTAC, CIMNE, ECCOMAS Lisboa, Portugal, 2024, ISSN: 2696-6999.
@inproceedings{Cardoso:2024:ECCOMAS,
title = {Aerostructural design of a medium-altitude medium-endurance fixed wing UAV},
author = {P. M. Cardoso and A. C. Marta and N. B. Matos},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Cardoso_2024_ECCOMAS.pdf, Full-text PDF
https://www.scipedia.com/sj/eccomas2024
https://eccomas2024.org/},
issn = {2696-6999},
year = {2024},
date = {2024-06-01},
booktitle = {Proceedings of the ECCOMAS Congress 2024 - 9th European Congress on Computational Methods in Applied Sciences and Engineering},
pages = {12},
address = {Lisboa, Portugal},
organization = {APMTAC, CIMNE, ECCOMAS},
abstract = {The UAV market is currently very populated, driving the manufacturers to design more efficient solutions to obtain a competitive edge. A cost-effective approach is to improve existing products using new technologies and design tools. This work addresses the desire of a UAV manufacturer to develop a growth version of an existing Medium-Altitude Medium-Endurance (MAME) Unmanned Aerial Vehicle (UAV). To that end, the aerostructural optimization of the wing is performed using coupled high-fidelity Computational Fluid Dynamics (CFD) and Computational Structural Dynamics (CSD). Gradient-based optimization fed with derivatives of functions of interest computed using the adjoint method are used for computational efficiency. The coupled problem is posed in the aerostructural optimization framework, targeting for maximum aircraft range, being the solution a result of the concurrent discipline analyses. The set of design variables include wing twist distribution, using the free-form deformation approach, material thicknesses and carbon fibre orientations. The optimized wing geometry exhibits a gain of 5% in aircraft range, with 2% better aerodynamic efficiency (L/D) and 63% wing weight reduction. The impact of multilayer composite manufacturing constraints, namely adjacency of ply angles in neighbouring regions and the orthogonality between ply angles, was found not to be significant. The studies identified weaknesses of the baseline wing and provided meaningful engineering insights for the next generation MAME UAV design.
Keywords: Multidisciplinary optimization, Aircraft design, Adjoint method, High-fidelity methods, Free-form deformation, Composite materials},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
The UAV market is currently very populated, driving the manufacturers to design more efficient solutions to obtain a competitive edge. A cost-effective approach is to improve existing products using new technologies and design tools. This work addresses the desire of a UAV manufacturer to develop a growth version of an existing Medium-Altitude Medium-Endurance (MAME) Unmanned Aerial Vehicle (UAV). To that end, the aerostructural optimization of the wing is performed using coupled high-fidelity Computational Fluid Dynamics (CFD) and Computational Structural Dynamics (CSD). Gradient-based optimization fed with derivatives of functions of interest computed using the adjoint method are used for computational efficiency. The coupled problem is posed in the aerostructural optimization framework, targeting for maximum aircraft range, being the solution a result of the concurrent discipline analyses. The set of design variables include wing twist distribution, using the free-form deformation approach, material thicknesses and carbon fibre orientations. The optimized wing geometry exhibits a gain of 5% in aircraft range, with 2% better aerodynamic efficiency (L/D) and 63% wing weight reduction. The impact of multilayer composite manufacturing constraints, namely adjacency of ply angles in neighbouring regions and the orthogonality between ply angles, was found not to be significant. The studies identified weaknesses of the baseline wing and provided meaningful engineering insights for the next generation MAME UAV design.
Keywords: Multidisciplinary optimization, Aircraft design, Adjoint method, High-fidelity methods, Free-form deformation, Composite materials
Keywords: Multidisciplinary optimization, Aircraft design, Adjoint method, High-fidelity methods, Free-form deformation, Composite materials
Gameiro, R. S.; Matos, N. M. B.; Marta, A. C.
Wing aerodynamic design for a MAME UAV using high-fidelity numerical tools Proceedings Article
In: Proceedings of the AeroBest 2023 - II ECCOMAS Thematic Conference on Multidisciplinary Design Optimization of Aerospace Systems, pp. 260-283, IDMEC, Instituto Superior Técnico Lisboa, Portugal, 2023, ISBN: 978-989-53599-4-3.
@inproceedings{Gameiro:2023:AEROBEST,
title = {Wing aerodynamic design for a MAME UAV using high-fidelity numerical tools},
author = {R. S. Gameiro and N. M. B. Matos and A. C. Marta},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Gameiro_2023_AEROBEST.pdf, Full-text PDF
https://www.eccomas.org/publications/conference-proceedings/
https://aerobest2023.idmec.tecnico.ulisboa.pt},
isbn = {978-989-53599-4-3},
year = {2023},
date = {2023-07-01},
booktitle = {Proceedings of the AeroBest 2023 - II ECCOMAS Thematic Conference on Multidisciplinary Design Optimization of Aerospace Systems},
pages = {260-283},
address = {Lisboa, Portugal},
organization = {IDMEC, Instituto Superior Técnico},
abstract = {The UAV market is currently very competitive, with the frequent launch of new products and a wide range of solutions already available, forcing manufacturers to explore the design space faster and more efficiently than in the past. A cost effective approach is to develop growth versions, improving an existing product with new technologies and design tools. Some of these tools include RANS based high fidelity computational fluid dynamics methods and discrete adjoint gradient-based optimization, which will be used in this work on a numerical design framework to explore the aerodynamic shape optimization of a wing, as part of the development of a growth version of a MAME UAV for a leading Portuguese manufacturer. A comprehensive aerodynamic analysis of the current UAV wing will be performed, followed by an optimization procedure to minimize drag subject to a prescribed lift coefficient constraint. To that end, two different starting geometries will be considered and parameterized with common design variables, including twist and chord distribution, sweep and airfoil shape. New optimized geometries for different sets of design variables will be obtained with a significant drag coefficient reduction from the starting geometry. The optimized geometries will approach an elliptical lift distribution, although not exactly considering the trade-offs needed between skin friction and induced drag. Despite the fact that the results obtained here are not considered the final design, as more shape parametrizations and design variables are yet to be explored, they provided a good insight on how the different parametrizations are handled by the design optimization framework and considerations that should be taken.
Keywords: Optimization, Adjoint method, Aircraft design, Computational fluid dynamics, Free-form deformation, Geometric parametrization},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
The UAV market is currently very competitive, with the frequent launch of new products and a wide range of solutions already available, forcing manufacturers to explore the design space faster and more efficiently than in the past. A cost effective approach is to develop growth versions, improving an existing product with new technologies and design tools. Some of these tools include RANS based high fidelity computational fluid dynamics methods and discrete adjoint gradient-based optimization, which will be used in this work on a numerical design framework to explore the aerodynamic shape optimization of a wing, as part of the development of a growth version of a MAME UAV for a leading Portuguese manufacturer. A comprehensive aerodynamic analysis of the current UAV wing will be performed, followed by an optimization procedure to minimize drag subject to a prescribed lift coefficient constraint. To that end, two different starting geometries will be considered and parameterized with common design variables, including twist and chord distribution, sweep and airfoil shape. New optimized geometries for different sets of design variables will be obtained with a significant drag coefficient reduction from the starting geometry. The optimized geometries will approach an elliptical lift distribution, although not exactly considering the trade-offs needed between skin friction and induced drag. Despite the fact that the results obtained here are not considered the final design, as more shape parametrizations and design variables are yet to be explored, they provided a good insight on how the different parametrizations are handled by the design optimization framework and considerations that should be taken.
Keywords: Optimization, Adjoint method, Aircraft design, Computational fluid dynamics, Free-form deformation, Geometric parametrization
Keywords: Optimization, Adjoint method, Aircraft design, Computational fluid dynamics, Free-form deformation, Geometric parametrization
Silva, V. M. T.; Matos, N. M. B.; Marta, A. C.
Wing structural design for a MAME UAV using high-fidelity numerical tools Proceedings Article
In: Proceedings of the AeroBest 2023 - II ECCOMAS Thematic Conference on Multidisciplinary Design Optimization of Aerospace Systems, pp. 240-259, IDMEC, Instituto Superior Técnico Lisboa, Portugal, 2023, ISBN: 978-989-53599-4-3.
@inproceedings{Silva:2023:AEROBEST,
title = {Wing structural design for a MAME UAV using high-fidelity numerical tools},
author = {V. M. T. Silva and N. M. B. Matos and A. C. Marta},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Silva_2023_AEROBEST.pdf, Full-text PDF
https://www.eccomas.org/publications/conference-proceedings/
https://aerobest2023.idmec.tecnico.ulisboa.pt},
isbn = {978-989-53599-4-3},
year = {2023},
date = {2023-07-01},
booktitle = {Proceedings of the AeroBest 2023 - II ECCOMAS Thematic Conference on Multidisciplinary Design Optimization of Aerospace Systems},
pages = {240-259},
address = {Lisboa, Portugal},
organization = {IDMEC, Instituto Superior Técnico},
abstract = {With the rapid growth of the UAV market, the search for more efficient solutions promotes a huge competitive advantage for manufacturers. With the implementation of optimization techniques and the use of high-fidelity analysis in aircraft design, it is possible to develop better solutions. This work addresses the desire of a leading UAV manufacturer to improve its fleet to remain competitive in the surveillance UAV market. For this, a structural analysis tool using the finite element method is demonstrated, which is then used as part of a structural optimization framework. For this demonstration, static analyses of the wingbox of an existing UAV model, with a CFRP material with different lay-ups in certain areas of the model, are carried out for cruise and 4g load case, obtaining results of deformation and failure of this wing. These results help to identify possible weaknesses of the wing, as well as evaluate how the wingbox structural behaviour changes. The goals of this work include the validation of the numerical design framework using available experimental data and the study of alternative wing structural solutions. The results of the experimental and computational analyses presented slight differences. This was the expected behaviour due to model simplifications, which allowed for the the framework to be validated and proven useful. Three new optimal wingbox solutions were found, having a theoretical mass reduction of about 50%, while respecting a safety factor of 1.5. The first was optimised without displacement constraints and the other two had a maximum allowed displacement. These two differ on the optimization starting point to check for possible local minima, which were found.
Keywords: Optimization, Design framework, Adjoint method, Finite element method, Composite materials, Fiber orientation},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
With the rapid growth of the UAV market, the search for more efficient solutions promotes a huge competitive advantage for manufacturers. With the implementation of optimization techniques and the use of high-fidelity analysis in aircraft design, it is possible to develop better solutions. This work addresses the desire of a leading UAV manufacturer to improve its fleet to remain competitive in the surveillance UAV market. For this, a structural analysis tool using the finite element method is demonstrated, which is then used as part of a structural optimization framework. For this demonstration, static analyses of the wingbox of an existing UAV model, with a CFRP material with different lay-ups in certain areas of the model, are carried out for cruise and 4g load case, obtaining results of deformation and failure of this wing. These results help to identify possible weaknesses of the wing, as well as evaluate how the wingbox structural behaviour changes. The goals of this work include the validation of the numerical design framework using available experimental data and the study of alternative wing structural solutions. The results of the experimental and computational analyses presented slight differences. This was the expected behaviour due to model simplifications, which allowed for the the framework to be validated and proven useful. Three new optimal wingbox solutions were found, having a theoretical mass reduction of about 50%, while respecting a safety factor of 1.5. The first was optimised without displacement constraints and the other two had a maximum allowed displacement. These two differ on the optimization starting point to check for possible local minima, which were found.
Keywords: Optimization, Design framework, Adjoint method, Finite element method, Composite materials, Fiber orientation
Keywords: Optimization, Design framework, Adjoint method, Finite element method, Composite materials, Fiber orientation
Portugal, M.; Marta, A. C.
Optimal multi-sensor obstacle detection system for small fixed-wing UAV Proceedings Article
In: Proceedings of the AeroBest 2023 - II ECCOMAS Thematic Conference on Multidisciplinary Design Optimization of Aerospace Systems, pp. 184-208, IDMEC, Instituto Superior Técnico Lisboa, Portugal, 2023, ISBN: 978-989-53599-4-3.
@inproceedings{Portugal:2023:AEROBEST,
title = {Optimal multi-sensor obstacle detection system for small fixed-wing UAV},
author = {M. Portugal and A. C. Marta},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Portugal_2023_AEROBEST.pdf, Full-text PDF
https://www.eccomas.org/publications/conference-proceedings/
https://aerobest2023.idmec.tecnico.ulisboa.pt},
isbn = {978-989-53599-4-3},
year = {2023},
date = {2023-07-01},
booktitle = {Proceedings of the AeroBest 2023 - II ECCOMAS Thematic Conference on Multidisciplinary Design Optimization of Aerospace Systems},
pages = {184-208},
address = {Lisboa, Portugal},
organization = {IDMEC, Instituto Superior Técnico},
abstract = {This work provides a solution for the safety enhancement of small fixed-wing UAVs regarding obstacle detection during flight. The main goal is to implement an optimal multi-sensor system configuration. To achieve it, preceding works regarding the integration of available sensors in such systems were studied. As a result, select sensors (ultrasonic sensor, laser rangefinder, LIDAR and RADAR) were modeled for collision detection and avoidance simulations using the potential fields method. An optimization study using a genetic algorithm was conducted to find the sets of sensors and respective orientation that result in the best collision avoidance performance. To do so, a set of randomly generated collision scenarios with both stationary and moving obstacles were generated. This study resulted in relatively simple detection configurations that still provided high collision avoidance success rate. The ultrasonic sensor revealed to be inappropriate given its short range, while the laser rangefinder benefited from long range but had very limited field-of-view. In contrast, both the LIDAR and the RADAR are the most promising, as they exhibit not only a significant range but also a broad field-of-view. The best multi-sensor configurations were either a front-facing LIDAR or RADAR, complimented by a pair of laser rangefinders pointing sideways at an angle of 10 or 63 degrees, respectively. Once the hardware that should integrate an optimal system was known and available, the assembly of the final system, including the sensors and a PixHawk flight controller, was designed. The appropriate software (PX4 and QGroundControl) was also built and adapted to the current work. To validate the proposed system, all sensors were first individually tested before assembling the complete system. The bench tests attested the accuracy of the sensor specifications and previous simulations. As such, ground tests using a simple rover shall follow. Once the system is validated under these conditions, flight tests may begin.
Keywords: Sense and avoidance, Collision avoidance, Sensor fusion, Optimization, Laser rangefinder, LIDAR},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
This work provides a solution for the safety enhancement of small fixed-wing UAVs regarding obstacle detection during flight. The main goal is to implement an optimal multi-sensor system configuration. To achieve it, preceding works regarding the integration of available sensors in such systems were studied. As a result, select sensors (ultrasonic sensor, laser rangefinder, LIDAR and RADAR) were modeled for collision detection and avoidance simulations using the potential fields method. An optimization study using a genetic algorithm was conducted to find the sets of sensors and respective orientation that result in the best collision avoidance performance. To do so, a set of randomly generated collision scenarios with both stationary and moving obstacles were generated. This study resulted in relatively simple detection configurations that still provided high collision avoidance success rate. The ultrasonic sensor revealed to be inappropriate given its short range, while the laser rangefinder benefited from long range but had very limited field-of-view. In contrast, both the LIDAR and the RADAR are the most promising, as they exhibit not only a significant range but also a broad field-of-view. The best multi-sensor configurations were either a front-facing LIDAR or RADAR, complimented by a pair of laser rangefinders pointing sideways at an angle of 10 or 63 degrees, respectively. Once the hardware that should integrate an optimal system was known and available, the assembly of the final system, including the sensors and a PixHawk flight controller, was designed. The appropriate software (PX4 and QGroundControl) was also built and adapted to the current work. To validate the proposed system, all sensors were first individually tested before assembling the complete system. The bench tests attested the accuracy of the sensor specifications and previous simulations. As such, ground tests using a simple rover shall follow. Once the system is validated under these conditions, flight tests may begin.
Keywords: Sense and avoidance, Collision avoidance, Sensor fusion, Optimization, Laser rangefinder, LIDAR
Keywords: Sense and avoidance, Collision avoidance, Sensor fusion, Optimization, Laser rangefinder, LIDAR
Matos, N. M. B.; Marta, A. C.
Fixed-wing UAV model identification for longitudinal motion using first-order models And limited flight testing Proceedings Article
In: Proceedings of the AeroBest 2023 - II ECCOMAS Thematic Conference on Multidisciplinary Design Optimization of Aerospace Systems, pp. 120-141, IDMEC, Instituto Superior Técnico Lisboa, Portugal, 2023, ISBN: 978-989-53599-4-3.
@inproceedings{Matos:2023:AEROBEST,
title = {Fixed-wing UAV model identification for longitudinal motion using first-order models And limited flight testing},
author = {N. M. B. Matos and A. C. Marta},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Matos_2023_AEROBEST.pdf, Full-text PDF
https://www.eccomas.org/publications/conference-proceedings/
https://aerobest2023.idmec.tecnico.ulisboa.pt},
isbn = {978-989-53599-4-3},
year = {2023},
date = {2023-07-01},
booktitle = {Proceedings of the AeroBest 2023 - II ECCOMAS Thematic Conference on Multidisciplinary Design Optimization of Aerospace Systems},
pages = {120-141},
address = {Lisboa, Portugal},
organization = {IDMEC, Instituto Superior Técnico},
abstract = {System identification plays an important role in the determination of an aircraft behaviour that helps predict and simulate different scenarios crucial for control, mission or safety assurance analysis. This work describes the system identification process of a medium sized UAV through the usage of limited flight test data and a non-linear model dynamic simulator. The proposed solution uses parameter based first-order models to describe the various aerodynamic properties of the UAV. The parameter estimation is based on a least square error optimization algorithm in a time-domain formulation starting from a low-fidelity aerodynamic analysis solution. The work focuses on the longitudinal motion by using routine flight test data of pitch down and pitch up manoeuvres to excite the longitudinal dynamics. The optimization geared towards parameter tuning used a combination of pitch and altitude UAV model response as measure of accuracy. Very significant improvements in the UAV model response are obtained with the resulting optimizer found relevant longitudinal aerodynamic and control derivatives. The pitching moment derivatives proved to be the most important parameters, as expected. The process hereby described is meant to be usable on any fixed-wing UAV with limited planned flight test data achieving reasonable accuracy.
Keywords: Aircraft design, Optimization, Aerospace, System identification, Aircraft dynamics},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
System identification plays an important role in the determination of an aircraft behaviour that helps predict and simulate different scenarios crucial for control, mission or safety assurance analysis. This work describes the system identification process of a medium sized UAV through the usage of limited flight test data and a non-linear model dynamic simulator. The proposed solution uses parameter based first-order models to describe the various aerodynamic properties of the UAV. The parameter estimation is based on a least square error optimization algorithm in a time-domain formulation starting from a low-fidelity aerodynamic analysis solution. The work focuses on the longitudinal motion by using routine flight test data of pitch down and pitch up manoeuvres to excite the longitudinal dynamics. The optimization geared towards parameter tuning used a combination of pitch and altitude UAV model response as measure of accuracy. Very significant improvements in the UAV model response are obtained with the resulting optimizer found relevant longitudinal aerodynamic and control derivatives. The pitching moment derivatives proved to be the most important parameters, as expected. The process hereby described is meant to be usable on any fixed-wing UAV with limited planned flight test data achieving reasonable accuracy.
Keywords: Aircraft design, Optimization, Aerospace, System identification, Aircraft dynamics
Keywords: Aircraft design, Optimization, Aerospace, System identification, Aircraft dynamics
Coelho, V. L.; Silva, P. A.; Sá, P. J.; Caetano, J. V.; Félix, L. F.; Afonso, F.; Marta, A. C.
Design of a tactical eVTOL UAV with a hydrogen fuel cell Proceedings Article
In: Proceedings of the ICUAS 2021 - International Conference on Unmanned Aircraft Systems, pp. 94-103, ICUAS Association, Inc. Dubrovnik, Croatia, 2022.
@inproceedings{Coelho:2022:ICUAS,
title = {Design of a tactical eVTOL UAV with a hydrogen fuel cell},
author = {V. L. Coelho and P. A. Silva and P. J. Sá and J. V. Caetano and L. F. Félix and F. Afonso and A. C. Marta},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Coelho_2022_ICUAS.pdf, Full-text PDF
https://ieeexplore.ieee.org/document/9836046
http://www.uasconferences.com/2022_icuas/},
doi = {10.1109/ICUAS54217.2022.9836046},
year = {2022},
date = {2022-06-01},
booktitle = {Proceedings of the ICUAS 2021 - International Conference on Unmanned Aircraft Systems},
pages = {94-103},
address = {Dubrovnik, Croatia},
organization = {ICUAS Association, Inc.},
abstract = {Batteries have a substantial impact on the weight of electric Unmanned Aerial Vehicles (UAVs). With fuel cells being considered a possible alternative to batteries, the present work aims to design a fixed-wing UAV with vertical take-off and landing capability and a fuel cell-based propulsion system. Building on the tactical requirements of the Portuguese Air Force, this project covers all phases of aerodynamic, structural, and propulsive design, in which Blade Element Theory, Computational Fluid Dynamics, and Finite Element Method analyses were used. The final design resulted in an inverted V-tail, 21.7kg take-off weight UAV capable of three hours of flight time, more than twice the endurance of a battery-propelled version. To further stimulate scientific knowledge sharing, the authors have made all simulations and Computational Aided Designs available to the public via repository.
Keywords: Electric propulsion, Long endurance, Hybrid VTOL, Aircraft Design, Systems integration, Preliminary design},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Batteries have a substantial impact on the weight of electric Unmanned Aerial Vehicles (UAVs). With fuel cells being considered a possible alternative to batteries, the present work aims to design a fixed-wing UAV with vertical take-off and landing capability and a fuel cell-based propulsion system. Building on the tactical requirements of the Portuguese Air Force, this project covers all phases of aerodynamic, structural, and propulsive design, in which Blade Element Theory, Computational Fluid Dynamics, and Finite Element Method analyses were used. The final design resulted in an inverted V-tail, 21.7kg take-off weight UAV capable of three hours of flight time, more than twice the endurance of a battery-propelled version. To further stimulate scientific knowledge sharing, the authors have made all simulations and Computational Aided Designs available to the public via repository.
Keywords: Electric propulsion, Long endurance, Hybrid VTOL, Aircraft Design, Systems integration, Preliminary design
Keywords: Electric propulsion, Long endurance, Hybrid VTOL, Aircraft Design, Systems integration, Preliminary design
Alves, B. M.; Marta, A. C.; Félix, L. F.
Multidisciplinary optimisation of an eVTOL UAV with a hydrogen fuel cell Proceedings Article
In: Proceedings of the ICUAS 2021 - International Conference on Unmanned Aircraft Systems, pp. 134-143, ICUAS Association, Inc. Dubrovnik, Croatia, 2022.
