Combustion of N-Decane+air Mixtures To Investigate Laminar Burning Velocity Measurements At Elevated Temperatures

ABSTRACT The combustion characteristics of n-decane+air mixtures are experimentally investigated through laminar burning velocity measurements at 1 atm pressure and higher initial temperatures using an externally heated diverging channel (EHDC) method. Up to 610 K mixture temperature over an equivalence ratio range of 0.7–1.4, laminar burning velocities are reported with an accuracy of ± 5%. The current measurements exhibit a good match with existing experimental measurements, and agree closely with the predictions of Zhao, LLNL and PoliMi mechanisms at different mixture temperatures. The present measurements show an excellent match of temperature exponent (α) variation with equivalence ratio (ϕ) with the predictions of distinct kinetic models as well as experimental measurements. This study reveals that a substantial scatter exists among the predictions of different kinetic models. A variation of 20–30 cm/s in the burning velocity is observed at 610 K mixture temperature. Reaction R16 (H + HO2 = H2 + O2), which inherently reduces the burning velocity becomes insignificant at an elevated mixture temperature of 610 K, and the reaction R15 (H + HO2 = 2OH) plays a dominant role in accelerating the flame propagation. From reaction pathway diagrams, it is clear that a higher burning velocity at 610 K is associated with the increased reaction rate. The elemental-flux value associated with the formation of C2H3 from C2H4 at 610 K mixture temperature is ≅34% higher in comparison to the 470 K mixture temperature.


