High-Tech AI to Improve Coal-Burning Power Plants
Sponsors: U.S. Department of Energy.
Hot Surface Ignition in an Aircraft Environment
Team: Mehmed Ulcay, Luke Dillard, Jay Gore.
Experimental Investigation of CO2 addition effects on turbulent premixed combustion
Team: Dong Han, Jupyoung Kim, Jay Gore.
Sponsors: Department of Energy, National Energy Technology Laboratory (NETL), and University Turbine Systems Research (UTSR) Program.
Large Eddy Simulation of Hydrogen Piloted CH4/Air Premixed Combustion with CO2 dilution
Large eddy simulation of peak-temperature-matched, hydrogen-piloted, turbulent lean premixed methane/air jet flames with varying amounts of CO2 addition are reported. Such flames are relevant to low NOx gas turbines with high hydrogen content fuels and exhaust gas recirculation. A newly designed burner called Piloted Axisymmetric Reactor Assisted Turbulent (PARAT) flame burner was utilized. The operating conditions in the experiment were selected to highlight the kinetic effects of CO2 addition by matching the Reynolds numbers, Lewis numbers and adiabatic flame temperatures. The LES simulations utilize a finite rate chemistry solver with DRM19 combustion mechanism with adaptive zoning and a dynamic structure turbulence model. A five-level adaptive mesh refinement improves the velocity and temperature gradient resolution. The LES predicts the experimentally observed increase in flame length with CO2 levels caused by a decrease in the turbulent flame speed. The computational results also capture the experimentally observed departure from the thin flame limit and a collapse of the RMS versus mean temperature profiles for the three levels of CO2. The flame structure analysis showed significant increase in CO concentration higher than equilibrium values due to non-equilibrium chemistry effects because of the external addition of CO2 for the EGR emulating flames.
Team: Veeraraghava Raju Hasti, Jay Gore.
Structure and Dynamics of Alternative Aviation Fuels' Fires
This experimental investigation aims to determine the dynamics of initiation and propagation of pool fires. A novel apparatus for examining the pool fires allows us to characterize the differences in ignition, flame spread, and burning rates of buoyant fires on a temperature controlled pool exposed to laboratory air environment. Pool fire studies are performed on commercial aviation fuel like Jet A as well as alternative fuels including Fischer Tropsch(FT) fuels, HRJ (hydro processed renewable jet fuel) and synthetic iso-paraffin. The effects of fuels' initial temperature on ignition and flame spread rate are also investigated in this study. Below is a video which shows laser ignition process.
Team: Vikrant Goyal, Yerbatyr Tursyn, Jupyoung Kim, Jay Gore.
Sponsors: Federal Aviation Administration.
Lean Blowout Simulation in a Realistic Gas Turbine Combustor
Numerical simulations are reported in this paper studying the flow-split through various parts of a realistic gas turbine combustor inclusive of swirler passages, dilution jets, and all effusion/slot cooling holes. Numerical simulations including hardware details such as liners with effusion holes are challenging due to: (a) highly complex meshing of a large number of tiny effusion holes, and (b) requirement for an optimized robust mesh for a faster turnaround time to support engineering design calculations. With the goal to optimize turn-around time and accuracy, flow-split studies are carried out using steady-state Reynolds Averaged Navier-Stokes (RANS) and Large Eddy Simulation (LES) approaches. A second-order spatial discretization scheme for momentum and the modified Pressure Implicit Splitting of Operators (PISO) scheme for pressure-velocity coupling are used. We start with a study for the determination of the mesh resolution required for a good estimation of the discharge coefficient (𝑪𝑫) for a generic single hole, representative of an effusion hole, by varying: (i) the base mesh, (ii) near wall cell size based on a non-dimensional near wall distance (𝒀 +) and (iii) velocity gradients. In the next part, component wise flow-split simulations for the whole combustor with and without the effects of “dilution” jets and “effusion cooling” holes are compared against the experimental data. The results help establish the efficacy of steady-RANS and LES simulations in the determination of componentwise and total flow splits in the computational framework. The method used in the numerical calculation of component-wise flow splits involves closing all other passages except the one of interest while maintaining the pressure drop and is similar to the method used in the experiments. The total flow split, in turn, is estimated with all flow passages open, allowing for an interaction with different flows. This study is aimed at establishing a method of simulations for capturing the flow-splits and the velocity fields which in-turn are essential for predictive gas turbine simulations.
Team: Veeraraghava Raju Hasti, Jay Gore.
AFRL Referee Combustor:
Bluff Body Stabilized Turbulent Premixed Flame
Large Eddy Simulations of a bluff-body stabilized premixed flame (Volvo test case) with an objective to identify best practices for reacting flow simulations using fully automated meshing and Adaptive Mesh Refinement (AMR) is conducted in this study. Flamelet Generated Manifold model (FGM) and Thickened Flame model (TFM) were evaluated on different grid resolutions in this study. The geometry and boundary conditions used in the current work follows the guidelines outlined by the MVP-2 workshop. In order to identify grid independent settings, the base mesh size is varied from 3 mm down to 2 mm with further mesh refinements using AMR based on second gradients of velocity and temperature. The dynamic structure SGS model is used to close the turbulence quantities. The well-known UCSD detailed reaction mechanism with 57 species and 268 reactions for propane/air combustion is used with the FGM model. A two step global chemistry model is used in the TFM simulations in the present work. The LES results are compared against experimental data for velocity, temperature and CO mass fraction to assess the grid sensitivity on predictions with FGM and TFM models.
Team: Veeraraghava Raju Hasti, Jay Gore.
CO Mass-fraction Contour:
Instantaneous Temperature Distributions in Flamelet Generated Manifolds (FGM) and Thickened Flame Model (TFM) Simulations
Ignition and Flame Kernel Development in Premixed Flowing H2/air Mixtures
Minimum ignition energy and flame kernel propagation were investigated in premixed mixtures at different equivalence ratio and flow velocity. The ignition experiment used a laser-induced spark as an ignition source with lean premixed H2/air mixtures. The velocity of flowing combustible mixtures was varied between 15 and 30 m/sec. High speed Schlieren Imaging and an infrared camera were used to investigate the flame kernel propagation. Particle image velocimetry was used to examine the velocity and the turbulence intensity in a flow field.
Team: Seunghyun Jo, Jupyoung Kim, Jay Gore.
Autoignition Studies of Gas Turbine Fuels and Emulsion
Land-based gas turbine deploy a lean premixed pre-vaporized combustion concept to improve fuel efficiency by improving mixing of fuel and air. In a gas turbine engine, air is compressed by a compressor which causes an increase in pressure and corresponding temperature rise. Fuel premixes with the hot air stream in this environment with a probability of auto-ignition at these elevated pressures and temperatures. This study aims to determine the conditions that can cause fuel to autoignite in the premixer section, an important design criterion that needs to be determined prior to designing of high performance low emission engines. A continuous flow, high pressure, high-temperature operable test apparatus has been designed and installed to study the effect of pressure, temperature, velocity, and other key engine relevant parameters on ignition probability of fuel introduced in this hot environment. The test rig has diagnostic equipment to capture and characterize ignition events at elevated pressures and temperatures.
Team: Venkat Athmanathan, Jay Gore.