This investigation focuses on the premixed conical flame, a mainstay of combustion diagnostics. While extensive studies have explored high-frequency oscillations in the canonical Bunsen flame, little attention has been given to low-frequency oscillations affecting flame fronts. This thesis aims to explain the origins of these oscillations and proposes mitigation strategies. The goal is to improve high-accuracy diagnostics of standard conical flames by generating stable flame plumes through oscillation reduction.
The thesis explores two theories for low-frequency oscillations: buoyancy variations and flow rate fluctuations. Image chemiluminescence and CARS spectroscopy validate these hypotheses. Spatial oscillations are detected using flame front edge detection algorithms, while CARS analysis indirectly confirms fluctuations. Preliminary experiments establish reference data sets, followed by a primary experimental campaign spanning Reynolds numbers from 500 to 1000 for constant equivalence ratios. This isolates the effect of buoyancy variations. The report concludes by suggesting simple design changes for the flame holder to reduce instability amplitudes.

Solid rocket motors power sounding rockets and their thrust profile are determined mostly by their internal geometry. This study investigates the effect of the changes in internal geometry by changing the aspect ratio, truncation and elongation of grains in multiple BATES grain solid rocket motors. BATES grains are propellant segments in a thick-walled cylinder shape. Longitudinal effects, like erosive burning, augment the burn rate over the motor length changing the thrust profile. As part of the thesis, a model was made accounting for longitudinal variation both in the grain regression and gas flow. This model was validated through data from T-Minus Engineering measuring the pressure at the forward closure. With a sensitivity study, the effect of aspect ratio and length was researched. This resulted in a recommendation for geometry changes of the DART booster of T-Minus Engineering B.V. . Furthermore, additional validation data was acquired of static pressure along the length of the motor, instead of single-point measurements, to provide better validation of one-dimensional models. This was done using a non-invasive and inexpensive method using strain gauges on the chamber.

With the onset of climate change it is crucial to eliminate emissions, particularly for aviation reliant on gas turbine combustion. Computational models aid combustor improvement, but experimental data needed for validation is difficult to obtain. Fortunately, laser diagnostics can now achieve non-intrusive, high-resolution combustion measurements, especially with hybrid femtosecond/picosecond (fs/ps) coherent Raman scattering (CRS). In this thesis, the goal was to develop hybrid fs/ps CRS spectroscopy for the CH₄ ν₂ vibrational mode. Methane merits investigation as a climate forcer alongside potential use in carbon-neutral hydrogen production. Time-resolved CRS was performed for collision-independent, collisional dephasing, and laminar CH₄-air flame measurements. A time-domain model well-reproduced experiments, along with extracting molecule-specific ν₂ Q-branch dephasing linewidth parameters. Additionally, fundamental physio-chemical processes were observed in a spectral molecular fingerprint region across the CH₄-air flame: CH₄ dissociation, H₂ synthesis and consumption, mass diffusion, and fuel/oxidizer mixing.

In this M.Sc. thesis project a model is developed and validated that simulates the coherent anti-Stokes Raman scattering (CARS) signal of CO₂ and O₂ in the spectral region between 1250cm^-1 and 1680cm^-1. The aim is to perform temperature and concentration measurements in a typical hydrocarbon-air combustion flame. The project makes use of a two-beam time resolved CARS setup, with ultraboadband generated light from the pump/Stokes beam to excite the Raman transitions in this spectral window. In the 1250-1680cm^-1 region, CARS signatures of both CO₂ and O₂ are visible making it possible to perform thermometry on both molecular species, while also offering the option of evaluating relative CO₂-O₂/O₂-CO₂ concentrations. The appearance of the oxygen ro-vibrational (O-, S- and Q-branch) spectrum at low temperatures, along with a strong manifestation of the CO₂ Fermi influenced Q-branch (with red-shifted peaks below 1300cm^-1 and blue shifted peaks above 1350cm^-1) at higher temperatures, makes it possible to perform thermometry in low and high temperature combustion environments on both the reactant and product side. A total of 256 vibrational levels for CO₂ are taken into account for the model to simulate the CO₂ Fermi polyad. From these 256 vibrational levels, 181 Raman transitions are possible that fulfill the criteria of Δv₁=1, Δv₂=0, Δv₃=0, Δl=0 and ΔJ=0 . Three different experiments are performed including an M-flame, a V-flame and two experiments conducted in air for model validation. The temperature analysis using CO₂ provided satisfactory results regarding temperature assessments. Depending on the experiment, standard deviations below 2.3% and mean temperatures to within 1% of the temperatures corresponding to the expected values. The O₂ analysis showed a good correspondence to the CO₂ temperature values, differing by 43-76K. The O₂ analysis showed low standard deviations for the air temperature assessment (3.27%) and reliably predicted the ambient temperature with a difference lower than 2K. The M-flame experiments showed the least correspondence to actual values. These unsatisfactory results can for one part be attributed to the high signal-to-noise ratios (SNR) and for one part due to the flickering of the flame. In terms of concentration assessment the model closely evaluates O₂-CO₂ concentrations in the ambient air and from an exhaled human breath, while the flame assessments had a mediocre correspondence to the predicted ones from \textit{chem1d}. All in all the project shows that CO₂ temperature and concentration measurements in this CARS spectral region is feasible. It offers promise from a combustion perspective due to the possibility of performing simultaneous (O₂) rotational and (CO₂/O₂) vibrational thermometry, which makes it possible to perform measurements on both the product and reactant side of the flame front, while including spectral signatures of possibly three major molecular combustion species: CO₂, O₂ and H₂. Further improvements to the model and the application of the technique, make more complex combustion studies possible and help achieve the goal to develop a ultrafast, multiplex, state-of-the-art laser diagnostic tool for gas phase combustion measurements.

