This manuscript presents the design and initial application of a high-pressure combustion chamber (HPCC). The HPCC exhibits several unique design attributes to enable high-fidelity studies of propellant-combustion physics at high pressures. The HPCC employs a flangeless and weldless design to provide a compact, easy to access, and relatively light weight (for its size and pressure capability) test chamber. It has a cylindrical test volume of 13.1 L and is capable of operating at pressures from approximately 0.4 mbar to 200 bar. The vessel is equipped with a ZnSe window to enable the laser ignition of propellants and energetic materials and 4 sapphire windows (2″ diameter and 4″ × 2″ slots) to enable the use of multiple optical diagnostics spanning the ultraviolet to mid-infrared. The sapphire windows are mounted in plugs with adjustable length to bring the windows inside of the test volume and facilitate line-of-sight optical measurements. The vessel can be accessed from the top and bottom via removable 5″ diameter plugs, and the bottom plug can be modified to enable studies of gaseous jets and flames. Some of the HPCC’s testing capabilities are demonstrated via high-speed IR imaging and laser-absorption-spectroscopy measurements of temperature and CO in laser-ignited HMX (i.e., 1,3,5,7-tetranitro-1,3,5,7-tetrazoctane) flames at pressures from 2 to 25 bar.

Combustion in practical energy and propulsion systems such as rockets, gas turbines, or internal combustion engines is performed at elevated pressures in order to increase thermodynamic efficiency. As a result, controlled studies of combustion at high pressures are needed to understand the complex combustion physics governing these systems. The harsh conditions and small spatial and time scales associated with combustion often render traditional intrusive measurement tools such as thermocouples or gas sampling techniques (e.g., gas chromatography mass spectrometry) incapable of providing the necessary data. For these reasons, optical diagnostics ranging from visual imaging to advanced laser-spectroscopy techniques (e.g., absorption, fluorescence, and Raman scattering) are typically used to characterize the combustion process via time-resolved temperature and species measurements.1–4 Unfortunately, deploying optical diagnostics in high-pressure combustion environments is challenging due to a variety of spectroscopic challenges and the thermomechanical difficulties associated with achieving large optical access in such environments.

Despite these challenges, laboratory studies of combustion physics have been performed in a wide range of high-pressure facilities, namely, shock tubes,3 rapid compression machines,5,6 and pressure chambers.7–19 The short test times associated with the former typically limit or preclude the study of flame physics; thus, this work will focus on the latter. One of the earliest pressure chambers designed for the combustion research was built in 1928 for the study of the effect of pressure on diffusion flame height.7 More recently, Carter et al.8 constructed a chamber from standard piping and bolt circle flanges to reduce cost. This chamber is constructed of 316 stainless steel, and is capable of operating at pressures up to 60 atm with four orthogonally oriented windows with a maximum clear aperture of 63.5 mm. The windows are mounted between flanges on a short section of a pipe that is welded to the exterior of the chamber. The chamber was built to contain a flat flame burner on an x-y translation stage for studying the spatial evolution of chemical species in laminar and turbulent burner flames.

Soot formation in flat flames (e.g., McKenna burner flames) was studied using another chamber constructed from stainless steel in a similar manner to that of Carter et al.,8 but capable of operating at pressures up to 110 atm with a 24 cm inner diameter, 60 cm height, and 27.1 L volume. However, this chamber weighs 700 kg and heavy machinery is required to move the large flanges to gain access to the chamber.9–11 McCrain and Roberts12 also studied soot formation in a Burke–Schumann laminar diffusion flame. The stainless steel chamber could sustain a pressure of 35 atm with an internal diameter of 20 cm, a height of 100 cm, and a volume of 31.4 L. Large optical access was provided via viewports with a diameter of 76 mm which were also held externally via flanges and welded pipe. The chamber walls are water cooled such that water condenses on the walls instead of on the windows.12,13

Most recently, Joo et al.14 constructed a chamber with 4 orthogonally oriented windows with a clear aperture of 90 mm. The cylindrical test section of the chamber is capable of operating at pressures up to 35.5 atm, and it has a volume of 25 L, a 25.4 cm inner diameter, and a height of 50 cm. Ignition can be achieved via hot-wire ignition or glow plugs. The burner in this chamber is interchangeable and is built into the lower flange of the chamber. The test volume of this chamber can be accessed without heavy machinery, and a standard engine hoist can be used to easily move it to a different lab space.

Strand burners (used to measure burning rates of solid and liquid propellants)16–19 typically utilize a single screw plug to gain entry to the chamber and use slot-shaped windows to provide high visibility in the vertical direction. While optically accessible chambers employing this design can operate at higher peak operating pressures (e.g., 344 bar,18 365 bar,19 and 3000 bar17), this has come at the cost of employing smaller test volumes (typically ≈1 l) and less optical access (compared to the design presented here) to provide lower stress levels and facilitate operation at such high pressures.15 Unfortunately, the use of smaller internal volumes leads to a larger pressure rise during a combustion test and also encourages product gas recirculation, both of which can alter flame physics and the accuracy of measurements acquired using path-integrated diagnostics (e.g., imaging and laser-absorption spectroscopy). Later in this manuscript, we demonstrate that the HPCC does not suffer from these drawbacks. Furthermore, many common burners (e.g., McKenna and Hencken) are too large to physically fit inside such chambers, thereby limiting the range of combustion experiments that can be conducted in such chambers.

