Large-scale Engineering Research Articles

Thermal Frictional Analysis of a Novel Vibrating Cylindrical Turbulator in Double Tube Heat Exchangers for Engine Cooling

This study focuses on enhancing the thermal efficiency of a double-tube heat exchanger used in cooling systems for large-scale internal combustion engines by incorporating a novel vibrating cylindrical turbulator. The experimental investigation examined inlet Reynolds numbers ranging from 1052 to 8430 and evaluated various turbulator configurations, including fixed close-ended, fixed open-ended, vibrating close-ended, and vibrating open-ended cylindrical turbulators. Additionally, the impact of turbulator length, varying from 10 to 30 cm, on thermal-frictional characteristics was analyzed. The optimal configuration was determined using the thermal enhancement factor (TEF). Results showed that close-ended cylindrical turbulators significantly outperformed the open-ended configurations in terms of heat transfer and TEF. Although the vibrating turbulator produced a higher pressure drop compared to the fixed turbulator, it achieved a much higher heat transfer rate and TEF, making it a viable option for heat exchangers. The study also found that increasing the turbulator length leads to increased heat transfer, TEF, and pressure drop. The maximum TEF was recorded with the vibrating close-ended cylindrical turbulator, where heat transfer and pressure drop were up to 385% and 95% greater than those of a plain tube heat exchanger, respectively, resulting in a perfect TEF value of 3.45.

In situ iron coating of amorphous boron and characterization of its energy release behavior

Boron was widely employed as metal fuel in solid propellants for its higher gravimetric and volumetric combustion enthalpies. However, the low melting point and high boiling point of B2O3 on the surface of boron particles significantly hinder its ignition and combustion performance, seriously preventing the full realization of its thermodynamic potential. To address this issue, B@Fe composite fuels with different iron contents from 1.59 % to 7.09 % were prepared via liquid phase reduction in this paper and the quasi-static high temperature oxidation, ignition and combustion reaction characteristics and corresponding mechanism were studied, in order to provide effective modification techniques in supporting the large-scale engineering application of boron as metal fuel. The study results show that B@Fe can effectively reduce the ignition temperature and maximum heat flow temperature of amorphous boron, and increase the maximum mass gain rate. The ignition temperature and maximum heat flow temperature have an exponential function with the iron coating content. Iron coated amorphous boron can effectively shorten the laser ignition delay time by more than 50 %, and significantly change the flame structure, flame evolution process and combustion duration. Both amorphous boron and iron-coated amorphous boron flame showed a multi-layer structure, including the outer green flame, the middle yellow flame and the central incandescent flame. However, with the increase of the iron coating content, the flame height increased from 5.5 cm to 6.3 cm, the flame structure changed from mushroom type to triangle type, and the central incandescent flame area increased significantly. The equivalent combustion duration increased from 975 ms to 1997 ms, and the flame temperature increased from 1288 to 1505 °C. The emission spectra of BO2 and BO lines are captured during combustion. B@Fe with the increase of iron content, a large number of small diameter micropores appear in the condensed combustion products(CCPs) of composite fuels, and a grid-like structure was formed. With the increase of iron coating content, the boron content of condensed combustion products decreases, and the oxygen content increases. The main components of CCP include B, B2O3, B6O, FeB, Fe2B and Fe3B. Based on the ignition combustion experiment of B@Fe composite fuel, ignition combustion analysis model of B@Fe composite fuel was established.

