About

Tarek Echekki is an Associate Professor and the Associate Department Head in the Department of Mechanical and Aerospace Engineering at NC State University. He also serves as the Director of Undergraduate Programs within the department. His teaching portfolio includes graduate courses such as Fluid Dynamics of Combustion I and II, Foundations of Fluid Dynamics, and Turbulence, as well as undergraduate courses like Engineering Thermodynamics I and II and Fluid Mechanics I. His research focuses on combustion, emphasizing its importance in solving contemporary engineering problems by integrating thermodynamics, heat transfer, and fluid mechanics. Dr. Echekki has contributed to advancing understanding in turbulent combustion, chemical kinetics, and combustion chemistry acceleration, with numerous publications in these areas.

Research topics

• Computer Science
• Artificial Intelligence
• Physics
• Chemistry
• Algorithm
• Mechanics
• Statistical physics
• Engineering
• Mechanical engineering
• Organic chemistry

Selected publications

• Ultra-stretchable superomniphobic surfaces via machine-learning-guided laser ablation

  Matter · 2026-02-16

  article
  PublisherDOI

• React-NIF: A neural implicit flow-based framework for complex fuel combustion chemistry acceleration

  Fuel · 2025-12-29

  articleOpen accessSenior authorCorresponding

  In combustion simulations, chemistry integration presents a significant bottleneck, often limiting their overall performance. In this work, we propose a novel framework for accelerating the integration of complex fuels’ combustion chemistry based on an extension of the neural implicit flow (NIF) architecture, called React-NIF. As in NIF, React-NIF features two coupled subnetworks: a ShapeNet and a ParameterNet. Our present framework is designed to map the solution of a reduced set of representative scalars of the thermochemical space to the same solution at a later time increment. The network allows for a range of time steps to be selected during this mapping. Two key extensions are implemented to construct robust predictions of the solution vector. The first extension replaces the physical time input to the ParameterNet with a subset of the representative scalars to capture the dynamics of the oxidation process. The second extension involves the implementation of a ShiftNet, which aligns self-similar solution profiles characteristic of complex fuel oxidation. React-NIF is validated on hydrogen high-temperature oxidation and n-dodecane low-temperature oxidation in 0D. The results demonstrate that the proposed React-NIF achieves accurate predictions of thermochemical scalar profiles in 0D simulations while using a significantly smaller architecture than React-DeepONet (one-eleventh of the parameters of React-DeepONet for n-dodecane). In comparison with React-DeepONet, React-NIF also demonstrates significantly reduced computational training requirements. The model achieves a significant speed-up in direct chemistry integration, a moderate speed-up compared to React-DeepONet, and demonstrates strong extrapolation capabilities. • We propose a neural implicit flow (NIF) framework for accelerating combustion chemistry. • The React-NIF maps a solution vector of representative thermochemical scalars across an adaptive time step. • The framework was validated for H 2 and n-dodecane oxidation, yielding excellent predictions of species profiles in 0D. • The framework provides a more compact architecture than React-DeepONet and enables faster training.

  PublisherDOI

• Gravity Effects on Backdraft Phenomena in an Enclosure with Varying Opening Geometries

  Microgravity Science and Technology · 2025-07-30

  articleOpen accessSenior author

  Abstract In this study, we investigated the backdraft phenomenon numerically under different gravity conditions and 4 openings using the Fire Dynamics Simulator (FDS) code. Four different opening geometries are studied under ten different gravity conditions. The rate at which oxygen reaches an ignition block in the enclosure in the presence of gravity currents plays an important role in the onset of ignition, the subsequent backdraft formation, and the maximum pressure built inside the enclosure before the onset of backdraft. This role also explains why these effects are different under different openings. We observe that the gravity strength affects the ignition time and the onset of backdraft non-linearly. Moreover, it is found that the smoke exiting the enclosure cannot be considered a reliable precursor for the onset of backdraft to allow people on the outside to undertake necessary precautions. The effect of backdraft in the form of heat flow and impact force at the exit is also studied. It is found that the effect of heat flow is more severe than that of the impact force.

