An efficient ray tracing algorithm for the simulation of light trapping effects in Si solar cells with textured surfaces

Optimizing the design of the surface texture is an essential aspect of Si solar cell technology as it can maximize the light trapping efficiency of the cells. The proper simulation tools can provide efficient means of designing and analyzing the effects of the texture patterns on light confinement in an active medium. In this work, a newly devised algorithm termed Slab-Outline, based on a ray tracing technique, is reported. The details of the intersection searching logic adopted in Slab-Outline algorithm are also discussed. The efficiency of the logic was tested by comparing the computing time between the current algorithm and the Constructive Solid Geometry algorithm, and its superiority in computing speed was proved. The validity of the new algorithm was verified by comparing the simulated reflectance spectra with the measured spectra from a textured Si surface.     

CMOS compatible thermoelectric infrared sensors

A new structure for CMOS compatible thermoelectric infrared sensors is proposed. By using micro-link structures to connect several floating membranes, the largest floating membrane area yet obtained and large output voltages have been realised. The characteristics of the sensors have been measured, and are compared with those of existing devices     

Pharmacokinetics of enrofloxacin after intravenous, intramuscular and oral administration in houbara bustard (Chlamydotis undulata macqueenii)

The in-vitro activity of enrofloxacin against 117 strains of bacteria isolated from bustards was determined. Minimum inhibitory concentrations for 72% of the Proteus spp., E. coli, Salmonella spp. and Klebsiella spp. (n = 61) and for 48% of the Streptococci spp. and Staphylococci spp. (n = 31) were < or = 0.5 microg/mL. The minimum inhibitory concentration (MIC) of 76% of Pseudomonas spp. (n = 25) was < or = 2 microg/mL. Fourteen strains were resistant to concentrations > or = 128 microg/mL. The elimination half-lives (t1/2 elim beta) (mean +/- SEM) of 10 mg/kg enrofloxacin in eight houbara bustards (Chlamydotis undulata) were 6.80 +/- 0.79, 6.39 +/- 1.49 and 5.63 +/- 0.54 h after oral (p.o.), intramuscular (i.m.) and intravenous (i.v.) administration, respectively. Enrofloxacin was rapidly absorbed from the bustard gastro-intestinal tract and maximum plasma concentrations of 1.84 +/- 0.16 microg/mL were achieved after 0.66 +/- 0.05 h. Maximum plasma concentration after i.m. administration of 10 mg/kg was 2.75 +/- 0.11 microg/mL at 1.72 +/- 0.19 h. Maximum plasma concentration after i.m. administration of 15 mg/kg in two birds was 4.86 microg/mL. Bioavailability was 97.3 +/- 13.7% and 62.7 +/- 11.1% after i.m. and oral administration, respectively. Plasma concentrations of enrofloxacin > or = 0.5 microg/mL were maintained for at least 12 h for all routes at 10 mg/kg and for 24 h after i.m. administration at 15 mg/kg. Plasma enrofloxacin concentrations were monitored during the first 3 days of treatment in five houbara bustards and kori bustards (Ardeotis kori) with bacterial infections receiving a single daily i.m. injection of 10 mg/kg for 3 days. The mean plasma enrofloxacin concentrations in the clinical cases at 27 and 51 h (3.69 and 3.86 microg/mL) and at 48 h (0.70 microg/mL) were significantly higher compared with the 3 h and 24 h time intervals from clinically normal birds. The maximum plasma concentration (Cmax)/MIC ratio was ranked i.v. (10/mg/kg) > i.m. (15 mg/kg) > i.m. (10 mg/kg) > oral (10 mg/kg), but it was only higher than 8:1 for i.v. and i.m. administrations of enrofloxacin at 10 mg/kg and 15 mg/kg, respectively, against a low MIC (0.5 microg/mL). A dosage regimen of 10 mg/kg repeated every 12 h, or 15 mg/kg repeated every 24 h, would be expected to give blood concentrations above 0. 5 microg/mL and hence provide therapeutic response in the bustard against a wide range of bacterial infections.     