In this manuscript, the design and application of a High-Pressure Combustion Chamber (HPCC) are discussed in detail. The primary novelty of the HPCC lies in its unique combination of (1) large test volume (13 l), (2) large optical access (e.g., 50.8 mm clear aperture round windows and 190.5 mm × 38.1 mm slot windows), (3) high peak operating pressure (up to 200 bar), and (4) ease of access and entry. To the best of our knowledge, the HPCC can provide more optical access (quantified in the total clear aperture) at pressures up to 200 bar than any combustion vessel that has been reported in the open literature to date. The remainder of this manuscript is focused on describing the design process in detail along with an explanation of design decisions and challenges. The manuscript concludes with presenting a demonstration of some of the HPCC’s testing capabilities. Specifically, high-speed IR images of laser-ignited HMX (i.e., 1,3,5,7-tetranitro-1,3,5,7-tetrazoctane) and mid-infrared laser-absorption-spectroscopy measurements of path-integrated flame temperature and CO column density are presented at elevated pressures.

The HPCC and laser-ignition system are shown in Fig. 1. The external height of the HPCC is 610 mm (24 in.), and it has an outer diameter of 356 mm (14 in.). The pressurized test volume is roughly cylindrical in shape with a height of 406 mm (16 in.) and a diameter of 203.2 mm (8 in.), yielding a test volume of approximately 13.1 l. The vessel is constructed out of 17-4 PH stainless steel to provide resistance to corrosive propellant-combustion products (e.g., HCl) while maintaining high strength. Stress calculations were performed using Solidworks’ finite-element solver and indicate that the HPCC can withstand pressures up to 200 bar with a safety factor of 2. The HPCC was hydrostatically proof tested (without windows) at a pressure up to 140 bar (limited only by the water pump). The body of the vessel has large openings for threaded plugs [128.9 mm (5.07″) diameter] on the top and bottom of the vessel to provide access for setting up experiments and cleaning the test volume. The bottom plug has a tapered thread port to enable hot-wire ignition. Gas delivery, exhaust, pressure sensors, and wire feedthroughs all connect to the HPCC body through tapered thread ports.

FIG. 1.

Photograph of the HPCC and experimental setup used to study laser-ignited HMX flames with IR imaging and laser-absorption spectroscopy.

FIG. 1.

Photograph of the HPCC and experimental setup used to study laser-ignited HMX flames with IR imaging and laser-absorption spectroscopy.

Close modal

Optical access is achieved via two round sapphire windows [clear aperture of 50.8 mm (2 in.)] and 2 sapphire slot windows [clear aperture of 190.5 mm (7.5 in.) by 38.1 mm (1.5 in.)] that are mounted in window plugs located on opposite sides of the chamber. All windows have a 1° wedge on their inner face to prevent the formation of etalons (i.e., optical fringes) in scanned-wavelength laser-absorption experiments. The flat face of each window is sealed to the shoulder of its window plug using EPO-TEK T7109 epoxy. If necessary, all windows can be extended inside the chamber to bring the windows directly adjacent to the flame and thereby mitigate the potential influence of line-of-sight non-uniformities on laser-absorption measurements. A ZnSe window with a 7.3 mm (0.5 in.) clear aperture is mounted in the HPCC’s top plug to enable laser-ignition of propellants using a CO2 laser emitting 90 W of optical power near 10.6 μm.

Optical diagnostics are widely used to characterize combustion science and systems due to their non-intrusive nature and ability to provide high-speed (kHz to MHz rate) measurements of gas properties (e.g., temperature and composition).1,3,20 Two of the most widely used laser diagnostics in combustion are laser-absorption spectroscopy (LAS) and planar laser-induced fluorescence (PLIF). This section reports the design steps that were taken to enable high-fidelity LAS, PLIF, and imaging measurements of combustion in the HPCC.

The use of LIF and other imaging diagnostics is enhanced by using large windows to maximize the field of view and light collection. Furthermore, LIF is typically performed by imaging orthogonal to the laser beam. To facilitate the use of these diagnostics, the HPCC was designed with four orthogonally oriented windows (see Figs. 2 and 3), all with a clear aperture of at least 38 mm (1.5 in.).

FIG. 2.

CAD model of the HPCC showing the exterior (a) and two radial cross sections [(b) and (c)]. (1) Top plug, (2) ZnSe laser ignition window, (3) O-ring seals, (4) slot window holder extended into test volume, (5) slot window holder flush with the inner wall, (6) gas inlet, (7) bottom plug, (8) round window holder flush with the inner wall, and (9) round window holder extended into test volume.

FIG. 2.

CAD model of the HPCC showing the exterior (a) and two radial cross sections [(b) and (c)]. (1) Top plug, (2) ZnSe laser ignition window, (3) O-ring seals, (4) slot window holder extended into test volume, (5) slot window holder flush with the inner wall, (6) gas inlet, (7) bottom plug, (8) round window holder flush with the inner wall, and (9) round window holder extended into test volume.

Close modal
FIG. 3.

