Masaya Muto

Negative Magnus lift on a rotating sphere at around the critical Reynolds number

Negative Magnus lift acting on a sphere rotating about the axis perpendicular to an incoming flow was investigated using large-eddy simulation at three Reynolds numbers of 1.0 × 104, 2.0 × 105, and 1.14 × 106. The numerical methods used were first validated on a non-rotating sphere, and the spatial resolution around the sphere was determined so as to reproduce the laminar separation, reattachment, and turbulent transition of the boundary layer observed in the vicinity of the critical Reynolds number. The rotating sphere exhibited a positive or negative Magnus effect depending on the Reynolds number and the imposed rotating speed. At Reynolds numbers in the subcritical or supercritical regimes, the direction of the Magnus lift force was independent of the rotational speed. In contrast, the lift force was negative in the critical regime when particular rotating speeds were imposed. This negative Magnus effect was investigated in the context of suppression or promotion of boundary layer transition around the...

Negative Magnus Effect on a Rotating Sphere at around the Critical Reynolds Number

Negative Magnus lift acting on a sphere rotating about the axis perpendicular to an incoming flow is investigated using large-eddy simulation at three Reynolds numbers of 1.0× 104, 2.0 × 105, and 1.14 × 106. The numerical methods adopted are first validated on a non-rotating sphere and the spatial resolution around the sphere is determined so as to reproduce the laminar separation, reattachment, and turbulent transition of the boundary layer observed at around the critical Reynolds number. In the rotating sphere, positive or negative Magnus effect is observed depending on the Reynolds number and the rotating speed imposed. At the Reynolds number in the subcritical or supercritical region, the direction of the lift force follows the Magnus effect to be independent of the rotational speed tested here. In contrast, negative lift is observed at the Reynolds number at the critical region when particular rotating speeds are imposed. The negative Magnus effect is discussed in the context of the suppression or promotion of boundary layer transition around the separation point.

Large-Eddy Simulation of Turbulent Combustion in Multi Combustors for L30A Gas Turbine Engine

When designing a combustor, numerical analysis should be used to effectively predict different performances, such as flame temperature, emission, and combustion stability. However, even with the use of numerical analysis, several problems cannot be solved by investigating single combustors because, in an actual engine, interactions occur between multiple combustors. Therefore, to evaluate the detailed phenomenon in an actual combustor, the interactions between all combustors should be considered in any numerical analysis. On the other hand, a huge amount of computational cost is required for this type of analysis. Here a large-eddy simulation employing a flamelet/progress variable approach is applied to the numerical analysis of industrial combustors. The combustor used for this study is the L30A from Kawasaki Heavy Industries, Ltd. Computations are conducted with a supercomputer (referred to as the “K-computer”) in the RIKEN Advanced Institute for Computational Science. All combustors in the L30A engine (from the compressor outlet to the turbine inlet) are simulated, including the fuel manifold. This engine has eight can combustors that are connected through the fuel manifold and compressed air housing unit. The total number of elements is approximately 140 million. The flow patterns for each combustor are similar in all cans. A swirling flow from the main burner is formed and accelerated by the supplemental burner. There is a high-temperature region before the supplemental burner. The flow field and temperature distribution in an actual combustor interacting with other combustor cans are simulated adequately. The mass flow rate of the air and those of the fuels are distributed equally for each can. Therefore, the outlet temperature difference for each can is also very small.Copyright