  PublisherOA PDFDOI

• Physics-constrained machine learning for reduced composition space chemical kinetics

  Data-Centric Engineering · 2025-01-01 · 1 citations

  articleOpen accessSenior authorCorresponding

  Abstract Modeling detailed chemical kinetics is a primary challenge in combustion simulations. We present a novel framework to enforce physical constraints, specifically total mass and elemental conservation, during the reaction of ML models’ training for the reduced composition space chemical kinetics of large chemical mechanisms in combustion. In these models, the transport equations for a subset of representative species are solved with the ML approaches, while the remaining nonrepresentative species are “recovered” with a separate artificial neural network trained on data. Given the strong correlation between full and reduced solution vectors, our method utilizes a small neural network to establish an accurate and physically consistent mapping. By leveraging this mapping, we enforce physical constraints in the training process of the ML model for reduced composition space chemical kinetics. The framework is demonstrated here for methane, CH 4 , and oxidation. The resulting solution vectors from our deep operator networks (DeepONet)-based approach are accurate and align more consistently with physical laws.

  PublisherOA PDFDOI

• A PINN-DeepONet framework for extracting turbulent combustion closure from multiscalar measurements

  Computer Methods in Applied Mechanics and Engineering · 2024-06-28 · 21 citations

  articleSenior authorCorresponding
  PublisherDOI

• A Data-Based Hybrid Chemistry Acceleration Framework for the Low-Temperature Oxidation of Complex Fuels

  Energies · 2024-02-04 · 2 citations

  articleOpen accessSenior authorCorresponding

  The oxidation of complex hydrocarbons is a computationally expensive process involving detailed mechanisms with hundreds of chemical species and thousands of reactions. For low-temperature oxidation, an accurate account of the fuel-specific species is required to correctly describe the pyrolysis stage of oxidation. In this study, we develop a hybrid chemistry framework to model and accelerate the low-temperature oxidation of complex hydrocarbon fuels. The framework is based on a selection of representative species that capture the different stages of ignition, heat release, and final products. These species are selected using a two-step principal component analysis of the reaction rates of simulation data. Artificial neural networks (ANNs) are used to model the source terms of the representative species during the pyrolysis stage up to the transition time. This ANN-based model is coupled with C0–C4 foundational chemistry, which is used to model the remaining species up to the transition time and all species beyond the transition time. Coupled with the USC II mechanism as foundational chemistry, this framework is demonstrated using simple reactor homogeneous chemistry and perfectly stirred reactor (PSR) calculations for n-heptane oxidation over a range of composition and thermodynamic conditions. The hybrid chemistry framework accurately captures correct physical behavior and reproduces the results obtained using detailed chemistry at a fraction of the computational cost.

  PublisherOA PDFDOI

• Modeling and Simulation of Turbulent Flame Kernel Evolution

  2024-09-06 · 6 citations

  book-chapter1st authorCorresponding

  The evolution of a premixed flame kernel in a turbulent flow field is investigated. Models for the the flame surface area and the flame surface density (fsd) are proposed. These models are compared to results from Direct Numerical Simulations (DNS) of the evolution of a cylindrical flame kernel in two-dimensional turbulence. The evolution of the flame surface area is modeled as a superpositon of flame surface wrinkling and propagation. An algebraic model of the fsd based on the evolution of an initially laminar flame in a turbulent flow is used to predict the decay of the peak value of the fsd. The DNS results show that the profiles of the fsd in the progress variable space are self-similar with a shift of the peak towards the burnt gases. The peak value of the fsd agrees very well with the model. The contribution of the various terms to the fsd-transport equation is also investigated. The results show that the dominant contribution to the transport of fsd is from turbulent strain and curvature.

  PublisherDOI

• Transfer learning for predicting source terms of principal component transport in chemically reactive flow

  Data-Centric Engineering · 2024-01-01 · 1 citations

  articleOpen access

  Abstract Transfer learning has been highlighted as a promising framework to increase the accuracy of the data-driven model in the case of data sparsity, specifically by leveraging pretrained knowledge to the training of the target model. The objective of this study is to evaluate whether the number of requisite training samples can be reduced with the use of various transfer learning models for predicting, for example, the chemical source terms of the data-driven reduced-order modeling (ROM) that represents the homogeneous ignition of a hydrogen/air mixture. Principal component analysis is applied to reduce the dimensionality of the hydrogen/air mixture in composition space. Artificial neural networks (ANNs) are used to regress the reaction rates of principal components, and subsequently, a system of ordinary differential equations is solved. As the number of training samples decreases in the target task, the ROM fails to predict the ignition evolution of a hydrogen/air mixture. Three transfer learning strategies are then applied to the training of the ANN model with a sparse dataset. The performance of the ROM with a sparse dataset is remarkably enhanced if the training of the ANN model is restricted by a regularization term that controls the degree of knowledge transfer from source to target tasks. To this end, a novel transfer learning method is introduced, Parameter control via Partial Initialization and Regularization (PaPIR), whereby the amount of knowledge transferred is systemically adjusted in terms of the initialization and regularization schemes of the ANN model in the target task.

