Search

Through the adiabatic temperature rise experiment, the adiabatic temperature rise of concrete with hydration time was recorded. Based on the maturity degree theory, the relationship between the hydration degree of the concrete and the equivalent age was determined. Then, the hydration degree prediction model of the concrete's early elastic modulus and tensile strength was established. The local temperature and humidity of the concrete were measured by the shrinkage experiment, and based on the capillary water tension theory, a temperature-humidity prediction model for the early shrinkage of the concrete was designed. According to the ratio of the creep deformation and elastic deformation of concrete which were obtained through the restraint ring experiment, a model for predicting the early creep coefficient of concrete was proposed. Based on the coupling effect of “hydration-temperature-humidity,” a prediction model of early cracking risk coefficient of concrete under multifield coupling was proposed. Finally, several groups of slab cracking frame experiments were carried out, and the cracking risk prediction results of concrete were consistent with the actual situation, which indicated the correctness of the early cracking risk prediction model of concrete.

The reinforced concrete column after corrosion of chloride salt was subjected to magnesium phosphate cement (MPC)-carbon fiber reinforced plastics (CFRP) reinforcement and electrochemical chloride extraction (ECE) integration experiment. The axial compression behavior of the reinforced concrete column was tested, and then the numerical analysis was carried out. Two types of the corroded reinforced concrete columns were prepared for the test. The columns in type 1 were naturally corroded, and the columns in type 2 were electrified to accelerate the corrosion until the theoretical mass corrosion rate of steel bars reached 15%. Afterward, the MPC-CFRP was used to strengthen the corroded reinforced concrete columns and to remove the inside chlorine by electrochemistry method. The test results showed that the ultimate compressive capacity of the reinforced concrete columns strengthened by the MPC-CFRP increased by more than 22.63%. However, after ECE (current density, 3 A/m2), the ultimate compressive capacity of the reinforced concrete columns strengthened by MPC-CFRP reduced by 3.07% − 3.66%, and the chloride ion content at the interface between steel bar and concrete reduced by 78.52%, which postponed the corrosion of the steel bars. In addition, considering the complex nonlinear bond slip relationship between the steel bar and the concrete interface, a 3D mesoscopic finite element plastic damage model of the reinforced concrete columns strengthened by the MPC-CFRP was established. The numerical results were in good agreement with the test results, which indicated the 3D mesoscopic finite element numerical model can predict the compressive behavior of the reinforced concrete columns adequately.

Two kinds of layered double hydroxides (LDHs), MgAl-NO2 (N-LDH) and MgAl-CO3 (C-LDH), were incorporated to study the durability of mortar and concrete. The LDH contents of mortar were 1%, 2%, and 4% by mass and the LDH contents of concrete were 0.5%, 1%, 2%, 4%, respectively. The effect of LDHs on sulfate resistance of mortar was studied through dry-wetting cycle test, compressive strength test, and flexural strength test. In addition, the effects of LDHs on pore structure, chloride resistance, carbonation resistance, shrinkage, and creep of concrete were investigated by SEM, mercury injection test, XRD, chloride ion diffusion coefficient test, chloride salt corrosion depth test, carbonation depth test, shrinkage test, and creep test. The results showed that LDHs can improve the ability of resisting ion corrosion, carbonization, shrinkage, and creep, reduce the pore content, and optimize the pore structure of mortar and concrete to some extent. Moreover, 4% LDHs had a better effect on improving the durability of mortar and concrete compared to 0.5%, 1%, and 2% LDHs, and the effect of C-LDH was better than N-LDH.

The magnesium phosphate cement (MPC)-carbon fiber reinforced plastics (CFRP) strengthening and electrochemical chloride extraction (ECE) integration experiment was carried out for reinforced concrete beams corroded by sodium chloride. Then the bending performance of the reinforced concrete beams was analyzed by test and numerical simulation. The reinforced concrete beams of the control group were energized to accelerate corrosion until the theoretical corrosion rate of the steel bars reached 10%. The second group of the beams were strengthened, and the other three groups of the beams were strengthened and the inside chloride was removed by electrochemistry method with different dechlorination current densities. The bending test results showed that the bending bearing capacity of the strengthened concrete beam increased by 18.22%. The bending bearing capacity of the strengthened and dechlorinated beam increased by 15.11%, 13.25% and 9.76%, respectively. The chloride ion content at the interface between steel bar and concrete reduced by 68.68%–82.64%. In addition, the numerical simulation method of “standard cube test block-central pull out test block-reinforced concrete beam” was proposed. A 3D mesoscopic finite element plastic damage model of the reinforced concrete beams strengthened by MPC-CFRP was established. The numerical results were in good agreement with the experimental results.