Introduction
The development of the future internal-combustion and aircraft engines is bound to strictly adhere to the international regulatory standards of minimal emission levels aided with an efficient performance.This is directly linked to the combustion phenomenon at the core of the engine.The combustion process associated with the fuel injected into the engine is quite perplexing, since these commercial fuels are composed of hundreds of components, thus, making the kinetic model development a challenging task for detailed engine simulations.The kinetic model development is simplified through the selection of appropriate surrogate mixtures.Surrogate mixtures are formulated using simpler hydrocarbons to emulate the physical and chemical properties of the chosen fuel.The validation of the developed kinetic model using surrogate mixtures is performed against the experimental results of laminar burning velocity, ignition delay times, species profiles, and extinction limits.
Various research works have been carried out to measure the mixture burning velocity of n-decane+air mixtures as a monocomponent surrogate for both aviation and transport fuels.(Skjoth-Rasmussen et al. 2003) determined the laminar burning velocity of n-decane +air flames employing the Bunsen burner method with a linear-extrapolation approach at 473 K mixture temperature.At 500 K mixture temperature and 1 atm pressure, (Zhao et al. 2004) determined the laminar burning velocities of n-decane+air mixtures using the stagnation jet-wall flame arrangement.Unstretched laminar burning velocities were obtained using the linear-extrapolation method.(K.Kumar and Sung 2007), measured the laminar burning velocities of n-decane/air mixtures using the counterflow (CF) twinflame configuration at 1 atm pressure conditions.For initial temperatures between 360 K and 470 K, the unstretched burning velocity is determined by employing the linearextrapolation approach.(Chunsheng et al. 2010) performed experiments to determine the laminar burning velocities of n-decane+air flames employing the CF configuration.The flow velocities were measured using the Laser Doppler Velocimetry.A non-linear extrapolation approach was employed to acquire the unstretched laminar burning velocities at pressure and mixture temperature of 1 atm and 403 K respectively.(Singh, Nishiie, and Qiao 2011) used the spherically expanding flame (SEF) method to measure the laminar burning velocities and markstein lengths of n-decane+air mixtures.The unstretched laminar burning velocities were then obtained using the linear and nonlinear extrapolation approaches with a maximum variation of 2 cm/s between them.Using the same method, (Moghaddas, Eisazadeh-Far, and Metghalchi 2012) reported the burning velocity measurements of n-decane+air flames for 0.7 ≤ ϕ ≤ 1 range.The measurements were reported for mixture temperatures between 350 K and 610 K and pressures between 0.5 atm and 8 atm.(Kim et al. 2013) measured laminar burning velocities of n-decane+air mixtures using SEF for pressures of 0.7-5 atm with initial temperature, T u = 400 K.The model predictions were consistent with the measured values only at 1 and 2 atm.At lean conditions, the reported values were lower than the measured values using the CF method and the values for rich mixtures were observed to lie between the two sets of CF measurements (linear and non-linear extrapolation).For initial temperatures ranging from 350 K to 470 K with 1 to 3 atm pressures, (Hui and Sung 2013) reported the laminar burning velocities of n-decane along with several fuel+air mixtures for 0.7 ≤ ϕ ≤ 1.3 using the CF setup.
(Munzar et al. 2013) investigated the premixed stagnation flames and reported the laminar burning velocities of n-decane+air flames.At 400 K mixture temperature and 1 atm pressure, the measured values were in fine agreement with the model predictions for equivalence ratios between 0.7 and 1.2.Overpredictions (underpredictions) were observed for various kinetic models, when compared with the measured values for lean (rich) mixtures.(Comandini, Dubois, and Chaumeix 2015) performed experiments to report the burning velocities at 403 K mixture temperature of n-decane, n-butylbenzene, and n-propylcyclohexane, both single and multi-component mixtures using the SEF facility at 1 bar pressure.For initial mixture temperatures of 338 K and 358 K at atmospheric pressure, the burning velocities of n-decane and binary kerosene surrogate mixture were reported by (Alekseev et al. 2017).The measurements derived from the heat-flux (HF) arrangement exhibited a good match with the predictions of the PoliMi detailed mechanism (Ranzi et al. 2014).The measured values agreed (disagreed) with the measurements using CF (SEF) methods.For the operating conditions of 400 K temperature, 0.1 MPa pressure and equivalence ratios between 0.6 and 1.3, the laminar burning velocities of n-decane+air mixtures were measured by (Wu et al. 2018) using a Bunsen burner and they reported an excellent agreement with other experimental dataset and the predictions of JeTSurF 2.0 (Wang et al. 2010).
A recent investigation by (Le Dortz et al. 2021) measured the burning velocity of n-decane at 400 and 425 K mixture temperatures using the SEF method.(Zhu et al. 2021) measured the laminar burning velocities of n-decane/air using a Bunsen burner.The measurements were performed at 1 atm pressure and 405 K mixture temperature for an equivalence ratio range of 0.6-1.4.The burning velocity data in conjunction with ignition delay measurements and chemical kinetics were used lately to investigate the combustion properties of n-decane coupled combustion (Zhang et al. 2021).In the pursuit of developing new aviation surrogate fuels of RP-3 kerosene, few investigations have been performed lately.The laminar burning velocity of a new five-component surrogate fuel comprising n-decane (14% molar basis) in its composition was measured using a SEF bomb (J.Liu et al. 2022).The measurements were in agreement with the RP-3 kerosene (J.Liu et al. 2022).In a different study (X.Liu et al. 2022), a three-component (n-decane, 1,3,5-trimethylbenzene and iso-dodecane) surrogate was formulated and, the validity of the developed skeletal mechanism of RP-3 kerosene was tested using the laminar burning velocity values and the concentration of measured species (X.Liu et al. 2022).A brief summary of the existing work on the n-decane+air mixture is presented in Table 1.
For a jet fuel surrogate formulation, n-decane has been widely used as a singlecomponent surrogate.In a recent laminar burning velocity measurement of n-decane (C 10 H 22 )+air flames for initial temperatures up to 650 K, (Xie et al. 2022) employed the EHDC method.The measurements have been reported with an accuracy of ±6% without any description of the α (temperature exponent) variation with ϕ (equivalence ratio).The current work presents the, a) laminar burning velocity measurements of n-decane+air mixtures are reported and comprehensively compared with existing experimental investigations and kinetic model predictions for an initial temperature range (338-610 K), b) α variation with ϕ to highlight its influence on the current as well as existing measurements, and c) impact of crucial reactions on the burning velocity with mixture temperature is discussed using sensitivity analysis and reaction pathway diagrams.

