永田 靖典

Attitude control for small satellites is crucial to enable high value missions. Active attitude control is challenging for nanosatellites, due to their small mass and power budgets. On the other hand, the air in low Earth orbit is a promising resource for passive aero-stabilisation of a satellite's orientation. Potential has increased with the development of miniaturised deployable aeroshells for atmospheric entry. The EGG nanosatellite, released into space from the ISS in 2017, is one example of an aeroshell-equipped satellite with no means of active attitude control. Flight data from the EGG mission is a convenient resource to evaluate the concept of passive aero-stabilisation. In this work, a comprehensive coupled atmosphere-orbit-attitude simulation platform for small satellites was developed. The objectives are: (i) to validate the simulation against EGG mission data, and (ii) to evaluate the impact of attitude disturbances on robustness of passive aero-stabilisation. The results provide qualitative validation of the simulation platform, and show that passive attitude control with aeroshell deployed is highly sensitive to initial spin and spacecraft geometric asymmetry. These findings suggest the need for hybrid active control to turn aeroshell-equipped capsules into a viable means of trans-atmospheric transport.

Aerodynamic instability in the attitude of an inflatable re-entry vehicle in the subsonic regime has been observed during suborbital re-entry. This causes significant problems for aerodynamic decelerators using an inflatable aeroshell; thus, mitigating this problem is necessary. In this study, we revealed the instability mechanism using a computational science approach. To reproduce the in-flight oscillation motion in an unsteady turbulent flow field, we adopted a large-eddy simulation approach with a forced-oscillation technique. Computations were performed for two representative cases at transonic and subsonic speeds that were in stable and unstable states, respectively. Pitching moment hysteresis at a cycle in the motion was confirmed for the subsonic case, whereas such hysteresis did not appear for the transonic case. Pressures on the front surface and in the wake of the vehicle were obtained by employing a probe technique in the computations. Pressure phase delays at the surface and in the wake were confirmed as the pitch angle of the vehicle increased (pitch up) and decreased (pitch down), respectively. In particular, we observed that the wake structure formed by a large recirculation behavior significantly affected the pressure phase delay at the rear of the vehicle. The dynamic instability at subsonic speed resulted from flows that could not promptly follow the vehicle motion. Finally, the damping coefficients were evaluated for the design and development of the inflatable vehicle.

© 2020, American Institute of Aeronautics and Astronautics Inc, AIAA. All rights reserved. In the engine combustion test using the hypersonic wind tunnel facility, high temperature, high pressure, and high velocity flow is generated by combustion heating upstream of the nozzle. However, in this process, the turbulence of the flow and the contamination by the combustion gas occur unavoidably. Therefore, it is a concern that the different situation is observed between the ground test and actual flight because of these unavoidable factors. In this study, we focus on the mixing process of the injected fuel and the free-stream and aim to numerically evaluate the influence of the free-stream turbulence on the jet mixing. The LES calculation with free-stream turbulence is performed by the inflow boundary condition applying the velocity, temperature, and pressure fluctuations based on the Random Flow Generation (RFG) method, strong Reynolds analogy (SRA), and an isentropic process. The input parameter for RFG can control the generated free-stream turbulence characteristic inside the calculation domain. The difference between the cases with and without free-stream turbulence, whose intensity is up to 2.2% of mean free-stream velocity, is not observed clearly on the mean and instantaneous value fields in the present calculation.

© 2020, American Institute of Aeronautics and Astronautics Inc, AIAA. All rights reserved. Turbulence grids were applied to Mach 2 supersonic wind tunnel to increase turbulence in a mainstream. We measured wall pressure, velocity by Laser Speckle Velocimetry (LSV) using acetone condensation nanoparticle and density by acetone Planer Laser Induced Fluorescence (PLIF). In this study, the test section has 12 mm width and 10 mm height at the exit of the nozzle and the turbulence grids, which consisted of tungsten wires having sub-mm diameter, was installed at the exit of the nozzle. Combination of the wire grid and tunnel wall expansion increased mainstream turbulence without flow unstart in the test section flow. In the case with 0.4-mm-diameter grid at 3 by 3 arrangement having a blockage of 21% of the nozzle exit area, the mainstream turbulence reached 8% of the mainstream velocity. Nitrogen gas was perpendicularly injected into the grid-generated supersonic turbulent flow and it mixing performance was investigated by using acetone-PLIF. Installation of the wire grid affected not only mainstream turbulence but also wall-bounded flow, resulting in thickening boundary layer. As a result, jet penetration increased with installing the wire grid. However, no remarkable improvement of the jet mixing was observed with installing the wire grid.