OBJECTIVE To investigate seasonality and temporal trends in the incidence of NEC. STUDY DESIGN A retrospective cohort study from two tertiary NICUs in northern and central Connecticut involving 16,761 infants admitted over a 28-year period. Various perinatal and neonatal risk factors were evaluated by univariate, multivariate, and spectral density analyses. RESULTS Incidence of NEC was unchanged over the 28 years of study. Gestational age, birth weight, and birth-months (birth in April/May) were independently associated with stage II or III NEC even after adjusting for confounding factors (p

Search engine marketing (SEM) is an important channel for the success of e-commerce. With the increasing scale of catalog items, designing an efficient modern industrial-level bidding system usually requires overcoming the following hurdles: 1. the relevant bidding features are of high sparsity, preventing an accurate prediction of the performances of many ads. 2. the large volume of bidding requests induces a significant computation burden to offline and online serving. In this article, we introduce an end-to-end structure of a multi-objective bidding system for search engine marketing for Walmart e-commerce, which successfully handles tens of millions of bids each day. The system deals with multiple business demands by constructing an optimization model targeting a mixture of metrics. Moreover, the system extracts the vector representations of ads via the Transformer model. It leverages their geometric relation to building collaborative bidding predictions via clustering to address performance features' sparsity issues. We provide theoretical and numerical analyzes to discuss how we find the proposed system as a production-efficient solution.

Feature selection is a crucial problem in efficient machine learning, and it also greatly contributes to the explainability of machine-driven decisions. Methods, like decision trees and Least Absolute Shrinkage and Selection Operator (LASSO), can select features during training. However, these embedded approaches can only be applied to a small subset of machine learning models. Wrapper based methods can select features independently from machine learning models but they often suffer from a high computational cost. To enhance their efficiency, many randomized algorithms have been designed. In this paper, we propose automatic breadth searching and attention searching adjustment approaches to further speedup randomized wrapper based feature selection. We conduct theoretical computational complexity analysis and further explain our algorithms' generic parallelizability. We conduct experiments on both synthetic and real datasets with different machine learning base models. Results show that, compared with existing approaches, our proposed techniques can locate a more meaningful set of features with a high efficiency.

The compositions and contents of phases in cement paste soaked in sodium sulfate solution changing with time were investigated in this study. First of all, the porosities in cement paste specimens over time were measured by an in situ X-ray computed tomography (X-ray CT) test. Cement hydration and sulfate erosion occur simultaneously during the soaking process of cement paste specimens in sodium sulfate solution. Therefore, the hydration behaviour and the erosion of sulfates in cement paste were combined in a new way to establish an X-CT–hydration–deterioration model, which was used to calculate the variations of the volume fractions of the phases in cement paste. Subsequently, the volume fractions of each phase were used to calculate the elastic moduli of cement paste specimens with the bulk modulus, shear modulus of each phase and a self-consistent (SC) method. Afterward, the calculated elastic moduli were compared with the experimental results obtained by uniaxial compression test and the error values were 1% for 28 d, 2·5% for 88 d and 5% for 148 d, which were all within 5%, indicating that the advanced X-CT–hydration–deterioration model combined with the SC method can successfully predict the elastic modulus of cement paste eroded in sodium sulfate solution.

Biomaterials and biomedical implants have revolutionized the way medicine is practiced. Technologies, such as 3D printing and electrospinning, are currently employed to create novel biomaterials. Most of the synthesis techniques are ad-hoc, time taking, and expensive. These shortcomings can be overcome greatly with the employment of computational techniques. In this paper we consider the problem of bone tissue engineering as an example and show the potentials of machine learning approaches in biomaterial construction, in which different models was built to predict the elastic modulus of the scaffold at given an arbitrary material composition. Likewise, the methodology was extended to cell-material interaction and prediction at an arbitrary process parameter.

The large model size, high computational operations, and vulnerability against membership inference attack (MIA) have impeded deep learning or deep neural networks (DNNs) popularity, especially on mobile devices. To address the challenge, we envision that the weight pruning technique will help DNNs against MIA while reducing model storage and computational operation. In this work, we propose a pruning algorithm, and we show that the proposed algorithm can find a subnetwork that can prevent privacy leakage from MIA and achieves competitive accuracy with the original DNNs. We also verify our theoretical insights with experiments. Our experimental results illustrate that the attack accuracy using model compression is up to 13.6% and 10% lower than that of the baseline and Min-Max game, accordingly., Machine Learning (cs.LG); Cryptography and Security (cs.CR); Machine Learning (stat.ML)

In this study, the cement paste corroded in sodium sulfate solution was regarded as a three-phase composites composing of unhydrated cement particles, reaction products and pores on meso-scale. First of all, based on the results of the in-situ X-ray Computed Tomography (X-ray CT) test, a three-dimensional (3D) meso-scale finite element (FE) model of the cement paste specimen was established. Subsequently, homogenization methods were used to calibrate the mechanical parameters (elastic modulus, Poisson's ratio and strength) of the model and the constitutive relations of unhydrated cement particles and reaction products were modified. Finally, the failure process of the cement paste specimens conducted by the uniaxial compression test was simulated. The results showed that the compressive strength of the cement paste specimens experienced a process of increasing first and then decreasing with the increase of corrosion time, caused by the combination of cement hydration and sulfate ion corrosion. Additionally, the simulation results and the uniaxial compression test results were in good agreement, indicating that the modified constitutive relations were effective and the 3D meso-scale numerical simulation can well predict the uniaxial compression test of cement paste specimen under the corrosion of sodium sulfate.