@inproceedings{Alves:2022:ICUAS,
title = {Multidisciplinary optimisation of an eVTOL UAV with a hydrogen fuel cell},
author = {B. M. Alves and A. C. Marta and L. F. Félix},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Alves_2022_ICUAS.pdf, Full-text PDF
https://ieeexplore.ieee.org/document/9836228
http://www.uasconferences.com/2022_icuas/},
doi = {10.1109/ICUAS54217.2022.9836228},
year = {2022},
date = {2022-06-01},
booktitle = {Proceedings of the ICUAS 2021 - International Conference on Unmanned Aircraft Systems},
pages = {134-143},
address = {Dubrovnik, Croatia},
organization = {ICUAS Association, Inc.},
abstract = {To explore the use of hydrogen fuel cells as a feasible alternative on Unmanned Aerial Vehicles (UAVs), a class I concept was designed at the Portuguese Air Force
Research Centre (CIAFA). This work focuses on the Multidisciplinary Design Optimisation (MDO) methodology that was used to improve the 3h endurance of the baseline concept that had a Maximum Take-Off Weight of 21.6 kg, using 148 g of hydrogen and a 800 W fuel cell to power conventional flight operations. Another propulsive system comprised of batteries and rotors is used for Vertical Take-Off and Landing (VTOL). MDO was performed with the aid of OpenAeroStruct, a low fidelity software that combines Finite Element Analysis (FEA) and Vortex Lattice Method (VLM) to model lifting surfaces. Initially, a cruise and a load flight conditions were used with structural parameters and geometric twist as design variables. In a second approach, complexity was increased by including taper, wing chord and span as design variables in the problem formulation. Lastly, a third flight condition was introduced to ensure stall requirements were met. The use of MDO led to a 21% increase in endurance with a smaller wing, while satisfying all imposed constraints. This work marks an important milestone in the development of a future prototype at the CIAFA.
Keywords: Aerostructural optimization, Electric propulsion, Long endurance, Hybrid VTOL, Aircraft design, Conceptual design},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
To explore the use of hydrogen fuel cells as a feasible alternative on Unmanned Aerial Vehicles (UAVs), a class I concept was designed at the Portuguese Air Force
Research Centre (CIAFA). This work focuses on the Multidisciplinary Design Optimisation (MDO) methodology that was used to improve the 3h endurance of the baseline concept that had a Maximum Take-Off Weight of 21.6 kg, using 148 g of hydrogen and a 800 W fuel cell to power conventional flight operations. Another propulsive system comprised of batteries and rotors is used for Vertical Take-Off and Landing (VTOL). MDO was performed with the aid of OpenAeroStruct, a low fidelity software that combines Finite Element Analysis (FEA) and Vortex Lattice Method (VLM) to model lifting surfaces. Initially, a cruise and a load flight conditions were used with structural parameters and geometric twist as design variables. In a second approach, complexity was increased by including taper, wing chord and span as design variables in the problem formulation. Lastly, a third flight condition was introduced to ensure stall requirements were met. The use of MDO led to a 21% increase in endurance with a smaller wing, while satisfying all imposed constraints. This work marks an important milestone in the development of a future prototype at the CIAFA.
Keywords: Aerostructural optimization, Electric propulsion, Long endurance, Hybrid VTOL, Aircraft design, Conceptual design
Research Centre (CIAFA). This work focuses on the Multidisciplinary Design Optimisation (MDO) methodology that was used to improve the 3h endurance of the baseline concept that had a Maximum Take-Off Weight of 21.6 kg, using 148 g of hydrogen and a 800 W fuel cell to power conventional flight operations. Another propulsive system comprised of batteries and rotors is used for Vertical Take-Off and Landing (VTOL). MDO was performed with the aid of OpenAeroStruct, a low fidelity software that combines Finite Element Analysis (FEA) and Vortex Lattice Method (VLM) to model lifting surfaces. Initially, a cruise and a load flight conditions were used with structural parameters and geometric twist as design variables. In a second approach, complexity was increased by including taper, wing chord and span as design variables in the problem formulation. Lastly, a third flight condition was introduced to ensure stall requirements were met. The use of MDO led to a 21% increase in endurance with a smaller wing, while satisfying all imposed constraints. This work marks an important milestone in the development of a future prototype at the CIAFA.
Keywords: Aerostructural optimization, Electric propulsion, Long endurance, Hybrid VTOL, Aircraft design, Conceptual design
Félix, L.; Coelho, V.; Cruz, G.; Mendes, P.; Oliveira, T.; Caetano, J.; Alves, B.; Silva, P.; Sá, P.; Marta, A. C.; Afonso, F.; Ferreira, F.
Desenvolvimento de uma aeronave não tripulada VTOL com utilização de hidrogénio como fonte de energia Proceedings Article
In: Proceedings of the ECM 2021 - 3º Encontro de Investigação e Desenvolvimento em Ciências Militares, CINAMIL, Academia Militar Lisboa, Portugal, 2021.
@inproceedings{Felix:2021:ECM,
title = {Desenvolvimento de uma aeronave não tripulada VTOL com utilização de hidrogénio como fonte de energia},
author = {L. Félix and V. Coelho and G. Cruz and P. Mendes and T. Oliveira and J. Caetano and B. Alves and P. Silva and P. Sá and A. C. Marta and F. Afonso and F. Ferreira},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Felix_2021_ECM.pdf, Full-text PDF
https://academiamilitar.pt/ecm2021.html},
year = {2021},
date = {2021-11-01},
urldate = {2021-11-01},
booktitle = {Proceedings of the ECM 2021 - 3º Encontro de Investigação e Desenvolvimento em Ciências Militares},
address = {Lisboa, Portugal},
organization = {CINAMIL, Academia Militar},
series = {ID 45},
abstract = {Apresentação de um projeto interno do Centro de Investigação da Academia da Força Aérea para o desenvolvimento de uma aeronave não tripulada classe I com capacidade de descolagem e aterragem vertical e energia fornecida por uma célula de combustível de hidrogénio. A aeronave deverá ter uma autonomia mínima de 2 horas e capacidade para 2 kg de payload. Não deve exceder os 25 kg de peso máximo à descolagem.
Keywords: UAV, VTOL, Célula de combustível de hidrogénio, Projeto aeronáutico},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Apresentação de um projeto interno do Centro de Investigação da Academia da Força Aérea para o desenvolvimento de uma aeronave não tripulada classe I com capacidade de descolagem e aterragem vertical e energia fornecida por uma célula de combustível de hidrogénio. A aeronave deverá ter uma autonomia mínima de 2 horas e capacidade para 2 kg de payload. Não deve exceder os 25 kg de peso máximo à descolagem.
Keywords: UAV, VTOL, Célula de combustível de hidrogénio, Projeto aeronáutico
Keywords: UAV, VTOL, Célula de combustível de hidrogénio, Projeto aeronáutico
Rocha, T.; Oliveira, A.; ca L. E,; Marta, A. C.
Estudo experimental de endplates para asas finitas Proceedings Article
In: Proceedings of the CNME 2020 - 12º Congresso Nacional de Mecânica Experimental, Instituto Politécnico de Leiria Leiria, Portugal, 2021.
@inproceedings{Rocha:2021:CNME,
title = {Estudo experimental de endplates para asas finitas},
author = {T. Rocha and A. Oliveira and ca L. E and A. C. Marta},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Rocha_2021_CNME.pdf, Full-text PDF
https://cnme2020.ipleiria.pt/},
year = {2021},
date = {2021-09-01},
urldate = {2021-09-01},
booktitle = {Proceedings of the CNME 2020 - 12º Congresso Nacional de Mecânica Experimental},
address = {Leiria, Portugal},
organization = {Instituto Politécnico de Leiria},
abstract = {Os apêndices aerodinâmicos dos carros Formula Student têm-se tornado cada vez mais complexos. No entanto, devido às velocidades relativamente baixas atingidas pelos carros e aos constrangimentos impostos pelas regras das competições, a força de sustentação (downforce) que os carros produzem acarreta consigo elevados valores da força de resistência. Um conceito alternativo para as endplates da asa traseira de um Formula Student é apresentado com o objetivo de reduzir a força de resistência global da mesma. Para esse efeito, perfis alares são usados como secção das endplates. Um modelo à escala 1:2,5 foi construído para testar em túnel de vento. Três configurações geométricas foram testadas: asa sem placas; asa com placas planas e placas com perfil. A utilização de duas instalações experimentais permitiu avaliar a influência do número de Reynolds no desempenho da solução proposta. Os resultados obtidos confirmaram o potencial do conceito e mostraram efeitos de escala significativos associados aos baixos números de Reynolds.
Keywords: Formula student, Asas finitas, Endplates, Túnel de vento, Sustentação e resistência},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Os apêndices aerodinâmicos dos carros Formula Student têm-se tornado cada vez mais complexos. No entanto, devido às velocidades relativamente baixas atingidas pelos carros e aos constrangimentos impostos pelas regras das competições, a força de sustentação (downforce) que os carros produzem acarreta consigo elevados valores da força de resistência. Um conceito alternativo para as endplates da asa traseira de um Formula Student é apresentado com o objetivo de reduzir a força de resistência global da mesma. Para esse efeito, perfis alares são usados como secção das endplates. Um modelo à escala 1:2,5 foi construído para testar em túnel de vento. Três configurações geométricas foram testadas: asa sem placas; asa com placas planas e placas com perfil. A utilização de duas instalações experimentais permitiu avaliar a influência do número de Reynolds no desempenho da solução proposta. Os resultados obtidos confirmaram o potencial do conceito e mostraram efeitos de escala significativos associados aos baixos números de Reynolds.
Keywords: Formula student, Asas finitas, Endplates, Túnel de vento, Sustentação e resistência
Keywords: Formula student, Asas finitas, Endplates, Túnel de vento, Sustentação e resistência
Alturas, N.; Marta, A. C.
Optimal conceptual design of obstacle detection system for small fixed-wing UAVs Proceedings Article
In: Proceedings of the AeroBest 2021 - ECCOMAS Thematic Conference on Multidisciplinary Design Optimization of Aerospace Systems, pp. 324-342, IDMEC, Instituto Superior Técnico Lisboa, Portugal, 2021, ISBN: 978-989-99424-8-6.
@inproceedings{Alturas:2021:AEROBEST,
title = {Optimal conceptual design of obstacle detection system for small fixed-wing UAVs},
author = {N. Alturas and A. C. Marta},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Alturas_2021_AEROBEST.pdf, Full-text PDF
https://www.eccomas.org/publications/conference-proceedings/
https://aerobest2021.idmec.tecnico.ulisboa.pt},
isbn = {978-989-99424-8-6},
year = {2021},
date = {2021-07-01},
booktitle = {Proceedings of the AeroBest 2021 - ECCOMAS Thematic Conference on Multidisciplinary Design Optimization of Aerospace Systems},
pages = {324-342},
address = {Lisboa, Portugal},
organization = {IDMEC, Instituto Superior Técnico},
abstract = {A solution for the enhancement of safety during the flight of small fixed-wing UAVs, regarding the detection of obstacles during flight, is presented. This was achieved by making a market study on available sensors to find the most suitable to equip a UAV and by modeling them, so that these models could be integrated into collision detection and avoidance simulations. Different tracking filters and sensor fusion techniques were studied, where the Converted Measurement Kalman Filter and the Weighted Filter technique were found to be the best to implement. In the simulations, the Potential Fields avoidance method was chosen for being computationally inexpensive and for providing feasible solutions in real time. Several parametric studies were conducted to test the performance of the selected sensors and to assess how their different parameters affect the success of the obstacle avoidance. An optimization study was also conducted, using a global optimizer, to find the orientation of sensors, for different sets of sensors, that results in the best performance for a set of randomly generated collision scenarios with both stationary and moving obstacles. Relatively simple detection configurations were found that still provide high collision avoidance success rate.
Keywords: Potential fields, Genetic algorithm, Kalman filter, Unbiased conversion, Sensor fusion},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
A solution for the enhancement of safety during the flight of small fixed-wing UAVs, regarding the detection of obstacles during flight, is presented. This was achieved by making a market study on available sensors to find the most suitable to equip a UAV and by modeling them, so that these models could be integrated into collision detection and avoidance simulations. Different tracking filters and sensor fusion techniques were studied, where the Converted Measurement Kalman Filter and the Weighted Filter technique were found to be the best to implement. In the simulations, the Potential Fields avoidance method was chosen for being computationally inexpensive and for providing feasible solutions in real time. Several parametric studies were conducted to test the performance of the selected sensors and to assess how their different parameters affect the success of the obstacle avoidance. An optimization study was also conducted, using a global optimizer, to find the orientation of sensors, for different sets of sensors, that results in the best performance for a set of randomly generated collision scenarios with both stationary and moving obstacles. Relatively simple detection configurations were found that still provide high collision avoidance success rate.
Keywords: Potential fields, Genetic algorithm, Kalman filter, Unbiased conversion, Sensor fusion
Keywords: Potential fields, Genetic algorithm, Kalman filter, Unbiased conversion, Sensor fusion
Alves, B. M.; Coelho, V. L.; Silva, P. A.; Marta, A. C.; Félix, L. F.; Afonso, F.; Sá, P. J.; Caetano, J. V.
Design of a hydrogen powered small electric fixed-wing UAV with VTOL capability Proceedings Article
In: Proceedings of the AeroBest 2021 - ECCOMAS Thematic Conference on Multidisciplinary Design Optimization of Aerospace Systems, pp. 290-304, IDMEC, Instituto Superior Técnico Lisboa, Portugal, 2021, ISBN: 978-989-99424-8-6.
@inproceedings{Alves:2021:AEROBEST,
title = {Design of a hydrogen powered small electric fixed-wing UAV with VTOL capability},
author = {B. M. Alves and V. L. Coelho and P. A. Silva and A. C. Marta and L. F. Félix and F. Afonso and P. J. Sá and J. V. Caetano},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Alves_2021_AEROBEST.pdf, Full-text PDF
https://aerobest2021.idmec.tecnico.ulisboa.pt
https://www.eccomas.org/publications/conference-proceedings/},
isbn = {978-989-99424-8-6},
year = {2021},
date = {2021-07-01},
booktitle = {Proceedings of the AeroBest 2021 - ECCOMAS Thematic Conference on Multidisciplinary Design Optimization of Aerospace Systems},
pages = {290-304},
address = {Lisboa, Portugal},
organization = {IDMEC, Instituto Superior Técnico},
abstract = {Most electric small UAVs require large batteries, which lead to increased weight and low endurance. With the current development of new energy sources and emerging technologies, the present work aims to design a fixed-wing UAV with vertical take-off and landing (VTOL) capability using a fuel cell-based propulsion system. The design requirements made by the Portuguese Air Force include a maximum take-off mass of 25 kg and a minimum flight time of two hours. To accomplish these, a conceptual design framework was developed, supported by fast estimates for the disciplines of aerodynamics, structures, propulsion and controls, and a multi-objective optimisation approach led to the initial UAV configuration and sizing. The different discipline models were coupled and multidisciplinary optimisation was conducted to find the UAV optimal design. This process led to a 22 kg aircraft, having a maximum endurance over 3 hours with a 7.2L hydrogen tank, assisted with batteries for VTOL and climb. The results obtained suggest that the application of the hydrogen-powered fuel cell system meets the requirements set, while also proving to be a feasible alternative to conventional solutions.
Keywords: Fuel cell, Green aircraft, MDO, Multi-objective optimisation, Multi-rotor, Pusher configuration},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Most electric small UAVs require large batteries, which lead to increased weight and low endurance. With the current development of new energy sources and emerging technologies, the present work aims to design a fixed-wing UAV with vertical take-off and landing (VTOL) capability using a fuel cell-based propulsion system. The design requirements made by the Portuguese Air Force include a maximum take-off mass of 25 kg and a minimum flight time of two hours. To accomplish these, a conceptual design framework was developed, supported by fast estimates for the disciplines of aerodynamics, structures, propulsion and controls, and a multi-objective optimisation approach led to the initial UAV configuration and sizing. The different discipline models were coupled and multidisciplinary optimisation was conducted to find the UAV optimal design. This process led to a 22 kg aircraft, having a maximum endurance over 3 hours with a 7.2L hydrogen tank, assisted with batteries for VTOL and climb. The results obtained suggest that the application of the hydrogen-powered fuel cell system meets the requirements set, while also proving to be a feasible alternative to conventional solutions.
Keywords: Fuel cell, Green aircraft, MDO, Multi-objective optimisation, Multi-rotor, Pusher configuration
Keywords: Fuel cell, Green aircraft, MDO, Multi-objective optimisation, Multi-rotor, Pusher configuration
Rocha, I. M. D.; Marta, A. C.
Aeroelastic wing analysis and design Proceedings Article
In: Proceedings of the SymComp 2019 - 4th International Conference on Numerical and Symbolic Computation, pp. 41-60, GIMOSM,ISEL Porto, Portugal, 2019, ISBN: 978-989-99410-5-2.
@inproceedings{Rocha:2019:SYMCOMP,
title = {Aeroelastic wing analysis and design},
author = {I. M. D. Rocha and A. C. Marta},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Rocha_2019_SYMCOMP.pdf, Full-text PDF
http://symcomp2019.isel.pt
https://www.eccomas.org/publications/conference-proceedings/},
isbn = {978-989-99410-5-2},
year = {2019},
date = {2019-04-01},
booktitle = {Proceedings of the SymComp 2019 - 4th International Conference on Numerical and Symbolic Computation},
pages = {41-60},
address = {Porto, Portugal},
organization = {GIMOSM,ISEL},
abstract = {Since the early days of aviation, aeroelastic problems have shown to be some of the most challenging to solve. With the development of numerical methods, the study of aircraft structures and their interaction with the surrounding air flow at different flight conditions has become easily accessible and, thus, is now mandatory in the design phase of an aircraft. This work focuses on the development of a numerical tool for aircraft wing fluid-structure interaction (FSI) analyses, in which the external airflow and the internal structure interact. A panel method was implemented for the aerodynamic analysis and a finite-element model using equivalent beam elements was implemented for the structural analysis, both coded in MATLAB R language. Each analysis models were successfully individually verified against other bibliographic sources and then the two disciplines were coupled into the FSI numerical tool. To validate the accuracy of the numerical tool to predict aeroelastic parameters, such as flutter and divergence speeds, a half wing prototype was built and tested in a wind tunnel. The wing shape was parameterized using area, airfoil cross-section shape, aspect ratio, taper ratio, sweep angle and dihedral angle. Before the optimization, a parametric study was conducted to study the influence of these parameters in the wing performance. The validated FSI tool was then used in an optimization framework to obtain an optimized wing shape with the objective of maximizing the lift-to-drag ratio whilst guaranteeing that flutter and divergence behavior are not worse than that of the baseline wing.
Keywords: Aircraft design, Flutter, Divergence speed, Fluid-structure interaction, Wind tunnel, Optimization},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Since the early days of aviation, aeroelastic problems have shown to be some of the most challenging to solve. With the development of numerical methods, the study of aircraft structures and their interaction with the surrounding air flow at different flight conditions has become easily accessible and, thus, is now mandatory in the design phase of an aircraft. This work focuses on the development of a numerical tool for aircraft wing fluid-structure interaction (FSI) analyses, in which the external airflow and the internal structure interact. A panel method was implemented for the aerodynamic analysis and a finite-element model using equivalent beam elements was implemented for the structural analysis, both coded in MATLAB R language. Each analysis models were successfully individually verified against other bibliographic sources and then the two disciplines were coupled into the FSI numerical tool. To validate the accuracy of the numerical tool to predict aeroelastic parameters, such as flutter and divergence speeds, a half wing prototype was built and tested in a wind tunnel. The wing shape was parameterized using area, airfoil cross-section shape, aspect ratio, taper ratio, sweep angle and dihedral angle. Before the optimization, a parametric study was conducted to study the influence of these parameters in the wing performance. The validated FSI tool was then used in an optimization framework to obtain an optimized wing shape with the objective of maximizing the lift-to-drag ratio whilst guaranteeing that flutter and divergence behavior are not worse than that of the baseline wing.
Keywords: Aircraft design, Flutter, Divergence speed, Fluid-structure interaction, Wind tunnel, Optimization
Keywords: Aircraft design, Flutter, Divergence speed, Fluid-structure interaction, Wind tunnel, Optimization
Cardoso, S. A.; Pereira, J. L.; Marta, A. C.
Design for crashworthiness of an eletric vehicle Proceedings Article
In: Proceedings of the SymComp 2019 - 4th International Conference on Numerical and Symbolic Computation, pp. 61-80, GIMOSM,ISEL Porto, Portugal, 2019, ISBN: 978-989-99410-5-2.
@inproceedings{Cardoso:2019:SYMCOMP,
title = {Design for crashworthiness of an eletric vehicle},
author = {S. A. Cardoso and J. L. Pereira and A. C. Marta},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Cardoso_2019_SYMCOMP.pdf, Full-text PDF
http://symcomp2019.isel.pt
https://www.eccomas.org/publications/conference-proceedings/},
isbn = {978-989-99410-5-2},
year = {2019},
date = {2019-04-01},
booktitle = {Proceedings of the SymComp 2019 - 4th International Conference on Numerical and Symbolic Computation},
pages = {61-80},
address = {Porto, Portugal},
organization = {GIMOSM,ISEL},
abstract = {Modern car design concepts should guarantee not only adequate dynamic characteristics but also meet strict environmental and safety standards. Regarding the later, some organizations and governments have defined regulations to include passive and active protection systems in vehicles as an effort to further reduce the number of road casualties. This work focused on the design of passive structures that maximize the vehicle crashworthiness, thus protecting its occupants in case of collisions. The two cases most common collision cases were considered, namely frontal impact and side impact, following the Euro NCAP frontal full width and the side pole impact test protocols, respectively. An optimal design process was adopted, based on the geometric parameterization of the structure, finite-element analysis models and the use a multi-objective genetic algorithm to find the best solutions. The solutions represent a trade-off between crash performance and weight, all subjected to additional compliance constraints. For the frontal impact case, four types of vehicles were analyzed. While for the lighter case a single primary structure was developed to absorb the corresponding energy during the impact, for the heavier vehicles a secondary structure was added. For the side impact case, several beam configurations were tested, being selected a multi-thickness beam with a quadrangular misaligned cross-sectional shape. The final geometries have proven to fulfill all the requirements, while still exhibiting a good trade-off between weight increase and energy absorption during the impact. The inclusion of a secondary structure for frontal impact proved to have satisfactory effects on the overall behavior during the crash events. It was shown that this methodology led to optimized beam configurations in an efficient matter, saving valuable engineering time in the iterative process.
Keywords: Collision, Crash performance, Energy absorption, Optimization, Pareto front, Genetic algorithm},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Modern car design concepts should guarantee not only adequate dynamic characteristics but also meet strict environmental and safety standards. Regarding the later, some organizations and governments have defined regulations to include passive and active protection systems in vehicles as an effort to further reduce the number of road casualties. This work focused on the design of passive structures that maximize the vehicle crashworthiness, thus protecting its occupants in case of collisions. The two cases most common collision cases were considered, namely frontal impact and side impact, following the Euro NCAP frontal full width and the side pole impact test protocols, respectively. An optimal design process was adopted, based on the geometric parameterization of the structure, finite-element analysis models and the use a multi-objective genetic algorithm to find the best solutions. The solutions represent a trade-off between crash performance and weight, all subjected to additional compliance constraints. For the frontal impact case, four types of vehicles were analyzed. While for the lighter case a single primary structure was developed to absorb the corresponding energy during the impact, for the heavier vehicles a secondary structure was added. For the side impact case, several beam configurations were tested, being selected a multi-thickness beam with a quadrangular misaligned cross-sectional shape. The final geometries have proven to fulfill all the requirements, while still exhibiting a good trade-off between weight increase and energy absorption during the impact. The inclusion of a secondary structure for frontal impact proved to have satisfactory effects on the overall behavior during the crash events. It was shown that this methodology led to optimized beam configurations in an efficient matter, saving valuable engineering time in the iterative process.
