Selected Work, R&D and Engineering Case Studies

Technical question

Can a controlled compression–expansion trajectory activate useful low- and intermediate-temperature oxidation chemistry while avoiding subsequent high-temperature oxidation that consumes the desired intermediates?

Contribution

• Detailed chemical kinetics linked fuel molecular reactivity to the transient pressure–temperature history
• Compression ratio, cycle speed and equivalence ratio were treated as reactor-control variables
• Formation and survival of oxygenated hydrocarbons, H₂O₂, formaldehyde, alkenes and related intermediates were examined across the C₁–C₇ n-alkane series
• Useful intermediate formation was separated from the competing high-temperature ignition and deeper-oxidation pathway

Decision value

The case shows how molecular kinetics and a transient temperature–pressure trajectory can be translated into reactor operating-window decisions: which feedstock, which thermochemical trajectory, which reaction regime and which boundary must be avoided.

Evidence boundary

The evidence is model-based and predominantly single-zone. Spatial gradients, product extraction and recovery, separation, process safety, durability, cycle stability and industrial energy performance were not experimentally demonstrated. Reported model yields therefore remain case-specific modelling results rather than validated conversion, selectivity or process-performance data.

Technical question

Can fuel-reactivity modification reduce selected gaseous emissions without creating unacceptable penalties in power, nitrogen oxides, operability or downstream emission-control requirements?

Contribution

• Detailed chemical kinetics coupled with a homogeneous multi-zone engine representation
• A diesel surrogate of 77 vol% n-dodecane and 23 vol% m-xylene used as the baseline fuel
• Additive fraction, air–fuel ratio and engine speed examined as coupled operating variables
• Emission response interpreted together with the predicted power penalty and possible interaction with EGR, compression-ratio control and after-treatment

Decision value

The case shows why an additive strategy should be screened against several outputs at the same time rather than judged by one pollutant alone. It supports decisions about whether fuel chemistry, engine calibration, EGR or exhaust after-treatment should carry the next development burden.

Evidence boundary

The evidence is numerical and experimentally unvalidated for the proposed additive strategy. The exact additive formulation is not sufficiently documented in the controlled public evidence for reproducibility. No measured tailpipe emissions, particulate mass or number, durability, fuel-standard compliance, material compatibility or after-treatment performance were demonstrated. Reported reductions therefore remain case-specific modelling results.

Technical question

How can a large detailed kinetic mechanism be made smaller and faster without removing reactions that exert strong control over the autoignition quantities that must remain accurate?

Contribution

• Reaction-rate analysis identifies reactions carrying significant kinetic activity along the autoignition trajectory
• Sensitivity analysis retains slower reactions that still exert strong kinetic control
• The two criteria are combined so that slow but rate-limiting pathways are not discarded simply because their instantaneous rates are small
• Reduction thresholds are treated as mechanism- and application-dependent choices that require validation against the parent detailed chemistry

Decision value

The case translates mechanism reduction into an engineering decision sequence: define the observables that must be preserved, identify kinetically indispensable reactions, reduce only within that target domain, quantify the computational benefit and revalidate before transfer to a new fuel, reactor or multidimensional simulation.

Evidence boundary

The published method was evaluated for selected n-heptane and isooctane autoignition applications using CHEMKIN-era detailed mechanisms. Reported computational speedups are historical and should not be transferred to current hardware or software. Accuracy is not established automatically for new fuels, flames, CFD, emissions targets or operating domains. Later reaction-network and adaptive-chemistry work is related but is scientifically distinct from the original reduction method and does not retroactively expand its validation domain.

Technical question

Can the changing topology of a combustion reaction network identify locally active species during a simulation, and can that information support adaptive chemistry without confusing structural connectivity with predictive kinetic importance?

Contribution

• Combustion chemistry was represented as a dynamic reaction network whose species connectivity changes with the local thermochemical state
• Degree centrality was used to identify a principal species at each selected simulation time or computational location
• The dynamically selected principal species was coupled with a directed relation graph reduction step inside an adjusted dynamic adaptive chemistry workflow
• Tests with C1–C3 alkane chemistry and n-heptane showed that centrality ranking alone should not be treated as a direct substitute for predictive mechanism-reduction criteria

Decision value

The case separates two questions that are often mixed: Which species is structurally active in the local reaction network? and Which species must be retained to preserve a target prediction? This distinction supports more disciplined use of network metrics for mechanism interpretation, target selection and adaptive-chemistry design.

Evidence boundary

Degree centrality is a structural network descriptor; a large value does not by itself prove causal control of ignition, heat release, flame speed or pollutant formation. The published implementation used Cantera, GRI-Mech 3.0 for C1–C3 chemistry, an LLNL n-heptane mechanism, DRG reduction and an adjusted DAC workflow. Useful real-CPU-time comparison was not established in the paper. Transfer to other mechanisms, targets, reactors or multidimensional CFD therefore requires separate validation.