Experimental method
The understanding of flame dynamics in externally heated diverging channel was the preliminary focus in refs.(Khandelwal and Kumar 2010; S. Kumar 2011).The investigations of (Akram, Minaev, and Kumar 2013) concluded that, for a premixed gaseous fuel-air mixture, a stabilized planar flame is formed in the diverging channel for certain conditions of wall temperature gradient and inlet velocity, which can be used to determine the laminar burning velocity of the mixture at elevated mixture temperature conditions.This technique was appropriately tailored by (Katoch et al. 2016) for measuring the laminar burning velocity of liquid fuel+air mixtures.The burning velocities of different liquid fuels using an externally heated diverging channel (EHDC) method have been provided in the previous investigations (Katoch et al. 2016(Katoch et al. , 2019;;R. Katoch, Millán-Merino, and Kumar 2018;Kumar and Kumar 2021;Kumar et al. 2018;R. Kumar et al. 2020;R. Kumar, Kishore Velamati, and Kumar 2021;Kumar, Singhal, and Kumar 2021) and a simplified diagram of the experimental setup is shown in Figure 1.
A quartz-made diverging channel with a 15 • divergence with inlet dimensions of 25 × 2 mm (aspect ratio (Λ) = 12.5) is used in the present experimental investigations.Air from the compressor gets heated, when it is passed through an air preheater.This heated air then mixes with a metered quantity of n-decane (coming from an infusion pump) at junction "J" as shown in Figure 1.The temperature of the flow line of 2 mm diameter (red-coloured) is set above the boiling point of n-decane fuel (preferably above 10 K) to ensure complete vaporization of the injected liquid fuel.The fuel (n-decane) is injected into this flow line using an injection needle of 0.5 mm diameter.This arrangement ensures complete atomization and vaporization of the fuel to form a homogenous fuel-air mixture (Katoch et al. 2016).This premixed fuel-air mixture then passes through the diverging channel, wherein an ignition is arranged at the channel exit.A flame is stabilized in the channel post-ignition depending on the mixture velocity.The heater placed below the channel helps establish nearly adiabatic conditions for the formation of a planar flame to nullify the heat-loss from the planar flame to the diverging channel walls and then to the surroundings.Additionally, the heater also helps in providing a linear positive temperature gradient along the flow direction.At the stabilized planar flame location, the laminar burning velocity (S u ) calculation utilizes the following relation: A in , T in and U in refer to the area, temperature and mixture velocity at the channel inlet.A f and T f refer to the area and temperature at the flame-stabilized location.The heating rate of the heater, equivalence ratio (ϕ) and mixture velocity govern the position of stabilized flame.A detailed analysis of different uncertainties associated with the experimental investigation showed that the measured S u is accurate to ± 5%.For a propane-air mixture, an investigation (Akram and Kumar 2012) on the effect of heat loss (Botha and Brian Spalding 1954) using this method was found to reduce burning velocity by 2-4% of the adiabatic burning velocity (Akram and Kumar 2012;Katoch et al. 2016), and is included in the calculation of the uncertainties in the present measurements.The hydrodynamic strain rate in the present experiments (40-60 s −1 ) corresponds to uncertainty, of < 5% (Konnov et al. 2018).The effect of strain rate reflects on the area of planar flame (which is included in final uncertainty calculations).More details on the estimation of the uncertainties of the present setup can be found in ref. (R. Kumar et al. 2020) (the approach is briefly discussed in the section S1 of the supplementary material).The planar flames were stabilized for inlet velocities (U in ) between 0.25 m/s and 1.35 m/s.Negatively (positively) stretched flames were obtained for mixture velocities lower (higher) than this mixture velocity range.The measured burning velocities were then compared to the existing data reported in the  Mechanism developed by (Zhao et al. 2004) for n-decane was based on the modifications in the mechanism of (Zeppieri, Klotz, and Dryer 2000) in terms of the data of thermochemistry, sub-mechanisms and elementary rate parameters.Apart from Zhao's mechanism (Zhao et al. 2004), other mechanisms used in the present work are summarized in Table 2.

Computational methodology
Governing equations in CANTERA (v 2.5.1 and above) (Goodwin, Moffat, and Speth 2009) are solved using either a non-stoichiometric method or a stoichiometric method.In a nonstoichiometric method, the mass conservation equation is treated separately whereas the stoichiometric method is more robust, but slower.CANTERA tries a non-stoichiometric method first, and turns to a stoichiometric method if the non-stoichiometric fails to converge.Various parameters such as ratio, slope and curve were set to 3, 0.05, and 0.12 to ensure grid convergence.A diagram tool, Graphviz (Ellson et al. 2001) was included in the python (Sanner 1999) script of CANTERA (Goodwin, Moffat, and Speth 2009) to obtain the reaction pathway diagrams.For better representation, a threshold value of elementalflux was carefully provided in the CANTERA (Goodwin, Moffat, and Speth 2009) script.