Size effect modellings of axial compressive failure of RC columns at low temperatures

This paper applied a thermal-mechanical sequential coupled mesoscopic simulation method to explore the axial compression performance and the corresponding size effect of Reinforced Concrete Columns confined by Stirrups (i.e., RCCS) at low temperatures, with considering the interaction between concrete meso-components and steel bars as well as the low-temperature effect of mechanical parameters. Based on the heat conduction analysis, the axial compression mechanical failure behavior of RCCS with four structural sizes (i.e., 267 × 267 × 801, 400 × 400 × 1200, 600 × 600 × 1800 and 800 × 800 × 2400 mm) and two stirrup ratios (i.e., 1.26% and 2.89%) at different temperatures (i.e., T = 20, −30, −60 and −90°C) was subsequently simulated. The effects of temperature, structural size and volume stirrup ratio on axial compression properties were quantitatively discussed. The results showed that the peak strength of RCCS increased with the decreasing temperature, and the smaller-sized RCCS showed a stronger effect of low-temperature enhancement. Both the residual strength and displacement ductility coefficient decreased with the decreasing temperature. The peak strength, residual strength and displacement ductility coefficient of RCCS decreased with the increasing structural size, showing obvious size effects. The size effect on peak strength increased with the decreasing temperature, (the maximum increase was nearly 140%), but the size effect on displacement ductility coefficient decreased (the maximum decrease was nearly 70%). The peak strength, residual strength and ductility were enhanced with the increasing volume stirrup ratio, which was helpful to reduce the influence of size effect. Finally, an improved size effect theoretical model was proposed, which can effectively predict the axial compressive strength of RCCS with different structural sizes and stirrup ratios at room and low temperatures. The present research results can provide reference for the large-scale engineering application of RCCS in low-temperature environments.

Research on Activation Mechanisms In Situ during Rapid Desorption of Mercury-Absorbing Activated Coke by the Heating Method.

The adsorption of active coke is a good method to control mercury in coal-fired power plants, but spent powdered activated coke (SPAC) will cause secondary pollution and waste of resources if it is not properly treated. The purpose of this study was to explore the desorption performance of SPAC when heated in a drop-tube reactor under different atmospheres. The carbon consumption was 0.13%, 0.24%, 0.17%, 0.16%, and 0.14%, under desorption conditions in N2, O2, H2O, CO2, and SFG (N2 + 12% CO2 + 6% H2O + 6% O2) atmospheres, respectively. The readsorption performances of PAC-N2, PAC-CO2, PAC-H2O, PAC-O2, and PAC-SFG were 40%, 72%, 77%, 58%, and 86%, respectively. In particular, compared with the original PAC, the physicochemical properties of PAC-SFG (optimum conditions) changed greatly; the surface area of PAC-SFG and the pore volume increased by 14.1% and 11.6%, respectively, and the average pore diameter decreased by 0.203 nm. Furthermore, the total acid content of PAC-SFG was 0.505, which was greater than those of the other samples. Additionally, the desorption and activation mechanisms in situ were determined under the optimum reaction conditions. The results showed that the activator gas reacted with C to form new functional groups, Lewis acid sites, and developed pore structures, which could increase Hg0 adsorption efficiency from 66% to 86% after the adsorption/desorption cycles. This study provides a novel idea for the reuse of SPAC, and this research is expected to provide valuable guidance for Hg0 removal and SPAC activated in situ during rapid desorption processes in future large-scale engineering applications.

Surface modification of fibres with silane for enhancing the durability of FRP composites as reinforcement for seawater sea sand concrete (SWSSC): A review

