221 lines
6.4 KiB
TeX
221 lines
6.4 KiB
TeX
\documentclass[10pt,a4paper]{article}
|
|
\usepackage[margin=18mm]{geometry}
|
|
\usepackage{amsmath}
|
|
\usepackage{siunitx}
|
|
\usepackage{chemfig}
|
|
\usepackage{mhchem}
|
|
|
|
% -------- Conditional switches --------
|
|
\newif\ifresearch
|
|
\researchtrue % set to \researchfalse to hide research section
|
|
|
|
\title{\vspace{-1.0cm}Optical Combustion Diagnostics Using UV/Visible Chemiluminescence}
|
|
\author{}
|
|
\date{}
|
|
|
|
\begin{document}
|
|
\maketitle
|
|
\vspace{-0.8cm}
|
|
|
|
\section*{Purpose}
|
|
|
|
This note proposes a low-cost optical route for combustion diagnostics using flame
|
|
chemiluminescence. The approach is not intended to replace existing oxygen or CO
|
|
measurement systems. Instead, it provides an additional diagnostic layer for flame
|
|
quality, instability, and air/fuel condition.
|
|
|
|
Initial tests have already shown that a transimpedance amplifier (TIA) front-end,
|
|
connected to UV-sensitive photodiodes, produces clear measurable signals from flames.
|
|
This supports the feasibility of extending the existing IR flame-detection PCB concept
|
|
towards UV and chemiluminescence sensing.
|
|
|
|
\section*{Relevant flame chemistry}
|
|
|
|
Hydrocarbon flames produce excited radical species during combustion. Two particularly
|
|
useful emitters are hydroxyl and methylidyne radicals:
|
|
|
|
\[
|
|
\ce{OH^{*} -> OH + h\nu}
|
|
\]
|
|
|
|
\[
|
|
\ce{CH^{*} -> CH + h\nu}
|
|
\]
|
|
|
|
The dominant useful bands are approximately:
|
|
|
|
\[
|
|
\ce{OH^{*}} \approx \SI{310}{nm}
|
|
\]
|
|
|
|
\[
|
|
\ce{CH^{*}} \approx \SI{430}{nm}
|
|
\]
|
|
|
|
The relative intensity of these bands is linked to combustion state. In particular,
|
|
the ratio
|
|
|
|
\[
|
|
R = \frac{I_{\ce{OH^{*}}}}{I_{\ce{CH^{*}}}}
|
|
\]
|
|
|
|
may provide an indication of rich/lean tendency, while time variation in the
|
|
UV signal may provide information about flame instability, lift-off, pulsation,
|
|
or incipient poor combustion.
|
|
|
|
\section*{Sensing concept}
|
|
|
|
The proposed sensing chain is:
|
|
|
|
\[
|
|
\text{Flame emission}
|
|
\rightarrow
|
|
\text{UV/visible optical filter}
|
|
\rightarrow
|
|
\text{photodiode}
|
|
\rightarrow
|
|
\text{TIA}
|
|
\rightarrow
|
|
\text{gain/filtering}
|
|
\rightarrow
|
|
\text{ADC/DSP}
|
|
\]
|
|
|
|
A simple photodiode TIA has already been tested with flame sources in the UV region.
|
|
The observed response confirms that the required optical signal is detectable using
|
|
low-cost analogue electronics. This gives a practical development path:
|
|
|
|
\[
|
|
\text{IR flame PCB}
|
|
\rightarrow
|
|
\text{UV flame PCB}
|
|
\rightarrow
|
|
\text{dual-channel OH*/CH* diagnostic sensor}
|
|
\]
|
|
|
|
This is an incremental development, not a new platform from scratch.
|
|
|
|
\clearpage
|
|
|
|
\section*{Diagnostic value}
|
|
|
|
The proposed sensor could provide:
|
|
|
|
\begin{itemize}
|
|
\item flame presence detection;
|
|
\item flame stability / flicker analysis;
|
|
\item rich/lean indication from the $OH^*/CH^*$ ratio;
|
|
\item early warning of poor combustion before conventional limits are reached;
|
|
\item possible correlation with NOx-forming conditions;
|
|
\item service and commissioning diagnostics;
|
|
\item with the addition of Swan bands ($C_2$), early indication of soot-forming conditions;
|
|
\item ratios and strengths of $OH^*$, $CH^*$ and $C_2$ may provide insight into the instantaneous composition of waste or syngas fuels.