Keywords: Collision, Crash performance, Energy absorption, Optimization, Pareto front, Genetic algorithm
Keywords: Collision, Crash performance, Energy absorption, Optimization, Pareto front, Genetic algorithm
Rodrigues, S. S.; Marta, A. C.
On the treatment of multirow interface in aerodynamic turbomachinery adjoint solvers Proceedings Article
In: Proceedings of the EngOpt 2018 - 6th International Conference on Engineering Optimization, pp. 879-887, IDMEC, Instituto Superior Técnico Lisboa, Portugal, 2018.
@inproceedings{Rodrigues:2018:ENGOPT,
title = {On the treatment of multirow interface in aerodynamic turbomachinery adjoint solvers},
author = {S. S. Rodrigues and A. C. Marta},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Rodrigues_2018_ENGOPT.pdf, Full-text PDF
https://engopt2018.tecnico.ulisboa.pt/
https://link.springer.com/chapter/10.1007/978-3-319-97773-7_76},
doi = {10.1007/978-3-319-97773-7_76},
year = {2018},
date = {2018-09-01},
booktitle = {Proceedings of the EngOpt 2018 - 6th International Conference on Engineering Optimization},
pages = {879-887},
address = {Lisboa, Portugal},
organization = {IDMEC, Instituto Superior Técnico},
abstract = {The currently available computational power and improvements of high-fidelity numerical simulations have lead to an increased use of computational fluid dynamics (CFD) in the analysis of turbomachinery flows, particularly in design environments. The optimization cases often contain up to thousands of design variables and gradient-based (GB) optimization algorithms are typically selected due to their efficiency. The adjoint method is key to efficiently compute the derivatives required by the GB algorithms, with a computational cost nearly independent of the number of design variables. In this paper we present the details of the development of an adjoint multirow interface based on the mixing-plane treatment to extend an already existing adjoint solver using the ADjoint approach. The mixing-plane treatment allows the steady simulation of multiple rows, taking their interaction between one another into account and thus providing more realistic results. A stator/rotor turbine stage of a commercial jet engine is analyzed and some representative sensitivity results are presented and discussed.
Keywords: Automatic differentiation, Derivatives, Gradient-based optimization, Shape optimization, Operating conditions},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
The currently available computational power and improvements of high-fidelity numerical simulations have lead to an increased use of computational fluid dynamics (CFD) in the analysis of turbomachinery flows, particularly in design environments. The optimization cases often contain up to thousands of design variables and gradient-based (GB) optimization algorithms are typically selected due to their efficiency. The adjoint method is key to efficiently compute the derivatives required by the GB algorithms, with a computational cost nearly independent of the number of design variables. In this paper we present the details of the development of an adjoint multirow interface based on the mixing-plane treatment to extend an already existing adjoint solver using the ADjoint approach. The mixing-plane treatment allows the steady simulation of multiple rows, taking their interaction between one another into account and thus providing more realistic results. A stator/rotor turbine stage of a commercial jet engine is analyzed and some representative sensitivity results are presented and discussed.
Keywords: Automatic differentiation, Derivatives, Gradient-based optimization, Shape optimization, Operating conditions
Keywords: Automatic differentiation, Derivatives, Gradient-based optimization, Shape optimization, Operating conditions
Moita, N.; Marta, A. C.
Optimization of the propeller-driven propulsion system for a small UAV Proceedings Article
In: Proceedings of the EngOpt 2018 - 6th International Conference on Engineering Optimization, pp. 1372-1384, IDMEC, Instituto Superior Técnico Lisboa, Portugal, 2018.
@inproceedings{Moita:2018:ENGOPT,
title = {Optimization of the propeller-driven propulsion system for a small UAV},
author = {N. Moita and A. C. Marta},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Moita_2018_ENGOPT.pdf, Full-text PDF
https://engopt2018.tecnico.ulisboa.pt/
https://link.springer.com/chapter/10.1007/978-3-319-97773-7_118},
doi = {10.1007/978-3-319-97773-7_118},
year = {2018},
date = {2018-09-01},
booktitle = {Proceedings of the EngOpt 2018 - 6th International Conference on Engineering Optimization},
pages = {1372-1384},
address = {Lisboa, Portugal},
organization = {IDMEC, Instituto Superior Técnico},
abstract = {Integrated in the LEEUAV project, the objective of this work was to optimize the propeller-driven propulsion system previously implemented. A propeller was parametrized in terms of planform and airfoil shape and the software QPROP used to evaluate the performance of in terms of thrust, power and thrust coefficient and propeller efficiency. Experimental tests were conducted for three different propellers to study study the performance sensitivity to propeller diameter and pitch, electric motors, and also to validate the numerical model. Following those tests, a multi-objective shape optimization using MATLAB, for cruise and climb conditions, was performed. At the end of this optimization, a system motor+propeller with an higher efficiency was obtained.
Keywords: Propeller, Optimization, Efficiency, Thrust, Electrical power},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Integrated in the LEEUAV project, the objective of this work was to optimize the propeller-driven propulsion system previously implemented. A propeller was parametrized in terms of planform and airfoil shape and the software QPROP used to evaluate the performance of in terms of thrust, power and thrust coefficient and propeller efficiency. Experimental tests were conducted for three different propellers to study study the performance sensitivity to propeller diameter and pitch, electric motors, and also to validate the numerical model. Following those tests, a multi-objective shape optimization using MATLAB, for cruise and climb conditions, was performed. At the end of this optimization, a system motor+propeller with an higher efficiency was obtained.
Keywords: Propeller, Optimization, Efficiency, Thrust, Electrical power
Keywords: Propeller, Optimization, Efficiency, Thrust, Electrical power
Rodrigues, P.; Marta, A. C.
Efficient aerodynamic optimization of aircraft wings Proceedings Article
In: Proceedings of the EngOpt 2018 - 6th International Conference on Engineering Optimization, pp. 897-910, IDMEC, Instituto Superior Técnico Lisboa, Portugal, 2018.
@inproceedings{RodriguesP:2018:ENGOPT,
title = {Efficient aerodynamic optimization of aircraft wings},
author = {P. Rodrigues and A. C. Marta},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_RodriguesP_2018_ENGOPT.pdf, Full-text PDF
https://engopt2018.tecnico.ulisboa.pt/
https://link.springer.com/chapter/10.1007/978-3-319-97773-7_78},
doi = {10.1007/978-3-319-97773-7_78},
year = {2018},
date = {2018-09-01},
booktitle = {Proceedings of the EngOpt 2018 - 6th International Conference on Engineering Optimization},
pages = {897-910},
address = {Lisboa, Portugal},
organization = {IDMEC, Instituto Superior Técnico},
abstract = {Multidisciplinary design and optimization is a promising methodology for the efficient design of complex systems, in particular when it combines coupled analyses with gradient-based optimization techniques. In this case, it requires the derivatives evaluation of the functions of interest with respect to the design variables, which is the most demanding computational task in the process, so the goal of this work is to develop an efficient optimization framework to solve aerodynamic design problems using exact gradient information. To this end, the aerodynamic model based on the panel method is reformulated into five smaller modules, in which the respective sensitivity analysis blocks are constructed using exact gradient estimation methods: automatic differentiation, symbolic differentiation and the adjoint method. After the aerodynamic and corresponding sensitivity analysis tools are verified numerically, aerodynamic optimization problems are solved using the new tool with remarkable success since, when compared to the finite-differences method, the optimization time can be reduced by 90%.
Keywords: Gradient-based optimization, Sensitivity analysis, Automatic differentiation, Adjoint method},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Multidisciplinary design and optimization is a promising methodology for the efficient design of complex systems, in particular when it combines coupled analyses with gradient-based optimization techniques. In this case, it requires the derivatives evaluation of the functions of interest with respect to the design variables, which is the most demanding computational task in the process, so the goal of this work is to develop an efficient optimization framework to solve aerodynamic design problems using exact gradient information. To this end, the aerodynamic model based on the panel method is reformulated into five smaller modules, in which the respective sensitivity analysis blocks are constructed using exact gradient estimation methods: automatic differentiation, symbolic differentiation and the adjoint method. After the aerodynamic and corresponding sensitivity analysis tools are verified numerically, aerodynamic optimization problems are solved using the new tool with remarkable success since, when compared to the finite-differences method, the optimization time can be reduced by 90%.
Keywords: Gradient-based optimization, Sensitivity analysis, Automatic differentiation, Adjoint method
Keywords: Gradient-based optimization, Sensitivity analysis, Automatic differentiation, Adjoint method
Alves, J.; Moutinho, A.; Marta, A. C.; Gamboa, P.
Path planning and collision avoidance algorithms for small RPA Proceedings Article
In: Proceedings of the ICEUBI 2017 - International Congress on Engineering, pp. 665-675, Universidade da Beira Interior Covilhã, Portugal, 2017, ISBN: 978-989-654-403-4.
@inproceedings{Alves:2017:ICEUBI,
title = {Path planning and collision avoidance algorithms for small RPA},
author = {J. Alves and A. Moutinho and A. C. Marta and P. Gamboa},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Alves_2017_ICEUBI.pdf, Full-text PDF http://iceubi.ubi.pt/?page_id=105#1549554660760-112f8940-549c},
isbn = {978-989-654-403-4},
year = {2017},
date = {2017-12-01},
booktitle = {Proceedings of the ICEUBI 2017 - International Congress on Engineering},
pages = {665-675},
address = {Covilhã, Portugal},
organization = {Universidade da Beira Interior},
abstract = {The development of Remotely Piloted Aircrafts (RPAS) for civil applications has been rapidly growing over the past years. This work presents a solution for the generation of optimal trajectories for RPAS subject to manoeuvrability and collision avoidance constraints. To achieve this task a two-layered approach is proposed. In the first stage, classical path planning techniques are implemented to generate safe and flyable paths in a known static environment. The A* algorithm and Ant Colony Optimization (ACO) are used to find an optimal sequence of waypoints in a discrete environment. To ensure that the path is flyable and complies with curvature constraints, an optimization of Rational Bezier curves is implemented. The second stage is developed for real-time implementation and potential fields methods are used to replan the initial path when new obstacles are detected. For the global path planning stage the best results were found to be provided by using ACO to optimize waypoint order, A* to connect the waypoints and rational Bezier curves with constraint restriction. The Potential Fields method is computationally inexpensive proving to be a feasible solution for real-time implementation. It is shown that the algorithms perform reasonably well in several scenarios.
Keywords: RPAS, Collision avoidance, Path planning, A* algorithm, ACO algorithm},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
The development of Remotely Piloted Aircrafts (RPAS) for civil applications has been rapidly growing over the past years. This work presents a solution for the generation of optimal trajectories for RPAS subject to manoeuvrability and collision avoidance constraints. To achieve this task a two-layered approach is proposed. In the first stage, classical path planning techniques are implemented to generate safe and flyable paths in a known static environment. The A* algorithm and Ant Colony Optimization (ACO) are used to find an optimal sequence of waypoints in a discrete environment. To ensure that the path is flyable and complies with curvature constraints, an optimization of Rational Bezier curves is implemented. The second stage is developed for real-time implementation and potential fields methods are used to replan the initial path when new obstacles are detected. For the global path planning stage the best results were found to be provided by using ACO to optimize waypoint order, A* to connect the waypoints and rational Bezier curves with constraint restriction. The Potential Fields method is computationally inexpensive proving to be a feasible solution for real-time implementation. It is shown that the algorithms perform reasonably well in several scenarios.
Keywords: RPAS, Collision avoidance, Path planning, A* algorithm, ACO algorithm
Keywords: RPAS, Collision avoidance, Path planning, A* algorithm, ACO algorithm
Baião, T.; Marta, A. C.; Moutinho, A.; Gamboa, P.
Energy monitoring system for low-cost UAVs Proceedings Article
In: Proceedings of the ICEUBI 2017 - International Congress on Engineering, pp. 1214-1223, Universidade da Beira Interior Covilhã, Portugal, 2017, ISBN: 978-989-654-403-4.
@inproceedings{Baiao:2017:ICEUBI,
title = {Energy monitoring system for low-cost UAVs},
author = {T. Baião and A. C. Marta and A. Moutinho and P. Gamboa},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Baiao_2017_ICEUBI.pdf, Full-text PDF http://iceubi.ubi.pt/?page_id=105#1549554660760-112f8940-549c},
isbn = {978-989-654-403-4},
year = {2017},
date = {2017-12-01},
booktitle = {Proceedings of the ICEUBI 2017 - International Congress on Engineering},
pages = {1214-1223},
address = {Covilhã, Portugal},
organization = {Universidade da Beira Interior},
abstract = {In the present, unmanned aerial vehicles, particularly low-cost models, lack intrinsic safety systems despite increasing interest by the civilian public for these platforms, posing a threat to other aircraft, people and property. Integrated in a larger project that addresses safety issues for this type of aircraft, this work aims to contribute to the enhancement of their safety features by proposing an energy monitoring system capable of providing updated estimates of the final state of energy of the onboard sources, enabling the operator to understand if the planned mission can be completed safely, given its energetic requirements and taking into account environmental conditions such as wind and solar radiation. The remaining energy estimate enables better energy awareness during mission planning and the online updates allow to account for unexpected disturbances and obstacle avoidance. The proposed energy monitoring system is qualitatively validated and three methods not previously considered in the literature are proposed to estimate the required energy to complete a given planned mission, and their performance is evaluated using simulation software. It is concluded that the methods discussed are very sensitive to the quality of the data and simulation tools available, and those available would be inadequate for simulating a real scenario. Nonetheless, solid foundations for future work are established.
Keywords: UAV safety, Mission feasibility, Energy requirements, Estimation models, Estimation models},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
In the present, unmanned aerial vehicles, particularly low-cost models, lack intrinsic safety systems despite increasing interest by the civilian public for these platforms, posing a threat to other aircraft, people and property. Integrated in a larger project that addresses safety issues for this type of aircraft, this work aims to contribute to the enhancement of their safety features by proposing an energy monitoring system capable of providing updated estimates of the final state of energy of the onboard sources, enabling the operator to understand if the planned mission can be completed safely, given its energetic requirements and taking into account environmental conditions such as wind and solar radiation. The remaining energy estimate enables better energy awareness during mission planning and the online updates allow to account for unexpected disturbances and obstacle avoidance. The proposed energy monitoring system is qualitatively validated and three methods not previously considered in the literature are proposed to estimate the required energy to complete a given planned mission, and their performance is evaluated using simulation software. It is concluded that the methods discussed are very sensitive to the quality of the data and simulation tools available, and those available would be inadequate for simulating a real scenario. Nonetheless, solid foundations for future work are established.
Keywords: UAV safety, Mission feasibility, Energy requirements, Estimation models, Estimation models
Keywords: UAV safety, Mission feasibility, Energy requirements, Estimation models, Estimation models
Rodrigues, S. S.; Marta, A. C.
Numerical verification of adjoint-based sensitivity analysis for multistage turbomachinery design Proceedings Article
In: Proceedings of the 4th LAETA Young Researchers Meeting, LAETA/AeroG/UBI Covilhã, Portugal, 2017.
@inproceedings{Rodrigues:2017:LAETA,
title = {Numerical verification of adjoint-based sensitivity analysis for multistage turbomachinery design},
author = {S. S. Rodrigues and A. C. Marta},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Rodrigues_2017_LAETA.pdf, Full-text PDF
https://ubibliorum.ubi.pt/handle/10400.6/12215},
year = {2017},
date = {2017-10-01},
urldate = {2017-10-01},
booktitle = {Proceedings of the 4th LAETA Young Researchers Meeting},
address = {Covilhã, Portugal},
organization = {LAETA/AeroG/UBI},
abstract = {With the exponential growth in computational power as well as improvements in the accuracy of computational fluid dynamics (CFD) tools, their use in turbomachinery design and analysis has seen a great increase, particularly in optimization environments, where gradient-based optimization algorithms are often selected for their efficiency. These algorithms require the computation of the sensitivities of the functions of interest to the design variables. The number of design variables in an optimization problem may be in the order of thousands. As such, the use of the adjoint approach for calculating the gradients is highly advantageous as it produces function sensitivities with computational cost that is nearly independent of the number of design variables. In the analysis of turbomachinery, accounting for the interaction between the multiple blade passages is of paramount importance if one wishes to increase the accuracy of the simulation. Many computational methods exist to address this interactions. The mixing-plane treatment is one of the most widely used methods in the steady analysis of multiple rows of a turbomachine. This paper describes improvements to a discrete adjoint solver of a proprietary CFD solver for multistage turbomachinery applications, namely the adjoint counterpart of the mixing-plane formulation of the direct solver. The adjoint solver is developed using the ADjoint approach, where the partial derivatives required for the assembly of the adjoint system of equations are obtained using automatic differentiation tools. A verification of the implementation of the mixing-plane against the finite-difference approximations is presented. Sensitivities of selected surface functions of interest, such as mass flow, to other selected design parameters, such as surface nodes or inflow boundary conditions, calculated with both methods are presented. The results show good agreement of both derivatives and emphasise the benefits of the adjoint approach versus finite-differences in terms of accuracy and computational cost.
Keywords: Turbomachinery, Discrete adjoint, Multistage, Gradient-based optimization, Parallel processing, Finite-differences},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
With the exponential growth in computational power as well as improvements in the accuracy of computational fluid dynamics (CFD) tools, their use in turbomachinery design and analysis has seen a great increase, particularly in optimization environments, where gradient-based optimization algorithms are often selected for their efficiency. These algorithms require the computation of the sensitivities of the functions of interest to the design variables. The number of design variables in an optimization problem may be in the order of thousands. As such, the use of the adjoint approach for calculating the gradients is highly advantageous as it produces function sensitivities with computational cost that is nearly independent of the number of design variables. In the analysis of turbomachinery, accounting for the interaction between the multiple blade passages is of paramount importance if one wishes to increase the accuracy of the simulation. Many computational methods exist to address this interactions. The mixing-plane treatment is one of the most widely used methods in the steady analysis of multiple rows of a turbomachine. This paper describes improvements to a discrete adjoint solver of a proprietary CFD solver for multistage turbomachinery applications, namely the adjoint counterpart of the mixing-plane formulation of the direct solver. The adjoint solver is developed using the ADjoint approach, where the partial derivatives required for the assembly of the adjoint system of equations are obtained using automatic differentiation tools. A verification of the implementation of the mixing-plane against the finite-difference approximations is presented. Sensitivities of selected surface functions of interest, such as mass flow, to other selected design parameters, such as surface nodes or inflow boundary conditions, calculated with both methods are presented. The results show good agreement of both derivatives and emphasise the benefits of the adjoint approach versus finite-differences in terms of accuracy and computational cost.
Keywords: Turbomachinery, Discrete adjoint, Multistage, Gradient-based optimization, Parallel processing, Finite-differences
Keywords: Turbomachinery, Discrete adjoint, Multistage, Gradient-based optimization, Parallel processing, Finite-differences
Marta, A. C.; Moutinho, A.; Gamboa, P. V.
Operational safety systems for small drones Proceedings Article
In: Proceedings of the 4th LAETA Young Researchers Meeting, LAETA/AeroG/UBI Covilhã, Portugal, 2017.
@inproceedings{Marta:2017:LAETA,
title = {Operational safety systems for small drones},
author = {A. C. Marta and A. Moutinho and P. V. Gamboa},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Marta_2017_LAETA.pdf, Full-text PDF
https://ubibliorum.ubi.pt/handle/10400.6/12215},
year = {2017},
date = {2017-10-01},
urldate = {2017-10-01},
booktitle = {Proceedings of the 4th LAETA Young Researchers Meeting},
address = {Covilhã, Portugal},
organization = {LAETA/AeroG/UBI},
abstract = {Recent incidents involving small drones have been raising concerns about their safe operation. It is foreseen that embedded safety systems will become mandatory in the near future, when stricter operational regulations will be put into place. This work is part of the on-going research aimed to address the current lack of intrinsic safety systems by developing three distinct but highly coupled subsystems: flight energy management (FEM), obstacle detection (OD) and mission planning (MP). The FEM is able to estimate the energy balance of the updated mission plan by making an assessment of the available energy, both stored and to be harvested in-flight (e.g., photovoltaic), and comparing it to the required energy, taking into account the aircraft performance and in-route weather conditions (e.g., wind and solar radiation). This energy balance is continuously updated in real-time. The OD addresses all aspects of obstacle detection, namely the identification of the required aircraft instrumentation and the respective measurements processing to assess if there is probability of collision. Preliminary work using a LIDAR shows this sensor is valid for detecting close range obstacles. Further work will include its integration with other visual information in order to increase the detection range. Lastly, the MP allows for both pre-flight mission planning as well as in-flight mission replanning, taking into account the data received from the FEM to attest the mission feasibility, and from the OD regarding the detection of new obstacles. The mission planning depends on a set of desired waypoints, a list of known obstacles (static, such as terrain and buildings, or dynamic, such as other aircraft), the vehicle performance capabilities and the rules of air, resulting in a well-defined reference path. This path can currently be optimized for different metrics, such as time (minimum for fast execution or maximum for extended endurance), distance (maximum for extended range) or energy (minimum for extended range). These three subsystems are meant to be integrated into the drone flight-controller to provide valuable data to the operator and, as a last resort, to automatically take action to avoid collisions or abort energy limited missions. A solar powered unmanned fixed wing aircraft, previously designed and built, will serve as one of the flight platforms to assess the operational capabilities of the developed safety subsystems. This vehicle has been flight tested to characterize its power requirements across the flight envelope thus providing key data used in the implemented algorithms’ simulations.
Keywords: Unmanned aerial vehicle, Flight energy management, Mission planning, Obstacle detection, Obstacle avoidance},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Recent incidents involving small drones have been raising concerns about their safe operation. It is foreseen that embedded safety systems will become mandatory in the near future, when stricter operational regulations will be put into place. This work is part of the on-going research aimed to address the current lack of intrinsic safety systems by developing three distinct but highly coupled subsystems: flight energy management (FEM), obstacle detection (OD) and mission planning (MP). The FEM is able to estimate the energy balance of the updated mission plan by making an assessment of the available energy, both stored and to be harvested in-flight (e.g., photovoltaic), and comparing it to the required energy, taking into account the aircraft performance and in-route weather conditions (e.g., wind and solar radiation). This energy balance is continuously updated in real-time. The OD addresses all aspects of obstacle detection, namely the identification of the required aircraft instrumentation and the respective measurements processing to assess if there is probability of collision. Preliminary work using a LIDAR shows this sensor is valid for detecting close range obstacles. Further work will include its integration with other visual information in order to increase the detection range. Lastly, the MP allows for both pre-flight mission planning as well as in-flight mission replanning, taking into account the data received from the FEM to attest the mission feasibility, and from the OD regarding the detection of new obstacles. The mission planning depends on a set of desired waypoints, a list of known obstacles (static, such as terrain and buildings, or dynamic, such as other aircraft), the vehicle performance capabilities and the rules of air, resulting in a well-defined reference path. This path can currently be optimized for different metrics, such as time (minimum for fast execution or maximum for extended endurance), distance (maximum for extended range) or energy (minimum for extended range). These three subsystems are meant to be integrated into the drone flight-controller to provide valuable data to the operator and, as a last resort, to automatically take action to avoid collisions or abort energy limited missions. A solar powered unmanned fixed wing aircraft, previously designed and built, will serve as one of the flight platforms to assess the operational capabilities of the developed safety subsystems. This vehicle has been flight tested to characterize its power requirements across the flight envelope thus providing key data used in the implemented algorithms’ simulations.
