Sequence Of Layers Research Articles

Simulation of the Dynamic Responses of Layered Polymer Composites under Plate Impact Using the DSGZ Model

This paper presents a numerical study on the dynamic response and impact mitigation capabilities of layered ceramic–polymer–metal (CPM) composites under plate impact loading, focusing on the layer sequence effect. The layered structure, comprising a ceramic for hardness and thermal resistance, a polymer for energy absorption, and a metal for strength and ductility, is analyzed to evaluate its effectiveness in mitigating the impact loading. The simulations employed the VUMAT subroutine of DSGZ material models within Abaqus/Explicit to accurately represent the mechanical behavior of the polymeric materials in the composites. The VUMAT implementation incorporates the explicit time integration scheme and the implicit radial return mapping algorithm. A safe-version Newton–Raphson method is applied for numerically solving the differential equations of the J2 plastic flow theory. Analysis of the simulation results reveals that specific layer configurations significantly influence wave propagation, leading to variations in energy absorption and stress distribution within the material. Notably, certain layer sequences, such as P-C-M and C-P-M, exhibit enhanced impact mitigation with a superior ability to dissipate and redirect the impact energy. This phenomenon is tied to the interactions between the material properties of the ceramic, polymer, and metal, emphasizing the necessity of precise material characterization and enhanced understanding of the layer sequencing effect for optimizing composite designs for impact mitigation. The integration of empirical data with simulation methods provides a comprehensive framework for optimizing composite designs in high-impact scenarios. In the general fields of materials science and impact engineering, the current research offers some guidance for practical applications, underscoring the need for detailed simulations to capture the high-strain-rate dynamic responses of multilayered composites.

MRUNet-3D: A multi-stride residual 3D UNet for lung nodule segmentation

Obtaining an accurate segmentation of the pulmonary nodules in computed tomography (CT) images is challenging. This is due to: (1) the heterogeneous nature of the lung nodules; (2) comparable visual characteristics between the nodules and their surroundings. A robust multi-scale feature extraction mechanism that can effectively obtain multi-scale representations at a granular level can improve segmentation accuracy. As the most commonly used network in lung nodule segmentation, UNet, its variants, and other image segmentation methods lack this robust feature extraction mechanism. In this study, we propose a multi-stride residual 3D UNet (MRUNet-3D) to improve the segmentation accuracy of lung nodules in CT images. It incorporates a multi-slide Res2Net block (MSR), which replaces the simple sequence of convolution layers in each encoder stage to effectively extract multi-scale features at a granular level from different receptive fields and resolutions while conserving the strengths of 3D UNet. The proposed method has been extensively evaluated on the publicly available LUNA16 dataset. Experimental results show that it achieves competitive segmentation performance with an average dice similarity coefficient of 83.47 % and an average surface distance of 0.35 mm on the dataset. More notably, our method has proven to be robust to the heterogeneity of lung nodules. It has also proven to perform better at segmenting small lung nodules. Ablation studies have shown that the proposed MSR and RFIA modules are fundamental to improving the performance of the proposed model.

Failure prediction and optimization for composite pressure vessel combining FEM simulation and machine learning approach

Failure assessment is one of the fundamental tasks for optimization of composite pressure vessels (CPVs). However, the extensive design space of composites usually leads to costly and repetitive work of failure assessment that hinders the design and optimization of CPVs. In this work, a combined FEM simulation and machine learning approach is used to predict the failure factors R of CPVs for further optimization. CPVs with various design parameters are automatically generated and analyzed in ABAQUS to obtain failure factors under various external loadings. Then, an assembled deep neural network (DNN) namely Multi-DNN is trained to map relations between design parameters and R of CPVs. Further, transfer learning (TL) is introduced to improve the extensibility of our approach by fine-tuning the pre-trained Multi-DNN using scarce data of new design space. Such a TL-Multi-DNN model can precisely predict R of CPVs with new design parameters. Additionally, combining the trained models and genetic algorithm, the optimization of CPVs is carried out with high efficiency and low computational cost. As a result, R of CPV with various given and new design parameters can be estimated directly by the trained models and the optimal layer sequence of CPV can be obtained efficiently.

