Toshiya Kimura

Local area metacomputing for multidisciplinary problems: a case study for fluid/structure coupled simulation

A distributed parallel computing for a multidisciplinary problem has been realized on a heterogeneous parallel computer cluster. A communication library for heterogeneous parallel computing, Stampi, has been newly developed and implemented on a parallel computer cluster. This library is based on MPI specification and has capability of creating processes on the slave parallel computer from the master parallel computer. From programmers, only the ordinary functions of MPI-2 are seen and the heterogeneity is hidden. A fluid/structure coupled simulation has been applied, and the distributed parallel computing for the coupled simulation has been performed between vector-parallel and scalar-parallel computers using this library. The 3-dimension Euler equations were solved in a moving grid by a finite difference method on the vector-parallel computer, Fujitsu VPP300. The structure equation of motion was solved by a finite element method on the scalar-parallel computer, Hitachi SR2201. These two disciplines were solved concurrently and independently, synchronizing and exchanging the boundary data through the network. As a test problem, the aeroelastic response of the 3-dimensional wing in a transonic flow was calculated. In the present simulation, the CFD code and the CSD code have shown higher performance on the vector-parallel and the scaIar-parallel computers, respectively. Hence, the distributed parallel computing will be one of effective methods to raise the total performance for multidisciplinary simulations. problems

Distributed Parallel Computing for Fluid-Structure Coupled Simulations on a Heterogeneous Parallel Computer Cluster

Distributed parallel computing for a fluid-structure coupled simulation has been performed on a heterogeneous parallel computer cluster. The fluid and the structure dynamics are simulated on different parallel computers connected by a high-speed local network. These dynamics are coupled by a loose coupling method exchanging the boundary data between the fluid and the structure domains through the network. The data communication among parallel computers is realized by using the new communication library, Stampi, which has been developed to enable communication in a heterogeneous environment. The performance evaluation on a heterogeneous parallel computer cluster has shown that the distributed parallel computing for fluid-structure coupled simulations has the advantage of increasing the performance compared with the parallel computing on a single parallel computer.

Effects of Deep Throttling on Rocket Engine Systems

Dependencies of injection pressure, injection velocity ratio, pump flow rate, and coolant physical condition at a cooling jacket on the engine thrust level were investigated quantitatively for typical rocket engine cycles, an expander-bleed cycle and a staged combustion cycle using a dynamic simulator for a rocket engine. For low injection pressure of oxygen in the expander-bleed cycle engine during deep throttling, the effect of gasification of the oxygen using a heat exchanger was examined numerically. We found that placement of the heat exchanger upstream of the oxygen injector would be effective to increase the injection pressure difference, even though some oscillation would occur.

Experimental and Numerical Study on Performance of Extendible Nozzle for Altitude Compensation

An extendible nozzle for altitude compensation is considered to be a feasible device to improve the performance of booster engines because it can provide higher thrust at sea level and higher specific impulse in vacuum. The booster engine with the extendible nozzle has to deploy its nozzle extension on engine firing. For the design of the extendible nozzle including its driving mechanics, it is required to clarify the transitional phenomena during the nozzle deployment. Based on the experimental results of the firing tests using a sub-scale model, the characteristics of thrust coefficient, the nozzle axial load, and the nozzle side load affected by the nozzle flow transition are examined. The backflow of combustion gas through the gap between the fixed nozzle and the nozzle extension was also examined with use of CFD analysis. As results, disadvantages in the nozzle performance and the nozzle loads in case of improper nozzle deployment condition are clarified.

Development of a Reusable LOX/LH2 Rocket Engine - Firing Tests and Lifetime Evaluation Analysis

The main engine of the Reusable Sounding Rocket has been developed at the Kakuda Space Center of the Japan Aerospace Exploration Agency. A total of 54 engine firing experiments were conducted between June 2014 and February 2015. Throughout these experiments, advanced capabilities and functions of this engine, such as wide range throttling, accurate controllability, and health monitoring, were proved to be feasible. This rocket is also required to be reusable for over 100 flights. To confirm the long-life durability of this engine for over 100 flights in a limited experiment period, sequential multiple-firing tests were planned and conducted. Numerical analysis to evaluate damage of the main chamber due to firings was also conducted. The results of durability tests and numerical analysis showed reuse of this engine for over 100 flights to be feasible.

CFD/CSD coupled simulation on a parallel computer cluster

A coupled simulation for the computational fluid dynamics (CFD) and the computational. structure dynamics (CSD) has been performed on a heterogeneous parallel computer cluster. Tlie main solvers for CFD and CSD and the grid generator are executed being distributed to different parallel computers, which are connected by a network. The fluid and the structure dynamics are, coupled by exchanging the boundary data between the fluid and: the structure domains through the network. The data communication among processes oti a parallel computer and on different parallel computers is realized by the heterogeneous communication library, Stampi. In this paper, the procedure of the distributed parallel computing for the fluid/structure coupled simulation will be presented, and it will be shown that the distributed parallel computing has the greater advantage of increasing the total performance than parallel computing on a single parallel computer.