In this research, Magnesium Phosphate Cement (MPC) was innovatively used to bond with Carbon Fiber Reinforced Plastic (CFRP) in order to form the MPC-CFRP as a composite material, adopted both for chloride ions extraction and reinforcement of concrete columns. First of all, a series of tests were conducted to evaluate the feasibility of the MPC-CFRP as anode of Electrochemical Chloride Extraction (ECE) system, including chloride ions concentration, Scanning Electron Microscope (SEM) and tensile strength. Then, the MPC-CFRP was used to wrap around reinforced concrete column for the sake of reinforcement and extraction of chloride ions. The test results indicated that the chloride extraction efficiency and the tensile resistance of the MPC-CFRP electrode were superior to the control group. The ECE process can decrease the interface bonding strength between the concrete and the steel rebar while the compression resistance of the columns could be increased remarkably. Therefore, the MPCCFRP composite material is capable of achieving the dual functions of strengthening and repairing the reinforced concrete construction.

Widespread availability of next-generation sequencing (NGS) technologies has prompted a recent surge in interest in the microbiome. As a consequence, metagenomics is a fast growing field in bioinformatics and computational biology. An important problem in analyzing metagenomic sequenced data is to identify the microbes present in the sample and figure out their relative abundances. In this article we propose a highly efficient algorithm dubbed as “Hybrid Metagenomic Sequence Classifier” (HMSC) to accurately detect microbes and their relative abundances in a metagenomic sample. The algorithmic approach is fundamentally different from other state-of-the-art algorithms currently existing in this domain. HMSC judiciously exploits both alignment-free and alignment-based approaches to accurately characterize metagenomic sequenced data. To demonstrate the effectiveness of HMSC we used 8 metagenomic sequencing datasets (2 mock and 6 in silico bacterial communities) produced by 3 different sequencing technologies (e.g., HiSeq, MiSeq, and NovaSeq) with realistic error models and abundance distribution. Rigorous experimental evaluations show that HMSC is indeed an effective, scalable, and efficient algorithm compared to the other state-of-the-art methods in terms of accuracy, memory, and runtime.Availability of data and materialsThe implementations and the datasets we used are freely available for non-commercial purposes. They can be downloaded from: https://drive.google.com/drive/folders/132k5E5xqpkw7olFjzYwjWNjyHFrqJITe?usp=sharing

Widespread availability of next-generation sequencing (NGS) technologies has prompted a recent surge in interest in the microbiome. As a consequence, metagenomics is a fast growing field in bioinformatics and computational biology. An important problem in analyzing metagenomic sequenced data is to identify the microbes present in the sample and figure out their relative abundances. Genome databases such as RefSeq and GenBank provide a growing resource to characterize metagenomic sequenced datasets. However, both the size of these databases and the high degree of sequence homology that can exist between related genomes mean that accurate analysis of metagenomic reads is computationally challenging. In this article we propose a highly efficient algorithm dubbed as "Hybrid Metagenomic Sequence Classifier" (HMSC) to accurately detect microbes and their relative abundances in a metagenomic sample. The algorithmic approach is fundamentally different from other state-of-the-art algorithms currently existing in this domain. HMSC judiciously exploits both alignment-free and alignment-based approaches to accurately characterize metagenomic sequenced data. Rigorous experimental evaluations on both real and synthetic datasets show that HMSC is indeed an effective, scalable, and efficient algorithm compared to the other state-of-the-art methods in terms of accuracy, memory, and runtime.

The sustainable green building material magnesium phosphate cement (MPC) is widely used in the fields of solidifying heavy metals and nuclear waste and repair and reinforcement. Magnesium potassium phosphate hexahydrate (MKP) is the main hydration product of MPC. The transport of water and ions in MKP nanochannels determines the mechanical properties and durability of MPC materials. Herein, the interface models of MKP crystals with sodium chloride solution in the [001], [010] and [100] direction were established by molecular dynamics. The interaction of the MKP interface with water and ions was studied and the durability of MPC in sodium chloride solution was explained at the molecular level. The results show that a large number of water molecules are adsorbed on the MKP crystal surface through hydrogen bonds and Coulomb interactions; the surface water molecules have the bigger dipole moment and the dipole vector of most of the water molecules points to the solid matrix, when the crystal surfaces of the three models all show hydrophilicity. In addition, plenty of sodium ions are adsorbed at the MKP interface, and some potassium ions are desorbed from the matrix. In the MKP[001] model, the amount of potassium ions separated from the matrix and diffused into the solution is the highest and the interface crystal is the most disordered. Due to the attack of water and ions, the K–Os bond loses its chemical stability and the order of the MKP crystal is destroyed, which explains the decline of MPC performance after the erosion of sodium chloride solution at the molecular level. Besides, in the three models, the Na–Cl ion bond is more unstable than the K–Cl ion bond due to the smaller radius of the sodium atom. The stability of ionic bonds in the models is as follows: MKP[010] > MKP[100] > MKP[001].

