Composite Model Research Articles

Substantial Global Radial Variations of Basalt Content Near the 660‐km Discontinuity

AbstractMid‐ocean ridges generate basalt and harzburgite, which are introduced into the mantle through subduction as a mechanical mixture, contributing to both lateral and radial compositional heterogeneity. The possible accumulation of basalt in the mantle transition zone has been examined, but details of the mantle composition below the 660‐km discontinuity (hereafter d660) remain poorly constrained. In this study, we utilize the subtle waveform details of S660S, the underside shear‐wave reflection off the d660, to interpret the seismic velocity, density, and compositional structure near, and particularly below, the d660. We identify a significant difference in S660S waveform shape in subduction zones compared to other regions. The inversion results reveal globally enriched basalt at the d660, with a notably higher content in subduction zones, consistent with the smaller impedance jump and S660S peak amplitude. The basalt fraction decreases significantly to less than 10% near 800‐km depth, forming a global harzburgite‐enriched layer and resulting in a steep seismic velocity gradient just below the d660, in agreement with 1D global reference models. The striking compositional radial variations near the d660 verify geodynamic predictions and challenge the applicability of homogeneous radial compositional models in the mantle. These variations may also affect the viscosity profile and, consequently, the dynamics at the boundary between the upper and lower mantle.

Enhancing Survival Outcome Predictions in Metastatic Non-Small Cell Lung Cancer Through PET Radiomics Analysis

(1) Background: Advanced-stage lung cancer poses significant management challenges. The goal of this study was to identify crucial clinical and PET radiomics features that enable prognostic stratification for predicting outcomes. (2) Methods: PET radiomics features of the primary lung lesions were extracted from 99 patients with stage IVB NSCLC, and the robustness of these PET radiomics features was evaluated against uncertainties stemming from extraction parameters and contour variation. We trained three survival risk models (clinical, radiomics, and a composite) through a penalized Cox model framework. We also created a Balanced Random Forest classification predictive model, using the selected features, to predict 1-year survival. (3) Results: We identified 367 common PET radiomics features that exhibited robustness to perturbations introduced by contour variation and extraction parameters. Our findings indicated that both the radiomics and the composite model outperformed the clinical model in stratifying the risk for survival with statistical significance. In predicting 1-year survival, the radiomics model and the composite model also achieved better predicting accuracies compared to the clinical model. (4) Conclusions: Robust PET radiomics analysis successfully facilitated the stratification of patient risk for survival outcomes and predicted 1-year survival in stage IVB NSCLC.

Bethe vectors and recurrence relations for twisted Yangian based models

We study Olshanski twisted Yangian based models, known as one-dimensional “soliton non-preserving” open spin chains, by means of algebraic Bethe Ansatz. The even case, when the bulk symmetry is \mathfrak{gl}_{2n}𝔤𝔩2n and the boundary symmetry is \mathfrak{sp}_{2n}𝔰𝔭2n or \mathfrak{so}_{2n}𝔰𝔬2n, was studied in [Ann. Henri Poincaré 20, 339 (2018)]. In the present work, we focus on the odd case, when the bulk symmetry is \mathfrak{gl}_{2n+1}𝔤𝔩2n+1 and the boundary symmetry is \mathfrak{so}_{2n+1}𝔰𝔬2n+1. We explicitly construct Bethe vectors and present a more symmetric form of the trace formula. We use the composite model approach and Y(\mathfrak{gl}_n)Y(𝔤𝔩n)-type recurrence relations to obtain recurrence relations for twisted Yangian based Bethe vectors, for both even and odd cases.

Composite underwater optical wireless channel modeling based on the electric field Monte Carlo method

In this study, we propose a simulation method to investigate the scattering characteristics of beams with specific phase structures in waters containing micrometer-scale particles and bubbles. This method is based on the electric field Monte Carlo (EMC) algorithm and utilizes the Mie theory and ray tracing method. We simulate the propagation of orbital angular momentum (OAM) beams in a composite underwater channel using the proposed method and analyze the impact of bubble density on the spiral phase distribution of the OAM beams. We also design an underwater experiment in a water tank to simulate the propagation of OAM beams in the composite channel. The results reveal that the power of the desired mode of the OAM beam decreases as the bubble density increases, which is consistent with the simulation results. This method facilitates a realistic simulation of structured light beam propagation in waters containing particles and bubbles and can provide reference data for studying the scattering characteristics of structured beams in composite underwater channels.

