Due to manufacturing tolerance and deterioration during operation, different blades in a fan assembly exhibit geometric variability. This leads to asymmetry which will be amplified in the running geometry by centrifugal and aerodynamic loads. This study investigates a phenomenon known as Alternate Passage Divergence (APD), where the blade untwist creates an alternating pattern in passage geometry and stagger angle around the circumference. After the formation of alternating tip stagger pattern, APD’s unsteady effect, APD-induced Non-Synchronous Vibration (APD-NSV, abbreviated as NSV), can cause the blades from one group to switch to the other creating a travelling wave pattern around the circumference. Thus, it can potentially lead to high cycle fatigue issues. More importantly, this phenomenon occurs close to, or at, peak efficiency conditions and can significantly reduce overall efficiency. Therefore, it is vital to attenuate the NSV behaviour. The random nature of mis-staggering patterns complicate the evolution of NSV significantly. Thus, machine learning techniques are used to analyse mis-stagger patterns to identify patterns that can lead to NSV and thus help avoid it. Numerical results from 113 numerical cases (1.6 million CPU hours) are used to train and test the classifier. From the results, two parameters contributing to NSV behaviour have been identified with one of them enhancing the understanding found in the previous study.

The influence of turbulence modeling approach by means of (U)RANS and LES on the overall modeling of turbulent condensing wet steam flows is investigated using the example of a low-pressure steam turbine cascade. For an accurate numerical treatment of turbulence in presence of shock waves, necessary for predictive scale-resolving computations, a hybrid flux treatment switches between a baseline non-dissipative central flux in energy consistent split form and a shock-capturing upwind flux in shocked regions based on a shock sensor. Condensation is realized by a mono-dispersed Euler-Euler source term model, the equation of state by the highly efficient and accurate SBTL tabulation. The numerical treatment is validated with a decay of homogeneous isotropic turbulence test case containing eddy shocklets. The measurement results of the condensing wet steam cascade are overall much better matched by LES compared to RANS and URANS. Analysis shows that the LES is much better able to account for the inherently unsteady nature of the spontaneous condensation process and its interaction with the trailing edge shock wave structure.

Accurate numerical prediction of surface heat transfer in the presence of film cooling within aero-engine sub-components, such as blade effusion holes and combustor liners, has long been a goal of the aero-engine industry. It requires accurate simulation of the turbulent mixing and reaction processes between freestream and the cooling flow. In this study, the Stress Blended Eddy Simulation (SBES) turbulence model is used together with the Flamelet Generated Manifold (FGM) combustion model to calculate the surface heat flux upstream and downstream of an effusion cooling hole. The SBES model employs a blending function to automatically switch between RANS and LES based on the local flow features, and thus significantly reduces the computational cost compared to a full LES simulation. All simulations are run using ANSYS Fluent ® , a commercial finite-volume CFD solver. The test case corresponds to an experimental rig run at MIT, which is essentially a flat plate brushed by a uniform freestream of argon with ethylene seeded inside, and is cooled by either a reacting air or non-reacting nitrogen jet inclined at 35 degrees to the freestream. Calculations are performed for both reacting and non-reacting jet cooling cases across a range of jet-to-stream blowing ratios, and compared with the experimental data. The effects of mesh resolution are also investigated. Calculations are also performed across a range of Damköhler number (i.e. flow to chemical time ratio) from zero to 30, with unity blowing ratio, and the differences in the maximum surface heat flux magnitude in the reacting and non-reacting cases at a specific location downstream of the hole are investigated. Results from these analyses show good correlation with the experimental heat flux data upstream and downstream of the cooling hole, including the heat flux augmentation due to local reaction. Results from the Damköhler number sweep also show a good match with the experimental data across the range investigated.

Localized nonlinearities due to the contact friction interfaces are widely present in the aero-engine structures. They can significantly reduce the vibration amplitudes and shift the resonance frequencies away from critical operating speeds, by exploiting the frictional energy dissipation at the contact interface. However, the modelling capability to predict the dynamics of such large-scale systems with these nonlinearities is often impeded by the high computational expense. Component mode synthesis (CMS) based reduced order modelling (ROM) are commonly used to overcome this problem in jointed structures. However, the computational efficiency of these classical ROMs are sometimes limited as their size is proportional to the DOFs of joint interfaces resulting in a full dense matrix. A new ROM based on an adaptive formulation is proposed in this paper to improve the CMS methods for reliable predictions of the dynamics in jointed structures. This new ROM approach is able to adaptively switch the sticking contact nodes off during the online computation leading to a significant size reduction comparing to the CMS based models. The large-scale high fidelity fan blade assembly is used as the case study. The forced response obtained from the novel ROM is compared to the state-of-the-art CMS based Craig-Bampton method. A parametric study is then carried out to assess the influence of the contact parameters on the dynamics of the fan assembly. The feasibility of using this proposed method for nonlinear modal analysis is also characterised.