Engineered Cementitious Composite (ECC) is an advanced composite material with strain-hardening and multiple-cracking behaviors. In this study, two types of ECC materials were troweled on masonry infill walls strengthened with expansion bolt and interfacial agent. The seismic performance of the unreinforced and the ECC reinforced masonry structures was evaluated by the reversed cyclic loading test. The results showed that the application of the ECC on the masonry infill walls can improve the ultimate bearing capacity (in plane), ductility, stiffness and accumulative energy dissipation of the structure. Compared with the masonry structure with one side reinforced by the ECC, the structure with both sides reinforced by the ECC could better improve the ultimate bearing capacity, stiffness and cumulative energy dissipation with little difference on the ductility improvement. Moreover, the expansion bolt and the interfacial agent used to improve the connection between the ECC and the masonry structure have the ability to greatly increase the mechanical properties of the structure under cyclic loading.

This paper presented an experimental-computational approach to investigate compressive strength of magnesium phosphate cement (MPC) with nanoindentation technique and finite element analysis. Firstly, the compressive strength, macroscopic elastic modulus, porosity, microscopic elastic modulus, chemical compositions and contents of the MPC were tested by press, MIP, nanoindentation, XRD and SEM-EDS. Secondly, the microscopic elastic modulus of magnesium potassium phosphate hexahydrate (MKP) and magnesium oxide (MgO) as the components of the MPC were calculated by homogenization method. Then, the MPC was scanned by X-ray Computed Tomography (X-CT) to obtain the 3D structure adopted as the finite element model. Finally, the input parameters of the model were calibrated based on the results of homogenization and the modified constitutive relations to simulate the uniaxial compression test of the MPC. It was found that the simulated load - displacement results were in good agreement with the experimental results.

This paper computationally and experimentally investigated the microwave healing performance of graphite (flake graphite and exfoliated graphite nanoplatelet (xGNP)) modified asphalt mixture. The original fracture energy and strength were first measured for the modified asphalt mixture through the disk-shaped compact tension test. Then the micro-wave healing performance (recovered fracture energy and strength) were further examined. All these were enhanced with the added carbon materials. The recovered fracture strengths were also compared with FE cohesive zone model (CZM) simulation with digital image correlation (DIC) calibrated parameters. The predicted recovered fracture strength had good agreement with the experimental measurement.

A wireless oceanographic system is an underwater wireless-networked sensing system for oceanographic data collection, remote monitoring, and control. It represents a cutting-edge technology that could eliminate the need of long and expensive subsea cables while featuring real-time acquisition, flexible, and convenient deployment. Mainly powered by batteries, the wireless oceanographic system calls for power efficient deployment plans to extend the system lifetime. In this work, we first propose an architecture of wireless oceanographic system and then introduce a framework for the deployment planning. Specifically, we investigate the power consumption profile of the sensor nodes in the system. We employ a battery model for more accurate power estimation. Further, we analyzed the system lifetime with the consideration of the sensor node interactions. As a case study, the framework is applied to the Ocean-TUNE Long Island Sound (LIS) testbed, a real wireless oceanographic system with practical system configurations, and real data.

This paper presented an integrated computational-experimental approach for evaluating elastic modulus of cement paste corroded in brine solution on nanoscale. In the first place, the cured specimens were corroded in brine solution at room temperature for 90 days. Subsequently, the elastic moduli of 32 points of the corroded cement paste specimen were measured by nanoindentation test. Then, a three-dimensional nanoindentation model was established with finite element method and modified constitutive relation. The input parameters of the model were calibrated based on the test results and linear interpolation method. Finally, the difference between the simulation results and the experimental results was compared with good agreement. Therefore, the nanoindentation model established in this paper can predict the nanoindentation experiment well and the modified constitutive relation is effective.

