大山 聖

© 2018, AIAA Aerospace Sciences Meeting. All rights reserved. Large-eddy simulations are conducted to investigate the control effects of a dielectric barrier discharge plasma actuator on a dynamic flowfield with a Reynolds number of 2.56 × 105 and a reduced frequency of 0.02π. The objective flowfields include dynamic stall phenomenon, flow separation and reattachment. First the flowfield without control is investigated and it is found that dynamic stall process can be classified into five stages; formation of a laminar separation bubble, breakdown of the laminar separation bubble which triggers formation of a dynamic stall vortex, convection of the dynamic stall vortex, full stall from the leading edge, and recovery to the attached state. Then the control effects with three burst frequencies (F +) of 0.5, 6, and 50 in nondimensionalized value are investigated. The DBD plasma actuator successfully enhances the cycle-integrated aerodynamic performances of the airfoil and major control effects are summarized into three; delay of dynamic stall, enhancement of aerodynamic forces during full stall by large vortices, and promotion of reattachment. The most effective burst frequency for each control effect differs from each other, showing that the best case for delaying the dynamic stall onset is the case with F + of 50 while the best case for promoting the reattachment is the case with F + of 6. The results show that the promoting the reattachment is effective for improving the cycle-integrated net damping and shortening the duration under the stall. For further improvement, the current results give a strong prospect of a closed-loop control in which F + is adapted to the change in the flowfield.

We investigate the properties of widely used constrained multi-objective optimization benchmark problems. A number of Multi-Objective Evolutionary Algorithms (MOEAs) for Constrained Multi-Objective Optimization Problems (CMOPs) have been proposed in the past few years. The C-DTLZ functions and Real-World-Like Problems (RWLPs) have frequently been used for evaluating the performance of MOEAs on CMOPs. In this paper, however, we show that the C-DTLZ functions and widely-used RWLPs have some unnatural problem features. The experimental results show that an MOEA without any Constraint Handling Techniques (CHTs) can successfully find well-approximated nondominated feasible solutions on the C1-DTLZ1, C1-DTLZ3, and C2-DTLZ2 functions. It is widely believed that RWLPs are MOEA-hard problems, and finding the feasible solutions on them is a very hard task. However, we show that the MOEA without any CHTs can find feasible solutions on widely-used RWLPs such as the speed reducer design problem, the two-bar truss design problem, and the water problem. Also, it is seldom that the infeasible solution simultaneously violates multiple constraints in the RWLPs. Due to the above reasons, we conclude that constrained multi-objective optimization benchmark problems need a careful reconsideration.

While a large number of multi-objective evolutionary algorithms (MOEAs) for many-objective optimization problems (MaOPs) have been proposed in the past few years, an exhaustive benchmarking study has never been performed. Moreover, most previous studies evaluated the performance of MOEAs based on nondominated solutions in the final population at the end of the search. In this paper, we exhaustively investigate the convergence performance of 21 MOEAs using an unbounded external archive that stores all nondominated solutions found during the search process. Surprisingly, the experimental results for the WFG functions with up to six objectives indicate that several recently proposed MOEAs perform significantly worse than classical MOEAs. Moreover, the performance rank among the 21 MOEAs significantly depends on the number of function evaluations. Thus, the previously reported performance of MOEAs on MaOPs as well as the widely used benchmarking methodology must be carefully reconsidered.

© 2017, American Institute of Aeronautics and Astronautics Inc, AIAA. All rights reserved. This study focuses on detailed structures of a separation bubble in controlled airfoil-flows using a DBD plasma actuator. Time-averaged surface pressure measurements and well-resolved PIV are conducted. The present PIV measurement enables the observation of the detailed flow structure near the leading edge by connecting three adjacent images of particle image velocimetry (PIV). The airfoil is NACA0015 and the Reynolds number based on the chord length is 63,000. The angle of attack is 12 deg. corresponding to the fully separated flow from the leading edge. Three types of actuation (the normal-mode case mode and burst modes with F+ = 1 and 6 ) are considered. The flow control mechanism related to a separation bubble is discussed for each case through time-averaged and phase-averaged flow fields.