Measurements and predictions of variations of laminar burning velocity with temperature ratio
Figure 3 shows the variation of the laminar burning velocity with temperature ratio (T u /T u,0 ) at different equivalence ratios.Here, T u is the mixture temperature and T u,0 refers to atmospheric or reference temperature.In this measurement, T u,0 is the reference temperature, taken as 300 K.The detailed mechanism predictions are represented using lines and the present measurements using scatter points.The short dashed pink coloured line shows the power-law fit of the present measurements (solid sphere).At ϕ = 1, the measurements are consistent and in excellent agreement with the predictions of (Zhao et al. 2004), LLNL (Sarathy et al. 2011) and PoliMi (Ranzi et al. 2014) up to 610 K mixture temperature (T u /T u,0 ≈2.03).Predictions of JeTSurF 2.0 (Wang et al. 2010) match well for T u /T u,0 ≥ 1.45 with the current measurements and this model predicts lower burning velocities for mixture temperatures < 435 K. Predicted values using the Aachen surrogate model (Honnet et al. 2009) are lowest

Temperature exponent variation with mixture equivalence ratio
For the experimental investigations performed at atmospheric pressure conditions, and mixture temperatures up to 610 K, the temperature exponent (α) variation over an equivalence ratio (ϕ) ranging from 0.7 to 1.4 is shown in Figure 4.The associated uncertainty in the temperature exponent values is well within ±4% (calculated using the least-square model (Alekseev, Christensen, and Konnov 2015)).Only the present values exhibit a variation similar to that obtained using different kinetic model predictions, wherein the minimum value is observed at ϕ ≃ 1.1 and these temperature exponent values increase as the mixtures turn rich or lean.At different mixture conditions, present values of temperature exponent are in agreement with the predictions of (Zhao et al. 2004) and

