Sonda elétrica e ótica para medição do escoamento de misturas bifásicas

Abstract

The fluids movement in pipelines is a wide relevance matter, present in several processes of pharmaceutical, energy, petrochemical, and food industries and the availability of tools and devices that support these activities has great importance, since the occurrence of failures in systems where mixtures flow can cause serious economic, environmental and social consequences. Experimental techniques for flow measurement such as those based on measurement of conductance, ultrasound and ionizing radiation have disadvantages such as the impossibility of measuring non-conductive mixtures, susceptibility to electromagnetic interference, and high attenuation with the transmission distance. The construction and evaluation of an electric and optical probe for the two-phase flows monitoring in industrial environments is presented in this research. The probe consists of two stainless steel electrodes positioned transverse to the flow used to measure the fluid electrical conductance. The electrodes are hollow and within them are inserted fiber Bragg gratings used to measure the flow deformation. The performance of the probe was verified by initial tests on a vertical bench, static calibration with various weights, and dynamics with different flow rates, two-phase flow tests and comparison between probe components and other transducers. In the initial tests, the bubbles mean velocity measured by the probe conductive component resulted in 0.11739 m/s with a standard deviation of 0.0376 m/s, while the optical component measured 0.11720 m/s and 0.0429 m/s. The bubble contact times with the optical component were between 0.66 s and 0.72 s, while the values of the conductive component were 0.68 s and 0.72 s. When comparing the performances of the two probe components, there were differences of 1.09% between the bubbles velocity measurements and 1.43% for the contact times. The independent variable measurement uncertainties of the probe optical component in the static calibration varied from 0.15 N to 0.24 N, with a mean value of 0.19 N and the dependent variable mean uncertainty was 56 pm. The adjustment lines correlation coefficients had an average value of 0.99872 and a mean sensitivity of 315 pm/N was obtained. The experimental data from the two calibrations are in good agreement with the theoretical values. The mathematical models show that changing parameters associated with the probe design can optimize the parameters range to be measured, for example by changing the probe sensitivity. The results show that the combination of the two techniques is feasible and allows data redundancy, self-calibration probe facilities, which is small, causing low flow load loss, easy to construct and robust structure allowing its use in an industrial environment.