We investigate the impact of three control parameters (the population size μ, the number of children λ, and the number of reference points H) on the performance of Nondominated Sorting Genetic Algorithm III (NSGA-III). In the past few years, many efficient Multi-Objective Evolutionary Algorithms (MOEAs) for Many-Objective Optimization Problems (MaOPs) have been proposed, but their control parameters have been poorly analyzed. The recently proposed NSGA-III is one of most promising MOEAs for MaOPs. It is widely believed that NSGA-III is almost parameter-less and requires setting only one control parameter (H), and the value of μ and λ can be set to μ = λ ≈ H as described in the original NSGA-III paper. However, the experimental results in this paper show that suitable parameter settings of μ, λ, and H values differ from each other as well as their widely used parameter settings. Also, the performance of NSGA-III significantly depends on them. Thus, the usually used parameter settings of NSGA-III (i.e., μ = λ ≈ H) might be unsuitable in many cases, and μ, λ, and H require a particular parameter tuning to realize the best performance of NSGA-III.

This paper experimentally investigates the effectiveness of a closed-loop flow control method using a DBD plasma actuator for a NACA0015 airfoil, in which the surface pressure fluctuation is fed back to the system; the actuator was driven when the pressure fluctuation exceeds the setup threshold. The Reynolds number based on the chord length is set to 63,000 and the angle of attack is in the range from 12 to 15 degrees. The actuator was installed on the surface at 5% of the chord length from the leading edge. The results show that the closed loop control worked better than the continuous operation. In the angle of attack of 12 and 14 degrees, the complete attached flow was attained by setting the appropriate threshold value of the pressure fluctuation. On the other hand, in 15 degrees, although the complete attached flow was not attained, the flow separation was partially suppressed.

© American Institute of Aeronautics and Astronautics Inc, AIAA. All rights reserved. The effects of disturbed boundary layer at the nozzle exit on acoustic waves of supersonic jets of the Mach and Reynolds numbers of 2.0 and 900,000, respectively, are investigated by large-eddy simulations. The high order compact schemes and sufficient grid points are used to solve the compressible Navier Stokes equations. The inflow is tripped by the method used in the previous study for a subsonic jet computation in which a random vortex is imposed inside the boundary layer of the nozzle. Two disturbed boundary layer cases (disturbed cases) with different disturbance strength and one laminar boundary layer case (laminar case) are investigated. The flow seems to be much disturbed by the tripping, and the slower growth of the shear layer thickness for the disturbed cases is observed than that for the laminar case. This slower growth for the disturbed case leads to its longer potential core length. The laminar case has stronger peaks inside the nozzle near the nozzle exit and it corresponds to the turbulent transition. With regard to the acoustic fields, the region where the most strong sound pressure level (SPL) is observed is the end of the potential core for the disturbed cases, while the laminar case have higher SPL around the transition region due to the strong Mach wave generation by the transition. The SPL of the laminar case is 5dB higher than disturbed cases at the far field, and the spectral of the laminar case is entirely higher than those of disturbed cases in the wide range of the frequency. Disturbance strength for disturbed case does not affect the flow and acoustic fields much in the range of disturbance strength we investigated.