Laminar burning velocity measurements at different mixture temperatures
Figure 5 displays the comparison of measured and predicted burning velocities at 338 and 470 K mixture temperatures.At 338 K, the present measurements show a good match with the predictions of the kinetic models of PoliMi (Ranzi et al. 2014) and (Zhao et al. 2004) at all mixture conditions.For ϕ ≤ 1.2, the measured values are also consistent with the measurements of (K.Kumar and Sung 2007) and (Alekseev et al. 2017), and slightly lower for ϕ > 1.2.Measurements of (Moghaddas, Eisazadeh-Far, and Metghalchi 2012) and, the predictions of the Aachen surrogate (Honnet et al. 2009) and JeTSurF 2.0 (Wang et al. 2010) are relatively lower than the current measurements at all equivalence ratios.The predictions of LLNL (Sarathy et al. 2011) match well only for rich mixtures (ϕ ≥ 1.2) and are relatively lower for ϕ < 1.2 in comparison to the current measurements up to 500 K mixture temperature.
At 470 K, the present measurements display a fairly good agreement with the various experimental measurements except for (Skjoth-Rasmussen et al. 2003).Predictions of the Aachen surrogate kinetic model (Honnet et al. 2009) are relatively lower than the other measured values and kinetic predictions.Excellent compliance is exhibited by the current measurements to the predictions of (Zhao et al. 2004) and PoliMi (Ranzi et al. 2014).Predictions of JeTSurF 2.0 (Wang et al. 2010) agree only for lean-to-stoichiometric mixtures.Predictions are lower than the present measured values for ϕ ≥ 1.Excluding the significantly higher velocities of (Skjoth-Rasmussen et al. 2003), a high scatter is observed for leaner mixtures and this scatter significantly reduces for rich mixtures.
The measured laminar burning velocities in this study show a similar variation to other measurements and kinetic model predictions with equivalence ratios at 360 K (Figure 6) and 338 K (Figure 5).Out of all other measurements reported in the literature, the current measurements show fine agreement with the predictions of (Zhao et al. 2004) and PoliMi Figure 6.Variation of the measured burning velocity measurements at 360 K mixture temperature.(Ranzi et al. 2014).Recent measurements of (Alekseev et al. 2017) on heat-flux arrangement also match well with the predictions of (Zhao et al. 2004) and PoliMi (Ranzi et al. 2014) for the equivalence ratio range, ϕ = 0.7-1.3 at 338 and 360 K mixture temperatures.
A significant amount of data on laminar burning velocities measurements is available at a mixture temperature of 400 K Figure 7. Initial measurements of (K.Kumar and Sung 2007) and (Singh, Nishiie, and Qiao 2011) depict a considerable scatter, especially at ϕ = 1.1, as high as 10 cm/s Figure 7a.However, over the years, numerous efforts have been made to reach a consensus and this aspect is quite evident in Figure 7a.All the predictions, except for the Aachen surrogate (Honnet et al. 2009), exhibit good consistency with the measured values at various mixture conditions.Predictions of the Aachen surrogate (Honnet et al. 2009) kinetic model match relatively well with various measurements for rich mixture conditions (ϕ > 1.2).The current measurements illustrate good compliance with the predictions of (Zhao et al. 2004) and PoliMi (Ranzi et al. 2014) kinetic models.The recent measurements of (Wu et al. 2018) at 400 K using the Bunsen burner method report lower velocities in comparison to the predictions of (Zhao et al. 2004) and PoliMi (Ranzi et al. 2014).However, a better match is observed with both the predictions at 400 K and for the measurements reported at 423 K mixture temperature.Figure 7b shows the comparison of laminar burning velocity measurements using the EHDC method with the predictions of various kinetic models.The measurements of (Xie et al. 2022) are slightly lower than the current measurements for lean mixtures.For rich mixtures, both measurements are quite consistent.Overall, the scatter among the predictions of kinetic models is higher at lean conditions in comparison to the rich mixtures.
The burning velocities at higher mixture temperatures (higher than the auto-ignition temperature of n-decane with reasoning discussed in (R. Kumar et al. 2020)) are shown in Figures 8 and 9.At 500 K Figure 8, the measurements of (Zhao et al. 2004) using the counterflow method comply well with the predictions of the PoliMi (Ranzi et al. 2014) kinetic model.Present measurements show a good consistency with the measurements of (Zhao et al. 2004) and the predictions of PoliMi (Ranzi et al. 2014) up to ϕ = 1.2.Lower velocities are reported for rich mixtures (ϕ ≥1.2).At all mixture conditions, an excellent agreement is evident between the present values and the predictions of (Zhao et al. 2004) kinetic model (Figures 8 and 9).Interestingly, the predictions of LLNL (Sarathy et al. 2011) show great consistency with the current measurements at mixture temperatures >600 K Figure 9.
Figure 9 (a) compares the current measurements with the predictions of distinct kinetic models.It is clear from the figure that the measurements obtained from this study are in good agreements with the predictions of (Zhao et al. 2004), LLNL (Sarathy et al. 2011) and PoliMi (Ranzi et al. 2014).(b) compares the present measurements with the measurements of (Xie et al. 2022) at 650 K using the EHDC method.The present values are extrapolated using the powerlaw correlation Figure 3 for obtaining the laminar burning velocity at 650 K. Overall, a good match is seen between both the measured values.It is worth noticing that the predictions of the  kinetic models of (Zhao et al. 2004), LLNL (Sarathy et al. 2011) and PoliMi (Ranzi et al. 2014) fall within the uncertainty range of both the measurements using EDHC method.