|
|
\end{itemize}
|
|
|
|
It should be stressed that this sensor does not directly measure CO or oxygen
|
|
concentration. Its value is as a complementary combustion-quality sensor, especially
|
|
where fast optical response provides information that slower gas probes may not.
|
|
|
|
\section*{Additional advantages}
|
|
|
|
\begin{itemize}
|
|
|
|
\item \textbf{Fast response:} The optical signal is generated directly at the reaction zone and is not subject to transport delay.
|
|
|
|
\item \textbf{Non-intrusive measurement:} No insertion into the flame or flue is required.
|
|
|
|
\item \textbf{Harsh environment suitability:} Optical sensing may be more robust than conventional probes in high-temperature or contaminated conditions.
|
|
|
|
\item \textbf{Early fault detection:} Chemiluminescence changes may precede measurable CO or O$_2$ changes.
|
|
|
|
\item \textbf{Dynamic information:} Temporal behaviour (oscillation, intermittency) becomes observable.
|
|
|
|
\item \textbf{Independent channel:} Provides plausibility checking against existing sensors.
|
|
|
|
\item \textbf{Low-cost replication:} Additional channels can be added at low cost.
|
|
|
|
\item \textbf{Variable fuel suitability:} Particularly relevant for syngas and mixed fuels.
|
|
|
|
\item \textbf{Degradation monitoring:} Optical fouling may be inferred from signal changes.
|
|
|
|
\end{itemize}
|
|
|
|
\clearpage
|
|
|
|
\section*{Commissioning support}
|
|
|
|
In industrial burner systems, commissioning involves stepping through firing rates
|
|
and storing actuator positions (fuel valve, fan VSD). The aim is stable operation
|
|
across the full range.
|
|
|
|
Currently this relies on:
|
|
\begin{itemize}
|
|
\item visual flame observation;
|
|
\item flue gas measurements (O$_2$, CO);
|
|
\item operator judgement.
|
|
\end{itemize}
|
|
|
|
These do not directly observe the reaction zone.
|
|
|
|
The proposed sensor could:
|
|
|
|
\begin{itemize}
|
|
\item provide real-time flame quality at each firing point;
|
|
\item identify marginal or poorly mixed conditions;
|
|
\item optimise fuel/air settings based on flame behaviour;
|
|
\item improve repeatability;
|
|
\item enable semi-automated commissioning using metrics such as $OH^*/CH^*$ and signal stability.
|
|
\end{itemize}
|
|
|
|
This extends the concept from monitoring to active commissioning support.
|
|
|
|
% -------- Research section (conditional) --------
|
|
\ifresearch
|
|
|
|
\clearpage
|
|
\section*{Research questions}
|
|
|
|
\begin{itemize}
|
|
|
|
\item \textbf{Information content:}
|
|
What combustion state information is recoverable from multi-band chemiluminescence?
|
|
|
|
\item \textbf{Separability:}
|
|
Can fuel effects be distinguished from air setting, turbulence, and fouling?
|
|
|
|
\item \textbf{Dynamics:}
|
|
Do temporal statistics indicate instability or blow-off proximity?
|
|
|
|
\item \textbf{Fuel inference:}
|
|
Can variable fuels (e.g. syngas) be characterised indirectly?
|
|
|
|
\item \textbf{Robustness:}
|
|
How sensitive are results to burner geometry and viewing conditions?
|
|
|
|
\item \textbf{Implementation:}
|
|
Is low-cost photodiode hardware sufficient for industrial deployment?
|
|
|
|
\end{itemize}
|
|
|
|
\fi
|
|
|
|
\section*{Proposed next step}
|
|
|
|
Build a multi-channel demonstrator including:
|
|
|
|
\[
|
|
\SI{310}{nm}\ \ce{OH^{*}} \quad \text{and} \quad \SI{430}{nm}\ \ce{CH^{*}}
|
|
\]
|
|
|
|
Additional channels may include:
|
|
\begin{itemize}
|
|
\item broadband visible;
|
|
\item green-filtered ($C_2$ Swan band);
|
|
\item broadband IR.
|
|
\end{itemize}
|
|
|
|
The selected microcontroller provides sufficient ADC channels and DSP capability.
|
|
Each IR channel requires one op-amp, while each UV channel requires two, which remains
|
|
compatible with the current architecture.
|
|
|
|
\end{document} |