On Thermoelastic Impact Modelling of Frozen Composite Target During Pre – Heated Projectile Penetration Starts of Motion

This study investigates a composite double-layer structure for improved thermal shock resistance. A modified Hugoniot elastic limit model is presented for the composite, followed by a 2D thermo-elastic impact simulation using commercial software. The simulation focuses on a composite material under initial extreme low temperature conditions with alternating metallic (Steel, Aluminum) and non-metallic layers (Kevlar 49, Graphite). The frozen target is subjected to pre- heated projectile. The objective is to optimize the composite's durability by strategically placing reinforcement particles within specific layers. The analysis explores the effect of different particle types (oil, water, Aluminum, Steel) and sizes (0.3mm, 0.5mm, 1mm) on the composite's stress response. It was found that aluminum and steel particles significantly reduce stress compared to fluid/gas particles, confirmed qualitatively by literature. Kevlar particles within the SiCp layer enhance its resistance, while Aluminum particles within the Kevlar layer offer weight reduction benefits. Moreover, for Kevlar, larger particles improve resistance, and vice versa for the SiCp case. Considering weight, a particle size of 0.5mm is chosen for both layers. Moreover, a finite element analysis of the optimized composite model subjected to thermoelastic impact loading demonstrates its superior performance compared to the non-reinforced composite. Specific layer combinations (SiCp with Kevlar particles, Graphite or Kevlar with Aluminum particles) show the most significant stress reduction. Finally, separate 3D ballistic analysis was performed for Tungsten having 600m/sec projectile into 5 layered target with thickness of 2.8mm each layer and appropriate interaction friction (SiCp - Steel 304 - Al 7075-T651 - Kevlar 49 - Graphite Crystalline) during penetration time of 0.006sec at 300K. The dynamic explicit transient analysis was confirmed with the predecessors' analytic calculations.

Distribution of bone fragments in angled shots: an experimental study conducted on composite models containing artificial bone plates.

In conventional gunshot injuries to targets containing bone the resulting osseous fragments do not precede but follow the bullet on its further way through adjacent soft tissues. The term "secondary projectiles" for the particles does not appear to be appropriate since they are not believed to have enough energy necessary for creating their own wound channels away from the temporary cavity. Former studies have shown that in angled shots to glass panes the bulk of splinters does not follow the bullet's trajectory: The majority of the glass fragments, especially the larger ones, move at right angles to the pane shot through. The aim of the presented study was to examine whether osseous fragments behave like glass splinters in angled shots to flat synthetic bone. In this context, it should also be assessed, whether the bone fragments might act as secondary projectiles in rare cases. To answer these questions, test shots were fired to composite models consisting of flat synthetic bone and ballistic gelatin. Pistol cartridges 9mm Luger were used to fire the shots which were video-documented with a high-speed camera. Afterwards, the composite models underwent CT examination and macroscopic inspection. Video-documentation revealed that the larger bone particles from the perforation site move at a roughly right angle from the osseous sheet into the gelatin, causing an eccentric bulge of the temporary cavity. The smaller bone fragments were also lodged along the bullet's path, predominantly in the cracks radiating from the permanent wound channel.

Umami-gcForest: Construction of a predictive model for umami peptides based on deep forest

Umami peptides have recently gained attention for their ability to enhance umami flavor, reduce salt content, and provide nutritional benefits. However, traditional wet laboratory methods to identify them are time-consuming, laborious, and costly. Therefore, we developed the Umami-gcForest model using the deep forest algorithm. It constructs amino acid feature matrices using ProtBERT, amino acid composition, composition-transition-distribution, and pseudo amino acid composition, applying mutual information for feature selection to optimize dimensions. Compared to other machine learning baseline, umami peptide prediction, and composite models, the validation results of Umami-gcForest on different test sets demonstrated outstanding predictive accuracy. Using SHapley Additive exPlanations to calculate feature contributions, we found that the key features of Umami-gcForest were hydrophobicity, charge, and polarity. Based on this, an online platform was developed to facilitate its user application. In conclusion, Umami-gcForest serves as a powerful tool, providing a solid foundation for the efficient and accurate screening of umami peptides.

Changes in 24-h Movement Behaviors During Relationship and Parenthood Transitions: A Compositional Data Analysis.