© 2016, American Institute of Aeronautics and Astronautics Inc, AIAA. All rights reserved. This study explores multiobjective aeroacoustic designs for flame deflector of a launch pad. The acoustic characteristics associated with deflector shapes are identified by multi- objective evolutionary computation with large eddy simulations. The objective functions in the multiobjective aeroacoustic design are designed to minimize (1) the spatial-averaged sound pressure level near the payload fairing, (2) the time-averaged maximum pressure on the curved surface of the frame deflector, and (3) the deviation of the curved surface from the at plate inclined at 45° The multiobjective evolutionary computation requires 2500 large eddy simulations. Evaluation of each single configuration required 130 nodes (1040 total cores) of a “K” supercomputer and 6 hour calculation. As the result of optimization, 146 nondominated solutions are obtained. The analysis of nondominated solutions clearly reveals various trade-off relations and correlations among the objective functions. The flow field analysis shows that as the curved surface around the impingement region becomes steeper, weaker acoustic waves are generated in the impingement region. This trend is related to the size of the separation bubble near the impingement region; as the surface steepens, the bubble shrinks. Furthermore, analysis of the effect of nozzle-to-launchpad distance on nondominated solutions are conducted to extract more useful knowledge for rocket launch site design.

© American Institute of Aeronautics and Astronautics Inc, AIAA. All rights reserved. The convective Mach number dependences of sound sources in a compressible mixing layer are investigated by direct numerical simulation. Characteristics of sound sources are analyzed using the source terms of the Lighthill equation. The acoustic waves become weaker with increasing the Mach number due to the weaker vortices by compressibility and the canceling out of the Reynolds stress term and the entropy term. Also, the smaller scale acoustic waves appears for Mc ≥ 1.5. The results suggest that those change in the characteristics of the sound sources are due to the appearance of shocklets for larger convective Mach number cases. Also, those cases show the higher turbulent Mach number over the wide range of flow field.

Two- (2D) and three-dimensional (3D) simulation of flow over a flat plate at various low Reynolds numbers (Re) are conducted. The predictability of instantaneous flow fields in the 2D simulation and variation of the flow field characteristics with Re are discussed in this study. The results show that 2D simulations can predict qualitative characteristics of averaged features such as surface pressure and skin friction coefficients. Moreover, two types of critical Re are specified; separation Re (Re-s) and bubble-length Re (Re-bl). Detailed analysis for the averaged pressure distribution predictability are performed by deriving a Reynolds averaged pressure gradient equation and budgeting each term.

In this paper, we describe how discretizing design variables on real-coded genetic algorithms (RCGAs) can influence the convergence and the diversity of Pareto optimal solutions. We use Non-dominated Sorting Genetic Algorithm II (NSGA-II) as an RCGA based on Pareto dominance, changing the number of significant digits after the decimal point for each design variable. Test problems and engineering problems are investigated. Computational results show that the use of a smaller number of significant figures instead of larger ones achieves better convergence that a larger number in many cases. In the DTLZ3 test problem, low applied precision avoids dominance-resistant solutions (DRSs) and improves both the generational distance (GD) and the inverted generational distance (IGD). On the other hand, in the DTLZ4 test problem, low digit precision improves GD, whereas it worsens IGD. This indicates that a minimum digit precision is required to maintain the diversity of Pareto optimal solutions in some problems. When we use RCGAs, it is critical to set the number of significant digits after the decimal point to realistically represent actual engineering problems.

The advantages of evolutionary computation with very large populations for many-objective optimization problems are investigated. The effects of a population size of up to 1,000,000 are studied, with the number of generations fixed at 100. To overcome difficulty in computational time, we use a many-objective evolutionary algorithm designed for massive parallelization (CHEETAH) on the K supercomputer. For unimodal test problems DTLZ2 and DTLZ4, the inverted generational distance (IGD) decreases as the population increases while the generational distance (GD) is saturated with a population size of 10,000. This means an evolutionary computation with massive population size mainly contributes to improvement of diversity of obtained non-dominated solutions. Even when the total number of evaluations is fixed, this conclusion is unchanged. For the multimodal test problems DTLZ1 and DTLZ3, GD and IGD are reduced with increasing population size of up to 10,000 but are not significantly improved with population sizes larger than this. This is probably due to the difficulty in obtaining good non-dominated solutions for DTLZ1 and DTLZ3 with current CHEETAH. Because CHEETAH is bases on NSGA-II (only the non-dominated sort portion is modified for more effective many-objective optimization and parallelization), we expect that the current conclusion qualitatively stays the same for other NSGA-II-based algorithms. To take advantage of the larger population size, development of operators such as selection and crossover designed for very large population size may be required.