Keywords: Unmanned aerial vehicle, Flight energy management, Mission planning, Obstacle detection, Obstacle avoidance
Keywords: Unmanned aerial vehicle, Flight energy management, Mission planning, Obstacle detection, Obstacle avoidance
Diogo, R.; Bras, M.; Marta, A. C.; Suleman, A.
Topology optimization of a high aspect ratio wing box using an ant colony optimization algorithm Proceedings Article
In: Proceedings of the CASI AERO 2017 - 63rd Aeronautics Conference of the Canadian Aeronautics and Space Institute, Canadian Aeronautics and Space Institute Toronto, Canada, 2017.
@inproceedings{Diogo:2017:CASI,
title = {Topology optimization of a high aspect ratio wing box using an ant colony optimization algorithm},
author = {R. Diogo and M. Bras and A. C. Marta and A. Suleman},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Diogo_2017_CASI.pdf, Full-text PDF https://theoac.ca/events/EventDetails.aspx?id=935861},
year = {2017},
date = {2017-05-01},
booktitle = {Proceedings of the CASI AERO 2017 - 63rd Aeronautics Conference of the Canadian Aeronautics and Space Institute},
address = {Toronto, Canada},
organization = {Canadian Aeronautics and Space Institute},
abstract = {Aircraft transportation is seeing a demand for leaner and greener technologies. The challenge for the next generation aircraft is to improve fuel economy by increasing the lift to drag ratio, which translates into an increase of the aspect ratio (AR) of the wing. To build high aspect ratio wings, lighter structures must be carefully designed and new materials and construction techniques must be developed to obtain the required wing box strength. The increase in weight due to the longer spans reduces the advantages brought by the introduction of a high AR wing, thus making the design of the wing box a fundamental aspect of the process.
In this work, a topology optimization of the wing box of a reference high aspect ratio cantilever wing is carried out. The topology optimization procedure is based on an Ant Colony Optimization (ACO) algorithm. The ACO is a metaheuristic biologically influenced algorithm that has been proven to be useful to solve NP-hard combinatorial optimization problems in an expedite way. It allows for an easy parallelization of the solution process and its application to topology optimization problems has been recently introduced. An example is initially studied that compares the ACO algorithm with a gradient based method. An external finite element solver is then coupled to study the wing box of the high AR wing. An external aerodynamic load case is computed for a trimmed cruise condition from a CFD analysis and the results of the optimization are discussed.
Keywords: Wing design, Light structures, Ant colony optimization, Finite element analysis, Aerodynamic loads},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Aircraft transportation is seeing a demand for leaner and greener technologies. The challenge for the next generation aircraft is to improve fuel economy by increasing the lift to drag ratio, which translates into an increase of the aspect ratio (AR) of the wing. To build high aspect ratio wings, lighter structures must be carefully designed and new materials and construction techniques must be developed to obtain the required wing box strength. The increase in weight due to the longer spans reduces the advantages brought by the introduction of a high AR wing, thus making the design of the wing box a fundamental aspect of the process.
In this work, a topology optimization of the wing box of a reference high aspect ratio cantilever wing is carried out. The topology optimization procedure is based on an Ant Colony Optimization (ACO) algorithm. The ACO is a metaheuristic biologically influenced algorithm that has been proven to be useful to solve NP-hard combinatorial optimization problems in an expedite way. It allows for an easy parallelization of the solution process and its application to topology optimization problems has been recently introduced. An example is initially studied that compares the ACO algorithm with a gradient based method. An external finite element solver is then coupled to study the wing box of the high AR wing. An external aerodynamic load case is computed for a trimmed cruise condition from a CFD analysis and the results of the optimization are discussed.
Keywords: Wing design, Light structures, Ant colony optimization, Finite element analysis, Aerodynamic loads
In this work, a topology optimization of the wing box of a reference high aspect ratio cantilever wing is carried out. The topology optimization procedure is based on an Ant Colony Optimization (ACO) algorithm. The ACO is a metaheuristic biologically influenced algorithm that has been proven to be useful to solve NP-hard combinatorial optimization problems in an expedite way. It allows for an easy parallelization of the solution process and its application to topology optimization problems has been recently introduced. An example is initially studied that compares the ACO algorithm with a gradient based method. An external finite element solver is then coupled to study the wing box of the high AR wing. An external aerodynamic load case is computed for a trimmed cruise condition from a CFD analysis and the results of the optimization are discussed.
Keywords: Wing design, Light structures, Ant colony optimization, Finite element analysis, Aerodynamic loads
Bras, M.; Suleman, A.; Diogo, R.; Marta, A. C.
Ant colony optimization method applied to topology optimization of aircraft structures Proceedings Article
In: Proceedings of the CANCAM 2017 - 26th Canadian Congress of Applied Mechanics, pp. 24-27, University of Victoria Victoria, Canada, 2017, ISBN: 9781510856783.
@inproceedings{Bras:2017:CANCAM,
title = {Ant colony optimization method applied to topology optimization of aircraft structures},
author = {M. Bras and A. Suleman and R. Diogo and A. C. Marta},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Bras_2017_CANCAM.pdf, Full-text PDF
https://cancam2017.me.uvic.ca},
isbn = {9781510856783},
year = {2017},
date = {2017-05-01},
booktitle = {Proceedings of the CANCAM 2017 - 26th Canadian Congress of Applied Mechanics},
pages = {24-27},
address = {Victoria, Canada},
organization = {University of Victoria},
abstract = {A structural Topology Optimization (TO) of the cross-section of a wing-box is carried out using an Ant Colony Optimization (ACO) method. ACO is a meta-heuristic biologically influenced algorithm that has been proven to be useful to solve NP-hard combinatorial optimization problems in an expedite way. Its application to solve topology optimization has been introduced recently. The algorithm is first used with a cantilever beam example and the results compared with the literature. The algorithm is then coupled with an external finite element solver to perform a topology optimization of the cross-section of a wing box. The external aerodynamic loads are computed from a CFD analysis for a specific flight condition and applied to the structural model. The advantages of using ACO to discrete TO problems are demonstrated and the results of the optimization are discussed.
Keywords: Heuristic methods, Ant colony optimization, Topology optimization, Aircraft structures},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
A structural Topology Optimization (TO) of the cross-section of a wing-box is carried out using an Ant Colony Optimization (ACO) method. ACO is a meta-heuristic biologically influenced algorithm that has been proven to be useful to solve NP-hard combinatorial optimization problems in an expedite way. Its application to solve topology optimization has been introduced recently. The algorithm is first used with a cantilever beam example and the results compared with the literature. The algorithm is then coupled with an external finite element solver to perform a topology optimization of the cross-section of a wing box. The external aerodynamic loads are computed from a CFD analysis for a specific flight condition and applied to the structural model. The advantages of using ACO to discrete TO problems are demonstrated and the results of the optimization are discussed.
Keywords: Heuristic methods, Ant colony optimization, Topology optimization, Aircraft structures
Keywords: Heuristic methods, Ant colony optimization, Topology optimization, Aircraft structures
Campos, L. M. B. C.; Marta, A. C.
On acoustic-vortical-entropy waves in propulsion systems Proceedings Article
In: Proceedings of the ICAS 2016 - 30th Congress of the International Council of the Aeronautical Sciences, International Council of the Aeronautical Sciences Daejeon, South Korea, 2016, ISBN: 978-3-932182-85-3.
@inproceedings{Campos:2016:ICAS,
title = {On acoustic-vortical-entropy waves in propulsion systems},
author = {L. M. B. C. Campos and A. C. Marta},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Campos_2016_ICAS.pdf, Full-text PDF
https://www.icas.org/ICAS_ARCHIVE/ICAS2016/data/index.html},
isbn = {978-3-932182-85-3},
year = {2016},
date = {2016-09-01},
booktitle = {Proceedings of the ICAS 2016 - 30th Congress of the International Council of the Aeronautical Sciences},
address = {Daejeon, South Korea},
organization = {International Council of the Aeronautical Sciences},
series = {ICAS2016-P4.1},
abstract = {The noise of jet and rocket engines involves the coupling of sound to swirling flows and to heat exchanges leading in the more complex cases of triple interactions to acoustic-vortical-entropy (AVE) waves. The present paper presents as far as the authors are aware the first derivation of the AVE equation for axisymmetric linear non-dissipative perturbations of a compressible, non-isentropic, swirling mean flow, with constant axial velocity and constant angular velocity. The axisymmetric AVE wave equation is obtained for the radial velocity perturbation, specifying its radial dependence for a given frequency and axial wavenumber. The AVE wave equation in the case of zero axial wavenumber has only one singularity at the sonic radius, where the isothermal Mach number for the swirl velocity is unity. The exact solution of the AVE wave equation is obtained as series expansions of Gaussian hypergeometric type valid inside, outside and around the sonic radius, thus: (i) covering the whole flow region; (ii) identifying the singularity at the sonic condition at the sonic radius; (iii) specifying near-axis and asymptotic solutions for small and large radius. Using polarization relations among
wave variables specifies exactly the perturbations of: (i,ii) the radial and azimuthal
velocity; (iii,iv) pressure and mass density; (v,vi) entropy and temperature. It is shown that
the dependence of the AVE wave variables on the radial distance can be: (a) oscillatory with
decaying amplitude; (b) monotonic with increasing amplitude. The case (b) of AVE wave amplitude increasing monotonically with the radial distance applies if the frequency times a function of the adiabatic exponent is less than the modulus of the vorticity (or twice the angular velocity). In the opposite case (a) the oscillatory nature of acoustic waves predominates over the tendency for monotonic growth of vortical perturbations. Associating sound with stable potential flows and swirl with unstable vortical flows suggests a criterion valid in non-isentropic conditions, that is in the presence of heat exchanges, that is a condition for stable combustion in a confined space: the peak vorticity (multiplied by a factor of order unity dependent on the adiabatic exponent) should be less than the lowest or fundamental frequency of the cavity.
Keywords: Aeroacoustics, Absolute/convective instability, Critical layers},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
The noise of jet and rocket engines involves the coupling of sound to swirling flows and to heat exchanges leading in the more complex cases of triple interactions to acoustic-vortical-entropy (AVE) waves. The present paper presents as far as the authors are aware the first derivation of the AVE equation for axisymmetric linear non-dissipative perturbations of a compressible, non-isentropic, swirling mean flow, with constant axial velocity and constant angular velocity. The axisymmetric AVE wave equation is obtained for the radial velocity perturbation, specifying its radial dependence for a given frequency and axial wavenumber. The AVE wave equation in the case of zero axial wavenumber has only one singularity at the sonic radius, where the isothermal Mach number for the swirl velocity is unity. The exact solution of the AVE wave equation is obtained as series expansions of Gaussian hypergeometric type valid inside, outside and around the sonic radius, thus: (i) covering the whole flow region; (ii) identifying the singularity at the sonic condition at the sonic radius; (iii) specifying near-axis and asymptotic solutions for small and large radius. Using polarization relations among
wave variables specifies exactly the perturbations of: (i,ii) the radial and azimuthal
velocity; (iii,iv) pressure and mass density; (v,vi) entropy and temperature. It is shown that
the dependence of the AVE wave variables on the radial distance can be: (a) oscillatory with
decaying amplitude; (b) monotonic with increasing amplitude. The case (b) of AVE wave amplitude increasing monotonically with the radial distance applies if the frequency times a function of the adiabatic exponent is less than the modulus of the vorticity (or twice the angular velocity). In the opposite case (a) the oscillatory nature of acoustic waves predominates over the tendency for monotonic growth of vortical perturbations. Associating sound with stable potential flows and swirl with unstable vortical flows suggests a criterion valid in non-isentropic conditions, that is in the presence of heat exchanges, that is a condition for stable combustion in a confined space: the peak vorticity (multiplied by a factor of order unity dependent on the adiabatic exponent) should be less than the lowest or fundamental frequency of the cavity.
Keywords: Aeroacoustics, Absolute/convective instability, Critical layers
wave variables specifies exactly the perturbations of: (i,ii) the radial and azimuthal
velocity; (iii,iv) pressure and mass density; (v,vi) entropy and temperature. It is shown that
the dependence of the AVE wave variables on the radial distance can be: (a) oscillatory with
decaying amplitude; (b) monotonic with increasing amplitude. The case (b) of AVE wave amplitude increasing monotonically with the radial distance applies if the frequency times a function of the adiabatic exponent is less than the modulus of the vorticity (or twice the angular velocity). In the opposite case (a) the oscillatory nature of acoustic waves predominates over the tendency for monotonic growth of vortical perturbations. Associating sound with stable potential flows and swirl with unstable vortical flows suggests a criterion valid in non-isentropic conditions, that is in the presence of heat exchanges, that is a condition for stable combustion in a confined space: the peak vorticity (multiplied by a factor of order unity dependent on the adiabatic exponent) should be less than the lowest or fundamental frequency of the cavity.
Keywords: Aeroacoustics, Absolute/convective instability, Critical layers
Marta, A. C.; Moutinho, A.; Gamboa, P. V.
Drones Safe Flight: on addressing operational safety for low cost small drones Proceedings Article
In: Proceedings of the CEM 2016 - Mechanical Engineering Conference, LAETA/IDMEC/FEUP Porto, Portugal, 2016.
@inproceedings{Marta:2016:CEM,
title = {Drones Safe Flight: on addressing operational safety for low cost small drones},
author = {A. C. Marta and A. Moutinho and P. V. Gamboa},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Marta_2016_CEM.pdf, Full-text PDF},
year = {2016},
date = {2016-06-01},
booktitle = {Proceedings of the CEM 2016 - Mechanical Engineering Conference},
address = {Porto, Portugal},
organization = {LAETA/IDMEC/FEUP},
abstract = {The recent exponential growth of small drones has been raising concerns about the safety of their operation. It is foreseen that embedded safety systems will become mandatory in the near future, when stricter operational regulations will be put in place. As such, the development of a low cost and low weight solution is paramount. This work is part of the on-going research aimed to address the current lack of intrinsic safety systems by developing three distinct but highly coupled subsystems: the flight energy management (FEM), the mission planning (MP) and the obstacle detection and avoidance (ODA). Although aimed primarily to fixed wing drones, its conceptual design is general to any type of drones. With the FEM, it is possible to estimate the energy balance of the current mission plan by making an assessment of the available energy, both stored and to be harvested in-flight (e.g., photovoltaic), and comparing it to the required energy, taking into account the aircraft performance and in-route weather conditions (e.g., wind and solar radiation). First principles of physics are used to obtain a baseline for the energy models without compromising future higher-fidelity models. The estimates can be done both pre-flight and continuously updated in-flight. This results in a real-time energy balance update that feeds the MP subsystem. The MP allows for both pre-flight mission planning as well as adaptive in-flight mission planning, taking into account the data received from the FEM to attest its feasibility. It is based on a set of waypoints and a list of known fixed obstacles, such as terrain and buildings. The mission can be optimized for different metrics, such as time (minimum for fast execution or maximum for extended endurance), distance (maximum for extended range). This subsystem provides data to the drone operator that can be used to redefine the mission given the energy balance update. The MP feeds the FEM with the planned mission for the update of the required energy. Lastly, the ODA comprises the hardware for sensing obstacles and the software for planning an evasive maneuver for collision avoidance. The detection of obstacles leads to an update of the list of known obstacles. Any maneuver triggered by the ODA is fed to the MP as it affects both the current mission and, consequently, the energy balance estimated by the FEM. The goal is to develop a low cost safety system that can ensure the drone can either execute the mission successfully or it can prematurely return to the base safely.
Keywords: Unmanned aerial vehicle, Flight energy management, Mission planning, Obstacle detection and avoidance},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
The recent exponential growth of small drones has been raising concerns about the safety of their operation. It is foreseen that embedded safety systems will become mandatory in the near future, when stricter operational regulations will be put in place. As such, the development of a low cost and low weight solution is paramount. This work is part of the on-going research aimed to address the current lack of intrinsic safety systems by developing three distinct but highly coupled subsystems: the flight energy management (FEM), the mission planning (MP) and the obstacle detection and avoidance (ODA). Although aimed primarily to fixed wing drones, its conceptual design is general to any type of drones. With the FEM, it is possible to estimate the energy balance of the current mission plan by making an assessment of the available energy, both stored and to be harvested in-flight (e.g., photovoltaic), and comparing it to the required energy, taking into account the aircraft performance and in-route weather conditions (e.g., wind and solar radiation). First principles of physics are used to obtain a baseline for the energy models without compromising future higher-fidelity models. The estimates can be done both pre-flight and continuously updated in-flight. This results in a real-time energy balance update that feeds the MP subsystem. The MP allows for both pre-flight mission planning as well as adaptive in-flight mission planning, taking into account the data received from the FEM to attest its feasibility. It is based on a set of waypoints and a list of known fixed obstacles, such as terrain and buildings. The mission can be optimized for different metrics, such as time (minimum for fast execution or maximum for extended endurance), distance (maximum for extended range). This subsystem provides data to the drone operator that can be used to redefine the mission given the energy balance update. The MP feeds the FEM with the planned mission for the update of the required energy. Lastly, the ODA comprises the hardware for sensing obstacles and the software for planning an evasive maneuver for collision avoidance. The detection of obstacles leads to an update of the list of known obstacles. Any maneuver triggered by the ODA is fed to the MP as it affects both the current mission and, consequently, the energy balance estimated by the FEM. The goal is to develop a low cost safety system that can ensure the drone can either execute the mission successfully or it can prematurely return to the base safely.
Keywords: Unmanned aerial vehicle, Flight energy management, Mission planning, Obstacle detection and avoidance
Keywords: Unmanned aerial vehicle, Flight energy management, Mission planning, Obstacle detection and avoidance
Carrolo, E.; Marta, A. C.
Gust wind load reduction in wind turbine blades Proceedings Article
In: Proceedings of the 10th International Conference on Composite Science and Technology, IDMEC, Instituto Superior Técnico Lisbon, Portugal, 2015.
@inproceedings{Carrolo:2015:ICCST,
title = {Gust wind load reduction in wind turbine blades},
author = {E. Carrolo and A. C. Marta},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Carrolo_2015_ICCST.pdf, Full-text PDF
http://www.dem.ist.utl.pt/iccst10/},
year = {2015},
date = {2015-09-01},
booktitle = {Proceedings of the 10th International Conference on Composite Science and Technology},
address = {Lisbon, Portugal},
organization = {IDMEC, Instituto Superior Técnico},
series = {ID 87},
abstract = {Large wind turbine blades have many advantages in terms of power efficiency, despite representing an hazard concerning the high loads applied on the structure. Since the end of the last century, some researchers have been discussing about passive control techniques. The implementation of this kind of aeroelastic response does not bring additional maintenance or weight, unlike active control, because there are no additional devices or complementary structures, and is very useful either to reduce fatigue loads or optimize energy output. The main purpose was to achieve an effective reduction in aerodynamic loading in a wind turbine blade applying the concept of bend-twist coupling. In the scope of this work, computational models were developed that simulated the fluid-structure interaction on a enhanced blade model. Coupled analysis demonstrated that this design can reduce aerodynamic load and maximum tip deflection in high wind speeds, thus proving to be a realistic passive control technique.
Keywords: Aeroelastic tailoring, Bend-twist coupling, Passive control, Fluid-structure interaction},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Large wind turbine blades have many advantages in terms of power efficiency, despite representing an hazard concerning the high loads applied on the structure. Since the end of the last century, some researchers have been discussing about passive control techniques. The implementation of this kind of aeroelastic response does not bring additional maintenance or weight, unlike active control, because there are no additional devices or complementary structures, and is very useful either to reduce fatigue loads or optimize energy output. The main purpose was to achieve an effective reduction in aerodynamic loading in a wind turbine blade applying the concept of bend-twist coupling. In the scope of this work, computational models were developed that simulated the fluid-structure interaction on a enhanced blade model. Coupled analysis demonstrated that this design can reduce aerodynamic load and maximum tip deflection in high wind speeds, thus proving to be a realistic passive control technique.
Keywords: Aeroelastic tailoring, Bend-twist coupling, Passive control, Fluid-structure interaction
Keywords: Aeroelastic tailoring, Bend-twist coupling, Passive control, Fluid-structure interaction
Campos, L. M. B. C.; Marta, A. C.
On thermo-acoustic acoustic-vortical-entropy waves and flow stability Proceedings Article
In: Proceedings of the 5th CEAS Air & Space Conference, TU Delft Delft, Netherlands, 2015.
@inproceedings{Campos:2015:CEAS,
title = {On thermo-acoustic acoustic-vortical-entropy waves and flow stability},
author = {L. M. B. C. Campos and A. C. Marta},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Campos_2015_CEAS.pdf, Full-text PDF
https://aerospace-europe.eu/digital-library/digital-library-ceas/papers-ceas-1/on-thermo-acoustic-acoustic-vortical-entropy-waves-and-flow-stability/},
year = {2015},
date = {2015-09-01},
booktitle = {Proceedings of the 5th CEAS Air & Space Conference},
address = {Delft, Netherlands},
organization = {TU Delft},
series = {ID 88},
abstract = {The noise of jet and rocket engines involves the coupling of sound to swirling flows and to heat exchanges leading in the more complex cases of triple interactions to acoustic-vortical-entropy (AVE) waves. The present paper presents as far as the authors are aware the first derivation of the AVE equation for axisymmetric linear non-dissipative perturbations of a compressible, non-isentropic, swirling mean flow, with constant axial velocity and constant angular velocity. The axisymmetric AVE wave equation is obtained for the radial velocity perturbation, specifying its radial dependence for a given frequency and axial wavenumber. The AVE wave equation in the case of zero axial wavenumber has only one singularity at the critical radius, where the isothermal Mach number for the swirl velocity is unity. The exact solution of the AVE wave equation is obtained as series expansions of Gaussian hypergeometric type valid inside, outside and around the critical layer, thus: (i) covering the whole flow region; (ii) identifying the singularity at the sonic condition at the critical layer; (iii) specifying near-axis and asymptotic solutions for small and large radius. Using polarization relations among wave variables specifies exactly the perturbations of: (i,ii) the radial and azimuthal velocity; (iii,iv) pressure and mass density; (v,vi) entropy and temperature. It is shown that the dependence of the AVE wave variables on the radial distance can be: (a) oscillatory with decaying amplitude; (b) monotonic with increasing amplitude. The case (b) of AVE wave amplitude increasing monotonically with the radial distance applies if the frequency times a function of the adiabatic exponent is less than the vorticity (or twice the angular velocity). In the opposite case (a) the oscillatory nature of acoustic waves predominates over the tendency for monotonic growth of vortical perturbations. Associating sound with stable potential flows and swirl with unstable vortical flows suggests a criterion valid in non-isentropic conditions, that is in the presence of heat exchanges, that is a condition for stable combustion in a confined space: the peak vorticity (multiplied by a factor of order unity dependent on the adiabatic exponent) should be less than the lowest or fundamental frequency of the cavity.