This review discusses the necessity of surface modification to effectively use fibre-reinforced polymers (FRPs) as reinforcements for seawater and sea sand concrete (SWSSC). FRPs can be a viable reinforcement material for SWSSC as an alternative to the traditional carbon steel reinforcement that will suffer unacceptably high rates of corrosion due to the very high chloride content of the SWSSC environment. FRPs exhibit mechanical properties similar to steels but suffer insignificant degradation/corrosion due to chloride. However, the fibres in certain FRPs (such as basalt FRPs and glass FRPs) suffer degradation due to attack by the alkaline environment of concrete, causing degradation in fibre-matrix interfacial strength and loss of FRPs’s overall mechanical strength and durability. Therefore, fibres used in such FRPs require suitable surface modification to ameliorate their degradation, enabling their practical use as reinforcements of SWSSC. While carbon fibre-reinforced polymers (CFRPs) demonstrate superior resistance to alkali-induced degradation, their widespread use in large-scale civil engineering is hindered by high costs. In this review, more economic alternatives, i.e., basalt fibre-reinforced polymers (BFRPs) and glass fibre-reinforced polymers (GFRPs) are critically reviewed. To enhance the durability of FRPs and mitigate alkali-induced degradation of the basalt/glass fibres, suitable surface modifications are necessary to safeguard the mechanical properties of FRPs. Underpinned by a comprehensive discussion on the necessity of surface modification of fibres for the use of FRPs in SWSSC environment, this review focuses on the potential of advanced approaches for surface modifications of the fibres, i.e., coatings of silanes, and silanes impregnated with graphene nanoplatelets (GNPs), graphene oxide (GO) and its functionalised derivatives or metal oxides (such as SiO2, ZrO2, Al2O3, TiO2 and CeO2). This review critically discusses the opportunities for suitable surface modification of basalt and glass fibres for enabling the durable use of FRPs as reinforcements for sweater sea sand concrete.

Modified Easley formula for elastic critical global shear buckling stress of corrugated steel webs considering real boundary conditions

The real juncture between corrugated steel webs (CSWs) and flanges follows a multi-segmented line, distinct from that of flat steel webs. Classic methods may yield significant deviations in predicting the elastic global shear buckling capacity of CSWs of various scales due to their failure to consider real boundary constraints. Therefore, a universally applicable formula for calculating the elastic critical global shear buckling stress of CSWs, which accounts for real boundary conditions, is proposed. This formula is pertinent to both large-scale engineering CSWs and small-scale testing CSWs. This study commenced with a comprehensive reassessment of the elastic global shear buckling calculation method. Subsequently, the influence of geometric parameter ratios on the elastic critical global buckling stress was examined. The primary parameter was identified and employed to improve the global buckling coefficient. The proposed calculation method was validated using different corrugation configurations, including 1000-type, 1200-type, 1600-type, 1800-type, and 2000-type CSWs, as well as other CSWs used in experimental settings. These results were compared with those obtained from other reference methods. Findings indicate that the accuracy of the classic theoretical method is affected by variations in both boundary conditions and geometric dimensions due to the constraint effect of real boundary conditions. Under the real boundary conditions, the elastic critical global shear buckling stress of CSWs with simply supported boundary conditions is close to that of CSWs with consolidated boundary conditions. The ratio of web height to corrugation depth primarily affects the elastic global shear buckling capacity, which decreases as the ratio increases. The Easley formula can be modified based on the web height to corrugation depth ratio. Comparisons of numerous numerical and theoretical results reveal that the proposed calculation method exhibits commendable computational precision. In comparison to alternative formulas, the proposed method demonstrates enhanced consistency for calculating CSWs with varying geometric dimensions and boundary conditions, thereby demonstrating its favorable applicability. These conclusions provide valuable reference for the shear design of CSWs.

Three-dimensional Change Detection and Description in Complex Construction Scenarios

Abstract. In large-scale engineering construction scenarios, construction objects and equipment are numerous; their spatial positions and postures change frequently and are complex and varied. Describing the changes between construction entities and the entities themselves in a three-dimensional (3D) geometric space and functional semantic space is challenging. In response to this challenging, we propose a novel workflow to detect and describe the changes in a construction site. First, we add the constraint of object surface continuity to the octree change detection (OCD), exclude the independently changing cells, and improve the change detection robustness. Second, based on the changes in entity elements that occur in the 3D geometric space, we introduce six-type semantics to characterize the changes that occur. Differing from existing change detection semantic description methods that focus on the attribute semantics of changed entity elements, our method focuses on the semantic description of the change process. Finally, for the variations in different semantic types, we propose specific quantization methods that can better quantify the variations in entity elements in 3D geometric space. For the proposed solution, we tested it using two sets of multi-period time-series point cloud data, which can accurately identify the changing construction entities in space and complete the calculation of bridge construction progress, including the building demolition and construction, and the change in the construction apparatus position.