Sensitivity analysis
Figures 10 and 11 show the sensitivity coefficient variation of key reactions affecting the laminar burning velocity of n-decane+air flames at 470 K and 610 K initial temperatures respectively.The PoliMi (Ranzi et al. 2014) kinetic model is employed to analyze the sensitivity of important reactions due to its closeness with the present measured values, which is evident from Figures 4-9.In contrast to PoliMi (Ranzi et al. 2014) kinetic model, the negative sensitivity coefficient values of reactions H+HCO=H 2 +CO and H+HO 2 =H 2 +O help explain the lower predictions of the LLNL (Sarathy et al. 2011) model (Fig. S1 of the supplementary material).Further, reactions OH+CO=H+CO 2 and H+O 2 (+M)=HO 2 (+M) show that the magnitude of the sensitivity coefficients is slightly higher for LLNL (Sarathy et al. 2011) kinetic model (Fig. S1).Reaction R5 (H+O 2 =O+OH) exhibits the highest positive sensitivity coefficient of all the reactions both at 470 K and 610 K mixture temperatures (Figures 10 and 11).This sensitivity variation tends to increase as the mixture turns richer.In Figure 10, the highest and least contribution of reaction R5 is observed for ϕ = 1.4 and ϕ = 0.7 respectively.A similar trend is noticed with the rise in mixture temperature to 610 K (Figure 11).At lean condition (ϕ = 0.7), the reaction R5 competes with an important chain terminating reaction R22 (H+O 2 (+M)=HO 2 (+M)) (Dayma et al. 2012) which results in a lower sensitivity of reaction R5 in comparison to the sensitivity at ϕ = 1.4 (Figures 10 and 11).As the mixture turns rich, the negative sensitivity of reaction R22 decreases or becomes zero (Figures 10 and 11).For any combustion process, reaction R5 is regarded as the vital chain branching step (Turányi 1997).Further, this reaction exhibits high sensitivities, leading to being termed as a rate-limiting step as it causes a substantial change in the overall reaction rate (Nowak and Warnatz 1987;Ray 1983).
The reaction between H and HO 2 (hydroperoxy radical) radicals (Figures 10 and 11) proceeds via two routes: R16 (H+HO 2 =H 2 +O 2 ) and R15 (H+HO 2 = 2OH).The reaction R16 retards the burning velocity, whereas the reaction R15 enhances the burning velocity as shown in Figure 10 at 470 K.With the rise in mixture temperature to 610 K, the reaction R16 no longer affects the burning velocity and the overall effect results in accelerating the flame due to reaction R15 (Figure 11).Interestingly, this effect is observed only in lean condition (ϕ = 0.7).The formation of stable species such as H 2 O and CH 4 via reactions R6 (H+OH+M=H 2 O+M) and R31 (H+CH 3 (+M)=CH 4 +(+M)) respectively exhibit negative sensitivity.The production of H radical from reactions R24 (OH+CO=H+CO 2 ) and R130 (HCO+M=H+CO+M) participates in increasing the burning velocity.

Reaction pathway diagram
(Figures 12 and 13) show the reduced reaction pathway diagrams of n-decane + air mixtures (for ϕ = 1) at mixture temperatures of 470 K and 610 K respectively.The PoliMi (Ranzi et al. 2014) kinetic model is used for the representation owing to its closeness with the present measurements at different conditions (Figures 4 and 9) The thickness of the arrowed line depicts its significance based on the elemental-flux (EF) value (number shown to the right of the pathways).The details of the EF are described by Turányi and Tomlin (Turányi and Tomlin 2014).The oxidation of n-decane proceeds with the formation of decyl radical (n-C 10 H 21 ) and HO 2 :n-C 10 H 22 + O 2 = HO 2 + n-C 10 H 21 (R11659).
The detailed reaction pathway diagram (including reaction R11659) at an initial temperature of 470 K is illustrated in the supplementary material (section S2), wherein the reactor temperature (T reactor ) and threshold elemental-flux (EF t ) are set to 1250 K and 0.008 kmol/m 3 -s respectively.For the same mixture temperature, when the reactor temperature is increased to 1550 K, CH 4 decomposes to form CH 3 as shown in the supplementary material (section S2).For a high-temperature system, reaction R32 (CH 4 +H=CH 3 +H 2 ) becomes important (Sutherland, M-C, and Michael 2001).A close scrutiny of the detailed path-flux analysis at 1550 K (section S2) shows that the forward reaction (formation of CH 3 radical) of R32 is more favoured.This is also prominent for a higher T reactor of 1850 K (as shown in Figures 12 and 13 For the clear depiction of reaction pathways with less ambiguity, these diagrams (Figures 12 and 13) are constrained by the T reactor and EF t of 1850 K and 0.02 kmol/m 3 -s respectively and begin with CH 4 decomposition (since the reactor temperature is higher than 1550 K).The python script to obtain these pathway diagrams are similar and reported in refs.(R. Kumar, Kishore Velamati, and Kumar 2021;Kumar, Singhal, and Kumar 2021).The inherent nature of both the active radicals, ethylene (C 2 H 4 ) and vinyl radical (C 2 H 3 ) increases the overall reaction rate (Davis and Law 1998).A higher reaction rate at 610 K is associated with a higher EF value for the formation of C 2 H 3 from C 2 H 4 .The EF value at 610 K is higher by ≅34% in comparison to the value at 470 K.This results in higher reactivity and ultimately higher burning velocity at 610 K.