© 2019, American Institute of Aeronautics and Astronautics Inc, AIAA. All rights reserved. Condensation nanoparticle planar laser light scattering imaging was conducted in a suction type Mach 1.9 supersonic flow. A series of the nanoparticle image pair was obtained by using the same optical components of PIV. The velocity data were obtained by the image pairs using the image-based correlation procedure as same as PIV. Acetone vapor instead of tracer particles in a conventional PIV was added to a mainstream gas in a reservoir. Acceleration process through a Laval nozzle automatically generated condensation particles of the acetone which were uniformly seeded into the entire flowfield. The increase in the additive concentration increased the number density of the particle and enabled a detail visualization of the vortex structures in the boundary layer. The increase in the additive concentration also increased the mean molecular weight of the acetone-seeded air. This decreased the flow speed. However, this is not a big matter because the heat release due to the condensation was negligible and the decrease in the flow speed was easily predicted from the thermodynamic properties of the gas. The particle size was difficult to be measured directly, so the tracer response time was estimated by the oblique shock test. The Stokes diameter of the particle was estimated to be 160 nm. Such a small diameter particle provides high traceability resulting in capturing shock wave with a few vector spacing and also provides high spatial resolution image resulting in capturing sub-mm scale vortices within the boundary layer. The mean and turbulent velocity fields evaluated from such high spatially and temporally resolved image pairs fairly agreed with previous measurement in the boundary layer.

Numerical detection of harmful vortices in pump sumps, such as an air-entraining vortex (AEV) and a submerged vortex (SMV), is crucially important to develop the drain pump machinery. We performed numerical simulations of the benchmark experiments of the pump sump conducted by Matsui et al. (2006 and 2016) using the OpenFOAM and compared the simulation results with the experimental data considering the effects of turbulence model, grid density and detection method of the vortices. We studied the threshold of the gas-liquid volume fraction of the VOF method and the second invariant of velocity gradient tensor to identify AEV and SMV. The methods proposed in the present paper were found to be very effective for the detection of the vortices, and the simulation results by RANS with the SST k-omega model successfully reproduced the experimental data. LES with the Smagorinsky model, however, was sensitive to the grid system and difficult to reproduce the experimental data even for the finest grid system having 3.7 million cells in the present study.

Turbulent flow through helical pipes with circular cross section is numerically investigated comparing with the experimental results obtained by our team. Numerical calculations are carried out for two helical circular pipes having different pitches and the same nondimensional curvature delta (=0.1) over a wide range of the Reynolds number from 3000 to 21,000 for torsion parameter beta (=torsion /root 2 delta = 0.02 and 0.45). We numerically obtained the secondary flow, the axial flow and the intensity of the turbulent kinetic energy by use of three turbulence models incorporated in OpenFOAM. We found that the change to fully developed turbulence is identified by comparing experimental data with the results of numerical simulations using turbulence models. We also found that renormalization group (RNG) k-epsilon turbulence model can predict excellently the fully developed turbulent flow with comparison to the experimental data. It is found that the momentum transfer due to turbulence dominates the secondary flow pattern of the turbulent helical pipe flow. It is interesting that torsion effect is more remarkable for turbulent flows than laminar flows.

Three-dimensional direct numerical simulations of a viscous incompressible fluid flow through a helical pipe with a circular cross section were conducted for three Reynolds numbers, Re (= 80, 300, and 1000), and two nondimensional curvatures, delta (= 0.1 and 0.05), over a wide range of torsion parameters, beta (= nondimensional torsion= /root 2 delta), from 0.02 to 2.8. Well-developed axially invariant regions were obtained where the friction factors were calculated, in good agreement with the experimental data obtained by Yamamoto et al. [Fluid Dyn. Res. 16, 237 (1995)]. It was found that the friction factor sharply increases as beta increases from zero, then decreases after taking a maximum, and finally slowly approaches that of a straight pipe when beta tends to infinity. It is interesting that a peak of the friction factor exists in the region 0.2 <= beta <= 0.3 for all the Reynolds numbers and curvatures studied in the present paper, which manifests the importance of the torsion parameter in helical pipe flow.

