Aero-optics refers to optical distortions due to index-of-refraction gradients that are induced by aerodynamic density gradients. At hypersonic flow conditions, the bulk velocity is many times the speed of sound and density gradients may originate from shock waves, compressible turbulent structures, acoustic waves, thermal variations, etc. Due to the combination of these factors, aero-optic distortions are expected to differ from those common to sub-sonic and lower super-sonic speeds. This report summarizes the results from a 2019-2022 Laboratory Directed Research and Development (LDRD) project led by Sandia National Laboratories in collaboration with the University of Notre Dame, New Mexico State University, and the Georgia Institute of Technology. Efforts extended experimental and simulation methodologies for the study of turbulent hypersonic boundary layers. Notable experimental advancements include development of spectral de-aliasing techniques for highspeed wavefront measurements, a Spatially Selective Wavefront Sensor (SSWFS) technique, new experimental data at Mach 8 and 14, a Quadrature Fringe Imaging Interferometer (QFII) technique for time-resolved index-of-refraction measures, and application of QFII to shock-heated air. At the same time, model advancements include aero-optic analysis of several Direct Numerical Simulation (DNS) datasets from Mach 0.5 to 14 and development of wall-modeled Large Eddy Simulation (LES) techniques for aero-optic predictions. At Mach 8 measured and predicted root mean square Optical Path Differences agree within confidence bounds but are higher than semi-empirical trends extrapolated from lower Mach conditions. Overall, results show that aero-optic effects in the hypersonic flow regime are not simple extensions from prior knowledge at lower speeds and instead reflect the added complexity of compressible hypersonic flow physics.
High-speed, optical imaging diagnostics are presented for three-dimensional (3D) quantification of explosively driven metal fragmentation. At early times after detonation, Digital Image Correlation (DIC) provides non-contact measures of 3D case velocities, strains, and strain rates, while a proposed stereo imaging configuration quantifies in-flight fragment masses and velocities at later times. Experiments are performed using commercially obtained RP-80 detonators from Teledyne RISI, which are shown to create a reproducible fragment field at the benchtop scale. DIC measurements are compared with 3D simulations, which have been ‘leveled’ to match the spatial resolution of DIC. Results demonstrate improved ability to identify predicted quantities-of-interest that fall outside of measurement uncertainty and shot-to-shot variability. Similarly, video measures of fragment trajectories and masses allow rapid experimental repetition and provide correlated fragment size-velocity measurements. Measured and simulated fragment mass distributions are shown to agree within confidence bounds, while some statistically meaningful differences are observed between the measured and predicted conditionally averaged fragment velocities. Together these techniques demonstrate new opportunities to improve future model validation.
A significant portion of the energy released by an explosive is contained in a postdetonation fireball. Quantitative characterizations of in situ optical properties are needed to verify predictive models of these environments. This work investigates the narrowband red, green, and blue (RGB) emissive and absorptive characteristics of lab scale explosive blasts. Experiments utilize a custom triple-bandpass filter to image blast emission with a high-speed color camera and calculate temperatures from the RGB band ratios. The measured emission is contingent on the optical density, which is also explored with a triple-band (RGB) optical density measurement using the color camera. Time histories of the calculated emissive temperature and optical depth provide insight into the interpretation of pyrometric measurements in optically dense, particle laden combustion environments. This is applicable to many other reacting systems.
Four Direct Numerical Simulation (DNS) datasets covering effective freestream Mach numbers of 8 through 14 are used to investigate the behavior of turbulence-induced aero-optical distortions in hypersonic boundary layers. The datasets include two from simulations of flat plate boundary layers (Mach 8 and 14) and two from simulations of flow over a sharp cone (Mach 8 and 14). Instantaneous three-dimensional fields of density from each DNS are converted to refraction index and integrated to produce distributions of the Optical Path Differences (OPD) caused by turbulence. These values are then compared to experimental data from the literature and to an existing model for the root-mean-square of the OPD. Although the model was originally developed for flows with Mach ≤ 5, it provides a basis to which we compare the hypersonic data.
Compressible wall modeled large-eddy simulations of a Mach eight turbulent boundary layer over a flat plate were carried out for the conditions of the Hypersonic Wind Tunnel at Sandia National Laboratories. Overall good agreement of the velocity and temperature profiles is obtained with reference data from a direct numerical simulation and a theoretical relationship. Profiles of the resolved root-mean-square velocity fluctuations are in adequate agreement with the reference data. The refractive index is calculated from the density field and integrated along an expected beam path to calculate the optical path length. Then, by subtracting a bilinear fit of the instantaneous optical path length, the optical path difference is obtained. The computed aero-optical path difference shows a similar dependence on the aperture size as in the literature. The normalized root-mean-square optical path difference from the present wall-modeled large-eddy simulations and a reference direct numerical simulation and experiment are in good agreement. The optical path distortion is slightly above the value predicted by a semi-analytical relationship from the literature. Finally, instantaneous snapshots of the flow are analyzed via proper orthogonal decomposition and the optical path distortion is computed from subsets of the modes. The optical path distortion converges quickly with increasing number of modes which suggests that the main contribution comes from large energetic flow structures.