There is scarcity of studies using device-based measures to examine how relationship and parenthood transitions modify 24-h movement behaviors. This study examined how the composition of 24-h movement behaviors changes during these life transitions. Young adults (n = 170, mean age 25.6 years, SD 0.6) from the Special Turku Coronary Risk Factor Intervention Project (STRIP) wore wrist-worn accelerometers for 1 week and completed questionnaire at ages 26 and 31 years. Participants were categorized by relationship status into single (16%), those transitioning from single to partnered (31%), partnered (47%), and separated (7%), and by parenthood status into non-parents (73%), new parents (19%), and parents (8%). Changes in daily movement behaviors, including sleep, sedentary behavior (SED), light physical activity (LPA), and moderate-to-vigorous physical activity (MVPA), were examined using compositional linear mixed models. In general, LPA and MVPA decreased relative to sleep and SED (p = 0.007). Differences emerged between LPA and MVPA in relationship and parenthood groups (p for group × time interaction 0.008 and 0.001). Those transitioning to partnership decreased MVPA by 17 min/day, while partnered and separated individuals showed no notable MVPA change but decreased LPA by 14 and 43 min/day. Single individuals and non-parents decreased LPA and MVPA in similar proportions. New parents decreased MVPA by 20 min/day, while parents increased it by 19 min/day. Becoming first-time parent and starting relationship was associated with decline of MVPA. Our results suggest the importance of considering these life transitions and providing guidance for maintaining physical activity despite changes in life situations.

A hybrid model for accurate prediction of composite longitudinal elastic modulus

This research presents a novel hybrid model that integrates a physical-based empirical model with an Artificial Neural Network (ANN) to accurately predict the longitudinal modulus of elasticity for composites under compression. The study focuses on a composite material with a pore inclusion within an ABS plastic matrix, exploring various pore volumes, orientations, and shapes. As part of the proposed hybrid model, a regression-type neural network was trained in MATLAB® to predict and correct discrepancies between the Generalized Stiffness Formulation (GSF) homogenization-based modeling method and the collected compression experimental test results. Using MATLAB® neural network, random error datasets were used to train the feed-forward neural network, and the remaining error datasets were used for validating the performance of the proposed hybrid modeling scheme.The hybrid model demonstrated superior performance, achieving the lowest Mean Error (ME) of 0.1684864, Mean Absolute Error (MAE) of 1.051846, Mean Squared Error (MSE) of 3.500952, and highest R-squared of 0.998797. The proposed hybrid model outperformed both the Generalized Stiffness Formulation (GSF) and standalone ANN models. The significant improvement in prediction accuracy underscores the novelty and robustness of the hybrid approach in composite material modeling. Furthermore, this method can be used to refine any existing physical model by focusing on improving these established models to match experimental results and reducing the discrepancies, which offers a more efficient and attractive strategy for accurate predictions.

Air leakage mechanism and hole sealing technology in directional long-drilled perimeter rock-borehole composite fissures: for gas enrichment in varying coal-seam thickness change areas

The issue of gas enrichment in the tectonic zone of coal seams, which poses significant threats to coal mine safety and leads to gas disasters, is primarily addressed through borehole extraction. The quality of this extraction is notably influenced by air leakage and sealing technologies used during drilling. To tackle the problem of gas leakage in directional long boreholes within the floor, a composite fissure leakage model for the surrounding rock and borehole is developed. This model inverts the characteristics of coal seam thickness variations and abnormal gas enrichment through heterogeneous geostatistical modeling. It enables the quantitative characterization of the air leakage mechanism in directional long boreholes of the bottom plate and proposes an effective sealing process. Due to damage and disturbance, fissures and irregular plastic zones form around the excavated roadway, which are major contributors to air leakage. Over time, the air content in the coal and rock seams gradually increases while the gas content and concentration decrease continuously. Compared to the polyurethane "two plugs and one injection" sealing process, the multi-stage grouting and sealing process provides a more stable gas extraction concentration and flow rate in the sealed drill holes. This process effectively seals the fissure zones of the roadway, reduces air leakage from the boreholes, and improves gas extraction concentration.