Keywords: Jet noise, Axisymmetric perturbations, Critical radius, Frobenius-Fuchs series, Gaussian hypergeometric functions},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
The noise of jet and rocket engines involves the coupling of sound to swirling flows and to heat exchanges leading in the more complex cases of triple interactions to acoustic-vortical-entropy (AVE) waves. The present paper presents as far as the authors are aware the first derivation of the AVE equation for axisymmetric linear non-dissipative perturbations of a compressible, non-isentropic, swirling mean flow, with constant axial velocity and constant angular velocity. The axisymmetric AVE wave equation is obtained for the radial velocity perturbation, specifying its radial dependence for a given frequency and axial wavenumber. The AVE wave equation in the case of zero axial wavenumber has only one singularity at the critical radius, where the isothermal Mach number for the swirl velocity is unity. The exact solution of the AVE wave equation is obtained as series expansions of Gaussian hypergeometric type valid inside, outside and around the critical layer, thus: (i) covering the whole flow region; (ii) identifying the singularity at the sonic condition at the critical layer; (iii) specifying near-axis and asymptotic solutions for small and large radius. Using polarization relations among wave variables specifies exactly the perturbations of: (i,ii) the radial and azimuthal velocity; (iii,iv) pressure and mass density; (v,vi) entropy and temperature. It is shown that the dependence of the AVE wave variables on the radial distance can be: (a) oscillatory with decaying amplitude; (b) monotonic with increasing amplitude. The case (b) of AVE wave amplitude increasing monotonically with the radial distance applies if the frequency times a function of the adiabatic exponent is less than the vorticity (or twice the angular velocity). In the opposite case (a) the oscillatory nature of acoustic waves predominates over the tendency for monotonic growth of vortical perturbations. Associating sound with stable potential flows and swirl with unstable vortical flows suggests a criterion valid in non-isentropic conditions, that is in the presence of heat exchanges, that is a condition for stable combustion in a confined space: the peak vorticity (multiplied by a factor of order unity dependent on the adiabatic exponent) should be less than the lowest or fundamental frequency of the cavity.
Keywords: Jet noise, Axisymmetric perturbations, Critical radius, Frobenius-Fuchs series, Gaussian hypergeometric functions
Keywords: Jet noise, Axisymmetric perturbations, Critical radius, Frobenius-Fuchs series, Gaussian hypergeometric functions
Mascarenhas, J.; Marta, A. C.; Infante, V.
Towards aeroacoustic-structural optimization of composite wind turbine blades Proceedings Article
In: Proceedings of the 10th International Conference on Composite Science and Technology, IDMEC, Instituto Superior Técnico Lisbon, Portugal, 2015.
@inproceedings{Mascarenhas:2015:ICCST,
title = {Towards aeroacoustic-structural optimization of composite wind turbine blades},
author = {J. Mascarenhas and A. C. Marta and V. Infante},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Mascarenhas_2015_ICCST.pdf, Full-text PDF
http://www.dem.ist.utl.pt/iccst10/},
year = {2015},
date = {2015-09-01},
booktitle = {Proceedings of the 10th International Conference on Composite Science and Technology},
address = {Lisbon, Portugal},
organization = {IDMEC, Instituto Superior Técnico},
series = {ID 100},
abstract = {The extension of an aeroacoustic wind turbine (WT) blade shape optimization to tackle composite structural mechanics is presented. The existing framework is briefly described, following a detailed description of the composite model developed using a finite element model (FEM) formulation. The internal structure of the WT blade is a spar box, for which a finite element code using shell elements is developed capable of handling isotropic, orthotropic and laminated composite materials. The FEM tool developed showed good agreement with comparable results obtained using a commercial software. The structural analysis capability developed is to be later coupled to the aeroacoustic design framework. This effort paves the way for the ultimate goal of optimizing a WT blade, both external shape for aerodynamic and aeroacoustic performance, and internal sizing for structural response, taking int account multiple disciplines, using a multidisciplinary optimization approach.
Keywords: Multidisciplinary design optimization, Composite materials, Wind turbine, Shape optimization},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
The extension of an aeroacoustic wind turbine (WT) blade shape optimization to tackle composite structural mechanics is presented. The existing framework is briefly described, following a detailed description of the composite model developed using a finite element model (FEM) formulation. The internal structure of the WT blade is a spar box, for which a finite element code using shell elements is developed capable of handling isotropic, orthotropic and laminated composite materials. The FEM tool developed showed good agreement with comparable results obtained using a commercial software. The structural analysis capability developed is to be later coupled to the aeroacoustic design framework. This effort paves the way for the ultimate goal of optimizing a WT blade, both external shape for aerodynamic and aeroacoustic performance, and internal sizing for structural response, taking int account multiple disciplines, using a multidisciplinary optimization approach.
Keywords: Multidisciplinary design optimization, Composite materials, Wind turbine, Shape optimization
Keywords: Multidisciplinary design optimization, Composite materials, Wind turbine, Shape optimization
Campos, L. M. B. C.; Marta, A. C.
On the stability of axisymmetric compressible, vortical non-isentropic flow Proceedings Article
In: Proceedings of the CleanAir 2015, 12th International Conference on Energy for a Clean Environment, IDMEC, Instituto Superior Técnico Lisbon, Portugal, 2015.
@inproceedings{Campos:2015:CleanAir,
title = {On the stability of axisymmetric compressible, vortical non-isentropic flow},
author = {L. M. B. C. Campos and A. C. Marta},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Campos_2015_CleanAir.pdf, Full-text PDF
https://cleanair2015.wixsite.com/cleanair2015},
year = {2015},
date = {2015-07-01},
booktitle = {Proceedings of the CleanAir 2015, 12th International Conference on Energy for a Clean Environment},
address = {Lisbon, Portugal},
organization = {IDMEC, Instituto Superior Técnico},
series = {Combustion Instabilities},
abstract = {The noise of jet and rocket engines involves the coupling of sound to swirling flows and to heat exchanges leading in the more complex cases of triple interactions to acoustic-vortical-entropy (AVE) waves. The present paper presents the derivation of the AVE equation for axisymmetric linear non-dissipative perturbations of a compressible, non-isentropic, swirling mean flow, with constant axial velocity and constant angular velocity. The axisymmetric AVE wave equation is obtained for the radial velocity perturbation, specifying its radial dependence for a given frequency and axial wavenumber. The AVE wave equation in the case of zero axial wavenumber has only one singularity at the critical radius, where the isothermal Mach number for the swirl velocity is unity. The exact solution of the AVE wave equation is obtained as series expansions of Gaussian hypergeometric type valid inside, outside and around the critical layer. Using polarization relations among wave variables specifies exactly the perturbations of: (i,ii) the radial and azimuthal velocity; (iii,iv) pressure and mass density; (v,vi) entropy and temperature. It is shown that the dependence of the AVE wave variables on the radial distance can be: (a) oscillatory with decaying amplitude; (b) monotonic with increasing amplitude. The case (b) of AVE wave amplitude increasing monotonically with the radial distance applies if the frequency times a function of the adiabatic exponent is less than the vorticity (or twice the angular velocity). In the opposite case (a) the oscillatory nature of acoustic waves predominates over the tendency for monotonic growth of vortical perturbations. Associating sound with stable potential flows and swirl with unstable vortical flows suggests a criterion valid in non-isentropic conditions, that is in the presence of heat exchanges, that is a condition for stable combustion in a confined space: the peak vorticity (multiplied by a factor of order unity dependent on the adiabatic exponent) should be less than the lowest or fundamental frequency of the cavity.
Keywords: Aeroacoustics, Absolute/convective instability, Critical layers},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
The noise of jet and rocket engines involves the coupling of sound to swirling flows and to heat exchanges leading in the more complex cases of triple interactions to acoustic-vortical-entropy (AVE) waves. The present paper presents the derivation of the AVE equation for axisymmetric linear non-dissipative perturbations of a compressible, non-isentropic, swirling mean flow, with constant axial velocity and constant angular velocity. The axisymmetric AVE wave equation is obtained for the radial velocity perturbation, specifying its radial dependence for a given frequency and axial wavenumber. The AVE wave equation in the case of zero axial wavenumber has only one singularity at the critical radius, where the isothermal Mach number for the swirl velocity is unity. The exact solution of the AVE wave equation is obtained as series expansions of Gaussian hypergeometric type valid inside, outside and around the critical layer. Using polarization relations among wave variables specifies exactly the perturbations of: (i,ii) the radial and azimuthal velocity; (iii,iv) pressure and mass density; (v,vi) entropy and temperature. It is shown that the dependence of the AVE wave variables on the radial distance can be: (a) oscillatory with decaying amplitude; (b) monotonic with increasing amplitude. The case (b) of AVE wave amplitude increasing monotonically with the radial distance applies if the frequency times a function of the adiabatic exponent is less than the vorticity (or twice the angular velocity). In the opposite case (a) the oscillatory nature of acoustic waves predominates over the tendency for monotonic growth of vortical perturbations. Associating sound with stable potential flows and swirl with unstable vortical flows suggests a criterion valid in non-isentropic conditions, that is in the presence of heat exchanges, that is a condition for stable combustion in a confined space: the peak vorticity (multiplied by a factor of order unity dependent on the adiabatic exponent) should be less than the lowest or fundamental frequency of the cavity.
Keywords: Aeroacoustics, Absolute/convective instability, Critical layers
Keywords: Aeroacoustics, Absolute/convective instability, Critical layers
Cardeira, A.; Marta, A. C.
Aeroelastic analysis of aircraft wings Proceedings Article
In: Proceedings of the CMN 2015 - Congress on Numerical Methods in Engineering, APMTAC/SEMNI Lisbon, Portugal, 2015.
@inproceedings{Cardeira:2015:CMN,
title = {Aeroelastic analysis of aircraft wings},
author = {A. Cardeira and A. C. Marta},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Cardeira_2015_CMN.pdf, Full-text PDF
http://www.dem.ist.utl.pt/cmn2015/},
year = {2015},
date = {2015-06-01},
booktitle = {Proceedings of the CMN 2015 - Congress on Numerical Methods in Engineering},
address = {Lisbon, Portugal},
organization = {APMTAC/SEMNI},
series = {ID 119},
abstract = {Aeroelasticity phenomena involve the study of the interaction between aerodynamic, inertial, and elastic forces (dynamic aeroelasticity). Modern aircraft structures, using more and more lightweight flexible composite materials, make the aeroelastic study an extremely important aspect in aircraft design. Flutter is a dynamic aeroelastic instability characterized by sustained oscillation of structure arising from interaction between those three forces acting on the body. The present work aims to study the flutter behavior on three-dimensional subsonic aircraft wings, using a computationally efficient method. For that, a computational aeroelasticity design framework is created using a custom developed panel method for the fluid flow analysis and a commercial software for the structural analysis. A validation of the flow solver is made using wind tunnel data, while the structural solver is verified using available tests. The coupling of the two domains is made using an adequate time discretization scheme. The results are presented for a reference wing. Following the wing baseline analysis, a parametric study under flutter conditions is performed, revealing some physically expected correlations: i) increasing the freestream velocity leads to higher vibration amplitude, whereas the frequency remains unchanged; ii) moving the wing spars aft or forward, causing the twist center to move away from the aerodynamic center, leads to instability; iii) decreasing the material density (weight) leads to higher flutter frequency and amplitude; iv) increasing the material stiffness (Young modulus) leads to higher frequency and smaller amplitude flutter. It is concluded that the framework shows very good agreement to the theoretical influences of the parameters studied. Despite the simplification of the fluid flow, which was assumed to be potential, this method proves to be a very useful tool in aircraft preliminary design.
Keywords: Aeroelasticity, Panel method, Fluid-structure interaction, Finite element method, Flutter, Divergence velocity},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Aeroelasticity phenomena involve the study of the interaction between aerodynamic, inertial, and elastic forces (dynamic aeroelasticity). Modern aircraft structures, using more and more lightweight flexible composite materials, make the aeroelastic study an extremely important aspect in aircraft design. Flutter is a dynamic aeroelastic instability characterized by sustained oscillation of structure arising from interaction between those three forces acting on the body. The present work aims to study the flutter behavior on three-dimensional subsonic aircraft wings, using a computationally efficient method. For that, a computational aeroelasticity design framework is created using a custom developed panel method for the fluid flow analysis and a commercial software for the structural analysis. A validation of the flow solver is made using wind tunnel data, while the structural solver is verified using available tests. The coupling of the two domains is made using an adequate time discretization scheme. The results are presented for a reference wing. Following the wing baseline analysis, a parametric study under flutter conditions is performed, revealing some physically expected correlations: i) increasing the freestream velocity leads to higher vibration amplitude, whereas the frequency remains unchanged; ii) moving the wing spars aft or forward, causing the twist center to move away from the aerodynamic center, leads to instability; iii) decreasing the material density (weight) leads to higher flutter frequency and amplitude; iv) increasing the material stiffness (Young modulus) leads to higher frequency and smaller amplitude flutter. It is concluded that the framework shows very good agreement to the theoretical influences of the parameters studied. Despite the simplification of the fluid flow, which was assumed to be potential, this method proves to be a very useful tool in aircraft preliminary design.
Keywords: Aeroelasticity, Panel method, Fluid-structure interaction, Finite element method, Flutter, Divergence velocity
Keywords: Aeroelasticity, Panel method, Fluid-structure interaction, Finite element method, Flutter, Divergence velocity
Rodrigues, S. S.; Marta, A. C.
Discrete adjoint mixing-plane formulation for multi-stage turbomachinery design Proceedings Article
In: Proceedings of the CMN 2015 - Congress on Numerical Methods in Engineering, APMTAC/SEMNI Lisbon, Portugal, 2015.
@inproceedings{Rodrigues:2015:CMN,
title = {Discrete adjoint mixing-plane formulation for multi-stage turbomachinery design},
author = {S. S. Rodrigues and A. C. Marta},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Rodrigues_2015_CMN.pdf, Full-text PDF
http://www.dem.ist.utl.pt/cmn2015/},
year = {2015},
date = {2015-06-01},
urldate = {2015-06-01},
booktitle = {Proceedings of the CMN 2015 - Congress on Numerical Methods in Engineering},
address = {Lisbon, Portugal},
organization = {APMTAC/SEMNI},
series = {ID 193},
abstract = {The use of computational fluid dynamics (CFD) tools in turbomachinery has seen an increase as a result of the exponential growth of computational power, as well as of improvements of the accuracy of numerical simulations. These tools are often used in optimization environments, where gradient-based optimization algorithms are the most common due to its efficiency. The optimization may contain a large number of design variables (typically in the order of thousands), such in cases of shape optimization. In these cases, the adjoint approach for calculating the gradients is beneficial, as it provides a way of obtaining function sensitivities with a computational cost that is independent of the number of design variables, as opposed to what happens with the well known finite-difference approach. The interaction between adjacent blade rows has an important impact on the whole performance of a multistage turbomachine. The most commonly used method to address these effects in the simulation of multiple rows is the mixing-plane treatment, that has become a standard industrial tool in the design environment. In this paper, improvements on the adjoint solver of a proprietary CFD solver for multistage turbomachinery applications are presented, namely the adjoint counterpart of the mixing-plane formulation used in the direct solver. The solver is developed using the discrete ADjoint approach, where the partial derivatives required for the assembly of the adjoint system of equations are obtained using automatic differentiation tools.
Keywords: Turbomachinery, Discrete adjoint, Multi-stage, Gradient-based optimization, Parallel processing, Design framework},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
The use of computational fluid dynamics (CFD) tools in turbomachinery has seen an increase as a result of the exponential growth of computational power, as well as of improvements of the accuracy of numerical simulations. These tools are often used in optimization environments, where gradient-based optimization algorithms are the most common due to its efficiency. The optimization may contain a large number of design variables (typically in the order of thousands), such in cases of shape optimization. In these cases, the adjoint approach for calculating the gradients is beneficial, as it provides a way of obtaining function sensitivities with a computational cost that is independent of the number of design variables, as opposed to what happens with the well known finite-difference approach. The interaction between adjacent blade rows has an important impact on the whole performance of a multistage turbomachine. The most commonly used method to address these effects in the simulation of multiple rows is the mixing-plane treatment, that has become a standard industrial tool in the design environment. In this paper, improvements on the adjoint solver of a proprietary CFD solver for multistage turbomachinery applications are presented, namely the adjoint counterpart of the mixing-plane formulation used in the direct solver. The solver is developed using the discrete ADjoint approach, where the partial derivatives required for the assembly of the adjoint system of equations are obtained using automatic differentiation tools.
Keywords: Turbomachinery, Discrete adjoint, Multi-stage, Gradient-based optimization, Parallel processing, Design framework
Keywords: Turbomachinery, Discrete adjoint, Multi-stage, Gradient-based optimization, Parallel processing, Design framework
Marta, A. C.; Gamboa, P. V.
Design of a solar RPAS for extended surveillance missions Proceedings Article
In: Proceedings of the 3rd LAETA Young Researchers Meeting, ADAI/FCT-UC Coimbra, Portugal, 2015.
@inproceedings{Marta:2015:LAETA,
title = {Design of a solar RPAS for extended surveillance missions},
author = {A. C. Marta and P. V. Gamboa},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Marta_2015_LAETA.pdf, Full-text PDF
https://3ejil.wordpress.com/submissoes/},
year = {2015},
date = {2015-05-01},
urldate = {2015-05-01},
booktitle = {Proceedings of the 3rd LAETA Young Researchers Meeting},
address = {Coimbra, Portugal},
organization = {ADAI/FCT-UC},
series = {ID Aero 07},
abstract = {The design of a very long endurance electric remotely piloted aircraft system (RPAS) accomplished by a multidisciplinary team of LAETA researchers is presented. The RPAS mission requirements are derived from civilian surveillance applications, such as forest, coast or border patrol. Very long endurance is achieved by an electric propulsion system assisted with solar cell arrays and a careful lightweight airframe design. The main steps of the design are covered, including numerical simulations, as well as construction, component and systems testing and flight testing. The feasibility of a green, low cost, small footprint RPAS is demonstrated.
Keywords: Hybrid propulsion, Solar energy, Long range, Fixed wing aircraft, Ground control station},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
The design of a very long endurance electric remotely piloted aircraft system (RPAS) accomplished by a multidisciplinary team of LAETA researchers is presented. The RPAS mission requirements are derived from civilian surveillance applications, such as forest, coast or border patrol. Very long endurance is achieved by an electric propulsion system assisted with solar cell arrays and a careful lightweight airframe design. The main steps of the design are covered, including numerical simulations, as well as construction, component and systems testing and flight testing. The feasibility of a green, low cost, small footprint RPAS is demonstrated.
Keywords: Hybrid propulsion, Solar energy, Long range, Fixed wing aircraft, Ground control station
Keywords: Hybrid propulsion, Solar energy, Long range, Fixed wing aircraft, Ground control station
Marta, A. C.; Gamboa, P.
Long endurance electric UAV for civilian surveillance missions Proceedings Article
In: Proceedings of the ICAS 2014 - 29th Congress of the International Council of the Aeronautical Sciences, International Council of the Aeronautical Sciences St.Petersburg, Russia, 2014, ISBN: 3-932182-80-4.
@inproceedings{Marta:2014b:ICAS,
title = {Long endurance electric UAV for civilian surveillance missions},
author = {A. C. Marta and P. Gamboa},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Marta_2014b_ICAS.pdf, Full-text PDF
https://www.icas.org/ICAS_ARCHIVE/ICAS2014/icas2014_materials.html},
isbn = {3-932182-80-4},
year = {2014},
date = {2014-09-01},
booktitle = {Proceedings of the ICAS 2014 - 29th Congress of the International Council of the Aeronautical Sciences},
address = {St.Petersburg, Russia},
organization = {International Council of the Aeronautical Sciences},
series = {ID 2014_0547},
abstract = {The design of a long endurance electric UAV is presented. The mission requirements are derived from civilian surveillance applications, such as forest, coast or border patrol. As such, long endurance is desired, which is tackled by a careful lightweight airframe design and an electric propulsion system assisted with solar cell arrays. The main steps of the design are covered, including numerical simulations, as well as construction, component and systems testing and flight testing. The feasibility of a green, low cost, small footprint UAV is demonstrated.
Keywords: Hybrid propulsion, Solar energy, Long range, Fixed wing, Ground control station},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
The design of a long endurance electric UAV is presented. The mission requirements are derived from civilian surveillance applications, such as forest, coast or border patrol. As such, long endurance is desired, which is tackled by a careful lightweight airframe design and an electric propulsion system assisted with solar cell arrays. The main steps of the design are covered, including numerical simulations, as well as construction, component and systems testing and flight testing. The feasibility of a green, low cost, small footprint UAV is demonstrated.
Keywords: Hybrid propulsion, Solar energy, Long range, Fixed wing, Ground control station
Keywords: Hybrid propulsion, Solar energy, Long range, Fixed wing, Ground control station
Rodrigues, S. S.; Marta, A. C.
Framework for low-noise wind turbine blade design Proceedings Article
In: Proceedings of the ICAS 2014 - 29th Congress of the International Council of the Aeronautical Sciences, International Council of the Aeronautical Sciences St.Petersburg, Russia, 2014, ISBN: 3-932182-80-4.
@inproceedings{Rodrigues:2014:ICAS,
title = {Framework for low-noise wind turbine blade design},
author = {S. S. Rodrigues and A. C. Marta},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Rodrigues_2014_ICAS.pdf, Full-text PDF
https://www.icas.org/ICAS_ARCHIVE/ICAS2014/icas2014_materials.html},
isbn = {3-932182-80-4},
year = {2014},
date = {2014-09-01},
booktitle = {Proceedings of the ICAS 2014 - 29th Congress of the International Council of the Aeronautical Sciences},
address = {St.Petersburg, Russia},
organization = {International Council of the Aeronautical Sciences},
series = {ID 2014_589},
abstract = {Power production from wind energy has been increasing over the past decades, with more areas being used as wind farms and larger wind turbines (WTs) being built. As the awareness of the impact of wind energy on the environment and human health as also increased, so has the interest in developing fast turnaround WT blade design frameworks capable of predicting both aerodynamic and aeroacoustic performance. In this paper, the development of such framework is described and the results of single and multi operating point optimizations of the blades of the commercial AOC 15/50 WT are presented and discussed. Noise reductions of up to 9.8% were achieved, with a cost of only 1% in energy production.
Keywords: Aeroacoustic, Multidisciplinary optimization, NURBS, Genetic algorithms},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Power production from wind energy has been increasing over the past decades, with more areas being used as wind farms and larger wind turbines (WTs) being built. As the awareness of the impact of wind energy on the environment and human health as also increased, so has the interest in developing fast turnaround WT blade design frameworks capable of predicting both aerodynamic and aeroacoustic performance. In this paper, the development of such framework is described and the results of single and multi operating point optimizations of the blades of the commercial AOC 15/50 WT are presented and discussed. Noise reductions of up to 9.8% were achieved, with a cost of only 1% in energy production.
Keywords: Aeroacoustic, Multidisciplinary optimization, NURBS, Genetic algorithms
Keywords: Aeroacoustic, Multidisciplinary optimization, NURBS, Genetic algorithms
Marta, A. C.; Shankaran, S.
On the optimal location of bleed ports for gas turbine cooling Proceedings Article
In: Proceedings of the ICAS 2014 - 29th Congress of the International Council of the Aeronautical Sciences, International Council of the Aeronautical Sciences St.Petersburg, Russia, 2014, ISBN: 3-932182-80-4.