Technical question

Can an adaptive chemistry workflow reduce the active mechanism as conditions evolve, preserve the declared accuracy targets, transfer across parent mechanisms and still deliver a useful end-to-end computational benefit?

Contribution

• Reduction accuracy, mechanism transfer and runtime performance were evaluated as separate evidence dimensions
• GRI-Mech 3.0 methane/air cases were used for the controlled accuracy baseline
• USC Mech II provided a second-mechanism transfer test using a mechanism-specific candidate domain rather than assuming the first mechanism's domain would transfer unchanged
• Runtime attribution separated chemistry-integration benefit from controller, rebuild and mechanism-management overhead

Decision value

The case demonstrates why reduced chemistry should not be judged by mechanism size or local solver speed alone. A useful adaptive implementation must preserve the required prediction, transfer only within a justified domain, control state and rebuild integrity, and reduce total wall-clock cost after all adaptation overhead is included.

Evidence boundary

This is computational validation within the declared benchmark cases, not experimental validation of combustion chemistry or a universal adaptive-chemistry result. Static reduction of the tested USC cases produced a computational speed benefit on the reported computer, but the complete adaptive Python/Cantera prototype was slower because management overhead dominated. The results do not establish universal scheduler success, universal mechanism transfer, CFD acceleration or hardware- independent speedup. New fuels, reactors and multidimensional applications require separate validation and performance engineering.

Technical question

Which physical, chemical and operational relationships must be preserved, re-measured or deliberately allowed to change when laboratory or pilot results are transferred toward an industrial process?

Contribution

• Governing chemistry, transport, thermodynamics and hydrodynamics are identified before choosing a scaling rule
• Similarity and dimensionless analysis are used selectively rather than assuming that every criterion can be preserved at the same time
• Fit-for-purpose modelling and representative experiments are linked to explicit questions that must be resolved before the next development stage
• Pilot evidence is connected to operability, materials, safety, process integration, economics and environmental constraints before industrial progression

Decision value

The case provides a practical sequence for deciding what to test next, which scale-dependent risks require evidence, when a more detailed model is justified, and when development should proceed, be modified, remain on hold or stop. It helps prevent attractive laboratory results from being treated as industrial proof before the controlling phenomena and complete process consequences are understood.

Evidence boundary

This is an engineering decision framework supported by open technical reviews and cross-domain process-development experience. It is not a universal scale-up equation, a validated design method for every technology, a substitute for representative pilot data, or an industrial performance guarantee. Each application still requires domain-specific modelling, experiments, hazard and regulatory assessment, techno-economic analysis and validation against its own operating window.

Technical question

Which mechanism actually controls process performance, and does a proposed intensification method address that bottleneck?

Contribution

• Reaction–transport regime classification
• Bottleneck-oriented interpretation
• Energy-normalized intensification perspective
• Validation and scale-up questions
• Executable benchmark framework for engineering screening

Decision value

Supports experimental design, process comparison, claims assessment and selection of chemical, transport, separation or equipment interventions.

Evidence boundary

The framework is diagnostic and screening-oriented. Illustrative calculations and dimensionless descriptors do not constitute independent process validation. It is not a universal performance predictor and does not replace representative testing or site-specific engineering.

Technical question

How strongly do background water constituents compete for radicals, and what does that imply for apparent pollutant removal?

Contribution

• Matrix-scavenging calculations
• Radical-utilization fractions
• Apparent kinetic and treatment-time estimates
• Carbonate and alkalinity utilities
• Automated software tests

Decision value

Supports matrix characterization, scenario comparison, interpretation of apparent AOP performance and preparation of laboratory or pilot validation programmes.

Evidence boundary

Software tests verify implemented calculations, not predictive validity for a real matrix, contaminant mixture, oxidant, catalyst, irradiation system or reactor.

Technical question

Which qualitative trends in archived simulations remain useful, and what must be resolved before the model can support design or scale-up?

Contribution

• Cantera-compatible mechanism and examples
• Curated moisture, temperature and atmosphere cases
• Explicit temperature-unit warning
• Source-model reconciliation and validation plan
• Responsible carbon-management boundaries

Decision value

Demonstrates how legacy scientific material can become a transparent, auditable development foundation before model complexity or industrial claims are increased.

Evidence boundary

The repository is a research reconstruction and screening foundation. Exact source-model reproduction remains unresolved, and the repository is not a validated feedstock-, particle- or reactor-specific industrial design model.

Scientific question

How might reversible precursor association progress toward the first persistent particles without assuming a universal inception mechanism?

Contribution

• Explicit association and dissociation terms
• Non-stabilizing loss pathways
• Bounded stabilization probability
• Executable zero-dimensional demonstration
• Testable validation questions

Scientific value

Provides a structured basis for formulating and testing persistence and stabilization hypotheses while keeping the conceptual model separate from established nanoparticle- inception mechanisms.

Evidence boundary

The illustrative parameters are not fitted to a specific experiment. Software execution demonstrates implementation integrity, not physical validation. The model does not yet predict particle number, size distribution, soot volume fraction or industrial performance.