Search

The existing database of transition measurements in hypersonic ground facilities has established that the onset of boundary layer transition over a circular cone at zero angle of attack shifts downstream as the nosetip bluntness is increased with respect to a sharp cone. However, this trend is reversed at suciently large values of the nosetip Reynolds number, so that the transition onset location eventually moves upstream with a further increase in nosetip bluntness. This transition reversal phenomenon, which cannot be ex- plained on the basis of linear stability theory, was the focus of a collaborative investigation under the NATO STO group AVT-240 on Hypersonic Boundary-Layer Transition Predic- tion. The current paper provides an overview of that e ort, which included wind tunnel measurements in three di erent facilities and theoretical analysis related to modal and nonmodal ampli cation of boundary layer disturbances. Because neither rst and second- mode waves nor entropy-layer instabilities are found to be substantially ampli ed to ini- tiate transition at large bluntness values, transient (i.e., nonmodal) disturbance growth has been investigated as the potential basis for a physics-based model for the transition reversal phenomenon. Results of the transient growth analysis indicate that disturbances that are initiated within the nosetip or in the vicinity of the juncture between the nosetip and the frustum can undergo relatively signi cant nonmodal ampli cation and that the maximum energy gain increases nonlinearly with the nose radius of the cone. This nding does not provide a de nitive link between transient growth and the onset of transition, but it is qualitatively consistent with the experimental observations that frustum transition during the reversal regime was highly sensitive to wall roughness, and furthermore, was dominated by disturbances that originated near the nosetip.

Although low-disturbance (“quiet”) hypersonic wind tunnels are believed to provide more reliable extrapolation of boundary-layer transition behavior from ground to flight, the presently available q...

The krypton tagging velocimetry (KTV) technique is applied to the turbulent boundary layer on the wall of the Mach 3 Calibration Tunnel at Arnold Engineering Development Complex (AEDC) White Oak. Profiles of velocity were measured with KTV and Pitotpressure probes in the Mach 2.75 turbulent boundary layer comprised of 99% N2/1% Kr at momentum-thickness Reynolds numbers of ReΘ = 800, 1400, and 2400. Agreement between the KTVand Pitot-derived velocity profiles is excellent. The KTV and Pitot velocity data follow the law of the wall in the logarithmic region with application of the Van Driest I transformation. Also, the velocity data in the wake region are consistent with data from the literature for a turbulent boundary layer with a favorable pressure gradient history. A modification of the Mach 3 AEDC Calibration Tunnel is described which facilitates operation at several discrete unit Reynolds numbers consistent with AEDC Hypervelocity Tunnel 9 run conditions of interest. Moreover, to enable near-wall measurement with KTV, an 800 nm longpass filter was used to block the reflection and scatter from the 760.2 nm read-laser pulse. With the longpass filter, the 819.0 nm emission from the re-excited Kr can be imaged to track the displacement of the metastable tracer without imaging the reflection and scatter from the read laser off of solid surfaces.

This paper presents new methodologies based on Mack’s amplitude method to estimate start of transition in hypersonic wind tunnels when 2 nd mode dominates the transition process. The methodologies rely on linear stability computations, measured or modeled tunnel noise, and experimentally determined 2 nd mode breakdown amplitude and receptivity correlations. The first technique uses the measured pitot pressure fluctuations and correlations for the receptivity coefficient to estimate the initial amplitudes. The second technique replaces pitot pressure measurements with a freestream noise spectral amplitude model based on the parameters found in Pate’s correlation. The rationale behind this new model and new insight into Pate’s correlation are provided. The methods can successfully correlate Tunnel 9 and Tunnel F cone measurements at Mach numbers between 9 and 10, unit Reynolds numbers between 1.8 and 30 million per meter and nosetip radii between 1.5 and 12.7 mm. The standard error between the measured and predicted start of transition is approximately 14% which corresponds to ~6.5 cm or 4% of the cone length. In comparison, predictions based on a constant transition N factor yield a standard error greater than 40% and fail to predict transition in several cases.