Development and characterization of microwave‐processed linear low‐density polyethylene based sisal/jute hybrid laminates

AbstractNatural fiber‐reinforced composites are gaining significant popularity for their biodegradability and eco‐friendliness. Fiber modifications and hybridization have been used to address issues like hydrophilicity and fiber inhomogeneity. However, efficient manufacturing is still a challenge for natural fiber‐reinforced polymer composites. The present study explores the potential of microwave processing of hybrid laminates composed of sisal and jute fibers. The laminates, consisting of linear low‐density polyethylene (LLDPE) as matrix and sisal and jute as reinforcement materials were subjected to microwave processing at specific power, time, frequency, and loads. Four types of laminates with different stacking sequences: sisal‐sisal‐sisal (SSS), sisal‐jute‐sisal (SJS), jute‐sisal‐jute (JSJ), and jute‐jute‐jute (JJJ) were developed. The developed composites showed ~3% void content, with the SJS and JSJ composite having the least and highest void content, respectively. The SJS composite showed the highest tensile and flexural strength of 15.27 and 20.40 MPa, respectively, out of all the configurations and an improvement of 42% and 85% over pure LLDPE. Hybrid composites having high‐strength fibers in the skin layer exhibited superior mechanical properties. Microscopic examination of the fractured specimens revealed that fiber pull‐out and fiber breakage were the primary failure mechanisms of failure. Lateral failure due to delamination between matrix and fiber was predominant with grip and gauge regions as frequent failure points. The incorporation of sisal and jute fiber reinforcement into the polymers doesn't change the thermal stability of the fabricated composites significantly. The above results show that microwave‐assisted processing is a promising method for producing natural fiber‐reinforced hybrid polymer composites.Highlights Development of hybrid laminates using microwave energy at 2.45 GHz. The mechanism of microwave‐based processing of polymer composites has been discussed. Effect of layering sequence on mechanical and thermal properties of the developed composite. Assessment of tensile strength with morphology, flexural strength and thermo‐gravimetric analysis.

Interply hybridization of hollow glass fiber reinforced vascular composites

AbstractThe use of hybrid reinforcement in polymer composites has been getting attention due to its potential for weight and cost reduction while achieving improved design flexibility for mechanical and thermal properties. In this study, an interply hybridization of epoxy laminates was performed using layers of solid and hollow glass fibers. The structural and thermal performance of the resulting hybrid vascular composites was investigated. All composite parts were manufactured using vacuum‐assisted resin transfer molding (VARTM), following an interply hybridization plan for the stacking sequence of hollow and solid fiber layers. An expected decrease in flexural properties and thermal conductivity was observed for the laminates with completely hollow fibers. On the other hand, hybridization had a positive effect on the laminate properties. It was shown that hybrid vascular composites could be manufactured without a significant reduction in mechanical and thermal performance by a judicious hybridization plan. The optimum hybridization of the laminate achieved a 3.1% increase in the specific flexural strength and a 5.3% decrease in thermal conductivity while containing a vascular network that can be used for active thermal control.Highlights A vascular interply composite structure consisting of hollow and solid fibers was manufactured and characterized for the first time. The I4 composite, vascularized through the middle layers of the laminate, has almost as much flexural strength as solid composite and much more specific flexural strength. The thermal conductivity of the I4 composite is almost as much as that of the solid composite. Functional composites for thermal management and smart applications can be manufactured using hollow fibers without sacrificing flexural strength and thermal conductivity with interply vascularization.