Climate science shows that the limitation of global warming requires a rapid transition towards net-zero emissions of green house gases (GHG) on a global scale. Expanding renewable power generation in a significant way is seen as an imperative measure within this transition. To compensate for the inherent volatility of wind- and solar-based power generation, flexible and dispatchable power generation technologies such as gas turbines are required. If operated with CO 2 -neutral fuels such as hydrogen or in combination with carbon capture plants, a GHG-neutral gas turbine operation could be achieved. An effective leverage to enhance carbon capture efficiency and a possible measure to safely burn hydrogen in gas turbines is the partial external recirculation of exhaust gas. By means of a model-based analysis of an industrial gas turbine, the present study initially assesses the thermodynamic impact caused by a fuel switch from natural gas to hydrogen. Although positive trends such as increasing net electrical power output and thermal efficiency can be observed, the overall effect on the gas turbine process is only minor. In a following step, the partial external recirculation of exhaust gas is evaluated and compared both for the combustion of natural gas and hydrogen, regardless of potential combustor design challenges. The influence of altering working fluid properties throughout the whole gas turbine process is thermodynamically evaluated for ambient temperature recirculation and recirculation at an elevated temperature. A reduction in thermal efficiency can be observed as well as non-negligible changes of relevant process variables. These changes are are more distinctive at a higher recirculation temperature.

Due to manufacturing tolerance and deterioration during operation, different blades in a fan assembly exhibit geometric variability. This leads to asymmetry which will be amplified in the running geometry by centrifugal and aerodynamic loads. This study investigates a phenomenon known as Alternate Passage Divergence (APD), where the blade untwist creates an alternating pattern in passage geometry and stagger angle around the circumference. After the formation of alternating tip stagger pattern, APD’s unsteady effect, APD-induced Non-Synchronous Vibration (APD-NSV), can cause the blades from one group to switch to the other creating a travelling wave pattern around the circumference. Thus, it can potentially lead to high cycle fatigue issues. More importantly, this phenomenon occurs close to, or at, peak efficiency conditions and can significantly reduce overall efficiency. Therefore, it is vital to attenuate the NSV behaviour. In this study, an redesign approach is investigated.

Due to manufacturing tolerance and deterioration during operation, fan blades in the aero-engine exhibit geometric variability. This leads to asymmetry in the assembly which will be amplified in the running geometry by centrifugal and aerodynamic loads. This study investigates a phenomenon known as Alternative Passage Divergence (APD), where the blade untwist creates an alternating pattern in passage geometry and stagger angle around the circumference, resulting in two groups of blades. This phenomenon occurs close to, or at, peak efficiency conditions and can significantly reduce overall efficiency. This study focuses on a type of non-integral vibration which occurs during APD. After the formation of alternating tip stagger pattern, APDs unsteady effect can cause the blades from one group to switch to the other, creating a travelling wave pattern around the circumference.It was found from numerical assessment on a randomly mis-staggered assembly that real engines can potentially experience such travelling disturbance and suffer fatigue damage. An idealised case is used to capture the bulk behaviour from the more complex cases in real engines and to decipher the underlying mechanism of this travelling disturbance. The results indicate that the driving force originates from the interaction between passage shock displacement and the passage geometry.

Piezoelectric-based vibration reduction has the potential to improve the lifetime and structural integrity of turbomachinery blades by reducing the risk of high-cycle fatigue. Semi-active techniques produce small, self-powered implementations that can meet the strict design requirements for rotating machinery. Most semi-active techniques switch piezoelectric transducers between an open circuit and a shunt circuit in a way that reduces vibration. However, these switches produce a small impulse on the structure due to the electromechanical coupling of the piezoelectric material. Since multiple-mode vibration reduction typically requires distributed sections of piezoelectric material, the switching impulse generated by one transducer may affect others. This study investigates the effect of the switching impulse on the non-switched piezoelectric transducers. Experimental testing shows the switching impulse induces a voltage response at the natural frequencies of the structure, with the strongest responses occurring at modes where both the switched and non-switched piezoelectric transducers have high electromechanical coupling. Furthermore, piezoelectric sections that lack coupling at the excitation frequency of the structure exhibit a more noticeable response to the switching impulse. This response enables remote sensing of switches, which may facilitate wireless coordinated switch timing.