The electrochemical accelerated corrosion of reinforced concrete (RC) columns was carried out with the target corrosion rate of longitudinal rebar as 15%, then the magnesium phosphate cement (MPC)-carbon fiber reinforced plastics (CFRP) as strengthening approach and electrochemical chloride extraction (ECE) method were integrated to process the RC columns. Among them, when the dechlorination period was 35d, the current density of dechlorination was selected as 1 A/m2, 2 A/m2, and 3 A/m2; when the periods of dechlorination were 21d, 35d and 49d respectively, the dechlorination current density was set as 2 A/m2. Finally, the seismic performance and dechlorination efficiency of RC columns under different dechlorination current densities and dechlorination periods were mainly analyzed by failure mode, hysteresis performance, lateral strength, ductility performance, stiffness degradation, residual deformation plus energy dissipation. Compared with the corroded RC columns, the cumulative energy dissipation capacity of the RC columns enhanced by MPC-CFRP promoted by 101.33%−206.36%. When the dechlorination period was 35d, the seismic performance of the RC columns dechlorided by the current density of 1 A/m2 and 2 A/m2 did not decrease significantly, and the seismic performance of RC columns with the dechlorination current density of 3 A/m2 greatly reduced. When 2 A/m2 was regarded as the dechlorination current density, the seismic performance of concrete columns with the dechlorination periods of 21d and 35d did not decrease significantly, and the seismic performance of the concrete columns with the dechlorination periods of 49d greatly reduced. However, it was worth noting that the chloride ion content of the interface between rebar and concrete abated significantly with the growth of the current density and the dechlorination period.

How the multi-walled carbon nanotubes (MWCNTs) influenced the Young’s modulus of C-S-H was investigated at multi-scale in this paper. The functionalized MWCNTs dispersed by ultrasonication were mixed with CaO and nanosilica to prepare MWCNTs/C-S-H composites with calcium to silicon ratios (C/SD) of 1.5 and 2.0 respectively, and the contents of MWCNTs were 1% and 3% of the total mass of CaO and nanosilica. The Young’s modulus of MWCNTs/C-S-H was evaluated by atomic force microscope (AFM) at multi-scale. The effect mechanism of MWCNTs on Young’s modulus of C-S-H was analyzed by various testing techniques and homogenization method. The BSE, TEM and N2 adsorption tests results showed that MWCNTs acted as a bridge to reduce the number of cracks caused by shrinkage of C-S-H. Meanwhile, MWCNTs can reduce the porosity of C-S-H by filling the pores. The AFM consequence illustrated that Young’s modulus of C-S-H was increased by MWCNTs at both nano and micro scales. At the nano scale, MWCNTs increased the Young’s modulus of C-S-H by increasing the mean chain length (MCL) of C-S-H. At the micro scale, MWCNTs increased the Young’s modulus of C-S-H by increasing the Young’s modulus of C-S-H globules and reducing the porosity and the amount of cracks.

In this paper, the rheological mechanism of fresh concrete material was investigated and the method of evaluating its rheological property was optimized by combining the slump test of five groups of fresh concrete and lattice Boltzmann method (LBM). The rheological parameters of fresh concrete samples based on Herschel-Bulkley (H–B) model and Bingham model were obtained by rheometer and equation derivation with nonlinear and linear fitting approaches. The comparison of Bingham and H–B models in the simulation results of the slump test demonstrated that the H–B model was more accurate in expressing the rheological characteristics of the fresh concrete. In the Bingham model, the errors of slump s, spread sf and the time of spread reaching 500 mm T500 ranged from 2.00% to 39.29%. The errors of s and sf in the H–B model were less than 2%, and the error of T500 was about 5%. Afterwards, the results of the slump test were predicted based on the H–B model. The effects of rheological parameters (power-law index consistency index and yield stress) on velocity, viscosity, shear strain rate, flow rate and external contour were revealed.

This paper presents an efficient experimental-numerical analysis of fracture mechanics in magnesium phosphate cement (MPC) based on the structural and mechanical properties of its constituents including potassium magnesium phosphate hexahydrate (MKP), magnesium oxide (MgO) and pores. At micro-scale, the fracture energy and material strength of solid phases were obtained relying on the combination of nanoindentation experiments and simulation. The X-ray computed tomography (XCT) image-based 3D meso-structure model of MPC beam was generated and incorporated with the finite element cohesive zone model to analyse the fracture process of MPC beam under three-point bending. The unknown fracture parameters of cohesive elements at the interface between MKP and MgO were determined via the model calibration process conditional to the experimental data in terms of relationship between macro-load and crack mouth opening displacement. The cohesive strengths obtained for MKP, MgO and MKP-MgO were found to be 5.8, 106 and 24 MPa, respectively. In the same order, the fracture energies were0.02, 0.08 and 0.04 N/mm, respectively.

