Compressible Model Research Articles

Swelling Induced Hardening and Degradation Induced Slow-Expansion Hydrogel for a Modified Cervical Spinal Cord Compression Animal Model.

This study presents the development of a polyurethane hydrogel (PUG) for use in a chronic cervical spinal cord compression animal model, leveraging microphase separation and dynamic covalent bonds to achieve swelling induced hardening and degradation-induced slow expansion. PUG-SS and PUG-SS-60% were synthesized with varying disulfide bond concentrations, offering controllable degradation rates and mechanical properties. The hydrogels demonstrated significant swelling-induced hardening and maintained compression above the cervical spinal cord's intrinsic modulus. MRI and histopathological analyses confirmed effective and sustained spinal cord compression, with PUG-SS-60% showing prolonged effects. Behavioral tests, including the BBB locomotor scale, von Frey pain test, and catwalk gait analysis, indicated quicker motor function recovery with PUG-SS and sustained compression with PUG-SS-60%. In vitro cytotoxicity assays showed no significant hydrogel-induced cell death. This study underscores the potential of PUG-SS-60% for providing controlled, sustained compression in chronic spinal cord compression models, paving the way for advanced nonsurgical treatment strategies and improved understanding of degenerative cervical myelopathy (DCM) pathology.

On Elastic Language Models

Large-scale pretrained language models have achieved compelling performance in a wide range of language understanding and information retrieval tasks. While their large scales ensure capacity, they also hinder deployment. Knowledge distillation offers an opportunity to compress a large language model to a small one, in order to reach a reasonable latency-performance tradeoff. However, for scenarios where the number of requests (e.g., queries submitted to a search engine) is highly variant, the static tradeoff attained by the compressed language model might not always fit. Once a model is assigned with a static tradeoff, it could be inadequate in that the latency is too high when the number of requests is large, or the performance is too low when the number of requests is small. To this end, we propose an elastic language model ( ElasticLM ) that elastically adjusts the tradeoff according to the request stream. The basic idea is to introduce a compute elasticity to the compressed language model, so that the tradeoff could vary on-the-fly along a scalable and controllable compute. Specifically, we impose an elastic structure to equip ElasticLM with compute elasticity and design an elastic optimization method to learn ElasticLM under compute elasticity. To serve ElasticLM , we apply an elastic schedule. Considering the specificity of information retrieval, we adapt ElasticLM to dense retrieval and reranking, and present an ElasticDenser and an ElasticRanker, respectively. Offline evaluation is conducted on a language understanding benchmark GLUE, and several information retrieval tasks including Natural Question, Trivia QA and MS MARCO. The results show that ElasticLM along with ElasticDenser and ElasticRanker can perform correctly and competitively compared with an array of static baselines. Furthermore, an online simulation with concurrency is also carried out. The results demonstrate that ElasticLM can provide elastic tradeoffs with respect to varying request stream.

Growth-adaptive distillation compressed fusion model for network traffic identification based on IoT cloud–edge collaboration

The development of the Internet of Things (IoT) has led to the rapid growth of the types and number of connected devices and has generated large amounts of complex and diverse traffic data. Traffic identification on edge servers solves the real-time and privacy requirements of IoT management and has attracted much attention, but still faces several problems: (1) traditional machine learning (ML) models rely on artificially constructed features, and the existing deep learning (DL) traffic identification models have reached their performance limit; and (2) insufficient computing resources of edge servers limit the possible improvement in the performance of deep learning models by increasing the number of parameters and structural complexity. To address these issues, we propose a lightweight fusion model. First, the Network-in-Network (NiN) model and Random Forest (RF) model are used on the cloud server to construct a traffic identification fusion model. The excellent representation extraction capability of the NiN compensates for the RF’s dependence on manual feature extraction, and its modular structure is suitable for the subsequent model compression operations. Then, the NiN was distilled. We propose Growth-Adaptive Distillation to lightweight the NiN model, which can reduce the operation of manually adjusting the structure of the student model and ensure the efficiency and low power consumption of the fusion model deployment. In addition, both the RF in the cloud and the distilled NiN are deployed on the edge server. Comparisons with multiple algorithms on two network traffic datasets show that the proposed model achieves state-of-the-art performance while ensuring the use of minimal computational resources.