An Integrated Multi-Source Dataset for Measuring Settlement Evolution in the United States from 1810 to 2020

Understanding changes in the built environment is vital for sustainable urban development and disaster preparedness. Recent years have seen the emergence of a variety of global, continent-level, and nation-wide datasets related to the current state and the evolution of the built environment, human settlements or building stocks. However, such datasets may face limitations like incomplete coverage, sparse building information, coarse resolution, and limited timeframes. This study addresses these challenges by integrating three spatial datasets to create an extensive, attribute-rich sequence of settlement layers spanning 200 years for the contiguous U.S. This integration process involves complex data processing, merging property-level real estate, parcel, and remote sensing-based building footprint data, and creating gridded multi-temporal settlement layers. This effort unveils the latest edition (Version 2) of the Historical Settlement Data Compilation for the U.S. (HISDAC-US), which includes the latest land use and structural information as of the year 2021. It enables detailed research on urban form and structure, helps assess and map the built environment’s risk to natural hazards, assists in population modeling, supports land use analysis, and aids health studies.

Optimal Re-Materialization Strategies for Heterogeneous Chains: How to Train Deep Neural Networks with Limited Memory

Training in Feed Forward Deep Neural Networks is a memory-intensive operation which is usually performed on GPUs with limited memory capacities. This may force data scientists to limit the depth of the models or the resolution of the input data if data does not fit in the GPU memory. The re-materialization technique, whose idea comes from the checkpointing strategies developed in the Automatic Differentiation literature, allows data scientists to limit the memory requirements related to the storage of intermediate data (activations), at the cost of an increase in the computational cost. This paper introduces a new strategy of re-materialization of activations that significantly reduces memory usage. It consists in selecting which activations are saved and which activations are deleted during the forward phase, and then recomputing the deleted activations when they are needed during the backward phase. We propose an original computation model that combines two types of activation savings: either only storing the layer inputs, or recording the complete history of operations that produced the outputs. This paper focuses on the fully heterogeneous case, where the computation time and the memory requirement of each layer is different. We prove that finding the optimal solution is NP-hard and that classical techniques from Automatic Differentiation literature do not apply. Moreover, the classical assumption of memory persistence of materialized activations, used to simplify the search of optimal solutions, does not hold anymore. Thus, we propose a weak memory persistence property and provide a Dynamic Program to compute the optimal sequence of computations. This algorithm is made available through the Rotor software, a PyTorch plug-in dealing with any network consisting of a sequence of layers, each of them having an arbitrarily complex structure. Through extensive experiments, we show that our implementation consistently outperforms existing re-materialization approaches for a large class of networks, image sizes and batch sizes.

Influence of layering sequence on performance of jute/wool epoxy hybrid composites: a comparative study with automotive plastic

Natural fibres such as Jute and Wool are increasingly being used in fibre reinforced polymer composites, hence progressively supplanting synthetic fibres. Combining these two unique natural fibers would improve the performance of composites used in secondary structural applications. This study is intended to investigate the effects of Jute and Wool fiber arrangement on mechanical properties in an epoxy matrix. In order to achieve this, hand layup was employed to fabricate composite laminates consisting of four layered fabric layers stacked in sequences, as Jute/Wool/Wool/Jute (JW1), Wool/Jute/Jute/Wool (JW2), Jute/Wool/Jute/Wool (JW3), and Jute/Jute/Wool/Wool (JW4). Effect of these stacking sequences on the properties of Natural Fibre Hybrid Composites (NFHCs) was evaluated by conducting a series of standardized tests such as Flexural, Interlaminar Shear Strength (ILSS), Impact and Single End Notch Bend (SENB) tests, as per ASTM standards. On the other hand an automobile interior thermoplastic material was evaluated for different mechanical properties and compared with hybrid Jute-Wool composite. Experimental results showed that the JW1 composite, exhibited superior mechanical properties because it has tailored with a Wool fabric at the core and Jute fabric on its outer surface. Hence, JW1 had a remarkable tensile strength of 40 MPa, a maximum flexural strength of 99 MPa and extremely high interlaminar shear strength (ILSS) of 3.09 MPa. In addition to this, JW1 also had high fracture toughness, measuring around 72 M P a m 1 /2, and impressive impact strength of 85 J/m2. Furthermore, scanning electron microscope (SEM) photo images confirmed that the presence of strong bonding between the fibres and matrix in JW1 composite exhibited and overall improvement of mechanical performance by 82% when compared with other combinations (JW2, JW3, and JW4) including automotive plastics.

