Characteristic time scale as optimal input in Machine Learning algorithms:Homogeneous autoignition

Considering temporally evolving processes,the search for optimal input selection in Machine Learning(ML)algorithms is extended here beyond(i)the readily available independent variables defining the process and(ii)the dependent variables suggested by feature extraction methods,by considering the time scale that characterizes the process.The analysis is based on the process of homogeneous autoignition,which is fully determined by the initial temperature T(0)and pressure p(0)of the mixture and the equivalence ratio𝜙that specifies the initial mixture composition.The aim is to seek the optimal input for the prediction of the time at which the mixture ignites.The Multilayer Perceptron(MLP)and Principal Component Analysis(PCA)algorithms are employed for prediction and feature extraction,respectively.It is demonstrated that the time scale that characterizes the initiation of the process𝜐𝑓(0),provides much better accuracy as input to MLP than any pair of the three independent parameters T(0),p(0)and𝜙or their two principal components.Indicatively,it is shown that using𝜐𝑓(0)as input results in a coefficient of determination R 2 in the range of 0.953 to 0.982,while the maximum value of R 2 when using the independent parameters or principal components is 0.660.The physical grounds,on which the success of𝜐𝑓(0)is based,are discussed.The results suggest the need for further research in order to develop selection methodologies of optimal inputs among those that characterize the process.

An experimental and modeling study of n-hexadecane autoignition under low-to-intermediate temperatures

N-hexadecane is a potential candidate of diesel surrogate fuels and is also the largest linear alkane(n-alkanes)with known chemical kinetic models.The objective of this study is to investigate the autoignition characteristics of n-hexadecane in the lowto-intermediate temperature region and to validate the existing kinetic models.In this study,the ignition delay times(IDTs)of nhexadecane were measured using a heated rapid compression machine(RCM)at two pressures of 7 and 10 bar,and over equivalence ratios ranging from 0.5 to 1.3.Two-stage ignition characteristic and the negative temperature coefficient(NTC)behavior of total ignition delay time were experimentally captured.This study paid special attention to the influence of pressure,equivalence ratio,and oxygen content on the IDTs of n-hexadecane.It is observed that both the total IDTs and the first-stage IDTs decrease with the rise of those parameters.It is worth noting that the first-stage IDT is found to show a greater dependence on temperature but a weaker dependence on other parameters compared to the total IDT.The observed IDT dependence in the lowtemperature region(LTR)were quantitatively described by ignition delay time correlations.The newly measured IDTs were then validated against two kinetic models(LLNL and CRECK).Simulation results show that both models underpredict the first-stage IDT but generally capture the temperature dependence.The CRECK model well predicts the total IDTs of n-hexadecane while the LLNL model significantly underpredicts the total IDTs at most investigated conditions.To the best of our knowledge,this study is the first investigation on n-hexadecane autoignition under low-to-intermediate temperatures,which deepens the understanding of large n-alkane oxidation and contributes to the improvement of the existing kinetic models.

Statistical characteristics of spray autoignition of transient kerosene jet in cross flow

To study statistical characteristics of the random spray autoignition,aviation kerosene was injected transiently into non-vitiated air crossflow in a flow reactor with optical accesses.The operating conditions were relevant to gas turbine combustor:the air crossflow pressure and temperature were in the range of 1.4-1.7 MPa and 830-947 K,respectively,and the jet-tocrossflow momentum flux ratios were 20,50 and 80.Statistical distributions of random ignition delay times with adequate convergence were estimated based on histograms.The dependences of the distributions on reactor pressure,temperature,and jet-to-crossflow momentum flux ratio were studied.The results show that the resulting distributions appear more concentrated with the increase of air temperature or jet-to-crossflow momentum flux ratio.And then the correlations for the mean and standard deviation of the ignition delay time sample data were developed based on the present results.Compared with the correlations of ignition delay time of homogeneous premixed gas-phase kerosene/air mixture reported in the literature,the results show a greater significance pressure dependence and lower temperature sensitivity of the ignition delay time of nonpremixed kerosene spray.