Geomechanical Classifications, Geotechnical Indexes, and Fractured Rock Media: the Influence of Discontinuities on the Rock Masses Description

Using geotechnical indexes and geomechanical classifications is crucial in estimating the quality and behaviour of fissured rock masses. These tools play a significant role in large-scale engineering works, particularly in underground projects where the rock mass and its response to excavation are critical for project safety and financial feasibility. The widespread adoption of these classification systems in geoengineering necessitates continuous development and improvement to enhance accuracy and reliability and align them with the evolving construction landscape. Discontinuities, in particular, profoundly impact the strength, deformability, and permeability of the rock mass, therefore defining its behaviour. Given that rock mass assessment forms the basis of geotechnical characterization and evaluation, it is essential to understand and evaluate its characteristics and their influence on the classification systems or indexes.

On application of the relative entropy concept in reliability assessment of some engineering cable structures

The main research problem studied in this work is an uncertain response and reliability assessment of the spatial cable structures due to the environmental stochasticity as well as material and geometrical imperfections. Some popular cable structures are analyzed for this purpose using the Stochastic Finite Element Method (SFEM) implemented with the use of three different techniques, namely the iterative generalized perturbation method, semi-analytical approach as well as the Monte-Carlo simulation. Uncertainty quantification delivered in this study is based on the series of FEM analyses of both static and dynamic structural problems. They enable the Least Squares Method determination of the structural polynomial responses linking extreme stresses and deformations with several uncorrelated uncertainty sources. Reliability assessment, fundamental in durability and Structural Health Monitoring, is completed using a comparison of the First Order Reliability Method (FORM) with probabilistic distance formulated by Bhattacharyya. Input uncertainties are assumed to be Gaussian according to the Maximum Entropy Principle. They have specific expected values following engineering design demands or the provisions of designing codes, whereas their standard deviations do not exceed the 10% level. The methods presented and the results obtained in this study may serve for further reliability analyses of large-scale civil engineering structures completed with both steel cables and also reinforced concrete plates like suspended bridges, for instance.

Multiple network embedding for anomaly detection in time series of graphs

The problem of anomaly detection in time series of graphs is considered, focusing on two related inference tasks: the detection of anomalous graphs within a time series and the detection of temporally anomalous vertices. These tasks are approached via the adaptation of multiple adjacency spectral embedding (MASE), a statistically principled method for joint graph inference. The effectiveness of the method is demonstrated for these inference tasks, and its performance is assessed based on the nature of detectable anomalies. Theoretical justification is provided, along with insights into its use. The approach identifies anomalous vertices beyond just large degree changes when applied to the Enron communication graph, a large-scale commercial search engine time series, and a larval Drosophila connectome.

Development of an improved mass transfer model for condensation with dynamic feedback regulation

In the field of condensation research, accurate and efficient numerical simulation models are highly desired. One of the widely used phase change mass transfer models is known as the Lee model. However, this model has limitations due to uncertainties in the mass transfer intensity factor, which can lead to simulation failures and impede large-scale engineering applications. This study presents an improved mass transfer model which employs a tuning function to dynamically and accurately modify the mass transfer intensity factor in each two-phase cell via feedback adjustment. This approach enables immediate and customized adjustment of the factor for each grid within the limit of the tuning function, significantly reducing the errors caused by repeated artificial adjustments and related uncertainties. The adjustment capabilities of the hyperbolic tangent and softsign functions are evaluated. The softsign function is superior in both regulation efficiency and accuracy. The model's accuracy is verified by comparing with previous experiments and simulations, as well as by the one-dimensional Stefan problem. The proposed model achieves high prediction accuracy, reduces the risk of simulation dispersion, and significantly enhances computational efficiency with the iteration number reduced by nearly half under certain conditions compared to the Lee model. Furthermore, the new model shows robustness as the prediction accuracy is insensitive to variations in critical parameters. The improved model is expected to provide valuable assistance for future research and applications in flow condensation with outstanding computational accuracy, efficiency, and stability.