@inproceedings{Marta:2014a:ICAS,
title = {On the optimal location of bleed ports for gas turbine cooling},
author = {A. C. Marta and S. Shankaran},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Marta_2014a_ICAS.pdf, Full-text PDF
https://www.icas.org/ICAS_ARCHIVE/ICAS2014/icas2014_materials.html},
isbn = {3-932182-80-4},
year = {2014},
date = {2014-09-01},
booktitle = {Proceedings of the ICAS 2014 - 29th Congress of the International Council of the Aeronautical Sciences},
address = {St.Petersburg, Russia},
organization = {International Council of the Aeronautical Sciences},
series = {ID 2014_517},
abstract = {A novel approach to determine the optimal location of bleed ports for gas turbine cooling is presented. It is based on the adjoint method, thus being computationally very efficient and producing an extensive sensitivity analysis to the control variables. This approach easily evaluates every point of all surfaces (hub, blade and casing), giving the designer a new level of information for determining the optimal location that delivers the most impact in cooling at the least expense of required bleed flow.
Keywords: Adjoint method, Sensitivity analysis, Source terms, Heat transfer},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
A novel approach to determine the optimal location of bleed ports for gas turbine cooling is presented. It is based on the adjoint method, thus being computationally very efficient and producing an extensive sensitivity analysis to the control variables. This approach easily evaluates every point of all surfaces (hub, blade and casing), giving the designer a new level of information for determining the optimal location that delivers the most impact in cooling at the least expense of required bleed flow.
Keywords: Adjoint method, Sensitivity analysis, Source terms, Heat transfer
Keywords: Adjoint method, Sensitivity analysis, Source terms, Heat transfer
Amândio, L. F.; Marta, A. C.; Afonso, F.; Vale, J. L.; Suleman, A.
Stochastic optimization in aircraft design Proceedings Article
In: Proceedings of the EngOpt 2014 - 4th International Conference on Engineering Optimization, pp. 267-272, IDMEC, Instituto Superior Técnico Lisbon, Portugal, 2014, ISBN: 9780429226687.
@inproceedings{Amandio:2014:ENGOPT,
title = {Stochastic optimization in aircraft design},
author = {L. F. Amândio and A. C. Marta and F. Afonso and J. L. Vale and A. Suleman},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Amandio_2014_ENGOPT.pdf, Full-text PDF
https://www.taylorfrancis.com/books/mono/10.1201/b17488
http://www.dem.ist.utl.pt/engopt2014/},
isbn = {9780429226687},
year = {2014},
date = {2014-09-01},
booktitle = {Proceedings of the EngOpt 2014 - 4th International Conference on Engineering Optimization},
pages = {267-272},
address = {Lisbon, Portugal},
organization = {IDMEC, Instituto Superior Técnico},
series = {ENGOPT2014-5623},
abstract = {This paper focuses on analysing the advantages and disadvantages of using stochastic optimization, especially in aircraft design problems. First, a literature review served as a starting point to choosing some of the most common and promising methods of robust design optimization, reliability based design optimization and robust and reliability based design optimization. The chosen methods were Monte Carlo, method of moments, Sigma point, reliability index approach, performance measure approach, sequential optimization and reliability assessment, and reliable design space. After implementing these methods, they were were firstly tested using an analytical function and their performances compared. Four of these methods were then chosen to be implemented in a multidisciplinary optimization tool specially tailored to solve aircraft optimization problems. To evaluate the chosen methods in a more realistic environment, two new reliability based test cases related to aircraft design were developed. In these test cases, surrogate models were employed instead of the more computationally expensive disciplinary analysis, with the main objective being the study of how the efficiency of each method changed with the number of uncertainty parameters. The obtained results revealed that the efficiency of each method is closely related to the type of problem solved. While in the analytical case, for high levels of uncertainty, the robust optimization method showed some difficulties in achieving the target reliability, in the aircraft design cases, it proved to be the best method in terms of the relation between accuracy and computational cost.
Keywords: Uncertainty propagation, Robust design optimization, Reliability-based design optimization, Multidisciplinary optimization, Benchmark methods},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
This paper focuses on analysing the advantages and disadvantages of using stochastic optimization, especially in aircraft design problems. First, a literature review served as a starting point to choosing some of the most common and promising methods of robust design optimization, reliability based design optimization and robust and reliability based design optimization. The chosen methods were Monte Carlo, method of moments, Sigma point, reliability index approach, performance measure approach, sequential optimization and reliability assessment, and reliable design space. After implementing these methods, they were were firstly tested using an analytical function and their performances compared. Four of these methods were then chosen to be implemented in a multidisciplinary optimization tool specially tailored to solve aircraft optimization problems. To evaluate the chosen methods in a more realistic environment, two new reliability based test cases related to aircraft design were developed. In these test cases, surrogate models were employed instead of the more computationally expensive disciplinary analysis, with the main objective being the study of how the efficiency of each method changed with the number of uncertainty parameters. The obtained results revealed that the efficiency of each method is closely related to the type of problem solved. While in the analytical case, for high levels of uncertainty, the robust optimization method showed some difficulties in achieving the target reliability, in the aircraft design cases, it proved to be the best method in terms of the relation between accuracy and computational cost.
Keywords: Uncertainty propagation, Robust design optimization, Reliability-based design optimization, Multidisciplinary optimization, Benchmark methods
Keywords: Uncertainty propagation, Robust design optimization, Reliability-based design optimization, Multidisciplinary optimization, Benchmark methods
Rodrigues, S. S.; Marta, A. C.
Design of after-market wind turbine blade add-ons for noise reduction Proceedings Article
In: Proceedings of the EngOpt 2014 - 4th International Conference on Engineering Optimization, pp. 245-249, IDMEC, Instituto Superior Técnico Lisbon, Portugal, 2014, ISBN: 9780429226687.
@inproceedings{Rodrigues:2014:ENGOPT,
title = {Design of after-market wind turbine blade add-ons for noise reduction},
author = {S. S. Rodrigues and A. C. Marta},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Rodrigues_2014_ENGOPT.pdf, Full-text PDF
https://www.taylorfrancis.com/books/mono/10.1201/b17488
http://www.dem.ist.utl.pt/engopt2014/},
isbn = {9780429226687},
year = {2014},
date = {2014-09-01},
booktitle = {Proceedings of the EngOpt 2014 - 4th International Conference on Engineering Optimization},
pages = {245-249},
address = {Lisbon, Portugal},
organization = {IDMEC, Instituto Superior Técnico},
series = {ENGOPT2014-5592},
abstract = {As a result of the continous growth of Wind Turbines (WTs) implementation worldwide, the problem of WT noise has become more relevant than ever. The increase of noise legal constraints and their non uniformity across different countries and/or regions make it important to address this problem early in the design phase of a WT. However, one might want to reduce the noise produced by WTs that are already in use to comply with new stricter noise limits. In the present work, this problem is addressed by using a WT blade optimization framework to obtain the shape of blade add-ons that could be attached to the blade of a WT (for example, by using some kind of adhesive) to reduce its noise without compromising the performance. Blade Element Momentum (BEM) theory was used to compute the aerodynamic performance of the WT and semi-empirical models were used for the airfoil self-noise and turbulence interaction noise prediction. The aerodynamic analysis of the cross sectional airfoil shapes, required for both the BEM calculations and noise predictions, is performed using the viscous-inviscid interactive code XFOIL. Non-dominated Sorting Genetic Algorithm-II was used as the optimization algorithm. Two NURBS curves were used to define the baseline cross sectional airfoil shape of the blade at certain control sections and another two to define a variation from the baseline, totalling up to 54 design variables. The framework was used to optimize the AOC 15/50 WT, a commercial downwind, three bladed WT. Optimal solutions were selected from the Pareto front and discussed in detail. These solutions ranged from an increase in energy production of 5.2% to a decrease in noise levels of 5.9%. The results demonstrated that, while it is preferable to address noise concerns in the design phase of the WT, it is possible to address them with favourable results after its construction..
Keywords: Noise reduction, Aeroacoustic analysis, Airfoil self-noise, Design optimization, Retrofitting, Multi-objective optimization},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
As a result of the continous growth of Wind Turbines (WTs) implementation worldwide, the problem of WT noise has become more relevant than ever. The increase of noise legal constraints and their non uniformity across different countries and/or regions make it important to address this problem early in the design phase of a WT. However, one might want to reduce the noise produced by WTs that are already in use to comply with new stricter noise limits. In the present work, this problem is addressed by using a WT blade optimization framework to obtain the shape of blade add-ons that could be attached to the blade of a WT (for example, by using some kind of adhesive) to reduce its noise without compromising the performance. Blade Element Momentum (BEM) theory was used to compute the aerodynamic performance of the WT and semi-empirical models were used for the airfoil self-noise and turbulence interaction noise prediction. The aerodynamic analysis of the cross sectional airfoil shapes, required for both the BEM calculations and noise predictions, is performed using the viscous-inviscid interactive code XFOIL. Non-dominated Sorting Genetic Algorithm-II was used as the optimization algorithm. Two NURBS curves were used to define the baseline cross sectional airfoil shape of the blade at certain control sections and another two to define a variation from the baseline, totalling up to 54 design variables. The framework was used to optimize the AOC 15/50 WT, a commercial downwind, three bladed WT. Optimal solutions were selected from the Pareto front and discussed in detail. These solutions ranged from an increase in energy production of 5.2% to a decrease in noise levels of 5.9%. The results demonstrated that, while it is preferable to address noise concerns in the design phase of the WT, it is possible to address them with favourable results after its construction..
Keywords: Noise reduction, Aeroacoustic analysis, Airfoil self-noise, Design optimization, Retrofitting, Multi-objective optimization
Keywords: Noise reduction, Aeroacoustic analysis, Airfoil self-noise, Design optimization, Retrofitting, Multi-objective optimization
Tojo, B. M.; Marta, A. C.
Aero-structural blade design of a high-power wind turbine Proceedings Article
In: Proceedings of the 9th AIAA Multidisciplinary Design Optimization Specialist Conference, American Institute of Aeronautics and Astronautics Boston, USA, 2013.
@inproceedings{Tojo:2013:AIAA,
title = {Aero-structural blade design of a high-power wind turbine},
author = {B. M. Tojo and A. C. Marta},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Tojo_2013_AIAA.pdf, Full-text PDF
https://arc.aiaa.org/doi/10.2514/6.2013-1531},
doi = {10.2514/6.2013-1531},
year = {2013},
date = {2013-04-01},
booktitle = {Proceedings of the 9th AIAA Multidisciplinary Design Optimization Specialist Conference},
address = {Boston, USA},
organization = {American Institute of Aeronautics and Astronautics},
series = {AIAA 2013-1531},
abstract = {The present work focused on the development of an aero-structural design framework for a high-power horizontal axis wind turbine. To achieve this, it was necessary to carefully characterize the blade, to develop a suitable fluid-structure interaction solver and, finally, to combine both with post-processing tools. A fluid-structure interaction solver was developed for OpenFOAM, dedicated to simulate wind turbine rotors. The fluid-structure coupling was achieved through a loose coupling strategy, which means that there are separated solvers for the flow and structure analysis, which are combined through the update of boundary conditions. To simulate the blades rotation movement, an approach based on the single rotating frame method was used, meaning that the whole domain rotated with the turbine rotor with a constant angular velocity. The simulations of the rotor produced valid and interesting results namely, correct flow fields and pressure distributions. Considering that it is expected horizontal axis wind turbine rotors to continue growing to even larger sizes than the one modeled, it was shown that blade displacements due to flow induced forces are definitely a problem that needs to be taken into account when designing new wind turbine blades.
Keywords: Wind turbine, Fluid-structure interaction, CFD development, High-fidelity analysis, Single rotating frame, OpenFOAM, ICEM-CFD},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
The present work focused on the development of an aero-structural design framework for a high-power horizontal axis wind turbine. To achieve this, it was necessary to carefully characterize the blade, to develop a suitable fluid-structure interaction solver and, finally, to combine both with post-processing tools. A fluid-structure interaction solver was developed for OpenFOAM, dedicated to simulate wind turbine rotors. The fluid-structure coupling was achieved through a loose coupling strategy, which means that there are separated solvers for the flow and structure analysis, which are combined through the update of boundary conditions. To simulate the blades rotation movement, an approach based on the single rotating frame method was used, meaning that the whole domain rotated with the turbine rotor with a constant angular velocity. The simulations of the rotor produced valid and interesting results namely, correct flow fields and pressure distributions. Considering that it is expected horizontal axis wind turbine rotors to continue growing to even larger sizes than the one modeled, it was shown that blade displacements due to flow induced forces are definitely a problem that needs to be taken into account when designing new wind turbine blades.
Keywords: Wind turbine, Fluid-structure interaction, CFD development, High-fidelity analysis, Single rotating frame, OpenFOAM, ICEM-CFD
Keywords: Wind turbine, Fluid-structure interaction, CFD development, High-fidelity analysis, Single rotating frame, OpenFOAM, ICEM-CFD
Coimbra, J. F. G.; Marta, A. C.
Aero-acoustic optimization of airfoils for wind turbines Proceedings Article
In: Proceedings of the 9th AIAA Multidisciplinary Design Optimization Specialist Conference, American Institute of Aeronautics and Astronautics Boston, USA, 2013.
@inproceedings{Coimbra:2013:AIAA,
title = {Aero-acoustic optimization of airfoils for wind turbines},
author = {J. F. G. Coimbra and A. C. Marta},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Coimbra_2013_AIAA.pdf, Full-text PDF
https://arc.aiaa.org/doi/10.2514/6.2013-1601},
doi = {10.2514/6.2013-1601},
year = {2013},
date = {2013-04-01},
booktitle = {Proceedings of the 9th AIAA Multidisciplinary Design Optimization Specialist Conference},
address = {Boston, USA},
organization = {American Institute of Aeronautics and Astronautics},
series = {AIAA 2013-1601},
abstract = {The subject of airfoil design, in the context of wind turbines, is approached with the objective of optimizing the geometry for the best aerodynamic and aero-acoustic trade-off. The work developed is made up of four stages: the aerodynamic analysis module, which consists of two parts, one responsible for airfoil design and parameterization and another dedicated to flow analysis; an aero-acoustic module, based in the semi-empirical model from Brooks, Pope and Marcolini and the turbulent inflow prediction scheme from Moriarty, Guidati and Migliore; the integration of both modules in one single computational tool; and the development of a multi-objective optimization framework. The airfoil geometry build tool is based on the mathematical description of Bezier curves and, the wind turbine dedicated computational tool Rfoil has been used for boundary-layer modeling and inclusion of rotational effects. The code developed integrated both developed modules in a single interactive shell in Python. The Python module pyOpt was selected as the interactive development environment in which the optimization took place. A genetic algorithm was selected to handle multiple local minima and multi-objective problems. Several airfoil families, commonly used in the wind turbine technology, were analyzed from the aerodynamic and aero-acoustic perspectives with the developed tools, and used as reference for general comparison. Optimized airfoil geometries, that either minimize noise emission or favour aerodynamic performance were obtained and classes of aero-acoustically optimized airfoils were identified in the resulting Pareto fronts.
Keywords: Airfoil design, Wind turbines, Aeroacoustics, Multi-objective optimization, Multidisciplinary optimization},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
The subject of airfoil design, in the context of wind turbines, is approached with the objective of optimizing the geometry for the best aerodynamic and aero-acoustic trade-off. The work developed is made up of four stages: the aerodynamic analysis module, which consists of two parts, one responsible for airfoil design and parameterization and another dedicated to flow analysis; an aero-acoustic module, based in the semi-empirical model from Brooks, Pope and Marcolini and the turbulent inflow prediction scheme from Moriarty, Guidati and Migliore; the integration of both modules in one single computational tool; and the development of a multi-objective optimization framework. The airfoil geometry build tool is based on the mathematical description of Bezier curves and, the wind turbine dedicated computational tool Rfoil has been used for boundary-layer modeling and inclusion of rotational effects. The code developed integrated both developed modules in a single interactive shell in Python. The Python module pyOpt was selected as the interactive development environment in which the optimization took place. A genetic algorithm was selected to handle multiple local minima and multi-objective problems. Several airfoil families, commonly used in the wind turbine technology, were analyzed from the aerodynamic and aero-acoustic perspectives with the developed tools, and used as reference for general comparison. Optimized airfoil geometries, that either minimize noise emission or favour aerodynamic performance were obtained and classes of aero-acoustically optimized airfoils were identified in the resulting Pareto fronts.
Keywords: Airfoil design, Wind turbines, Aeroacoustics, Multi-objective optimization, Multidisciplinary optimization
Keywords: Airfoil design, Wind turbines, Aeroacoustics, Multi-objective optimization, Multidisciplinary optimization
Campos, L. M. B. C.; Marta, A. C.
Fundamental bending frequencies of tapered wings Proceedings Article
In: Proceedings of the 54th AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics, and Materials Conference, American Institute of Aeronautics and Astronautics Boston, USA, 2013.
@inproceedings{Campos:2013:AIAA,
title = {Fundamental bending frequencies of tapered wings},
author = {L. M. B. C. Campos and A. C. Marta},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Campos_2013_AIAA.pdf, Full-text PDF
https://arc.aiaa.org/doi/10.2514/6.2013-1633},
doi = {10.2514/6.2013-1633},
year = {2013},
date = {2013-04-01},
booktitle = {Proceedings of the 54th AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics, and Materials Conference},
address = {Boston, USA},
organization = {American Institute of Aeronautics and Astronautics},
series = {AIAA 2013-1633},
abstract = {The bending frequencies of a wing are calculated based on the model of a beam clamped at the root and free at the tip; since the mass and the area moment of inertia (per unit span) vary along the span, a non-uniform beam is considered. For a sweptback wing with straight leading- and trailing-edges, the chord is a linear function of the span; the same linear function of the span applies to thickness, in the case of constant thickness-to-chord ratio. Thus, the bending modes of a non-uniform beam are considered, with mass and area moment of inertia respectively quadratic and quartic functions of the span. There is no exact solution expressible in finite terms using elementary functions, and thus power series expansions are used. The boundary conditions, that the wing is clamped at the root and free at the tip, lead to the natural bending frequencies. The fundamental bending frequency is calculated for a delta wing, and compared with a rectangular wing, with the same span, mean chord and thickness, mass density and Young modulus. It is shown that the fundamental frequency is higher by a factor 4.96 for the delta wing, that is, it is stiffer because it has a higher proportion of the mass near the root; it is also shown that the case of the tapered wing is intermediate between the delta and the rectangular wing. Lastly, the analytical results obtained are used to validate same numerical modal analyses of rectangular and delta wing beams using high-fidelity finite-element model software.
Keywords: Cantilever beam, Modal analysis, Transverse vibrations, Bending frequencies, Analytical solution, Power series expansion, Frobenius–Fuchs series},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
The bending frequencies of a wing are calculated based on the model of a beam clamped at the root and free at the tip; since the mass and the area moment of inertia (per unit span) vary along the span, a non-uniform beam is considered. For a sweptback wing with straight leading- and trailing-edges, the chord is a linear function of the span; the same linear function of the span applies to thickness, in the case of constant thickness-to-chord ratio. Thus, the bending modes of a non-uniform beam are considered, with mass and area moment of inertia respectively quadratic and quartic functions of the span. There is no exact solution expressible in finite terms using elementary functions, and thus power series expansions are used. The boundary conditions, that the wing is clamped at the root and free at the tip, lead to the natural bending frequencies. The fundamental bending frequency is calculated for a delta wing, and compared with a rectangular wing, with the same span, mean chord and thickness, mass density and Young modulus. It is shown that the fundamental frequency is higher by a factor 4.96 for the delta wing, that is, it is stiffer because it has a higher proportion of the mass near the root; it is also shown that the case of the tapered wing is intermediate between the delta and the rectangular wing. Lastly, the analytical results obtained are used to validate same numerical modal analyses of rectangular and delta wing beams using high-fidelity finite-element model software.
Keywords: Cantilever beam, Modal analysis, Transverse vibrations, Bending frequencies, Analytical solution, Power series expansion, Frobenius–Fuchs series
Keywords: Cantilever beam, Modal analysis, Transverse vibrations, Bending frequencies, Analytical solution, Power series expansion, Frobenius–Fuchs series
Campos, L. M. B. C.; Marta, A. C.
Analytical natural frequencies of tapered wings Proceedings Article
In: Proceedings of the ICAS 2012 - 28th Congress of the International Council of the Aeronautical Sciences, International Council of the Aeronautical Sciences Brisbane, Australia, 2012, ISBN: 978-0-9565333-1-9.
@inproceedings{Campos:2012:ICAS,
title = {Analytical natural frequencies of tapered wings},
author = {L. M. B. C. Campos and A. C. Marta},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Campos_2012_ICAS.pdf, Full-text PDF
https://www.icas.org/ICAS_ARCHIVE/ICAS2012/},
isbn = {978-0-9565333-1-9},
year = {2012},
date = {2012-09-01},
booktitle = {Proceedings of the ICAS 2012 - 28th Congress of the International Council of the Aeronautical Sciences},
address = {Brisbane, Australia},
organization = {International Council of the Aeronautical Sciences},
series = {ICAS2012-P3.2},
abstract = {The bending frequencies of a wing are calculated based on the model of a beam clamped at the root and free at the tip; since the mass and the moment of inertia (per unit span) vary along the span, a non-uniform beam is considered. For a sweptback wing with straight leading- and trailing-edges, the chord is a linear function of the span; the same linear function of the span applies to thickness, in the case of constant thickness-to-chord ratio. Thus, the bending modes of a non-uniform beam are considered, with mass and moment of inertia respectively quadratic and quartic functions of the span. There is no exact solu-
tion expressible in finite terms using elementary functions, and thus power series expansions are used. The boundary conditions, that the wing is clamped at the root and free at the tip, lead to the natural bending frequencies. The fundamental bending frequency is calculated for a delta wing, and compared with a rectangular wing, with the same span, mean chord and thickness, mass density and Young modulus. It is shown that the fundamental frequency is higher by a factor 11.32 for the delta wing., i.e., it is stiffer because it has a higher proportion of the mass near the root; it is also shown that the case of the tapered sweptback wing is intermediate between the delta and the rectangular wing.
Keywords: Bending, Vibration, Non-uniform beam, Frobenius-Fuchs series, Eigenvalues},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
The bending frequencies of a wing are calculated based on the model of a beam clamped at the root and free at the tip; since the mass and the moment of inertia (per unit span) vary along the span, a non-uniform beam is considered. For a sweptback wing with straight leading- and trailing-edges, the chord is a linear function of the span; the same linear function of the span applies to thickness, in the case of constant thickness-to-chord ratio. Thus, the bending modes of a non-uniform beam are considered, with mass and moment of inertia respectively quadratic and quartic functions of the span. There is no exact solu-
tion expressible in finite terms using elementary functions, and thus power series expansions are used. The boundary conditions, that the wing is clamped at the root and free at the tip, lead to the natural bending frequencies. The fundamental bending frequency is calculated for a delta wing, and compared with a rectangular wing, with the same span, mean chord and thickness, mass density and Young modulus. It is shown that the fundamental frequency is higher by a factor 11.32 for the delta wing., i.e., it is stiffer because it has a higher proportion of the mass near the root; it is also shown that the case of the tapered sweptback wing is intermediate between the delta and the rectangular wing.
