Visualization Approaches for Evaluating Ridge Regression Estimators in Mixture and Mixture‐Process Experiments

When the component proportions in mixture experiments are restricted by lower and upper bounds, the input space of a designed experiment space can become an irregular region that can induce multicollinearity problems when estimating the component proportion parameters. Thus, ridge regression provides a beneficial means of stabilizing the coefficient estimates in the fitted model. Previous research has focused on using prediction variance as a metric for determining an appropriate value of the ridge constant, k. We use visualization techniques to illustrate and evaluate ridge regression estimators and the robustness of estimation with respect to the variance and the bias. The addition of bias allows better balancing between the stability of the estimators and minimally changing the estimates. We illustrate the graphical methods with mixture and mixture‐process examples from the literature. Copyright © 2014 John Wiley & Sons, Ltd.     

Materials and Wireless Microfluidic Systems for Electronics Capable of Chemical Dissolution on Demand

Electronics that are capable of destroying themselves, on demand and in a harmless way, might provide the ultimate form of data security. This paper presents materials and device architectures for triggered destruction of conventional microelectronic systems by means of microfluidic chemical etching of the constituent materials, including silicon, silicon dioxide, and metals (e.g., aluminum). Demonstrations in an array of home‐built metal‐oxide‐semiconductor field‐effect transistors that exploit ultrathin sheets of monocrystalline silicon and in radio‐frequency identification devices illustrate the utility of the approaches.     

The effects of residual polymerization solvent on the physical properties of poly(2‐cyano‐p‐phenylene terephthalamide)

Poly(2‐cyano‐p‐phenylene terephthalamide) (CY‐PPTA) was obtained by the polycondensation of terephthaloyl dichloride and 2‐cyano‐p‐phenylene diamine in the mixture of N‐methyl‐2‐pyrrolidone (NMP) and calcium chloride (CaCl₂). Washing the polymerized product with water and drying at the elevated temperature inevitably left a small amount of polymerization residues which could be eliminated only by additional washing with acetone. The thermogravimetric and ¹H‐/¹³C‐NMR analyses revealed that the residues were largely composed of NMP which existed as a complex with the polymer. The complex was broken up between 200 and 300 °C and evolved 5 wt % of gaseous products, which had an adverse effect on the physical properties of as‐spun CY‐PPTA fibers obtained by dry jet‐wet spinning. The heat treatment of the as‐spun fibers including residual NMP exhibited some porous morphology on the fiber surface due to the evolved gases. However, the existence of the residual NMP had little effect on the intrinsic viscosity and liquid crystalline phase behavior of the polymer. Both rheological and optical properties exhibited the critical concentration at 3 wt % with the clear schlieren texture of nematic liquid crystalline phase. The inclusion of residual NMP decreased dynamic viscosity and yield stress. © 2016 Wiley Periodicals, Inc. J. Appl. Polym. Sci. 2016 , 133, 43672.     

Effects of K2O Evaporation on the Structural Properties of KSbO3 Compounds

A specimen having a stoichiometric composition of KSbO₃·(KSb) calcined at 800°C has an R rhombohedral structure (RS), and changes to a Pn cubic structure (CS) when calcined at 1100°C. Finally, a <111>‐oriented rhombohedral phase is formed in the specimen calcined at 1230°C. K/Sb ratio decreases from 1.0 in RS, 0.93 in CS, and finally to 0.85 in <111>‐oriented rhombohedral phases. On the other hand, a specimen having a K‐excess composition of K_1.1SbO₃ calcined at 800°C shows a RS that is maintained in the K‐excess specimen calcined at 1230°C. The composition of these specimens is very close to KSb. Therefore, the RS with a space group of R is a stable form of KSbO₃. The formation of Pn cubic and <111>‐oriented R phases can be explained by the evaporation of K₂O during the calcination process at temperatures above 1100°C.     