Versatile filamentous fungal host highly-producing heterologous natural products developed by genome editing-mediated engineering of multiple metabolic pathways

Natural secondary metabolites are medically, agriculturally, and industrially beneficial to humans. For mass production, a heterologous production system is required, and various metabolic engineering trials have been reported in Escherichia coli and Saccharomyces cerevisiae to increase their production levels. Recently, filamentous fungi, especially Aspergillus oryzae, have been expected to be excellent hosts for the heterologous production of natural products; however, large-scale metabolic engineering has hardly been reported. Here, we elucidated candidate metabolic pathways to be modified for increased model terpene production by RNA-seq and metabolome analyses in A. oryzae and selected pathways such as ethanol fermentation, cytosolic acetyl-CoA production from citrate, and the mevalonate pathway. We performed metabolic modifications targeting these pathways using CRISPR/Cas9 genome editing and demonstrated their effectiveness in heterologous terpene production. Finally, a strain containing 13 metabolic modifications was generated, which showed enhanced heterologous production of pleuromutilin (8.5-fold), aphidicolin (65.6-fold), and ophiobolin C (28.5-fold) compared to the unmodified A. oryzae strain. Therefore, the strain generated by engineering multiple metabolic pathways can be employed as a versatile highly-producing host for a wide variety of terpenes.

The field-enriched finite element method with gravity effects for simulating the cracking behaviors of large-scale engineering rock masses

The field-enriched finite element method uses a scalar field defined as a field variable to describe cracks and characterize their impact on the displacement field and stress field of the solution model. It is capable of avoiding remeshing and employing level set functions to describe cracks when simulating the propagation of cracks. In this work, a field-enriched finite element model with gravity effects is proposed to simulate the large-scale failure process of engineering rock masses, and several numerical cases of geotechnical engineering are successfully analyzed. First, by introducing the unified tensile fracture criterion into the numerical model, the large-scale failure process of the intact slope is simulated. Second, the sliding process of rock slopes containing en echelon joints is numerically investigated. Third, the cracking process of the concrete dam is analyzed. Finally, the effects of joint and bedding plane inclination angles on the stability of tunnel chamber in transversely isotropic rock mass are studied. The numerical results indicate that the numerical method proposed in this work can accurately solve the large-scale failure process of rock masses.

Dams, dams and more dams: Issues in evaluating business cases for dam expansion in Australia

Abstract Conventional interventions for mitigating flood risk and increasing water security often focus on large-scale engineering solutions despite evidence of significant cost overruns and negative environmental impacts. This study is an example of participatory research in water economics, motivated (and requested) by concerns from downstream communities and Indigenous stakeholders that impacts of a proposed dam upgrade and storage increase were not adequately considered. We reviewed the preliminary benefit-cost analysis (BCA) of the proposed AUD$2.1 billion investment to expand the Wyangala Dam in Australia and conducted a new BCA incorporating downstream community perspectives. The original economic analysis favoured the proposed investment, but it underestimated costs and overestimated benefits. Our economic analysis was complemented with a qualitative analysis of downstream community perspectives, plus a quantitative analysis indicating that project costs were underestimated by a minimum of 116%. In comparison, benefits were overestimated by 56%. Neglecting the potential impact of climate change also severely overestimated the original benefit-cost ratio. Based on our calculations, expanding the dam was unlikely to yield a net social benefit. In 2023, a new government decided the dam expansion was too expensive. We recommend policymakers prioritise independent evaluations and community engagement for BCAs on large-scale water infrastructure projects to ensure equitable investment decisions that maximise social welfare and adequate environmental assessment.

A comprehensive investigation of the adsorption behaviour and mechanism of industrial waste sintering and bayer red muds for heavy metals.