Analysis of Tensile and Bending Strength of Coconut Fiber Reinforcement Composite on Quasi Isotropic Laminates Stacking Sequence

A composite is a material structure composed of two or more combinations of constituents combined macroscopically, where the two combinations do not dissolve each other. The first phase is called reinforcement, while the second phase is called matrix. This study aims to analysis the tensile and bending strength values of coconut fiber reinforced composites with a sequence of layers of quasi-isotropic fiber laminates. Experimental studies were carried out on polyester resin composites reinforced with coconut fiber in the fiber layer sequences of [0°/0°/0°], [+45°/0°/-45°] and [+60°/0°/-60°]. Tensile test specimen standards refer to ASTM D-3039 and bending test standards refer to ASTM D-7264. The results of the blistering test obtained a maximum stress of 18.3298 N/mm2 in the fiber layer sequence [0°/0°/0°] and a minimum stress of 10.8966 N/mm2 in the fiber layer sequence [+60°/0°/ -60°]. Meanwhile, the bending test results showed that the maximum bending stress was 76.065 N/mm2 in the fiber layer sequence [0°/0°/0°] and the minimum stress was 30.256 N/mm2 in the fiber layer sequence [+45°/0°/-45 °].

Sequence Stratigraphy of Subcontinental Reservoirs of the Upper Part of the Tanopcha Formation of the Minkhov Field (North of Western Siberia)

Abstract ––The purpose of this paper is to present the results of sequence stratigraphic modeling of subcontinental reservoirs of the Tanopcha Formation upper part (Lower Cretaceous) of the Minkhov field, located in the northern part of the Messov oil and gas region of the Gydan oil and gas region. The study is vital due to the need for additional exploration and commissioning of hydrocarbon reserves in the lower part of the Pokur suprahorizon in the north of Western Siberia. The modeling is based on sedimentological columns for four wells, in which the object of study was characterized by a core with sufficient completeness. The structure of seven sequences has been studied. Sequences PK 2-2, 2-3, 2-4, 2-5, 2-6, and PK 3-1, in accordance with the classical concept, are divided into three systems tracts: lowstand (LST), transgressive (TST), highstand (HST). The boundaries of the sequences are outlined along the erosive surfaces in the bottom of the alluvial complexes. LST tops – on contacts of alluvial facies with tidal, marsh, lake sediments, indicating flooding of the coastal plain at the beginning of transgressions. TST tops – on contacts of facies of the tidal plain with alluvial deposits, indicating the beginning of normal regressions. The main reservoirs of the object are large belts and covers of alluvium of amalgamating channels as part of the LST. Additional reservoirs are thin bodies of tidal genesis in the composition of TST, and alluvial genesis in the composition of HST. The bottom of the PK 3-2 sequence was outlined on the contacts of coastal-marine TST sandstones with tidal and marsh facies, indicating the event transgression of the coastline. One of the main reservoirs of the field, the TP1 formation, is connected to the TST of the PK 3-2 sequence. The boundaries of the selected sequences can be traced on seismic sections of acoustic impedance. Depending on the temporary thickness in the impedance field, the sequences are displayed as two-, three- and four-layer sequences of layers, which opens up the possibility of their further study by the method of wave field inversion.

A numerical extension of White's theory of P-wave attenuation to non-isothermal poroelastic media.