Transverse cross section of the HPCC CAD model. (1) Chamber body, (2) bottom plug, (3) round window mount, (4) round window holder, (5) slot window mount, (6) slot window holder, (7) tapered thread port, (8) hot wire ignition port, (9) extended round window holder, (10) sapphire windows, and (11) extended slot window holder.

FIG. 3.

Transverse cross section of the HPCC CAD model. (1) Chamber body, (2) bottom plug, (3) round window mount, (4) round window holder, (5) slot window mount, (6) slot window holder, (7) tapered thread port, (8) hot wire ignition port, (9) extended round window holder, (10) sapphire windows, and (11) extended slot window holder.

Close modal

The accuracy of LIF, LAS, and IR imaging measurements can be reduced by line-of-sight non-uniformities and optical trapping, for example, in product gases located outside the flame. This can be a concern when conducting measurements within a closed chamber where recirculation zones can transport combustion products away from the flame and then back into the measurement path but now at a drastically different temperature. As a result, it is desirable to minimize the distance over which the emission or laser light must travel outside the primary region of interest (i.e., the flame). To mitigate this risk, the HPCC employs a large test volume and was designed to enable the windows to protrude inside the HPCC. A similar approach was taken by Smith et al.21 and Girard et al.22 who used sapphire rods to direct laser light to and from flat flames in a low-pressure chamber. However, since this configuration is not desirable for all types of experiments, the window design is modular to enable many different window locations to be employed.

Both the round and slot shaped window assemblies employ a window holder and window mount (see Fig. 3). The window mount is bolted to the outside of the chamber via a bolt circle and sealed to the HPCC body with a face seal O-ring. The window holder is fixed to the inner side of the window mount using bolts for the slot window and an ISO 11926-1 port for the round window and sealed using face seal and ISO 11926-1 port O-rings, respectively. The sapphire windows are sealed directly to the window holders using a thermally removable epoxy (EPO-TEK T7109). Sealing the windows using epoxy instead of clamping the window between gaskets reduces the risk of breaking windows during installation and testing at high pressures. This method also allows wedged windows to be used to prevent etalons (i.e., optical fringes) from forming in scanned-wavelength laser-absorption measurements. The window holders can be made to different lengths allowing the windows to protrude inside the chamber, thereby adjusting the optical path length through the test volume. The window holders are relatively small and inexpensive to make, facilitating the use of different window sizes and locations. To the best of our knowledge, this modular window design is a unique attribute of the HPCC compared to other combustion chambers published in the literature.

FEA (finite element analysis) simulations of the initial design (performed using Solidworks) of the window holders and mounts indicated the presence of large stress concentrations. The geometry of the window holders, window mounts, and chamber body was iterated until a design with a minimum safety factor of 2 was achieved at 200 bar. Furthermore, FEA simulations of each component attached to the HPCC were performed to ensure that part-chamber interactions do not adversely affect the strength of each part or the HPCC. For the round windows, the required window thickness was determined using the following equation:

(1)

Here, Th (in.) is the window thickness, P (psi) is the chamber pressure, R (in.) is the unsupported radius, SF is the desired safety factor, and MR (psi) is the modulus of rupture of the window material. FEA simulations of windows mounted in the window holder were performed for the wedged slot windows to obtain a more accurate result due to their more complex geometry. The minimum acceptable SF of the windows was chosen to be 4 due to the increased risk of fracturing window materials. Both wedged round and wedged slot windows have been tested successfully in the chamber. The round and slot windows have a maximum thicknesses of 5 mm and 7.6 mm and SF = 4 pressure ratings of 46.2 and 68.9 bar, respectively. To tolerate the maximum HPCC chamber pressure of 200 bar, the thickness of the round and slot sapphire windows would need to be 10.6 mm and 13 mm, respectively. During operation at elevated pressures, all personnel wear eye protection and polycarbonate shields are placed around each of the windows to protect personnel in the event of a window failure.

The HPCC was designed to be suitable for studies of propellant flames and burner flames at both low and high pressures. The chamber is capable of operating at pressures as low as 0.4 mbar using a standard vacuum roughing pump and at pressures up to 200 bar. The former enables studies of low-pressure flat flames (e.g., for studying flame kinetics), while the latter enables propellant burn rate measurements to be performed at pressures relevant to rocket motors.

Securing safe operation at pressures as high as 200 bar while maintaining large optical access and a large internal volume was extremely challenging. This is because the large internal volume and large holes providing optical access in the HPCC body lead to large stress concentrations. To withstand the high stress, the stainless steel alloy 17-4 PH (ISO 15156, EN/DIN 1.4542 X5CrNiCuNb16.4, annealed condition A) was selected. This is because it has almost 3.5 times the yield strength of the more commonly used 316 or 304 stainless steel, yet it is still easily available, machinable, and relatively inexpensive. The high strength of this material enabled a minimum SF of 2 at 200 bar while also keeping the body to a reasonable size and weight (360 kg) which enables the HPCC to be relocated using a typical engine hoist.

Despite the high strength of the material used, the high stress created by the large test volume and high pressures required careful design of the chamber geometry. The most severe stress concentrations were found near the top inner edge of the window ports made for the slot windows. To reduce stress, the walls of the chamber were thickened to 76.2 mm (3 in.) enabling them to support extreme loads despite the presence of large window ports. If the window ports were removed, this wall thickness would enable the chamber to withstand a pressure of 1216 bar with a safety factor of 2.1. Somewhat counterintuitively, stress was further reduced by widening the slot window ports. While this decreased the amount of steel available for support, it greatly increased the area where loads were previously concentrated near the top of the slots, thus reducing the stress. Stress and safety factors were determined using the Solidworks simulation’s FEA package as mentioned previously.