Inverse feedforward control of piezoelectric actuators using optimized composite neural network-based Hammerstein model

This paper utilizes the optimized composite neural network (OCNN)-Hammerstein model to directly identify the dynamic inverse hysteresis effect, employing it as a feedforward controller to compensate for the rate-dependent and amplitude-dependent dynamic hysteresis of the piezoelectric actuator. The OCNN-Hammerstein model consists of an optimized composite neural network and an auto-regressive exogenous model in series. The OCNN comprises convolutional neural network layer and radial basis function neural network layer, with its hyperparameters optimized using a modified grey wolf optimizer algorithm. Compared to the existing dynamic Prandtl-Ishlinskii-Hammerstein and dynamic Bouc-Wen-Hammerstein models in the literature, the feedforward controller based on the proposed model in this paper demonstrates superior performance, with an RSME below 0.45 μm and a 74.5% reduction in relative hysteresis at the condition of 300 Hz and maximum amplitude. The feedforward controller exhibits universality across all amplitude ranges and within the 50–300 Hz frequency range, while also providing predictive compensation for dynamic hysteresis at 300–400 Hz. The feedforward controller based on the OCNN-Hammerstein model in this paper provides a crucial methodology for open-loop control of piezoelectric actuators.

SRNAdeep: a novel tool for bacterial sRNA prediction based on DistilBERT encoding mode and deep learning algorithms

BackgroundBacterial small regulatory RNA (sRNA) plays a crucial role in cell metabolism and could be used as a new potential drug target in the treatment of pathogen-induced disease. However, experimental methods for identifying sRNAs still require a large investment of human and material resources.MethodsIn this study, we propose a novel sRNA prediction model called sRNAdeep based on the DistilBERT feature extraction and TextCNN methods. The sRNA and non-sRNA sequences of bacteria were considered as sentences and then fed into a composite model consisting of deep learning models to evaluate classification performance.ResultsBy filtering sRNAs from BSRD database, we obtained a validation dataset comprised of 2438 positive and 4730 negative samples. The benchmark experiments showed that sRNAdeep displayed better performance in the various indexes compared to previous sRNA prediction tools. By applying our tool to Mycobacterium tuberculosis (MTB) genome, we have identified 21 sRNAs within the intergenic and intron regions. A set of 272 targeted genes regulated by these sRNAs were also captured in MTB. The coding proteins of two genes (lysX and icd1) are implicated in drug response, with significant active sites related to drug resistance mechanisms of MTB.ConclusionIn conclusion, our newly developed sRNAdeep can help researchers identify bacterial sRNAs more precisely and can be freely available from https://github.com/pyajagod/sRNAdeep.git.

Multisource Data-Driven Carbon Price Composite Forecasting Model: Based on Feature Selection and Multiscale Prediction Strategy

Accurate prediction of carbon trading prices is crucial for guiding investors to make informed decisions and assisting governments in formulating scientific carbon trading policies. This paper introduces a multisource data-driven carbon price composite forecasting model, aimed at enhancing prediction accuracy through advanced data processing and analysis methods. The model initially employs a secondary decomposition strategy, including Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (CEEMDAN) and Variational Mode Decomposition (VMD) methods, to decompose the original data series into three sub-sequences of different frequencies. Subsequently, it utilizes the Partial Autocorrelation Function (PACF) and Random Forest algorithm for feature selection to determine the input variables for different frequency sequences and conducts in-depth analysis and selection of influencing factors, including unstructured data. Furthermore, the model employs a multiscale forecasting strategy, combining Particle Swarm Optimization (PSO) enhanced Bidirectional Long Short-Term Memory (BiLSTM) and Extreme Gradient Boosting (XGBoost) models, along with the Autoregressive Integrated Moving Average (ARIMA) method, to predict based on the characteristics of different frequency components. Finally, the forecasts are integrated using PSO-BiLSTM to form a comprehensive forecast. Notably, given the high correlation between the trend series and influencing factor variables, the model jointly predicts them. A case study based on the Guangdong carbon market in China demonstrates that the proposed composite forecasting model outperforms other benchmark models, with Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and Mean Absolute Percentage Error (MAPE) values of 0.4009, 0.2699, and 0.5183%, respectively. This forecasting model provides an effective tool for predicting and analyzing carbon price fluctuations, offering new insights for precise carbon market price predictions.

Containerized Freight Index Prediction Integrating Deep Learning and Time Series Analysis

In the context of globalization, the shipping industry, as a core pillar of international trade, is crucial for global economic stability. This study focuses on the China Containerized Freight Index (CCFI) and proposes a composite deep learning model combining CEEMDAN, CNN, LSTM, and SENet for CCFI analysis and prediction. Initially, the CEEMDAN algorithm is used to decompose the CCFI time series, extracting useful signals from noise. Then, CNN extracts features, LSTM model sequences, and SENet enhances feature representation. The CEEMDAN-CNN-LSTM-SENet model achieves superior predictive performance with MAE of 29.5171, RMSE of 33.7638, and MAPE of 1.56%. Compared to other models like ECL (MAE 77.0951, RMSE 57.3559, MAPE 3.25%), our model shows approximately 61.7% improvement in MAE. This research offers a robust tool for shipping market analysis and policy-making, providing new perspectives for handling complex economic data.