Uniaxial compression performance of anti-tetrachiral structures considering the effects of cell size and boundary conditions

Anti-tetrachiral structures (AS) are typical metamaterials known for their negative Poisson's ratio, and have great potential application in reducting the damage of the ship caused by collisions. The existing analysis of the mechanical properties of AS is conducted by applying the energy method to a unit cell with periodic boundary conditions (PBC). In available works, the shear force at the structure's boundaries is neglected. But is it permissible to disregard the impact of shear forces at the boundaries on the structure's equivalent mechanical properties? By examining the deformation relationship between the ribs and the nodal rings, we developed a uniaxial compression model for AS under both free and constrained boundary conditions, which accurately predicts the mechanical properties of the AS structure. This model was validated through numerical simulations and experiments. The findings reveal that the equivalent mechanical properties of AS exhibit a size dependence related to the cell size. For example, for AS with equivalent density and identical overall dimensions, the equivalent Young's modulus of an AS with 2×2 cells will be twice that of an AS with 4×4 cells. Furthermore, the size effect of the structure can be neglected when the number of cells larger than 8×8. Moreover, it is found that the present model considering boundary conditions exhibits an equivalent Young's modulus 25 % higher than the model neglecting boundary conditions. The study's findings indicate that the presence of boundary conditions can disrupt PBC, leading to significant discrepancies between theoretical derivations and practical applications.

Insights into the characteristics of sheet/cloud cavitation and tip-leakage cavitation based on a compressible Euler-Lagrange model

Insights into the characteristics of sheet/cloud cavitation and tip-leakage cavitation based on a compressible Euler-Lagrange model

Growth chamber gas dynamics in Cz silicon single crystal growth process

Mathematical simulation results for inert gas (argon) dynamics in the rarefied atmosphere of Redmet-10 Cz single crystal silicon growth furnace chamber have been reported. Gas flows in cold (room temperature) and hot (growth process) chambers have been analyzed. Convective gas flows have been considered for two growth chamber gas supply options: basic supply, i.e., through the top central inlet hole only, and combined, i.e., with additional gas supply through the lateral inlet hole. Gas outlet is through the growth chamber bottom outlet hole for both options. For the cold chamber option, forced convection has been studied using the incompressible ideal gas model. For hot chamber, both the nonisothermal compressible ideal gas model and the weakly compressible gas model in the Boussinesq approximation have been used for studying the vortex structure produced by a joint action of forced and thermal-gravitational convection. The effect of different gas inlet methods on the convective transport of silicon monoxide evaporating from the free melt surface has been discussed.

Effects of Structure Parameters on Static Performance of Gas Foil Bearings Based on a New Fully Coupled Elastic–Aerodynamic Model

Abstract: Accurately predicting the performance of gas foil bearings (GFBs) is important. This paper presents a new fully coupled elastic–aerodynamic model for analyzing the static performance of gas foil journal bearings (GFJBs). The gas compressible lubrication model was solved in MATLAB to obtain the gas pressure. The foil structure deformation was solved in COMSOL by considering the Coulomb friction and allowing the contact surfaces to be separated from each other. Under given load and rotational speed conditions, the calculated minimum gas film thickness and attitude angle match well with the literature data, validating the accuracy of the developed model. Based on the model developed, a comprehensive and systematic analysis of the effects of the structural parameters on the static performance was performed. The results showed that as the bump height, top foil thickness, and bump foil thickness increased, the load capacity could be improved to different degrees. The bump foil thickness had the greatest effect on the load capacity. These results provide theoretical guidance for the structural design and practical applications of GFJBs.