Face seal O-rings were used wherever possible to provide low leak rates and to relax machining tolerances. For the slot window ports, the O-ring seat was recessed into the outside of the chamber such that there would be a smaller area for pressure to act laterally upon the sides of the slot and, in effect, pull it apart (see Fig. 3). Recessing the seat too far into the side of the chamber reduced the area on the top and bottom of the slots and increased stress. An optimal distance for recessing the seal was found by performing multiple FEA simulations and interpolating to the optimal seat depth. This design provides a leak rate of 6.5 and 0.46 μbar/min (at vacuum) to be achieved with windows and without windows (using solid window blanks), respectively.

Gases are delivered and removed from the HPCC via a high-pressure gas manifold shown in Fig. 4. The manifold is connected to a vacuum pump to evacuate the chamber. The manifold also contains needle valves for controlling the flow of gases into and out of the HPCC and an adjustable relief valve to prevent over-pressurizing the test volume. The manifold can be connected to three gas-supply lines at once. The chamber can be vented directly into a fume hood to bring the chamber to atmospheric pressure and/or flush corrosive gases out of the chamber into an exhaust hood. The chamber can also be pumped down to vacuum through the gas manifold. The pressure in the manifold is measured using capacitance manometers with a range of 0–1000 Torr and 0–10 000 Torr (both MKS Baratron 627F series). A 0–70 bar pressure transducer (Unik 5000 PDCR5011-TA-A3-CB-H0-PF-70BARA, 0.07 ms time response) is mounted directly to the HPCC to record the chamber pressure during a test.

FIG. 4.

Plumbing and instrumentation diagram of the HPCC and gas manifold.

FIG. 4.

Plumbing and instrumentation diagram of the HPCC and gas manifold.

Close modal

Threaded plugs, similar to the Crawford bomb design,15,16 are used to gain entry to the HPCC from the top and bottom (see Fig. 2). These plugs utilize a piston-seal O-ring for sealing and weigh only 13 kg. This design was chosen over a more commonly used flange-based design11–13 to facilitate chamber access. For example, flanges that are suitable for this design and operating pressure weigh 39 kg each, which would require assistance from machinery to install and remove, thereby severely complicating access to the chamber. Furthermore, this would increase the weight of the HPCC by 53% and would require a stronger and heavier support system.

A total of 7 NPT ports (2 × 1/4″, 4 × 3/8″, 1 × 1/2″) at various locations around the chamber provide feedthrough access to the test volume. These ports can all support various types of feedthroughs which can be used for gas delivery, wires for sensors and hot-wire ignition, and power for control of actuators.

A custom table was designed to support the HPCC and enable the removal of the bottom plug. When fully assembled, the HPCC weights 360 kg (793.4 lbs) and an additional 50 kg of aluminum breadboards surround the HPCC to mount optics and the other test equipment. Furthermore, a 177.8 mm (7″) diameter hole in the table is required to comfortably remove the bottom plug. To support this weight, the table top was made from a 1524 mm × 914 mm × 19 mm (60 in. × 36 in. × 0.75 in.) slab of aluminum 6061-T6. The aluminum slab is supported by a ThorLabs non-isolating 800 mm support frame with a load capacity of 700 kg.

A portion of the HPCC’s testing capabilities were demonstrated by performing laser-absorption-spectroscopy and IR imaging measurements of laser-ignited HMX flames at elevated pressures.

Figures 1 and 5 illustrate the experimental setup used to characterize laser-ignited flames of HMX with laser-absorption spectroscopy and infrared imaging. A continuous-wave (CW) CO2 laser (Coherent C70A) emitting up to 90 W of power near 10.6 μm was used to ignite pellets of pressed HMX (6 mm diameter, 3–5 mm tall). The laser beam was directed to the pressure vessel using a series of uncoated gold mirrors (R = 98.82% at 10.6 μm), and the beam path was enclosed using aluminum beam tubes. An anti-reflection-coated ZnSe window was used as a beamsplitter to combine a visible (532 nm) alignment beam (produced by a 5 mW CW diode laser) onto the CO2 laser’s beam path. A 25.4 mm diameter, AR-coated ZnSe lens with a focal length of 1000 mm was used to reduce the diameter of the CO2 laser’s beam to 6 mm at the location of the HMX pellet which was located approximately 3.5 m from the CO2 laser. Either air or N2 was used as the bath gas.

FIG. 5.

Cross-sectional view of the HPCC model illustrating the location of the CO2-laser beam and QCL beam used for laser-ignition of HMX pellets and laser-absorption-spectroscopy measurements of flame temperature and CO column density, respectively.

FIG. 5.

Cross-sectional view of the HPCC model illustrating the location of the CO2-laser beam and QCL beam used for laser-ignition of HMX pellets and laser-absorption-spectroscopy measurements of flame temperature and CO column density, respectively.