Molecular dynamics study on mechanical properties and plastic deformation mechanism of GNPs/Mg composites

This study utilizes molecular dynamics methods to embedded GNPs (graphene nanosheets) of uniform size and thickness into single crystal Mg, establishing a GNPs/Mg composite model. The mechanical properties and plastic deformation behavior of the composite under various compression conditions (volume fraction of GNPs, strain rate, and temperature) are investigated. The results indicate that uniformly distributed GNPs enhance load transfer and provide dispersion strengthening, effectively improving the composite's mechanical properties. As the volume fraction of GNPs increases, the yield stress of the composite first increases and then decreases. When the volume fraction of GNPs is 3.9%, both the yield stress and Young's modulus of the composite reach their maximum values of 9.16GPa and 108.83GPa, respectively. The uniformly dispersed GNPs constrain dislocations on prismatic and pyramidal planes, thereby suppressing twinning deformation. Consequently, dislocation slip predominantly occurs as 1/3<-1100> partial dislocations on the (0001) basal plane. Young's modulus is not sensitive to strain rate; as the strain rate increases from 0.0025Å·ps⁻¹ to 0.0125Å·ps⁻¹, the yield stress rises from 8.50GPa to 10.42GPa. At high strain rates, only some dislocation sources are activated, and the number of 1/3<-1100> partial dislocations decreases with increasing strain rate. Within the temperature range of 100K to 500K, both the elastic modulus and yield stress of the GNPs/Mg composite decrease with increasing temperature.

Experimental and numerical investigation on composite beam–column joint connection behavior using different types of connection schemes

Abstract Over the past few decades, composite joints have been regarded as one of the most promising approaches. After reviewing the research articles dealing with composite beam-column joints, it can be withdrawn that few researchers studied and developed different types of composite beam-column joints. The objectives of this paper to study the effect and behavior of different connection types on the composite beam – column joint under repeated loading as well as to investigate new forms of slab connection with column flange in composite connection joints, aiming to enhance the behaviors of the joints and make the slab endure a part of the load in a more efficient way. Five specimens with various connection types were tested to failure; the composite model number two (CM2) was considered the control specimen. Regarding the ultimate moment, the result found that CM5 has moment 50.47% more than control specimen, the lowest rotation value was in CM5. The general behaviours of the FEM were in good accordance with the experiment results. According to the parametric a higher yield strength is more rigid stiffer, the effect of web thickness on the load is too small and increase in load resistance with an increase in beam flange thickness.

Prospects for rank-reduced CCSD(T) in the context of high-accuracy thermochemistry.

Obtaining sub-chemical accuracy (1 kJ mol-1) for reaction energies of medium-sized gas-phase molecules is a longstanding challenge in the field of thermochemical modeling. The perturbative triples correction to coupled-cluster single double triple [CCSD(T)] constitutes an important component of all high-accuracy composite model chemistries that obtain this accuracy but can be a roadblock in the calculation of medium to large systems due to its O(N7) scaling, particularly in HEAT-like model chemistries that eschew separation of core and valence correlation. This study extends the work of Lesiuk [J. Chem. Phys. 156, 064103 (2022)] with new approximate methods and assesses the accuracy of five different approximations of (T) in the context of a subset of molecules selected from the W4-17 dataset. It is demonstrated that all of these approximate methods can achieve sub-0.1 kJ mol-1 accuracy with respect to canonical, density-fitted (T) contributions with a modest number of projectors. The approximation labeled Z̃T appears to offer the best trade-off between cost and accuracy and shows significant promise in an order-of-magnitude reduction in the computational cost of the CCSD(T) component of high-accuracy model chemistries.