Work is presented on a novel hybrid fs/ps two beam CARS application for spatiotemporal gas phase hydrogen thermometry, capable of sampling at a rate of 1 kHz. The aim of the research is to assess the feasibility of directly probing hydrogen for accurate temperature measurements in a combustion environment. Typically, nitrogen is used in applied CARS thermometry because of the inert nature of the gas and the availability on both the products- and reactants side of the flame. Nitrogen is well understood in CARS applications and is therefore used as a benchmark for the results obtained through hydrogen thermometry. No hydrogen was available in the lab, so a rich premixed methane-air flame was used in a regular Bunsen burner. At high equivalence ratio’s sufficient amounts of hydrogen are formed as an intermediate species to be probed in the flame. Quantum-mechanicalmodels were used to predict the molecular responses of the corresponding molecules, after which they are compared to the experimental data. The feasibility of hydrogen thermometry was assessed through a series of three different experiments: using the standard output pulse of the regenerative amplifier at 35 fs for spatiotemporal measurements, using the 35 fs pulse for point measurements, and using a frequency broadened 10 fs pulse produced through filamentation for point measurements. Nitrogen and hydrogen spectra were recorded simultaneously to ensure proper data comparison. The 1 kHz acquisition rate enabled the evaluation of temporal correlation in the temperature data. The spatiotemporal CARS measurements clearly depicted flame dynamics at 1 Hz and 12Hz, but the signal to noise ratio in the hydrogen signal was too weak to properly assess the data. The point measurements using the 35 fs pulse yielded similar average temperatures, but the precision in the hydrogen temperatures was poor (25%) due to the fact that only the first two S-branch transition in hydrogen were excited. In the experiments using the 10 fs pulse, the first four S-branch transitions in hydrogen were excited. This greatly improved the precision for the temperature measurements (2%), surpassing that of the nitrogen thermometry (4%). The nitrogen and hydrogen temperatures also showed great temporal correlation (0.85-0.9). However, there was a very consistent temperature deficit in the hydrogen temperatures of 550K, most like caused by a wavelength dependence in the optics (which are optimized for narrowband applications) which severely affects the higher Raman shifted hydrogen signal. Lastly, an experiment was done using the 10 fs pulse to probe O-branch transitions in hydrogen, and a serious discrepancy between the CARS and CSRS spectra were found, again likely caused by the optics in the setup. These results approximated nitrogen temperatures much better, within 7% at 2000 K. Overall, the results were promising, but not yet conclusive and more research is needed.

Hybrid fs/ps coherent anti-Stokes Raman CARS is a powerful laser diagnostic technique for thermometry and concentration measurements in chemically-reacting flows. The use of SHBC to generate the ps probe pulse is simple, robust, and relatively efficient. However, the presence of spectral sidebands in this pulse determines an additional modulation of the CARS spectrum, preventing comparisons to synthetic models. A solution is to employ a 4F-filter, to clean the spectrum, but dumps ~70% of the probe pulse energy. Fs/ps CARS thermometry can greatly benefit from an investigation of the spectrochronographic properties of the ps pulses, removing the filter. The probe pulse was characterized through frequency- and time-correlation measurements: its spectrum is discretized and the relative phase of each spectral segment is determined by measuring its arrival time. The fs pump/Stokes pulse is employed as an active temporal gate to perform time-correlation measurements. The phase information is further tuned using a genetic algorithm optimization. The resulting model is capable of rendering the additional modulation of experimental CARS spectra.