Close modal

A CW quantum-cascade laser (QCL) emitting 30 mW of optical power near 4.8 μm was used to acquire measurements of gas temperature and CO column density in HMX flames using scanned-wavelength direct absorption (scanned-DA) or scanned-wavelength-modulation spectroscopy with first-harmonic-normalized second-harmonic detection (scanned-WMS-2f/1f). The laser beam was directed through the HPCC at a location corresponding to 4 mm or 8 mm above the initial surface of the HMX pellet. The laser beam was collected by a 25.4 mm diameter, AR-coated CaF2 lens with a focal length of 40 mm and was focused onto a thermoelectrically cooled, mercury-cadmium telluride (MCT) detector (Vigo Systems PVI-4TE-5) with a bandwidth of 10 MHz. A bandpass filter centered near 2060 cm−1 with a full-width at half-maximum of 40 cm−1 was used to reduce the infrared background emission.

In scanned-DA experiments, the wavelength of the QCL was scanned across the P(0, 20) and P(1, 14) absorption transitions near 2059.9 and 2060.3 cm−1, respectively, at 500 Hz via injection-current tuning. The laser current was scanned below its threshold current to enable the background emission to be recorded between each scan. The wavelength scanning was characterized using a solid germanium etalon with a free-spectral range near 0.0163 cm−1. A Voigt profile was least-squares fit to the measured spectra and the best-fit integrated absorbances were used to determine the path-integrated gas temperature and column density of CO (χCOL).23 

Scanned-WMS-2f/1f spectral fitting24 was used to measure the flame temperature and CO column density at high-pressures (10 bar) since the increased collisional broadening and beamsteering prevented in situ determination of the incident light intensity (Io) and reduced signal-to-noise ratio (SNR), thereby severely complicating scanned-DA measurements. A similar scanned-WMS technique was recently used to measure CO in a high-pressure rocket motor.30Figure 6 shows simulated absorbance spectra for the absorption-transitions-of-interest at pressures of 1, 2, and 10 bar. At 10 bar, the spectra are significantly blended and a non-absorbing wavelength does not exist within the wavelength-tuning range of the QCL. In scanned-WMS-2f/1f experiments, the wavelength of the QCL was sinusoidally scanned between the peaks of CO’s P(0, 20) and P(1, 14) absorption transitions at 500 Hz while simultaneously modulated at 100 kHz. The scan and modulation amplitudes were 0.26 and 0.12 cm−1, respectively. During post-processing, the raw detector signal was passed through digital lock-in filters with a bandwidth of 5 kHz to extract the measured scanned-WMS-2f/1f spectra. The measured spectra were then passed to the spectral-fitting routine developed by Goldenstein et al.24 to determine the integrated absorbance and collisional width of each transition. The gas temperature and column density of CO were then calculated using the best-fit integrated absorbances in the same manner as done in scanned-DA measurements.

FIG. 6.

Simulated absorbance spectra of CO near 2059.5 cm−1 at various pressures for a gas temperature of 2750 K, a CO mole fraction of 0.20, and a path length of 1 cm. Absorbance spectra were calculated using the HITEMP2010 database.25 

FIG. 6.

Simulated absorbance spectra of CO near 2059.5 cm−1 at various pressures for a gas temperature of 2750 K, a CO mole fraction of 0.20, and a path length of 1 cm. Absorbance spectra were calculated using the HITEMP2010 database.25 

Close modal

High-speed mid-infrared imaging was performed simultaneously orthogonal to the QCL’s beam path. Images were obtained using a Telops FAST-IR 2K camera which employs a cryogenically cooled InSb focal plane array with 256 × 320 pixels. Images were recorded at 2000 frames-per-second (FPS) with an integration time of 7.5 μs. A bandpass filter (Spectrogon NB-3300-066 nm) centered near 3300 nm with a FWHM of 66 nm was used to pass emission from primarily carbonaceous species (e.g., HMX, HCN, and formaldehyde), although some emission from H2O’s ν3 fundamental vibrational band is also detected in this spectral window.

Figure 7 illustrates select IR images acquired during laser-ignition of HMX in N2 at 2 bar. The first image (a) illustrates the IR emission originating from the boiling surface of HMX and the warm HMX vapor coming off the surface. The second image (b) illustrates the formation of an ignition kernel located above the surface, and the third image (c) illustrates the flame front propagating downward to the boiling surface. The last image (d) illustrates the flame structure as it burns in a quasi-steady state. Figure 8 illustrates the structure of HMX flames at 2, 10, and 25 bar. At all pressures, the flames exhibit structural characteristic of laminar premixed flames and the flames propagate linearly down the pellet with little edge burning observed. At 2 bar, a reaction front is visible approximately 2 mm above the surface. Surface boiling and miniature surface eruptions are observed throughout the burn period. At 10 and 25 bar, the reaction front is too close to the surface to observe with the optics employed here. Furthermore, the IR image of the flame at 25 bar indicates significant narrowing of the flame core, which may result from alterations in the local flame temperature or concentration of various species emitting in this spectral window. That said, care must be taken in interpreting these images as the intensity of the infrared emission observed exhibits complex dependencies on gas temperature, pressure, and composition.26 The combination of these results illustrates the ability of IR imaging (unlike visible imaging) to spatially and temporally resolve the entire laser-ignition process, from surface heating through flame holding.

FIG. 7.