This work presents a novel strategy to numerically predict CO emissions in gas turbines that operate under part-load conditions employing fuel-staging concepts. In multi-burner systems, fuel can be redistributed to solely run a fraction of the available burners. The situation of active burners interacting with air from adjacent cold burners may lead to quenching effects. Our group recently published a flamelet-based combustion model for low-reactive conditions. Furthermore, a model was proposed for the prediction of CO beyond the assumption of thin reaction zones. These models are adopted in this work and further extended in order to capture quenching. All models are implemented and applied to a simple geometry for the purpose of demonstrating basic mechanisms that are relevant for fuel-staged gas turbines operating at part load conditions. Furthermore, validation is performed in a silo combustor that comprises 37 burners. Here, burner groups are switched off during part load, leading to intense interaction between hot and cold burners. Major improvement in comparison to CO predictions from the flamelet-based combustion model is achieved. It is demonstrated that the model is able to predict the correct values of global CO emissions. Furthermore, the models capacity of handling fuel-staging mechanisms like the CO drop during a burner switch-off event is shown.

The present paper describes atmospheric experimental tests of a new Ansaldo Energia full scale burner which was designed to burn fuels byproduct of steel making processes (mixtures of Blast-Furnace Gas (BFG) and Coke-Oven Gas (COG)), characterized by very low heating values (LHV∼2–3.5 MJ/kg) and very low stoichiometric air/fuel ratios (∼0.5–1 kg/kg). In particular, flame stability and blow-out margins were assessed for different burner variants and fuel compositions such as pure BFG, blends of BFG with increasing content of COG, and also a synthetic mixture of natural gas, hydrogen and nitrogen (NG/H 2 /N 2 ). Except for pressure, all burner inlet conditions were simulated as in the actual gas turbine engine. The best performing burner among those tested demonstrated an excellent burning stability behavior over a wide operating range and stably burned pure BFG without any supplementary fuel. Furthermore, considering that in most operating concepts gas turbine engines for Ultra-Low BTU applications require a back-up fuel (such as oil, propane or natural gas) to ignite and ramp up or to perform load-rejections, the present atmospheric tests also assessed maneuvers to switch from natural gas operation to syngas operation. Also in this type of dual-fuel operation the burner demonstrated a wide flame stability range.

Hybrid turbulence modeling is a practical approach to efficiently model the wall-bounded turbulent flows. In this paper, a stress-blended eddy simulation (SBES) model is used with the flamelet generated manifold model (FGM) for modeling turbulent combustion. In the current SBES, the near-wall region is modeled using a two-equation k-ω Reynolds-averaged Navier-Strokes (RANS) formulation, and switches to a large eddy simulation (LES) model in the core region using a blending function. Similarly, the turbulence-related combustion modeling parameters, such as the variances in scalar transport equations and scalar dissipation, are also blended using the same blending function. This combined hybrid FGM-SBES approach is implemented into ANSYS Fluent software and then used to model a swirl-stabilized flame. The flame used is a methane-fueled burner, developed at DLR Stuttgart as the PRECCINSTA combustor. The experimental data for this combustor are available for multiple operating conditions. A stable operating point (φ = 0.83, P = 30 kW) is chosen. The current FGM-SBES results are compared with experimental data as well as with FGM-LES computations. Differences in predictions of mean and variance of reaction progress and mixture fraction in the core versus the near wall region are analyzed and quantified. The impact of the differences in these parameters is then evaluated by comparing temperature and species mass fractions. The findings from the current work, in terms of accuracy, validity and best practices when modeling wall-bounded flows with FGM-SBES are discussed and summarized.

