Session T08: Experimental Techniques: Aerodynamics/Wind Tunnel (8:00am - 8:45am CST)

Characterization of a multi-fan-array wind tunnel studying insect flight behavior.

How flying insects navigate dynamic wind environments with turbulence and shear flow remains an activate area of investigation. To study their behavior in a controlled manner, we are developing a multi-fan-array wind tunnel with a 90x45x45 cm cubed working section. The fan array consists of 36 80mm fans laid out in a 6x6 grid, and each fan can be independently controlled to provide wind speeds between 0 and 50 cm/s. The fans can be operated in three different configurations designed to produce laminar, shear, and turbulent flow. To characterize the fluid flow in each of these conditions, we will use a hot wire anemometer mounted to a 2D plotter to measure with wind speed and turbulent intensity in several cross sections of the wind tunnel, for each of these three configurations. Preliminary results indicate that after the addition of a horizontal dividing plate we could measure a significant change in wind speed between the top and bottom sections of 20cm/s. After characterizing the fluid flow we will determine how each of the three conditions influences the flight behavior of fruit flies using a 3D multi-camera tracking system. Our results will lay a foundation for studying how insects behave in turbulent wind environments, with a particular future interest in how turbulence modulates their odor plume tracking behavior.   

Drift compensation in hot-wire anemometry

The drift of hot-wire (or other thermal anemometry-based) sensors continues to be a concern for those using constant temperature anemometers (CTAs). Although most previous work on this subject focused on compensating for changes in fluid temperature, the current work, based on that of Hewes et al. (Meas. Sci. Tech. 2020), focuses on compensating for changes in the wire's cold resistance, which may occur as a hot-wire ages, for example. We first show that this drift may be significant, especially for certain combinations of hot-wire sensors and CTAs. We then demonstrate that this drift can be compensated, using a modified form of the relationship between the output voltage of the CTA and the fluid velocity, which takes into account the sensor resistance and other resistances involved in the Wheatstone bridge (top resistance, operating resistance, cable and probe resistances). Moreover, we show that this method can compensate for drift caused by small changes in fluid temperature, even when the temperature is not known, which is of notable practicality. Application of this method is expected to reduce the frequency of calibrations needed to maintain satisfactory degrees of accuracy and repeatability in experimental measurements of fluid flows made using CTAs.   

Characterization of tunable active grid generated turbulence in a water tunnel facility

In the past quarter-century, extensive research of wind-tunnel active grid turbulence generating devices has allowed researchers to fine-tune turbulent flow in experimental settings. That research and capability are rarely available in experimental facilities that use water as the preferred fluid medium. Our group has successfully employed a Makita-style active grid turbulence generator that is comprised of ten rotating shafts with a dedicated stepper motor to control each shaft. A total of sixty rotating winglets rotate using various control protocols to produce controlled and elevated levels of free-stream turbulence For the current experimental campaign, we vary the shaft angular frequency ranging from 0.25 to 3.0Hz to vary the vane Rossby number; the free stream velocities are also varied between 0.10m/s-0.83m/s to vary the mesh Reynolds number. Besides, the winglet blockage was varied by altering the size and solidity of the winglets. We will present the results of our experiments by presenting flow statistics such as turbulence intensity, Taylor Reynolds number, flow anisotropy, and integral length scale for the various cases tested. A comparison will be also be done with wind-tunnels tests over the same parameter range.   

Delay in transition in a plane channel flow using a wind tunnel with large contraction

Measurements were conducted for various friction Reynolds numbers (Re$_{\mathrm{\tau }})$ ranging from laminar through transition to turbulence in a phane channel. The flow was established using a wind tunnel of dimensions much larger than that of the channel, and using two nozzles with contraction ratios of 8.3 and 13 respectively, separated by a settling chamber. Skin friction was determined using friction velocity obtained from wall pressure measurements using a micromanometer and mean velocity from Pitot measurement of the velocity profile. The flow was found to be laminar till Re$_{\mathrm{\tau }}=$ 70, much higher than the usual value of 45, which is observed to the start of transition from various other experiments and direct numerical simulation (DNS). Also, the flow attained the state of fully developed turbulence at Re$_{\mathrm{\tau }}=$ 175, again much higher than the usual value of 70 observed by others. This delay in transition and turbulence is attributed to having a large effective contraction of around 108, that is expected to have largely reduced the non-uniformities in the flow, hence reducing the stress and skin friction.   

Fan array wind tunnels: of mice and mars (and machine learning too)

The development, evolution, and ongoing maturity of fan array wind tunnel (FAWT) technologies at Caltech (2011-present) has made it a tool of choice for real world modeling of problems not easily studied with traditional wind tunnels. The technology has promulgated across a unique and varied set of disciplines, largely due to a) the testing options afforded by an increased parameter space b) the prescriptible and reproducible nature of the flows generated and c) the overall decreased mixing lengths inherent to the design. Two recent implementations have greatly enhanced our understanding of the scope of application of this technology and will briefly be highlighted, namely a small, near-silent FAWT to aid in the neurobiological studies of mouse olfaction and a chamber-safe FAWT able to replicate the wind on Mars to supplement testing of the Mars Helicopter project for Nasa/JPL-Caltech. To highlight our approach for higher-order problems, we explore the combination of machine learning and FAWT technologies through use of reinforcement learning to generate `self-healing', two-dimensional, uniform flow profiles that dynamically adapt to disturbances (such as inlet blockages) to maintain uniformity as measured downstream.   

Shock Train Analysis of Varying Deck Plate Configurations for a Multi Stream Rectangular Nozzle

The interactions in a supersonic flow involve a level of complexity that oftentimes does not match natural intuition that is built concerning the usual incompressible flows. Thus, experimentation is key to making improvements in aircraft design. Concerning supersonic aircrafts, one may analyze specifically the jet outlet and the compressible interaction of the flow which is emanated. By employing a supersonic jet with a single expansion ramp nozzle, this flow can be simulated. Then, by utilizing Particle Image Velocimetry (PIV), interactions can be further identified and characterized. From these PIV images, we can observe shockwaves that form from the jet outlet which continue downstream after reflections off shear layers. These shear layers develop as the result of mixing between the supersonic flow and the ambient surroundings. By adjusting the shape of a deck plate which is positioned at the exit of the jet, the profile of the resulting shock train can be altered to achieve advantages in acoustics. This study compares the shock trains which emanate from varying deck plate configurations to a nominal case. The deck plates of interest are; a twice nominal deck plate, an infinite in span deck plate, and a triangular trailing edge deck plate.   

Dispersed Pressure Sensing for Flow Field Estimation

Fan array wind tunnels offer an ideal test environment for small unmanned aerial vehicles (UAV) due to their high configurability and ability to simulate a multitude of different flows. However, flows currently must be manually measured and adjusted to achieve desired conditions. Measurements have to be taken by sweeping across which can be inaccurate and does not provide a temporally synchronized reading of the entire field. The development of an array of pitot-static tubes can provide accurate pressure and velocity data across an entire flow field at a single instance, giving a better picture of the entire field. This measurement technique has significant potential for machine learning applications on fan arrays, particularly as a feedback system to generate and maintain desired flows. This pitot-static tube array was implemented on a small fan array in Caltech's Center for Autonomous Systems and Technology (CAST) with the primary design conditions of minimizing interference with the flow produced by maintaining a small profile and structural integrity and stability.