Keywords: Bending, Vibration, Non-uniform beam, Frobenius-Fuchs series, Eigenvalues
tion expressible in finite terms using elementary functions, and thus power series expansions are used. The boundary conditions, that the wing is clamped at the root and free at the tip, lead to the natural bending frequencies. The fundamental bending frequency is calculated for a delta wing, and compared with a rectangular wing, with the same span, mean chord and thickness, mass density and Young modulus. It is shown that the fundamental frequency is higher by a factor 11.32 for the delta wing., i.e., it is stiffer because it has a higher proportion of the mass near the root; it is also shown that the case of the tapered sweptback wing is intermediate between the delta and the rectangular wing.
Keywords: Bending, Vibration, Non-uniform beam, Frobenius-Fuchs series, Eigenvalues
Marta, A. C.; Shankaran, S.
Exploring the use of adjoint methods for detailed sensitivity analysis on turbomachinery Proceedings Article
In: Proceedings of the EngOpt 2012 - 3rd International Conference on Engineering Optimization, COPPE, UFRJ Rio de Janeiro, Brazil, 2012, ISBN: 9788576503439.
@inproceedings{Marta:2012:ENGOPT,
title = {Exploring the use of adjoint methods for detailed sensitivity analysis on turbomachinery},
author = {A. C. Marta and S. Shankaran},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Marta_2012_ENGOPT.pdf, Full-text PDF
http://www.swge.com.br/engopt/},
isbn = {9788576503439},
year = {2012},
date = {2012-07-01},
booktitle = {Proceedings of the EngOpt 2012 - 3rd International Conference on Engineering Optimization},
address = {Rio de Janeiro, Brazil},
organization = {COPPE, UFRJ},
series = {ENGOPT2012-377},
abstract = {During the last decade, significant progress has been made in the field of optimization using high-fidelity models. The increased use of adjoint methods has allowed the computation of sensitivity information, required by gradient-based optimizers, in a very efficient manner. In terms of fluid dynamics, while the first applications were focused on aerodynamic shape design, recent approaches to the development of adjoint solvers, namely with the use of automatic differentiation tools, have made possible to extend their capabilities far beyond that. The present paper briefly describes a discrete adjoint method implementation for a generic CFD solver and lays down the steps to estimate sensitivities of functions of interest with respect to any variables handled by the flow solver. The applications presented are based on turbomachinery blade design problems. Two different capabilities are illustrated: one more traditional geared toward shape optimization that focuses on estimating gradients of some turbomachinery aerodynamic performance parameters with respect to blade geometry, and another more innovative geared toward estimating the effect of the inlet or exit boundary conditions on some aerothermal performance parameters. The computational cost required by this method in terms of CPU time is considerably reduced compared to the popular finite-difference method, at the expense of a more complex implementation. The detailed sensitivity information obtained is discussed from a designer perspective. Other possible applications of adjoint methods are listed and their development implications are also described.
Keywords: Turbomachinery, Shape, Boundary conditions, Optimization, Adjoint method, Sensitivity},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
During the last decade, significant progress has been made in the field of optimization using high-fidelity models. The increased use of adjoint methods has allowed the computation of sensitivity information, required by gradient-based optimizers, in a very efficient manner. In terms of fluid dynamics, while the first applications were focused on aerodynamic shape design, recent approaches to the development of adjoint solvers, namely with the use of automatic differentiation tools, have made possible to extend their capabilities far beyond that. The present paper briefly describes a discrete adjoint method implementation for a generic CFD solver and lays down the steps to estimate sensitivities of functions of interest with respect to any variables handled by the flow solver. The applications presented are based on turbomachinery blade design problems. Two different capabilities are illustrated: one more traditional geared toward shape optimization that focuses on estimating gradients of some turbomachinery aerodynamic performance parameters with respect to blade geometry, and another more innovative geared toward estimating the effect of the inlet or exit boundary conditions on some aerothermal performance parameters. The computational cost required by this method in terms of CPU time is considerably reduced compared to the popular finite-difference method, at the expense of a more complex implementation. The detailed sensitivity information obtained is discussed from a designer perspective. Other possible applications of adjoint methods are listed and their development implications are also described.
Keywords: Turbomachinery, Shape, Boundary conditions, Optimization, Adjoint method, Sensitivity
Keywords: Turbomachinery, Shape, Boundary conditions, Optimization, Adjoint method, Sensitivity
Shankaran, S.; Marta, A. C.; Venugopal, P.; Barr, B.; Wang, Q.
Interpretation of adjoint solutions for turbomachinery flows Proceedings Article
In: Proceedings of the ASME Turbo Expo 2012: Power for Land, Sea and Air, pp. 2229-2242, ASME International Gas Turbine Institute Copenhagen, Denmark, 2012, ISBN: 978-0-7918-4474-8.
@inproceedings{Shankaran:2012a:IGTI,
title = {Interpretation of adjoint solutions for turbomachinery flows},
author = {S. Shankaran and A. C. Marta and P. Venugopal and B. Barr and Q. Wang},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Shankaran_2012a_IGTI.pdf, Full-text PDF
https://asmedigitalcollection.asme.org/GT/proceedings-abstract/GT2012/44748/2229/365438},
doi = {10.1115/GT2012-69588},
isbn = {978-0-7918-4474-8},
year = {2012},
date = {2012-06-01},
booktitle = {Proceedings of the ASME Turbo Expo 2012: Power for Land, Sea and Air},
pages = {2229-2242},
address = {Copenhagen, Denmark},
organization = {ASME International Gas Turbine Institute},
series = {GT2012-69588},
abstract = {While the mathematical derivation of the adjoint equations and its numerical implementation is well established, there is scant discussion on the understanding of the adjoint solution by itself. As this is a field solution of similar resolution of the flowfield, there is wealth of data that can be used for design guidance. This paper addresses this specific topic. In particular, we take representative cases from turbomachinery aerodynamic problems and use the adjoint solution to identify the “physical insight” it provides. We aim to tie the adjoint solution to the flow-field which has physical properties. Towards this end, we first look at three problems 1) a fan, 2) a compressor rotor and stator, 3) a low pressure turbine. In all three of them, we focus on changes related to geometry, but one can also realize the changes using other inputs to the flow solver (eg. boundary conditions). We show how the adjoint counter-part of the density, the velocity fields and the turbulence quantities can be used to provide insights into the nature of changes the designer can induce to cause improvement in the performance metric of interest. We also discuss how to use adjoint solutions for problems with constraints to further refine the changes. Finally, we use a problem where it is not immediately apparent what geometry changes need to be used for further evaluation with optimization algorithms. In this problem, we use the adjoint and flow solution on a turbine strut, to determine the kind of end-wall treatments that reduce the loss. These changes are then implemented to show that the loss is reduced by close to 8%.
Keywords: Design optimization, Flow dynamics, Fan, Compressor, Turbine, Endwall profile},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
While the mathematical derivation of the adjoint equations and its numerical implementation is well established, there is scant discussion on the understanding of the adjoint solution by itself. As this is a field solution of similar resolution of the flowfield, there is wealth of data that can be used for design guidance. This paper addresses this specific topic. In particular, we take representative cases from turbomachinery aerodynamic problems and use the adjoint solution to identify the “physical insight” it provides. We aim to tie the adjoint solution to the flow-field which has physical properties. Towards this end, we first look at three problems 1) a fan, 2) a compressor rotor and stator, 3) a low pressure turbine. In all three of them, we focus on changes related to geometry, but one can also realize the changes using other inputs to the flow solver (eg. boundary conditions). We show how the adjoint counter-part of the density, the velocity fields and the turbulence quantities can be used to provide insights into the nature of changes the designer can induce to cause improvement in the performance metric of interest. We also discuss how to use adjoint solutions for problems with constraints to further refine the changes. Finally, we use a problem where it is not immediately apparent what geometry changes need to be used for further evaluation with optimization algorithms. In this problem, we use the adjoint and flow solution on a turbine strut, to determine the kind of end-wall treatments that reduce the loss. These changes are then implemented to show that the loss is reduced by close to 8%.
Keywords: Design optimization, Flow dynamics, Fan, Compressor, Turbine, Endwall profile
Keywords: Design optimization, Flow dynamics, Fan, Compressor, Turbine, Endwall profile
Shankaran, S.; Marta, A. C.
Robust optimization for aerodynamic problems using polynomial chaos and adjoints Proceedings Article
In: Proceedings of the ASME Turbo Expo 2012: Power for Land, Sea and Air, pp. 2217-2227, ASME International Gas Turbine Institute Copenhagen, Denmark, 2012, ISBN: 978-0-7918-4474-8.
@inproceedings{Shankaran:2012b:IGTI,
title = {Robust optimization for aerodynamic problems using polynomial chaos and adjoints},
author = {S. Shankaran and A. C. Marta},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Shankaran_2012b_IGTI.pdf, Full-text PDF
https://asmedigitalcollection.asme.org/GT/proceedings-abstract/GT2012/44748/2217/365430},
doi = {10.1115/GT2012-69580},
isbn = {978-0-7918-4474-8},
year = {2012},
date = {2012-06-01},
booktitle = {Proceedings of the ASME Turbo Expo 2012: Power for Land, Sea and Air},
pages = {2217-2227},
address = {Copenhagen, Denmark},
organization = {ASME International Gas Turbine Institute},
series = {GT2012-69580},
abstract = {The central theme of this paper is to show how one can combine Polynomial Chaos Expansions (PCE) and adjoint theory to efficiently obtain sensitivities for robust optimal control and optimization. A non-intrusive PCE method is used to analyze the constraint equations for the state (which depends on uncertain inputs), namely the governing equations of the dynamical system. Adjoint solutions are constructed for each of the polynomial basis functions used in the approximate expansion. The combination of the gradient for each basis-adjoint pair is used to form the overall gradient. The resulting gradient can be used to improve an initial guess in an iterative optimization procedure. The repeated use of the non-intrusive PCE method, the adjoint solver and the gradient estimate can be used to determine optimal control laws for the governing system in the presence of uncertainties. The formulation of the optimal control problem is presented in the context of the flow equations where the expected value of a functional is to be minimized. The boundary shape is the control. Using an analytical problem to determine the trade-off between cost and accuracy of some PCE methods, the optimization algorithm is applied to an airfoil optimization problem in external flow. The optimal solutions are compared against a multi-point design approach and shown to result in better designs in the constrained and unconstrained case. Finally, the approach is used to reduce the mean of Loss of a low pressure turbine blade. The associated cost of this approach in an optimization setting is equal to the cost of a PCE analysis (≈ Q deterministic simulations) plus Q (number of unknowns in the PCE expansion) adjoint solves for each iteration of a steepest-descent algorithm. However, this cost can be further reduced for certain objective functions using an intrusive formulation for the adjoint equations.
Keywords: Sensitivity analysis, Polynomials chaos expansion, Adjoint method, Airfoil optimization, Multi-point design},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
The central theme of this paper is to show how one can combine Polynomial Chaos Expansions (PCE) and adjoint theory to efficiently obtain sensitivities for robust optimal control and optimization. A non-intrusive PCE method is used to analyze the constraint equations for the state (which depends on uncertain inputs), namely the governing equations of the dynamical system. Adjoint solutions are constructed for each of the polynomial basis functions used in the approximate expansion. The combination of the gradient for each basis-adjoint pair is used to form the overall gradient. The resulting gradient can be used to improve an initial guess in an iterative optimization procedure. The repeated use of the non-intrusive PCE method, the adjoint solver and the gradient estimate can be used to determine optimal control laws for the governing system in the presence of uncertainties. The formulation of the optimal control problem is presented in the context of the flow equations where the expected value of a functional is to be minimized. The boundary shape is the control. Using an analytical problem to determine the trade-off between cost and accuracy of some PCE methods, the optimization algorithm is applied to an airfoil optimization problem in external flow. The optimal solutions are compared against a multi-point design approach and shown to result in better designs in the constrained and unconstrained case. Finally, the approach is used to reduce the mean of Loss of a low pressure turbine blade. The associated cost of this approach in an optimization setting is equal to the cost of a PCE analysis (≈ Q deterministic simulations) plus Q (number of unknowns in the PCE expansion) adjoint solves for each iteration of a steepest-descent algorithm. However, this cost can be further reduced for certain objective functions using an intrusive formulation for the adjoint equations.
Keywords: Sensitivity analysis, Polynomials chaos expansion, Adjoint method, Airfoil optimization, Multi-point design
Keywords: Sensitivity analysis, Polynomials chaos expansion, Adjoint method, Airfoil optimization, Multi-point design
Marta, A. C.; Cadete, B. J. P.
Aero-structural optimization of sailplane wings Proceedings Article
In: Proceedings of the 2nd LAETA Young Researchers Meeting, IDMEC/FEUP Porto, Portugal, 2012.
@inproceedings{Marta:2012:LAETA,
title = {Aero-structural optimization of sailplane wings},
author = {A. C. Marta and B. J. P. Cadete},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Marta_2012_LAETA.pdf, Full-text PDF},
year = {2012},
date = {2012-04-01},
booktitle = {Proceedings of the 2nd LAETA Young Researchers Meeting},
address = {Porto, Portugal},
organization = {IDMEC/FEUP},
series = {ID 14},
abstract = {This paper presents a framework for the multi-disciplinary design analysis and optimization of sailplane wings. The approach used in the multi-disciplinary optimization framework uses a multi-disciplinary feasible architecture. The geometric parametrization method employed follows a free-form deformation method. To solve the aero-structural problem, a panel method coupled with a finite-element solver is implemented. The coupled non-linear system is solved using an approximate Newton-Krylov approach. The optimization algorithm uses sequential quadratic programming, where the gradients are evaluated using the adjoint method. A real sailplane wing, based on the LET L-23 Super Blanı́k from the Portuguese Air Force, is used as test case. Single disciplinary analyses assess the capabilities of the disciplinary modules of the framework. Results are presented for a drag minimization problem using aerodynamic and multi-disciplinary optimizations. They reveal important trade-offs between disciplinary optimum and multi-disciplinary optimum at the preliminary design stage.
Keywords: Multi-disciplinary optimization, Adjoint method, Coupled analysis, Free-form deformation method, Panel method, Finite-element method},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
This paper presents a framework for the multi-disciplinary design analysis and optimization of sailplane wings. The approach used in the multi-disciplinary optimization framework uses a multi-disciplinary feasible architecture. The geometric parametrization method employed follows a free-form deformation method. To solve the aero-structural problem, a panel method coupled with a finite-element solver is implemented. The coupled non-linear system is solved using an approximate Newton-Krylov approach. The optimization algorithm uses sequential quadratic programming, where the gradients are evaluated using the adjoint method. A real sailplane wing, based on the LET L-23 Super Blanı́k from the Portuguese Air Force, is used as test case. Single disciplinary analyses assess the capabilities of the disciplinary modules of the framework. Results are presented for a drag minimization problem using aerodynamic and multi-disciplinary optimizations. They reveal important trade-offs between disciplinary optimum and multi-disciplinary optimum at the preliminary design stage.
Keywords: Multi-disciplinary optimization, Adjoint method, Coupled analysis, Free-form deformation method, Panel method, Finite-element method
Keywords: Multi-disciplinary optimization, Adjoint method, Coupled analysis, Free-form deformation method, Panel method, Finite-element method
Marta, A. C.; Shankaran, S.; Stein, A.
Blade shape optimization using RANS discrete adjoint solver Proceedings Article
In: Proceedings of the 1st LAETA Young Researchers Meeting, IDMEC, Instituto Superior Técnico Lisboa, Portugal, 2010.
@inproceedings{Marta:2010:LAETA,
title = {Blade shape optimization using RANS discrete adjoint solver},
author = {A. C. Marta and S. Shankaran and A. Stein},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Marta_2010_LAETA.pdf, Full-text PDF},
year = {2010},
date = {2010-11-01},
booktitle = {Proceedings of the 1st LAETA Young Researchers Meeting},
address = {Lisboa, Portugal},
organization = {IDMEC, Instituto Superior Técnico},
series = {ID 24},
abstract = {Recent developments in numerical tools for turbomachinery design have made practical the use of gradient-based optimization using high-fidelity computational fluid dynamic (CFD) simulations. Such has been made possible with the use of adjoint solvers, that can efficiently provide the gradients of the functions of interest with respect to the design variables required by the optimizer, at a cost almost independent of the number of variables. The derivation and implementation of the discrete adjoint solver for a legacy Reynolds Average Navier–Stokes (RANS) CFD solver are briefly explained. The adjoint-based gradients of some functions of interest, such as mass flow, pressure ratio and efficiency, with respect to shape parameters are computed and benchmarked against finite-difference approximations and excellent agreement is demonstrated. The outline of the integration of such adjoint tool in an engineering design framework is presented and discussed. The adjoint-based design framework is tested on a shape optimization problem using a set of Hicks-Henne bump functions superimposed on the baseline shape as design variables. A simple design problem is presented: a compressor rotor blade passage is setup as an unconstrained maximization problem, where the efficiency is increased by tweaking the camberline angle distribution.
Keywords: Adjoint, Design, Optimization, Blade, Turbomachinery},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Recent developments in numerical tools for turbomachinery design have made practical the use of gradient-based optimization using high-fidelity computational fluid dynamic (CFD) simulations. Such has been made possible with the use of adjoint solvers, that can efficiently provide the gradients of the functions of interest with respect to the design variables required by the optimizer, at a cost almost independent of the number of variables. The derivation and implementation of the discrete adjoint solver for a legacy Reynolds Average Navier–Stokes (RANS) CFD solver are briefly explained. The adjoint-based gradients of some functions of interest, such as mass flow, pressure ratio and efficiency, with respect to shape parameters are computed and benchmarked against finite-difference approximations and excellent agreement is demonstrated. The outline of the integration of such adjoint tool in an engineering design framework is presented and discussed. The adjoint-based design framework is tested on a shape optimization problem using a set of Hicks-Henne bump functions superimposed on the baseline shape as design variables. A simple design problem is presented: a compressor rotor blade passage is setup as an unconstrained maximization problem, where the efficiency is increased by tweaking the camberline angle distribution.
Keywords: Adjoint, Design, Optimization, Blade, Turbomachinery
Keywords: Adjoint, Design, Optimization, Blade, Turbomachinery
Marta, A. C.; Shankaran, S.; Stein, A.
Blade shape optimization using RANS discrete adjoint solvers Proceedings Article
In: Proceedings of the EngOpt 2010 - 2nd International Conference on Engineering Optimization, IDMEC, Instituto Superior Técnico Lisbon, Portugal, 2010.
@inproceedings{Marta:2010:ENGOPT,
title = {Blade shape optimization using RANS discrete adjoint solvers},
author = {A. C. Marta and S. Shankaran and A. Stein},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Marta_2010_ENGOPT.pdf, Full-text PDF
http://www1.dem.ist.utl.pt/engopt2010/},
year = {2010},
date = {2010-09-01},
booktitle = {Proceedings of the EngOpt 2010 - 2nd International Conference on Engineering Optimization},
address = {Lisbon, Portugal},
organization = {IDMEC, Instituto Superior Técnico},
series = {ENGOPT2010-1410},
abstract = {Recent developments in numerical tools for turbomachinery design have made practical the use of gradient-based optimization using high-fidelity computational fluid dynamic (CFD) simulations. Such has been made possible with the use of adjoint solvers, that can efficiently provide the gradients of the functions of interest with respect to the design variables required by the optimizer, at a cost almost independent of the number of variables. The derivation and implementation of the discrete adjoint solver for a legacy Reynolds Average Navier–Stokes (RANS) CFD solver are briefly explained. The adjoint-based gradients of some functions of interest, such as mass flow, pressure ratio and efficiency, with respect to shape parameters are computed and benchmarked against finite-difference approximations and excellent agreement is demonstrated. The outline of the integration of such adjoint tool in an engineering design framework is presented and discussed. The adjoint-based design framework is tested on a shape optimization problem using a set of Hicks-Henne bump functions superimposed on the baseline shape as design variables. A simple design problem is presented: a compressor rotor blade passage is setup as an unconstrained maximization problem, where the efficiency is increased by tweaking the camberline angle distribution.
Keywords: Adjoint, Design, Optimization, Blade, Turbomachinery},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Recent developments in numerical tools for turbomachinery design have made practical the use of gradient-based optimization using high-fidelity computational fluid dynamic (CFD) simulations. Such has been made possible with the use of adjoint solvers, that can efficiently provide the gradients of the functions of interest with respect to the design variables required by the optimizer, at a cost almost independent of the number of variables. The derivation and implementation of the discrete adjoint solver for a legacy Reynolds Average Navier–Stokes (RANS) CFD solver are briefly explained. The adjoint-based gradients of some functions of interest, such as mass flow, pressure ratio and efficiency, with respect to shape parameters are computed and benchmarked against finite-difference approximations and excellent agreement is demonstrated. The outline of the integration of such adjoint tool in an engineering design framework is presented and discussed. The adjoint-based design framework is tested on a shape optimization problem using a set of Hicks-Henne bump functions superimposed on the baseline shape as design variables. A simple design problem is presented: a compressor rotor blade passage is setup as an unconstrained maximization problem, where the efficiency is increased by tweaking the camberline angle distribution.
Keywords: Adjoint, Design, Optimization, Blade, Turbomachinery
Keywords: Adjoint, Design, Optimization, Blade, Turbomachinery
Marta, A. C.; Shankaran, S.; Holmes, D. G.; Stein, A.
Development of adjoint solvers for engineering gradient-based turbomachinery design applications Proceedings Article
In: Proceedings of the ASME Turbo Expo 2009: Power for Land, Sea and Air, pp. 385-397, ASME International Gas Turbine Institute Orlando, USA, 2009, ISBN: 978-0-7918-4888-3.
@inproceedings{Marta:2009:IGTI,
title = {Development of adjoint solvers for engineering gradient-based turbomachinery design applications},
author = {A. C. Marta and S. Shankaran and D. G. Holmes and A. Stein},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Marta_2009_IGTI.pdf, Full-text PDF
https://asmedigitalcollection.asme.org/GT/proceedings-abstract/GT2009/48883/385/344458},
doi = {10.1115/GT2009-59297},
isbn = {978-0-7918-4888-3},
year = {2009},
date = {2009-06-01},
booktitle = {Proceedings of the ASME Turbo Expo 2009: Power for Land, Sea and Air},
pages = {385-397},
address = {Orlando, USA},
organization = {ASME International Gas Turbine Institute},
series = {GT2009-59297},
abstract = {High-fidelity computational fluid dynamics (CFD) are common practice in turbomachinery design. Typically, several cases are run with manually modified parameters based on designer expertise to fine-tune a machine. Although successful, a more efficient process is desired. Choosing a gradient-based optimization approach, the gradients of the functions of interest need to be estimated. When the number of variables greatly exceeds the number of functions, the adjoint method is the best-suited approach to efficiently estimate gradients. Until recently, the development of CFD adjoint solvers was regarded as complex and difficult, which limited their use mostly to academia. This paper focuses on the problem of developing adjoint solvers for legacy industrial CFD solvers. A discrete adjoint solver is derived with the aid of an automatic differentiation tool that is selectively applied to the CFD code that handles the residual and function evaluations. The adjoint-based gradients are validated against finite-difference and complex-step derivative approximations.