Conclusions
The current work measures the laminar burning velocity of n-decane+air flames and presents the data at 1 atmospheric pressure up to 610 K mixture temperatures over an equivalence ratio range of 0.7-1.4.The experiments were performed using an externally heated diverging channel (EHDC) method.The crucial findings of this work are summarized as: (1) The laminar burning velocities of n-decane+air flames are reported up to 610 K mixture temperature over an equivalence ratio range of 0.7-1.4 for the first time with an accuracy of ± 5%.Prior to this work, the measurements of n-decane + air flames were limited only to 470 K mixture temperature.
(2) The current measurements exhibit a good agreement with the predictions of (up to 610 K) (Zhao et al. 2004), LLNL (Sarathy et al. 2011) and PoliMi (Ranzi et al. 2014) mechanisms at different mixture temperatures, and also with the other experimental measurements.
(3) The present measurements show an excellent match between the temperature exponent (α) variation with equivalence ratio and predictions from various kinetic models.(4) Interestingly, with an increase in mixture temperature, a significant disparity is observed among the predictions of different mechanisms.For instance, at a mixture temperature of 610 K, a variation of 20-30 cm/s in burning velocity is observed over an equivalence ratio range of 0.7-1.4.(5) Reaction R16 (H + HO 2 = H 2 + O 2 ) forms stable species, lowering the burning velocity at 470 K mixture temperature.With the rise in initial temperature to 610 K, reaction R16 no longer becomes significant, and reaction R15 (H + HO 2 = 2OH) becomes dominant in accelerating the flame.(6) The elemental-flux value associated with the formation of C 2 H 3 from C 2 H 4 at a mixture temperature of 610 K is higher by ≅34% when compared to the value at 470 K. Thus resulting in higher values of reaction rate and burning velocity at 610 K.

Figure 1 .
Figure 1.Experimental specifics of an externally heated diverging channel (EHDC) method (G: the gap between the diverging channel and heater and O: the overlap distance between the heater and quartz channel).
literature and predictions of different reaction mechanisms.A direct image of a n-decane +air planar flame is shown in Figure2.
among all the predictions and measurements.Similar observations for the lean mixture (ϕ = 0.7) are shown in Figure 3b.At ϕ = 1.4, the lowest values are observed for the predictions of the Aachen surrogate (Honnet et al. 2009) and JeTSurF 2.0 (Wang et al. 2010) Figure 3c.Considerable scatter tends to prevail for mixture temperatures >540 K (T u /T u,0 = 1.8) among the predictions of different reaction mechanisms.This was absent for lean and stoichiometric conditions.

Figure 3 .
Figure 3. Measurements and predictions variations of laminar burning velocity at different equivalence ratios.

Figure 4 .
Figure 4. Temperature exponent variation with equivalence ratio of the experimental measurements and their comparison with various kinetic model predictions.

Figure 5 .
Figure 5. Variation of laminar burning velocity measurements at initial temperatures of 338 and 470 K.

Figure 7 .
Figure 7. Variation of the measured burning velocity at 400 K mixture temperature.

Figure 8 .
Figure 8. Variation of the burning velocity at 500 K initial temperature.

Figure 9 .
Figure 9. Variation of the laminar burning velocity measurements for initial temperatures of (a)610 K, and (b) 650 K

Figure 12 .
Figure 12.Reduced reaction pathway diagram of stoichiometric n-decane + air mixture at 470 K.

Figure 13 .
Figure 13.Reduced reaction pathway diagram of stoichiometric n-decane + air mixture at 610 K.

Table 1 .
Details of laminar burning velocity measurements of n-decane + air flames.

Table 2 .
List of mechanisms of n-decane + air mixtures.