Search

Classification of stellar spectra from voluminous spectra is a very important and challenging task. In order to better classify stellar spectra, inspired by the principle of deep convolutional neural network (CNN), we propose a supervised algorithm for stellar spectra classification based on 1D stellar spectra convolutional neural network (1D SSCNN). In 1D SSCNN, we modify the traditional 2D convolutional neural network into 1D network to adapt to the spectral classification. On the basis of using convolution algorithm, the spectral features are extracted and used for classification. We first use the stellar spectra data to train a 1D SSCNN to obtain a well-trained model, and then we apply the well-trained model to classify the unknown spectra. To evaluate the performance of the proposed algorithms, we apply 1D SSCNN to classify three spectral types: F-type spectra, G-type spectra, and K-type spectra and 10 subclasses of K-type spectra: A0-type, A5-type, F0-type, F5-type, G0-type, G5-type, K0-type, K5-type, M0-type, and M5-type spectra from Sloan Digital Sky Survey (SDSS). Our 1D SSCNN algorithm obtain higher classification accuracy compared with support vector machine (SVM), random forest (RF), and artificial neural network (ANN). [ABSTRACT FROM AUTHOR]

Primary combustion products of boron-based propellants are incomplete combustion products that are emitted from the gas generator of a solid ducted rocket. Studying the composition of primary combustion products provides valuable information about the primary combustion process and also helps to better understand the secondary combustion process. The particle size of the primary combustion products is analyzed by a laser particle size analyzer. The qualitative analysis of the sample composition is performed by using x-ray diffraction, x-ray photoelectron spectroscopy, and thermogravimetry-differential scanning calorimetry experiments. Based on these results, an integrated quantitative analysis of the sample composition is conducted. The quantitative analysis methods include tube furnace heating, ion chromatography, infrared spectroscopy, and inductively coupled plasma chromatography. In addition, scanning electron microscopy and energy dispersive spectrometry are also used to analyze the micro-morphology and distribution of different components in the sample. The primary combustion products mainly contain B, C, BC, HBO, BO, BN, Mg, MgCl, and NHCl. BC (22-24%), HBO (20%), and B (16.8%) are the three major components, while BC, B, and C (9.8-11.8%) are the three combustible components present in the highest amounts. The oxidant NHClO is completely consumed during the primary combustion, while the metal additive Mg does not show much reactivity. The micro-morphology and distribution of BC, HBO (or BO), B, Mg, and C in the sample are investigated. Some components in the primary combustion products are found to be agglomerated, while some components are dispersed. Large particles in the sample mainly include B and Mg, while BC, HBO (or BO), and C particles are small. In general, the combustion completeness of the primary combustion products is rather low. Therefore, better understanding and controlling of the secondary combustion process is very important to improve the performance of B-based propellants. [ABSTRACT FROM AUTHOR]

We investigate the strain relaxation and surface morphology of epitaxial SrTiO3 (STO) films grown on (001)O and (110)O planes of orthorhombic NdGaO3 (NGO), and (001) plane of cubic (LaAlO3)0.3(Sr2AlTaO6)0.7 (LSAT) substrates. Although the average lattice mismatches are similar, strikingly regular crosshatched surface patterns can be found on STO/NGO(001)O[(110)O] films, contrary to the uniform surface of STO/LSAT(001). Based on the orientation and thickness dependent patterns and high-resolution x-ray diffractions, we ascribe the crosshatch morphology to the anisotropic strain relaxation with possibly the 60° misfit dislocation formation and lateral surface step flowin STO/NGOfilms, while an isotropic strain relaxation in STO/LSAT. Further, we show that the crosshatched STO/NGO(110)O surface could be utilized as a template to modify the magnetotransport properties of epitaxial La0.6Ca0.4MnO3 films. This study highlights the crucial role of symmetry mismatch in determining the surface morphology of the perovskite oxide films, in addition to their epitaxial strain states, and offers a different route for designing and fabricating functional perovskite-oxide devices. [ABSTRACT FROM AUTHOR]

We present in situ measurements of the unit-cell volume of a natural terrestrial ilmenite (Jagersfontein mine, South Africa) and a synthetic reduced ilmenite (FeTiO3) at simultaneous high pressure and high temperature up to 16 GPa and 1273 K. Unit-cell volumes were determined using energy-dispersive synchrotron X-ray diffraction in a multi-anvil press. Mössbauer analyses show that the synthetic sample contained insignificant amounts of Fe3+ both before and after the experiment. Results were fit to Birch-Murnaghan thermal equations of state, which reproduce the experimental data to within 0.5 and 0.7 GPa for the synthetic and natural samples, respectively. At ambient conditions, the unit-cell volume of the natural sample [V0 = 314.75 ± 0.23 (1σ) ų] is significantly smaller than that of the synthetic sample [V0 = 319.12 ± 0.26 ų]. The difference can be attributed to the presence of impurities and Fe3+ in the natural sample. The 1 bar isothermal bulk moduli KT0 for the reduced ilmenite is slightly larger than for the natural ilmenite (181 ± 7 and 165 ± 6 GPa, respectively), with pressure derivatives K'0 = 3 ± 1. Our results, combined with literature data, suggest that the unit-cell volume of reduced ilmenite is significantly larger than that of oxidized ilmenite, whereas their thermoelastic parameters are similar. Our data provide more appropriate input parameters for thermo-chemical models of lunar interior evolution, in which reduced ilmenite plays a critical role. [ABSTRACT FROM AUTHOR]