Split-spectrum Method Research Articles

Split-spectrum intensity-based optical fiber sensors for measurement of microdisplacement, strain, and pressure

A self-referencing technique compensating for fiber losses and source fluctuations in reflective air-gap intensity-based optical fiber sensors is described. A dielectric multilayer short-wave-pass filter is fabricated onto or attached to the output end face of the lead-in-lead-out multimode fiber. The incoming broadband light from a white light or a light-emitting diode is partially reflected at the filter. The transmitted light through the filter projects onto a mirror. The light returning from the reflecting mirror is recoupled into the lead-in-lead-out fiber. These two reflections from the filter and the reflecting mirror are spectrally separated at the detector end. The power ratio of these two reflections is insensitive to source fluctuations and fiber-bending loss. However, because the second optical signal depends on the air-gap separation between the end face of the lead-in-lead-out fiber and the reflecting mirror, the ratio provides the information on the air-gap length. A resolution of 0.13 µm has been obtained over a microdisplacement measurement range of 0-254 µm. The sensor is shown to be insensitive to both fiber-bending losses and variations in source power. Based on this approach, a fiber-strain sensor was fabricated with a multilayer interference filter directly fabricated on the end face of the fiber. A resolution of 13.4 microstrain was obtained over a measurement range of 0-20,000 microstrain with a gauge length of 10 mm. The split-spectrum method is also incorporated into a diaphragm displacement-based pressure sensor with a demonstrated resolution of 450 Pa over a measurement range of 0-0.8 MPa.

A MULTIPLE-TIME-SCALE TURBULENCE MODEL BASED ON VARIABLE PARTITIONING OF THE TURBULENT KINETIC ENERGY SPECTRUM

A multiple-time-scale turbulence model based on a single point closure and a simplified split-spectrum method is presented. In the model the effect of the ratio of the production rate to the dissipation rate on the eddy viscosity is modeled using multiple time scales and a variable partitioning of the turbulent kinetic energy spectrum. Example problems included are a wall-jet flow, a wake-boundary-layer interaction flow, a backward-facing step flow, and a confined coaxial swirling jet. The multiple-time-scale turbulence model yields significantly improved computational results compared with those obtained using the k-c turbulence model. It is also shown that the present turbulence model can resolve the viscous superlayer of turbulent flows.