Performance of an anisotropy-resolving hybrid LES/RANS (HLR) model was investigated. An important feature of this HLR model is the introduction of an extra anisotropic term in a sub-grid scale (SGS) model for large eddy simulation (LES) to represent the SGS stress anisotropy more correctly. Although the basic performance of this model was validated in some previous studies, it is still unclear how the model works for very high Reynolds-number (Re) turbulent flows, in which no-slip wall conditions are no more applicable. Thus, to investigate the predictive performance of this HLR model, it was applied to very high Re turbulent channel flows up to Re_τ = 6 × 10⁷, together with a conventional wall function as the wall-boundary condition. The computational results obtained by the present anisotropic HLR model were carefully compared with those by an isotropic HLR model.

In this paper, the characteristics of a transitionally rough turbulent flow over a real rough surface are examined. In our research, high-resolution large eddy simulations over a scanned marine painted rough surface were carried out, and the friction Reynolds number and roughness Reynolds number (inner-scaled roughness height) were systematically varied. Away from the rough surface, the mean velocity and Reynolds stresses are unaffected by the mean roughness height. Further, the similarity away from the rough surface is clearly confirmed when we use the effective wall-normal distance. The effective wall-normal distance is defined as the wall-normal integral of the plane-porosity (void fraction). Moreover, near the rough surface, the Reynolds stresses asymptotically decay toward the bottom of the rough surface when plotted against the inner-scaled effective distance. The profiles of the mean velocity and Reynolds stresses against the effective distance reveal that the mean velocity profile can be characterized by the roughness Reynolds number (inner-scaled roughness length scale). However, the other parameters should be considered when characterizing the near-wall Reynolds stress behavior. When the budget terms in the plane and Reynolds averaged momentum equations are analyzed, it is found that the drag force term dominates the momentum transfer in the vicinity of the bottom of the rough surface. As the roughness Reynolds number increases, the pressure drag contribution to the skin friction coefficient increases, while the viscous contribution decreases.

In order to drastically accelerate the development processes of advanced heat exchangers, a new design framework integrating shape optimization, rapid prototyping and experimental validation is proposed. For the optimal design of heat transfer surfaces, a new adjoint-based shape optimization algorithm taking into account unsteady turbulent transport is developed. The present shape optimization algorithm is applied to two different conventional pin-fin arrays with circular cross sections so as to maximize the analogy factor, i.e., the ratio of heat transfer and pumping power for driving the fluid. The resultant optimal fin shapes are elongated in the streamwise direction and also characterized by bump-like structures formed on the upstream side of the pins. Investigation of numerical results reveals that the pressure drop of the optimal shape is significantly reduced by the suppression of vortex shedding behind the fin, whereas the heat transfer performance is maintained by the extended surface. The optimal shapes are fabricated by a resin-based additive manufacturing technique. A single-blow method allows to evaluate the heat transfer coefficient of low-thermal conductivity materials by measuring the inlet and outlet air temperature only, while the pressure loss is estimated from the pressure measurements at the upstream and downstream of the text matrix by Pitôt-tubes. As a result, significant improvement of thermal hydraulic performance is experimentally confirmed for the optimal pin-fin arrays as predicted by numerical analyses.

Pulsating turbulent channel flows under constant temperature difference (CTD) condition and uniform heat flux heating (UHF) condition are studied by direct numerical simulation (DNS). The main objective of the present study is to clarify how the dissimilarity between momentum transfer and heat transfer appeared in the pulsating flows, in which the dissimilarity may originate from the CTD condition dissimilar to the no-slip condition of the velocity field rather than from the thermo-fluid physics under pulsation. Simulations have been performed for three pulsation frequencies under the friction Reynolds number at steady-state, Re_τs = 300. Comparing the phase-averaged quantities under CTD and UHF conditions, it is found that the frequency dependence of the temperature oscillations in the near-wall region is almost the same regardless of the thermal boundary condition although the time-averaged temperature profiles are different. As a result, the ratio of Stanton number to friction factor, which works as a barometer of the analogy, changes during the pulsation period at both CTD and UHF conditions. Besides, the oscillation amplitude becomes larger as the pulsation frequency increases. Therefore, it was confirmed that the dissimilarity appears regardless of the thermal boundary condition. In addition, turbulent Prandtl number shows similar cyclic behavior to the ratio of Stanton number to friction factor. Time variations of each component constituting turbulent Prandtl number reveal that increasing dissimilarity at the high frequency is mainly attributed to the amplified oscillation of velocity gradient near the wall, where Reynolds shear stress and turbulent heat flux are kept at around the time-averaged values because the near-wall vortex structures cannot follow the rapid change of flow rate.