A series of Particle image velocimetry (PIV) experiments have been performed in the AEDC White Oak Mach 3 Calibration Wind Tunnel to investigate the effect of the Reynolds number based on the momentum thickness, Reθ = 1673 − 7130, on mean streamwise and wall normal fluctuating velocity and Reynolds stress profiles in a turbulent boundary layer. High Reθ cases show the Morkovin transformed wall normal fluctuating velocity profile in good agreement with Klebanoff incompressible data as well as DNS above ~0. 4δ, and streamwise fluctuating velocity agreement above 0. 1δ. No conclusive fluctuating velocity magnitude dependence on Reθ is observed. The greater error at lower spatial resolution and higher Stokes number indicates that the spatial resolution and particle response have a greater influence on fluctuating velocity profiles. Estimates of the power spectra density from PIV measurements reveal truncation of energy at high wavenumbers due to particle lag and spatial filtering. The flatter spectra of the wall normal velocity, compared to the streamwise direction, leads to increased energy at high wavenumbers which could explain why the wall normal fluctuating velocity profile is more susceptible to measurement error from particle lag and spatial filtering.

The boundary-layer transition and stability characteristics of sharp cones at angle-of-attack are investigated with measurements at Mach 10 in the Arnold Engineering Development Complex (AEDC) Hypervelocity Wind Tunnel 9 on a 1.5-m long, 7-deg cone at unit Reynolds numbers between 1.8 and 15 million per meter. The transition location is determined with coaxial thermocouples and temperature sensitive paint, and stability measurements are obtained using high-frequency response pressure sensors. The measurements are used to validate the STABL-3D linear stability theory (LST) code at angles-of-attack up to 6-deg. The computations are found to reproduce the experimental trends regarding the effect of angle-of-attack on the growth of 2 mode waves. The amplitude of the 2 mode waves near breakdown on the leeward and windward meridians scale linearly with edge Mach number. The initial amplitudes estimated using linear stability computations are found to scale with Pitot noise in the unstable 2 mode frequency band. Linear stability computations along with the ability to correlate initial 2 mode amplitudes to tunnel noise and to correlate maximum 2 mode amplitudes to edge Mach number enables the use of Mack’s amplitude method to predict 2 mode transition. This methodology is investigated to accurately predict sharp cone boundary layer transition at 0-deg AoA. Extension to sharp cones at angle-of-attack is expected to be straightforward.

Hypervelocity Wind Tunnel No. 9, located at the White Oak, MD site of the Arnold Engineering Development Complex (AEDC), has long been recognized as a unique world class ground-test facility. The facility was developed in the early 1970s to provide critical low-altitude, high Mach number data in support of the Navy's reentry development programs. Since its inception, Tunnel 9 has maintained a leading role in hypersonic ground testing by continually expanding its operational capabilities to match the needs of current and projected programs, maintaining data quality, and understanding customer requirements. Tunnel 9 started with a unique design built around a state-of-the-art supply heater that provided a clean, high-pressure, high-temperature nitrogen supply. Initial operation of Tunnel 9 realized a Mach 10 and 14 aerodynamic simulation capability. Additional Mach 7 and 8 high Reynolds number capabilities were subsequently developed. Each upgrade to Tunnel 9 during the past 40 years of operation has been in response to various sponsors or hypersonic basic research requirements. These capability enhancements have helped maintain Tunnel 9's position as a core DoD hypersonic test and evaluation (T&E) ground-test facility, which has been identified as a leading facility in all major hypersonic facility studies. Recent improvements and modernization over the past 10 years have focused on test article measurements and have significantly changed the types, quantity, and quality of test data that are readily acquired in a Tunnel 9 test entry. Recent advancements include major system changes such as a completely new control room and high-speed data system to entirely new measurement capabilities such as global heattransfer measurements using Temperature Sensitive paint technology. These along with other incremental improvements have allowed Tunnel 9 to provide new insights into the physics associated with complex hypersonic flows.

