Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “data aggregation”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,333 records · Page 74
All sourcesNASA NTRSDOE OSTI
DOE OSTI2020

ECP ST Capability Assessment Report (CAR) for VTK-m (FY20)

The ECP/VTK-m project is providing the core capabilities to perform scientific visualization on Exascale architectures. The ECP/VTK-m project fills the critical feature gap of performing visualization and analysis on processors like graphics-based processors. The results of this project will be delivered in tools like ParaView, Vislt, and Ascent as well as in stand-alone form. Moreover, these projects are depending on this ECP effort to be able to make effective use of ECP architectures. One of the biggest recent changes in high-performance computing is the increasing use of accelerators. Accelerators contain processing cores that independently are inferior to a core in a typical CPU, but these cores are replicated and grouped such that their aggregate execution provides a very high computation rate at a much lower power. Current and future CPU processors also require much more explicit parallelism. Each successive version of the hardware packs more cores into each processor, and technologies like hyper threading and vector operations require even more parallel processing to leverage each core's full potential. VTK-m is a toolkit of scientific visualization algorithms for emerging processor architectures. VTK-m supports the fine-grained concurrency for data analysis and visualization algorithms required to drive extreme scale computing by providing abstract models for data and execution that can be applied to a variety of algorithms across many different processor architectures. The ECP/VTK-m project is building up the VTK-m codebase with the necessary visualization algorithm implementations that run across the varied hardware platforms to be leveraged at the Exascale. We will be working with other ECP projects, such as ALPINE, to integrate the new VTK-m code into production software to enable visualization on our HPC systems.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

Building-to-Grid Modeling Framework with a Case Study of Battery Systems

Coordinating electricity load shifting operations of multiple buildings has the potential to achieve significant benefits for both building owners and grid operation. In support of this research, this paper develops a novel building-to-grid modeling framework, which uses EnergyPlus to develop physics-based building energy models and GridLAB-D to develop distribution-level grid models. The data exchanges between buildings and distribution-level grid are conducted using Python script and HELICS co-simulation platform. Then, by using this modeling framework, this paper conducts a case study of battery systems in a connected community, which consists of 10 multi-family buildings with appropriately sized batteries. The three scenarios, baseline operation, tolerant to price fluctuations, and opportune usage, are studied. The result displays that, by adopting onsite batteries, both electricity costs for building owners and aggregated peak power demands are reduced.

Building-to-grid, connected community, co-simulati↗
DOE OSTI2018

Formation of Silica-Lysozyme Composites Through Co-Precipitation and Adsorption

Interactions between silica and proteins are crucial for the formation of biosilica and the production of novel functional hybrid materials for a range of industrial applications. The proteins control both precipitation pathway and the properties of the resulting silica–organic composites. Here, we present data on the formation of silica–lysozyme composites through two different synthesis approaches (co-precipitation vs. adsorption) and show that the chemical and structural properties of these composites, when analyzed using a combination of synchrotron-based scattering (total scattering and small-angle X-ray scattering), spectroscopic, electron microscopy, and potentiometric methods vary dramatically. We document that while lysozyme was not incorporated into nor did its presence alter the molecular structure of silica, it strongly enhanced the aggregation of silica particles due to electrostatic and potentially hydrophobic interactions, leading to the formation of composites with characteristics differing from pure silica. The differences increased with increasing lysozyme content for both synthesis approaches. Yet, the absolute changes differ substantially between the two sets of composites, as lysozyme did not just affect aggregation during co-precipitation but also particle growth and likely polymerization during co-precipitation. Our results improve the fundamental understanding of how organic macromolecules interact with dissolved and nanoparticulate silica and how these interactions control the formation pathway of silica–organic composites from sodium silicate solutions, a widely available and cheap starting material.