This paper presents an integrated investigation of elastic modulus of corroded cement paste by experiments and homogenization calculation. First of all, cured cement paste samples were immersed in brine solution with specific chemical compositions for 60 d. Then the elastic moduli of the samples were measured by the nanoindentation test. The test results demonstrated that the elastic moduli of the nanoindentation points decreased, compared with the original cement paste. In addition, X-ray Diffraction (XRD) and Scanning Electron Microscope-Energy Dispersive Spectroscope (SEM-EDS) were combined to analyze the compositions of the corroded cement paste quantitatively. As a result, volume fractions of nine compositions of the corroded cement paste in each nanoindentation point were calculated. Afterward, an advanced homogenization method combining Mori-Tanaka (MT) method with self-consistent (SC) method was proposed to calculate the elastic moduli of the nanoindentation points. The calculation results had good agreement with the experimental results due to small relative difference and standard deviation. The back-scattered electron (BSE) image was used to analyze the elastic moduli distribution of the nanoindentation points as well. The result indicated that the elastic moduli of the nanoindentation points were related to the elastic moduli of their main compositions.

Fly ash (FA) is one of the common supplementary cementitious materials used in cement and concrete. Because of the complexity of morphology, component, hydration and other factors of FA, it is very difficult to measure the content of FA in hardened concrete. According to the characteristic of the unburned carbon in FA neither dissolving in concrete nor participating in the chemical reaction, three different kinds of FA were used to prepare FA-cement pastes with different volumes. Then selective dissolution method and thermal analysis were used to quantitatively analyze the content of FA in hardened paste. The results indicated that the differences of testing data were less than 1.5%, compared with the theoretical values by the method of selective dissolution combined with thermal analysis. Therefore, it is concluded that the method of selective dissolution combined with the thermal analysis can measure the content of FA in hardened FA-cement system more accurately.

This paper presents a laboratory investigation of the thermal, electrical, rheological and mechanical properties and performance of control and graphite (flake graphite and exfoliated graphite nanoplatelet (xGNP)) modified asphalt binder and mixture. For the graphite modified asphalt binder, the rolling thin film oven (RTFO) test and pressure aging vessel (PAV) test were utilized to simulate the short-term and long-term aging process of control and graphite modified asphalt binder, respectively. The bending beam rheometer (BBR) test and dynamic shear rheometer (DSR) test were conducted to evaluate the rheological properties of the control and graphite modified asphalt binder at low and high temperatures, respectively. The Fourier transform infrared spectroscopy (FTIR) was used to evaluate the oxidation group content in these asphalt binders. The thermal conductivity of the graphite modified asphalt binder increased with graphite content. For the graphite modified asphalt mixtures, both thermal and electrical conductivities also increased with added graphite modifiers. The measured dynamic modulus results of mixture performance tests indicated that the added graphite particles were capable of increasing their moduli at both high and low temperatures. The Hamburg wheel tracking device (HWTD) test results also showed an improved rutting resistance. As a result, the graphite modified asphalt mixture can improve multiple physical properties and high-temperature performance as promising conductive materials for many applications.

This presented research investigated the physical properties and accelerated sunlight-healing performance of graphite modified asphalt materials. Two types of graphite materials, flake graphite and exfoliated graphite nanoplatelets (xGNP), were added to asphalt with different contents by weight. Asphalt binder tests were conducted to evaluate the performance of modified asphalt, including rotational viscosity, light absorbance, and thermal conductivity. The tests results of graphite modified asphalt showed increased viscosities, decreased activation energies, and increased light absorbance and thermal conductivity in comparison to the control asphalt. Afterwards, the graphite modified asphalt mixture beams (5% flake graphite and 2% xGNP) were prepared and used for cyclic fracture-light healing tests. The digital image correlation (DIC) was utilized to characterize the displacement changes in the fracture zone during the light healing process. The DIC results indicated that the improved healing performance of graphite modified asphalt mixtures and decreased healing displacement with light healing cycles. Finally, the measured recovered strength after each cycle was used to evaluate the healing performance of both the graphite modified and control samples. The results of cyclic fracture-light healing tests indicated that the graphite modified asphalt materials have significantly improved healing performance and therefore these materials have promises in promoting new light healing applications.

Feature selection is a core problem in machine learning. It plays an important role in making efficient and explainable machine-driven decisions. Embedded feature selection methods, such as decision trees and LASSO, suffer from learner dependency and cannot be applied well to many popular learners. Wrapper methods, which fit arbitrary learning models, are receiving growing interests in many scientific fields. In order to effectively search relevant features in wrapper methods, many randomized schemes have been proposed. In this paper, we present efficient randomized feature selection algorithms empowered by automatic breadth searching and attention searching adjustments. Our schemes are generic and highly parallelizable in nature and can be easily applied to many related algorithms. Theoretical analysis proves the efficiency of our algorithms. Extensive experiments on synthetic and real dataset show that our techniques achieve significant improvements in the selected features' quality and selection time.