Mesoscopic P-wave attenuation in layered, partially saturated thermo-poroelastic media is analyzed by combining the theories of Biot poroelasticity and Lord-Shulman thermoelasticity (BLS). The attenuation is quantified by estimating the quality factor Q. The mesoscopic attenuation effect, commonly referred to as wave-induced fluid flow (WIFF), is the process that converts fast compressional and shear waves into slow diffusive Biot waves at mesoscopic heterogeneities larger than the pore scale, but much smaller than the dominant wavelengths. This effect was first modeled in White's isothermal theory by quantifying the seismic response of a periodic sequence of planar porous layers that are alternately saturated with gas or water. This work presents a numerical extension of White's theory for the non-isothermal case in this type of sequence. For this purpose, an initial-boundary-value problem (IBVP) for the BLS wave propagation equations is solved using the finite element method, where the particle velocity field is recorded at uniformly distributed receivers. The quality factor is estimated using spectral-ratio and frequency-shift methods. The Q-estimates show that thermal effects influence the attenuation of the P-wave and the velocity dispersion compared to the isothermal case.

Theoretical magnetotelluric response of stratiform earth consisting of alternative homogeneous and transitional layers

Abstract The magnetotelluric (MT) responses are explicitly solved for a stratiform earth containing multiple transitional layers in which the conductivity varies linearly with depth. In the model under consideration, any one homogeneous layer with constant conductivity or transitional one may be absent in the geometry. The traditional one-dimensional (1D) models with sharp boundaries will be obtained if all the transitional layers are absent in the geometry, while a special 1D model consisting of a sequence of contiguous transitional layers may be obtained if all the homogeneous layers (except the basement layer) are removed from the geometry. The tangential electric and magnetic fields as well as the surface impedance are analytically expressed by Airy functions. The analytical formula is validated in three theoretical examples by comparing with the results from available codes. The apparent resistivity and impedance phase on the surface of three different transitional models are illustrated to analysis the influence of the transitional layers on MT responses. The new formula provides an alternative way to obtain the analytic MT responses for the special layered earth.

Nonintrusive Load Monitoring (NILM) Using a Deep Learning Model with a Transformer-Based Attention Mechanism and Temporal Pooling

Nonintrusive load monitoring (NILM) is an important technique for energy management and conservation. In this paper, a deep learning model based on an attention mechanism, temporal pooling, residual connections, and transformers is proposed. This article presents a novel approach for NILM to accurately discern energy consumption patterns of individual household appliances. The proposed method entails a sequence of layers, including encoders, transformers, attention, temporal pooling, and residual connections, offering a comprehensive solution for NILM while effectively capturing appliance-specific energy usage in a household. The proposed model was evaluated using UK-DALE, REDD, and REFIT datasets in both seen and unseen cases. It shows that the proposed model in this paper performs better than other methods stated in other papers in terms of F1-score and total error of the results (in terms of SAE). This model achieved an F1-score equal to 92.96 as well as a total SAE equal to −0.036, which shows its effectiveness in accurately diagnosing and estimating the energy consumption of individual home appliances. The findings of this research show that the proposed model can be a tool for energy management in residential and commercial buildings.

Fresnel coefficients, coherent optical scattering, and planar waveguiding

The Fresnel formalism of transmission and reflection of coherent waves on the abrupt interface between two different optical media is revisited and extended to study optical phenomena concerning guided waves. Using the Fresnel formalism, the propagation of light through different sequences of parallel layers of constant refractive index is analysed, revealing that, in all cases, the solutions corresponding to waveguiding share major features. In the analysis, aiming to connect waveguiding with coherent optical scattering, the possibility of the system having a resonance is examined, and it is shown that guided modes always emerge as singular solutions of the associated coherent scattering configuration. Invariably, the resonance condition that must be satisfied to obtain waveguiding corresponds to the pole of the associated total reflection and transmission coefficients. Even more, the fulfilment of the requirements to meet the resonance condition is a clear indication of the possibility of waveguiding.

