Designing optical glasses by machine learning coupled with a genetic algorithm

Engineering new glass compositions have experienced a sturdy tendency to move forward from (educated) trial-and-error to data- and simulation-driven strategies. In this work, we developed a computer program that combines data-driven predictive models (in this case, neural networks) with a genetic algorithm to design glass compositions with desired combinations of properties. First, we induced predictive models for the glass transition temperature (T_g) using a dataset of 45,302 compositions with 39 different chemical elements, and for the refractive index (n_d) using a dataset of 41,225 compositions with 38 different chemical elements. Then, we searched for relevant glass compositions using a genetic algorithm informed by a design trend of glasses having high n_d (1.7 or more) and low T_g (500 °C or less). Two candidate compositions suggested by the combined algorithms were selected and produced in the laboratory. These compositions are significantly different from those in the datasets used to induce the predictive models, showing that the used method is indeed capable of exploration. Both glasses met the constraints of the work, which supports the proposed framework. Therefore, this new tool can be immediately used for accelerating the design of new glasses. These results are a stepping stone in the pathway of machine learning-guided design of novel glasses.     

电气化铁路既有线深路堑扩堑石方控制爆破

介绍电气化铁路既有线无隔墙台阶爆破扩堑方法。沿既有线方向设置低台阶 ,边界处布置光爆孔 ,采用“同列同段和列间微差”的起爆网路。采用自制的“炮被”和架设“钢管排架”阻挡飞石、滚石和滑石 ,确保了既有线的安全。文中还概述了“炮被”和“钢管排架”的制做以及作者的认识和体会。     

某金矿尾矿提纯石英应用对比试验研究

针对某金矿，在回收金银的基础上，对金银浮选尾矿进行云母和长石浮选，以及提纯石英试验，最终获得SiO2品位99.91%的石英产品。提纯石英产品与比利时矽比科矿业有限公司石英产品进行对比试验，通过化学成分、粒度组成、熔制、粘度和热膨胀系数测试对比试验，得出金矿提纯石英达到甚至超过矽比科进口石英产品的各项技术指标。     

Nanocrystallization and optical properties of CsPbBr3−xIx perovskites in chalcogenide glasses

The confinement of CsPbX₃ (X = Cl, Br, and I) perovskite nanocrystals (NCs) in a stabilized inorganic glass matrix is a new strategy for improving their long-term stability and promoting their applications in the optoelectronic field. Here, in situ nanocrystallization strategy is developed to precipitate CsPbBr_3?xI_x NCs with arbitrary I/Br ratio among an elaborately designed GeS₂–Sb₂S₃-based chalcogenide glass matrix. Spherical CsPbBr_3?xI_x NCs are homogeneously distributed in the glass matrix after thermal treatment. The photoluminescence (PL) spectra show that the emission peaks of CsPbBr_3?xI_x NCs can be tuned from 570 nm to 722 nm with the replacement of Br by I. The fs transient absorption (TA) spectra reveal that there exists some structural defects in the NCs, leading to short PL decay life. This work would shed light on confining CsPbX₃ NCs into glassy matrices, facilitating their future applications in photoelectronic fields.     

Development of silica based organic slurries for stereolithographic printing process

In this study, silica based slurries for stereolithographic printing of glass structures are developed and characterized. Stereolithography has the potential to print complex structures with high resolution. Therefore, acrylate based photocurable slurries have been developed and their viscosities are examined as a function of the solid loading. A critical shear rate can be derived, which must not be exceeded during the printing process. Therefore, rheological characterizations provide important insights into the printing process and the ability to produce samples with precise structures. Other properties such as polymerization time and curability kinetic were investigated with time dependent attenuated total reflection infrared spectroscopy (ATR-IR). Afterwards, the slurries were printed on a commercial printer operating with visible light. For debinding the printed green bodies, the decomposition temperatures were derived from thermogravimetric analysis in order to obtain stable and transparent samples.     