The hydration rate, microstructure, mechanical properties and durability of Magnesium phosphate cement (MPC) prepared by different kinds of phosphates are quite different. These differences in macroscopic properties are essentially determined by microscopic chemical reactions. Therefore, in this paper, the adsorption mechanism of MgO interface to water and ions in the initial hydration process of MPC and the hydration structure of different ions were discussed at the molecular level. Herein, the structure and dynamics properties of water, NH4+, Na+, K+ and Cl− ions at MgO interface were analyzed by molecular dynamics method. The results show that the water molecules near the MgO interface can be connected to the matrix oxygen through hydrogen bonding and have a Coulomb interaction with the Mg on the interface. Therefore, the water molecules accumulate and stratify at the interface and the density distribution curve of water molecules forms a peak near the interface. This explains the stage of MgO adsorption of water molecules during the initial hydration of MPC. Besides, a large amount of ammonium ion, sodium ion and potassium ion are adsorbed on the MgO surface, and the cation density distribution curve also forms a peak near the interface. In addition, the radius of sodium ion is smaller and the Na-Os bond is stronger than that of NH4+ and K+ (Os represents the oxygen atom in MgO). Hence, the number of sodium ions adsorbed to the interface is the largest, the radial distribution function (RDF) of Na-Os forms a high-strength and sharp peak, The Os coordination number of Na+ is more than that of NH4+ and K+. This also explains the experimental phenomenon that sodium magnesium phosphate cement has a faster hydration rate and shorter setting time than ammonium magnesium phosphate cement and potassium magnesium phosphate cement. Due to the adsorption of the interface, the mean square displacement of cations has decreased to some extent compared with that in the corresponding pure solution models. This study provides a basic understanding of the initial hydration mechanism of MPC at the molecular level.

Numerical simulation has been widely employed in investigating stone-based composite behaviors. Model reconstruction is a prerequisite step to conduct numerical simulations. The objective of this paper is to provide a comprehensive review on the development of heterogeneous model reconstruction of stone-based composites, in particular, Portland cement concrete and asphalt mixtures. The numerical models mainly refer to discrete element models and finite element models. According to the review results, there are two types of heterogeneous models for stone-based materials based on modeling methods, the image based model and computer generated model. The image based model is obtained through image processing on X-ray or optical images by identifying the different phases in the composite. The computer generated model is obtained by placing computer generated aggregate particles into asphalt or cement matrix. Some subcategories for both the image based and computer generated models were detailed as well. The pros and cons of the image based models and computer generated models were also stated and some suggestions were provided. In general, image based models can capture the detailed geometrical information of each phase, but it is costly and time consuming, and the model accuracy is highly dependent on the image processing techniques. Compared to image based model, the computer generated model is more cost effective and much easier to implement, but the main concern is the accuracy of the aggregate model shapes. Future research directions are also provided based on the authors’ views.

This paper presents an integrated computational–experimental approach for evaluating recovered fracture strength after induction healing of asphalt concrete beams containing steel wool fibers. The cyclic beam fracture-healing test was utilized to measure the recovered fracture strength of asphalt concrete samples at temperatures of 60 °C, 80 °C, and 100 °C as reported in previous paper. The 2D multiphase bilinear cohesive zone models (CZM) were employed to predict the fracture strength of original and healed (after fracture) beam samples. For the original sample, the CZM with measured fracture parameters has good prediction on both crack path and sample force–displacement relations. With regards to healed samples, the healed fracture energy was proportionally calibrated with peak stress ratios based on bilinear cohesive law. All the nine tested samples were simulated with eight fracture-healing test cycles. The predicted fracture behaviors of these samples showed favorable comparison with the measured fracture strength after each fracture-healing cyclic test. These comparison study indicated that the integrated computational–experimental tools can be applied to evaluate the sample healing performance.

At present, it is unclear how the microstructure and multi-scale mechanical properties of concrete change after ultra-high-rise pumping. To this end, this study applied BSE image analysis method and nanoindentation test to investigate the influence of ultra-high-rise pumping on the porosity, thickness and elastic modulus of interface transition zone (ITZ) in concrete, firstly. The results showed when increasing the pumping height, the elastic modulus and uniaxial compressive strength of concrete and mortar increased slightly, the thickness and porosity of ITZ decreased, and the elastic modulus of ITZ increased. Secondly, based on the porosity test results of ITZ and paste, the volume fractions of pores, mortar and equivalent body (aggregate + ITZ) in concrete with different pumping heights were determined, and then a 3D model of concrete was established based on the X-ray Computed Tomography test results. Thirdly, according to the mechanical properties of the mortar before and after ultra-high-rise pumping, the constitutive relationship of the mortar in the ultra-high-rise pumping concrete was modified, and the mechanical parameters of the equivalent body in the model considered different pumping heights were calculated by the homogenization method. Finally, the uniaxial compression of ultra-high-rise pumping concrete was accurately simulated using the finite element method. On this basis, according to the relationship between pumping measurement index (PMI) and the performance of ultra-high-rise pumping concrete, the uniaxial compressive stress–strain curve of ultra-high-rise pumping concrete was precisely predicted.