<p>Numerical prediction of air-entraining and submerged vortices in pump sumps is important for engineering applications. The validation of pump sump simulations, however, still is not enough, because the simulations is very complicated; for examples, treatment of gas-liquid interface, detection method of the vortices and selection of turbulence model etc. We conducted numerical simulations of the benchmark experiments of the pump sump conducted by Matsui et al. (2006, 2016) and compared the simulation with the experimental data to investigate the effects of turbulence model, grid density and detection method of the vortices. We determined a threshold of the gas-liquid fraction function of VOF method (<i>α</i>) and the second invariant of velocity gradient tensor (<i>Q</i><sub>2</sub>) to detect the air-entraining and submerged vortices by using vorticity, respectively. This method well detected the vortices and well reproduced the experiments for the RANS simulation using SST <i>k-ω</i> model. Large eddy simulation using Smagorinsky model, however, was sensitive to the grid system and difficult to reproduce the experimental vortex structures even for the finest grid system having 3.7 million cells.</p>

Fluctuating pressure (p' ) of a large-scale vortical structure generated in a semiconductor single wafer spin cleaner was detected by using microphone array. Twelve microphones were installed on the exhaust cover under the rotating disk of the cleaner with their interval of 7.5° or 15°. Power spectrum densities (PSD) of p' were compared with those of fluctuating velocity measured by PIV for various rotation angular velocities to identify fluctuations due to convection of the large-scale vortical structure. Good agreement of PSDs indicates that the large-scale structure could be detected by using microphone. Cross-correlation of p' measured at different positions revealed that the large-scale structure convected to the downstream in the rotational direction of the disk. The convection speed was about 12 % of the angular velocity of the rotating disk. Number of the vortex in the large-scale structure was also evaluated from the time-series p' data. Time-space contour map was made for p' based on the data measured at the different angular position, and showed periodical swept strip patterns. Presences of the strip patterns indicate the pressure disturbances were stably convected to the downstream. From this time-space map, two-dimensional Fourier transform efficiently extracted the number of vortices in the large-scale structure.

<p>In the electrodynamic flow control, weakly-ionized plasma flow behind the strong shock wave could be controlled by the applied magnetic field around a reentry vehicle. To control the flow field, a very strong magnetic field is required and it could be applied by the superconducting magnet, which is too large and heavy for a reentry capsule. Thus, in the present study, to avoid the use of the superconducting magnet, the electrodynamic effect from the combination of multiple weaker magnetic source such as permanent magnets is numerically investigated. According to the MHD simulation, the influence on the drag force caused by the multiple magnetic source, which is placed equiangularly around the body axis, could be diminished by the Hall effect. When the Hall effect is significant, the induced electric current intensity is very small because the electric field is generated much weaker than the one of the single magnet case. Therefore, the present magnetic configuration using multiple magnetic source might not be effective under the high Hall parameter condition such as the reentry flight.</p>

We experimentally and numerically investigated large-scale structures formed by vortices in a single wafer spin cleaner. The Q-criterion identified the vortices developed in the cleaner as the flow regions with positive second invariant of the velocity gradient tensor obtained by both the PIV and LES. The time-series two-components PIV data shows that small-vortices were clustered near and under the edge of the rotating disk and were periodically emanated from there to the housing wall of the cleaner. The emanation frequency was increased with increasing in the angular velocity of the rotating disk. Three-dimensional LES reveal that six longitudinal vortices were spirally developed from under the edge of the rotating disk to the housing wall. This structure stably rotated slower than the disk speed. Fourier analysis of the LES data agreed with that of the PIV data. This supports that the passages of the stable spiral vortices on the PIV measurement region resulted in the periodical emanation of the clustered small-vortices observed in the PIV. Such a very large-scale spiral structure will induce reattachment of contaminants on the wafer surface, and should be destructed for development of much higher efficient cleaner.

An inflatable decelerator is promising as a next-generation atmospheric-entry system owing to its reduced aerodynamic heating and high packing efficiency. In this study, a suborbital reentry demonstration of a flare-type thin-membrane aeroshell sustained by a single inflatable torus using an S-310-41 sounding rocket was carried out. An experimental vehicle was specially developed for this reentry demonstration; it was equipped with a 1.2-m-diam flare-type thin-membrane aeroshell and had a total mass of 15.6 kg. In the flight test, the aeroshell with an inflatable torus was deployed at an altitude of 100 km during a suborbital flight under the conditions of zero-gravity and near vacuum. The experimental vehicle reentered Earth's atmosphere from an altitude of 150 km. During free fall, it accelerated to a Mach number of 4.5 (1.32 km/s) because of gravity force. After that, it started decelerating because of aerodynamic force at an altitude of 70 km. According to the flight data, the vehicle remained intact during the reentry and the aeroshell achieved the expected decelerating performance. This reentry demonstration proves that the flare-type thin-membrane aeroshell sustained by the inflatable torus works well as a decelerator for atmospheric-entry vehicles. Further, the drag coefficient of the experimental vehicle in the supersonic, transonic, and subsonic regimes under free-flight conditions was estimated from the flight trajectory.