Keywords: Sensitivity analysis, Design optimization, Computational fluid dynamics, Adjoint method, Finite-differences},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
High-fidelity computational fluid dynamics (CFD) are common practice in turbomachinery design. Typically, several cases are run with manually modified parameters based on designer expertise to fine-tune a machine. Although successful, a more efficient process is desired. Choosing a gradient-based optimization approach, the gradients of the functions of interest need to be estimated. When the number of variables greatly exceeds the number of functions, the adjoint method is the best-suited approach to efficiently estimate gradients. Until recently, the development of CFD adjoint solvers was regarded as complex and difficult, which limited their use mostly to academia. This paper focuses on the problem of developing adjoint solvers for legacy industrial CFD solvers. A discrete adjoint solver is derived with the aid of an automatic differentiation tool that is selectively applied to the CFD code that handles the residual and function evaluations. The adjoint-based gradients are validated against finite-difference and complex-step derivative approximations.
Keywords: Sensitivity analysis, Design optimization, Computational fluid dynamics, Adjoint method, Finite-differences
Keywords: Sensitivity analysis, Design optimization, Computational fluid dynamics, Adjoint method, Finite-differences
Mader, C. A.; Martins, J. R. R. A.; Marta, A. C.
Towards aircraft design using an automatic discrete adjoint solver Proceedings Article
In: Proceedings of the 18th AIAA Computational Fluid Dynamics Conference, American Institute of Aeronautics and Astronautics Miami, USA, 2007, ISBN: 978-1-62410-129-8.
@inproceedings{Mader:2007:AIAA,
title = {Towards aircraft design using an automatic discrete adjoint solver},
author = {C. A. Mader and J. R. R. A. Martins and A. C. Marta},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Mader_2007_AIAA.pdf, Full-text PDF
https://arc.aiaa.org/doi/10.2514/6.2007-3953},
doi = {10.2514/6.2007-3953},
isbn = {978-1-62410-129-8},
year = {2007},
date = {2007-06-01},
booktitle = {Proceedings of the 18th AIAA Computational Fluid Dynamics Conference},
address = {Miami, USA},
organization = {American Institute of Aeronautics and Astronautics},
series = {AIAA-2007-3953},
abstract = {The ADjoint method is applied to a three-dimensional Computational Fluid Dynamics (CFD) solver to generate the sensitivities required for aerodynamic shape optimization. The ADjoint approach selectively uses Automatic Differentiation (AD) to generate the partial derivative terms in the discrete adjoint equations. By selectively applying AD techniques, the computational cost and memory overhead incurred by using AD are significantly reduced, while still allowing for the the accurate treatment of arbitrarily complex governing equations and boundary conditions. Once formulated, the discrete adjoint equations are solved using the Portable, Extensible Toolkit for Scientific computation (PETSc). With this approach, the computed adjoint vector can be used to calculate the total sensitivities required for aerodynamic shape optimization of a complete aircraft configuration. The resulting sensitivities are compared with complex-step derivatives to establish their accuracy. The tools developed are applied to an infinite wing test case to demonstrate the accuracy and efficiency of the method.
Keywords: Aerodynamic shape optimization, Computational fluid dynamics, Aircraft design, Reynolds-averaged Navier-Stokes, Euler equations, Sensitivity analysis},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
The ADjoint method is applied to a three-dimensional Computational Fluid Dynamics (CFD) solver to generate the sensitivities required for aerodynamic shape optimization. The ADjoint approach selectively uses Automatic Differentiation (AD) to generate the partial derivative terms in the discrete adjoint equations. By selectively applying AD techniques, the computational cost and memory overhead incurred by using AD are significantly reduced, while still allowing for the the accurate treatment of arbitrarily complex governing equations and boundary conditions. Once formulated, the discrete adjoint equations are solved using the Portable, Extensible Toolkit for Scientific computation (PETSc). With this approach, the computed adjoint vector can be used to calculate the total sensitivities required for aerodynamic shape optimization of a complete aircraft configuration. The resulting sensitivities are compared with complex-step derivatives to establish their accuracy. The tools developed are applied to an infinite wing test case to demonstrate the accuracy and efficiency of the method.
Keywords: Aerodynamic shape optimization, Computational fluid dynamics, Aircraft design, Reynolds-averaged Navier-Stokes, Euler equations, Sensitivity analysis
Keywords: Aerodynamic shape optimization, Computational fluid dynamics, Aircraft design, Reynolds-averaged Navier-Stokes, Euler equations, Sensitivity analysis
Mader, C. A.; Marta, A. C.; Martins, J. R. R. A.
Aerodynamic shape optimization of an oblique wing using the ADjoint approach Proceedings Article
In: Proceedings of the 15th Conference of the CFD Society of Canada, CFD Society of Canada Toronto, Canada, 2007.
@inproceedings{Mader:2007:CFD,
title = {Aerodynamic shape optimization of an oblique wing using the ADjoint approach},
author = {C. A. Mader and A. C. Marta and J. R. R. A. Martins},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Mader_2007_CFD.pdf, Full-text PDF
http://www.cfdcanada.ca/},
year = {2007},
date = {2007-05-01},
booktitle = {Proceedings of the 15th Conference of the CFD Society of Canada},
address = {Toronto, Canada},
organization = {CFD Society of Canada},
abstract = {The ADjoint method is applied to a three-dimensional Computational Fluid Dynamics (CFD) solver to gen-
erate the sensitivities required for aerodynamic shape optimization. The ADjoint approach selectively uses Automatic Differentiation (AD) to generate the partial derivative terms in the discrete adjoint equations. By selectively applying AD techniques, the computational cost and memory overhead incurred by using AD are significantly reduced, while still allowing for
the accurate treatment of arbitrarily complex governing equations and boundary conditions. Once formulated, the discrete adjoint equations are solved using the Portable Extensible Toolkit for Scientific computation (PETSc). With this approach, the adjoint vector is used to compute the total sensitivities required for aerodynamic shape optimization of a complete aircraft configuration. The resulting sensitivities are compared with finite difference derivatives to verify accuracy.
Keywords: Adjoint method, Automatic differentiation, Computational fluid dynamics, Aircraft design, Sensitivity analysis},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
The ADjoint method is applied to a three-dimensional Computational Fluid Dynamics (CFD) solver to gen-
erate the sensitivities required for aerodynamic shape optimization. The ADjoint approach selectively uses Automatic Differentiation (AD) to generate the partial derivative terms in the discrete adjoint equations. By selectively applying AD techniques, the computational cost and memory overhead incurred by using AD are significantly reduced, while still allowing for
the accurate treatment of arbitrarily complex governing equations and boundary conditions. Once formulated, the discrete adjoint equations are solved using the Portable Extensible Toolkit for Scientific computation (PETSc). With this approach, the adjoint vector is used to compute the total sensitivities required for aerodynamic shape optimization of a complete aircraft configuration. The resulting sensitivities are compared with finite difference derivatives to verify accuracy.
Keywords: Adjoint method, Automatic differentiation, Computational fluid dynamics, Aircraft design, Sensitivity analysis
erate the sensitivities required for aerodynamic shape optimization. The ADjoint approach selectively uses Automatic Differentiation (AD) to generate the partial derivative terms in the discrete adjoint equations. By selectively applying AD techniques, the computational cost and memory overhead incurred by using AD are significantly reduced, while still allowing for
the accurate treatment of arbitrarily complex governing equations and boundary conditions. Once formulated, the discrete adjoint equations are solved using the Portable Extensible Toolkit for Scientific computation (PETSc). With this approach, the adjoint vector is used to compute the total sensitivities required for aerodynamic shape optimization of a complete aircraft configuration. The resulting sensitivities are compared with finite difference derivatives to verify accuracy.
Keywords: Adjoint method, Automatic differentiation, Computational fluid dynamics, Aircraft design, Sensitivity analysis
Mader, C. A.; Martins, J. R. R. A.; Marta, A. C.
Towards aerodynamic shape optimization of an oblique wing using the ADjoint approach Proceedings Article
In: Proceedings of the CASI Aero 2007 Conference, Canadian Aeronautics and Space Institute Toronto, Canada, 2007.
@inproceedings{Mader:2007:CASI,
title = {Towards aerodynamic shape optimization of an oblique wing using the ADjoint approach},
author = {C. A. Mader and J. R. R. A. Martins and A. C. Marta},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Mader_2007_CASI.pdf, Full-text PDF
http://www.casi.ca/},
year = {2007},
date = {2007-04-01},
booktitle = {Proceedings of the CASI Aero 2007 Conference},
address = {Toronto, Canada},
organization = {Canadian Aeronautics and Space Institute},
abstract = {The ADjoint method is applied to a three-dimensional Computational Fluid Dynamics (CFD) solver to generate the sensitivities required for aerodynamic shape optimization. The ADjoint approach selectively uses Automatic Differentiation (AD) to generate the partial derivative terms in the discrete adjoint equations. By selectively applying AD techniques, the computational cost and memory overhead incurred by using AD are significantly reduced, while still allowing for the the accurate treatment of arbitrarily complex governing equations and boundary conditions. Once formulated, the discrete adjoint equations are solved using the Portable, Extensible Toolkit for Scientific computation (PETSc). With this approach, the computed adjoint vector can be used to calculate the total sensitivities required for aerodynamic shape optimization of a complete aircraft configuration. The resulting sensitivities are compared with finite-difference derivatives to verify accuracy. Once formulated, these derivatives can then be used to perform gradient based aerodynamic optimization.
Keywords: Adjoint method, Automatic differentiation, Computational fluid dynamics, Aircraft design, Sensitivity analysis},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
The ADjoint method is applied to a three-dimensional Computational Fluid Dynamics (CFD) solver to generate the sensitivities required for aerodynamic shape optimization. The ADjoint approach selectively uses Automatic Differentiation (AD) to generate the partial derivative terms in the discrete adjoint equations. By selectively applying AD techniques, the computational cost and memory overhead incurred by using AD are significantly reduced, while still allowing for the the accurate treatment of arbitrarily complex governing equations and boundary conditions. Once formulated, the discrete adjoint equations are solved using the Portable, Extensible Toolkit for Scientific computation (PETSc). With this approach, the computed adjoint vector can be used to calculate the total sensitivities required for aerodynamic shape optimization of a complete aircraft configuration. The resulting sensitivities are compared with finite-difference derivatives to verify accuracy. Once formulated, these derivatives can then be used to perform gradient based aerodynamic optimization.
Keywords: Adjoint method, Automatic differentiation, Computational fluid dynamics, Aircraft design, Sensitivity analysis
Keywords: Adjoint method, Automatic differentiation, Computational fluid dynamics, Aircraft design, Sensitivity analysis
Marta, A. C.; Alonso, J. J.
High-speed MHD flow control using adjoint-based sensitivities Proceedings Article
In: Proceedings of the 14th AIAA/AHI Space Planes and Hypersonic Systems and Technologies Conference, American Institute of Aeronautics and Astronautics Canberra, Australia, 2006, ISBN: 978-1-62410-050-5.
@inproceedings{Marta:2006c:AIAA,
title = {High-speed MHD flow control using adjoint-based sensitivities},
author = {A. C. Marta and J. J. Alonso},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Marta_2006c_AIAA.pdf, Full-text PDF
https://arc.aiaa.org/doi/10.2514/6.2006-8009},
doi = {10.2514/6.2006-8009},
isbn = {978-1-62410-050-5},
year = {2006},
date = {2006-11-01},
booktitle = {Proceedings of the 14th AIAA/AHI Space Planes and Hypersonic Systems and Technologies Conference},
address = {Canberra, Australia},
organization = {American Institute of Aeronautics and Astronautics},
series = {AIAA-2006-8009},
abstract = {As advances in integrated hypersonic flight vehicles materialize, it is worth devoting some effort to optimize their design. The plasma generated at high Mach numbers permits the use of electromagnetic actuators to control the flow. One can view this situation as an optimization problem in which a cost function is to be minimized by varying a set of control variables, while satisfying a specific set of constraints. The flow under the presence of a magnetic field is typically modeled with the equations of magneto-hydrodynamics (MHD) and the scientific community has already devoted much effort into their analysis. The focus of this work is to tackle the problem of efficiently estimating the sensitivity of cost functions and/or constraints with the use potential of a gradient-based optimizer in mind. This paper outlines the implementation of a scalable tool to compute sensitivities of cost functions that arise in such optimization problems. The sensitivities are computed using the discrete adjoint method that has already been successfully tested by the authors for this class of problems. The derivation of the discrete adjoint equations is done with the help of Tapenade, an automatic differentiation (AD) tool, that only requires some minor re- writing of the original flow solver code but avoids the tedious hand-differentiation approach. This strategy has proved to be much faster to implement and much less error-prone than a hand-differentiation approach and the resulting sensitivities are in exact agreement with the discrete sensitivities that would be produced by exact finite-differences of the original code. The linear adjoint system of equations obtained is solved using the Portable, Extensible Toolkit for Scientific computation (PETSc). The low magnetic Reynolds number MHD governing equations are used to model the flow but this methodology can be transparently applied to any set of governing equations, cost functions and design variables. Sensitivities of aerodynamic coefficients such as lift, drag and moment, with respect to the electrical conductivity of the flow are computed in a multi-block grid using this approach and their values are verified, totaling more than 180,000 design variables.
Keywords: Electromagnetic control, Hypersonic flight actuators, Control theory, Adjoint method, Automatic differentiation},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
As advances in integrated hypersonic flight vehicles materialize, it is worth devoting some effort to optimize their design. The plasma generated at high Mach numbers permits the use of electromagnetic actuators to control the flow. One can view this situation as an optimization problem in which a cost function is to be minimized by varying a set of control variables, while satisfying a specific set of constraints. The flow under the presence of a magnetic field is typically modeled with the equations of magneto-hydrodynamics (MHD) and the scientific community has already devoted much effort into their analysis. The focus of this work is to tackle the problem of efficiently estimating the sensitivity of cost functions and/or constraints with the use potential of a gradient-based optimizer in mind. This paper outlines the implementation of a scalable tool to compute sensitivities of cost functions that arise in such optimization problems. The sensitivities are computed using the discrete adjoint method that has already been successfully tested by the authors for this class of problems. The derivation of the discrete adjoint equations is done with the help of Tapenade, an automatic differentiation (AD) tool, that only requires some minor re- writing of the original flow solver code but avoids the tedious hand-differentiation approach. This strategy has proved to be much faster to implement and much less error-prone than a hand-differentiation approach and the resulting sensitivities are in exact agreement with the discrete sensitivities that would be produced by exact finite-differences of the original code. The linear adjoint system of equations obtained is solved using the Portable, Extensible Toolkit for Scientific computation (PETSc). The low magnetic Reynolds number MHD governing equations are used to model the flow but this methodology can be transparently applied to any set of governing equations, cost functions and design variables. Sensitivities of aerodynamic coefficients such as lift, drag and moment, with respect to the electrical conductivity of the flow are computed in a multi-block grid using this approach and their values are verified, totaling more than 180,000 design variables.
Keywords: Electromagnetic control, Hypersonic flight actuators, Control theory, Adjoint method, Automatic differentiation
Keywords: Electromagnetic control, Hypersonic flight actuators, Control theory, Adjoint method, Automatic differentiation
Marta, A. C.; Alonso, J. J.
Discrete adjoint formulation for the ideal MHD equations Proceedings Article
In: Proceedings of the 3rd AIAA Flow Control Conference, American Institute of Aeronautics and Astronautics San Francisco, USA, 2006, ISBN: 978-1-62410-036-9.
@inproceedings{Marta:2006b:AIAA,
title = {Discrete adjoint formulation for the ideal MHD equations},
author = {A. C. Marta and J. J. Alonso},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Marta_2006b_AIAA.pdf, Full-text PDF
https://arc.aiaa.org/doi/10.2514/6.2006-3345},
doi = {10.2514/6.2006-3345},
isbn = {978-1-62410-036-9},
year = {2006},
date = {2006-06-01},
booktitle = {Proceedings of the 3rd AIAA Flow Control Conference},
address = {San Francisco, USA},
organization = {American Institute of Aeronautics and Astronautics},
series = {AIAA-2006-3345},
abstract = {For over twenty years there has been computational work to analyze and control hypersonic flows using electromagnetic effects but no true effort has been pursued to automate the flow control process. The lack of a design framework that provides automated multi-disciplinary optimization (MDO) capabilities for this class of problems is the principal motivation for this work. This paper extends the foundation of a MDO component previously developed by the authors to the ideal MHD equations that govern the three-dimensional flow of an inviscid compressible perfectly conducting fluid with an externally imposed magnetic field. The gas is assumed perfect and chemically frozen, as the focus is on the discrete adjoint derivations rather than on the flow solver. Control theory has already been proved successful dealing with both aerodynamic shape and magnetohydrodynamics (MHD) optimization using an inviscid low magnetic Reynolds number model by the authors. The discrete adjoint is the best suitable option to deal with the complex equations that govern MHD, and with the nature of the cost functions that may be used for relevant design
problems. At this point, the derivation of the adjoint system of equations is done by hand differentiation of the flow solver. The sensitivities computed using the discrete adjoint formulation are matched against values obtained using finite-differences and a sample design problem is presented. On going work includes the incorporation of non-ideal effects in the governing equations, parallelization of both the flow and the adjoint solvers and the use of automatic differentiation tools to effortlessly compute the adjoint system of equations from the coded flow solver. Once all this is accomplished, the investigation of meaningful design problems and the definition of significant cost functions will finally be tackled.
Keywords: Electromagnetic flow control, Hypersonic flight actuators, Control theory, Adjoint method, Automatic differentiation},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
For over twenty years there has been computational work to analyze and control hypersonic flows using electromagnetic effects but no true effort has been pursued to automate the flow control process. The lack of a design framework that provides automated multi-disciplinary optimization (MDO) capabilities for this class of problems is the principal motivation for this work. This paper extends the foundation of a MDO component previously developed by the authors to the ideal MHD equations that govern the three-dimensional flow of an inviscid compressible perfectly conducting fluid with an externally imposed magnetic field. The gas is assumed perfect and chemically frozen, as the focus is on the discrete adjoint derivations rather than on the flow solver. Control theory has already been proved successful dealing with both aerodynamic shape and magnetohydrodynamics (MHD) optimization using an inviscid low magnetic Reynolds number model by the authors. The discrete adjoint is the best suitable option to deal with the complex equations that govern MHD, and with the nature of the cost functions that may be used for relevant design
problems. At this point, the derivation of the adjoint system of equations is done by hand differentiation of the flow solver. The sensitivities computed using the discrete adjoint formulation are matched against values obtained using finite-differences and a sample design problem is presented. On going work includes the incorporation of non-ideal effects in the governing equations, parallelization of both the flow and the adjoint solvers and the use of automatic differentiation tools to effortlessly compute the adjoint system of equations from the coded flow solver. Once all this is accomplished, the investigation of meaningful design problems and the definition of significant cost functions will finally be tackled.
Keywords: Electromagnetic flow control, Hypersonic flight actuators, Control theory, Adjoint method, Automatic differentiation
problems. At this point, the derivation of the adjoint system of equations is done by hand differentiation of the flow solver. The sensitivities computed using the discrete adjoint formulation are matched against values obtained using finite-differences and a sample design problem is presented. On going work includes the incorporation of non-ideal effects in the governing equations, parallelization of both the flow and the adjoint solvers and the use of automatic differentiation tools to effortlessly compute the adjoint system of equations from the coded flow solver. Once all this is accomplished, the investigation of meaningful design problems and the definition of significant cost functions will finally be tackled.
Keywords: Electromagnetic flow control, Hypersonic flight actuators, Control theory, Adjoint method, Automatic differentiation
Marta, A. C.; Alonso, J. J.; Tang, L.
Automatic magnetohydrodynamic control of hypersonic flow using a discrete adjoint formulation Proceedings Article
In: Proceedings of the 44th AIAA Aerospace Sciences Meeting and Exhibit, American Institute of Aeronautics and Astronautics Reno, USA, 2006, ISBN: 978-1-62410-039-0.
@inproceedings{Marta:2006a:AIAA,
title = {Automatic magnetohydrodynamic control of hypersonic flow using a discrete adjoint formulation},
author = {A. C. Marta and J. J. Alonso and L. Tang},
url = {https://mdo.tecnico.ulisboa.pt/wp-content/uploads/publications_Inproceedings_Marta_2006a_AIAA.pdf, Full-text PDF
https://arc.aiaa.org/doi/10.2514/6.2006-370},
doi = {10.2514/6.2006-370},
isbn = {978-1-62410-039-0},
year = {2006},
date = {2006-01-01},
booktitle = {Proceedings of the 44th AIAA Aerospace Sciences Meeting and Exhibit},
address = {Reno, USA},
organization = {American Institute of Aeronautics and Astronautics},
series = {AIAA-2006-370},
abstract = {A significant amount of work in the analysis of high-speed magnetohydrodynamics (MHD) has been carried out during the past decade. However, the fact remains that very little effort has been devoted to design applications of these analysis tools. The main reasons for this lack of design focus have been that the analysis tools were in the process of maturing and that the cost of the simulations was sufficiently large that design applications were beyond the reach of existing computing resources. The absence of such a design framework that provides automated multi-disciplinary optimization (MDO) capabilities for this class of high-speed problems is the principal motivation for this work. This paper develops the foundation of one of the components of such an MDO environment. Control theory has already proved successful in dealing with both aerodynamic shape and aero-structural optimization problems and in this work it is extended to MHD. The discrete adjoint approach emerges as the best suitable option to deal with the complex equations that govern MHD and with the nature of the cost and constraint functions that may be used for relevant design problems. The equations governing the three-dimensional flow of a compressible conducting fluid in a magnetic field using the low magnetic Reynolds number approximation are used. At this stage, other simplifications are assumed, such as a frozen chemical state, so that the soundness of the basic derivations can be established. The details of the theory and implementation of the discrete adjoint solver for the MHD equations are presented in this paper. The gradients obtained using the discrete adjoint approach are validated against finite-difference approximations and shown to be very accurate. A demonstration of the design capabilities is also included with a simple design problem using several design variables and constraints.
Keywords: Electromagnetic flow control, Hypersonic flight actuators, Control theory, Adjoint method, Automatic differentiation},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
A significant amount of work in the analysis of high-speed magnetohydrodynamics (MHD) has been carried out during the past decade. However, the fact remains that very little effort has been devoted to design applications of these analysis tools. The main reasons for this lack of design focus have been that the analysis tools were in the process of maturing and that the cost of the simulations was sufficiently large that design applications were beyond the reach of existing computing resources. The absence of such a design framework that provides automated multi-disciplinary optimization (MDO) capabilities for this class of high-speed problems is the principal motivation for this work. This paper develops the foundation of one of the components of such an MDO environment. Control theory has already proved successful in dealing with both aerodynamic shape and aero-structural optimization problems and in this work it is extended to MHD. The discrete adjoint approach emerges as the best suitable option to deal with the complex equations that govern MHD and with the nature of the cost and constraint functions that may be used for relevant design problems. The equations governing the three-dimensional flow of a compressible conducting fluid in a magnetic field using the low magnetic Reynolds number approximation are used. At this stage, other simplifications are assumed, such as a frozen chemical state, so that the soundness of the basic derivations can be established. The details of the theory and implementation of the discrete adjoint solver for the MHD equations are presented in this paper. The gradients obtained using the discrete adjoint approach are validated against finite-difference approximations and shown to be very accurate. A demonstration of the design capabilities is also included with a simple design problem using several design variables and constraints.
Keywords: Electromagnetic flow control, Hypersonic flight actuators, Control theory, Adjoint method, Automatic differentiation
Keywords: Electromagnetic flow control, Hypersonic flight actuators, Control theory, Adjoint method, Automatic differentiation