36 MATERIALS SCIENCE↗
DOE OSTI2025

Defining a Platform Approach and Market Participation: Data Driven Business Models for Solid State Transformer-Based Synthetic Inertia and Voltage Stability Controls (CRADA Final Report, Project 1, Mod 1)

The primary objective of this project is to determine the incremental value created with the medium voltage solid-state transformer (MV SST) technology to different stakeholders in view of the updated DER grid regulations. This includes studying the benefits of the MV SST technology in a range of use cases for EV and DER penetration including (1) “corridor charging” for EVs and (2) solar plus storage (FERC 2222). The potential customers of this technology include utilities for EV charging, DER installers who must meet utility interconnection requirements, balancing authorities, and DER aggregators. The traditional transformers on the grid could be a limiting factor for the EV-grid integration as the distribution transformers were not designed to handle the dynamic and fluctuating EV charging loads. Thus, the issues such as voltage fluctuations, increased losses and reduced efficiency [1] can negatively impact the grid operation. To address these challenges, transformers with flexibility and adaptability become imperative to meet the evolving energy demands. In this regard, the concept of Medium Voltage Solid-State Transformers.

14 SOLAR ENERGY↗
DOE OSTI2019

Transcriptomic atlas of mushroom development reveals conserved genes behind complex multicellularity in fungi

The evolution of complex multicellularity has been one of the major transitions in the history of life. In contrast to simple multicellular aggregates of cells, it has evolved only in a handful of lineages, including animals, embryophytes, red and brown algae, and fungi. Despite being a key step toward the evolution of complex organisms, the evolutionary origins and the genetic underpinnings of complex multicellularity are incompletely known. The development of fungal fruiting bodies from a hyphal thallus represents a transition from simple to complex multicellularity that is inducible under laboratory conditions. We constructed a reference atlas of mushroom formation based on developmental transcriptome data of six species and comparisons of >200 whole genomes, to elucidate the core genetic program of complex multicellularity and fruiting body development in mushroom-forming fungi (Agaricomycetes). Nearly 300 conserved gene families and >70 functional groups contained developmentally regulated genes from five to six species, covering functions related to fungal cell wall remodeling, targeted protein degradation, signal transduction, adhesion, and small secreted proteins (including effector-like orphan genes). Several of these families, including F-box proteins, expansin-like proteins, protein kinases, and transcription factors, showed expansions in Agaricomycetes, many of which convergently expanded in multicellular plants and/or animals too, reflecting convergent solutions to genetic hurdles imposed by complex multicellularity among independently evolved lineages. Finally, this study provides an entry point to studying mushroom development and complex multicellularity in one of the largest clades of complex eukaryotic organisms.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2015

A dynamically driven, universal thermal profile of galaxy groups and clusters

Large scale structures such as groups and clusters of galaxies show a universal, nearly linear entropy radial profile K(r). Using deprojected 16 clusters and 12 groups from the literature, we find that K∝r{sup 0.96±0.01}, consistent with the mean power-law index (0.9–1.1) of previous studies. A similarly good fit is given by a τ∝r{sup 0.72±0.01} ratio between cooling and free-fall times. Both profiles slightly flatten at small radii, as τ becomes of order unity. The entropy profile is usually attributed to self-similar shock accretion (shown to be inconsistent with the data), to non-standard heat conduction, or to turbulent heating. We argue that a dynamical mechanism is needed to sustain such a universal profile, oblivious to the temperature peak at the edge of the core and to the virial shock at the outskirts, and robust to the presence of ongoing cooling, merger, and active galactic nucleus activity. In particular, we show that such a profile can be naturally obtained in a spiral flow, which is likely to underlie most galaxy aggregates according to the ubiquitous spiral patterns and cold fronts observed. Generalizing a two-phase spiral flow model out to the virial radius surprisingly reproduces the thermal profile. A generalized Schwarzschild criterion indicates that observed spiral patterns must involve a convective layer, which may regulate the thermal profile.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2017

Ensembles of novelty detection classifiers for structural health monitoring using guided waves

Guided wave structural health monitoring uses sparse sensor networks embedded in sophisticated structures for defect detection and characterization. The biggest challenge of those sensor networks is developing robust techniques for reliable damage detection under changing environmental and operating conditions. To address this challenge, we develop a novelty classifier for damage detection based on one class support vector machines. We identify appropriate features for damage detection and introduce a feature aggregation method which quadratically increases the number of available training observations.We adopt a two-level voting scheme by using an ensemble of classifiers and predictions. Each classifier is trained on a different segment of the guided wave signal, and each classifier makes an ensemble of predictions based on a single observation. Using this approach, the classifier can be trained using a small number of baseline signals. We study the performance using monte-carlo simulations of an analytical model and data from impact damage experiments on a glass fiber composite plate.We also demonstrate the classifier performance using two types of baseline signals: fixed and rolling baseline training set. The former requires prior knowledge of baseline signals from all environmental and operating conditions, while the latter does not and leverages the fact that environmental and operating conditions vary slowly over time and can be modeled as a Gaussian process.

