David M. Lineberry

Experimental Cold Flow Characterization of a Swirl Coaxial Injector Element

A full-scale swirl coaxial injector element has been designed as part of an effort to examine liquid oxygen and liquid methane (LOX-LCH4) combustion instability phenomena under lunar ascent engine operating conditions. The scope of the LOX-LCH4 study encompassed cold flow, low pressure combustion, and high pressure combustion experimentation in an effort to establish a fast response analysis methodology for evaluating injector performance. As a baseline investigation, the spray characteristics of the injector’s central LOX post, exclusive of the exterior LCH4 annulus, were evaluated at constant ambient back pressure, across a range of steady mass flow rates. The effects of mass flow rate variance on the swirling sheet’s free cone spray angle and penetration length were assessed. Droplet velocity and diameter measurements were mapped within the spray’s primary and secondary atomization regions. The impact of throttling on the injector’s atomization quality was also surveyed.

Pulsator Designs for Liquid Rocket Injector Research

Experimental studies of injectors and other elements of Liquid Rocket Engine (LRE) behavior in the presence of strong pressure pulsations, typical for real conditions of operating an LRE, are problematic due to a lack of generators of strong periodical pulsation, particularly in high frequency range, which is typical for high frequency combustion instability (HFCI). In order to overcome these problems, devices for excitation of periodic pressure pulsation in gaseous and liquid media were evaluated, designed, built and studied. Two methods of generating pulsation were selected for further evaluation; a hydro-mechanical pulsator and an inertial pulsator. Both devices generate pulsation within the feed lines for injector elements. The hydro-mechanical pulsator has the advantage of not requiring a powerful drives and is able to produce strong periodic pulsations with amplitudes up to 70% of the mean pressure. However, it requires elevated expenses of pressurized model liquid, is comprised of rather complicated rotational machine parts, bearings and seals, has a multitude of rotating parts which must be dynamically balanced, and does not produce an ideal harmonic pressure signal. On the other hand, the inertial pulsator is capable of producing a sinusoidal pressure signal without complex machining requirements or excess expenditure of operational liquid. The primary disadvantage to the inertial pulsator is that it requires a large financial investment in its requirement for a powerful drive. CFD-modeling of the pulsating flow inside the hydro-mechanical pulsator showed the possibility to control the form of pulsation based on the geometry of the orifices of the rotating valve. An analytical model of the inertial pulsator was built to define its main parameters and determine optimum geometry. The hydro-mechanical pulsator showed to be more convenient to use in model cold flow tests with low to moderate pressure conditions while the inertial pulsator is more suitable for dynamic testing in high pressure conditions. The inertial pulsator can deal with fluids such as cryogenics or toxic liquids, but does not work for pressure excitation in gases. Experiments showed the possibility to define response functions of such LRE elements as injectors, valves, dampers and control units using the designed pulsation generators.

Propellant Throttling Effects on Self-pulsation of Liquid Rocket Swirl-coaxial Injection

An experimental assessment was conducted to identify the sensitivity of self-pulsation injection and spray characteristics to throttling of propellant simulants in a liquid rocket swirl-coaxial injector element. A set of test conditions were scaled to preserve flow similitude from a single case of spontaneous self-pulsation, wherein the inner LOX post was configured with no recess. The natural dynamics of pressure-swirl injection were examined to assess possible influence on the overall stability characteristics of self-pulsating swirlcoaxial injection. Pressure fluctuations in both the liquid and gas manifolds, as well as external acoustic signatures in the injector far-field, were compared with time-resolved pattern recognition of spray dynamics via backlit high speed cinematography. While discrete coupling in frequency was identified between the liquid manifold, acoustic radiance, and strong spray field fluctuations during self-pulsation, spectra in the liquid manifold was dominated by low frequency activity in all cases. Low frequency structures had slight, but persistent, influence in the spray field even when high frequency self-pulsation was encountered; however, these low frequency effects were not found to be explicitly coupled to manifold fluctuations. LOX post resonance is thought to have been excited by self-pulsation in several test cases.