With the increase of computational power, more sophisticated computational methods can be used, larger systems simulated, and complex phenomena predicted more reliably. Nevertheless, up to now, when turbomachinery systems are numerically optimized, each of the components, i.e., the compressor, combustor, and turbine, is simulated separately from the other two. While this approach allows the use of highly dedicated simulation tools, it does not account for the interactions between the different components. With the purpose to meet the future requirements in terms of low emissions, high reliability and efficiency, a novel, highly efficient, fully-coupled, approach based on the Reynolds-Averaged Navier-Stokes equations (RANS) has been developed, enabling a steady or time-accurate simulation of a full aero-engine within a single code. One of the advantages of a steady, fully coupled approach over a steady component-by-component approach, is that the boundary conditions at the interfaces do not need to be guessed. A fully coupled, time-accurate simulation has furthermore the advantage that the effect of the non-uniform temperature distribution at the outlet of the combustor is accounted for in the determination of the thermal field of the turbine. A Smart Interface methodology permits a direct coupling between the different engine components, compressor-combustor-turbine, and allows the Computational Fluid Dynamics (CFD) models to vary between each component within the same code. This allows the user to switch off, for instance, the combustion model in the turbine and compressor blocks. For the simulation of the combustion process, the Flamelet Generated Manifold (FGM) method is applied. While the approach is superior to classical tabulated chemistry approaches and reliably captures finite-rate effects, it is computationally inexpensive since it only requires the solution of a few extra scalars and the look-up of a combustion table. The model has been extended so that high-speed compressible flows can be simulated and the potential effects between the combustor and the adjacent blade rows can be accounted for. The Nonlinear Harmonic (NLH) method is used to model the unsteady interactions between the blade rows as well as the influence of the inhomogeneities at the combustor outlet on the downstream turbine blade rows. Compared to conventional time-accurate RANS simulations (URANS), this method is two to three orders of magnitude faster and makes time-accurate turbomachinery simulations affordable. With the aim of ensuring thermodynamic consistency between the different components of the engine, the same form of the energy equation is solved in all engine elements. Furthermore, the same thermodynamic coefficients, which are used to describe the reacting processes in the combustor, are used for a caloric description of the fluid in the compressor and turbine blocks. The thermodynamic data between the blocks is transferred using the OpenLabs™ module. The developed approach is described in detail and the potential of the novel full-engine methodology is exploited on the KJ66 micro-turbine gas engine case. The results of both the steady and the time-accurate, fully coupled approaches are analyzed and the interaction between the different components of the KJ66 engine discussed.

Phosphor thermometry has been developed for wall temperature measurements in gas turbines and gas turbine model combustors. An array of phosphors has been examined in detail for spatially and temporally resolved surface temperature measurements. Two examples are provided, one at high pressure (8 bar) and high temperature and one at atmospheric pressure with high time resolution. To study the feasibility of this technique for full scale gas turbine applications a high momentum confined jet combustor at 8 bar was used. Successful measurements up to 1700 K on a ceramic surface are shown with good accuracy. In the same combustor, temperatures on the combustor quartz walls were measured, which can be used as boundary conditions for numerical simulations. An atmospheric swirl-stabilized flame was used to study transient temperature changes on the bluff body. For this purpose, a high-speed setup (1 kHz) was used to measure the wall temperatures at an operating condition where the flame switches between being attached (M-flame) and being lifted (V-flame) (bistable). The influence of a precessing vortex core (PVC) present during M-flame periods is identified on the bluff body tip, but not at positions further inside the nozzle.

Turbomachinery blade technology has recently trended towards the use of monolithic bladed disks. Although offering a wealth of benefits, this construction removes the blade attachment interface present in the conventional design, thus unintentionally removing a source of friction-based damping needed to counteract large vibrations during resonance passages. This issue is further exacerbated for blade mistuning, which is well-known to induce vibration localization with correspondingly larger vibration magnitudes. Recently, an alternative method to reduce vibration, termed Resonance Frequency Detuning (RFD), utilizes the variable stiffness properties of piezoelectric materials embedded on-blade to detune the response when approaching a resonance crossing, thus resulting in reduced vibration. For a single-degree-of-freedom (SDOF) system, the vibration reduction performance and the optimal stiffness state switching is well-defined. Previously, RFD has been experimentally validated on a representative blade for a sufficiently well-separated vibration mode, thus satisfying the SDOF assumption. No such experimental validation currently exists for a system with closely-spaced modes or, more specifically, applied to blade mistuning. This work utilizes an academic blisk machined in the form of 8 blades attached to a central hub. Each blade incorporates two collocated piezoelectric patches located near the blade root: one patch provides the stiffness state modulation, while the other patch provides actuation to mimic engine order excitations. For the forcing configuration studied, experimental results show qualitative agreement to numerical results with the vibratory response associated with the optimal stiffness state switch showing reductions across all blades.