For thermometry and concentration measurements in the gas-phase two-beam hybrid femtosecond/picosecond (fs/ps) coherent Raman spectroscopy (CRS) is the benchmark. To acquire additional information on major combustion species like for example carbon dioxide, oxygen and molecular hydrogen a supercontinuum is required. In this work in-situ femtosecond-laser-induced filamentation is proposed. This is a non-linear optical process where the pulse self-focusses. Through self-phase modulation and non-linear dispersion in the plasma medium a supercontinuum is created. In-situ generation of the supercontinuum allows for a great simplification of the experimental setup as no additional chirped mirrors are required to deliver a near-transform-limited ultrabroadband pulse to the probe volume. The work presents an experimental campaign split into three phases to characterise the filamentation process, the probe volume dimensions, and the effects of the filamentation process on thermometry.

There is a tremendous push to explore alternative fuels such as hydrogen or synthetic fuels to combat the effects of climate change. An important step in the development of alternative fuels is the characterisation of its combustion characteristics. There are various diagnostic tools which aid in the determination of flame properties. Optical diagnostics play a crucial role by providing a non-invasive form of measurement. In the current project the idea is to utilise the passive chemiluminescence from flames to determine flame properties. The focus is on the emission from the OH* radical present in flames. This radical is responsible for radiation in the ultraviolet region ranging from 280 nm to 350 nm. This radical was also chosen as it is a radical common to both hydrocarbon as well as hydrogen flames. The objective of the project was to develop an emission spectra model of the OH¤ radical and to test its feasibility in predicting flame temperature by comparing it with emission from a laminar premixed methane/air Bunsen flame. The underlying physics regarding the various aspects of the project are discussed in detail in the initial part of the report. In the latter part, the work done for the project is explained. An emission model was developed for the OH* radical in MATLAB. The model is capable of calculating the emission spectra at different temperatures and line shapes. Then a spectrometer was built to capture the radiation from the flames in the UV region. The obtained data was post processed to get the flame spectra. The flame spectra obtained from the flame was compared against the emission model to determine the best temperature fit. It was observed that the temperatures were much higher than the gas temperatures observed in methane/air flames. This is attributed to a non thermal population distribution in the excited molecules. There however exists a unique mapping to the equivalence ratio which can be utilised for equivalence ratio determination when properly calibrated. There are many possible directions uncovered to take the work further by exploring the applicability of the tool for other fuels and flame regimes.

This master thesis investigates the assessment of Flame-Wall Interaction using 1D broadband RCARS thermometry. Flame-Wall Interactions (FWI) have been shown to adversely effect combustion processes in terms of both efficiency and maximum performance. Heat and momentum losses cause incomplete combustion and allow for unburnt hydrocarbons (UHCs) and unoxidised radicals such as carbon monoxide to leave a combustion engine. This is not accurately represented within presently used combustion models, which raises the need for accurate data on essential combustion parameters, such as temperature, during a FWI. In order to assess thermal energy losses, CARS thermometry has been found to be an accurate, non-intrusive temperature measuring technique which can be used for in situ probing of combustion processes. Due to increasing power densities of newly developed combustion engines as well as continuously increasing aerial traffic, the adverse effects of FWI on combustion affect the efficiency and environmental impact of aircraft engines only increases. By curbing the losses induced by maintaining combustion in a confined space, depleting fuel reserves are better conserved and adverse environmental effects curbed. This study focuses on the evaluation of temperature data obtained by broadband RCARS thermometry of a turbulent FWI as well as the delay of the probe pulse with respect to the pump/Stokes beams. Since the dephasing of the excited rotational states is not constant across the rotational states, the measured relative population of each state changes with increasing delay. In order to compensate for the resulting spectral heating issue, the pulse delay needs to be accounted for when generating the spectral libraries used to obtain temperature information from the spectral data captured by the camera. The integrity of these spectral libraries has been investigated by performing a spectral fitting routine of ideal CARS spectra computed by a Matlab code. The spectral libraries consist of Nitrogen CARS spectra ranging from 200K to 2400K in steps of 100K. The probe pulse delays investigated range from 50 ps to 450 ps. The temperature error measured in this study reduces with an increase in temperature and alternates between overestimating and underestimating the actual temperature of the investigated spectrum.

