Flow to Candidate: Temporal Knowledge Graph Reasoning With Candidate-Oriented Relational Graph.

Reasoning over temporal knowledge graphs (TKGs) is a challenging task that requires models to infer future events based on past facts. Currently, subgraph-based methods have become the state-of-the-art (SOTA) techniques for this task due to their superior capability to explore local information in knowledge graphs (KGs). However, while previous methods have been effective in capturing semantic patterns in TKG, they are hard to capture more complex topological patterns. In contrast, path-based methods can efficiently capture relation paths between nodes and obtain relation patterns based on the order of relation connections. But subgraphs can retain much more information than a single path. Motivated by this observation, we propose a new subgraph-based approach to capture complex relational patterns. The method constructs candidate-oriented relational graphs to capture the local structure of TKGs and introduces a variant of a graph neural network model to learn the graph structure information between query-candidate pairs. In particular, we first design a prior directed temporal edge sampling method, which is starting from the query node and generating multiple candidate-oriented relational graphs simultaneously. Next, we propose a recursive propagation architecture that can encode all relational graphs in the local structures in parallel. Additionally, we introduce a self-attention mechanism in the propagation architecture to capture the query's preference. Finally, we design a simple scoring function to calculate the candidate nodes' scores and generate the model's predictions. To validate our approach, we conduct extensive experiments on four benchmark datasets (ICEWS14, ICEWS18, ICEWS0515, and YAGO). Experiments on four benchmark datasets demonstrate that our proposed approach possesses stronger inference and faster convergence than the SOTA methods. In addition, our method provides a relational graph for each query-candidate pair, which offers interpretable evidence for TKG prediction results.

Research on the Mechanical Model and Hysteresis Performance of a New Mild Steel-Rotational Friction Hybrid Self-Centering Damper.

A mild steel-friction self-centering damper with a hybrid energy-dissipation mechanism (MS-SCFD) was proposed, which consisted of a mild steel, frictional, dual-energy-dissipation system and a disc spring resetting system. The structure and principle of the MS-SCFD were explained in detail while the restoring force model was established. The hysteretic behavior of the MS-SCFD under low-cycle reciprocating loading was modeled. Then, the influence of parameters such as the disc spring preload, the friction coefficient, and the soft-steel thickness on the mechanical properties of the MS-SCFD was investigated. The results indicate that the simulation results are basically consistent with the theoretical prediction results, with a maximum error of only 9.46% for the key points of bearing capacity. Since the MS-SCFD is provided with a hysteretic curve in the typical flag type, it will obtain the capacity of excellent self-centering performance. It can effectively enhance the stiffness, bearing capacity, and self-centering capability of the damper after the pre-pressure of the disc spring is increased. The energy-dissipation capacity of the MS-SCFD increases with the increase in the friction coefficient. However, it also increases the residual deformation of the MS-SCFD. The energy dissipation of the MS-SCFD is particularly sensitive to the thickness of mild steel. After being loaded, all components of the MS-SCFD are not damaged except for the plastic deformation caused by the yielding of the mild steel. The normal function of the MS-SCFD can be restored simply by replacing the mild steel plates after the earthquake. Therefore, it can significantly enhance the economy and applicability of the damper.

Study on Capillary Water Absorption of Waterborne-Polyurethane-Modified Recycled Coarse Aggregate Concrete.

The reuse of construction and demolition waste as a substitute for natural coarse aggregate in the production of recycled concrete has been widely used. In order to study the capillary water absorption performance of waterborne-polyurethane-modified recycled aggregate concrete (WPUMRC), the effects of different curing systems, polymer-cement ratios, and waterborne polyurethane addition methods on the cumulative water absorption and the rate of capillary water absorption of WPUMRC were analyzed, and through MIP tests, the micro modification mechanism of waterborne polyurethane in recycled concrete was analyzed. The results indicate that the optimal curing system for both DC (waterborne polyurethane is added separately from water) and HC (waterborne polyurethane is mixed with some effective water and then added) is the 14 d standard curing-14 d indoor natural drying curing system. Waterborne polyurethane can fill the pores and micro-cracks inside WPUMRC or interweave with the hydration products of cement to form a spatial network structure, reducing the porosity, and thereby improving the capillary water absorption performance of WPUMRC. Based on the MIP test results, the grey correlation method was used to establish the relationship between capillary water absorption and the pore structure of WPUMRC under the optimal curing system. In addition, the prediction model of capillary water absorption in recycled concrete was established according to the test results, which can be used to predict WPUMRC's capillary water absorption performance.

Dynamic Behavior of a Precast and Partial Steel Joint under Various Shear Span-to-Depth Ratios.

The dynamic behavior of a PPSRC beam-column joint is related to constraint effect, strength deterioration and strain rate effect. Then, it can be assessed by bearing capacity, stiffness degradation, displacement ductility and energy consumption. The results show that the increased strain rate causes growth in ring stiffness, bearing capacity and energy consumption of PPSRC beam-column joints. However, the influence of shear span-to-depth ratio on dynamic mechanical properties of PPSRC beam-column joints is more obvious than that of strain rate. Regardless of strain rate, the bearing capacity, initial stiffness, ring stiffness and energy consumption of PPSRC beam-column joints decrease as the shear span-to-depth ratio increases. Moreover, the ring stiffness under reverse direction is smaller than that the under forward direction at each displacement level. However, the stiffness degradation under a lower shear span-to-depth ratio is more obvious than that under a higher shear span-to-depth ratio. Moreover, the displacement ductility with a higher shear span-to-depth ratio is better than that with a lower shear span-to-depth ratio. Finally, the mechanical properties of PPSRC beam-column joints are affected by the extension length of partial steel plate, and the reasonable extension length of the partial steel plate in the column is affected by the shear span-to-depth ratio.

Strain Conditions Monitoring on Corroded Prestressed Steel Strands in Beams Based on Fiber Bragg Grating Sensors.

Fiber Bragg Grating (FBG) sensors, with excellent properties, have been widely adopted to monitor the mechanical parameters in civil engineering in recent years. On the other hand, the current study on fatigue performance of corroded prestressed steel strands is still limited, and this is mainly because the long-term strain conditions monitoring is difficult to conduct. Based on the aforementioned considerations, a total of six beam specimens were fabricated in this study. The loading mode of four points bending was adopted in the form of sinusoidal waves in the experiments. On basis of the experimental results, it can be concluded that the fatigue life of the beam decreases sharply with the increase of the corrosion rate of steel strands. Besides, with the increase of the maximum fatigue load, the fatigue life of the beam will decrease significantly. Furthermore, the existing fatigue damage of steel strand inside the beam before corrosion may further accelerate its fatigue failure. As a result, the fatigue life of the beam is reduced because of the stress concentration. Under the same external load, the strain increment and the residual strain of steel strands in the stages of loading and unloading after corrosion increase significantly compared with other stages, while the existing residual strain always shows an increasing trend at various static loading stages. Therefore, the corrosion of steel strand seriously affects not only its mechanical properties, but also its fatigue performance. Finally, the FBG sensors are capable of measuring the steel strand strain, as well as the long-term strain conditions.