Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

ChemZIP: Accelerated Modeling of Complex Aerothermochemical Interactions in Novel Turbomachines for Sustainable High-Temperature Chemical Processes

Dylan Rubini, Budimir Rosic

TL;DR

This paper tackles the computational bottleneck of modeling aerochemical interactions in high-power turbomachines by introducing ChemZIP, a multifidelity ML surrogate that compresses high-dimensional kinetic space into a small latent manifold and learns fast source-term mappings. By coupling a linear chemistry encoder with a nonlinear decoder and a latent-space dynamics model, ChemZIP dramatically accelerates reacting-flow simulations in 3D CFD while achieving high accuracy (R^2 > 0.95 in 1D tests and within 10% of Fluent in 3D tests). Verification across 1-D and 3-D scenarios demonstrates substantial speedups (up to ~580× versus direct integration) and robust predictive capability for propane pyrolysis chemistry under turbomachinery-like heat loads. The platform enables aerochemically guided design optimization of turbo-reactors, with training data generation completed in hours and substantial potential for larger, more detailed mechanisms in industrial design cycles.

Abstract

This paper introduces a new platform to accelerate the modeling of complex aerothermochemical interactions in new turbomachines, turbo-reactors, to decarbonise chemical processes. While previous work has aerothermally demonstrated the potential to decarbonize the heat input to the reaction, optimizing the reaction efficiency has been a challenge. This is because measuring reaction performance with aerochemical simulations is computationally prohibitive due to the uniquely complex aerodynamics and chemistry within turbomachines. To address this, we introduce a new multifidelity machine-learning-assisted methodology, called ChemZIP, to mitigate this bottleneck. Although data-driven methodologies exist for combustion, modeling reactive flows along the bladed path of a turbomachine poses new challenges. This has led to a novel training data generation process, which allows rich dynamic responses of the chemical system to be embedded into the training dataset at a fraction of the cost of reacting flow simulations. The resulting high-dimensional composition vector is compressed into a low-dimensional basis using an autoencoder-like neural network, inspired by but more universal than traditional flamelet-generated manifolds. Verification against 10,000 unseen one-dimensional test conditions shows an R2 score exceeding 95% across all quantities of interest. Following this, ChemZIP is coupled into a fully-fledged viscous computational fluid dynamics solver. For a set of process-relevant three-dimensional configurations entirely different from the training data, the predictive accuracy of the thermochemical state remains within 10% of an industry-standard solver while convergence is achieved 50 times faster, even for a small mechanism. Therefore, numerical computations are sufficiently fast that aerothermochemical optimization is now feasible for the first time in the design cycle

ChemZIP: Accelerated Modeling of Complex Aerothermochemical Interactions in Novel Turbomachines for Sustainable High-Temperature Chemical Processes

TL;DR

This paper tackles the computational bottleneck of modeling aerochemical interactions in high-power turbomachines by introducing ChemZIP, a multifidelity ML surrogate that compresses high-dimensional kinetic space into a small latent manifold and learns fast source-term mappings. By coupling a linear chemistry encoder with a nonlinear decoder and a latent-space dynamics model, ChemZIP dramatically accelerates reacting-flow simulations in 3D CFD while achieving high accuracy (R^2 > 0.95 in 1D tests and within 10% of Fluent in 3D tests). Verification across 1-D and 3-D scenarios demonstrates substantial speedups (up to ~580× versus direct integration) and robust predictive capability for propane pyrolysis chemistry under turbomachinery-like heat loads. The platform enables aerochemically guided design optimization of turbo-reactors, with training data generation completed in hours and substantial potential for larger, more detailed mechanisms in industrial design cycles.

Abstract

This paper introduces a new platform to accelerate the modeling of complex aerothermochemical interactions in new turbomachines, turbo-reactors, to decarbonise chemical processes. While previous work has aerothermally demonstrated the potential to decarbonize the heat input to the reaction, optimizing the reaction efficiency has been a challenge. This is because measuring reaction performance with aerochemical simulations is computationally prohibitive due to the uniquely complex aerodynamics and chemistry within turbomachines. To address this, we introduce a new multifidelity machine-learning-assisted methodology, called ChemZIP, to mitigate this bottleneck. Although data-driven methodologies exist for combustion, modeling reactive flows along the bladed path of a turbomachine poses new challenges. This has led to a novel training data generation process, which allows rich dynamic responses of the chemical system to be embedded into the training dataset at a fraction of the cost of reacting flow simulations. The resulting high-dimensional composition vector is compressed into a low-dimensional basis using an autoencoder-like neural network, inspired by but more universal than traditional flamelet-generated manifolds. Verification against 10,000 unseen one-dimensional test conditions shows an R2 score exceeding 95% across all quantities of interest. Following this, ChemZIP is coupled into a fully-fledged viscous computational fluid dynamics solver. For a set of process-relevant three-dimensional configurations entirely different from the training data, the predictive accuracy of the thermochemical state remains within 10% of an industry-standard solver while convergence is achieved 50 times faster, even for a small mechanism. Therefore, numerical computations are sufficiently fast that aerothermochemical optimization is now feasible for the first time in the design cycle

Paper Structure

This paper contains 34 sections, 16 equations, 21 figures, 4 tables.

Table of Contents

  1. Introduction
  2. Motivation and Background
  3. Contributions in this Paper
  4. The Ultra-High Power Density Turbo-Reactor Concept
  5. Modeling Aerochemical Interactions
  6. Challenges
  7. Limitations of Existing Accelerations Approaches
  8. Physics-Driven
  9. Data-driven Acceleration Approaches
  10. Why Machine Learning?
  11. Existing Implementations
  12. ChemZIP: An Aerochemical Modeling Methodology
  13. Training Data Generation
  14. Sample Space
  15. One-Dimensional Plug Flow Reactor Model
  16. ...and 19 more sections

Figures (21)

  • Figure 1: Electrified turbo-reactor could decarbonise many hard-to-abate industrial processes.
  • Figure 2: High-fidelity numerical simulation showing the instantaneous temperature gradients present in the flow.
  • Figure 3: Schematic illustration of the reaction response to representative temperature gradients in the turbo-reactor compared to those in a tubular pipe of a conventional furnace.
  • Figure 4: Schematic illustration of a low-dimensional manifold in composition phase space.
  • Figure 5: Schematic of the training data generation workflow for the ChemZIP methodology.
  • ...and 16 more figures