A Method of Integrating Length Constraints into Encoder-Decoder Transformer for Abstractive Text Summarization

Text summarization aims to generate a concise version of the original text.The longer the summary text is,themore detailed it will be fromthe original text,and this depends on the intended use.Therefore,the problem of generating summary texts with desired lengths is a vital task to put the research into practice.To solve this problem,in this paper,we propose a new method to integrate the desired length of the summarized text into the encoder-decoder model for the abstractive text summarization problem.This length parameter is integrated into the encoding phase at each self-attention step and the decoding process by preserving the remaining length for calculating headattention in the generation process and using it as length embeddings added to theword embeddings.We conducted experiments for the proposed model on the two data sets,Cable News Network(CNN)Daily and NEWSROOM,with different desired output lengths.The obtained results show the proposed model’s effectiveness compared with related studies.

Study on embedded length of piles for slope reinforced with one row of piles

The embedded length of anti-slide piles for slope is analyzed by three-dimensional elastoplastic shear strength reduction method. The effect of embedded pile length on the factor of safety and pile behavior is analyzed. Furthermore, the effects of pile spacing, pile head conditions, pile bending stiffness and soil properties on length and behavior of pile are also analyzed. The results show that the pile spacing and the pile head conditions have significant influences on the critical embedded length of pile. It is found that the critical embedded length of pile, beyond which the factor of safety does not increase, increases with the decrease in pile spacing. The smaller the pile spacing is, the larger the integrity of the reinforced slope will be. A theoretical analysis of the slip surface is also conducted, and the slip surface determined by the pressure on piles, considering the influences of both soil and piles for slope, is in agreement with the ones in previous studies.

Laboratory pull-out tests on fully grouted rock bolts and cable bolts:Results and lessons learned

Laboratory pull-out tests were conducted on the following rock bolts and cable bolts:steel rebars,smooth steel bars,fiberglass reinforced polymer threaded bolts,flexible cable bolts,IR5/IN special cable bolts and Mini-cage cable bolts.The diameter of the tested bolts was between 16 mm and 26 mm.The bolts were grouted in a sandstone sample using resin or cement grouts.The tests were conducted under either constant radial stiffness or constant confining pressure boundary conditions applied on the outer surface of the rock sample.In most tests,the rate of displacement was about 0.02 mm/s.The tests were performed using a pull-out bench that allows testing a wide range of parameters.This paper provides an extensive database of laboratory pull-out test results and confirms the influence of the confining pressure and the embedment length on the pull-out response(rock bolts and cable bolts).It also highlights the sensitivity of the results to the operating conditions and to the behavior of the sample as a whole,which cannot be neglected when the test results are used to assess the bolt-grout or the grouterock interface.

Tensile bond anchorage properties of Australian 500N steel bars in concrete

In order to investigate the tensile bond anchorage properties of Australian 500N steel bars in concrete, 111 pullout tests were conducted. The precise bond slip values have been gained by using the laser displacement sensor with high resolution, including the complete bond-slip curves. How the main anchorage factors such as concrete strength, bar diameter （8, I0, 12, 16, 20, 24, 28, 32 and 36 mm） the concrete covered, embedded length and transverse reinforcement influencing the bond anchorage properties was studied under tensile condition. The process of the tensile force-slip failure for Australian 500N reinforcing steel can be divided into five stages： elastic stage, local slip stage, slip in ascent stage, slip in descent stage and remnant stage. The formula for calculating the tensile bond strength of Australian 500N reinforcing bar in concrete was proposed according to the test results, including the consistent model for tensile bond-slip relationship.

Numerical Analysis of the Composite Connection of Steel Joist Embedded in Concrete Girder

So far, numerous numerical studies have been conducted on the behavior of Composite Reinforced Concrete-Steel （RCS） beam-to-column connections. However, the lack of studies regarding the steel joist-concrete girder connection has yet to be addressed through comprehensive finite element methods to get an understanding of influential parameters. Hence, in this paper, composite connection of embedded steel joist in concrete girder is investigated with an appropriate finite element software, namely, ABAQUS. The validity of the proposed model is examined by the comparison made with the test data in literature. Results indicate that maximum bending capacity of the connection is achieved when embedment ratio is 1.78. Moreover, double web angles in the embedment region significantly reduce the embedment length required to achieve the maximum bending capacity. Finally, damage analyses show that bending capacity of concrete girder is slightly reduced in the connection zone.

DEM investigation of rock/bolt mechanical behaviour in pull-out tests

Rock bolt anchorage performance is crucial for tunnel support safety.We investigate the mechanical behaviour of reinforced rock and the bolts that reinforce it from the micro-scale to the macro-scale.Bolt pull-out tests were performed on soft rock using the distinct element method,in which a new contact model that considers bond size,is employed to constrain the main rock mechanical behaviour.The minimum sample width and height values for which the boundary effect can be neglected are first proposed through numerous tests on the influence of sample size on peak load and bond breakage.The influence of sample width is substantially greater than that of sample height.We then select an appropriate sample size to study the influence of bolt embedment length and confining pressure on the mechanical behaviours of the rock and bolt.The results show that increased rock bolt embedment length and confining pressure can increase the peak load;however,the bolt length effect is limited when exceeding the critical anchorage length.In cases without confining pressure,bond breakage occurs in the rock around the grout-rock interface and the breakage zone is rectangular,whereas in cases under confining pressure,the breakage zone presents an inverted cone shape.We use our results to discuss the influence of bond strength at the bolt-grout interface on the peak load and failure mode.The failure mode changes gradually from complex failure to single failure along the bolt-grout interface with decreasing interfacial bond strength.