First-of-their kind datasets from a high-speed X-ray tomography system were collected, and a novel numerical effort utilizing temporal information to reduce measurement uncertainty was shown. The experimental campaign used three high-speed X-ray imaging systems to collect data at 100 kHz of a scene containing high-velocity objects. The scene was a group of known objects propelled by a 12-gauge shotgun shell reaching speeds of hundreds of meters per second. These data represent a known volume where the individual components are known, with experimental uncertainties that can be used for reconstruction algorithm validation. The numerical effort used synthetic volumes in MATLAB to produce projections along known lines of sight to perform tomographic reconstructions. These projections and reconstructions were performed on a single object at two orientations, representing two timesteps, to increase the reconstruction accuracy.
Measurements are presented of the aero-optic distortion produced by a Mach 8 turbulent boundary layer in the Sandia Hypersonic Wind Tunnel. Flat optical inserts installed in the test section walls enabled a double-pass arrangement of a collimated laser beam. The distortion of this beam was imaged by a high-speed Shack-Hartmann sensor at a sampling rate of up to 1 MHz. Analysis is performed using two processing methods to extract the aero-optic distortion from the data. A novel de-aliasing algorithm is proposed to extract convective-only spectra and is demonstrated to correctly quantify the physical spectra even in case of relatively low sampling rates. The results are compared with an existing theoretical model, and it is shown that this model under-predicts the experimentally measured distortions regardless of the processing method used. Possible explanations for this discrepancy are presented. The presented results represent to-date the highest Mach number for which aero-optic boundary layer distortion measurements are available.
Aluminum particle combustion is a critical component in solid propellant operation. Understanding these processes is essential for improving specific impulse and other performance metrics. Prior studies of aluminum particle combustion in the literature have focused on spatial and temperature statistics for a single propellant strand size, which is typically significantly smaller than the full grain size used in aerospace and defense applications. In this work, we aim to determine the effect of increasing propellant strand size on several key properties of aluminum particle combustion at atmospheric pressure. To accomplish this, we use simultaneous high speed holography and imaging pyrometry to obtain temporally resolved spatial and temperature information. Here, we discuss how agglomerate size, velocity, and temperature statistics vary as a function of propellant strand size from 6 mm up to 19 mm in diameter. By understanding how the statistics scale as a function of strand size, we can determine how to extrapolate lab-scale experimental data to full-scale propellant burns.
The character of aero-optical distortions produced by turbulence is investigated for subsonic, supersonic, and hypersonic boundary layers. Data from four Direct Numerical Simulations (DNS) of boundary layers with nominal Mach numbers ranging from 0.5 to 8 are used. The DNS data for the subsonic and supersonic boundary layers are of flow over flat plates. Two hypersonic boundary layers are both from flows with a Mach 8 inlet condition, one of which is flow over a flat plate while the other is a boundary layer on a sharp cone. Density fields from these datasets are converted to index-of-refraction fields which are integrated along an expected beam path to determine the effective Optical Path Lengths that a beam would experience while passing through the refractions of the turbulent field. By then accounting for the mean path length and tip/tilt issues related to bulk boundary layer effects, the distribution of Optical Path Differences (OPD s) is determined. Comparisons of the root-mean-squares of the OPDs are made to an existing model. The OPDr m s values determined from the subsonic and supersonic data were found to match the existing model well. As could be expected, the hypersonic data does not match as well due to assumptions like the Strong Reynold Analogy that were made in the derivation of the model. Until now, the model has never been compared to flows with Mach numbers as high as included herein or to flow over a sharp cone geometry.
This paper validates the concept of a spatially filtered wavefront sensor, which uses a convergent-divergent beam to reduce sensitivity to aero-optical distortions near the focal point while retaining sensitivity at large beam diameters. This sensor was used to perform wavefront measurements in a cavity flow test section. The focal point was traversed to various spanwise locations across the test section, and the overall OPDRMS levels and aperture-averaged spectra of wavefronts were computed. It was demonstrated that the sensor was able to effectively suppress the stronger aero-optical signal from the cavity flow and recover the aero-optical signal from the boundary layer when the focal point was placed inside the shear region of the cavity flow. To model these measured quantities, additional collimated beam wavefronts were taken at various subsonic speeds in a wind tunnel test section with two turbulent boundary layers, and then in the cavity flow test section, where the signal from the cavity was dominant. The results from the experimental model agree with the measured convergent-divergent beam results, confirming that the spatial filtering properties of the proposed sensor are due to attenuating effects at small apertures.