Water accumulating in the fuel tanks of any aircraft presents an unacceptable safety hazard. This accumulation, which is due to a combination of physical phenomena, is difficult to assess and therefore not well known. The purpose of this project is the development of a lightweight computational prediction method for water accumulation in helicopter fuel tanks. This model delivers a better understanding of the physical mechanisms involved and allows prediction of the accumulated water quantity under given conditions.

The worldwide gas emissions regulations are getting more stringent. Together with the fluctuations of the finite fuel resources, it is expected that the use of unconventional and renewable fuels will increase. Hydrogen fuel is a zero-emission, renewable energy source. It has a great potential as one of the principal sustainable energy sources of the future. A gradual conversion to a hydrogen based energy system, if the necessary government framework is provided, is far more likely to happen. This opportunity can be found in the natural gas grid of the Netherlands. This natural gas infrastructure is accessible for both the industry and Dutch households. The IEA HGH RD programme has conducted research about this gradual transition, with promising results. It is currently therefore a key agenda point for the Dutch government. In order to safely make a transition to natural gas with hydrogen addition, it is of importance to know the fundamental combustion properties of these fuel mixtures. These properties are needed for the prediction of and modeling of combustion systems. Traditionally, the quantification of the combustion process is done by focusing on measuring properties such as laminar burning velocity, ignition and extinction characteristics, flammability limits and emissions. The laminar burning velocity is an important property of combustion mixtures. It contains fundamental information on reactivity, exothermicity and diffusivity.

The developments in combustion technology has to be catered to in parallel by the development in the instrumentation employed in collection of combustion data, to get a better picture of the complex processes involved. In this regard, the flame temperature and equivalence ratio are the most important parameters that has to be studied to infer the combustion efficiency and emissions from combustion. Combustion reactions in flames give rise to the formation of various intermediate radicals, some of which are chemiluminescent by nature. The relative intensity of optical emissions from chemiluminescent radicals such as OH*(the asterisk is used as a notation for electronically excited states), CH* and C2* can be used for flame thermometry and spatial analysis of flame structure in hydrocarbon flames. The C2* radical and CH* radical is emphasized upon in this study for the analysis of laminar premixed hydrocarbon flames. Spatial distribution of these radicals is mapped from their chemiluminescence signals and recorded across various zones of the flame using a fiber optic bundle to capture incoming photons, a Czerny-Turner spectrometer encompassing a diffraction grating for optical dispersion and a CCD detector to record the spectrum. Local flame spectra recorded in the reaction zone of methane/air flames are known to exhibit strong emission signals from OH*, CH* and C2* radicals. The peak intensity ratios of these radicals in the reaction zone are functions of local equivalence ratio. To sum up, spectral intensity measurements of such chemiluminescent radicals serve as tools for analyzing the local flame stoichiometry and for the measurement of local flame temperature.
In this thesis work, an experimental set-up is proposed to capture the spectral signals of a premixed, laminar, methane-air Bunsen flame at various spatial locations across the geometry of the flame. Using the emission spectra of this flame, the flame characteristics like relative intermediate radical species concentration, the spatial distribution of radicals, the flame equivalence ratio, the rotational and vibrational flame temperature and their degree of agreement with the adiabatic flame temperature are studied. It has been found that a definite relationship exists between the spectral signals emitted by these chemiluminescent radicals and the flame parameters such as equivalence ratio and adiabatic flame temperature.

The characteristics of Dutch natural gas (DNG) have been quantified by measuring the laminar flame speed of a premixed Bunsen flame over a range of equivalence ratios ϕ, varied from ϕ=0.80 to ϕ=1.55. The laminar flame speed has been calculated from the OH* chemiluminescence images recorded with a high speed camera with an acquisition rate of 1 kHz. Both the semi-cone angle method and mass conservation method were used in deriving the magnitude of the laminar burning velocity. From the data it was shown that the laminar flame speed of DNG is consistently lower compared to natural gas from Pittsburgh, Abu Dhabi and Indonesia. Moreover the data has been compared using the laminar flame speed of methane mixtures to validate the experimental setup. This showed that the laminar flame speed of DNG was marginally lower than the laminar flame speed of methane. The uncertainty within the measurements has been quantified by deriving the probability density function of the data.