Search

Automotive manufacturers have been trying to improve fuel efficiency without compromising the structural integrity and one of the ways to resolve the issue is to expand usage of tailor welded blanks (TWBs) in car body structure. Friction stir welding (FSW) is a joining process, which can be well fitted to obtaining aluminum tailored blanks when compared to other conventional joining processes. This paper presents an experimental and numerical study on TWBs produced by FSW with dissimilar aluminum 5083-H32 and 6061-T6 alloy sheets. The quality of the friction stir welded dissimilar joints was evaluated in terms of metallographic observations, hardness studies and tensile tests. Moreover, the local property changes in the weld regions were observed through digital image correlation (DIC) method. Formability of friction stir welded blanks was evaluated in the biaxial stretch forming mode using the limiting dome height (LDH) test. The failure location and the LDH values of the formed blanks were correlated to the hardness and local properties across the welds. The FE simulation of the LDH tests was also conducted incorporating Yld2000-2d anisotropy constitutive properties of the parent metal and further considering the properties of the non-homogeneous welded zone. The Marciniak-Kuczynski model was applied to predict the failure successfully in the FE model, and the results were validated with experimental data.

The distributions, densities and energies of defects near InGaN/GaN quantum wells in blue light-emitting diodes were simultaneously determined by utilizing capacitance–voltage (C–V) measurements. By combining the modulation frequency dependency, temperature dependency and C–V depth profiling with additional laser illumination, the densities and the locations of the defective layers could be determined. The relative defect densities of the devices were directly compared by monitoring the magnitude of the frequency dependence. This frequency dependency varies distinctly as the sample temperature changes. The activation energies of defects are then determined by analyzing the frequency dependency of C–V with temperature. We found that three different defects states were formed in a low-temperature-grown un-doped GaN (LT-GaN) layer inserted under the active layer. The activation energies of those defects were determined to be 3.96, 12.1 and 45.9 meV. The formation of additional defects states in the active layers induced by the insertion of LT-GaN layer was also observed.

We propose and demonstrate to derive the Auger recombination coefficient by fitting efficiency–current and carrier lifetime–current curves simultaneously, which can minimize the uncertainty of fitting results. The obtained Auger recombination coefficient is cm6s−1 in the present sample, which contributes slightly to efficiency droop effect.

The thermal effect and athermal effect such as electro-plastic effect of metallic materials induced by high density current can dramatically reduce the flow stress, which is beneficial to the forming process of less formable metal. In this paper, pulse current-assisted deep drawing of the magnesium alloy is proposed due to lower energy consumption and higher efficiency. In this process, the metal sheet is designed in series in a pulse current circuit and heated directly by the pulse current. In addition, the insulated mould is employed to avoid the current leaking. Experiments were conducted to demonstrate the feasibility and advantages of the proposed process. An experimental process system was established and the electrical-assisted Erichsen cupping tests and rectangular cup drawing tests were performed. The experiments showed that the forming load was reduced and the cupping height and associated principal strains were increased in the Erichsen cupping and deep drawing process assisted by high-density electric current.

A 4×4 semiconductor optical amplifier gate switch matrix was fabricated by integrating undoped upper-cladded waveguides with buried ridge stripe semiconductor optical amplifiers. The undoped upper cladding was used to prevent the high propagation loss of the conventional p-cladded waveguides. The 4×4 switch showed positive fiber-to-fiber gain with only two integrated optical amplifiers per transmission path and all sixteen transmission paths showed uniform characteristics.