The electrodynamic effect on the partially ionized flow around magnetized bodies was numerically investigated. In particular, the double-cone model was considered because, unlike a simple blunt-nosed model, it generates a complex flow including shock-shock and shock-boundary-layer interactions. Such flows are significantly affected by an applied magnetic field. The modification of the flowfield due to the applied magnetic field causes drag force enhancement and/or the mitigation of the aerodynamic heating. This is enhanced with increasing magnetic field intensity. Furthermore, local flow features such as the separation bubble and the local peak heating that appears near the kink point of the double-cone model are significantly affected. The size of the separation bubble increases with increasing magnetic field intensity and is clearly influenced by the configuration of the magnetic field. The local peak heating is reduced with increasing magnetic field intensity, but the effect of the magnetic field configuration is weak.

The magnetoaerodynamic force exerted on a magnetized model in a weakly ionized flow was investigated. In particular, the effects of the applied magnetic field's orientation with respect to the model axis were examined by rotating a spherical permanent magnet installed in the nose of the model. The axial force increase due to the applied magnetic field is clearly influenced by the orientation of the applied magnetic field and takes on a maximum value when the line connecting the magnetic poles (magnetic pole line) is perpendicular to the incoming flow direction, whereas it takes on a minimum value when the magnetic pole line is moderately inclined against the incoming flow direction. Furthermore, the side force appears when the orientation of the applied magnetic field is apart from the incoming flow direction, whereas it disappears when the magnetic pole line is perpendicular to the incoming flow direction.

An inflatable decelerator is promising as atmospheric entry systems in the next generation. Although the various kinds of inflatable decelerators were proposed and researched in the past, we focus especially on a flare-type membrane aeroshell sustained by an inflatable torus. The flare-type membrane aeroshell may easily make re-entry systems simple and light-weighted because the aeroshell can be composed mainly of light-weighted membrane. As an important milestone of the development of the vehicle, a re-entry flight demonstration of the vehicle with the flare-type membrane aeroshell was carried out using Japanese S-310 sounding rocket in August, 2012. This flight experiment was successfully accomplished and various data, including the attitude of the vehicle during the supersonic atmospheric re-entry, was acquired. The attitude dynamics was qualitatively reconstructed base on the flight data, which shows that the vehicle entered into the atmosphere with high angle of attack and experienced a drastic attitude motion caused by the aerodynamic force. After the initial significant motion phase, the attitude behavior of the vehicle calmed down before the successive deceleration phase. That is, the vehicle made a flight with a normal attitude and, as planned, its attitude was almost stable after the atmospheric-entry. In all, it was confirmed that the attitude stability of vehicles with a flare-type membrane aeroshell could be realized on a supersonic atmospheric-entry. © 2013 by the American Institute of Aeronautics and Astronautics, Inc. All rights reserved.

An inflatable decelerator is promising as atmospheric-entry systems in the next generation thanks to the aerodynamic heating relaxation and its packing efficiency. Our group focuses on a flare-type membrane aeroshell sustained by an inflatable torus, especially. As an important milestone of our development, a re-entry demonstration of the flare-type membrane aeroshell was carried out using a Japanese S-310 sounding rocket. The experimental vehicle which has a 1.2-meter-diameter membrane aeroshell and 15.6kg in total weight was developed for the re-entry demonstration. In this flight test, the membrane aeroshell with the inflatable torus was deployed at 100km in altitude during a suborbital flight under the zero gravity and vacuum condition, and the experimental vehicle re-entered the earth atmosphere from 150km in altitude. The experimental vehicle accelerated to 1.32km/s and Mach Number 4.5 due to the gravity force and started decelerating due to the aerodynamic force at 70km in altitude. According to the flight data, the experimental vehicle kept intact during the re-entry and the flare type membrane aeroshell achieved the expected decelerating performance. This re-entry demonstration proves that the flare-type membrane aeroshell sustained by the inflatable torus works well as a decelerator for atmospheric-entry vehicles. © 2013 by the American Institute of Aeronautics and Astronautics, Inc. All rights reserve.