1. 13reuer et a . ·- all( am et a .· rnvc prcscnte similar device designs that achieves thermal isolation using a vacuum cavity achieved \'ia a wafer-bond, thinbJck process. The thermal sensor consists of a 1500 Athick x 4 )1111-widex 200 ;011-long platinum sensing element on top of a 1500 A-thick silicon nitride membrane which seals a 200 J-Dn-diameter and 10 pmdeep vacuum cavity.25Two gold leads at each end of the sensmg element permit 4-point probe characterization exclusive of the effects of the biasing circuitry and to isolate the sensor performance. Fig. 2.4 is a plan-view SEM of the active area of the shear-stress sensor. Advantages of platinum-based sensors over polycrystalline silicon-based sensors include highc-r TCR, higher thermal operating range, reduced lifnoise, and no piezoresistive-induced pressure sensitivity. The platinum-sensing element demonstrated a factor of two increase in TCR over that of polycrystalline silicon. A TCR of 0.029 "C' possessing a maximum non-linearity of 2.7% over a temperature range of 20 °C to 400 °C was observed. The thermal isolation was also improved over previous designs due to the order of magnitude decrease in membrane thickness. Specifically, the sensor dissipated roughly /0 mW under zero flow conditions at a thermal overheat ratio of /.0 (sensor temperature of 327 °C). Static wall shear stress sensitivity experiments were performed in CC mode of excitation using a laminar flow cell for thermal overheats of 0.2 - 1.0 and wall shear stresses from{) Pa - I.7 Pa (Fig. 2.5). The static sensitivity increased with higher thermal overheat with a maximum sensitivity of 11 mV/Pa at an overheat of /.0. Dynamic wall shear stress sensitivities were obtained at multiple mean shear stress levels and overheats, using a novel, in-situ dynamic calibration technique. This technique provides known sinusoidal shear-stress perntrbations generated via acoustic plane-wave excitation.20·27These calibrations were performed in a constant-current mode of excitation for thermal overheats of 0.6 - 1.0 and for mean shear stress levels of 0.03 Pa to 0.06 Pa. A constant amplitude acoustic excitation or /05 dB SPL (rel. 20 pPa) was used for the calibration. The dynamic shear stress FRF is a function of the static sensitivity as shown in (Fig. 2.6). The sensor exhibited ;40- dB/decnde roll-off with a corner frequency of;600 H: indicative of a highly damped 2ndorder system. Previous work uses a l/2-order system to model the response of a sensor on a semi-infinite medium. The difference in response is explained by the presence or a scaled vacuum cavity that drastically reduces the unsteady heat conduction into the substrate and limits the dissipation due to conduction losses into the thin membrane. Finally, the dynamic range of the sensor is ultimately limited by its cross scnsiti\'itics to non-shearstress inputs and the device noise floor. Experiments were performed to quantitatively measure both the pressure sensiti\'ity and the noise floor. Experimental results indicate negligible pressure scnsiti\'ity (< / pV/Pa). up to /() klh. The noise floor spectra at zero mean flow and multiple overheats were also experimentally determined to be < 100 11Vlt/iI: In summary, the dynamic range of operation of the sensor was experimentally verified (9 pPa-1.7 Pa) and the frequency response-function obtained from /00 Hz to 8 kHz.25-40-
2. I -50 - 100 10-00
3. Thin-oil-film techniques are based on the beha,·ior of the oil film when shear acts upon it. Squire-1-1and Tanner & Blows40identified that the motion of the oil film would be sensitive to gravity, pressure gradients, and surface shear. Recently, Zilliac,19 Brown & Naughton50,and Celie & Zilliac51have conclusively demonstrated that the oil motion is sensitive to surface tension, but, in most cases, only in a very minor way. Thus, in those cases where the pressure gradient is small and gravity is negligible (horizontal surfaces for instance), the oil flow is dominated by the surface shear stress. This suggests that the flow of the oil should reveal characteristics of the shear acting upon it.
4. (3-1) where t is time. Eq. 3-1may be generalized for twodimensional flow of the oil: