In recent years, due to the shortage of timber resources, it has become necessary to reduce the excessive consumption of forest resources. Non-destructive testing technology can quickly find wood defects and effectively improve wood utilization. Deep learning has achieved significant results as one of the most commonly used methods in the detection of wood knots. However, compared with convolutional neural networks in other fields, the depth of deep learning models for the detection of wood knots is still very shallow. This is because the number of samples marked in the wood detection is too small, which limits the accuracy of the final prediction of the results. In this paper, ResNet-34 is combined with transfer learning, and a new TL-ResNet34 deep learning model with 35 convolution depths is proposed to detect wood knot defects. Among them, ResNet-34 is used as a feature extractor for wood knot defects. At the same time, a new method TL-ResNet34 is proposed, which combines ResNet-34 with transfer learning. After that, the wood knot defect dataset was applied to TL-ResNet34 for testing. The results show that the detection accuracy of the dataset trained by TL-ResNet34 is significantly higher than that of other methods. This shows that the final prediction accuracy of the detection of wood knot defects can be improved by TL-ResNet34.

To study the interaction between sand particles and geogrids, pullout tests were performed on one type of geogrid in Pingtan medium sand. The sand motion around the ribs was captured by high-resolution digital camera and CCD camera. The images were analysed with the aid of digital photography measurement and micro-image analysis. Microscopic particle motion and displacement fields around the transverse ribs were then obtained to verify the macro behaviour of the sand–rib interaction. It was found that the thickness of shear band increases with the normal stress applied, and the shear zone is asymmetrical to the soil–rib interface. A near wedge shape shear zone was observed by analysing the shear strain around the ribs, which indicated that a punching shear failure mechanism may be applicable to the tested geogrid in medium-dense sand. Comparison of various models for bearing capacity of geogrids shows that the size and shape of bearing members should be considered in the calculation of bearing resistance of geogrids.

In this paper, a metal–insulator–metal (MIM) waveguide structure consisting of a side-coupled rectangular cavity (SCRC), a rightward opening semi-ring cavity (ROSRC), and a bus waveguide is reported. The finite element method is used to analyze the transmission characteristics and magnetic-field distributions of the structure in detail. The structure can support triple Fano resonances, and the Fano resonances can be tuned independently by altering the geometric parameters of the structure. Moreover, the structure can be applied in refractive index sensing and biosensing. The maximum sensitivity of refractive index sensing is up to 1550.38 nm/RIU, and there is a good linear relationship between resonance wavelength and refractive index. The MIM waveguide structure has potential applications in optical on-chip nano-sensing.

A centrifuge test was performed on a reinforced embankment backfilled with lime-stabilized soil on soft clay installed with wick drains. The finite-element program PLAXIS was employed to simulate the centrifuge test on the basis of the test dimensions. Verifications of the numerical model were taken by comparing the numerical results with the measurements obtained from the centrifuge test, and it was found that both were in good agreement. Three cases of unreinforced, one-layer reinforced, and two-layer reinforced embankments were simulated and compared on the basis of the numerical model. A centrifuge similarity drainage relationship was also proposed in this paper on the basis of the equal degree of consolidation between the model and prototype drains.