This chapter aims to summarize the current research and findings on performance enhancement of the nanomodification on asphalt mixture. The study first summed up the main nanomaterials applied for the modification, including the nanomineral and nanocarbon materials. The dispersion protocol for these nanomaterials into the asphalt binder was further introduced, including mixing temperature, shear speed, and mixing time. Then the enhanced performance of the nanomodified asphalt mixture was concluded based on the reference review, including mechanical, electrical, thermal, and optical performance. It is found that the added nanomaterials can enhance the stiffness and high temperature stability of the asphalt mixture and reduce its permanent deflection. Similar findings on mechanical performance can be found for the asphalt mixture with different nanomaterials mainly due to the similar reinforcing mechanism (filler effect). The added carbon nanomaterials can further enhance the thermal conductivity, electrical conductivity, and absorbance for sunlight, which can lead to promising self-sensing and self-healing materials.

The location-based services for Internet of Things (IoTs) have attracted extensive research effort during the last decades. Wi-Fi fingerprinting with received signal strength indicator (RSSI) has been widely adopted in vast indoor localization systems due to its relatively low cost and the potency for high accuracy. However, the fluctuation of wireless signal resulting from environment uncertainties leads to considerable variations on RSSIs, which poses grand challenges to the fingerprint-based indoor localization regarding positioning accuracy. In this paper, we propose a top-down searching method using a deep reinforcement learning agent to tackle environment dynamics in indoor positioning with Wi-Fi fingerprints. Our model learns an action policy that is capable to localize 75% of the targets in an area of 25000m2 within 0.55m.

Singular Value Decomposition (SVD) is a fundamental problem in linear algebra and it has applications in many fields. For example, as a data reduction technique, SVD is widely used for Principal Component Analysis (PCA). It also has extensive applications in the machine learning domain. For example, it is widely used in feature selection, data clustering, classification, and regression. Traditional sequential SVD algorithms take cubic time which may not be affordable especially when the input matrix is large. Therefore, there is a pressing need to develop efficient sequential and parallel SVD algorithms. Jacobi iteration, as one of the most common SVD algorithms, has attracted a lot of attention. Among many algorithms that have been proposed, there are some relaxation schemes that have been proven to be very effective in parallel Jacobi-SVD computing. Such relaxation schemes are platform and architecture independent. In this paper, we propose some novel techniques to further generalize and speedup these relaxation schemes. Theoretical analysis proves the efficiency of our algorithms. Experimental results also show that our algorithms outperform the existing ones.

This paper presents an integrated computational-experimental approach for evaluating recovered fracture strength after light-healing of graphite-modified asphalt mixture beams. The cyclic fracture and light-healing tests were conducted to measure the recovered fracture strength of the asphalt mixture samples as reported in another paper. A two-dimensional (2D) multiphase bilinear cohesive zone model (CZM) was employed to predict the original and recovered fracture strength of the asphalt mixture samples. The digital image correlation (DIC) method was used to analyze the crack displacement variation of the samples during the fracture and light-healing processes. The relative strain ratio from DIC, originally measured fracture energy, and peak separation stress were combined to calibrate the recovered fracture energy of beams after light-healing processes. Two input parameters, calibrated fracture energy and peak separation stress, were used to simulate the sample fracture behavior and predict the r...

The purpose of this study is to evaluate the healing capacity of electroactive asphalt mastic and concrete beam samples with induction heating. The electrically conductive steel wool fibers were mixed with asphalt materials to heat the surrounding binders through induction energy. By introducing Newtonian binder flow in the heated asphalt materials, the microcrack healing performance can be significantly improved with the accelerated process. To investigate the induction healing performance, asphalt mastic and concrete beam samples were prepared by incorporating Type 1 steel wool fibers with an approximate length of 6.5 mm. Subsequently, the mastic beams were tested with fracture-healing cycles under the three-point bending test and induction healing process. Meanwhile, the concrete beams were tested with fracture-healing cycles using the modified three-point bending test with an elastic foundation support and healing procedure. Prior to the fracture tests, all samples were conditioned in the freezer for 6 h at −20 °C to limit the viscoelastic and unrecovered deformation. The samples were loaded until the peak value was obtained for fracture tests. During the healing process, the samples were heated at different temperatures (60 °C, 80 °C and 100 °C) within a short period of time. The result shows that the mastic sample can be fully healed at these three temperatures based on the recovered peak loads from the cyclic test. For the asphalt concrete beams, the test results showed that the healing performance increases with the heating temperatures (60 °C, 80 °C and 100 °C). The temperature distribution in the samples at the end of the healing procedure is also captured. Overall, it was found that the asphalt mixture samples still maintained at least half of the original fracture strength after six fracture-healing cycles. The experimental results indicate that the induction healing techniques have promises in elongating the pavement service life.

The Earth is a water planet. The ocean is used for nature resource exploitation, fishery, etc., and it also plays critical roles in global climate regulation and transportation. Consequently, it is extremely important to keep track of its condition. And thus ocean observation systems have received increasing attentions.