Parameters for Roughness Pattern and Directionality

The average flow model is widely used on the derivations of average Reynolds type equation.There are arguments on the use of Peklenik number ( ^P ), or Bhushan number ( ^B ). In this paper, the orientation angle ( _r ) of the representative asperities (the pattern directionality) as well as the Peklenik number defined in principal directions () is utilized as the parameters to define the texture of surface roughness. An experimental procedure based on the least square method is then proposed to identify the two parameters ( _r and ). The present procedure avoids the above argument on distinguishing the isotropic asperity with the anisotropic asperity oriented with 45°. Only one additional parameter ( _r ) is needed.     

Coupled effects of surface roughness and flow rheology on elastohydrodynamic lubrication

The coupled effects of surface roughness and flow rheology on elastohydrodynamic lubrication (EHL) circular contact problems are analyzed and discussed. The averaged type Reynolds equation utilizing the average flow model on the interactions between couple stress fluids and surface roughness, the elastic deformation equation, the viscosity–pressure and density–pressure relations equations, and the force balance equation are solved numerically by the multilevel multi-integration (MLMI) algorithm to calculate the pressure distributions and film thickness shapes. The results show that the transverse type roughness and standard deviation of composite roughness enhance the pressure and film thickness in the central contact region. Moreover, the longer the characteristic length of the couple stress fluids is, the smaller the pressure distribution is in the central contact region and the greater the film thickness is in all regions.     

Probability-Based Channel Pruning for Depthwise Separable Convolutional Networks

Channel pruning can reduce memory consumption and running time with least performance damage, and is one of the most important techniques in network compression. However, existing channel pruning methods mainly focus on the pruning of standard convolutional networks, and they rely intensively on time-consuming fine-tuning to achieve the performance improvement. To this end, we present a novel efficient probability-based channel pruning method for depthwise separable convolutional networks. Our method leverages a new simple yet effective probability-based channel pruning criterion by taking the scaling and shifting factors of batch normalization layers into consideration. A novel shifting factor fusion technique is further developed to improve the performance of the pruned networks without requiring extra time-consuming fine-tuning. We apply the proposed method to five representative deep learning networks, namely MobileNetV1, MobileNetV2, ShuffleNetV1, ShuffleNetV2, and GhostNet, to demonstrate the efficiency of our pruning method. Extensive experimental results and comparisons on publicly available CIFAR10, CIFAR100, and ImageNet datasets validate the feasibility of the proposed method.     

Phase Transformation of Calcium Phosphates by Electrodeposition and Heat Treatment

The effect of heat treatment on the calcium phosphate deposited on Ti-6Al-4V substrate using an electrolytic process is investigated. The calcium phosphate was deposited in a 0.04 M Ca(H₂PO₄)₂·H₂O (MCPM) solution on a Ti-6Al-4V substrate at 333 K (60 °C), 10 V, and 80 Torr for 1 hour, and calcined at various temperatures for 4 hours. The X-ray diffraction (XRD) results demonstrate that the phases are dicalcium phosphate (CaHPO₄, DCPD) and hydroxyapatile [Ca(PO₄)₆ (OH)₂, HAP] for the as-deposited samples. When the deposited sample was calcined at 873 K (600 °C) for 4 hours, the XRD results show that the transformation of DCPD to HAP occurs. Moreover, HAP converts to β-TCP, CPP, and CaO. For the sample calcined at 1073 K (800 °C) for 4 hours, the scanning electron microscopy (SEM) micrograph reveals that the crack of the calcined sample propagates with a width of about 3 μm. This result is due to HAP becoming decomposed and converting to β-TCP, CPP, CaO, and H₂O. The vaporization of H₂O within the calcined sample promotes the crack propagation and growth.     

Electrochemical Properties of Joints Formed Between Sn-9Zn-1.5Ag-1Bi Alloys and Cu Substrates in a 3.5 wt.% NaCl Solution

The electrochemical properties of the joints formed between Sn-9Zn-1.5Ag-1Bi alloys and Cu substrates in a 3.5 wt.% NaCl solution have been investigated by potentiodynamic polarization, X-ray diffraction, and scanning electron microscopy. For the Sn-9Zn-1.5Ag-1Bi/Cu joints in a 3.5 wt.% NaCl solution, corrosion current (I _corr), corrosion potential (E _corr) and corrosion resistance (R _p) are 2.46 × 10⁻⁶ A/cm², −1.18 V, and 7.54 × 10³ Ωcm², respectively. Cu₆Sn₅, Cu₅Zn₈, and Ag₃Sn are formed at the interface between the Sn-9Zn-1.5Ag-xBi solder alloy and Cu substrate. The corrosion products of ZnCl₂, SnCl₂ and ZnO are formed at the Sn-9Zn-1.5Ag-xBi/Cu joints after polarization in a 3.5 wt.% NaCl solution. Pits are also formed on the surface of the solder alloys.