A step function based recursion method for 0/1 deep neural networks

The deep neural network with step function activation (0/1 DNNs) is a fundamental composite model in deep learning which has high efficiency and robustness to outliers. However, due to the discontinuity and lacking subgradient information of the 0/1 DNNs model, prior researches are largely focused on designing continuous functions to approximate the step activation and developing continuous optimization methods. In this paper, by introducing two sets of network node variables into the 0/1 DNNs and by exploring the composite structure of the resulted model, the 0/1 DNNs is decomposed into a unary optimization model associated with the step function and three derivational optimization subproblems associated with the other variables. For the unary optimization model and two derivational optimization subproblems, we present a closed form solution, and for the third derivational optimization subproblem, we propose an efficient proximal method. Based on this, a globally convergent step function based recursion method for the 0/1 DNNs is developed. The efficiency and performance of the proposed algorithm are validated via theoretical analysis as well as some illustrative numerical examples on classifying MNIST, FashionMNIST and Cifar10 datasets.

Carbon dioxide injection into the Achimov formations using reinjection technology by the example of Ach3-4 formation in the Novo-Urengoy area of the Urengoy field

The prerequisites for the study are the calculation results for the cycling process, where carbon dioxide is proposed as the injection agent into the Achimov formations instead of dry gas, with the goal of increasing the condensate recovery factor. The work is focused on the efficiency assessment of carbon dioxide reinjection technology and reducing carbon footprint at a late stage of field development. The research object is the Аch3-4 formation within the Novo-Urengoy license area of the Urengoy field. The leading method to identify this problem is the results of the full-scale composite dynamic model in the ECLIPSE 300 format. The model takes into account the history of field development on depletion. The articles deals with two schemes for injecting carbon dioxide into the formation. In the first scheme, pure carbon dioxide is injected in a closed-loop system, but carbon neutrality through storage is not achieved. In the second scheme, carbon dioxide is injected using reinjection technology. Once injection begins, gas production stops. Only the condensate separated from the formation gas during low-temperature separation is sold and sent for further processing. After allocation of the condensate, the mixture of natural gas and carbon dioxide, in a specific proportion, is sent to the compressor station for reinjection into the formation in a gaseous state. Injecting pure carbon dioxide achieves a condensate recovery factor similar to that of gas injection with a 30 % carbon dioxide mixture. However, this option is less economically viable compared to the base and other scenarios due to high capital costs for upgrading the existing gas processing equipment (requiring the construction of an amine treatment unit). With carbon dioxide injection using reinjection technology, in addition to recovering extra condensate that had condensed during natural depletion, a reduction in the carbon footprint is also achieved. To maximize the condensate recovery factor, the optimal concentration of carbon dioxide in the injection mixture has been determined. The optimal timing for the start of injection was identified to maximize gas recovery. Economic efficiency is expected from the additional recovery of condensate trapped in the reservoir and from achieving carbon neutrality through the monetization and storage of carbon dioxide.

A composite analytical model to predict the thermal performance of borehole ground heat exchangers within stratified ground

The thermal performance of borehole ground heat exchangers is of significance to the efficacy of the ground source heat supply systems. This paper presents a new composite analytical model to evaluate the thermal performance of borehole ground heat exchangers within stratified ground. The model integrates critical factors influencing heat transfer, encompassing ground stratification, thermal interactions among boreholes, variability of heat flux along borehole depth, and flexible layouts of boreholes. It enables efficient computation of thermal data pertinent to heat transfer both inside and outside the boreholes during operation. Implemented in the TRNSYS platform, the model offers versatile applicability to ground-source heat pump systems, borehole thermal energy storage, and district heating and cooling networks. Model validations were carried out through four distinct experiments, covering varied borehole quantities and ground conditions, to substantiate its accuracy. Employing this model, a case study further investigated the thermal performance of a borehole field with 16 boreholes in a three-layer ground (comprising backfill soil, clay, and fine sand) extending to a depth of 63 m. Over a 60-day warm water injection period at 30 °C, the effects of borehole layout and ground stratification on the thermal performance of individual and collective boreholes were analysed. Key findings in this analysis include: (1) Across the square, L-shape, and linear layouts, the impact of layout on the thermal performance of borehole fields diminishes as the spacing increases from 1 m, 2 m, and 3 m, becoming negligible at 4-m spacing, and the positions of the most and least affected boreholes by thermal interferences vary by layout; (2) Designing a GSHP system in stratified ground using a homogenous ground model with equivalent thermal properties would overestimate the heat injection performance of the borehole field and lead to undersizing of the borehole ground heat exchangers; and (3) ground stratification should be considered when estimating the thermal performance of a borehole field, particularly when the boreholes are sparsely arranged, e.g., with spacing of 3 m or more.