© 2016, American Institute of Aeronautics and Astronautics Inc, AIAA. All rights reserved. A folding wing is an effective deployment mechanism for the airplane that is used for Mars exploration. A spring loaded hinge is considered as a deployment actuator for Mars airplane in Japan. A hinge torque is one of the primitive design variables to control the aerial deployment behavior. The required hinge torque for deployment is directly concerned with the deployment mechanism mass. Since the Mars airplane requires thorough mass reduction, it is necessary to reduce the required hinge torque while keeping high robustness for the aerial deployment. This paper investigates the robustness of the aerial deployment behavior especially focused on the effect of the hinge axis tilting. The hinge axis of the non- tilted hinge axis design is defined to be parallel to the X-axis of the center body coordinates. The conditions to judge whether the deployment succeeded or failed are defined for the state of the airplane. The margins of the airplane state for the conditions are set to the evaluation functions of the safety. The robustness of the safety is evaluated using the sigma level where the sigma level is a function of the average and standard deviation of the evaluation functions. For the robustness evaluation, this study deals with four dispersive parameters: drop velocity, surrounding gust velocity, initial pitch angle, and height. The robustness of several tilted and non-tilted hinge axis designs are calculated and then compared. The result clearly shows that the tilted hinge axis design can deploy with lower torque than the torque of the non-tilted hinge axis design. The motions of the individual cases are then studied to reveal the effect of the hinge axis tilting. It is clarified that the tilted hinge axis design is able to set the angle of attack of the outer wing positive under the wide range of conditions. Therefore, the aerodynamic force assists the deployment. In the appropriate condition, the wings deploy without torque of the deployment actuator.

Copyright © 2016 by the International Astronautical Federation (IAF). All rights reserved. Mars is the next milestone in our exploration of the solar system. The presence of an atmosphere on Mars signifies that an airplane could travel in its atmosphere using the aerodynamic forces of flight. The airplane allows for a platform that can cover a larger area of exploration than is currently available. A reconnaissance airplane offers the possibility to obtain high-resolution data on a regional scale of several hundreds to thousands of kilometers, which cannot be achieved with rovers or satellites. There is an extremely high demand for the exploration of Mars using an airplane that can fly in its atmosphere. One of the big problems for a Mars Airplane is the very low atmospheric density on Mars. So, it is difficult to obtain the required lift, as the wing area required to generate enough lift is inversely proportional to the density. So in order to reduce the required lift, thorough weight reduction is needed. Even so, a Mars Airplane needs a large wing area, which leads to another problem. To transport to Mars, a Mars Airplane must be small and compact. As a way to solve this conflicting problem, the Mars Airplane needs some deployment mechanisms. Various hurdles, including those described above, must be overcome in order to realize the flight exploration of Mars and all of them require innovative technological solutions. Hence, the Mars Airplane Working Group was established in 2010 with the aim of conducting flight technology validations for the MELOS1 mission using a compact airplane in JAXA/ISAS. The working group aims to realize Mars exploration using an airplane for the first time ever. At present, the mission being considered for the Mars exploration plane is "flying over a range of about 100 km to capture ground surface images and observing high-resolution images of residual magnetic fields." Cameras and magnetic field observation equipment are mounted as payloads on the airplane, which is expected to fly over a range of 100 km at a speed of 60 m/s. In the conceptual design, the weight of the airplane is about 4.0 kg, the span length is about 2.5 m, and the total length is about 2.0 m. This paper provides a summary of the Mars airplane development being considered by the working group in Japan and discusses the technical issues addressed in order to realize a Mars airplane.