Effect of Posterior Decompression on the Spinal Cord of Thoracic Ossification of the Posterior Longitudinal Ligament: A Finite Element Analysis

Thoracic ossification of the posterior longitudinal ligament (T-OPLL) causes myelopathy. Although posterior decompression for T-OPLL has shown positive results, patients with kyphotic curvatures often endure poor outcomes. Posterior decompression with fusion (PDF) has demonstrated better results compared to posterior decompression alone. This study aims to evaluate the effects of the posterior procedures for T-OPLL. A 3-dimensional finite element model of the C2-T12 spine, created from medical images, was used to develop the following T3-T4 OPLL compression models: an intact model (no surgery), 25% canal occupancy ratio (COR) OPLL, a discontinuous 25% COR OPLL, a continuous 50% COR OPLL, and a discontinuous 50% COR OPLL. These models were analyzed to evaluate the effects of posterior decompression (laminectomy [LN]) with varied fixation lengths (LN T3-T4, PDF T3-T4, LN T2-T5, and PDF T2-T5) in neutral, flexion, and extension positions. Increased discontinuity in OPLL led to increased stress on the spinal cord. Posterior decompression reduced spinal cord stress in the neutral posture. However, in flexion and extension, spinal cord stress increased for LN T3-T4, LN T2-T5, and PDF T3-T4 compared to the neutral posture. Notably, PDF T2-T5 prevented an increase in spinal cord stress during these motions. Effective management of intervertebral mobility and the appropriate length of decompression are crucial for addressing the thickness and mobility of T-OPLL.

All-polymer syntactic foams: Linking large strain cyclic experiments to Quasilinear Viscoelastic modelling for materials characterisation

The time-dependent behaviour of polymeric composites is critical in a broad range of applications, including those in marine, aerospace, and automotive environments. In the present study, we assess the validity of the quasi-linear viscoelastic (QLV) model to fit the stress–strain behaviour of all-polymer syntactic foams under large cyclic compressional strain in a novel experimental configuration. These syntactic foams were manufactured by adding hollow polymer microspheres of various sizes and wall thicknesses into a polyurethane matrix. These materials are known for their relatively large initial stiffness, and strong recoverability after large strains. In the QLV model, several strain energy functions (SEFs) were employed, including neo-Hookean, Ogden type I, and type II. The bulk and shear moduli are presented in the form of a Prony series. By estimating these experimental data using optimisation, the natural viscoelastic material properties and coefficients associated with the SEF were determined. The influence of the microsphere filling fraction was also explored. We show that at the strain rate considered here of 0.013 s−1, the compressible QLV model coupled with the Ogden-I SEF is capable of providing an excellent fit to experimental data. Critically, this fit can be achieved over a range of cycles via model optimisation to the first cyclic response only.

Dynamic Constitutive Model of Tungsten‐Whisker‐Reinforced Aluminum/Polytetrafluoroethylene Composite Material and Quantitative Determination of Impact Reaction Release Energy in Vacuum Environment

To meet the strength requirements of aluminum/polytetrafluoroethylene (Al/PTFE) in practical applications, tungsten (W) whisker is added into the traditional formula Al/PTFE (26.5%/73.5%) to improve the dynamic compressive strength of the reactive material. Mechanical properties of reactive material specimens prepared by sintering are tested. In the results, it is shown that the dynamic compressive strength of Al/PTFE‐reactive material reinforced by tungsten whisker with volume fraction of 13.1% can reach 141 MPa at strain rate of 1800 s−1. The constitutive model of dynamic compression for tungsten‐whisker‐enhanced Al/PTFE‐reactive material is constructed, and the results are basically consistent with the experimental results. The quantitative impact energy release of reactive material specimens under vacuum environment are performed by using the two‐stage light gas gun loading system, and the quantitative impact release energies of Al/PTFE‐reactive material reinforced by tungsten whisker under different whisker percentage content and diameter are obtained.