  PublisherOA PDFDOI

• Numerical investigation of gravity effects on backdraft phenomena in an enclosure with varying geometries

  arXiv (Cornell University) · 2024-07-29

  preprintOpen accessSenior author

  In this study, we investigated numerically the backdraft phenomenon under different gravity conditions and 4 openings using Fire Dynamics Simulator (FDS) code. Four different opening geometries are studied under ten different gravity conditions. As demonstrated by an earlier study in our group, backdraft is established under different gravity conditions from 0.1 g to 1 g. The rate at which oxygen reaches an ignition block in the enclosure in the presence of the gravity current plays an important role in the onset of ignition, the subsequent backdraft formation, and the maximum pressure built inside the enclosure before the onset of backdraft. This role also explains why these effects are different under different openings. We observe that the gravity strength affects the ignition time and the onset of backdraft nonlinearly. Moreover, the smoke exiting the enclosure is a precursor for the onset of backdraft, albeit with a short warning, allowing people on the outside to take necessary precautions. The effect of backdraft in the form of heat flow and impact force at the exit is also studied. It is found that the effect of heat flow is more severe than that of impact force.

  PublisherOA PDFDOI

• Comparisons of Different Representative Species Selection Schemes for Reduced-Order Modeling and Chemistry Acceleration of Complex Hydrocarbon Fuels

  Energies · 2024-05-28 · 4 citations

  articleOpen accessSenior authorCorresponding

  The simulation of engine combustion processes, such as autoignition, an important process in the co-optimization of fuel-engine design, can be computationally expensive due to the large number of thermo-chemical scalars needed to describe the full chemical system. Yet, the inherent correlations between the different chemical species during oxidation can significantly reduce the complexity of representing this system. One strategy is to select a subset of representative species that accurately captures the combustion process at a fraction of the computational cost of the full system. In this study, we compare the performance of four different techniques to select these species. They include the two-step principal component analysis (PCA) approach, directed relation graphs (DRGs), the global pathway selection (GPS) approach, and the manifold-informed species selection method. A parametric study of the representative species selection is carried out on data from the simulation of homogeneous and perfectly stirred reactors by investigating seven cumulative variances and 47 different cut-off percentages for the two-step PCA, and 65 and 51 thresholds for the DRGs and GPS, respectively. Results show that these selection methods capture key important species that can accurately describe the chemical system and track each stage of oxidation. The two-step PCA is sensitive to the cumulative variance, and DRGs and GPS are sensitive to the choice of target variables. By selecting key representative species and reducing the number of thermo-chemical scalars, these three methods can be used to develop computationally efficient hybrid chemistry schemes.

  PublisherOA PDFDOI

Recent grants

• Multiscale Turbulent Reacting Flows and Data-Based Modeling

  NSF · $200k · 2012–2016

• Computational Methods for Multiscale Turbulent Reacting Flows

  NSF · $217k · 2009–2013

• EAGER: An Experiment-Based Framework for Turbulent Combustion Modeling

  NSF · $100k · 2019–2022

• Computational and Experimental Studies of Turbulent Premixed Flame Kernels

  NSF · $30k · 2008–2013

Frequent coauthors

• Rishikesh Ranade

  15 shared

• Jacqueline H. Chen

  15 shared

• Hessam Mirgolbabaei

  University of Minnesota, Duluth

  12 shared

• Sultan Alqahtani

  King Khalid University

  7 shared

• Kevin M. Gitushi

  North Carolina State University

  7 shared

• Bhargav Ranganath

  Maharaj Vijayaram Gajapathi Raj College of Engineering

  6 shared

• Mohammad Affan Khalil

  6 shared

• Opeoluwa Owoyele

  Argonne National Laboratory

  6 shared

Education

• PhD, Mechanical Engineering

  Stanford University

  1993

• MS, Mechanical Engineering

  Stanford University

  1987

• BS, Mechanical Engineering

  Washington University in Saint Louis

  1985