Accurate knowledge of post-detonation fireball temperatures is important for understanding device performance and for validation of numerical models. Such measurements are difficult to make even under controlled laboratory conditions. In this work temperature measurements were performed in the fireball of a commercial detonator (RP-80, Teledyne RISI). The explosion and fragments were contained in a plastic enclosure with glass windows for optical access. A hybrid femtosecond-picosecond (fs-ps) rotational coherent anti-Stokes Raman scattering (CARS) instrument was used to perform gas-phase thermometry along a one-dimensional measurement volume in a single laser shot. The 13-mm-thick windows on the explosive-containment housing introduced significant nonlinear chirp on the fs lasers pulses, which reduced the Raman excitation bandwidth and did not allow for efficient excitation of high-J Raman transitions populated at flame temperatures. To overcome this, distinct pump and Stokes pulses were used in conjunction with spectral focusing, achieved by varying the relative timing between the pump and Stokes pulses to preferentially excite Raman transitions relevant to flame thermometry. Light scattering from particulate matter and solid fragments was a significant challenge and was mitigated using a new polarization scheme to isolate the CARS signal. Fireball temperatures were measured 35-40 mm above the detonator, 12-25 mm radially outward from the detonator centerline, and at 18 and 28 μs after initiation. At these locations and times, significant mixing between the detonation products and ambient air had occurred thus increasing the nitrogen-based CARS thermometry signal. Initial measurements show a distribution of fireball temperatures in the range 300-2000 K with higher temperatures occurring 28 μs after detonation.
Three-beam rotational coherent anti-Stokes Raman scattering (CARS) measurements performed in highly scattering environments are susceptible to contamination by two-beam CARS signals generated by the pump–probe and Stokes–probe interactions at the measurement volume. If this occurs, differences in the Raman excitation bandwidth between the two-beam and three-beam CARS signals can add significant errors to the spectral analysis. This interference, to the best of our knowledge, has not been acknowledged in previous three-beam rotational CARS experiments, but may introduce measurement errors up to 25% depending on the temperature, amount of scattering, and differences between the two-beam and three-beam Raman excitation bandwidths. In this work, the presence of two-beam CARS signal contamination was experimentally verified using a femtosecond–picosecond rotational CARS instrument in two scattering environments: (1) a fireball generated by a laboratory-scale explosion that contained particulate matter, metal fragments, and soot, and (2) a flow of air and small liquid droplets. A polarization scheme is presented to overcome this interference. By rotating the pump and Stokes polarizations +55◦ and −55◦ from the probe, respectively, the two-beam and three-beam CARS signals are orthogonally polarized and can be separated using a polarization analyzer. Using this polarization arrangement, the Raman-resonant three-beam CARS signal amplitude is reduced by a factor of 2.3 compared to the case where all polarizations are parallel. This method is successfully demonstrated in both scattering environments. A theoretical model is presented, and the temperature measurement error is studied for different experimental conditions. The criteria for when this interference may be present are discussed.
Accurate knowledge of post-detonation fireball temperatures is important for understanding device performance and for validation of numerical models. Such measurements are difficult to make even under controlled laboratory conditions. Here, temperature measurements were performed in the fireball of a commercial detonator (RP-80, Teledyne RISI). The explosion and fragments were contained in a plastic enclosure with glass windows for optical access. A hybrid femtosecond-picosecond (fs-ps) rotational coherent anti-Stokes Raman scattering (CARS) instrument was used to perform gas-phase thermometry along a one-dimensional measurement volume in a single laser shot. The 13-mm-thick windows on the explosive-containment housing introduced significant nonlinear chirp on the fs lasers pulses, which reduced the Raman excitation bandwidth and did not allow for efficient excitation of high-J Raman transitions populated at flame temperatures. To overcome this, distinct pump and Stokes pulses were used in conjunction with spectral focusing, achieved by varying the relative timing between the pump and Stokes pulses to preferentially excite Raman transitions relevant to flame thermometry. Light scattering from particulate matter and solid fragments was a significant challenge and was mitigated using a new polarization scheme to isolate the CARS signal. Fireball temperatures were measured 35–40 mm above the detonator, 12–25 mm radially outward from the detonator centerline, and at 18 and 28 µs after initiation. At these locations and times, significant mixing between the detonation products and ambient air had occurred thus increasing the nitrogen-based CARS thermometry signal. Initial measurements show a distribution of fireball temperatures in the range 300–2000 K with higher temperatures occurring 28 µs after detonation.
Three ultra-high-speed, 10 MHz, cameras imaged the time-resolved decay of laser-induced incandescence (LII) from soot in a turbulent non-premixed ethylene jet flame. Cameras were equipped with a stereoscope allowing each CMOS array to capture two separate views of the flame. The resulting six views were reconstructed into a volumetric soot decay series using commercially available DaVis tomographic software by LaVision. Primary soot particle sizes were estimated from the decay time history on a per voxel basis by comparing measured signals to an LII model. Experimentally quantified soot particle sizes agree with existing predictions and previous measurements.
In extreme supersonic, hypersonic, or explosive environments, the presence of gas-phase shocks cause coherent imaging distortions and inhibit object tracking. In this work, we aim to remove these distortions by measuring the relative phase of the light using digital phase-sensitive holography techniques including a two-step phase-shifting technique and a single-shot polarization phase-shifting technique. Once the phase of the shock-wave is acquired, the distortion is numerically canceled from the image and the un-distorted object image is recovered. This work discusses the theory, provides simulation results, and presents preliminary experimental data showing how this concept can be applied to remove shock-wave phase distortions created by supersonic air jets.