The issue of heavy metal pollution is a critical global concern that requires urgent solution. However, conventional heavy metal adsorbents are too costly to be applied in large-scale engineering. In this study, adsorption behavior and mechanism of sintering red mud (RM-A) and bayer red mud (RM-B) for heavy metals were investigated to address the disposal of red mud as industrial waste and remediation of heavy metal pollution. Batch adsorption experiments were conducted to explore the adsorption performances of RM-A and RM-B under various conditions. Characterization of RM-A and RM-B before and after adsorption by XRD, FTIR and SEM-EDX was applied to investigate the specific adsorption behavior and mechanism. Adsorption experiments of both RM-A and RM-B fitted pseudo-second-order kinetic model and Langmuir isotherm model, with estimated maximum adsorption capacity of 21.96 and 25.19 mg/g for Cd2+, 21.47 and 26.06 mg/g for Cu2+ and 55.47 and 59.65 mg/g for Pb2+, respectively. Precipitation transformation of calcite was the primary adsorption mechanism for RM-A, whereas ion exchange of cancrinite, surface coordination compounds of hematite and minor precipitation transformation of calcite accounted for the adsorption mechanism for RM-B. Overall, RM-A and RM-B exhibited best adsorption performance for Pb2+, with RM-B showing greater adsorption capacity attributed to its higher specific surface area. This study compared the adsorption properties of RM-A and RM-B for the first time and demonstrated that both red muds can be effectively applied to remove heavy metals, thereby contributing to the sustainable industrial waste management and resourceful reuse.

Nonlinear dynamics of wing-like structures using a momentum subspace-based Koiter-Newton reduction

Cantilevers find a wide range of applications in the design of scientific equipment and large-scale engineering structures such as aircraft wings. Analysis techniques based on linearization approximations are unable to capture the large amplitude oscillation behaviour of such structures and thus, necessitates development of dedicated nonlinear methods. In this work, the recent developments in the Koiter-Newton model reduction method are utilized to obtain nonlinear reduced order models (ROMs) from full finite element structural models in order to simulate large amplitude dynamics of cantilevers. The method describes a nonlinear system of governing equations comprising quadratic and cubic terms which are obtained as higher order derivatives of the in-plane strain energy. To ensure that the large rotations in cantilevers and the resultant foreshortening effect is also accounted for, a ROM updating algorithm is adopted where the ROM parameters are varied with the structural deflections. Linear eigenmodes of the structure are utilized to form the reduction subspace. To validate the methodology, the ROM solution is compared against experimental results and a convergence study is conducted to identify the number of modes needed to replicate the nonlinear response. Finally, a composite wingbox structure is considered for which time domain simulations are conducted and frequency response curves, obtained through a frequency sweep, are presented.

An efficient flux-reconstructed lattice boltzmann flux solver for flow interaction of multi-structure with curved boundary

Recently, the escalating computational capability provided by advanced GPU technology for numerical simulations is well-suited for tackling large-scale engineering challenges. The lattice Boltzmann flux solver (LBFS), as a relatively new fluid-solving method, combines the parallelization characteristics of the lattice Boltzmann method (LBM) with the ability to handle non-uniform grids. Building upon these advantages, this study aligns its flux computational steps with the demands of the GPU parallel computing environment, thus developing an efficient flux-reconstructed lattice Boltzmann flux solver (FRLBFS), the improved algorithm not only ensures precision but also achieves a significant enhancement in efficiency, reaching up to a remarkable 500-fold improvement. Additionally, this work combines the immersed boundary method, significantly improving the efficiency of addressing hydrodynamic problems with multiple structures. Through numerical validation, under identical GPU hardware conditions, the proposed method in this paper outperform LBM in terms of computational efficiency. Lastly, simulations of flow past an array of eight cylinders at different Reynolds numbers are conducted, instantaneous contour plots at various dimensionless time points and time-dependent curves of drag and lift coefficients for different cylinders are provided, to demonstrate the complex vortex shedding surrounding multiple cylinders and its extraordinary computational efficiency.