A Strain-Controlled Finite Strain Model for CRD Consolidation of Saturated Clays Considering Non-Linear Compression and Permeability Relationships

Consolidation is the combined phenomenon of the compression and groundwater seepage of clay. Accurate evaluation of the consolidation characteristic is essential for the design, construction, and long-term stability of geotechnical structures. In this study, a strain-controlled non-linear finite strain model for constant rate-of-deformation (CRD) consolidation was developed for quickly and reliably predicting the consolidation behavior of clay soils. The model can account for any form of non-linear compression and permeability relationships, thus considering variations in the coefficient of consolidation. Being strain-controlled, it overcomes the limitations of stress-controlled models which require complex numerical iteration. The validity and accuracy of this model were verified through rigorous comparisons with both numerical simulations and experimental data. For normally consolidated soils, a non-linear e-lgσ′compression model was used instead of a linear compression model. For overconsolidated soils, the Harris function compression model was determined to be recommended to overcome the discontinuities in total stress and pore pressure caused by the traditional piecewise e-lgσ′ model. It was also found that determining the steady state of consolidation for normally consolidated soils should use the non-linear method, while the linear method is suggested to be adopted for overconsolidated soils.

Lightweight Detection of Train Underframe Bolts Based on SFCA-YOLOv8s

Improving the accuracy and detection speed of bolt recognition under the complex background of the train underframe is crucial for the safety of train operation. To achieve efficient detection, a lightweight detection method based on SFCA-YOLOv8s is proposed. The underframe bolt images are captured by a self-designed track-based inspection robot, and a dataset is constructed by mixing simulated platform images with real train underframe bolt images. By combining the C2f module with ScConv lightweight convolution and replacing the Bottleneck structure with the Faster_Block structure, the SFC2f module is designed for feature extraction to improve detection accuracy and speed. It is compared with FasterNet, GhostNet, and MobileNetV3. Additionally, the CA attention mechanism is introduced, and MPDIoU is used as the loss function of YOLOv8s. LAMP scores are used to rank the model weight parameters, and unimportant weight parameters are pruned to achieve model compression. The compressed SFCA-YOLOv8s model is compared with models such as YOLOv5s, YOLOv7, and YOLOX-s in comparative experiments. The results indicate that the final model achieves an average detection accuracy of 93.3% on the mixed dataset, with a detection speed of 261 FPS. Compared with other classical deep learning models, the improved model demonstrates superior performance in detection effectiveness, robustness, and generalization. Even in the absence of sufficient real underframe bolt images, the algorithm enables the trained network to better adapt to real environments, improving bolt recognition accuracy and detection speed, thus providing technical references and theoretical support for subsequent related research.

Kinetic characteristics of bornite and chalcopyrite dissolution in nitric acid

The kinetic characteristics of dissolution of copper-bearing sulfides – chalcopyrite (CuFeS₂) and bornite (Cu₅FeS₄) – in nitric acid were studied. The kinetics of the dissolution process was described using a compressible nucleus model. Chalcopyrite of the Vorontsovskoye deposit and bornite of the Karabash deposit were used as research objects. Solution and cake samples were analyzed by optical emission spectrometry and X-ray fluorescence analysis, respectively. The results obtained were processed in the MS Excel software package. The influence of various factors, including temperature, solvent concentration, particle size, and process duration on the dissolution degree of minerals was studied. The process parameters were varied as follows: temperature – from 35 to 95°C; HNO₃ concentration – from 1 to 9 mol/dm³; particle size – from +0.1 to 0.056 mm; duration – from 0 to 60 min. It was established that an increase in temperature and acid concentration leads to a significant increase in the degree of dissolution of both chalcopyrite and bornite. A decrease in particle size also contributes to a more efficient dissolution of both minerals in nitric acid. The calculated activation energy values were 55 kJ/mol for chalcopyrite and 43 kJ/mol for bornite, which is characteristic of the kinetic region of the process. The reaction orders in terms of reactant were determined: 1.62 for chalcopyrite and 1.57 for bornite. In terms of particle size, these were -1.16 for chalcopyrite and -2.53 for bornite. On this basis, generalized equations of dissolution kinetics for both minerals were derived. The results obtained allow an assumption about the kinetic nature of dissolution of chalcopyrite and bornite under the studied conditions.

