Heat Flux Measurement in Shock Heated Combustible Gases and Clarification of Ignition Delay Time

Abstract

Correct understanding of the ignition and combustion processes in the combustion chambers are critical for modeling advanced schemes of engines of high-speed aircraft and promising spacecraft. Moreover, experimental data on the ignition delay time are a universal basis for the development and testing of combustion kinetic models. Moreover, the higher the temperature of the fuel mixture, the smaller this time value and the more important its correct determination. The use of a thermoelectric detector allows to measure ignition delay times and record heat fluxes with a high time resolution (to tenths of μs) during ignition in propane–air mixtures. Due to the faster response time, the use of it allows refining the ignition delay time of the combustible mixture, and the detector itself can serve as a useful device that allows a more detailed study of the ignition processes.

1. Introduction

The correct description of combustible mixtures ignition plays an important role in modeling detonation processes in the combustion chambers of jet engines [1]. This is especially true for the high-temperature region (of the order of 2000 K and above), where the ignition delay time is about a microsecond. Traditional measurement methods (piezoelectric and optical) do not allow accurate recording of such small values of the ignition delay time due to their inertness [2]. In this regard, the search for new, more accurate methods for measuring this quantity is relevant.

Shock tubes are a usable tool for studying various high-temperature gas dynamic processes [3]. A large number of shock tubes are involved for measuring the ignition delay time in high-temperature combustible mixtures [4,5]. The purpose of this work is registration of heat fluxes in the process of the combustible mixture ignition behind a reflected shock wave and improving the accuracy of the ignition delay time measurement at very short timescales. The authors failed to find works describing the use of heat flux detectors to study the ignition of hydrocarbon fuels in shock tubes. This is due to the fact that the use of such devices is difficult, since due to high temperatures and pressures, most of these detectors fail and stop to function. In addition, the data obtained in such aggressive conditions are very difficult to interpret due to the fact that the environmental parameters are far from the operating range of the normal functioning of the detectors. Temperature and heat flux values are obtained in such experiments indirectly. Usually, they are calculated using various computer programs. At the same time, the recently developed thermoelectric detector demonstrated good data in experiments with a reflected shock wave of low and high intensities [6,7]. This work is devoted to its application on measuring heat flux and the ignition delay time in combustible mixtures in a shock tube.

2. Experimental Setup

The inner diameter of the shock tube sections is 50 mm; the lengths of the driver and driven tubes are 1.0 and 3.7 m, respectively (Figure 1). A copper diaphragm D with calibrated notches was installed between the chambers. By varying the thickness of the diaphragm, the depth of the notches, and the pressure in the driver and driven tubes, it is possible to achieve the necessary conditions behind the reflected shock wave. The facility is equipped with systems for pumping out and preparing and filling gas mixtures. The preparation system is used to prepare the test mixture consisting of propane and air (21% O₂/79% N₂). The prepared mixture is fed to the driven tube through the filling system. Helium is used as a driver gas. The preliminary pumping out of the shock tube sections was carried out to a residual pressure of 10⁻³ Torr. The leakage of the sections did not exceed 10⁻⁴ Torr/min.

The walls of the driven tube (side and end) are made of stainless steel. P1, P2, and P3 are PCB 113B24 pressure sensors; OF1 and OF2 are optical waveguides, which were tuned to certain radiation wavelengths for recording of OH* emission; TD is thermoelectric detector. When the diaphragm separating the shock tube sections breaks, an incident shock wave is formed in the test gas, the velocity of which is determined directly in front of the endwall. For these purposes, P1 and P2 are used that located at a distance of 50 mm from each other. P1, P2, and OF2 were located on the side cylindrical wall of the shock tube, P3, OF1, and TD were located on the endwall (Figure 1b). All device surfaces were flush mounted. During the experiment, the Agilent 54624A oscilloscope recorded the DT, H2, F2, and F1 data in the range of 10–100 MHz per channel.

TD is based on anisotropic film radiation detectors made of high-temperature materials, which are obtained using vacuum technology at an inclined condensation angle. The design of the device is a heat-conducting substrate made of high-resistance silicon, on which contact pads are applied for thermo-emf receiving. A thermosensitive anisotropic film based on Cr (thickness ~0.3 μm) is formed on top. Such films make it possible to obtain a sufficiently high sensitivity to instantaneous thermal action. The sensors were calibrated using lamp calibration, laser heating, and reflected shock tube experiments. A lamp calibration was performed to evaluate the overall sensitivity of the fabricated sensor without obtaining a calibration factor. Laser heating of the sensor was carried out using a laser diode at a power in the range of 5–30 W at 970 nm and made it possible to determine the calibration dependence. Shock tube experiments allowed to refine the obtained coefficient under conditions close to real laboratory ones. The discrepancy between the characteristics of the sensor obtained with different procedures was no more than 5% in absolute scale. The sensitivity obtained for TD used was $2.4\times {10}^{-8}\mathrm{V}·{\mathrm{m}}^2/\mathrm{W}$. More information about experimental setup, calibration procedures, and uncertainty analysis can be found in [6,7,8].