Dib, Gerges (ORCID:0000000255004047)↗
DOE OSTI2025

HydraGNN v4.0

The new version of HydraGNN v4.0 provides additional core capabilities, such as: Inclusion of multi-body atomistic cluster expansion MACE, polarizable atom interaction neural network PAINN, and equivariant principal neighborhood aggregation (PNAEq) among the message passing layers supported -Inclusion of graph transformers to directly model long-range interactions between nodes that are distant in the graph topology Integration of graph transformers with message passing layers by combining the graph embedding generated by the two mechanisms, which allows for an improved expressivity of the HydraGNN architecture Improved re-implementation of multi-task learning (MTL) to allow its use for stabilized training across imbalanced, multi-source, multi-fidelity data Introduction of multi-task parallelism, a newly proposed type of model parallelism specifically for MTL architectures, which allows to dispatch different output decoding heads to different GPU devices Integration of multi-task parallelism with pre-existing distributed data parallelism to enable a 2D parallelization for distributed training Improved portability of the distributed training across Intel GPUs, which has been testes on ALCF exascale supercomputer Aurora Inclusion of 2-level fine-grained energy profilers portable across NVIDIA, AMD, and Intel GPUs to monitor the power and energy consumption associated with different functions executed by the HydraGNN code during data pre-load and training Restructuring of previous examples and inclusion of new sets of examples to illustrate the download, preprocess, and training of HydraGNN models on new large-scale open-source datasets for atomistic materials modeling (e.g., Alexandria, Transition1x, OMat24, OMol25)

Lupo Pasini, Massimiliano [Oak Ridge National Labo↗
DOE OSTI2021

CrossCheck: Rapid, Reproducible, and Interpretable Model Evaluation

Evaluation beyond aggregate performance metrics, e.g. F1-score, is crucial to both establish an appropriate level of trust in machine learning models and identify future model improvements. In this paper we demonstrate CrossCheck, an interactive visualization tool for rapid crossmodel comparison and reproducible error analysis. We describe the tool and discuss design and implementation details. We then present three use cases (named entity recognition, reading comprehension, and clickbait detection) that show the benefits of using the tool for model evaluation. CrossCheck allows data scientists to make informed decisions to choose between multiple models, identify when the models are correct and for which examples, investigate whether the models are making the same mistakes as humans, evaluate models’ generalizability and highlight models’ limitations, strengths and weaknesses. Furthermore, CrossCheck is implemented as a Jupyter widget, which allows rapid and convenient integration into data scientists’ model development workflows.

Arendt, Dustin L.↗
DOE OSTI2017

Characterizing the Laboratory Market

Laboratories are estimated to be 3-5 times more energy intensive than typical office buildings and offer significant opportunities for energy use reductions. Although energy intensity varies widely, laboratories are generally energy intensive due to ventilation requirements, the research instruments used, and other health and safety concerns. Because the requirements of laboratory facilities differ so dramatically from those of other buildings, a clear need exists for an initiative exclusively targeting these facilities. The building stock of laboratories in the United States span different economic sectors, include governmental and academic institution, and are often defined differently by different groups. Information on laboratory buildings is often limited to a small subsection of the total building stock making aggregate estimates of the total U.S. laboratories and their energy use challenging. Previous estimates of U.S. laboratory space vary widely owing to differences in how laboratories are defined and categorized. A 2006 report on fume hoods provided an estimate of 150,000 laboratories populating the U.S. based in part on interviews of industry experts, however, a 2009 analysis of the 2003 Commercial Buildings Energy Consumption Survey (CBECS) generated an estimate of only 9,000 laboratory buildings. This report draws on multiple data sources that have been evaluated to construct an understanding of U.S. laboratories across different sizes and markets segments. This 2016 analysis is an update to draft reports released in October and December 2016.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗
DOE OSTI2023

