A Numerical Calibration Approach to Obtain Cutting Fluid Droplet Sizes in a Turning Process via an Imaging System

Abstract

AbstractAirborne inhalable particulate in the workplace can represent a significant health hazard, and one of the primary sources of particles is mist produced through the application of cutting fluids in machining operations. The atomization process is one of the principal mechanisms associated with cutting fluid mist formation and generates droplets from fifty to a few thousand micrometers in size. These particles subsequently undergo vaporization and settling effects resulting in an aerosol to which workers may be exposed. While a variety of equipment is available to characterize the fine particulate in the breathing zone, standard equipment to measure the size of the atomized droplets is not available. In this paper, an imaging system is employed to characterize the large droplets produced by atomization in turning. One of the drawbacks of such a system is the time‐consuming experimental calibration procedure that is required to improve the accuracy of the droplet size measurements and extend the depth of field of the imaging system. With this in mind, an approach is introduced to predict droplet diameter based on measurement data without physical system calibration. The relationship between the actual diameter and the measured diameter is established based on an imaging system simulation model that includes a three dimensional point spread function and an image formation relationship grounded in the principles of geometric optics. These two components are combined using convolution integral theory to derive an image intensity profile. The introduction of halo width into the simulation greatly extends the image depth of field, which is a critical factor in capturing more droplets in one image and also minimizing particle size distribution bias towards larger droplets. The model predicts droplet diameter as a function of measured diameter and halo width. Model behavior of predicted diameters from the simulation compares well with those from a physical calibration of the system. The numerical calibration model is then used in the study of cutting fluid atomization in a turning process, and the measured droplet size distribution compares favorably with droplet sizes predicted by a mechanistic atomization model.