Truncating the exit of a discrete passage centrifugal compressor diffuser is observed to enhance a research compressor’s stall line. By interrogating the experimental data along with a set of well-designed Reynolds-Averaged Navier Stokes computations, this improvement is traced to reduced impact of secondary flows on the truncated diffuser’s boundary layer growth. The secondary flow system is characterized by counter-rotating streamwise vortex pairs that persist throughout the diffuser passage. The vortices are traced to two sources: background vortices resulting from impeller exit flow non-uniformity, and incidence vortices resulting from flow separation off the leading edge cusps unique to a discrete passage diffuser. The incidence vortices detrimentally impact the diffuser pressure rise capability by accumulating high loss flow along the diffuser wall near the plane of symmetry between the vortices. This contributes to a large passage separation in the baseline diffuser. Using reduced order flow modeling, the impact of the vortices on the boundary layer growth is shown to scale inversely with diffuser aspect ratio, and thus the separation extent is reduced for the higher aspect ratio truncated diffuser. Because the diffuser incidence angle influences the strength and location of the vortices, this mechanism can affect the slope of the compressor’s pressure rise characteristic and impact its stall line. Stall onset for the baseline diffuser configuration is initiated by the transition of the vortex location and corresponding passage separation between diffuser pressure and suction sides with increased cusp incidence. Conversely, because the extent of the passage separation in the truncated diffuser is diminished, the switch in separation from pressure to suction side does not immediately initiate instability.

This paper presents a conceptual design of a tail-cone thruster system which is operating under an axisymmetric inlet distortion. An effort to realize the targeted fuel burn saving that was proposed in NASA’s STARC_ABL aircraft design is made through a CFD based design approach. This method employs three iterative steps to exploit the CFD tools until the design requirements are met: a quasi-2D through-flow model to design the fan/EGV, a 3-D RANS simulation of the single blade row to account for the inlet/fan and the EGV/nozzle interaction, and a 3-D RANS simulation of the airframe with a propulsor installed – propulsion airframe integration (PAI). The design requirements which include the thrust, and shaft power of the propulsor are matched throughout the evaluations coming from two CFD domains, i.e., the turbo-machinery and the PAI. During the switch between these different computational domains, the inlet and exit profiles are matched via the correction factors of the body-force model. The present tail-cone thruster (TCT) aerodynamic design leverages a low-pressure ratio fan (FPR = 1.2∼1.25) of which the camber-line angles are predicted by a quasi-2D through-flow model. The quasi-2D model is derived to analyze the radially distorted flow resulting from the ingested boundary layer at the inlet. It also estimates the appropriate velocity vectors of the metal angles of the fan and EGV which is subjected to different types of vortex at the fan exit. The baseline geometry is revisited and its internal flow-path and exhaust cone are redesigned to illustrate the strong correlation among the components of the propulsor in the PAI domain. The peak efficiency point of the fan/EGV with respect to the blade counts, a.k.a. solidity, and rotational speed is chosen for the cruise condition via parametric studies. The corresponding performance maps are presented. The resulting performance metrics of the new conceptual design of the BLI propulsor are analyzed and compared with these of the baseline in the PAI aspect. Finally, ideas of the CFD based design of a BLI propulsor are discussed based on the observations drawn from the numerical results.

We examine the shifts in linear decay rates and growth rates, with and without control devices on a simple thermoacoustic system. When the system is stable, we measure the decay rate by pulsing with a loudspeaker. This improves on the experimental techniques developed by Rigas et al. (J. Fluid Mech., 2016, vol. 787, R1 [1]) and Jamieson et al. (Int. J. Spray and Comb. Dyn., accepted, 2016 [2]), to develop a more efficient method of experimental sensitivity analysis to apply in thermoacoustic systems. When the system is unstable, we use feedback control to bring it to a non-oscillating state. We then switch off the feedback control and measure the growth rate. The results suggest that both methods are suitable for use in the experimental sensitivity analysis of thermoacoustic systems. Our experimental set up is automated and we find that we can obtain thousands of decay rates in 1/12 the time compared with the previous work.