This investigation concerns the design, build, and testing of a new class of direct-measuring skin friction sensors capable of performing favorably in sustained hypervelocity flow conditions and detecting boundary layer transition effects. A rigorous, multi-step approach was developed to systematically test the sensors through various bench test setups and wind tunnel facilities. First, the sensors underwent NIST-traceable force calibrations to ensure that any measured wall shear values taken in testing were obtained and reported with well-documented experimental uncertainties. The calibration process was conducted through a series of tests to characterize the new sensor's static, thermal, pressure, and dynamic responses. Validation testing was then conducted in the medium-scale, blow-down Supersonic Tunnel at Virginia Tech under Mach 4.0 flow conditions. Once completed, the final capstone investigations were conducted in AEDC Wind Tunnel 9, a unique hypervelocity blowdown wind tunnel that uses a nitrogen gas test medium. One of the skin friction sensors was integrated into a steel 155.6 cm-(61.27 in.-) long, 7-degree half-angle, cone model. Flow was nominally maintained at Mach 10 and a stagnation temperature of 1000 K (1800 °R). The stagnation pressure and unit Reynolds number were varied over the range of 2.3 to 43.4 MPa (330 to 6300 psia) and 1.6 to 30.3x10/m (0.5 to 9.24x10/ft), respectively. Skin friction was measured over several boundary layer states including near-laminar, transitional, and turbulent flows. Wall shear measurements ranged between 0.92 and 340 Pa (0.02 and 7.1 psf), while the skin friction coefficients ranged between 0.0003 and 0.0060. The total uncertainty of the skin friction sensor remained at ±9.2% of the measurement for a 95% confidence level. The magnitude of skin friction depended on the state of the boundary layer, the Reynolds number of the flow, the nose tip radius of the cone model, and angle-of-attach under pitching runs.

Particle image velocimetry (PIV) experiments have been performed in the AEDC White Oak Mach 3 Calibration Wind Tunnel to validate a new seeding methodology applicable to Tunnel 9 for hypersonic boundary layer measurements. This methodology involved tangential wall injection of a highly concentrated Polyalphaolefins (PAO-4) aerosol. Validation was performed in a well characterized 7.5 to 9 mm thick Mach 2.7 turbulent boundary layer with a 590 m/s edge velocity. The effect of mass injection on the velocity, as determined from Pitot surveys with and without wall injection and PIV measurements with global and local seeding is small, Δu≈0.7-1.3%. Velocity measurements derived from Pitot pressure and traditional globally seeded PIV agree with the locally seeded PIV measurements. Streamwise velocity fluctuations follow the same trend as DNS data and previous experiments. Image processing techniques were used to perform a quantitative comparison of the particle concentration across the boundary layer. The analysis shows that the local seeding methodology provides a good seeding uniformity over 80% of the boundary layer thickness as well as twice as much near wall particles.

In this work, we present the excitation strategy, experimental setup, and results of an implementation of krypton tagging velocimetry (KTV) as applied to an underexpanded jet of three mixtures. We demonstrate that the KTV technique can be employed with gas mixtures of relatively low krypton concentration (0.5% Kr/99.5% N2) and conclude that the KTV technique shows promise as a velocimetry diagnostic with krypton as an inert, dilute, long-lifetime tracer in gas-phase flows.

Boundary-layer transition and stability data were obtained at Mach 10 in the Arnold Engineering Development Complex (AEDC) Hypervelocity Wind Tunnel 9 on a 1.5-m long, 7-deg cone at unit Reynolds numbers between 1.8 and 31 million per meter. A total of 24 runs were performed at angles-of-attack between 0 and 10-deg on sharp and blunted cones with nose radii between 5.1 and 50.8-mm. The transition location was determined with coaxial thermocouples and temperature sensitive paint while stability measurements were obtained using high-frequency response pressure sensors. Mean flow and boundary layer-stability computations were also conducted and compared with the experiment. The effect of angle-of-attack and bluntness on the transition location displays similar trends compared to historical hypersonic wind tunnel data at similar Mach and Reynolds numbers. The N factor at start of transition on sharp cones increases with unit Reynolds number. Values between 4 and 7 were observed. The N factor at start of transition significantly decreases as bluntness increases and is successfully correlated with the ratio of transition location to entropy layer swallowing length. Good agreement between the computed and measured spatial amplification rates and most amplified 2 mode frequencies are obtained for sharp and moderately blunted cones. For large bluntness, where the ratio of transition to entropy swallowing length is below 0.1, 2 mode waves were not observed before the start of transition on the frustum.