Infrared images of an HMX pellet (outlined by white box) undergoing laser ignition in an N2 bath at 2 bar within the HPCC. Plume of HMX vapor prior to ignition (a), ignition kernel (b), flame front propagating towards surface (c), and quasi-steady flame (d).

FIG. 7.

Infrared images of an HMX pellet (outlined by white box) undergoing laser ignition in an N2 bath at 2 bar within the HPCC. Plume of HMX vapor prior to ignition (a), ignition kernel (b), flame front propagating towards surface (c), and quasi-steady flame (d).

Close modal
FIG. 8.

IR images of the quasi-steady HMX flame in N2 at 2 bar (a), 10 bar (b), and 25 bar (c). The images illustrate how increasing the bath gas pressure brings the primary reaction front closer to the surface and narrows the flame.

FIG. 8.

IR images of the quasi-steady HMX flame in N2 at 2 bar (a), 10 bar (b), and 25 bar (c). The images illustrate how increasing the bath gas pressure brings the primary reaction front closer to the surface and narrows the flame.

Close modal

Figure 9(a) shows a representative single-scan measurement of CO’s absorbance spectra near 2060 cm−1 for a line-of-sight directed through the HMX flame approximately 8 mm above the initial pellet surface at a pressure of 2 bar. The best-fit Voigt profiles were determined using the integrated area, transition linecenter, and collisional width of each transition as free parameters. The Doppler width was initially fixed to the value corresponding to an initial guess temperature and then updated according to the measured temperature (obtained from the two-color ratio of integrated areas) obtained from the previous iteration. The incident light intensity was inferred from fitting a 3rd-order polynomial to the non-absorbing regions of the transmitted intensity time history. Background emission was measured in between scans and subtracted from the measured light intensity prior to calculating the absorbance using Beer’s law.

FIG. 9.

Scanned-DA results acquired in HMX flames in air at 2 bar. Measured and best-fit absorbance spectra of CO’s P(0, 20) and P(1, 14) transitions (a), measured temperature (b) and CO column density (c) time histories. All measurements were acquired at 500 Hz using a line-of-sight located approximately 8 mm above the initial surface location.

FIG. 9.

Scanned-DA results acquired in HMX flames in air at 2 bar. Measured and best-fit absorbance spectra of CO’s P(0, 20) and P(1, 14) transitions (a), measured temperature (b) and CO column density (c) time histories. All measurements were acquired at 500 Hz using a line-of-sight located approximately 8 mm above the initial surface location.

Close modal

Figures 9(b) and 9(c) show time histories of the measured temperature and CO column density (χCOL), respectively, for a laser-ignited HMX flame in air at 2 bar. After ignition, the flame temperature rises up to near 2750 K which agrees well with radiation-corrected thermocouple measurements acquired at 1 bar using a similar laser fluence for ignition and surface heating.27 Approximately 300 ms after ignition, the path-integrated flame temperature begins to oscillate at approximately 21 Hz with an amplitude of ±80 K. The 21 Hz oscillation is clearly noticeable in visible and infrared imaging of the HMX flame, and it is well known that HMX flames exhibit natural oscillations.27,28 However, to the best of our knowledge, the magnitude of the temperature oscillation has not been quantified previously. After accounting for the actual temperature oscillation, the 1-σ precision of the temperature measurement is ±10 K and efforts to rigorously quantify the measurement accuracy are undergoing. The CO column density oscillates in phase with the temperature oscillation. This is intuitive given that CO is a primary combustion product of HMX. In addition, a larger CO column density corresponds to a thicker flame which could encourage greater heat release and help insulate the flame core.

Figure 10(a) illustrates a portion of the scanned-WMS-2f/1f signal time history and best-fit signals acquired in laser-ignited HMX flames in N2 at 10 bar. The best-fit scanned-WMS-2f/1f spectra typically agree within 2% of the peak signal, thereby supporting the accuracy of the measurements. The best-fit spectra for a given up-scan and down-scan were obtained simultaneously to improve the accuracy of the best-fit spectroscopic parameters. This was beneficial here since the wavelength of the QCL could not be scanned across the entirety of the pertinent absorption lines due to a combination of (1) the limited wavelength-tuning capabilities of the QCL and (2) the larger collisional broadening encountered at 10 bar. In addition, to further improve the accuracy of the spectral-fitting routine, the collisional width of both absorption transitions was forced to be equal. This is justified by theory and experimental results. For example, the collisional-broadening parameters provided by the HITEMP2010 database25 indicate that the collisional widths for these transitions should agree within 0.5% for a mixture of 20% CO in air at 2750 K. Furthermore, the best-fit collisional widths measured here in HMX flames at 2 bar using scanned-DA typically agree within 1%.

FIG. 10.

Scanned-WMS-2f/1f results acquired in HMX flames in N2 at 10 bar. Measured and best-fit scanned-WMS-2f/1f spectra for four consecutive scans across CO’s P(0, 20) and P(1, 14) transitions (a), measured temperature (b) and CO column density (c) time histories. All measurements were acquired at 500 Hz using a line-of-sight located approximately 4 mm above the initial surface location.

FIG. 10.

