Abstract. A two-dimensional flow generated by the interaction of two opposing, symmetric curved wall jets is investigated experimentally. The overall flow field can be divided into the curved wall jet region, the interaction region, and the merged jet region; thus, the results of the measurement are discussed to characterize these three distinct regions. For the curved wall jet region, the Reynolds stress distribution, the correlation coefficient, fiu v', and the ratio of normal stresses, v2/u 2, are presented and the effects of curvature and adverse pressure gradient on these distributions are discussed. The Reynolds stress distributions in the interaction region are analyzed in detail to illuminate the negative production of the turbulent kinetic energy. The developing jet in this region is found to accelerate owing to the very high pressure arising from the collision of the two wall jets. A counter-gradient shear flow situation is also observed in this interacting region. Measured data in the merged jet region are often compared to those of plane jets and the development of the merged jet is discussed in that respect. The spreading rate of the present merged jet is found to be much larger than that of the plane jets. To account for the larger spreading rate, the intermittency distribution is also investigated.

SUMMARYThe problem of on-line optimal tracking control is considered for an aircraft terrain-following system. An on-line optimal control law which reflects the non-linear airborne vehicle dynamics with bounded control input is obtained by using a pointwise optimization technique called optimal decision strategy.The stability of the resulting system is also discussed. To demonstrate the feasibility and effectiveness of the proposed scheme, simulations have been performed for a realistic terrain-following system.

We previously introduced the simplified sampling biorthogonal direction implicit (SSB-ADI) method for 2-D Maxwell's equations. This paper extends it to a 3-D SSB-ADI method. Similar to the original method, the 3-D SSB-ADI method also partially uses the sampling property and compact support of the SBTD algorithm to display spatial discretization. Dispersion and stability analysis are presented. A numerical example shows that the SSB-ADI method is superior to the SB-ADI method in terms of reducing the CPU time and memory requirement of the numerical simulation.

No abstract