NIST Internship
During the summer of 2022, I participated in the SURF Program at NIST, the National Institute of Standards Technology. At NIST, I worked with Dr. Gregory Vogl in the Engineering Lab on a research project where we monitored the thermal drift of machine tools using image analysis. By the end of the summer, my coding skills in MATLAB had grown significantly, I understood the process of scientific research, and I presented my work. You can read my published paper here.
Abstract
Precision manufacturing is increasing in importance as manufacturers seek to make their processes faster and cheaper while still producing quality parts. A major hurdle in the quest for precision machining is thermal drift, which can account for up to 70% of machining accuracy errors. This study focuses on developing and testing a new method to monitor the thermal drift of machine tools using videos of dot calibration grids taken with a high-magnification wireless microscope. To do this, a dot calibration grid is fixed at each corner of the machine tool worktable and a wireless microscope camera is mounted in the tool holder of the spindle. Since the machine tool can move in three dimensions (X, Y, and Z) and the spindle can rotate, videos with various types of machine motion can be captured and analyzed to determine the thermal drift of the spindle relative to the worktable. Next, these videos are parsed into images to segment the dots and track the individual X and Y dot center positions. The Z dot position is estimated by finding the Z position that yields the maximum contrast of dot edges. The tracked positions are then inputted into a two-dimensional (X and Y) or three-dimensional (X, Y, and Z) linear model to solve for the 2D or 3D changes within the machine tool. For example, the 2D model contains six thermally induced variables that are solved in a least-squares manner: X translation, Y translation, Z rotation, XY squareness change, and XY planar thermal growth coefficients. Experiments were conducted to test the ability of this new technique to measure commanded translations as large as 50 micrometers. Data analysis revealed that this method was accurate at a micrometer-level and is an effective way of measuring thermal drift. Overall, a relatively cheap wireless microscope and small calibration grids can be used to track thermal drift in near real time. Such an approach has significant future commercial potential for diagnosing and correcting thermal drift in machine tools which will advance manufacturing precision and efficiency.