Instantaneous velocity measurements are necessary to improve the understanding of hypersonic turbulence and shock/turbulent boundary-layer interactions (STBLI). Such measurements are needed to improve the computational fluid dynamic codes required to reduce the developmental risks of new hypersonic systems. Particle image velocimetry (PIV) has been previously used in small-scale experimental facilities to provide global measurements of velocity fluctuations and Reynolds stresses in hypersonic turbulent boundary layers. However, PIV measurements in large-scale test & evaluation hypersonic facilities such as Tunnel 9 are challenging due to the facility scale, reduced test time, and seeding requirements. Based on an extensive literature review, a PIV development strategy has been developed for Tunnel 9. Using 0.25 micron liquid droplets injected at the wall, Stokes numbers of 0.4 and 0.02 are predicted upstream and downstream of the STBLI region produced by a 33-deg flare at Mach 10 for a free-stream Reynolds number of /ft. PIV development work will begin in the Mach 3 Calibration Laboratory before being transferred to Tunnel 9 for PIV measurements on a large scale hollow-cylinder-flare test article. Schlieren images obtained in the new Mach 3 Calibration Lab indicate that an 11 mm thick turbulent boundary layer can be obtained without tripping.

This work focuses on the development of a hybrid force balance technology that uses piezoelectric sensors in series with conventional strain gauge balance technology to enable broad frequency range force measurements in hypersonic wind tunnel test facilities. Here, sensitivity analyses are carried out to examine the effects of design parameters, including the thickness of mounting plates and location of calibration loading points on the reliability of force and moment measurements.

Laminar stagnation region heating augmentation is investigated in the AEDC Tunnel 9 at Mach 10 by performing high frequency surface pressure and heat transfer measurements on the Orion CEV capsule at zero degree angle-of-attack for unit Reynolds numbers between 0.5 and 15 million per foot. Heating augmentation increases with Reynolds number, but is also model size dependent as it is absent on a 1.25-inch diameter model at Reynolds numbers where it reaches up to 15% on a 7-inch model. Heat transfer space-time correlations on the 7-inch model show that disturbances convect at the boundary layer edge velocity and that the streamwise integral scale increases with distance. Therefore, vorticity amplification due to stretching and piling-up in the stagnation region appears to be responsible for the stagnation point heating augmentation on the larger model. This assumption is reinforced by the f(exp -11/3) dependence of the surface pressure spectrum compared to the f(exp -1) dependence in the free stream. Vorticity amplification does not occur on the 1.25- inch model because the disturbances are too large. Improved free stream fluctuation measurements will be required to determine if significant vorticity is present upstream or mostly generated behind the bow shock.

A miniature high temperature, high frequency, fiber optic pressure sensor capable of operation above 1000K and frequencies above 1MHz has been developed and tested. The sensor accuracy, pressure resolution, temporal resolution and high temperature capability were first demonstrated through a series of laboratory experiments. The sensors were then tested on the forebody and inside the combustor of a full-scale X-51 configuration at fully duplicated flight conditions in the LENS II reflected shock tunnel at the CUBRC facility in Buffalo, NY. The pressure sensors displayed good run-torun repeatability and durability during the 10 run test program. Forebody and tar combustor pressure measurements respectively displayed a ω -0.8 and ω -1 decay in the power spectrum for frequencies related to fluctuations within the log layer. At low frequency, the spectra were essentially flat. In the combustor, a significant increase in the power-law decay was observed for combusting runs. On the forebody, good spectral agreement was obtained at moderate frequencies between the fiber optic and Kulite piezoresistive pressure sensors. This sensing technology enables improved measurement capabilities during ground testing where high frequency pressure measurements in high temperature high speed flows are required.