Performance of unreinforced and geogrid-reinforced pile-supported embankments under localized surface loading: Analytical investigation

An analytical solution is proposed to identify the performance of unreinforced and geogrid-reinforcement pile-supported embankments under localized surface loading at working stress conditions based on the total efficacy and efficacy induced by soil arching alone, average strain of geogrid reinforcement, and average settlement of subsoil. This solution considered interactively soil arching within the embankment fill, the load-deflection behavior of geogrid reinforcement (if existed), and subsoil settlement. Specifically, the soil arching consisted of a structural arch with different stress states (evaluated by the elastoplastic state coefficient K) and a frictional arch. The load-deflection behavior of geogrid reinforcement was modeled by a membrane, with due consideration of the skin friction between the geogrid and soil. The subsoil was idealized as a one-dimensional compression model. The effectiveness of the proposed solution was verified by comparisons with results from the collected literature. It is shown that geogrid reinforcement improved the performance of embankments with low subsoil stiffness significantly more than that of those with high stiffness subsoil. A high tensile stiffness geogrid was found to be inefficient because its contribution to reducing the subsoil settlement and enhancing the load transfer efficacy was minimal. This paper provides a significant reference for optimizing these embankment design.

Axial compression stress-strain relationship of lithium slag rubber concrete

Replacing cement with lithium slag and fine aggregate with rubber in concrete solves waste disposal, reduces material consumption, boosts sustainability, and enhances concrete performance. A set of prismatic concrete specimens with varying proportions were designed and experimentally tested in order to study the compressive stress-strain behavior of lithium slag rubber concrete (LSRC). The main factors affecting the specimens were lithium slag substitution ratio (SL=0%, 10%, 20%, 30%) and rubber substitution ratio (SR=0%, 5%, 10%, 15%). The results demonstrated that the LSRC exhibited good integrity during the damage. Furthermore, the incorporation of lithium slag (LS) was found to effectively compensate for the reduction in compressive strength due to the incorporation of rubber. When 10% of the fine aggregate was replaced with rubber and 20% of the cement was substituted with lithium slag, the axial compressive strength, elastic modulus, and peak strain of the tested specimens increased by 21.57%, 6.92%, and 17.26%, respectively. Compared with ordinary concrete, LSRC has good toughness, impact resistance and durability with minimal loss of strength, and has broad application prospects in engineering fields (such as airports, highways, housing expansion joints, concrete floors and railway concrete sleepers, etc.). Based on the experimental data, simplified modified equations to predict the compressive strength, elastic modulus, peak strain and axial stress-strain constitutive model of LSRC were proposed, so as to promote the development of LSRC.

Интеллектуальная обработка текстовой информации: обзор автоматизированных методов суммаризации

Interest in innovative technological strategies and modern digital tools has increased significantly due to the need to manage large amounts of unstructured data. This paper reviews current paradigms and services for automated summarization, developed based on interdisciplinary research in linguistics, computer technologies, and artificial intelligence. It focuses on syntactic and lexical techniques employed by neural network models for text compression. The paper presents performance examples of such AI-powered services as QuillBot, Summate.it, WordTune, SciSummary, Scholarcy, and OpenAI ChatGPT. The contemporary automated models proved effective in using extractive and abstractive methods to generate summaries of varying quality and length. The extractive approach relies on identifying the most significant sentences from the original text, while abstractive algorithms create new sentence structures that preserve the main idea of the original content. Automated summarizers effectively utilize text compression techniques that are inherent to human approach to text processing, e.g., they exclude redundant information, simplify complex structures, and generalize data. These technologies provide high accuracy and coherence in the generated summaries, though each summarization model has its limitations. Optimal results depend on the specifics of the task at hand: extractive models provide brevity and precision while abstractive ones allow for deeper semantic processing. Automated summarization is becoming an important tool in various fields that require effective analysis and processing of large text data.