Integrated fiber paper-based composites enabling photo-Fenton synergistic degradation and switchable adsorption for efficient and multitasking pollutants removal

Single catalytic systems have received wide attention for eradicating pervasive contaminants in water but lack multifunctional integrated biobased capturing and purifying system (bio-CPs). In this work, a universal design of integrated bio-CPs with layered structure driven from fiber filtering paper (FP) was demonstrated for multitasking environmental remediation. Fe-rich composite catalysts (Fe-BiOBr/TiO2) and polyphenol-modified lignin system (PLPR) were herein first developed for the full decoration of FP substrates, achieving the facile construction of layered structure with highly active surface and photocatalytic layer. Profiting from this, a photo-Fenton synergetic degradation system on composites was designed, facilitating a remarkable degradation capability (99.24 % for MEB) and full-time operation ability. Driven by highly enhanced interfacial interactions mediated by polyphenol, this material was capable of efficient capture pollutants (165.29 mg·g-1 for Cr (VI)). And assisted with well-maintained high water flux feature, this bio-CPs that efficient remove pollutants by flowthrough capture were realized, showing the switchable application. Significantly, the convenience of surface deposition and self-assembly supported the potential applications in real system based on extended preparation of such bio-CPs. This work provided a novel and universal strategy to develop the universal and multifunctional integrated bio-CPs to support for large-scale, cost-effective, and sustainable wastewater remediation engineering.

The past, present and future of multi-scale modelling applied to wave-structure interaction in ocean engineering.

Concepts and evolution of multi-scale modelling from the perspective of wave-structure interaction have been discussed. In this regard, both domain and functional decomposition approaches have come into being. In domain decomposition, the computational domain is spatially segregated to handle the far-field using potential flow models and the near field using Navier-Stokes equations. In functional decomposition, the velocity field is separated into irrotational and rotational parts to facilitate identification of the free surface. These two approaches have been implemented alongside partitioned or monolithic schemes for modelling the structure. The applicability of multi-scale modelling approaches has been established using both mesh-based and meshless schemes. Owing to said diversity in numerical techniques, massively collaborative research has emerged, wherein comparative numerical studies are being carried out to identify shortcomings of developed codes and establish best-practices in numerical modelling. Machine learning is also being applied to handle large-scale ocean engineering problems. This paper reports on the past, present and future research consolidating the contributions made over the past 20 years. Some of these past as well as future research contributions have and shall be actualized through funding from the Newton International Fellowship as the next generation of researchers inherits the present-day expertise in multi-scale modelling. This article is part of the theme issue 'Celebrating the 15th anniversary of the Royal Society Newton International Fellowship'.

Domain Decomposition for Data-Driven Reduced Modeling of Large-Scale Systems

This paper focuses on the construction of accurate and predictive data-driven reduced models of large-scale numerical simulations with complex dynamics and sparse training datasets. In these settings, standard, single-domain approaches may be too inaccurate or may overfit and hence generalize poorly. Moreover, processing large-scale datasets typically requires significant memory and computing resources, which can render single-domain approaches computationally prohibitive. To address these challenges, we introduce a domain-decomposition formulation into the construction of a data-driven reduced model. In doing so, the basis functions used in the reduced model approximation become localized in space, which can increase the accuracy of the domain-decomposed approximation of the complex dynamics. The decomposition furthermore reduces the memory and computing requirements to process the underlying large-scale training dataset. We demonstrate the effectiveness and scalability of our approach in a large-scale three-dimensional unsteady rotating-detonation rocket engine simulation scenario with more than 75 million degrees of freedom and a sparse training dataset. Our results show that compared to the single-domain approach, the domain-decomposed version reduces both the training and prediction errors for pressure by up to 13% and up to 5% for other key quantities, such as temperature, and fuel, and oxidizer mass fractions. Lastly, our approach decreases the memory requirements for processing by almost a factor of four, which in turn reduces the computing requirements as well.