No abstract

With the rapid development of social economy, problems such as volatile organic compound (VOC) pollution and the excessive consumption of global petroleum resources have become increasingly prominent. People are beginning to realize that these problems not only affect the ecological environment, but also hinder the development of the organic polymer material industry based on raw fossil materials. Therefore, the modification and application of bio-based materials are of theoretical and practical significance. In this study, a series of vegetable oil-based acrylate prepolymers were synthesized by one-step acrylation using palm oil, olive oil, peanut oil, rapeseed oil, corn oil, canola oil, and grapeseed oil as raw materials, and the effect of different double bond contents on the product structure and grafting rate was investigated. Furthermore, the as-prepared vegetable oil-based acrylate prepolymers, polyurethane acrylate (PUA-2665), trimethylolpropane triacrylate (TMPTA), and photoinitiator (PI-1173) were mixed thoroughly to prepare ultraviolet (UV)-curable films. The effect of different grafting numbers on the properties of these films was investigated. The results showed that as the degree of unsaturation increased, the acrylate grafting number and the cross-linking density increased, although the acrylation (grafting reaction) rate decreased. The reason was mainly because increasing the double bond content could accelerate the reaction rate, while the grafted acrylic groups had a steric hindrance effect to prevent the adjacent double bonds from participating in the reaction. Furthermore, the increase in grafting number brought about the increase in the structural functionality of prepolymers and the cross-linking density of cured films, which led to the enhancement in the thermal (glass transition temperature) and mechanical (tensile strength, Young’s modulus) properties of the cured films.

No abstract

Abstract-In this paper, we provide a new architecture by using the programmable graphics processing unit (GPU) to move all electromagnetic computing code to graphical hardware, which significantly accelerates Graphical electromagnetic computing (GRECO) method. We name this method GPUECO. The GPUECO method not only employs the hidden surface removal technique of graphics hardware to identify the surfaces and wedges visible from the radar direction, but also utilizes the formidable of computing power in programmable GPUs to predict the scattered fields of visible surfaces and wedges using the Physical Optical (PO) and Equivalent Edge Current (EEC). The computational efficiency of the scattered field in fragment processors is further enhanced using the Z-Cull and parallel reduction techniques, which avoid the inconsistent branching and the addition of the scattered fields in CPU, respectively. Numerical results show excellent agreement with the exact solution and measured data and, the GPUECO method yields approximately 30 times faster results.

Cephalometric landmark detection is a crucial step in orthodontic and orthognathic treatments. To detect cephalometric landmarks accurately, we propose a novel multi-head attention neural network (CephaNN). CephaNN is an end-to-end network based on the heatmaps of annotated landmarks, and it consists of two parts, the multi-head part and the attention part. In the multi-head part, we adopt multihead subnets to gain comprehensive knowledge of various subspaces of a cephalogram. The intermediate supervision is applied to accelerate the convergence. Based on the feature maps learned from the multi-head Part, the attention part applies the multi-attention mechanism to obtain a refined detection. For solving the class imbalance problem, we propose a region enhancing (RE) loss, to enhance the efficient regions on the regressed heatmaps. Experiments in the benchmark dataset demonstrate that CephaNN is state-of-the-art with the detection accuracy of 87.61% in the clinically accepted 2.0-mm range. Furthermore, CephaNN is efficient in classifying the anatomical types and robust in a real application on a 75-landmark dataset.