Enhanced photocatalytic activity of (La, N) co-doped TiO_2 by TiCl_4 sol-gel autoigniting synthesis

（La, N） co-doped TiO2 photocatalysts were synthesized using TiC14 sol-gel autoignidng synthesis （SAS） starting from a complex compound system of TiCl4-La（NO3）3-citric acid-NH4NO3-NHyH2O, in which the （La, N） co-doped process was accompushed in the formation of TiO2 nanocrystals. The prepared samples were characterized by using X-ray diffraction （XRD）, X-ray photoemission spectroscopy （XPS） and UV-vis diffuse reflectance spectra. The results indicated that nitrogen and lanthanum were incorporated into the lattice and interstices of titania nanocrystals, which resulted in narrowing the band gap and promoting the separation of photoexcited hole-electron pairs, respectively, and showing expected red-shifts and enhanced photocatalytic activity under visible light. The mechanism on nitrogen doping and enhancement in photocatalyfic activity of （La, N） co-doped titania by SAS was discussed in detail.

Preparation and densification of BaTiO_3 nanopowder by sol-gel autoigniting synthesis

Autoigniting synthesis of gel from Ba(NO_3)_2, TiO(NO_3)_2 and C_6H_8O_7centre dot H_2O aqueous solution was investigated at an initial temperature of 600 deg C andtetragonal BaTiO_3 nanopowder with particle size of 80nm was prepared. It is indicated that thespecific surface area of the combustion product before and after calcinations is 14.74 m^2/g and12.49 m^2/g, respectively. The combustion wave is composed of solid phase reaction zone and gaseousphase flame reaction zone. The combustion flame temperature is 1 123 K derived from thermocouplemeasurement. The characteristics and densification behavior of the sol-gel autoigniting synthesizedBaTiO_3 nanopowder were investigated.

Stabilization mechanisms of lifted flames in a supersonic stepped-wall jet combustor

Flame stabilization is the key to extending scramjets to hypersonic speeds;accordingly,this topic has attracted much attention in theoretical research and engineering design.This study performed large eddy simulations(LESs)of lifted hydrogen jet combustion in a stepped-wall combustor,focusing on the flame stabilization mechanisms,especially for the autoignition effect.An assumed probability density function(PDF)approach was used to close the subgrid chemical reaction source.The reliability of the solver was confirmed by comparing the LES results with experimental data and published simulated results.The hydrogen jet and the incoming stream were first mixed by entraining large-scale vortices in the shear layer,and stable combustion in the near-wall region was achieved downstream of the flame induction region.The autoignition cascade is a transition of fuel-rich flame to stoichiometric ratio flame that plays a role in forming the flame base,which subsequently causes downstream flame stabilization.Three cases with different jet total temperatures are compared,and the results show that the increase in the total temperature reduces the lift-off distance of the flame.In the highest total temperature case,an excessively large scalar dissipation rate inhibits the autoignition cascade,resulting in a fuel-rich low-temperature flame.

Measuring the Adiabatic Ignition Delay of n-Pentane Mixture using Rapid Compression Machine

Recent combustion research has focused on low temperature combustion to meet engine emission regulations and to advance the development of low temperature homogenous compression ignition engines.Autoignition studies in this temperature regime are primarily performed by Rapid Compression Machines(RCMs)which are sensitive to the heat transfer characteristics of the experimental device.RCMs are widely used to measure autoignition data such as ignition delay and species concentration.Measured ignition delays from RCMs are typically reported at an adiabatic condition;however,this assumption may produce a systematic error in ignition delay measurement as heat transfer is observed to reduce the pressure and temperature during the autoignition process,e.g.,a longer ignition delay has a greater pressure and temperature drops.RCMs are custom built and have unique design characteristics that affect the heat transfer during the autoignition process.In addition,depending on the diluent composition(e.g.,helium versus nitrogen or argon),different heat transfer characteristics are expected.As a result,autoignition results at similar conditions may vary from facility to facility or depending on the used diluent.The dependency of the measured data on the used facility or diluent may produce uncertainty in the data which impact the development of high-fidelity combustion mechanisms.In this work,a new method is developed and utilized to eliminate heat transfer from the ignition delay data.To evaluate the new method,the autoignition of n-pentane mixtures in the low temperature regime were investigated using an RCM.To vary the heat transfer,the compression ratio of the RCM was changed and the ignition delays were measured at similar pressure and temperature conditions.The tests were performed at an equivalence ratio of approximately one and nitrogen and argon as diluents.By applying the new method,the effect of heat transfer on the ignition delays were eliminated successfully and ignition delays at adiabatic condition were determined.A detailed kinetic model of n-pentane was used to simulate the measured adiabatic ignition delay,which agreed well with the experimental data.