Impingement heat transfer by jet issuing from a cross-shaped nozzle

Heat transfer characteristics and flow patterns were measured over a plate for various separation distances between the nozzle exit and target plate when air issues from a sharp-edged cross-shaped nozzle and impinges on a plate. The local heat transfer coefficients in the radial direction for different circumferential positions were calculated using the wall temperatures measured by means of thermocouples, and flow patterns were observed using an oil-titanium IV oxide method. The isotherms of the infrared images were also measured using an infrared radiometer with a two-dimensional array of indium-antimony (InSb) sensors. The geometric axes were switched as a result of the self-induced velocity of a vortex filament; the convex corners became flat and the concave corners generated outward ejection. The distributions of the iso-heat transfer coefficient contours correspond well to the flow pattern and the isotherm contours. These contours extended diagonally and demonstrated the St. Andrew's cross pattern for short separations, subsequently changing to an octagonal pattern, and then becoming circular at large separations. The correspondence of the heat transfer characteristics to the flow behavior, as well as the heat transfer mechanism are also described. © 1998 Scripta Technica, Heat Trans Jpn Res, 27(3): 192–204, 1998     

Turbulent burning velocity of hydrogen–air premixed propagating flames at elevated pressures

Lewis number represents the thermo-diffusive effects on laminar flames. That of hydrogen–air mixture varies extensively with the equivalence ratio due to the high molecular diffusivity of hydrogen. In this study, the influences of pressure and thermo-diffusive effects on spherically propagating premixed hydrogen–air turbulent flames were studied using a constant volume fan-stirred combustion vessel. It was noted that the ratio of the turbulent to unstretched laminar burning velocity increased with decreasing equivalence ratio and increasing mixture pressure. Turbulent burning velocity was dominated by three factors: (1) purely hydrodynamic factor, turbulence Reynolds number, (2) relative turbulence intensity to reaction speed, the ratio of turbulence intensity to unstretched laminar burning velocity, and (3) sensitivity of the flame to the stretch due to the thermo-diffusive effects, Lewis and Markstein numbers. A turbulent burning velocity correlation in terms of Reynolds and Lewis numbers is presented.     

Fabrication of low-loss and wide-bandwidth v.a.d. optical fibres at 1.3 ?m wavelength

A low-loss and wide-bandwidth optical fibre in the long-wavelength region has been fabricated by the v.a.d. method. Loss at 1.3 ?m is 0.98 dB/km and 6 dB down bandwidth is 1600 MHz km at 1.26 ?m.     

Temperature distribution in Li2O pellets under neutron irradiation

The center temperature of Li₂O pellets under neutron irradiation was measured using an encapsuled pin. The surface temperatures of the pellets and the pellet/cladding heat transfer coefficients were estimated on the basis of the well-known conductivity integral.     

基于无线传感器网络的分散目标跟踪：实际测试平台的开发应用（英文）

In the real world, centralized tracking in a largescale wireless sensor network (WSN) may not be feasible due to the possible failure of fusion centre and the large communication delay in forwarding measurement data to the fusion centre. Distributed target tracking techniques can be employed by tasking sensor nodes near to the target to perform sensing, target state estimation and selection of future tasking sensor nodes. In this paper, the development and implementation of a prototype ultrasonic WSN testbed to demonstrate distributed target tracking using the Extended Kalman Filter (EKF) algorithm is described. In the testbed, a mobile robot is used to simulate the moving target, and static／mobile sensor nodes are deployed to detect and track the target. The sensor nodes and robots are equipped with sonar and MICAZ to receive and process instructions. Experimental evaluation of a number of sensor scheduling schemes are reported which shows the superior tracking performance of our distributed competition based sensor scheduling scheme.     

Effects of PbTiO 3 Seed Layer on the Characteristics of RF-Sputtered Pb(Zr 0.5 ,Ti 0.5 )O 3 Thin Films

The effects of seed layers on the characteristics of rf-sputtered lead zirconium titanate thin films were investigated. Prior to sputtering, PbTiO 3 seed layers (100 nm) were deposited onto the Pt/Ti/Si and Pt/Si substrates by sol-gel (spin coating) processing method. Structure-property relation was studied as functions of substrate temperature and sputtering conditions. Special efforts were given in optimizing the deposition parameters to prepare the films in the perovskite phase without post deposition annealing. Dielectric constant and loss tangent of the films were in the range 800-950 and 0.04 -0.06, respectively. Remanent polarization and coercive field were 23.1 w C/cm 2 and 75kV/cm, respectively, for the films without PbTiO 3 seed layer, where as the corresponding quantities for in situ -deposited perovskite Pb(Zr, Ti)O 3 films on PbTiO 3 seed layer were 28 w C/cm 2 and 65 kV/cm, respectively.     

Exploiting the relationships among several binary classifiers via data transformation

The structural resemblance among several existing classifiers has motivated us to investigate their underlying relationships. By exploring into the mapping solutions of these classifiers, we found that they can be linked by simple feature data scaling. In other words, the key to these relationships lies upon how the replica of feature data are being scaled. This finding leads us directly to an exploration of novel classifiers beyond existing settings. Based on an extensive empirical evaluation, we show that the proposed formulation facilitates a tuning capability beyond existing settings for classifier generalization.     

Face recognition with radial basis function (RBF) neural networks

A general and efficient design approach using a radial basis function (RBF) neural classifier to cope with small training sets of high dimension, which is a problem frequently encountered in face recognition, is presented. In order to avoid overfitting and reduce the computational burden, face features are first extracted by the principal component analysis (PCA) method. Then, the resulting features are further processed by the Fisher's linear discriminant (FLD) technique to acquire lower-dimensional discriminant patterns. A novel paradigm is proposed whereby data information is encapsulated in determining the structure and initial parameters of the RBF neural classifier before learning takes place. A hybrid learning algorithm is used to train the RBF neural networks so that the dimension of the search space is drastically reduced in the gradient paradigm. Simulation results conducted on the ORL database show that the system achieves excellent performance both in terms of error rates of classification and learning efficiency.