Tool Path Generation for Arbitrary Pockets with Islands

Avoiding gouge is crucial when converting cutter contact (CC) data into cutter location (CL) data for arbitrary pockets involving islands. This investigation presents a dynamic method for eliminating global gouges. Moreover, this study proposes using pair-wise bridges to merge pockets and islands. The proposed pair-wise bridge method leaves very small sized scallops. Additionally, an alternative Boolean method is designed to solve the problem of pocket containing numerous islands. Received: January 2005 / Accepted: September 2005     

Nonconforming quadrilateral finite elements:¶a correction

The object of this note is to correct an error in defining the basis for some nonconforming finite element methods when quadrilaterals occur in the partition of a two-dimensional domain. Received: March 2000 / Accepted: April 2000     

The joint strength and microstructure of fluxless Au/Sn solders in InP-based laser diode packages

The joint strength and microstructure of fluxless Au/Sn solders in InP-based laser-diode packages after thermal-aging testing were studied experimentally and numerically. Specimens were aged at 150°C for up to 64 days. The joint strength decreased as aging time increased. The microstructure and fracture surface of the Au/Sn solder joints showed that the joint strength decrease was caused by both the enlargement of the initial voids and an increase in the number of voids as the aging time increased. Finite-element method (FEM) simulations of joint strength were in good agreement with experimental measurements. Both experimental and numerical results indicate that the enlargement of the initial voids and an increase in the number of voids, caused by stress concentration as the aging period increased, resulted in the weakness of joint strength. The effect of temperature-cycling testing on the power variation of the InP laser diodes using fluxless Au/Sn solders was also studied. It was shown that the laser diodes operated in the stable condition up to 500 cycles.     

A Tight BER Upper Bound for Bit-Interleaved Coded Modulation with Square QAM and Gray Labeling

Bit-interleaved coded modulation (BICM) is a bandwidth-efficient coded system with diversity order higher than that of Ungerboeck's trellis-coded modulation on fading channels. In this paper, we investigate the BER (bit error rate) performance of BICM in the additive white Gaussian and Rayleigh fading channels. A new upper bound is given for the square QAM constellation with gray labeling, which constitutes a large portion of practical applications of BICM systems. The new upper bound is tighter than the well-known BICM union bound proposed in G. Caire et al. (1998)     

A microfluidic flow-converter based on a double-chamber planar micropump

A microfluidic flow-converter that transforms an oscillatory flow into a steady-like flow in a reciprocating-type pumping device is successfully developed in this study. The flow quality at the outlet is found to be significantly improved. The present micro-device is composed of two single-chamber PZT micropumps in parallel arrangement and can be fabricated using simple micro-electro-mechanical-system (MEMS) techniques. Based on the concept of the electronic bridge converter, the flow rectification is supported by four passive planar valves. Two operation modes, in-phase and anti-phase, were used to test the performance of the present device. In addition, the flow characteristics at the outlet were examined by an externally triggered micro-PIV system. The results reveal that the current flow-converter provided both high volume and smoothly continuous flow rates at the outlet when it was in anti-phase mode. Moreover, the volume flow rate was linearly proportional to the excitation frequency within a specific frequency regime. This indicates that the flow-converter was easily operated and controlled. The present microfluidic flow-converter has great potential for integration into future portable micro- or bio-fluidic systems.     

The principle cause of crack defects in optoelectronic materials with phosphorus-containing underlayer

The principle cause of cracks in laser-welded, Au-coated, optoelectronic materials with a phosphorus (P)-containing underlayer is studied experimentally and numerically modeled. Experimental results find that the crack formation is due to the existence of a P-containing underlayer and is not due to the thickness of the Au plating layer. The P-containing underlayer introduced by a electroless plating process may generate a low melting, P-rich segregation layer during solidification. A finite-element method analysis is performed to evaluate the residual stresses variation of the low melting P-rich segregation layer. Results show that the high residual tensile stresses of the P-rich segregation layer are generated by solidification shrinkage. A crack may be initiated by this residual stress. Both experimental observations and numerical calculations indicate that the crack formation mechanism in laser-welded Au-coated optoelectronic materials is directly related to the low melting P-rich segregation layer and its associated high tensile stresses. Based on these results, a Ni underlayer with P-free electroplating, instead of P-containing electroless plating, should be used prior to plating Au on optoelectronic materials to prevent crack formation in laser welded Au-coated optoelectronic materials.     

On the performance evaluation, selection, and design ofunique-words for time-division multiple access systems

The issues of performance evaluation, selection, and design of unique-words for time-division multiple-access systems are addressed for both additive white Gaussian noise and time-varying multipath fading channels. An efficient algorithm is proposed to compute the probability of a missed detection, which serves as the performance index of unique-words. Also, a simple upper bound on the probability of a missed detection is derived. This bound has been shown very tight for a large range of bit error rate. Based on this tight upper bound, a two-step algorithm is proposed for easy selection and design of good unique-words. For time-varying multipath fading channels, the performance of unique-words is evaluated for Rayleigh and Rician fading characteristics, which are suitable models for mobile and personal communication environments. The results show that the degradation caused by the fading effect can be larger than 10 dB for the performance of practical interest