Polydimethylsiloxane Fluidic Device with Polydopamine‐Coated Inner Channel for Production of Uniform Droplets

A polydimethylsiloxane (PDMS)‐based fluidic device with two flow channels is fabricated by using a rapid prototyping method. The PDMS‐based fluidic device is used to produce water‐in‐oil emulsion (W/O) droplets due to its intrinsic hydrophobicity. To produce uniform oil‐in‐water (O/W) emulsion droplets, the inner channel of the PDMS fluidic device is coated with polydopamine (PDA) by flowing a dopamine precursor in the water channel of the fluidic device. The PDA coating is confirmed by an increase in morphological roughness and nitrogen content. In addition, the contact angle of the PDMS surface decreases from 95° to 30° during PDA coating, suggesting that the inner surface of the fluidic device is hydrophilic. Uniform W/O and O/W emulsion droplets are produced by the pristine PDMS and PDA‐coated PDMS fluidic devices, respectively.

     

Pruning deep convolutional neural networks for efficient edge computing in condition assessment of infrastructures

Health monitoring of civil infrastructures is a key application of Internet of things (IoT), while edge computing is an important component of IoT. In this context, swarms of autonomous inspection robots, which can replace current manual inspections, are examples of edge devices. Incorporation of pretrained deep learning algorithms into these robots for autonomous damage detection is a challenging problem since these devices are typically limited in computing and memory resources. This study introduces a solution based on network pruning using Taylor expansion to utilize pretrained deep convolutional neural networks for efficient edge computing and incorporation into inspection robots. Results from comprehensive experiments on two pretrained networks (i.e., VGG16 and ResNet18) and two types of prevalent surface defects (i.e., crack and corrosion) are presented and discussed in detail with respect to performance, memory demands, and the inference time for damage detection. It is shown that the proposed approach significantly enhances resource efficiency without decreasing damage detection performance.     

Wear resistance characteristics of materials used in oil sands plants

The surge in demand for natural resources has shifted the focus of the international community toward the development of oil sands, shale oil, shale gas and other non-traditional energy sources. In extreme environments, materials used in petroleum gas plant modules are accompanied by various problems caused by low-temperature brittleness such as damage, corrosion and wear. Many researchers have been conducting studies to discover a suitable material whose lifespan could be improved by performing characteristics analyses and performance assessments. In this study, a material characteristics assessment was conducted based on a wear resistance test on materials that are commonly used at oil sands plants. Prior to a wear resistance test, a chemical composition analysis was performed on each of the specimens, and tensile, impact, hardness and corrosion tests were carried out to examine the correlation between their results with the results of the wear resistance test. Each test was performed according to ASTM G 105 standards, and the change in weight according to wear length was analysed for each material to determine the related tendencies. In addition, the results of the wear test were derived by analysing the change in the mass of the specimen before and after the test, and the surface roughness was assessed to analyse the performance related to wear and define the service life. The aim was to use these results to select a material that would be suitable for the abrasive environment of the key equipment and materials of plants.     

Low‐Temperature Sintering and Piezoelectric Properties of CuO‐Added KNbO3 Ceramics

Dielectric and piezoelectric properties of CuO‐added KNbO₃ (KN) ceramics were investigated. The CuO reacted with the Nb₂O₅, formed a CuO–Nb₂O₅‐related liquid phase during the sintering, and assisted the densification of the KN ceramics at low temperatures. Moreover, some of the Cu²⁺ ions replaced the Nb⁵⁺ ions in the matrix and behaved as a hardener. The dielectric and piezoelectric properties of the KN ceramics were considerably influenced by the relative density. The 1.0 mol% CuO‐added KN ceramic sintered at 960°C for 1.0 h, which showed a maximum relative density, exhibited a high phase angle of 86.9°, P_r of 14.8 μC/cm², and E_c of 1.8 kV/mm. This specimen also exhibited good dielectric and piezoelectric properties: ε^T₃₃/ε_o of 364, d₃₃ of 122 pC/N, k_p of 0.29, and Q_m of 611.