Experimental Pulsator Characterization for Liquid Injector Research

The Propulsion Research Center at the University of Alabama in Huntsville (PRC) has designed and built a hydro-mechanical pulsator to simulate the effects of high frequency combustion instability (HFCI) in liquid rocket engines (LRE). The response characteristics (output pressure fluctuation, amplitude, and frequency) of the pulsator were evaluated in an atmospheric test rig with filtered de-ionized water as the primary fluid. The outlet of the pulsator was connected to a swirl injector LOX post to provide some downstream flow resistance. The pulsator control variables (primary flow bypass throttle, back pressure throttle, motor drive frequency and steady state injector pressure) were systematically varied to assess the dependence of the output flow characteristics on the control variables. For each test, the average mass flow rates of the waste water, seal water leakage, and fluid delivered through the injector were measured. Dynamic pressure was measured at the pulsator exit and the mean static pressure was measured at both the injector and at the pulsator inlets. High frequency pressure measurements show a periodic pressure pulsation with maximum peak to peak amplitude of 2% of the injector pressure when the pulsator is activated. The preliminary characterization of the pulsator was insufficient to fully characterize the pulsator; however, general trends showing the relationship between the control variables are visible in the data. The relationships between the variables show that higher order curve fits are required but the amount of data collected in this initial characterization is insufficient to generate those relationships.

Effects of Variable Chamber Pressure on Swirl Coaxial Injection: A Cold Flow Study

A geometrically full-scale swirl injector element was subjected to cold flow testing to explore the influence of steady chamber pressure on spray characteristics and selfatomization at conditions that reflect real liquid rocket combustor conditions. Propellant simulant flow rate was held constant at 0.82 kg/s over a chamber pressure range of 0.95 MPa to 3.45 MPa. Imaging studies observed free cone spray angle to deviate from the theoretical value when conditioned with positive variation in chamber pressure. Additionally, sheet penetration length decreased as a function of chamber pressure conditions. Phase Doppler Particle Analyzer measurements resolved droplet velocity and diameter profiles through the spray envelope. Analysis showed mean velocities and diameters in the primary atomization region to behave inversely to chamber pressure.

Experimental Investigation of Cavitation Induced Feedline Instability from an Orifice

This paper details the results of an experimental investigation into the cavitation instabilities created by a circular orifice conducted at the University of Alabama in Huntsville Propulsion Research Center. This experiment was conducted in concert with a computational simulation to serve as a reference point for the simulation. Testing was conducted using liquid nitrogen as a cryogenic propellant simulant. A 1.06 cm diameter thin orifice with a rounded inlet was tested in an approximately 1.25 kg/s flow with inlet pressures ranging from 504.1 kPa to 829.3 kPa. Pressure fluctuations generated by the orifice were measured using a high frequency pressure sensor located 0.64 tube diameters downstream of the orifice. Fast Fourier Transforms were performed on the high frequency data to determine the instability frequency. Shedding resulted in a primary frequency with a cavitation related subharmonic frequency. For this experiment, the cavitation instability ranged from 153 Hz to 275 Hz. Additionally, the strength of the cavitation occurred as a function of cavitation number. At lower cavitation numbers, the strength of the cavitation instability ranged from 2.4 % to 7 % of the inlet pressure. However, at higher cavitation numbers, the strength of the cavitation instability ranged from 0.6 % to 1 % of the inlet pressure.

Near-field Film Thickness Measurements of an LPRE Swirl Injector Spray

A novel approach to the measure of liquid film thickness in the near-field of a liquid rocket swirl injector spray was developed. Archived data were analyzed, wherein liquid mass flow was throttled over a range of 0.064 to 0.082 kg/s—corresponding to injection pressure drops of 0.80 to 1.26 MPa. For these conditions, digital images of the swirling liquid sheet were deconvoluted with an Abel transform technique and processed with an objective film thickness measurement algorithm. Measurements were benchmarked against theoretical values of the injection film thickness, as well as empirical predictions of film thickness at the injection plane and slightly upstream within the injector nozzle. Results suggest this approach successfully discerns film thickness; however, uncertainties associated with the measurements limit definitive physical insight on film thickness behavior as a function of the particular test conditions presented in this work.

