Public reproducibility package for the manuscript "Beyond direct combustion: a reversible benchmark for heat delivery from chemical fuels".
This repository contains the public model code, generated CSV outputs, rendered figure assets, and Aspen exported audit material for inspecting and reproducing the manuscript results.
python -m venv .venv
. .venv/bin/activate
python -m pip install --upgrade pip
python -m pip install -e .
./run.sh --dry-run
./run.sh --smoke-testThe smoke test checks that the public model layers import and that the expected generated CSVs and Aspen export artifacts are present and readable.
- Layer 2 reversible benchmark: theorem-derived upper-bound heat delivery from chemical fuels across source temperature, delivery temperature, and fuel identity.
- Layer 4 Brayton heat-pump architecture-space support: a bounded, semi-realistic R134a architecture-family model used to show how a specific implementation family occupies part of the reversible design space.
- Rendered manuscript and supplementary figures: public PNG, PDF, and SVG outputs under
figures/, with generation scripts undermodels/figures/. - Aspen worked-case audit package: exported CSV tables, reports, and flowsheet imagery for the methane Brayton heat-pump worked case.
| Path | Contents |
|---|---|
models/independent/ |
Layer 2 reversible benchmark model and generated outputs. |
models/thermodynamic/ |
Layer 4 thermodynamic support model outputs and Brayton architecture-space model. |
models/figures/ |
Figure scripts and composite-figure assembly utilities. |
data/ |
Public mirrored CSV outputs and data/manifest.csv. |
figures/ |
Rendered public figure assets. |
assets/aspen/ |
Aspen exported audit PDFs, flowsheet image, and comparison CSVs. |
tests/smoke/ |
Public package integrity tests. |
The public runner supports local-scale model reproduction, data refreshes, figure generation, and package checks.
./run.sh --model independent
./run.sh --model brayton
./run.sh --data
./run.sh --figures
./run.sh --smoke-test./run.sh with no arguments runs the independent model, the local-scale Brayton analyses, refreshes public data mirrors, and runs the smoke tests.
Use PYTHON=/path/to/python ./run.sh ... if you want the runner to use a specific interpreter.
The independent model computes the thermodynamic ceiling:
Q_H,max = |Delta G_reaction(T_C)| / eta_Carnot
eta_Carnot = 1 - T_C / T_H
Its main output is the dimensionless multiplier Q_H,max / |Delta H|. The model includes methane, syngas, carbon, hydrogen, and ammonia fuel-property definitions in models/independent/core.py.
Run it directly with:
python -c "from models.independent import run_model; run_model()"Generated files:
models/independent/output/benchmark_surface.csvmodels/independent/output/temperature_sweep.csvmodels/independent/output/fuel_comparison.csv
These files are mirrored in data/independent/.
The Brayton architecture-space model is a bounded implementation-family model that couples methane fuel input to mechanical work and an R134a heat-pump cycle, with sink temperature emerging from the configured heat-transfer closure. It provides architecture-space support alongside the reversible benchmark.
Key files:
models/thermodynamic/brayton_heat_pump_architecture_space/system_config.yamlmodels/thermodynamic/brayton_heat_pump_architecture_space/core/physics_engine.pymodels/thermodynamic/brayton_heat_pump_architecture_space/core/refrigerant_properties.pymodels/thermodynamic/brayton_heat_pump_architecture_space/analysis/
Run the local public analyses with:
./run.sh --model braytonGenerated files include:
multivariate_design_space.csvregime_validity_sweep.csvsensitivity_data.csvefficiency_distribution_high_res*.csvlayer2_layer4_comparison.csvstress_test_results.csv
These files are mirrored in data/brayton_heat_pump_architecture_space/.
data/manifest.csv maps public CSV files to their generated source paths and gives a short description of each dataset. The mirrored data/ tree is the easiest place for downstream inspection; the corresponding models/**/output/ folders are the generated source of record.
Rendered figures are stored in figures/. Figure scripts write to models/figures/output/; the public figures/ directory contains the released copies. To regenerate available public figures:
./run.sh --figuresSome figure scripts consume Aspen exported CSVs, so keep assets/aspen/export/ intact when regenerating figures.
The Aspen audit package provides exported material for inspecting the worked-case boundary and key comparison quantities.
Included Aspen artifacts:
assets/aspen/ModelsMethaneHeatPumpPureBrayton.pdfassets/aspen/ResultsMethaneHeatPumpPureBrayton_A2.pdfassets/aspen/OverallFlowsheet.PNGassets/aspen/export/current_base_case_key_outputs.csvassets/aspen/export/base_case_comparison_aspen_vs_python.csvassets/aspen/export/aspen_comparator_calculations.csvassets/aspen/export/aspen_hybrid_figure_inputs.csv
The public smoke test verifies that these exported audit files are present and non-empty.
- The reversible benchmark is a thermodynamic upper bound for heat delivery from chemical fuels.
- The Brayton architecture-space model is an implementation-family support model with explicit validity labels; use
statusandis_validfields when interpreting its outputs. - The high-resolution CSVs included in the package are manuscript-preparation outputs. The public runner regenerates local-scale analyses by default.
- Aspen support is provided through exported tables, reports, and flowsheet imagery.
This repository is archived on Zenodo, and the manuscript preprint is available on ResearchGate:
- Repository DOI: 10.5281/zenodo.20324897
- Preprint DOI: 10.13140/RG.2.2.15559.66720
- Source repository: https://github.com/Dal1308/BeyondDirectCombustionRepo
See CITATION.cff for citation metadata.
This package is distributed under the GNU General Public License v3.0. See LICENSE.