Optimizing multi-recycled concrete for sustainability: Aggregate gradation, surface treatment methods and life cycle impact assessment

Utilizing demolished waste as recycled aggregate in concrete is a sustainable practice, however, the concerns over inconsistent quality, such as hardened mortar content and water absorption, hinder global adoption. This study evaluates the physical, microstructural properties and gives insights into Life Cycle Impact Assessment of three generations of recycled aggregate concrete. The microstructure analysis yielded compelling insights into the role of Interfacial Transition Zone (ITZ) of reclaimed aggregate on concrete. The study demonstrated that the effective proportioning of coarse and fine aggregate particles using Compressible Packing Model (CPM), resulted in a reduction of dosage of cement in concrete volume, reducing the Binder Intensity Index (BII), and an improvement in the strength and stability of concrete. Additionally, surface treatment methods, including water soaking, mechanical, and thermo-mechanical treatments, were explored. The thermo-mechanical treatment method proved to be effective in reducing 40–60 % of residual mortar. Life cycle assessment of treated and untreated recycled aggregate concrete was performed, considering five environmental impact indicators viz., Global Warming Potential (GWP), Acidification Potential (AP), Eutrophication Potential (EP), Abiotic Depletion Potential (ADP) and Human Toxicity (HT). These indicators illustrated that CPM method of mix proportioning accompanied with thermo-mechanical treatment of repurposed aggregate resulted in lowering environmental impact. These findings underscore the potential for sustainable concrete construction through the use of treated recycled aggregate, offering valuable insights into methods for augmenting its quality and performance.

Comparison of rotor hovering aerodynamic performance in free-surface and ground effects using numerical methods

The aerodynamic performance of the rotor hovering on the air–water free-surface, which is significant for cross-medium unmanned aerial vehicles, is merely studied. In this study, a compressible two-phase flow model is used to compare the aerodynamic performance in the free-surface effect (FSE) and the ground effect (GE) with various dimensionless distances, γ, between the rotor and the ground (or free-surface). According to the results, the vortex core in FSE moves further in both vertical and radial directions than in GE for the early stages. Additionally, the blade surface is separated into three parts. In zone I, the aerodynamic performance is mostly determined by proximity effects. For both FSE and GE, the downward induced velocity at the rotor disk rises with increasing γ, leading to a decrease in the sectional thrust coefficient CT,S. By the way, CT,S is larger in FSE. In zone III, the aerodynamic performance is mostly governed by the blade tip vortex. The trend of aerodynamic performance with γ is reversed compared with zone I. The above-mentioned two opposing tendencies result in a smaller rotor thrust in FSE than in GE within the range of 0.60≤γ≤3.00, but a higher rotor thrust in FSE within the range of γ≤0.60.

A compressible semi-resolved computational fluid dynamics-discrete element method coupling model for fluid–solid systems with heat transfer

A compressible semi-resolved computational fluid dynamics-discrete element method coupling model for fluid–solid systems with heat transfer

Improving mortar properties using traditional ceramic materials ground to precisely controlled sizes

The present work investigates the impact of particle size reduction of traditional ceramic materials as partial substitutes for Portland cement in mortars. Ceramic brick, ceramic tile, and stoneware were selected, with three particle sizes (D50 of 1, 5, and 15 μm) achieved through grinding operations adapted to each material grindability. The reactivity of ceramic powders was assessed via dissolution in saturated lime solution. Mortars were prepared with 10 % and 20 % cement mass replaced by ceramic powders ground to each fineness. The packing density of mortars was evaluated using the Compressible Packing Model. Compressive strength was measured at 1, 3, 7, and 28 days, and pore size distribution was analyzed by mercury intrusion porosimetry. Results indicated that ceramic tile required less grinding energy than brick and stoneware. High-energy grinding slightly altered the crystalline structure and increased amorphous content, enhancing reactivity with lime. Increased cement replacement with finer ceramic powders (D50 about 1 μm) improved strength, increased mesopores (50 nm), and reduced pore size threshold, attributed to filler and pozzolanic effects. A multiple linear regression model effectively described the influence of various variables on mortar strength with the interaction terms demonstrating the complexity of the interplay of the variables.