Comparative analysis of ground thermal conductivity and thermal resistance of borehole heat exchanger in different geological layered sequence

Borehole heat exchanger usually passes through many geological layers of the different thermo-physical properties from one layer to another in practical project of ground-coupled heat pump system. Considering the effect of heterogeneity in geological structures on heat transfer of borehole heat exchanger, it is necessary for in-situ thermal response test analysis to investigate the effect of geological layered sequence on ground thermal conductivity and borehole thermal resistance. Firstly, based on the validated numerical layered borehole heat exchanger model and numerical homogenous borehole heat exchanger model, the simulated distributions of water temperatures in U-pipe and heat flux along the depth in thermal response test are respectively compared and analysed. Then, the effective thermal conductivity and effective borehole thermal resistance estimated by the infinite line source model are respectively obtained based on the simulated thermal response tests carried out in the borehole heat exchanger installed in the six types of geological layered structures. The borehole thermal resistances in the two numerical model were respectively calculated by integrating the local thermal resistance along the depth. Finally, the ground thermal conductivities and borehole thermal resistances are compared in six types of geological layered structures. The results show that the maximum relative error of 7 % between the effective thermal conductivity and the thickness-weighted thermal conductivity in the numerical layered borehole heat exchanger model is located in the sequence of S-L-C due to BOTTOM 60m-thickness clay with lowest thermal conductivity. The relative error between effective borehole thermal resistance and borehole thermal resistance in the numerical layered borehole heat exchanger model are 3.5 %∼7.9 % in the six types of layered geological structures, and the maximum relative error between the two borehole thermal resistances corresponding to the two numerical models is only 1.1 %, which proves that effect of geological layered sequence on the deviation is limited in the two numerical models.

Local coupled Markov chain model for simulating varied stratigraphy

The coupled Markov chain (CMC) model is widely employed to simulate geological uncertainty based on sparse borehole data, owing to its simplicity, parameter efficiency, and computational effectiveness. However, the existing CMC methods exhibit limitations in accurately predicting the orientation and significant dip angle changes of soil stratigraphy. These limitations stem from their reliance on a globally constant Walther constant, resulting in generated layer boundaries converging to a fixed angle. To address these issues, this study develops a local CMC model to enhance the accuracy of predicting layer sequences and describe the geological uncertainty of complex geological structures using limited borehole and surface data. In the proposed local CMC method, the global geological profile is decomposed into interconnected fragments, guided by reference boreholes. For each CMC fragment, the likelihood of observational scenarios is determined to identify the layer orientation and optimal Walther’s constant corresponding to the local geological profile. By combining the most likely realizations from all CMC fragments, the final geological profile and stratigraphic uncertainty are obtained. Illustrative case studies demonstrate that the proposed approach effectively addresses the limitations of conventional CMC models in accurately simulating significant changes in layer orientation and dip angle variations, given the same conditions of sparse borehole data. The employed localization strategy enables the prediction of geological profiles that encompass special layers such as soil interlayers, anticlines, and synclines, which cannot be well achieved by conventional CMC models. The local CMC model significantly improves the matching accuracy of geological model in situations with limited boreholes. Therefore, it plays a crucial role in the analysis of geological structures under data-scarce conditions.

Geophysical and geotechnical investigations of the site of a collapsed two-storey building in Modomo, Ile-Ife, Southwestern Nigeria