A Mechanistic Assessment of Swirl Injection and Atomization by X-ray Radiographic and Optical Techniques

Detailed x-ray radiographic and optical experiments were conducted to evaluate injection and atomization characteristics of a liquid rocket swirl injector. Sprays issued from a single swirl-coaxial element—with, and without, co-annular gas flow—were exposed to focused, monochromatic x-rays. Two dimensional time-averaged attenuation data were collected, revealing projected mass distribution of liquid in the injector near-field, as well as mass-weighted axial velocities integrated over time. Measurements of liquid film thickness and spray cone angle were also inferred from the x-ray radiography data. Intra-set comparisons of the x-ray data revealed the mechanistic influence of gaseous co-flow on the liquid sheet. Similar measurements were gathered from visible light imagery and compared to those resolved from x-ray data. Contrasts between these data sets shed light on the sensitivities of the image processing techniques implemented in the study.

Scaling a Single Element Atmospheric Combustor

This research evaluated a method of scaling a single element sub-scale combustor to match the combustion instability modes of a full-scale liquid rocket engine. The experiments used a shear-coaxial injector in an atmospheric chamber using gaseous oxygen and a heated fuel mixture of gaseous methane and nitrogen. The sub-scale flow conditions matched the full-scale equivalence ratio, injection velocities and propellant volumetric flow rates. The first set of experiments empirically determined the effect of chamber diameter on chamber temperature. The results were used to calculate the dimensions of the sub-scaled combustion chamber that would match the transverse frequencies of the full-scale engine. The scaled chamber was tested at two fixed injector locations: center of the chamber (r/rc = 0%) and 0.25 in. from the wall (r/rc = 83.7%). The centered test displayed evidence of coupling between the 1L chamber response and the injector oxygen post at 880 Hz. Injector coupling was also observed with the full-scale engine. With the injector 0.25 in. from the wall, the average chamber temperature dropped about 350 C from the centered test. As a consequence, the frequencies of the transverse responses were lower than the full-scale values. The transverse responses included a 1T at 4,000 Hz, a 2T at 6,400 Hz and a 1R at 8,200 Hz. The amplitudes of the chamber responses varied from 0.001% to 0.01% of chamber pressure. In the sub-scale testing, no major difference was found between the stable and unstable set points of the full-scale engine. A translating stage was used to determine the influence of injector position on instability modes. The results show that the 1L chamber response is present at every location and transverse responses appear as the injector moves near the wall.

Numerical Studies on Frequency Response to Mass Flow Rate Variations in a Hydromechanical Pulsator

Understanding injector dynamics in the presence of strong pressure pulsations in Liquid Fueled Rocket Engines (LFRE) is important due to its strong association with the combustion instability phenomena. Lack of generators of strong periodical pulsations, particularly in the high frequency range, is problematic. To overcome this deficiency, a hydro-mechanical pulsator has been built to generate controlled periodic pressure oscillation, which simulates pressure pulsations in the feed lines of rocket motor injector systems. The present numerical analysis effort utilizes the unsteady Reynolds Averaged Navier-Stokes (URANS) approach, to model the pulsating turbulent flow inside the hydro-mechanical pulsator as a preliminary design study to estimate optimum operating conditions for the proposed experimental study. Numerical results obtained from the study using sliding/deforming grids coupled with a two-layer low-Reynolds number RNG k-e model indicated that the form of pulsation could be controlled based on the geometry of the orifices of the rotating valve in order to define its main parameters, and in the determination of the optimum geometry for obtaining a sinusoidal harmonic response. The results of the computations to evaluate the effect of varying mass flow rates to examine the influence of change in amplitude and phase characteristics of standard injectors on the boundaries, frequencies and amplitude of high frequency combustion instability are presented in the paper.