Scanned-WMS-2f/1f results acquired in HMX flames in N2 at 10 bar. Measured and best-fit scanned-WMS-2f/1f spectra for four consecutive scans across CO’s P(0, 20) and P(1, 14) transitions (a), measured temperature (b) and CO column density (c) time histories. All measurements were acquired at 500 Hz using a line-of-sight located approximately 4 mm above the initial surface location.

Close modal

Figures 10(b) and 10(c) illustrate scanned-WMS-2f/1f measurements of temperature and CO column density, respectively, acquired in HMX flames in N2 at 10 bar. Approximately 200 ms after ignition, the flame temperature reaches a quasi-steady value of ≈2775 K and oscillations at a distinct frequency were not detected. The scan-to-scan measurement precision is typically within 15 K, thereby highlighting the precision of this diagnostic despite the pronounced beamsteering encountered at higher pressures. Unlike the data acquired at 2 bar, the column density of CO measured at 10 bar reaches a quasi-steady plateau and oscillations at a distinct frequency were not detected. This results from a combination of effects, namely, (1) reduced oscillation in the flame structure and thickness and (2) reduced edge burning which caused the flame thickness to grow significantly during tests in air. Both of these effects result from the higher gas pressure and the absence of oxygen in the bath gas. For this test, visible and IR imaging indicates that the flame thickness was approximately equal to the pellet diameter [see Fig. 8(b)]. Assuming a uniform line-of-sight, the quasi-steady column density of ≈0.13 cm would correspond to a CO mole fraction of ≈0.22 which is consistent with what is expected in the far-field of HMX flames.27,29

In each experiment, the pressure increased by approximately 3.9% during the burn period, which indicates that a constant-pressure assumption is appropriate for these experiments. In addition, window extensions were not used to reduce the optical path to include only the flame since there was no evidence of combustion products being transported back into the optical line-of-sight (e.g., due to recirculation) during the test duration. For example, immediately after the flame extinguished, CO could not be detected in the chamber via the laser-absorption sensor for at least 5 s (data was acquired for more than 5 s). This results from a combination of the large test volume of the HPCC and the relatively small HMX pellets burned here.

The authors thank Purdue University and the Air Force Office of Scientific Research (Grant No. FA9550-18-1-0210) with Dr. Mitat Birkan as program manager for supporting the development and testing of the HPCC, as well as the Defense Threat Reduction Agency (Grant No. HDTRA1-17-1-0023) with Dr. D. Allen Dalton as program manager for supporting the development of the WMS temperature and CO diagnostic. We also thank Monique McClain, Morgan Ruesch, and Professor Steven Son for insightful discussions regarding the design of the HPCC.