Gas turbine engines are widely used as the marine main power system. However, they can’t reverse like diesel engine. If the reversal is realized, other ways must be adopted, for example, controllable pitch propeller (CPP) and reversible gearing. Although CPP has widespread use, the actuator installation inside the hub of the propeller lead to the decrease in efficiency, and it takes one minute to switch “full speed ahead” to “full speed astern”. In addition, some devices need to be added for the reversible gearing, and it takes five minutes to switch from “full speed ahead” “to “full speed astern”. Based on the gas turbine engine itself, a reversible gas turbine engine is proposed, which can rotate positively or reversely. Most important of all, reversible gas turbine engine can realize operating states of “full speed ahead”, “full speed astern“ and “stop propeller”. And, it just takes half of one minute to switch “full speed ahead” to “full speed astern”. Since reversible gas turbine engines have compensating advantages, and especially in recent years computational fluid dynamics (CFD) technology and turbine gas-dynamics design level develop rapidly, reversible gas turbine engines will be a good direction for ship astern. In this paper, the power turbine of a marine gas turbine engine was redesigned by three dimensional shape modification, and the flow field is analyzed using CFD, in order to redesign into a reverse turbine. The last stage vanes and blades of this power turbine were changed to double-layer structure. That is, the outer one is reversible turbine, while the inner is the ahead one. Note that their rotational directions are opposite. In order to realize switching between rotation ahead and rotation astern, switching devices were designed, which locate in the duct between the low pressure turbine and power turbine. Moreover, In order to reduce the blade windage loss caused by the reversible turbine during working ahead, baffle plates were used before and after the reversible rotor blades. This paper mainly studied how to increase the efficiency of the reversible turbine stage, the torque change under different operating conditions, rotational speed and rotational directions, and flow field under typical operating conditions. A perfect profile is expected to provide for reversible power turbine, and it can decrease the blade windage loss, and increase the efficiency of the whole gas turbine engine. Overall, the efficiency of the newly designed reversible turbine is up to 85.7%, and the output power is more than 10 MW, which can meet requirements of no less than 30% power of rated condition. Most importantly, the shaft is not over torque under all ahead and astern conditions. Detailed results about these are presented and discussed in the paper.

The thermoacoustic oscillations of a partially premixed flame stabilized in a backward facing step combustor are studied at a constant equivalence ratio in long and short combustor configurations corresponding to with and without acoustic feedback respectively. We perform simultaneous time-resolved particle image velocimetry (TR-PIV) and chemiluminescence for selected flow conditions based on the acoustic characterization in the long combustor. The acoustic characterization shows a transition in the dominant pressure amplitudes from low to high magnitudes with an increase in the inlet flow Reynolds number. This is accompanied by a shift in the dominant frequencies. For the intermittent pressure oscillations in the long combustor, the wavelet analysis indicates a switch between the acoustic and vortex modes with silent zones of relatively low-pressure amplitudes. The short combustor configuration indicates the presence of the vortex shedding frequency and an additional band comprising the Kelvin Helmholtz mode. Next, we apply the method of finite-time Lyapunov exponent (FTLE) to the time-resolved velocity fields to extract features of the Lagrangian coherent structures (LCS) of the flow. In the long combustor post transition with the time instants with dominant acoustic mode, a large-scale modulation of the FTLE boundaries over one cycle of pressure oscillation is evident. Further, the FTLEs and the flame boundaries align each other for all phases of the pressure oscillation. In the short combustor, the FTLEs indicate the presence of small wavelength waviness that overrides the large-scale vortex structure, which corresponds to the vortex shedding mode. This behaviour contrasts with the premixed flame in the short combustor reported earlier in which such large scales were found to be seldom present. The presence of the large-scale structures even in the absence of acoustic feedback in a partially premixed flame signifies its inherent unstable nature leading to large pressure amplitudes during acoustic feedback. Lastly, the FTLE boundaries provide the frequency information of the identified coherent structure and also acts as the surrogate flame boundaries that are estimated from just the velocity fields.

A staged injector developed by JAXA and fueled with kerosene is studied in a high-pressure combustion experiment. With a stable pilot fuel flow rate, the fuel flow rate in the main stage is progressively increased. A high-speed OH-PLIF system is used to record the flame motion at 10,000 fps. In the beginning of the recording, POD modes shows that the flame behavior is dominated by relatively low-frequency rotation due to the swirling motion of the flow. These rotational motions then coexist with a thermo-acoustic instability around 475 Hz which increases the amplitude of the pressure fluctuations inside the chamber. DMD analyses indicate that this instability is associated with a widening of the flame occurring when the pressure fluctuations are the highest, giving the instability a positive feedback. The instability frequency then abruptly switches to 500 Hz while retaining the same driving mechanisms. Potential candidates for this frequency change are proposed.