The unsteadiness of the bow-shock wave in front of a 7-inch diameter Apollo-shaped capsule was measured in the T5 hypervelocity shock tunnel at GALCIT. The pressure uctuations in the reservoir generate expansion and compression waves traveling down the nozzle which interact with the bow-shock in front of the model. The resulting bow-shock displacement is measured using a high-speed Schlieren system where the shock position is extracted using a sub-pixel tracking algorithm. Oscillations of the shock position with amplitudes between 1% to 5% of the stando distance are measured. The power spectrum of the bow-shock motion displays frequencies in the 10 kHZ range corresponding to the frequency content of the pressure uctuations measured in the reservoir. The RMS free stream pressure disturbances in that frequency range are estimated to be between 3.5% and 7%. Simulations of the eect of acoustic waves on spheres were used to give a crude estimate of the facility noise level in the 10 kHz range.

An investigation of the nozzle expansion of air from a reservoir at high specific enthalpy is presented. The objective is to test the chemical models assumed in nonequilibrium numerical computations and obtain a better understanding of the conditions at the nozzle exit as a function of the reservoir conditions. Static pressure measurements at the nozzle exit are performed in the T5 free-piston shock tunnel at the California Institute of Technology and compared with computations using DPLR CFD code which solves the chemically reactive Navier Stokes equations. Experiments are performed at four conditions with a range of stagnation pressures between 20 and 60 MPa and stagnation enthalpies between 10 and 17 MJ/kg. The measured static pressures are between 10% to 17% lower than those computed which seems to indicate that less recombination occurs in the experiment.

The uniformity of the convective heating due to a free stream of hypervelocity air expanded through a conical nozzle is evaluated in the T5 free-piston reflected shock tunnel at Caltech. The stagnation-point heat flux is measured on a rake of ten one-inch-diameter spheres located at the nozzle exit plane. A total of eight shots (two shots at each of the four conditions) were performed with stagnation enthalpies varying from 9.5 to 21 MJ/kg and stagnation pressures from 20 to 60 MPa. Good uniformity was found as the standard deviation among gages ranged between ±4% and ±7% when the two gages located near the nozzle wall were excluded. Good heat flux shot-to-shot repeatability was also found as the standard deviation in stagnation-point beat flux between shots at a given condition was generally below ±4%. The nozzle expansion and flow over a single sphere were simulated using DPLR CFD code with a fully catalytic wall. Very good agreement was found between DPLR and the experiment as the differences between computation and experiment are under 8% for each condition which is significantly below the experimental uncertainties.

ux gages. Simulations of the heat conduction inside the gage using Gasp solver show, that the response is sensitive to geometry of the junction which sits on epoxy. The epoxy acts as an insulator creating a local overheating that increases with the junction aspect ratio and the step height between the inner and outer electrodes and decreases when epoxy is present next to the step. The increase of the overheating with the junction aspect ratio is due to the thermal insulation of the epoxy layer located under the junction used to electrically insulate the two electrodes of the thermocouple. When the junction protrudes signicantly above the surface (about 10 microns), the heat transfer is enhanced from the larger surface area whereas, the lateral conduction around the epoxy layer decreases, which further contributes to the junction overheating. To assess those eects, a calibration method using a reected shock in argon is developed and experiments are performed in the T5 free-piston shock tunnel which is used as a shock tube by replacing the nozzle throat insert with a closed end-wall insert. Five gages are simultaneously calibrated at a high heat ux for 30 microseconds. For microsecond time scales, it is found that the eective value of the thermal product, p ck , is approximately 30% lower than the average of chromel and constantan. The standard deviation of 10% among the ve gages indicates that the gage-to-gage variation of the junction geometry is relatively small as the temperature is greatly sensitive to the geometry on microsecond time scales.