Small Reactors in Microgrids: Technology Modeling and Selection (Net-Zero Microgrid Program Project Report)

This report demonstrates the capabilities of the net-zero microgrid (NZM) Xendee platform for modeling an SR module with electricity, heat extraction and thermal storage in microgrids configurations. The model effectively captures the most important technical and economic considerations for SR technology specific analysis: cost and operational characteristics of SR technology and financial costs and incentives. The model can analyze multiple scenarios to establish metrics for cost-competitive and zero-carbon microgrids connected to the grid or completely isolated. The model is fully integrated within the Xendee platform for modeling and analysis of clean energy microgrids with storage and generation from renewable energy sources. The model captures the capabilities, constraints, and nuances of SR by incorporating parameters related to plant economics, design efficiency and performance, plant operation and component and fuel lifespan. The cost and operational parameters modeled in the SR module are specific to the technology selected for integration in the microgrid. Cost parameters recognize advanced nuclear technology for modular production and installation based on economies of scale from factory manufacture and related commissioning, and cost reduction through technology maturation—first-of-a-kind (FOAK) and nth-of-a-Kind (NOAK). The cost parameters include installation, operations and maintenance (O&M), fuel refueling cycle, and reactor life. Installation cost reflects economies of scale due to unit sizing at scale and colocation. O&M economies of scale for both fixed- and variable-cost fuel life-cycle costs are incurred at every refueling interval, with separate front- and back-end fuel costs, as well as waste-handling and disposition costs. This report investigates key characteristics of different SR technologies suitable for microgrid applications, including design principles, sizing, coolant properties, temperature ratings, fuel structures, and life-cycle considerations. This also includes fuel technologies applicable to these SR systems, alongside strategies for nuclear-waste and spent-fuel management and approaches to address safety, security, and proliferation challenges. Four primary groups of SR technologies are examined: water-cooled, liquid-metal-cooled, high-temperature gas-cooled, and molten-salt-cooled systems. In this report, an initial guideline for technology selection is established, aligning the characteristics of the technologies with the requirements of microgrids. The selection of technology in a microgrid is influenced by various factors, including financial capacity, location and accessibility, demand type and characteristics, reliability and resilience requirements, area constraints, and the lifespan of the microgrid. The types of electrical and non-electrical applications within the microgrid also play a significant role in technology selection. The characteristics of SRs, such as their smaller size, modularity, transportability, long refueling interval, improved safety features, ability to operate in autonomous or semi-autonomous mode, and provision of high-grade heat, are particularly appealing for microgrids. Furthermore, a list of considerations for implementing SRs in microgrids is outlined. The SR model is created to be continuously improved with the acquisition of actual data on investment and operational costs, experience with supply chains, production at scale, and field deployments. In the near term, performance data on applications in microgrids will become available from lessons learned from laboratory tests, such as those planned for the Microreactor Applications Research Validation and Evaluation Project (MARVEL), led by Idaho National Laboratory (INL). The SR model incorporates scenario data and known SR design specifications, enabling technoeconomic analysis for SR deployment in microgrids. It specifically considers the distinctive attributes of SRs as generators in technoeconomic studies. SRs can be modeled and analyzed with generation from renewable-energy sources, energy storage, and flexible loads over a range of functionality and applications. This offers a comprehensive tool for feasibility studies, scenario development, and sensitivity analysis for “what-if” consideration of any range of assumptions about SRs in microgrids and other aggregations of distributed-energy resources, including virtual power plants.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗
DOE OSTI2017

Deep Learning for Unsupervised Insider Threat Detection in Structured Cybersecurity Data Streams

Situational awareness data is a prototypical example of structured heterogeneous streaming data with asynchronous timing that can quickly scale beyond the cognitive power of a human analyst. We present deep learning methods that can train in real time from unlabeled streams and provide probabilistic anomaly scores. We compare our novel deep and recurrent neural network models to three baseline systems using threat detection recall on the CERT Insider Threat Dataset v6.2. In our experiments, our deep neural network and our recurrent neural network systems outperformed all baselines, and we found that using aggregated activity count features yielded the best performance. Our best model produces on average an anomaly score of 95.53 percentile for the day, suggesting it can facilitate a great reduction in analyst workload. Our models are also interpretable, in that they break down the anomaly score into individual features of user behavior, which could further aid analysts in reviewing potential cases of insider threat.

