CNT‐Based Artificial Hair Sensors for Predictable Boundary Layer Air Flow Sensing

Abstract

While numerous flow sensor architectures mimic the natural cilia of crickets, locusts, bats, and fish, the prediction of sensor output for given flow conditions based on the sensor properties has not been achieved. Challenges include difficulty in determining the electromechanical properties of the sensors, limited working knowledge of the boundary layer, low sensitivity to small hair deflections, and lack of models for large deflections. Within this work, hair sensors are fabricated using piezoresistive arrays of carbon nanotubes (CNTs) without traditional microelectromechanical processing. While correlating the CNT array electromechanical properties to synthesis conditions remains a challenge, a consistent, proportional, and predictable response to steady, boundary-determined air flow is obtained using theory and measurement for various lengths of hairs. The moment sensitivity is shown to scale inversely with the CNT length and stiffness to a typical maximum of 1.3 ± 0.4% resistance change nN −1 m −1 . The normalized CNT piezoresistivity is constant (1.1 ± 0.2) for a majority of the more than two dozen sensors examined despite the orders-of-magnitude variability in both sensitivity and CNT compressive modulus. The sensor sensitivity and noise both distinctly change as the flow transitions from steady and laminar to turbulent, suggesting the sensor may be capable of detecting flow transitions.

High-Bandwidth and Sensitive Air Flow Sensing Based on Resonance Properties of CNT-on-Fiber Hairs

“…This sensor was one of 25 sensors previously examined in steady flow conditions [4]. The magnitude of the Fourier transformed response is plotted versus the frequency in Figure 2a for each length.…”

“…Sensitivity has been previously discussed and includes both the ability to measure a small amplitude of air perturbation and the magnitude of response per change in air velocity past this threshold [1][2][3][4]. The frequency bandwidth of useful operation has been discussed to a smaller extent.…”

“…Due to the strong resonance, higher-order resonant modes are accessible at even higher frequencies. The fabrication, sensitivity to steady aerodynamics, internal mechanics, and integration of arrays into aerodynamic structures of these CNT-based sensors have been previously discussed [1,4,13]. Essentially, the external aerodynamic loading induces a moment inside the capillary which compresses the radially grown CNT array, providing a sigmoidal resistance change (decrease) as a function of flow velocity or deflection.…”

“…While the total length of the hair (L T ) includes the portion of fiber enclosed by the capillary, this model most closely matches these devices if the beam is considered only as the length of the fiber external to the capillary (L E ) and the hair is assumed to be fixed at the opening of the capillary. Going further, it has been previously shown that these sensors can be more accurately modeled as beams supported by an elastic foundation [4,15,16]. In the case of the dynamic response of the sensor, a stiffness constant k appears in the time-varying Euler beam equation:…”

“…The nanotubes surrounding the glass fiber inside the capillary act as the elastic base, supporting the fiber along the length of hair internal to the capillary. The value of the stiffness constant can be determined by characterizing the bending and deflection of the hair under mechanical loading or by measuring the electrical response under various loading conditions, noting that the electrical response is uniquely a function of the bending moment on the hair [4]. It has been shown that the relationship between the elastic base stiffness and the compressive modulus of the CNTs (κ) can be approximated as…”