Conjugate radiation-convection-conduction simulation was conducted for a solar volumetric receiver of silicon carbide honeycomb for high temperature heat absorption at 1,000°C and higher. Simulation was made for three cases of channel cell size: 0.6mm; 1.5mm; 2.9mm. At two levels of incident heat flux 1,400 kW m⁻² and 4,200 kW m⁻², air mass flux was changed variously for optimization of working conditions. When the cell size is reduced from d = 2.9 mm to 0.6 mm, the receiver efficiency together with the air temperature at the receiver exit increase at each level of incident heat flux. At 1,400 kW m⁻², the receiver efficiency exceeds 0.8 when the air temperature is as high as 1000°C in the case of the smallest cell size: d = 0.6 mm. At 4,200 kW m⁻² , the efficiency surpasses 0.80 when the air temperature is almost 1500°C in the case of d = 0.6 mm. The heat losses from the receiver was analyzed through budget of energy balance equation. It was found that the thermal radiation was attenuated by reduction of channel cell size which resulted in enhancement of the receiver efficiency. The mean temperature at the top edge of the receiver decreased with the reduction of channel size in consistency with the attenuation of thermal radiation. The numerical result demonstrated that the reducing cell size is essential to absorb concentrated solar light at very high temperatures beyond 1000°C and higher.

Natural convection of a paramagnetic fluid through a partially-heated vertical channel is numerically studied in the presence of a magnetic field from two block magnets placed behind the heated wall. Magnets are aligned with opposite orientations. This magnet orientation induces strong magnetic force normal to the magnet at the magnet junction due to short distance between poles. When the temperature-dependent magnetic susceptibility changes due to wall heating near the magnet, the magnetothermal force is induced remarkably near the magnet junction. This additional force overlaps to the natural convection along the heated wall and results in the changes of heat and fluid flow along the heated wall. It is found that, flow becomes slow and the local heat transfer is suppressed below the elevation of the magnet junction, and the flow acceleration and heat transfer enhancement are observed above the junction elevation, of which effects depend on the relative magnet elevation to the heated wall. It is also found that the transition to vibrating flow occurs at the specific magnet elevation. The time-averaged Nusselt number suggests this vibrating convection has the potential to enhance the heat transfer remarkably because this magnetically-induced flow vibration continues along the heated wall up to the outlet.

High-enthalpy dissociating aerodynamic environment during high-speed spacecraft entries brings new challenges to the prediction of heat transfer characteristics on the surface of thermal protection system. This study numerically dealt with the coupling modeling of chemically reacting interface in carbon-oxygen dissociating environment in order to accurately and reliably predict the Mars entry heating load. Computational fluid dynamics, computational heat transfer and interface balance with proper coupling strategy were involved in the coupling algorithm to take into account the surface catalysis, material ablation and structural thermal response. Numerical simulation shows that the interfacial reaction model has influences on coupling evolution by exchanging various patterns of heat transfer on the interface, including that from temperature gradient, that caused by chemical reaction, and that carried by the injection kinetic energy. The types and energy barriers of interfacial reaction were found to change the aerodynamic heating enhancement and its evolution over coupling time, which is the most remarkable difference from the perfect gas result. Gas-solid interaction involving interfacial reaction exhibits three distinct temporal intervals: the initial, developing and fully developed stages, and the chemical dynamics and heat transfer characteristics vary at different temporal scales. Related research provides important technical support for the design of thermal protection system for space vehicles.

Six degrees of freedom turning flight simulation is presented for a tilt-rotor aircraft represented by V-22 Osprey, considering interaction of fluid and rigid body in a coupled manner. A tilt-rotor aircraft has a hovering function like a helicopter by turning axes of rotor toward the sky during takeoff or landing. On the other hand, it behaves as a reciprocating aircraft by turning axes of rotor forward in flight. The tilt-rotor aircraft is known to be susceptible to instable state compared to conventional aircraft. For realizing Digital Flight of turning flight of the aircraft, combination with the Moving Computational Domain (MCD) method and the multi-axis sliding mesh approach is applied. In the MCD method, the whole of the computational domain itself moves with the bodies included inside the domain, which makes an airplane possible to fly freely in the physical space without any restriction of region size. Moreover, this method is also applied to rotation of rotors. The multi-axis sliding mesh approach is computational technique to enable us to deal with multiple rotating axes of different direction, and it is used to rotate two rotors and change flight attitude of the aircraft. As a result of the coupled computation between flow field and rigid body using above approach, the airplane gained lift and propulsion by rotating the rotor and flew in turning by operating flight control surfaces such as flaperons, elevator and rudders. Moreover, the manipulating variables of flight control surfaces needed for turning flight, flight attitude of the aircraft and generated lift were found. Differences of fluid flow between straight flight and turning flight were also captured.