Tuor, Aaron R.↗
DOE OSTI2015

At least 10% shorter C–H bonds in cryogenic protein crystal structures than in current AMBER forcefields

High resolution protein crystal structures resolved with X-ray diffraction data at cryogenic temperature are commonly used as experimental data to refine forcefields and evaluate protein folding simulations. However, it has been unclear hitherto whether the C–H bond lengths in cryogenic protein structures are significantly different from those defined in forcefields to affect protein folding simulations. This article reports the finding that the C–H bonds in high resolution cryogenic protein structures are 10–14% shorter than those defined in current AMBER forcefields, according to 3709 C–H bonds in the cryogenic protein structures with resolutions of 0.62–0.79 Å. Also, 20 all-atom, isothermal–isobaric, 0.5-μs molecular dynamics simulations showed that chignolin folded from a fully-extended backbone formation to the native β-hairpin conformation in the simulations using AMBER forcefield FF12SB at 300 K with an aggregated native state population including standard error of 10 ± 4%. However, the aggregated native state population with standard error reduced to 3 ± 2% in the same simulations except that C–H bonds were shortened by 10–14%. Furthermore, the aggregated native state populations with standard errors increased to 35 ± 3% and 26 ± 3% when using FF12MC, which is based on AMBER forcefield FF99, with and without the shortened C–H bonds, respectively. These results show that the 10–14% bond length differences can significantly affect protein folding simulations and suggest that re-parameterization of C–H bonds according to the cryogenic structures could improve the ability of a forcefield to fold proteins in molecular dynamics simulations. - Highlights: • Cryogenic crystal structures are commonly used in computational studies of proteins. • C–H bonds in the cryogenic structures are shorter than those defined in forcefields. • A survey of 3709 C–H bonds shows that the cryogenic bonds are 10–14% shorter. • The 10–14% differences affect molecular dynamics simulations of protein folding. • Shortening C–H bonds by 10–14% can improve protein folding simulations.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2025

Spatially explicit terrestrial carbon densities for calibrating the carbon cycle in human-Earth system Models

Soil and vegetation carbon stocks play a critical role in human-Earth system models. These stocks (denominated as densities in MgC/ha) affect variables such as land use change emissions and also influence land use change pathways under climate forcing scenarios where terrestrial carbon is assigned a carbon price. Here we present reharmonized soil and vegetation carbon densities both at the 5-arcmin resolution grid cell level and also aggregated to 235 water sheds for 4 land use types (Cropland, Grazed land, Urban land and unmanaged vegetation) and 15 unmanaged land cover types. Moreover, we use the distribution of carbon within and across pixels to define statistical "states" of carbon, once again differentiated by land type. These statistical states are used to define a range of possible carbon values that can be used for defining initial conditions of soil and vegetation carbon in human-Earth system models. We implement these data in a state-of-the-art multi sector dynamics model, namely the Global Change Analysis Model (GCAM), and show that these new data improve several land use responses, especially when terrestrial carbon is assigned a carbon price.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2015

The architecture of amyloid-like peptide fibrils revealed by X-ray scattering, diffraction and electron microscopy

The aggregation process and the fibril state of an amyloidogenic peptide suggest monomer addition to be the prevailing mechanism of elongation and a model of the peptide packing in the fibrils has been obtained. Structural analysis of protein fibrillation is inherently challenging. Given the crucial role of fibrils in amyloid diseases, method advancement is urgently needed. A hybrid modelling approach is presented enabling detailed analysis of a highly ordered and hierarchically organized fibril of the GNNQQNY peptide fragment of a yeast prion protein. Data from small-angle X-ray solution scattering, fibre diffraction and electron microscopy are combined with existing high-resolution X-ray crystallographic structures to investigate the fibrillation process and the hierarchical fibril structure of the peptide fragment. The elongation of these fibrils proceeds without the accumulation of any detectable amount of intermediate oligomeric species, as is otherwise reported for, for example, glucagon, insulin and α-synuclein. Ribbons constituted of linearly arranged protofilaments are formed. An additional hierarchical layer is generated via the pairing of ribbons during fibril maturation. Based on the complementary data, a quasi-atomic resolution model of the protofilament peptide arrangement is suggested. The peptide structure appears in a β-sheet arrangement reminiscent of the β-zipper structures evident from high-resolution crystal structures, with specific differences in the relative peptide orientation. The complexity of protein fibrillation and structure emphasizes the need to use multiple complementary methods.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
DOE OSTI2023