1.
C. S.
Goldenstein
,
R. M.
Spearrin
,
J. B.
Jeffries
, and
R. K.
Hanson
, “
Infrared laser-absorption sensing for combustion gases
,”
Prog. Energy Combust. Sci.
60
,
132
176
(
2017
).
2.
K.
Kohse-Höinghaus
,
R. S.
Barlow
,
M.
Aldén
, and
J.
Wolfrum
, “
Combustion at the focus: Laser diagnostics and control
,”
Proc. Combust. Inst.
30
,
89
123
(
2005
).
3.
R. K.
Hanson
and
D. F.
Davidson
, “
Recent advances in laser absorption and shock tube methods for studies of combustion chemistry
,”
Prog. Energy Combust. Sci.
44
,
103
114
(
2014
).
4.
T.
Dreier
,
R.
Chrystie
,
T.
Endres
, and
C.
Schulz
, “
Laser-based combustion diagnostics
,” in
Encyclopedia of Analytical Chemistry
(
Wiley Online Library
,
2016
), pp.
1
44
.
5.
M. T.
Donovan
,
X.
He
,
B. T.
Zigler
,
T. R.
Palmer
,
M. S.
Wooldridge
, and
A.
Atreya
, “
Demonstration of a free-piston rapid compression facility for the study of high temperature combustion phenomena
,”
Combust. Flame
137
,
351
365
(
2004
).
6.
E. F.
Nasir
and
A.
Farooq
, “
Time-resolved temperature measurements in a rapid compression machine using quantum cascade laser absorption in the intrapulse mode
,”
Proc. Combust. Inst.
36
,
4453
4460
(
2017
).
7.
S.
Burke
and
T.
Schumann
, “
Diffusion flames
,”
Ind. Eng. Chem.
20
,
998
1004
(
1928
).
8.
C. D.
Carter
,
G. B.
King
, and
N. M.
Laurendeau
, “
A combustion facility for high-pressure flame studies by spectroscopic methods
,”
Rev. Sci. Instrum.
60
,
2606
2609
(
1989
).
9.
K.
Thomson
,
O.
Gulder
,
E.
Weckman
,
R.
Fraser
, and
D.
Snelling
, “
A new diffusion flame burner and pressure vessel for high pressure soot formation study
,” in
Canadian Section, Spring Technical Meetings
(
Combustion Institute
,
Windsor, ON
,
2002
).
10.
H. I.
Joo
and
Ö. L.
Gülder
, “
Soot formation and temperature structure in small methane-oxygen diffusion flames at subcritical and supercritical pressures
,”
Combust. Flame
157
,
1194
1201
(
2010
).
11.
A. M.
Vargas
and
Ö. L.
Gülder
, “
A multi-probe thermophoretic soot sampling system for high-pressure diffusion flames
,”
Rev. Sci. Instrum.
87
,
055101
(
2016
).
12.
L. L.
McCrain
and
W. L.
Roberts
, “
Measurements of the soot volume field in laminar diffusion flames at elevated pressures
,”
Combust. Flame
140
,
60
69
(
2005
).
13.
Y.
Li
, “
Applications of transient grating spectroscopy to temperature and transport properties measurements in high-pressure environments
,” Ph.D. thesis,
North Carolina State University
,
2001
.
14.
P. H.
Joo
,
J.
Gao
,
Z.
Li
, and
M.
Aldén
, “
Experimental apparatus with full optical access for combustion experiments with laminar flames from a single circular nozzle at elevated pressures
,”
Rev. Sci. Instrum.
86
,
035115
(
2015
).
15.
J. W.
Taylor
, “
A melting stage in the burning of solid secondary explosives
,”
Combust. Flame
6
,
103
107
(
1962
).
16.
B. L.
Crawford
,
C.
Huggett
,
F.
Daniels
, and
R. E.
Wilfong
, “
Direct determination of burning rates of propellant powders
,”
Anal. Chem.
19
,
630
633
(
1947
).
17.
W.
McBratney
and
J.
Vanderhoff
, “
High pressure windowed chamber burn rate determination of liquid propellant XM46
,” Technical Report ARL-TR-442,
Army Research Laboratory
,
1994
.
18.
E.
Boyer
and
K.
Kuo
, “
High-pressure combustion behavior of nitromethane
,” AIAA Paper No. 99-2358,
1999
.
19.
R.
Carro
,
M.
Stephens
,
J.
Arvanetes
,
A.
Powell
,
E.
Petersen
, and
C.
Smith
, “
High-pressure testing of composite solid propellant mixtures: Burner facility characterization
,” AIAA Paper 2005-3617,
2005
.
20.
C.
Schulz
and
V.
Sick
, “
Tracer-LIF diagnostics: Quantitative measurement of fuel concentration, temperature and fuel/air ratio in practical combustion systems
,”
Prog. Energy Combust. Sci.
31
,
75
121
(
2005
).
21.
C. H.
Smith
,
C. S.
Goldenstein
, and
R. K.
Hanson
, “
A scanned-wavelength-modulation absorption-spectroscopy sensor for temperature and H2O in low-pressure flames
,”
Meas. Sci. Technol.
25
,
115501
(
2014
).
22.
J.
Girard
,
R.
Spearrin
,
C.
Goldenstein
, and
R.
Hanson
, “
Compact optical probe for flame temperature and carbon dioxide using interband cascade laser absorption near 4.2 μm
,”
Combust. Flame
178
,
158
167
(
2017
).
23.
R. M.
Spearrin
,
C. S.
Goldenstein
,
I. A.
Schultz
,
J. B.
Jeffries
, and
R. K.
Hanson
, “
Simultaneous sensing of temperature, CO, and CO2 in a scramjet combustor using quantum cascade laser absorption spectroscopy
,”
Appl. Phys. B
117
,
689
698
(
2014
).
24.
C. S.
Goldenstein
,
C. L.
Strand
,
I. A.
Schultz
,
K.
Sun
,
J. B.
Jeffries
, and
R. K.
Hanson
, “
Fitting of calibration-free scanned-wavelength-modulation spectroscopy spectra for determination of gas properties and absorption lineshapes
,”
Appl. Opt.
53
,
356
367
(
2014
).
25.
L. S.
Rothman
,
I.
Gordon
,
R.
Barber
,
H.
Dothe
,
R. R.
Gamache
,
A.
Goldman
,
V.
Perevalov
,
S.
Tashkun
, and
J.
Tennyson
, “
HITEMP, the high-temperature molecular spectroscopic database
,”
J. Quant. Spectrosc. Radiat. Transfer
111
,
2139
2150
(
2010
).
26.
R. K.
Hanson
,
R. M.
Spearrin
, and
C. S.
Goldenstein
,
Spectroscopy and Optical Diagnostics for Gases
(
Springer International Publishing Switzerland
,
2016
).
27.
C.-J.
Tang
,
Y. J.
Lee
,
G.
Kudva
, and
T. A.
Litzinger
, “
A study of the gas-phase chemical structure during CO2 laser assisted combustion of HMX
,”
Combust. Flame
117
,
170
188
(
1999
).
28.
J. C.
Finlinson
,
T.
Parr
, and
D.
Hanson-Parr
, “
Laser recoil, plume emission, and flame height combustion response of HMX and RDX at atmospheric pressure
,”
Symp. (Int.) Combustion
25
,
1645
1650
(
1994
).
29.
M. W.
Beckstead
,
K.
Puduppakkam
,
P.
Thakre
, and
V.
Yang
, “
Modeling of combustion and ignition of solid-propellant ingredients
,”
Prog. Energy Combust. Sci.
33
,
497
551
(
2007
).
30.
D. D.
Lee
,
F. A.
Bendana
,
S. A.
Schumaker
, and
R. M.
Spearrin
, “
Wavelength modulation spectroscopy near 5 μm for carbon monoxide sensing in a high-pressure kerosene-fueled liquid rocket combustor
,”
Appl. Phys. B
124
,
1
10
(
2018
).