This paper presents the over-current quench test and post-quench operation results of a 7 T 78 mm winding diameter multi-width (MW), no-insulation (NI) magnet in a bath of liquid helium at 4.2 K. The MW-NI magnet consists of 13 double-pancake (DP) coils wound with GdBCO tapes having five different widths ranging from 4.1–8.1 mm. After the magnet reached 7.3 T at 253 A, the magnet current was further increased purposely until the magnet quenched at 312 A, corresponding to a current density of 895 A mm−2 for the central DP coils of the narrowest 4.1 mm tape. The NI DP coils showed a fast magnetically coupled quench propagation from the quenched DP to the rest of the 'healthy' DP coils. The stored magnetic energy of 25.4 kJ was completely dissipated in 0.3 s with an average dissipation power rate of 85 kW. The post-quench magnet, operated sequentially in baths of liquid nitrogen at 77 K and in liquid helium at 4.2 K, showed no discernable changes from the pre-quench magnet in their key parameters, except the magnet characteristic resistance, pre-1.4 mΩ versus post-3.6 mΩ. Thus, a forced quench of the magnet, thanks to the NI winding technique, kept the integrity—mechanical, electrical, and magnetic—of this NI magnet intact.

We report continuous-wave (CW) operation of an InAs/InGaAsP quantum-dot (QD) heterostructure laser diode (LD) with an output power of 1.1?W at room temperature. This is the first observation on the laser output over 1?W from InAs QD-LDs fabricated on InP substrates. Also, the high-power lasing emission was successfully achieved at up to 60??C. The improvement in the lasing characteristics can be attributed to enhancement in modal gain of QD-LDs because of the increase in spatial carrier confinement around localized QD states by using a dot-in-a-well structure and a thin GaAs strain-modulating layer.

This paper reports the transient characteristics of two types of magnetically coupled GdBCO racetrack pancake coils: (1)?a pair of insulated coils (INS?INS coils) and (2)?a pair of non-insulated coils (NI?NI coils). In order to investigate the electromagnetic behavior of the INS?INS and NI?NI coils, charging/sudden discharging tests were performed with respect to the distance between primary and secondary coils of the coil pairs. The charging/sudden discharging test results show that the amount of induced current and measured axial magnetic field decreased with increasing distance between the primary and secondary coils. Additionally, charge and discharge delays were clearly observed in the NI?NI coils. The simulated results show good agreement with the experimental results, which validates the proposed approach involving an equivalent circuit model.

We investigated the effect of double air gaps embedded between sapphire and undoped GaN on the strain reduction in the InGaN/GaN-based green LED structure. Selective GaN growth and electrochemical etching were exploited to achieve embedded air gaps. Raman spectroscopy and photoluminescence were employed to demonstrate the relation between strain relaxation and indium incorporation. The double air gaps caused strain relaxation and led to a higher In incorporation in InGaN layers, which in turn caused a redshift of the PL spectra. As a result, three different peak wavelengths according to the existence of air gaps were observed.

This paper reports experimental and analytical studies on the capture efficiency of a superconducting high-gradient magnetic separation (HGMS) system, employing ferromagnetic filters with various wire diameters and packing fractions. The magnetization force was inversely proportional to wire diameter because the magnetic field gradient increased with decreasing wire diameter. In addition, the capture efficiency of the HGMS system was increased with decreasing wire diameter and increasing packing fraction. The experimental results from the filtration tests via the superconducting HGMS system were in reasonably good agreement with the capture efficiency using the magnetization force obtained by magnetic field analysis.

Recently, ultra-thin chips with thicknesses of under 35??m have emerged as an option for thinner, high performance electronic devices. For reliable electronic devices and high throughput packaging processes, the mechanical properties of ultra-thin chips need to be accurately understood. In this study, the fracture strength of an ultra-thin flash memory chip was measured using a ball-on-ring (BOR) test. To evaluate and validate the bi-axial strength in the BOR test, a finite element analysis was performed. It was compared with the analytical solution based on Hertzian contact theory. Flash memory chip specimens with different thicknesses were prepared and their bi-axial strengths were tested with respect to various wafer thinning process parameters such as grinding speed and polishing time. Raman spectroscopy was used to characterize the residual stress generated during the wafer thinning process. The surface roughness of the silicon wafer was measured using an atomic force microscope under various wafer thinning conditions. From the study, the fracture strength characteristics of the ultra-thin chip could be established as a function of the wafer thinning parameters.