The parameters of the experiments are presented in

Table 1. Experimental parameters.

# Exp.	Composition	ϕ	P0, kPa	VSW, m/s	P5, MPa	T5, K
1	2.1% C3H8/20.56% O2/77.34% N2	0.5	20.02	1289	2.12	1644
2	4.2% C3H8/20.12% O2/75.68% N2	1	17	1302	1.85	1670

. ϕ is the equivalence ratio of mixture, P₀ is the driven tube initial pressure in the driven section, V_SW (incident shock wave velocity). P₅ and T₅ (pressure and temperature behind the reflected shock) were calculated using the GASEQ program [9].

3. Results and Discussion

Figure 2 and Figure 3 show heat flux data from the thermoelectric detector along with the P3 pressure and OH* emission from OF1 and OF2. The conversion of TD sensor readings from mV to MW/m² was carried out according to the linear factor defined in previous section. Pressure coefficient from manual was applied for P3 data conversion to atm. Under the considered parameters, the thermoelectric detector records well the history of heat flux changes.

TD heat flux readings have sharper edges. Thus, the rise of the primary disturbance is 0.1–0.5 μs, depending on the amplitude. The first response time of P3 is 1 μs or more. Moreover, the nature of TD data indicates low inertia of the sensor. On the scale of fractions of a microsecond, a thermoelectric detector is capable of registering abrupt changes in the thermal gradient. It should be noted that with such short time scales large heat flux values are available for registration—more than 4 MW/m². In addition, the signal-to-noise ratio is at a high level, which makes it possible to recognize all the features that occur when the mixture is ignited and after. Taking into account the inertia of the sensors, the nature of pressure and heat flux in Figure 2 and Figure 3 coincide quite well.

Chemiluminescence emission from excited particles such as OH* or CH*, as well as pressure, are convenient and effective diagnostic tools for monitoring the ignition delay time in mixtures heated by reflected shock. However, at short times (few microseconds), the precise determination of such delay can be difficult due to the insufficiently low inertia of the pressure sensor and the limited numerical aperture of the optical fiber. More details about this can be found in [5]. Data received allow clarifying the ignition delay time. They also show that ignition starts behind the reflected shock near the endwall (before the reflected shock passes OF2). At the time, as the reflected shock with ignited mixture behind passes OF2, it also shows OH* emission after 8 μs.

Table 2. Rise moments from sensors data.

	Exp. #1 (Figure 2), μs			Exp. #2 (Figure 3), μs
first rise	P3	OF1	TD	P3	OF1	TD
	0.2	4	0	0.1	4	0
second rise	N/D	N/D	2.1	N/D	N/D	2

demonstrates comparison of beginning of rises from sensor histories.

It can be said that despite the relatively low response time of the pressure sensor, the P3 has some kind of second rise (after 2.2 μs). However, the exact definition of its start becomes difficult due to constant fluctuations. However, it speaks for the validity of the TD data. In addition, the beginning of the first rise fronts perfectly coincides with the pressure sensor.

Due to the advantages of the thermoelectric detector, the difference between the data from P3, OF1, and TD exists, and it allows us to determine the values of the ignition delay time more accurately. In exp. #1, it is 2.05 μs vs. 4 μs (OF1). The same goes from data of exp. #2—2 μs vs. 4 μs. That is, the refinement of delay is more than doubled. The results agree with measurements by other authors in this temperature range [8,10,11] and with heat flux calculation studies in shock tubes and detonation engines under similar conditions [12,13,14].

As described earlier in the previous section, the calibration procedure for the thermoelectric detector for the heat flux behind the reflected shock was carried out in [6] for values less than 1 MW/m². In [7], under a similar technique, the sensor recorded values of more than 45 MW/m² in high-temperature xenon. However, the calibration characteristic can be nonlinear when going to high values, which were observed in [7] and in this work. When measuring the ignition delay time, the most important thing is the response speed of the sensor; obtaining absolute values does not play a big role. However, we can say that this device allows measuring the heat flux behind the reflected shock in a wide range of values. In addition, TD functions well in such aggressive environments (high pressures and significant heat loads).

4. Conclusions

In the experiments carried out, the thermoelectric detector was used for the first time to register the parameters of the ignition of combustible mixtures in a reflected shock. It was mounted simultaneously on a flange along with pressure transducer and optical fiber for OH* emission recording. The response time to a sharp thermal disturbance caused by the arrival of an incident shock or the ignition start is shorter in comparison to the high-frequency pressure sensor. It also successfully operates at high pressures (more than 20 atm behind the reflected shock). The nature of the device inertia makes it possible to record the history of the heat flux on a scale of hundreds of nanoseconds. The use of TD makes it possible to refine the value of the ignition delay time of the mixture in short time ranges (up to several microseconds). The advantages of the thermoelectric detector make it possible to use it for measuring such periods in various installations (shock tubes, models of various engines, etc.); there is no need to use additional devices, such as pressure sensors or optical recording systems.