Measuring local droplet parameters using single optical fiber probe

Abstract

This article proposes a new single optical fiber probe (OFP) sensor method for measuring local droplet parameters such as droplet fraction, droplet velocity, and droplet diameter. To this end, we developed a stretched conical optical fiber probe (SC-OFP) sensor that minimizes the wetting of the sensing tip at a high droplet velocity, a discrimination algorithm based on the first and second derivatives of the output of the raw signal of the OFP sensor, and a method that uses neural network training to determine local droplet velocity independently of droplet size. In the experiments conducted, the developed SC-OFP sensor measured local droplet velocity with a mean absolute percentage error (MAPE) smaller than 4.1% for velocities ranging from 2 to 32 m/s, and droplet diameter with a MAPE smaller than 4.9% for diameters ranging from 0.29 to 1.78 mm.

Bayesian optimization of single-atom alloys and other bimetallics: efficient screening for alkane transformations, CO2 reduction, and hydrogen evolution

“…Some studies have used BO for material discoveries in computational chemistry; however, only few of these studies use BO for adaptive experimental design. Also, there are relatively few studies that involve search spaces with discrete quantities 28 as opposed to continuous functions. Other Bayesian techniques such as the Bayesian information criterion 29 and Bayesian model averaging 30 have been employed by previous studies; 31 however, these techniques are used to aid in creating models on an existing dataset rather than for recommending new data points (i.e., adaptive experimental design), which is the main purpose of BO.…”