Building the I (Interoperability) of FAIR for performance reproducibility of large-scale composable workflows in RECUP

Abstract-Scientific computing communities increasingly run their experiments using complex data- and compute-intensive workflows that utilize distributed and heterogeneous architectures targeting numerical simulations and machine learning, often executed on the Department of Energy Leadership Computing Facilities (LCFs). We argue that a principled, systematic approach to implementing FAIR principles at scale, including fine-grained metadata extraction and organization, can help with the numerous challenges to performance reproducibility posed by such workflows. We extract workflow patterns, propose a set of tools to manage the entire life cycle of performance metadata, and aggregate them in an HPC-ready framework for reproducibility (RECUP). We describe the challenges in making these tools interoperable, preliminary work, and lessons learned from this experiment.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2019

Isothermal Microcalorimetry: Insight into the Impact of Crystallite Size and Agglomeration on the Lithiation of Magnetite, Fe 3 O 4

Magnetite, Fe 3 O 4 , holds significant interest as a Li-ion anode material because of its high theoretical capacity (926 mAh/g) associated with multiple electron transfers per cation center. Notably, both crystallite size and agglomeration influence ion transport. This report probes the effects of crystallite size (12 and 29 nm) and agglomeration on the reactions involved with the formation of the surface electrolyte interphase on Fe 3 O 4 . Isothermal microcalorimetry (IMC) was used here to determine the parasitic heat evolved during lithiation by considering the total heat measured, cell polarization, and entropic contributions. Interestingly, the 29 nm Fe 3 O 4 -based electrodes produced more parasitic heat than the 12 nm samples (1346 vs 1155 J/g). This observation was explored using scanning electron microscopy (SEM) and X-ray fluorescence (XRF) mapping in conjunction with spatially resolved X-ray absorption spectroscopy (XAS). SEM imaging of the electrodes revealed more agglomerates for the 12 nm material, affirmed by XRF maps. Further, XAS results suggest that Li + transport is more restricted for the smaller crystallite size (12 nm) material, attributed to its greater degree of agglomeration. These results rationalize the IMC data, where agglomerates of the 12 nm material limit solid electrolyte interphase formation and parasitic heat generation during lithiation of Fe 3 O 4 .

25 ENERGY STORAGE↗
DOE OSTI2021

Estimation of constituent properties of concrete materials with an artificial neural network based method

Multi-scale models are developed for heterogeneous concrete materials to estimate their macroscopic mechanical properties in terms of micro-structural data. One crucial challenge of those models is the identification of local properties of constituent phases. In this paper, we present an efficient method based on Artificial Neural Networks (ANN). Typical concrete materials are taken as example. A macroscopic analytical strength criterion is established from three steps of nonlinear homogenization procedure. The macroscopic strength of materials is determined as a function of the frictional coefficient and cohesion of solid cement particles at nanometer scale, intra-particle pores, inter-particle pores and aggregates (inclusions). The objective is to identify the nanoscopic frictional coefficient and cohesion of cement particle from measured macroscopic values of uniaxial compression and tensile strengths. For this purpose, a numerical method based on the ANN is developed. With the analytical macroscopic strength criterion, sensitivity studies are first realized to identify the most important micro-structural parameters influencing the macroscopic strength of concrete. A simplified analytical macroscopic strength criterion is then proposed. A large dataset is further constructed through the inversion of the analytical strength criterion by using the aggregates volume fraction, porosity, macroscopic uniaxial tensile and compressive strengths as input variables and the frictional coefficient and cohesion of cement particles as output unknowns. An ANN model containing four hidden layers and 100 neurons in each layer is constructed and trained by using this dataset. Various types of validation of the ANN model are performed. It is found that the proposed ANN based model can effectively predict the frictional coefficient and cohesion of porous cement paste at the microscopic scale with a very good accuracy.

36 MATERIALS SCIENCE↗