The lattice deformation of a-ZnO films on r-plane Al2O3 substrates, grown by plasma-assisted molecular-beam epitaxy (PAMBE), is strongly dependent on growth temperature: i) the unit cell volume is smaller than that of a reference ZnO bulk and it increases with the increase in growth temperature; ii) a-plane lattice constants decrease and c-plane lattice constants increase with the increase in growth temperature; and iii) residual strain decreases with the increase in growth temperature, irrespective of the direction of a-, m-, and c-axes. It is proposed that the ZnO lattices in the a-ZnO/r-Al2O3 system are individually relaxed along in-plane axes by releasing the residual strain in terms of compensating thermal mismatch.

{101̄1} semipolar GaN-based light-emitting diodes (LEDs) grown on 1° miscut m-plane sapphires substrates via metal organic chemical vapor deposition showed undulated surface morphology with ridges and valleys. On the ridge regions, two dominant emission peaks, one at a shorter wavelength (∼438 nm) and one at a longer wavelength (∼490 nm), were observed using electroluminescence and micro-photoluminescence. In the valley regions, the longer peak was observed to be significantly quenched due to the grain boundary. The origin of the longer peak is believed to be not only inhomogeneous distribution of In composition in multiple quantum wells (MQWs) but also strong localization around the ridge region. Moreover, thickness variation of faceted MQWs could be associated with the peak broadening in 101̄1 semipolar LEDs. The results were also confirmed by cathodoluminescence and cross-sectional transmission electron microscopy.

An efficient pump beam combining technique using single-mode fibers was introduced. Well-defined modes formed from a single-mode fiber bundle were efficiently coupled into an output multimode fiber. The coupling rate can be maintained up to an input fiber number of 331. The total power of a multimode output fiber combined with single-mode fibers can greatly exceed the output power of a multimode laser diode pigtailed with the same multimode fiber.

We report the electrical properties of single-walled carbon nanotube bundles aligned between two electrodes with a gap of 8 µm by the alternating current dielectrophoresis method. The resistance increased exponentially as the number of structural defects increased from the oxygen plasma treatment. However, no noticeable change was observed in the carrier concentration and tube–tube interaction. The resistance dependence on the plasma treatment time was explained on the basis of the localization of electron states at defect sites. The aspects of the defects were examined by measuring the effects of ammonia adsorption on the conductance of the device.

Stacked superconducting tapes are an important element for fabricating large current density, superconducting fault current limiters (SFCLs). Various methods are used to stack superconducting tapes and the selection of the most appropriate method for an SFCL may affect its quench and recovery characteristics. In the stacked structure, the thermal/cooling condition of each tape is affected by its position and the inserted insulation material. In this paper, six samples with distinctive interposing methods were fabricated and a short-circuit test was performed to determine the quench and recovery characteristics. The test results of the six samples were compared and the individual performance of each sample in the short-circuit condition was also evaluated.

In general, conventional high-Tc superconducting fault current limiters (SFCLs) are operated by cooling systems with a liquid cryogen, such as liquid nitrogen (LN2). However, in the fault mode, LN2 evaporates because of joule heating in the SFCL module, so the SFCL system experiences an enormous increase in nitrogen gas volume. In this case, the thermal stability and protection of the system become the primary concerns for the design of the SFCL cooling system. In order to enhance the thermal stability and safety of the system, an SFCL cooled by solid nitrogen (SN2) as a large heat capacitor has been proposed as an alternative. In this paper we report the quench/recovery characteristics of a YBCO-coated conductor (CC) evaluated in an SN2 cooling system for the SFCL. A feasibility study is also conducted on the reference design codes and thermal requirements for the optimal design of the SN2 cooling system. The results demonstrate that the improved thermal contact obtained between the SN2 cooling system and the SFCL module renders the proposed system a suitable cryogen for the SFCL module. Detailed experimental results for the LN2 and SN2 cooling systems are presented and discussed.