Surveying of Na2O3–BaO–PbO–Nb2O5–SiO2–Al2O3 glass-ceramics system in terms of alpha,proton, neutron and gamma protection features by utilizing GEANT4 simulation codes

In this study, nuclear shielding qualities of glass-ceramics with chemical composition Na₂O₃–BaO–PbO–Nb₂O₅–SiO₂–Al₂O₃ containing different amount of BaO and PbO were investigated. The μ_ρ values were simulated using GEANT4 toolkit at 0.015–20 MeV wide energy range and the obtained results were verified by theoretical WinXCOM results. The variables such as μ_ρ, HVL, MFP, Z_eff, N_eff, EBF and EABF were computed to determine the gamma-ray shielding performances of studied glass ceramics. The results revealed that increase in PBO and BaO percentages in glass samples has caused to decrease the HVL, MFP, EBF and EABF values and increase μ_ρ, Z_eff values. It has been seen that N28 and S24 samples own superior protection ability against gamma radiation. In addition, the shielding capacity of these glass ceramics against charged and uncharged particles were predicted by determination of MSP and PR values for alpha, proton and ΣR values for neutrons. It has been concluded that PbO and BaO addition improve radiation shielding competences of glass ceramics. The data obtained from this study will be beneficial for designing glass ceramics shields for radiation protection enforcements.     

Impact behavior and fractographic study of carbon nanotubes grafted carbon fiber-reinforced epoxy matrix multi-scale hybrid composites

Carbon nanotubes are the most promising reinforcement for high performance composites. Multiwall carbon nanotubes were directly grown onto the carbon fiber surface by catalytic thermal chemical vapor deposition technique. Multi-scale hybrid composites were fabricated using the carbon nanotubes grown fibers with epoxy matrix. Morphology of the grown carbon nanotubes was investigated using field emission scanning electron microscopy and transmission electron microscopy. The fabricated composites were subjected to impact tests which showed 48.7% and 42.2% higher energy absorption in Charpy and Izod impact tests respectively. Fractographic analysis of the impact tested specimens revealed the presence of carbon nanotubes both at the fiber surface and within the matrix which explained the reason for improved energy absorption capability of these composites. Carbon nanotubes presence at various cracks formed during loading provided a direct evidence of micro crack bridging. Thus the enhanced fracture strength of these composites is attributed to stronger fiber–matrix interfacial bonding and simultaneous matrix strengthening due to the grown carbon nanotubes.     

Mechanical damage characterization of glass fiber-reinforced polymer laminates by ultrasonic maps

A method for simultaneous measurement of the thickness and density for Glass Fiber-Reinforced Polymer (GFRP) laminate plates with ultrasonic waves in C-Scan mode is presented in the form of maps. The method uses three different signals in immersion pulse-echo C-Scan mode. The maps obtained based on the density show the heterogeneity of the material at high resolution at the pixel level (1 × 1 mm²) and therefore they represent an efficient tool to assess and evaluate the damage of the composite structures after manufacturing and after an applied mechanical loading.     

Operational time-series data modeling via LSTM network integrating principal component analysis based on human experience

Today’s information technologies involve increasingly intelligent systems, which come at the cost of increasingly complex equipment. Modern monitoring systems collect multi-measuring-point and long-term data which make equipment health prediction a “big data” problem. It is difficult to extract information from such condition monitoring data to accurately estimate or predict health statuses. Deep learning is a powerful tool for big data processing that is widely utilized in image and speech recognition applications, and can also provide effective predictions in industrial processes. This paper proposes the Long Short-term Memory Integrating Principal Component Analysis based on Human Experience (HEPCA-LSTM), which uses operational time-series data for equipment health prognostics. Principal component analysis based on human experience is first conducted to extract condition parameters from the condition monitoring system. The long short-term memory (LSTM) framework is then constructed to predict the target status. Finally, a dynamic update of the prediction model with incoming data is performed at a certain interval to prevent any model misalignment caused by the drifting of relevant variables. The proposed model is validated on a practical case and found to outperform other prediction methods. It utilizes a powerful deep learning analysis method, the LSTM, to fully process big condition monitoring series data; it effectively extracts the features involved with human experience and takes dynamic updates into consideration.