Geophysical and the geotechnical methods were used to study the site of a collapsed two-storey building in orderto determine the sequence and competence of the subsurface layers and determine the cause of the collapse of the building. The geophysical method employed the electrical resistivity method involving one dimensional (1D) Vertical Electrical Sounding (VES) and the two dimensional (2D) imaging techniques. Three VES stations were occupied using Schlumberger electrode configuration. 2D imaging data were acquired along four traverses using dipole-dipole electrode configuration. The geotechnical method involved Cone Penetration Test (CPT). CPT data was acquired at two points with the aid of 2.5 ton Dutch CPT machine. The results showed four subsurface layers based on the 1D VES results and 2D imaging results namely: topsoil, weathered layer, fractured basement rock and fresh basement rock. The topsoil resistivity varied from 116 Ωm to 191 Ωm and has thickness of about 1.5 m. It was classified as moderately competent. The second layer is weathered rock with resistivity varying between 92 Ωm and 327 Ωm and of about 8 m to 13 m thickness. This layer is also moderately competent. The third layer is fractured basement rock. It underlies the weathered rock and have resistivity varying from 391 Ωm to 405 Ωm. Its thickness varied from about 2 m to 20 m. This layer is competent. The fourth layer is the fresh basement rock having resistivity varying from 1365 Ωm to 12348 Ωm and is highly competent. The subsurface materials in the area are competent and can sustain the foundation of the building. Hence, the study concluded that the building collapse did not result from incompetent subsurface materials. Factors other than incompetent subsoil material that is human factors are believed to be responsible. This study has demonstrated the effectiveness of geophysical and geotechnical investigations in determining the competence of subsurface materials at engineering sites.

A Tree-Structured Multitask Model Architectures Recommendation System.

Neural networks with branched architectures, namely, tree-structured models, have been employed to jointly tackle multiple vision tasks in the context of multitask learning (MTL). Such tree-structured networks typically start with a number of shared layers, after which different tasks branch out into their own sequence of layers. Hence, the major challenge is to determine where to branch out for each task given a backbone model to optimize for both task accuracy and computation efficiency. To address the challenge, this article proposes a recommendation system that, given a set of tasks and a convolutional neural network-based backbone model, automatically suggests tree-structured multitask architectures that could achieve a high task performance while meeting a user-specified computation budget without performing model training. Extensive evaluations on popular MTL benchmarks show that the recommended architectures could achieve competitive task accuracy and computation efficiency compared with state-of-the-art MTL methods. Our tree-structured multitask model recommender is open-sourced and available at https://github.com/zhanglijun95/TreeMTL.

Side-Channel Attack Analysis on In-Memory Computing Architectures

In-memory computing (IMC) systems have great potential for accelerating data-intensive tasks such as deep neural networks (DNNs). As DNN models are generally highly proprietary, the neural network architectures become valuable targets for attacks. In IMC systems, since the whole model is mapped on chip and weight memory read can be restricted, the pre-mapped DNN model acts as a “black box” for users. However, the localized and stationary weight and data patterns may subject IMC systems to other attacks. In this paper, we propose a side-channel attack methodology on IMC architectures. We show that it is possible to extract model architectural information from power trace measurements without any prior knowledge of the neural network. We first developed a simulation framework that can emulate the dynamic power traces of the IMC macros. We then performed side-channel leakage analysis to reverse engineer model information such as the stored layer type, layer sequence, output channel/feature size and convolution kernel size from power traces of the IMC macros. Based on the extracted information, full networks can potentially be reconstructed without any knowledge of the neural network. Finally, we discuss potential countermeasures for building IMC systems that offer resistance to these model extraction attack.

Effects of thermal conditions on fatigue behaviour of laminated glass/epoxy plates under tension-tension cycle

The fatigue behavior study on laminated glass/epoxy composite plates at elevated temperature is attempted in the present research work. INSTRON 8862 servo-electric universal testing machine with hydraulic suspension is utilized to perform low cyclic tension-tension fatigue tests at 0.5Hz and 0.7Hz frequencies. Bluehill universal software compatible with INSTRON 8862 is employed to obtain the relationship of fatigue stress upon cycles to failure (S – N) for each specimen. The parametric investigation is done for the loading frequency, lamination sequence and number of layers to understand their effects on fatigue behavior of composite laminated plates under ambient and thermal environment. The elicited results lead to the conclusion that above parameters greatly influenced the fatigue behavior of composite laminated plates under thermal loading. The rising temperature has significant adverse effects on fatigue life. The present research is beneficial for the analysis and design of laminated composite plate or plate-like structures in the domain of fatigue analysis.